Dejian Shen

Cracking Control on
Early-Age Concrete
Through Internal Curing

@seismicisolation
@seismicisolation
Cracking Control on Early-Age Concrete Through
Internal Curing

@seismicisolation
@seismicisolation
Dejian Shen

Cracking Control
on Early-Age Concrete
Through Internal Curing

@seismicisolation
@seismicisolation
Dejian Shen
College of Civil and Transportation
Engineering
Hohai university
Nanjing, China

ISBN 978-981-19-8397-9 ISBN 978-981-19-8398-6 (eBook)

https://doi.org/10.1007/978-981-19-8398-6

Jointly published with Science Press, Beijing, China

The print edition is not for sale in China mainland. Customers from China mainland please order the print
book from: Science Press

© Science Press 2023

This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse
of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publishers, the authors, and the editors are safe to assume that the advice and information in this book
are believed to be true and accurate at the date of publication. Neither the publishers nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publishers remain neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore

@seismicisolation
@seismicisolation
Preface

Concrete has large annual consumption in China due to diverse sources, good dura-
bility, and relatively low cost. However, early-age cracking of concrete remains a
problem in many practical engineering projects, e.g., urban rail transits, tunnels, high-
rise buildings, and long-span bridges. Cracks may occur and develop quickly when
the restrained stress of concrete induced by shrinkage deformation exceeds its tensile
capacity, which will destroy the integrity of the concrete structures. The aggressive
substances such as chlorides or sulfates can easily penetrate concrete through cracks,
affecting the quality of concrete and causing many security problems. Therefore, for
the long-term sustainability of the concrete structures is important, investigations on
the methods to increase the early-age cracking resistance of HSC are needed. Internal
curing can increase the early-age cracking resistance of concrete. The internal curing
material acts as an internal “reservoir” after absorbing water in advance. When there
is insufficient water due to self-desiccation effect in the concrete, it can timely provide
water from the interior, so as to control the drop of internal relative humidity and
the development of shrinkage in the concrete. However, systematics investigations
on the cracking control of internally cured concrete are lacking.
This monograph summarizes and outlines the main theoretical and experimental
findings obtained in the decade studies conducted at the College of Civil and Trans-
portation Engineering, Hohai University, Nanjing city, China, with a specific focus on
the early-age cracking resistance of internally cured concrete. In this monograph, the
technique of internal curing for cracking resistance of concrete with SAPs and LWAs
is further developed. It establishes models for the IRH of early-age internally cured
concrete considering the effect of internal curing water content, working w/c ratio
and critical time, and curing humidity, respectively. Models for predicting the auto-
genous shrinkage of early-age internally cured concrete are also established based
on the IRH or ultrasonic velocity of concrete. The variation law and mechanism
of early-age tensile creep of internally cured concrete are revealed. Furthermore,
the variation law and mechanism of early-age crack resistance of internally cured
concrete under continuous restrained condition or uniaxial restrained condition at
early age are explored. The integrated criterion of cracking resistance is proposed
to evaluate the early-age cracking behavior of internally cured concrete. The author

v
@seismicisolation
@seismicisolation
vi Preface

hopes it can provide certain help and reference for professionals, practitioners, and
also for undergraduate or graduate students.
The financial supports from the National Natural Science Foundation of China
(51879092), Science and Technology Project of Ministry of Housing and Urban-
Rural Development of the People’s Republic of China (2014-K3-011, 2018-K5-
015, and 2018-K8-010), Transportation Science and Technology Project of Jiangsu
Province (2015T55-1), Project of Natural Science Foundation of Jiangsu Province
(BK2010512), and Special Fund for Water Conservation Research in the Public
Interest (200701014 and 201101014) are gratefully acknowledged.
The draft for Chap. 1 was prepared by Dr. Jiacheng Kang, Dr. Ci Liu, Dr. Zhizhuo
Feng, Dr. Chuyuan Wen, and Dr. Chengcai Li. The draft for Chap. 2 was prepared
by Dr. Ci Liu. The draft for Chap. 3 was prepared by Dr. Chuyuan Wen. The draft
for Chap. 4 was prepared by Dr. Chengcai Li and Ms. Tingting Zhang. The draft for
Chap. 5 was prepared by Dr. Zhizhuo Feng. The draft for Chap. 6 was prepared by
Dr. Jiacheng Kang. Graduate students in the author’s research group contributed a
great deal to the research work and completion of this monograph, including Ms.
Dabao Cheng, Mr. Ying Chen, Mr. Xiaojian Tang, Mr. Jiaxin Shen, Mr. Shucheng
Deng, Mr. Tao Wang, Mr. Xudong Wang, Mr. Panpan Yao, Mr. Xiaoguang Zhao, Mr.
Yongqiang Shen, Mr. Shuaishuai Zhu, Mr. Yong Ji, Mr. Huafeng Shi, Mr. Yuhua Wu,
Mr. Mingliang Wang, Mr. Jinliang Jiang, Mr. Kaiqi Liu, Mr. Baizhong Zhou, and Ms.
Wengting Wang. Their hard-working, dedication, and intelligence have contributed
considerably to the work presented in this monograph.

Nanjing, China Dejian Shen

May 2022

@seismicisolation
@seismicisolation
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Internal Relative Humidity of Early-Age Internally Cured
Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3 Autogenous Shrinkage of Early-Age Internally Cured Concrete . . . . 105
4 Tensile Creep of Early-Age Internally Cured Concrete
with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
5 Cracking Resistance of Internally Cured Concrete
by Restrained Ring Test at Early-Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
6 Cracking Resistance of Internally Cured Concrete Under
Uniaxial Restrained Condition at Early-Age . . . . . . . . . . . . . . . . . . . . . . 269

vii
@seismicisolation
@seismicisolation
Abbreviations

CTE Coefficient of thermal expansion

GGBFS Ground granulated blast furnace slag
HPC High-performance concrete
HSC High-strength concrete
IC Internal curing
ICC Internally cured concrete
IRH Internal relative humidity
LWAs Lightweight aggregates
PP fiber Polypropylene fiber
RH Relative humidity
SAPs Superabsorbent polymers
TSTM Temperature-stress testing machine
w/b ratio Water-to-binder ratio
w/c ratio Water-to-cement ratio
wic /c ratio Internal curing water-to-cement ratio

ix
@seismicisolation
@seismicisolation
Chapter 1
Introduction

Contents

1.1 Early-Age Cracking of Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.1.1 Significance of Early-Age Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.2 Causes of Early-Age Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.3 Measures for Controlling Early-Age Cracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2 State-of-the-Art of Internal Curing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2.1 What is Internal Curing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.2.2 How Does Internal Curing Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.3 How to Prepare Internally Cured Concrete . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.4 How Does Internal Curing Work on Internal Relative Humidity . . . . . . . . . . . . . 17
1.2.5 How Does Internal Curing Work on Autogenous Shrinkage . . . . . . . . . . . . . . . . . 21
1.2.6 How Does Internal Curing Work on Tensile Creep . . . . . . . . . . . . . . . . . . . . . . . . 23
1.2.7 Cracking Resistance by Restrained Ring Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
1.2.8 Cracking Resistance Under Uniaxial Restrained Condition . . . . . . . . . . . . . . . . . 32
1.3 Objectives and Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

Concrete structure is widely used in civil engineering, transportation engineering,

hydraulic engineering, nuclear engineering, ocean engineering, and agricultural engi-
neering around the world [1–4]. Concrete with a low w/c ratio exhibits superior
mechanical properties and durability [5, 6]. However, high shrinkage of concrete
is revealed due to a decrease in w/c ratio, which is prone to trigger the early-age
cracking under the restrained condition [7–10]. The cracking provides the path for
deleterious substances such as water and chloride into matrix and potentially result in
the serviceability problems of concrete structures [11, 12]. Moreover, concrete struc-
ture with cracks is costly to maintain and repair [13]. Therefore, cracking control
at early age is significant for increasing durability and prolonging the service life
as well as improving the safety of concrete structure. Many measures have been
employed to control the cracking at early age, such as the incorporation of fibers,
chemical admixtures, and IC materials [14, 15]. The introduction of IC materials
with high water absorption prevents the reduction of relative humidity and genera-
tion of capillary pressure inside the cement matrix, thereby decreasing the shrinkage

© Science Press 2023 1

D. Shen, Cracking Control on Early-Age Concrete Through Internal Curing,
https://doi.org/10.1007/978-981-19-8398-6_1
@seismicisolation
@seismicisolation
2 1 Introduction

of concrete [16, 17]. Therefore, IC is an effective mitigation method to enhance the

cracking resistance of concrete at early age.
This chapter introduces the significance and causes of early-age cracking behav-
iors, and some methods for controlling early-age cracks are summarized. An
overview on IC as well as the objectives and scope of this monograph are also
presented.

1.1 Early-Age Cracking of Concrete

1.1.1 Significance of Early-Age Cracking

Concrete is a composite material composed of fine and coarse aggregate bonded

together with a fluid cement (cement paste) that hardens (cures) over time. Concrete
is the most used human-made material with 14 billion cubic meters (equivalent to
32 billion tons of concrete) produced every year for use due to diverse sources, good
durability, and relatively low cost [18]. In addition, a large amount of concrete is
widely used in the building structures, bridges, dams, roads, tunnels, substructures,
high buildings, nuclear reactors, etc. [19]. However, early-age cracking may develop
and propagate due to the low tensile strength of concrete. “Early age” is a concept
closely related to its performance state. For concrete with different compositions, the
definitions of “early age” are different according to their performance. Early age is
defined as the period before the performance of concrete is stable. Generally, the early
age refers to the period before 28 days. At early age, as the temperature of concrete
first increases due to hydration heat and decreases, rapid thermal shrinkage may lead
to cracking. In addition, in concrete with low water to cement ratio, hydration of
cement may cause self-desiccation and autogenous shrinkage.
The occurrence of cracks in concrete not only affects the appearance, functionality,
and durability of the structure, but also reduces the safety of the structure [20].
In addition, the development of cracks will also increase the maintenance cost of
buildings during service. The consequences of early-age cracking are summarized
as follows.
Firstly, the occurrence of cracks will affect the appearance and aesthetic perfor-
mance of the concrete structure and raise public concern and attention regarding
safety issues [21]. There is no rational basis for defining the aesthetic implications
of cracking. Campbell-Allen proposes a graphical aid for determining the aestheti-
cally acceptable crack width in relation to the nature of the structure and the viewing
distance after reviewing the work on human reactions to cracks [22], as shown in
Fig. 1.1.
Secondly, the existence of cracks will affect the functionality of concrete structures
with the development of age, especially in some structures such as the side wall of
basement, reservoirs, water channels and pipelines, sluice, sewage treatment tank of
waterworks, as well as roofs [23]. Some early-age cracks may be continuous through

@seismicisolation
@seismicisolation
1.1 Early-Age Cracking of Concrete 3

Fig. 1.1 Aesthetically 12

Monumental buildings
acceptable crack widths [22] 11 Prestige public buildings
10 Domestic buildings
Factory and commercial buildings
9 Parking stations and garages
Little used and scarcely seen areas

Viewing distance (m)

8
7
6
5
4
3
2
1
0
0.0 0.1 0.2 0.3 0.4
Crack width (mm)

the full thickness of a section and can lead to seepage or leakage of these structures,
which will also cause harm to the users and their property, especially electronic
equipment [21].
Thirdly, cracks are not allowed in some special functional structures, such as
nuclear power plants, chemical and radioactive waste repositories, liquefied natural
gas storage tanks, and waste treatment structures. If cracks occur in these concrete
structures, chemicals leak through the cracks, which will be a disaster [24].
Fourthly, early-age cracking, if left unchecked, may lead to long-term mainte-
nance issues. These cracks impair the durability of the concrete structures by letting
water with chemical agents and gases infiltrate inducing carbonation, steel corro-
sion, chemical attack and/or freeze–thaw damage, and also cause huge resources and
economic damage [25]. If the cracks in concrete continue to develop, which may
affect the bearing capacity of the structure and accelerate fatigue failure [26, 27].
Additionally, the cracking of concrete will also increase the maintenance costs.
These cracks affect the appearance of structures and raise public concern and attention
regarding safety issues [21]. Therefore, the cracks should be repaired as soon as
possible, however, improper remedial work leads to further problems, such as water
ponding [21]. When the development of cracks affects the functionality and durability
of the structure, some materials need to be used to block the cracks in time. In
addition, when cracks reduce the bearing capacity of the structure, strengthening
the structure in time is very necessary. The occurrence of cracks not only brings
economic loss to users, but also increases the maintenance cost during the service of
concrete structures.

@seismicisolation
@seismicisolation
4 1 Introduction

1.1.2 Causes of Early-Age Cracking

Generally, the concrete cracks can be grouped according to the characteristics of

cracking, such as random, map, transverse, longitudinal, corner, and re-entrant cracks
[20]. The grouping of cracks according to the crack characteristics was not only
limited to early-age hardened concrete, but also included hardened concrete at later
ages.
Early-age cracking may be caused by several causes such as thermal expan-
sion, overloading, restraint, chemical reactions, construction measures, and curing
methods. The main causes of early-age cracking can be grouped as follows.

1. Tensile strength

The tensile strength of early-age concrete is relatively low although it increases with
age [28]. The values of mean tensile strength at 3 d are around 60% of the values at
28 d in accordance with data of EN1992-1-1 [29]. Micro-cracking may develop and
propagate at early age due to the low tensile strength of concrete. Cracking occurs
when the restrained tensile stress exceeds the tensile strength of concrete. However,
the failure stress caused by the temperature and restraint is lower than the direct
or split tensile strength at the same age. Some premature failures are found even
when the tensile stress approaches 0.50 of the direct tensile strength [30]. Besides
the stress approach, the strain-based approach has also been adopted to assess early-
age cracking. Strain capacity in tension defines the allowable magnitude of tensile
strain of concrete before cracking. If the restrained strain exceeds the tensile strain
capacity, the crack would occur [31].

2. Temperature

As the hydration reaction is exothermic, and the thermal conductivity of concrete

is relatively low, it normally endures temperature rises that have special relevance
in concrete structures [32]. Temperature rise is more obvious in mass concrete [33,
34]. Two types of relevant thermal gradients can be identified in mass concrete: (1)
one is relative to time, i.e. a given point of the structure has its temperature varying
throughout time; (2) another is a spatial gradient that corresponds to the temperature
difference, at a given instant, between two different points of the structure. Thermal
deformation is associated with the hydration rate, maximum temperature, decreasing
rate of temperature. Thermal deformation is the result of either differential expansion
within an element during heating causing internal restraint or by external restraint to
contraction of an element on cooling from a temperature peak. Early-age cracking
occurs within a few days in thinner sections, but it may take longer to develop in
thicker sections which cool more slowly. Thermal strains driven by high temperature
gradients occur between the interior and surface of structural elements since concrete
has poor thermal conductivity. The partial and fully restrained movement from the
cooling or other temperature change induces tensile stress in concrete. If such strains
or stresses reach a certain limit, undesirable cracks can occur [35].

@seismicisolation
@seismicisolation
1.1 Early-Age Cracking of Concrete 5

3. Shrinkage

Shrinkage mainly includes plastic shrinkage, chemical shrinkage, autogenous

shrinkage, drying shrinkage, and carbonation shrinkage, which have overlapping
results but with different mechanisms [36]. It is observed that autogenous shrinkage
accounts for the most significant volume change in HPC at early ages compared to
other types of shrinkage.
Autogenous shrinkage is the macroscopic volume reduction of cementitious mate-
rials when cement hydrates after initial setting [8]. It happens when the hydra-
tion process consumes water and causes internal drying, resulting in a decrease
in the material volume [37]. The autogenous shrinkage of concrete is a complex
phenomenon influenced by many factors including: fineness of cement, cement type,
supplementary cementitious material, aggregate, fiber, w/c ratio, admixtures, and
curing method [38, 39]. At the early stage, when the concrete is still soft, autogenous
shrinkage is only attributed to the chemical changes driven by cement hydration, but
at a later stage (after approximately 5 h), it is mostly due to self-desiccation, which is
intensified by certain mix ingredients (silica fume and superplasticizer) and low w/c
ratio. During the early age, tensile strength of concrete is too low to resist the crack
propagation caused by shrinkage stresses. Since there is volume unbalance between
reactants and products of reaction, autogenous shrinkage imposes additional strains
to concrete that may cause the cracking tendency to increase. Autogenous shrinkage
leads to micro-cracks in the matrix and lowers the durability of the structure.
4. Restraint

Early-age cracking is greatly affected by restraint [39]. If there are no restraint,

concrete will not crack. The internal restraint can be generated on the surface of the
aggregate particles contained in the concrete. The external restraint can be generated
from adjacent older concrete elements, sidewalls, baseplates, subsoils, subgrades, etc.
[40]. Internal restraint alone occurs only in the totally unrestrained concrete member.
However, all structural concrete elements are more or less externally restrained.
Hence, internal and external restraint nearly always coexist.
If the free deformations are hindered, restraint will arise. The magnitude of
hindrance is often described by a restraint factor. Nearly all structural members
are restrained. The degree of restraint varies, it depends on several parameters which
will be dealt with later. Restraint to movements occur both externally and internally.
External restraint can be mainly grouped into two types, as shown in Fig. 1.2 [29]:
(1) End restraint: Restraint may arise at the end and at inner supports of beams,
slabs, frames, etc., if the deformation is impeded there by neighboring members;
(2) Continuous restraint: If a slab or foundation is cast directly on to the subsoil,
restraint actions are built-up from the ends of the member. Especially on rock, the
restraint may become high. On clayey soil and gravel, the shear resistance of the soil
and other factors influence the magnitude of restraint.
Internal restraint arises when one part of the freshly placed section expands or
contracts differently to another part of the same section. In very thick sections, internal
restraint tends to dominate, while in thinner sections external restraint is dominant.

@seismicisolation
@seismicisolation
6 1 Introduction

(a) Continuous restraint (b) End restraint

Fig. 1.2 External restraint

In a few situations both types of restraint need to be considered. Restrained ring

test is designed to simulate the continuous restraint [41], and TSTM is designed to
simulate the end restraint [42].

1.1.3 Measures for Controlling Early-Age Cracking

Early-age shrinkage induced cracking has become one of the major challenges
faced for the popularization and application of concrete [43]. Therefore, increasing
the cracking resistance is essential to ensure the safety and durability of concrete
structures [15]. Taking effective measures to reduce the shrinkage and increase the
cracking resistance of concrete is an important task. At present, many methods to
control early-age cracks have been used in practical concrete engineering. Measures
can be taken mainly from the aspects of mitigating the drop of IRH [44, 45], control-
ling the change of temperature in concrete [46], decreasing the early-age shrinkage
[47], increasing the tensile strength of concrete [48], decreasing the restraint degree
of concrete structure [49], etc. The main measures are summarized as follows.

1. Mitigating the drop of IRH in concrete

The shrinkage-induced cracking of concrete is associated with the drop of IRH caused
by moisture loss [44]. Furthermore, the shrinkage deformation increases proportion-
ally with the drop of IRH [45]. The moisture loss of concrete is related to ambient
drying and cement hydration. Drying shrinkage occurs as the reason that the moisture
in concrete diffuses into the outside ambient when the ambient humidity is lower than
IRH in concrete. Cement hydration consumes moisture and causes the occurrence
of self-desiccation, which will increase the autogenous shrinkage [50, 51].
Curing agents, plastic sheeting, mono-molecular films, water fogging, or wind
breaks in conjunction with properly designed concrete mixtures are used to prevent
the evaporation of moisture in the concrete [28], which can not only prevent the plastic

@seismicisolation
@seismicisolation
1.1 Early-Age Cracking of Concrete 7

shrinkage cracking, but also increase the cracking resistance under the influence of
autogenous shrinkage or drying shrinkage. In addition, some materials such as SAPs
[44], pre-wetted LWAs [45], and rice husk ash [52] have been used to mitigate the
drop of IRH in the concrete, so as to reduce the drying shrinkage and autogenous
shrinkage of the concrete.
2. Controlling the change of temperature in concrete
The cracks of concrete are more likely to occur when the temperature deformation
is restrained [53]. Meanwhile, the self-induced stress of massive concrete structure
due to the formwork removal or cold wave effect also has a great influence on the
cracking resistance of concrete, leading to a surface crack which may further develop
into a through crack [46]. The early-age cracking can be minimized by controlling
the change of temperature in concrete.
Some measures aimed at reducing the high maximum temperature caused by the
hydration heat are taken. Very low cement contents, limitation of lift-heights, and
pipe cooling has been used to reduce the maximum temperature in mass concrete
[54]. The ratio of sand is reasonably selected to increase the proportion of aggregate
as much as possible and reduce the amount of cement, and this measure is taken under
the condition of ensuring the workability and strength of concrete. Admixtures such
as fly ash and GGBFS are used to replace part of cement, which can delay the peak
temperature due to the slow rate of hydration of fly ash [8, 55]. In addition, liquid
nitrogen and cold water spraying are used to reduce the temperature of raw materials
to control the pouring temperature of concrete, which can reduce the maximum
temperature [53].
The hydration rate is reduced to control the thermal cracking. Some chemical
admixtures such as temperature rising inhibitor are used to reduce the hydration rate
of concrete [56]. The use of temperature rising inhibitor can reduce the hydration
rate, reduce the maximum temperature of concrete, prolong the time to reach the
maximum temperature, and reduce the cooling rate of concrete after reaching the
maximum temperature. Additionally, the phase change materials are used in concrete
to help keep the heat of hydration within the acceptable limit and thus reduce the
rate of deformation and stress development, decreasing the risk of early-age concrete
cracking [57].
Thermal insulation measures such as covering straw bags and quilts are usually
used to reduce the decrease rate of temperature in concrete surface in winter, which
can reduce the temperature gradient and temperature difference between the interior
and the surface of concrete. This measure can be taken to reduce the stress gradient
caused by temperature gradient, so as to increase the cracking resistance of early-age
concrete [53].
3. Decreasing the early-age shrinkage
The early-age shrinkage of concrete mainly includes include plastic shrinkage, auto-
genous shrinkage, drying shrinkage, chemical shrinkage, and carbonation shrinkage.
The cracking resistance of concrete can be improved by decreasing the early-age
shrinkage [53], and some common measures are summarized as follows.

@seismicisolation
@seismicisolation
8 1 Introduction

Different kinds of fibers can be incorporated into concrete to decrease the early-age
shrinkage, which also prevents the propagation of micro-cracks in concrete. Fibers
such as steel fiber (hooked-end fiber, straight fiber, undulated fiber, etc.), polyvinyl
alcohol fiber, polypropylene fiber, and cellulose fiber have been used to decrease the
early-age shrinkage of concrete [58].
Adding an appropriate amount of expansion agent can generate micro expansion
in concrete. The micro expansion can compensate for the shrinkage in the process of
cement hydration and fill the small cracks, so as to control and reduce the generation
and development of cracks [59].
Shrinkage reducing admixture significantly reduces the shrinkage of HSC at early
age by reducing the surface tension of pore solution [60]. Additionally, the shrinkage
of concrete is reduced by reducing the content of cement. An appropriate amount of
fly ash or GGBFS with reasonable fineness is used to replace part of cement, which
can decrease the early-age autogenous shrinkage of concrete [8, 55].

4. Increasing the tensile strength of concrete

Improving the tensile strength of concrete is a direct measure to control the early-age
cracking of concrete. Nano-sized materials have been incorporated into cement-based
composites to modify their properties and increase the tensile strength of concrete
[61]. Besides, some supplementary cementitious material such as silica fume is used
to improve the tensile strength of concrete [62].
It has been shown that fibers, even in low volume, will reduce crack widths and
increase the cracking resistance of concrete [28]. The addition of short randomly
distributed fibers has been shown to arrest cracking. Many types of fiber such as
steel fiber, carbon fiber, polyethylene fiber, glass fiber, and polypropylene fiber have
been widely used in engineering to increase the tensile strength of concrete [15, 53].

5. Decreasing the restraint degree of concrete structure

The early-age shrinkage may lead to the development of stress and contribute to
the occurrence of cracking when the concrete is restrained internally by aggregates
and externally by other components of the building structure [58]. The restraint
condition is also one of the most crucial factors for estimation of the early-age
concrete cracking.
The measures to reduce the degree of internal restraint can be achieved by selecting
aggregates with reasonable elastic modulus.
The reduction of external restraint is one of the most economic methods of
increasing the early-age cracking resistance and this can be achieved through the
sequence and timing of the construction. Several methods have been developed to
reduce the degree of end restraint. Firstly, a reasonable section form can reduce
the degree of end restraint. Secondly, for mass concrete structures, the sequence
method can be used for construction, which is to divide the building foundation or
large-area concrete plane structure into several areas and pour concrete in layers
and blocks [21]. Additionally, the horizontal or vertical construction joints or the
post-cast strip can be reasonably set at appropriate positions to reduce the degree of

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 9

end restraint. Meanwhile, some measures have been taken to reduce the degree of
continuous restraint. The sliding layer, such as asphalt glue, is usually set between
mass concrete foundation and rock foundation or between the foundation and thick
concrete cushion to reduce the degree of continuous restraint.
Although free shrinkage measurements can be used to directly compare different
mixture designs, they do not provide sufficient information to determine if a concrete
mixture will crack in service. The accurate prediction of early-age cracking is
complex, since it is dependent on several factors including the free shrinkage (rate
and magnitude), time-dependent material property development, stress relaxation
(creep), fracture resistance, structural geometry, and the degree of structural restraint
[28]. Revealing the development law of restrained stress of concrete under different
restraint degrees can accurately predict the cracking risk.
In summary, the occurrence of early-age cracks has a great impact on the appear-
ance, functionality, safety, and durability of concrete structure. The maintenance
cost also has to be increased to repair the cracks. Many measures have been devel-
oped to control the occurrence and propagation of cracks. The development of new
controlling measures is of great significance to ensure the functionality, safety, dura-
bility, and the requirements of low-cost maintenance of buildings with the continuous
emergence of new structural forms and the demand of rapid construction. As a new
measure to control the early-age cracking in concrete, IC technology has been applied
in a certain range in practical engineering [50]. Investigations on IRH, autogenous
shrinkage, tensile creep, and cracking resistance under the continuous and uniaxial
restrained conditions of early-age ICC are necessary for better analysis, design,
construction, and management for cracking prevention of concrete structures.

1.2 State-of-the-Art of Internal Curing

1.2.1 What is Internal Curing

In order to reduce the early-age cracking potential of concrete, the curing of concrete
is usually carried out, which is generally called as traditional curing. However, the
low w/c ratio of concrete makes the internal structure denser than that of ordinary
concrete, resulting in the poor permeability, as shown in Fig. 1.3. The water provided
by traditional curing is difficult to penetrate into the concrete, and the curing effect
cannot achieve the expectation [63]. Therefore, the early-age cracking of concrete
cannot be effectively alleviated with traditional curing.
IC is being used mainly to reduce early-age cracking by maintaining a high rela-
tive humidity within the hydrating cement paste by releasing the water absorbed by
IC materials. IC can delay the drop in the critical pore size that remains saturated
by providing readily-available water from IC materials to fill the voids created by
chemical shrinkage. In 2003, the International Union of Laboratories and Experts
in Construction Materials, Systems and Structures (RILEM) described the IC of

@seismicisolation
@seismicisolation
10 1 Introduction

Curing water

Traditional curing
Penetration
Internal curing area

Initial phase Curing phase

Normal Internal Curing

aggregate curing area
material

Fig. 1.3 Traditional curing versus IC [66]

concrete as the introduction of additional water into the concrete for curing [64].
American Concrete Institute (ACI) defined IC in its ACI Terminology Guide in 2010
as ‘supplying water throughout a freshly placed cementitious mixture using reser-
voirs, via pre-wetted LWAs, that readily release water as needed for hydration or to
replace moisture lost through evaporation or self-desiccation’. In 2011, Bentz and
Weiss [65] provided a state-of-the-art review of the subject of IC, first addressing
its history and theory, and then proceeding to summarize published guidance on
implementing IC in practice and published research on its influence on the perfor-
mance properties of concrete. In 2017, National Concrete Pavement Technology
Center proposed a guide specification for ICC [66]. In 2021, Yang et al. [67] gave an
overview of factors affecting the effectiveness of IC in cement-based materials from
three aspects: the amount of IC water, characteristics of IC materials, and migration
distance of IC water.
With the development of IC, many materials can be utilized, such as SAPs [68],
LWAs [63], and rice husk ash [52]. SAPs are a kind of hydrogels which can absorb
water as high as 500–1500 g/g whereas the absorption capacity of common hydro-
gels is no more than 1000 g/g [69, 70]. Generally, the SAPs can be divided into
two categories based on the mechanism of water absorption: chemical and physical
absorption; according to the different raw materials, it can be divided into about
six categories: (1) starch [71]; (2) cellulose such as cardiomyopathy cellulose-based
SAPs [72]; (3) protein [73]; (4) synthetic of polymers such as the copolymers of
acrylic acid and acrylamide [74]; (5) chitosan [75]; and finally (6) the blends and

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 11

composite such as organic–inorganic hybrid materials-based super-absorbent poly-

mers [76, 77]. It can also be classified into two types by the method of cross-linking:
the formation of network structure through addition of the cross-linker, such as poly
(acrylic acid) cross-linked by starch/acryl ate cross-linked [78]; the self-crosslinking
SAPs such as polyacrylates [79] and polyacrylamides [80].
In the 1970s, the first SAPs were developed in the Northern Regional Labora-
tory of the US Department of Agriculture, which was synthesized from starch-graft-
hydrolyzed acrylonitrile products [32]. In 1978, Japan began the commercial produc-
tion of SAPs for female napkins [33]. After further developments, in 1980s, Germany
and France employed SAPs materials in baby diapers [34]. In late 1990, world
production of SAPs resins reached one million tons. In 2001, the concept of SAPs was
firstly proposed by Jensen and Hansen [68] for the prevention of self-desiccation in
hardening cement-based materials. In 2007, RILEM Technical Committee 225-SAP
coordinated research efforts and compiled the results of studies into mechanisms of
internal curing using SAPs and effects of SAPs addition on HSC properties in the
fresh and hardened state. In 2008, Lura et al. [81] proposed the preliminary results
about 3D distribution of SAPs in mortars. In 2012, Wyrzykowski et al. [82] proposed
a mechanistic-type numerical model of cementitious materials. In 2014, RILEM
Technical Committee 260-RSC collected and analyzed information on applications
of SAPs in concrete as well as initiate necessary research activities for clarifying
open questions related with practical use of SAPs. In 2019, International Conference
on the Application of SAPs and Other New Admixtures Towards Smart Concrete
gathered the papers devoted in major part to the studies of the behavior of SAPs in
cement-based materials, mainly concrete (in particular the absorption behavior) as
well as the effect of the SAPs on fresh and hardened concrete properties. In 2020,
Boshoff et al. [83] presented two commercially available SAPs samples with different
chemical compositions tested in terms of their ability to mitigate plastic shrinkage
cracking of concrete, which was based on the inter-laboratory study performed by
six international research groups in the framework of RILEM Technical Committee
260-RSC. In 2022, Schröfl et al. [84] summarized the recent progress in SAPs for
concrete and focused on the relationships between absorptivity in terms of absorp-
tion/desorption kinetics and retention of the absorbed liquid as well as the polymer
physico-chemical principles.
LWAs are a kind of porous material, including coarse aggregate with bulk density
less than 1100 kg/m3 and fine aggregate with bulk density less than 1200 kg/m3
[65]. Clay ceramsite, shale ceramsite, lytag, and pumice, are commonly used for IC
[85]. The differences in IC are mainly related to their pore structure and porosity
which affects the efficiency of IC [86]. High porosity increases water absorption
rate and curing effectiveness, but reduces the strength of concrete, especially at
early age. Usually, water absorption of high strength LWA at 1 h is under 10%,
but water absorption of ordinary LWAs is about 15–20% [87]. Large pores or low
water absorption will reduce the efficiency of IC [88]. LWAs, therefore, should
have an appropriate porosity and pore structure to achieve better internal curing
effectiveness and higher strength. By the way, low open porosity, which can achieve

@seismicisolation
@seismicisolation
12 1 Introduction

a low absorption rate and a small thermal conductivity coefficient, is as a main target
in the preparation of LWAs.
The first published recognition of the IC potential of LWAs is likely that of Paul
Klieger in 1957 [89], who wrote “Lightweight aggregates absorb considerable water
during mixing which apparently can transfer to the paste during hydration”. In 2001,
Bentur et al. [90] investigated restrained autogenous shrinkage in HSC with pre-
wetted LWAs. In 2003, Lura investigated the autogenous shrinkage of concrete with
pre-wetted LWAs. In 2004, Zhutovsky et al. [91] verified that LWAs acted as an
internal water reservoir preventing reduction of relative humidity in the cementitious
matrix, and this kind of curing could be successfully applied to obtain improved HPC
with reduced sensitivity to cracking. In 2009, Henkensiefken et al. [92] evaluated the
water absorption in internally cured mortar made with water-filled LWAs. In 2014,
Lura et al. [93] investigated the IC performance of pre-saturated LWAs produced
from biomass-derived waste. In 2017, Zhutovsky and Kovler [245] demonstrated
that w/c ratio had a considerable impact on cracking potential of ICC with LWAs.
In 2021, Kovler organized the RILEM Technical Committee 196-ICC to develop
practical recommendations for utilizing of advanced methods of LWAs in concretes.
The IC material acts as an internal “reservoir” after absorbing water in advance.
When there is insufficient water due to self-desiccation effect in the concrete, it
can timely provide water from the interior, so as to control the decline of internal
relative humidity and the development of shrinkage in the concrete [68]. However,
the investigations on the cracking control of ICC with LWAs and SAPs are still
limited, which facilitates the further studies.

1.2.2 How Does Internal Curing Work

The water absorbed by the IC agents such as LWAs and SAPs before casting can be
used to supply the water for hydration of cement. When the w/c ratio of concrete is
less than 0.42, the proportion of SAPs will increase the hydration degree of cement
and promote the formation of voids in concrete [94]. However, at the same time,
the introduction of water into SAPs will increase the porosity whether the pores are
filled with water or not [65]. Due to the water absorption of SAPs, when the water
required for cement hydration is insufficient, SAPs will gradually release water to
the surrounding cement, promote cement hydration, increase the degree of cement
hydration and improve the strength of concrete [95]. Mechtcherine [64] believed that
the shrinkage of concrete would be greatly reduced due to the addition of SAPs, so the
density of concrete would be improved and the generation of internal microcracks
would be reduced. However, with the increase of SAPs proportion, the IC water
gradually increases, which will have a negative effect on the splitting tensile strength
of concrete.
Cement or mortar with lower density is floated in concrete, while normal coarse
aggregates are submerged [96]. When concrete is mixed with LWAs, LWAs are
floated due to the lower density [97]. The floatation of the LWAs not only reduces

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 13

the effective evaporation area, but also reduces bleeding [97]. The fraction of the
paste within LWAs also increases with the increase in the replacement rate, which
makes the paste obtain water supply more effectively. As a result, within the certain
LWAs proportion range, the microstructure of the paste became denser and stronger
to resist cracking, thus postponing the time to cracking, as reported in [98]. However,
when the proportion of LWAs exceeded the limit value, due to the low strength of
LWAs, the axial tensile strength of concrete decreased, and the time to cracking
was not postponed. Although the high proportion of LWAs had an adverse effect on
the performance of concrete, LWAs could still be considered as the advantageous
material in IC to reduce the shrinkage and postpone the time to cracking of concrete.

1.2.3 How to Prepare Internally Cured Concrete

Firstly, two common IC agents (i.e., SAPs and LWAs) are recommended with the
proper type. Secondly, water absorption and release of IC agents are introduced.
Thirdly, the mixture proportioning with IC is presented with a detailed calculation
process. Finally, the mixing process of ICC with SAPs and LWAs is summarized.
1. Materials

Many IC agents have been invented [67] including SAPs, LWAs, and high-absorption
porous materials such as rice husk ash [52], miscanthus combustion ash [99], and
cellulose fibers [100]. Among them, the most widely investigated and used IC agents
are SAPs and LWAs [101–104]. SAPs can be made by solution or suspension poly-
merization to obtain particles of different sizes and shapes [105, 106]. For application
in concrete: covalently cross-linked polymers of acrylic acid and acrylamide, neutral-
ized by alkali hydroxide, have been proven to be efficient [107]. Due to the presence
of pores in the LWAs, the weight of high-strength lightweight aggregate concrete
is 20–40% lower than conventional concrete [108], with density values of less than
2000 kg/m3 [109]. LWAs can be divided into three types according to the source:
natural LWAs such as pumice, scoria and porous tuff; artificial LWAs such as clay
ceramsite, shale ceramsite, lytag, expanded perlite and expanded vermiculite; and
industrial waste slag such as expansion slag beads, spontaneous combustible coal
gangue and cinder [85]. Among them, artificial LWAs are the most suitable IC mate-
rial because the pore structure and performance can be designed to meet the practical
requirements through the composition of raw materials and burning conditions [85].

2. Water absorption and release of internal curing agents

Two common IC agents are SAPs and LWAs. SAPs absorb additional IC water before
the addition to the mixture or absorb free water after the addition to the mixture, and
then SAPs gradually release water needed for the second hydration reaction within the
concrete, which can contribute to the formation of dense and uniform gel and improve
the compactness and durability of concrete with SAPs [110]. The release of IC water

@seismicisolation
@seismicisolation
14 1 Introduction

from SAPs causes the uniform and rapid redistribution of pore water in cement paste
at early age, significantly increasing the IRH and improving humidity distribution
[82]. LWAs are used as a water reservoir that can provide water to replenish the
empty pore volume that is created by chemical shrinkage during hydration. Since
water is removed from large pores to small pores, the ideal LWAs would have pore
sizes larger than the pore size that develops in the cement paste [111]. A majority
of water is lost at a high RH (RH > 96%) implying the pores in LWAs are large and
the water is available to be lost at high relative humidities, which is preferred for IC.
Also, the samples release almost all (99%) of their moisture when an RH of 92% is
reached, which implies that the water will leave the pores of LWAs if a large enough
suction pressure (or a low enough IRH) exists [96]. The water absorption capacity of
IC agents can be tested by direct or indirect methods, including the tea-bag method
[112–114], the equal slump test [43], the filtration method [115], the hydration heat
method [68], the laser diffraction particle size analysis [116], and the traditional
air-void analysis [117].
The absorption capacity (ACtb ) at a specific time through the tea-bag method could
be calculated with Eq. (1.1), as reported in [115].

m3 − m2 − m1
ACtb = (1.1)
m1

The absorption capacity (ACf ) at a specific time through a filtration method could
be calculated by Eq. (1.3), as reported in [115].

mb − mc
ACf = (1.2)
m1

where m1 is the mass of the dry SAPs, in g; m2 is the mass of the pre-wetted tea-bag,
in g; m3 is the mass of the tea-bag (with the hydrogel inside) at a specific time, in g;
mb is the mass of filtered fluid at a specific time, in g; and mc is the mass of added
test fluid, in g.
3. Mixture proportioning with internal curing

The mass of LWAs needed to supply water for IC can be calculated with Eq. (1.3),
as reported in [16, 63].

Cf · CS · αmax
MLWA = (1.3)
S × φLWA

where MLWA is the mass of (dry) LWAs needed per unit volume of concrete, in kg/m3 ;
Cf is the cement factor (content) for concrete mixture, in kg/m3 ; CS is the chemical
shrinkage of cement, in g of water/g of cement; αmax is the maximum expected degree
of hydration of cement, as shown in Fig. 1.4; S is the degree of saturation of aggregate
(0–1); and φLWA is the absorption of lightweight aggregate, in kg of water/kg of dry
LWA.

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 15

Fig. 1.4 Calculation of the 1.0

maximum cement hydration
degree αmax
0.8

0.6

α max
0.4

0.2

0.0
0.0 0.1 0.2 0.3 0.4 0.5
w/c ratio

The mass of SAPs in the mixtures was calculated with Eqs. (1.4) and (1.5), as
reported in [114].

Cf · CS · αmax
MSAP = (1.4)
ϕSAP,30
MSD,TB+SAP − MSD,TB − MD,SAP
ϕSAP,30 = (1.5)
MD,SAP

where MSAP is the mass of the dry SAP, in kg/m3 ; Cf is the cement content of
the mixture, in kg/m3 ; CS is the chemical shrinkage of the cement, in mL water/g
cement; αmax is the expected maximum degree of hydration, which ranges from 0
to 1 (unitless); ϕSAP,30 is the 30 min absorption of SAPs in pore solution, in kg
of solution/kg of dry SAP; MSD,TB+SAP is the mass of the teabag with SAPs in the
soaked surface dry condition, in g; MSD,TB is the mass of the teabag alone in the
soaked surface dry condition, in g; and MD,SAP is the mass of the dry SAP, in g.

4. Concrete mixing
For ICC with SAPs, two mixing methods are utilized to add SAPs into concrete:
SAPs pre-wetting process and SAPs dry-mixing process [44, 118, 119]. SAPs tend
to agglomerate during mixing after pre-absorption, which leads to poor dispersion
in concrete and influences the actual IC efficiency [125]. The pre-wetting process
of SAPs can achieve higher spread and better workability, whereas the dry-mixing
process of SAPs can ensure uniform dispersion of the polymer [120]. The processing
of concrete internally cured with SAPs is summarized in Table 1.1.
For ICC with LWAs, the mixing procedure was conducted as per ASTM C192-19
[127]. Before mixing, all LWAs were oven-dried, air-cooled, and then submerged
in water for 24 ± 1 h. The volume of water used to submerge the LWAs included
both mixing water and the water absorbed by the LWAs in 24 h [96]. The excess

@seismicisolation
@seismicisolation
16 1 Introduction

Table 1.1 Processing of concrete internally cured with SAPs [118]

Mixing processes Moulding References
(1) Adding and mixing SAPs and all 100 mm × 100 mm × 100 mm Justs et al. [121]
dry cementitious materials for 20 s (compressive test)
(2) Adding 90% of water during the 25 mm × 25 mm × 100 mm
period of 30 s (flexural test and static elastic
(3) Wet mixing for 60 s modulus test)
(4) Adding superplasticizer, the rest
of water and mixing for 90 s
(5) Adding fiber during the period
of 45 s
(6) Mixing for 120 s
(1) Pre-mixing cement powder, 150 mm × 150 mm × 150 mm Assmann et al. [122]
oven-dried aggregates and dry (compressive test)
SAPs for 30 s Dog-bone shaped specimens
(2) Adding the first half of water (uniaxial tensile test)
and mixing for 120 s ϕ150 mm × 300 mm cylinders
(3) Adding the second half of water (elastic modulus test)
together with the superplasticizer
and mixing for 5 min
(1) Mixing dry cement and 100 mm × 100 mm × 100 mm Mönning [120]
aggregates for 30 s (splitting tensile and compressive
(2) Adding approximately 50% by test)
mass of the water and mixing for ϕ150 mm × 300 mm cylinders
90 s (Young’s modulus test)
(3) Adding the SAPs and mixing
for 60 s
(4) Adding superplasticizer or air
entraining agent and the remaining
water and mixing for 180 s
(1) Pre-wetting the SAPs 100 mm × 100 mm × 300 mm Wang et al. [123]
(2) Mixing the pre-wetted SAPs (autogenous shrinkage test)
and other dry materials for 60 s 150 mm × 150 mm × 150 mm
(3) Adding water and mixing for (internal relative humidity test)
60 s
(1) Mixing cement, sand and 80 mm × 100 mm × 300 mm Kong et al. [124, 125]
aggregate for 1 min (shrinkage test)
(2) Adding water together with the 100 mm × 100 mm × 100 mm
pre-soaked SAPs to the dry mixture (water absorption, moisture
and mixing for 3 min content, chloride ion penetration
and compressive test)
(1) Mixing cementitious materials 100 mm × 100 mm × 100 mm Azari Jafari [126]
and aggregates for 60 s (compressive test and electrical
(2) Adding water and resistivity test)
superplasticizer and mixing for 200 mm × 200 mm × 200 mm
5 min (water absorption test)
(3) Adding the pre-saturated SAPs 75 mm × 75 mm × 280 mm
to the mixture and mixing for 5 min (shrinkage test)

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 17

water (water not absorbed in 24 h) was decanted from the aggregate and then used
as the mixing water [128]. Both the water and cement were conditioned at room
temperature for a minimum of 24 ± 1 h [128]. The processing of concrete internally
cured with LWAs is summarized in Table 1.2.

1.2.4 How Does Internal Curing Work on Internal Relative

Humidity

The loss of water due to self-desiccation caused by cement hydration may result in a
decrease of IRH in concrete [132–135]. Moreover, when the concrete is exposed to
the external drying environment, the IRH will be further reduced by moisture diffu-
sion [136, 137]. Once the w/c ratio is less than a critical value, the decrease of IRH
will induce shrinkage [96, 138, 139]. Normally, the shrinkage deformation increases
in proportion to the decrease of moisture content [140–142]. Concrete may crack
when the tensile stress induced by restrained shrinkage inside concrete develops
beyond the tensile strength of concrete [6]. Moreover, the IRH in concrete affects
the cement hydration rate, strength, and durability of concrete [143]. Experimental
results reported in [144] demonstrate that the early age is one of the most important
period in the lifetime of cementitious materials. The decrease of IRH will increase
shrinkage, affect creep, and indirectly affect the cracking resistance of concrete at
early age [145]. Therefore, the variations of IRH should be measured for under-
standing the cracking resistance of concrete in depth at early age. In 2015, Shen et al.
[44] investigated the effect of IC with SAPs on the IRH of early-age concrete, the
schematic diagram of concrete IRH measurement set-up is shown in Fig. 1.5.

1. Internal relative humidity of early-age concrete internally cured with SAPs

During water mixing, SAPs absorb water and swell. An internal water reservoir is
built into fresh concrete and serves as a curing agent by gradually releasing the
water absorbed during the hydration process [147]. Additional curing water is thus
produced through SAPs, and the decrease in the IRH within concrete is counteracted
[68]. In view of this finding, investigation into the variations of IRH in concrete
with SAPs addition is fundamental to evaluate the cracking performance of ICC. In
HSC under sealed conditions, the drop in internal humidity as a result of the self-
desiccation effect is notably postponed by the addition of SAPs [124]. The moisture
distribution in low-strength concrete with a high w/c ratio is mainly influenced by
moisture diffusion as a result of drying rather than as a result of self-desiccation. By
contrast, self-desiccation considerably influences moisture distribution in HSC with
a low w/c ratio [132]. Therefore, the changes in the IRH of ICC under the effects
of self-desiccation and moisture diffusion are of considerable practical importance
[137].

@seismicisolation
@seismicisolation
18 1 Introduction

Table 1.2 Processing of concrete internally cured with LWAs [118]

Mixing processes Moulding References
(1) Mixing the normal fine and 100 mm × 100 mm × 300 mm Zhang et al. [129]
coarse aggregates (internal relative humidity and
(2) Adding cement followed by the temperature test)
required water with the
supperplasticizer mixed in, and the
mixing for 2 min
(3) Adding pre-soaked lightweight
aggregates and mixing for another
2 min
(1) Loading the dry, fine, ϕ100 mm × 200 mm cylinders Schlitter et al. [130]
normal-weight aggregate and the (split tensile, compressive strength,
pre-wetted fine LWAs into the and elastic modulus tests)
“buttered” mixer
(2) Decanting the mixing water from
the fine LWAs where applicable
(3) Starting the mixer and adding
50% of the remaining mixture water
(4) Adding the cement and
remaining mixture water containing
the high-range water-reducing
admixture
(5) Mixing the mortar for 3 min,
rested for 3 min while the sides of
the mixer were scraped, and then
mixed for a final 2 min
(1) Oven drying all aggregates and ϕ100 mm × 200 mm cylinders Golias et al. [120]
cooling for 24 h before mixing (compressive strength test, and
(2) When mixed “Dry”, mixing the water sorptivity test)
oven-dry LWA, NWA and IC water,
resting for 5 min, and then
continuing mixing
(3) When mixed using the LWAs as
24 h pre-wetted (Wetted),
submerging the LWAs in both the
mixing and IC water to absorb water
for 24 h ± 1 h
(4) When using the LWAs as
vacuum saturated, oven-drying the
LWAs and placing in a desiccator
under a 738 mm Hg vacuum for
15 min. Then adding both the
mixing and IC water to the LWA.
The aggregate and water were kept
under vacuum for 3 h with periodic
agitation
(continued)

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 19

Table 1.2 (continued)

Mixing processes Moulding References
(1) Mixing the dry constituents 25 mm × 25 mm × 285 mm Dang et al. [131]
manually for about 60 s (drying shrinkage test)
(2) Adding water along with the ϕ50.8 mm × 101.6 mm cylinder
dissolved superplasticizer, further (strength test)
mixing the wet mixture for 120 s. ϕ50.8 mm × 50.8 mm cylinder
(3) Mixing the dry constituents (water absorption test)
(excluding the fine LWA) for 60 s ϕ50.8 mm × 10 mm disk (Gas
(4) Adding water with the dissolved permeability test)
superplasticizer, and mixing the wet
mixture for 120 s
(5) Adding the fine LWAs and
mixing for another 90 s until a
homogeneous mixture is obtained

2 2
6 8 6 7
1. Temperature and humidity sensor
5 5 2. Connection to data collector
75 mm 75 mm 3. Plastic mould
9 9 4. Film
1 3 1 3 5. Spacer rod
4 4 6. Polyvinyl chloride tube
7. Surface exposed to air
150 mm 150 mm 8. Film and tinfoil
9. Concrete
(a) (b)

Fig. 1.5 Schematic diagram of concrete IRH measurement set-up: (a) sealed concrete specimens;
(b) unsealed concrete specimens [44, 146]

Several prediction models for autogenous shrinkage of concrete at early age are
proposed using the experimental data on IRH variations in ordinary concrete [8,
148–150]. Therefore, the prediction model for IRH is important for studying the
autogenous shrinkage and then evaluating the cracking resistance of early-age ICC.
The IRH variations in concrete are significantly affected by IC w/c ratio [94] and the
length of the water–vapor saturated stage with 100% RH (defined as critical time in
[94]) in HSC is affected by the addition of SAPs [44]. The IC water is much more
effective in increasing the IRH in concrete than the corresponding additional mixing
water [124]. The working w/c ratio [45] (the w/c ratio that actually affects the IRH in
concrete) may be higher than the total w/c ratio (effective w/c ratio [151] + IC w/c
ratio [124]), which will affect the change of IRH in concrete. The model for IRH in
early-age concrete internally cured with SAPs considering the amount of IC water
is proposed in [44], and the prediction model for IRH in early-age ICC with SAPs
in consideration of working w/c ratio and critical time is also meaningful.

@seismicisolation
@seismicisolation
20 1 Introduction

2. Internal relative humidity of early-age concrete internally cured with pre-wetted

LWAs
LWAs can also serve as IC water reservoirs to provide curing water to compensate
for the moisture loss of hydrating cement paste [152–154], and thus reduce the
effects of self-desiccation and prevent the IRH from decreasing in concrete [128].
Therefore, investigation on the IRH variations in early-age ICC with pre-wetted
LWAs is fundamental to evaluate its cracking resistance. The IRH in concrete is
also influenced by the IRH decrease rate [155]. The critical time and IRH decrease
rate in early-age ICC with pre-wetted LWAs under unsealed condition are important
indices when proposing the prediction model for IRH in concrete. A linear model
for the prediction of IRH in hardened concrete is proposed in [156], and this model
is improved by adjusting some parameters in [157]. Models for the prediction of
IRH in concrete are also proposed in some literatures [147, 149, 150] considering
the w/c ratio and concrete age. Moreover, the theoretical models for the prediction of
IRH and cement hydration in concrete are also proposed in [143, 158]. The models
utilized for description of phenomena taking place during IC are proposed in [82,
153], and a model for IRH in ICC with SAPs considering amounts of IC water is
also proposed in [44]. Therefore, proposing the model for the IRH in early-age ICC
with pre-wetted LWAs in consideration of working w/c ratio is meaningful.

3. Internal relative humidity of early-age concrete under different curing humidity

conditions

As an important parameter of the environment, the external humidity has a great effect
on the moisture diffusion and variations of IRH in concrete [159–161], which will
affect the autogenous shrinkage, thus affecting the early-age cracking resistance of
concrete. Researches on the change law of the temperature and humidity of concrete
under the conditions of constant or cycling temperature and humidity have been
conducted in [162], and the constant temperature and humidity are 20 °C and 40%,
respectively. Researches on the variations of IRH in concrete are conducted under
the conditions of sealing and unsealing, and the curing humidity is 50% ± 3% [51],
about 30% [163], and 35% ± 5% RH [161], respectively. Moisture diffusion of
ordinary concrete under wetting or drying conditions at three temperature levels
(20, 40, and 60 °C) is studied in [164], the equilibrium relationship among pore
pressure, saturated vapor pressure, and absolute vapor pressure is established. More
recently, a universal and mechanism based prediction model for the distribution
of moisture in concrete at early age is developed in [160], and the environmental
humidity of 0.1, 0.3, 0.5, 0.7, 0.9, and 1.0 is used in the model to investigate the
effect of environmental humidity on the progress of IRH in concrete under drying,
respectively. Some more sophisticated hygro-thermos-chemo -mechanical numerical
models based on hybrid mixture theory have been developed gradually in recent
years, as reported in [165–167]. These models are proposed based on theoretical
or numerical analysis. A simplified prediction model for IRH in early-age concrete
based on experimental results considering the effect of curing humidity is necessary
to be proposed.

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 21

1.2.5 How Does Internal Curing Work on Autogenous

Shrinkage

The autogenous shrinkage is defined as the volume change of cementitious materials

resulting from the capillary pressure induced by the cement hydration and self-
desiccation without transferring water to the surroundings from the macroscopic
point of view [168]. In this section, firstly, the measurement method for autoge-
nous shrinkage is presented and compared. Then, influential factors of autogenous
shrinkage of ICC are discussed with the results in the literature. Finally, predic-
tion models for autogenous shrinkage of ICC are introduced including empirical or
semi-empirical models and mechanistic models.

1. Measurement of autogenous shrinkage

Measurements of autogenous strain have been carried out in two fundamentally

different ways: measurement of volumetric strain and measurement of linear strain
[169]. Volumetric measurements are frequently performed by casting the fresh
cement paste in a rubber balloon that is immersed in a fluid. The volume change
of the cement paste is measured, for example, by monitoring the weight change (i.e.,
buoyancy change) of the immersed sample [10]. Linear measurements are frequently
performed by casting the cement paste in a rigid mold with low friction and recording
the length change by a displacement transducer [169]. The volumetric method can
start the measurements immediately after casting; however, the lack of a steady
contact between the rubber membrane and the cement paste, as well as osmosis
through the membrane used in volumetric measurements are considerable disadvan-
tages [170]. Since the aggregate may damage the rubber bag, the volumetric measure-
ments are not suitable for concrete. Linear measurements are usually adopted, and
prisms or cylinders are often used.
The measurement methods of linear measurements include contact measure-
ment and non-contact measurement. The sensors commonly used in the test device
of contact measurement include linear variable differential transducers [171], dial
gauge [172], and embedded resistance strain gauge [173], while the sensors used
for the non-contact measurement include eddy-current displacement sensor [174],
laser-based sensor [175], and potentiometer sensor [176]. The contact measurement
method starts the measurement after the final setting time. However, the non-contact
measurement method overcomes the shortcomings and can measure the deforma-
tion of concrete continuously and automatically after casting. Thus, the non-contact
measurement method has better convenience and higher accuracy in measuring the
early-age autogenous shrinkage of concrete.

2. Influential factors of autogenous shrinkage of internally cured concrete

The IC method has been used as an effective technique for shrinkage mitigation by
introducing water into concrete to maintain and promote further cement hydration
[177]. The released water contributes to restoring part of the water that has been

@seismicisolation
@seismicisolation
22 1 Introduction

lost through internal or external drying [178]. Some studies have been conducted on
autogenous shrinkage of ICC.
The addition of SAPs is efficient in reducing the autogenous shrinkage of HPC
[133]. Assmann et al. [122] reported that the autogenous shrinkage of reference
concrete with w/c ratio of 0.36 and 0.42 was 95 με and 85 με, respectively, whereas
the autogenous shrinkage of concrete modified with SAPs with w/c ratio of 0.42 was
35 με, indicating that the addition of SAPs in concrete could effectively reduce the
autogenous shrinkage regardless of whether additional water was introduced. LWAs
can also mitigate the autogenous shrinkage of ICC. Cusson et al. [151] reported
that the autogenous shrinkage of ICC with pre-wetted LWAs decreased compared to
that of normal concrete when the total w/c ratios are the same. Akcay et al. [179]
suggested that the use of pre-soaked LWAs as water reservoirs effectively mitigated
the autogenous shrinkage of concrete. The w/c ratio influences the magnitude of
autogenous shrinkage [94]. Shen et al. [14] indicated that the addition of pre-wetted
LWAs reduced the autogenous shrinkage of concrete with different basic w/c ratios.
The cracking potential of concrete can also be reduced by adding fibers [180, 181]
or supplementary cementitious materials including GGBFS and fly ash [182–184].
Kaufmann et al. [185] reported that the effect of 1% PP fiber was remarkable on
the reduction of autogenous shrinkage. Previous works showed that the addition of
steel fibers markedly suppressed the shrinkage of concrete [186, 187]. Reportedly,
as the steel fiber content exceeded a certain amount, the cohesiveness increased and
the mobility of HPC was reduced in return. Huang et al. [186] reported that volume
expansion was observed when GGBFS cement was used. Malhotra [188] showed that
fly ash had an inhibiting effect on autogenous shrinkage due to the deceleration of
the reduction of internal humidity in the matrix. The effect of fiber or supplementary
cementitious materials on autogenous shrinkage of normal concrete has been investi-
gated in-depth. IC plays an important role in mitigating the autogenous shrinkage of
concrete. Thus, investigations on the effect of fiber and supplementary cementitious
materials on the autogenous shrinkage of early-age ICC are necessary.
3. Prediction models for autogenous shrinkage of internally cured concrete
Investigations on the prediction and estimation of the autogenous shrinkage of
concrete [189] based on several classical models [190] have been conducted. The
classical models for predicting the autogenous shrinkage of concrete include Model
Code 2010 [191], EN-1992 [190], ACI 209 model [192], Tazawa-Miyazawa model
[193], Hiroshi Hashida and Nobuyuki Yamazaki model [194], Hua-Acker-Eriacher
model [195], B3 model [196], B4 model [197] and Persson model [148]. Existing
prediction methods for autogenous shrinkage can be divided into two main categories:
(1) empirical and semi-empirical models and (2) mechanistic models. Empirical or
semi-empirical models are models obtained via regression analysis of experimental
data [196, 198, 199]. The parameters usually taken into account are the mixture prop-
erties such as the w/c ratio, mineral compositions of cement and aggregates, stiffness
of aggregates, paste or aggregate volume fraction, and the compressive strength of the
concrete [198, 200]. In mechanistic models, poromechanics approaches for unsatu-
rated porous materials based on changes in pore fluid pressure as the main driving

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 23

force are commonly used to predict autogenous shrinkage (e.g., [132, 195, 201]).
Most of the models for predicting the pore fluid pressure are based on the Kelvin-
Laplace equation [202]. This equation is valid both when changes in capillary pres-
sure or in disjoining pressure are considered as the physical mechanism causing
shrinkage [203, 204].
Investigations on the prediction model of early-age autogenous shrinkage of ICC
are important. Firstly, due to the differences in concrete materials, the existing models
for predicting the autogenous shrinkage do not consider the expansion of ICC at an
early age. Thus, a modification to the existing models for predicting the autoge-
nous shrinkage of ICC is necessary. Secondly, existing several theoretical models
have been established considering quite many factors such as cement composition,
admixture type, compressive strength, and the water-cement ratio of concrete, thus
making the calculation complex and difficult. There is no fast method to obtain early-
age autogenous shrinkage of concrete on site. Ultrasonic velocity measurement is
non-destructive and easy to evaluate the integrity, quality, and strength of concrete.
If the prediction model of early-age autogenous shrinkage strain based on test results
of ultrasonic velocity is proposed, the autogenous shrinkage strain of ICC can be
predicted easily and quickly in actual projects [102]. Thirdly, the measurement of
autogenous shrinkage of concrete is complex in practical engineering, while the
determination of IRH is relatively easier. If the prediction model of early-age auto-
genous shrinkage based on test results of IRH is proposed, the autogenous shrinkage
can be easily estimated [205].

1.2.6 How Does Internal Curing Work on Tensile Creep

The tensile creep evolves gradually deformation under constant tensile stress, which
can relieve approximately 50% of the restrained stress [37] and hinder the formation
of cracking in early-age engineering structures. Accordingly, cracking potential is
highly dependent on creep behavior [206–208]. The creep of concrete occurs in
connection with water movement, which indicates that any interference in the water
regime of concrete will affect creep [122]. The water released by SAP during cement
hydration may influence the creep. In this section, firstly, the measurement for the
early-age tensile creep is presented. Secondly, the effect of SAPs on the tensile creep
is discussed. Finally, the effect of initial strength/stress ratio and loading age on the
tensile creep of ICC with SAPs is discussed, respectively.
1. Measurement of tensile creep
Difficulties in obtaining a uniform stress distribution have restricted the experimental
investigation of the phenomenon of tensile creep. Some test methods are proposed
for the tensile creep of hardened concrete. In 1969, Cook et al. [209] proposed a
device for long-term tension testing, and load is applied manually to the specimen
by tightening the loading nut. The steel bar is embedded in the concrete, and the
tensile force is exerted on the concrete by stretching the steel bar. The test object is

@seismicisolation
@seismicisolation
24 1 Introduction

Fig. 1.6 Lever system

loading method [212]

the concrete of 7 d. In 1974, Domone [210] proposed an axial tensile device to test
the tensile creep of concrete at the age of 28 d. The specimen is dog-bone shaped, and
axial tension is applied by clamping the end of the specimen with a special anchor. In
1977, Brooks et al. [211] proposed a method to test the tensile creep, acoustic gauge
is embedded in the bobbin specimen mold to record the deformation of concrete at
the age of 28 d and 56 d. For testing the tensile creep of concrete before 7 d, in 1985,
Reinhardt et al. [212] proposed lever system loading method, as shown in Fig. 1.6. A
constant tensile load is applied to the concrete through the lever and weight, and the
deformation is recorded through the displacement sensor fixed on the test specimen.
In 1990, Aly [213] proposed spring system loading method, as shown in Fig. 1.7.
The test principle is basically the same as the lever method. A constant tensile load
is applied to the concrete through spring, and the deformation is recorded through
the displacement sensor fixed on the gripped end.

Fig. 1.7 Spring system loading method [213]

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 25

In 1994, Kovler [249] proposed the TSTM system for the determination of tensile
creep of concrete at an age of 1 d. The load cell and the gripped end were connected
through a universal joint to ensure that a uniaxial tensile load was exerted on the
restrained shrinkage specimen by the stepper motor. TSTM can directly test the
deformation after concrete pouring without formwork removal, and can test the
tensile creep at very early age.
In 2001, Altoubat [214] utilized the uniaxial tensile loading device developed
originally to test restrained shrinkage to measure basic creep of concrete. The system
tested two identical dog-bone-shaped samples: one loaded and the other free of
load, which is similar to TSTM system. In 2009, Kamen et al. [215] proposed a
developed TSTM based on Kovler’ research to investigate the tensile creep of ultra
high performance fibre reinforced concrete, and the loading age is 3 d. In 2013,
Ranaivomanana et al. [216] proposed a rigid frame with a hinged lever arm to test
tensile creep of concrete. The specimen is loaded by using calibrated weights stacked
on a platen. In 2018, Al-Manaseer et al. [217] utilized a custom-made tensile testing
machine to test tensile creep. The equipment consists of a stiff frame with two
stainless steel grips: a bottom stationary one and a top mobile grip that is attached to
a closed-loop load cell which can maintain constant load.

2. Tensile creep of early-age ICC with SAPs

The tensile creep is influenced by the mechanical properties which are mainly
influenced by w/c ratio of concrete. Research studies on tensile creep of ordinary
concrete utilizing different w/c ratios under restrained condition and constant tensile
or compressive load have been conducted [139]. Previous experimental results show
that tensile creep of concrete increases [218] or decreases [139] as w/c ratio increases.
In the direct tensile test, the tensile creep increases as w/c ratio increases [219].
However, the tensile creep of concrete decreases as w/c ratio increases under axial
restrained condition at early age [39]. The total w/c ratio increased as the IC water
increased, which made the internal structure of concrete loose and increased the
capillary aperture, as reported in [220]. The tensile creep of concrete under axial
restrained condition decreases as SAPs content increases at early age [221]. Through
direct tensile testing, the basic tensile creep of concrete with a w/b ratio of 0.35 and
containing 0.3% SAPs (by mass of binder) loaded at early ages is studied in [222],
and test results show that concrete with SAPs exhibits higher basic tensile creep
than the reference concrete without SAPs when first being loaded at 1, 6, and 14 d.
Study on the early-age tensile creep of HSC utilizing different SAPs contents under a
constant initial stress/strength ratio (i.e., under a constant tensile load) is meaningful
for evaluating the early-age cracking resistance of ICC.

3. Tensile creep of early-age ICC with SAPs under different initial stress/strength
ratios

The initial stress/strength ratio is defined as the ratio of applied constant tensile
stress to the axial strength of concrete at the loading age, which can reflect the tensile
creep under different stress level and then be used to evaluate the early-age cracking

@seismicisolation
@seismicisolation
26 1 Introduction

resistance of concrete. Early-age tensile creep of normal concrete increases as the

initial stress/strength ratio increases and corresponding value of initial stress/strength
ratio ranges from 19% to 45% [223], 44% to 60% [224], 54% to 80% [225], and 30%
to 50% [216], respectively. Results reported in [224] show that the specific tensile
creep and tensile creep coefficient in ultra-high performance concrete at 14 days of
loading increase by 11% and 44% when the initial stress/strength ratio increases from
40% to 60%, respectively. The initial stress/strength ratio of 60% [139] and ranging
from 25% to 35% [222] are used to investigate the tensile creep of ICC with SAPs,
respectively. The nonlinear tensile creep occurs when the initial stress/strength ratio
exceeds a certain range. The nonlinear tensile creep of ordinary concrete at early
age occurs when the initial stress/strength ratio is larger than 50% [226], 60% [227],
35% [215], and 30% [207]. Investigations on the nonlinear tensile creep of ICC with
SAPs remain lacking. Thus, studies on the tensile creep of ICC with SAPs under
different stress levels, and the nonlinear properties of tensile creep are necessary to
predict the tensile creep and cracking resistance of concrete at early age.

4. Tensile creep of early-age ICC with SAPs under different loading ages

Tensile creep is influenced significantly by loading age, particularly at early ages,

which is necessary to evaluate the cracking resistance of concrete at early age.
Research studies on the tensile and compressive creep of ordinary concrete loaded
at varying ages have been conducted [217, 228–232], and the initial loading ages are
1 d [230], 2 d [207], 3 d [233, 234], 4 d [229], and 7 d [235]. Investigation on the
tensile creep of concrete loaded between the setting time and 1 d remain lacking.
The tensile creep of concrete with SAPs loaded at 28 d is studied in [139]. The basic
tensile creep of ICC with SAPs loaded at 1, 6, and 14 d is studied in [222], and test
results show that concrete with SAPs exhibits higher basic tensile creep than the
reference concrete without SAPs.
The tensile creep of early-age HSC with the addition of fly ash or GGBFS exhibits
a nonlinear creep characteristic before 5 d or 2 d [230], respectively. Great tensile
creep occur when the loading age is before or at 1 d, and the relationship between
initial stress/strength ratio and tensile creep is nonlinear when loading occurs at 1 d
[223]. The specific tensile creep of concrete is roughly identical at the loading age
of 0.67 d and 1 d and decreased with the increasing loading age ranging from 2 to
7 d [236], respectively. The early-age specific tensile creep of concrete loaded at
the age of 7 d is significantly lower than that loaded at 1 d [219]. The initial basic
tensile creep rate is large during the first 10–20 h after loading, and the tensile creep
becomes age-independent after a few days [214]. The basic tensile creep of ICC
with SAPs loaded at 1 d is significantly larger than that loaded at 6 d and 14 d [222].
Accordingly, investigations on the tensile creep under different loading ages and
loaded before 1 d are necessary to evaluate the tensile creep and cracking resistance
of ICC with SAPs at early age.

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 27

1.2.7 Cracking Resistance by Restrained Ring Test

The effects of shrinkage and continuous restrained condition on cracking resistance

of ICC are considered in this section. The review of restrained ring test is provided.
The significance of investigations on cracking resistance of ICC with SAPs and
LWAs by restrained ring test condition is highlighted. The effects of GGBFS and PP
fibers on cracking resistance of concrete by restrained ring test are reviewed.
1. Testing method of restrained ring test for internally cured concrete
Restrained ring test is a prevalent experimental method for investigating restrained
shrinkage and cracking behavior of cementitious materials in consideration of its
simplicity and versatility [36, 237]. In 1942, Carlson [238] firstly proposed the
restrained ring test to investigate the cracking behavior of concrete, as illustrated
in Fig. 1.8. In 1967, Malhotra and Zoldners [239] demonstrated that restrained ring
test can be conducted to evaluate the tensile strength of concrete. In 1979, Swamy
and Stavrides [240] investigated the fiber reinforced concrete with restrained ring test
by monitoring the strain in steel ring. In 1989, Grzybowski and Shah [241] employed
the restrained ring test to investigate strain development in plain and fiber reinforced
concrete using strain gages on the concrete surface for the calibration of a modeling
approach.
In 2005 and 2018, the standards of the restrained ring test including AASHTO
[242] and ASTM C1581 [243] have been established respectively. In 2005, The
method of ring test is also concluded in Chinese Standard CCES01, which provides
the geometrics dimensions and testing procedure in details [244]. With shrinkage of
concrete, volume change of concrete will be hindered by the restraint deposed by
steel ring, and circumferential tensile stress develops at the same time. Restrained
concrete ring is prone to trigger harmful cracking as tensile strength is insufficient
for existent tensile stress. The restrained ring specimen proposed in ASTM C1581
[243] is illustrated in Fig. 1.9 and geometric dimension is shown in Table 1.3.

Fig. 1.8 A view of the ring

testing with an automatic
hydraulic punping unit [238]

@seismicisolation
@seismicisolation
28 1 Introduction

Outer Ring

Concrete Ring
Specimen
Seal

Bolt with
eccentric
washers D

E E
A
Nonabsorptive
Steel Ring B base
C

Nonabsorptive
Base
Plan View Section E - E

Fig. 1.9 Restrained ring specimen in ASTM C1581 [243]

Table 1.3 Geometric

Figure dimensions SI units (mm)
dimensions of restrained ring
specimen proposed in ASTM A 12.5 ± 0.13
C1581 [243] B 330 ± 3
C 406 ± 3
D 150 ± 6

In 2004, Hossain and Weiss [245] firstly described how the ring test might be
used to provide quantitative information about stress development that may be used
to assess the potential for cracking in concrete. An analytical stress formulation
is presented to compute the actual residual stress level in the concrete using only
the measured strain from the steel ring, as illustrated in Fig. 1.10. In 2013, Gao
et al. [246] employed the ring test to determine the stress relaxation parameters of
concrete in tension at early age. Continuously monitoring the strain that develops in
the steel ring from the time of casting enables the effects of autogenous shrinkage
to be determined as well as the effects of drying shrinkage. In 2016, Shen et al. [41]
utilized the restrained ring test to investigate the effect of SAPs on cracking resistance
of concrete. Recently, in 2020, Dopko et al. [247] utilized restrained ring test to
quantify the cracking resistance of concrete mixtures with carbon microfibers. In
2021, Kanavaris et al. [248] proposed a thermo-chemo-hygro-mechanical simulation
approach for modelling the concrete behavior during restrained shrinkage ring tests
under two distinct drying conditions. The restrained ring specimen generally utilized
is illustrated in Fig. 1.11.
Some other ring specimens are also utilized for investigation on cracking resis-
tance of cementitious materials. In 1993, Kovler et al. [249] utilized a new core steel
ring with a higher thermal expansion coefficient to investigate the cracking behav-
iors. The investigation by Kovler et al. [249] firstly applied active restraint until a

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 29

σθ Pres.
Steel
ring
Pres.

Concrete
ring

Fig. 1.10 Restrained ring specimen [248]

Fig. 1.11 Restrained ring

specimen [41]

specific time at which a tensile load was applied to the concrete ring by introducing
a temperature rise in the specimen assembly. In Norway, special testing ring was
developed by Dahl [250] to be used for plastic shrinkage cracking evaluation in
1994. In this test, the concrete is firstly restrained between two steel rings, with the
external one having flat ribs which act as stress raiser. This ring is fit with special
attachment to allow drying under a stream of hot air, simulating dry wind. In 2003,
the concept of dual restrained ring test gradually takes shape in the investigation by
Bentur and Kovler [251], as illustrated in Fig. 1.12. In 2003, elliptical ring specimens
was developed by He et al. [252] for accelerating crack propagation of cement-based
materials and automatically recording the starting point of a crack, as illustrated in
Fig. 1.13.
Although the restrained ring test has been widely employed for ordinary concrete,
cracking resistance of ICC should be further studied by restrained ring test for better
understanding.

2. Cracking resistance of internally cured concrete by restrained ring test

SAPs can effectively reduce shrinkage [253], cracking potential [112, 254, 255], and
change the rheology of the fresh material [256, 257]. The effects of IC on concrete
shrinkage can help in the effort to extend the life of concrete structures as well as
reducing costly repairs or replacements [96]. In 2020, Lefever et al. [258] investigated

@seismicisolation
@seismicisolation
30 1 Introduction

Thermal
boundaries
Inner Invar ring
Strain gages

Outer Invar rings Concrete ring

Fig. 1.12 Dual ring specimen [251]

Restraint

Elliptical concrete ring

Threaded bolt

Fig. 1.13 Elliptical ring specimen [252]

the effect of SAPs on the restrained shrinkage of mortar by means of the restrained
shrinkage ring test. However, few investigations focus on the effect of SAPs on ICC
by restrained ring test. Therefore, investigations on the cracking resistance of ICC
with SAPs through restrained ring test were necessary for better understanding.
The concrete cured with the external water demonstrates the high cracking poten-
tial. The reason for this phenomenon is that the concrete with low w/c ratio shows
the low level of penetration, which results in the unsatisfactory curing effect [259].
Therefore, IC can be utilized to improve the curing effect of concrete comprehen-
sively at early age [63, 95]. IC is an effective method owing to its unique advantages
that water preserved in the absorptive material can be released and further promote
cement hydration. The IC water can be supplied continuously by the absorptive mate-
rial inside the concrete for cement hydration, and the external curing water penetrates
into the concrete only several millimeters [259]. IC has the good effect on reducing
or even eliminating shrinkage at early age [260, 261]. LWAs are extensively utilized
in IC [262]. An experimental study utilizing the dual ring test shows that the mortar
containing pre-wetted LWAs has the later time to cracking compared with that of
mortar without pre-wetted LWAs [259]. The concrete utilizing IC with pre-wetted
LWAs demonstrates the decrease in stress rate and shrinkage, and the time to cracking
will be accordingly postponed [263]. Some investigations focus on the shrinkage of

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 31

ICC with LWAs [264–266]. However, the cracking resistance of ICC with LWAs by
restrained ring test needs to be further studied.
3. Influential factors of cracking resistance of internally cured concrete by restrained
ring test

(1) Addition of GGBFS

Nowadays, GGBFS has been used widely in the mixture of concrete as a kind of
mineral admixture. The addition of GGBFS can effectively decrease the cement
hydration and improve the properties of concrete, such as workability, long-term
strength, and durability [267]. Using GGBFS as a partial replacement takes advantage
of the energy saving in Portland Cement [268]. The cracking potential of concrete
decreases with the increase of GGBFS content [269]. The addition of GGBFS in
HPC reduces the total shrinkage, porosity, and mobility of chloride ions [270]. The
rate of initial hydration of GGBFS when mixed with water is slower than that of
Ordinary Portland Cement [271]. The study in [272] shows that a partial replacement
of Portland Cement with GGBFS decreases the free drying shrinkage compared with
that of mixtures with 100% Portland Cement. Approximately 500,000 t of CO2 will
be reduced when GGBFS replaces 50% of cement of each ton. However, few studies
investigate the effect of GGBFS on cracking resistance of ICC. The investigation on
effect of GGBFS on cracking resistance of ICC is significant for the application in
practice. Therefore, the early-age cracking resistance of ICC with GGBFS needs to
be further investigated by restrained ring test.

(2) Addition of PP fiber

The cracking potential can be decreased mainly by restricting the shrinkage and
increasing tensile strength of concrete. Fiber reinforced concrete is widely utilized
owing to its high tensile strength and excellent toughness. Adding PP fibers like
Barchip fibers in concrete is another effective method to decrease the cracking poten-
tial at early age [273]. Adding fibers with good alkali resistance, high tensile strength,
and high ultimate elongation can make up for these shortcomings and improve the
performance of concrete. While concrete can be considered as a relatively fragile
material, the toughness of concrete reinforced with randomly distributed short fibers
preventing or controlling the initiation, expansion or coalescence of cracks can be
improved [274]. In addition, fibers in concrete can effectively decrease the shrinkage
of concrete. When the tensile strength of concrete is insufficient, the concrete in the
cracking area will withdraw from work. At the same time, fibers perpendicular to
the crack hindering the further development of the crack begin to manifest tensile
strength. However, although the ring test has been implemented to assess the cracking
potential of ICC [150, 221], more investigations on the cracking potential of fiber
reinforced concrete with IC through the ring test need to be further studied.

@seismicisolation
@seismicisolation
32 1 Introduction

1.2.8 Cracking Resistance Under Uniaxial Restrained

Condition

TSTM can test the temperature process, autogenous shrinkage, restrained stress,
tensile creep, and cracking resistance of concrete at early age simultaneously. The
influence of SAPs, LWAs, GGBFS, w/c ratio, and PP fiber length on the early-age
cracking resistance of ICC under uniaxial restrained condition was revealed.

1. Testing method of TSTM for internally cured concrete

The cracking frame with solid crossheads (rigid cracking frame made of steel) was
developed by Springenschmid [275] in 1973. This frame can be used in tests on
concrete specimens with a cross section of 150 × 150 mm2 , as shown in Fig. 1.14.
Between the two crossheads the length of the specimen must be kept constant even if
the ambient temperature changes. In order to achieve this, the crossheads are screwed
to the bars in a definite distance of the bar ends. The reaction force in the massive steel
bars causes a small elastic deformation of the bars which is recorded continuously
by the strain gauges. By this also the stress-development in the specimen is obtained
continuously. Due to the small elastic deformation of the steel bars the degree of
restraint of the concrete specimen lowers from 100% for the fresh mix to about
80% for the hardened concrete. The formwork can be connected with an external
heating/cooling system. Thus cracking frame can be used to measure thermal stresses
for an arbitrary course of concrete temperature.
In order to solve the problem that the restraint degree of cracking frame is not
adjustable, in 1976, Paillère and Serrano [277] from France developed a kind of
uniaxial restraint test device by controlling the movement of adjustable beams to
ensure that the length of the specimen was unchanged, and reached 100% constraint

Fig. 1.14 Principle of a rigid cracking frame [276]

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 33

Fig. 1.15 Principle sketch of the first TSTM [276]

degree. In this device, the deformation of concrete specimens with 175 mm × 175 mm
load surface can be measured from the early age.
In 1985, the TSTM was developed by Springenschmid to investigate the restrained
stress caused by the hydration heat [276]. As shown in Fig. 1.15 As soon as the
deformation of restrained specimen reaches the threshold value, an additional force
provided by the motor will be applied on the specimen to return the specimen to the
original position, then 100% restraint is fulfilled as the total deformation of restrained
specimen is zero.
In 1994, a closed loop restraining system of TSTM with two specimens (one free
specimen and one restrained specimen) was developed by Kovler [278], and the appli-
cation of automatic controlling system has greatly improved the accuracy of TSTM.
In 2001, Altoubat and Lange [214] from the United States used the similar test device
to investigate the tensile basic creep of early-age concrete. In 2001, Sule et al. [279]
from Delft University studied the early-age cracking of reinforced concrete members
using a similar self-developed test device. In 2002, Qin et al. from Tsinghua Univer-
sity developed the TSTM based on the equipment principle of Springenschmid. In
2015, Shen et al. [98] from Hohai University used the TSTM to evaluate the early-age
cracking resistance of ICC.
Compared with the traditional test method, TSTM provides different degrees of
restraint and an adiabatic condition to concrete specimens for better simulating the
actual working conditions of the mass concrete. Uniaxial constant restraint degree

@seismicisolation
@seismicisolation
34 1 Introduction

can be achieved and the tensile creep can be obtained by free and restrained shrinkage
tests simultaneously using TSTM. Usually, TSTM consists of four main systems:
the temperature controlling system, the displacement measuring system, the load
controlling system and the displacement controlling system, as shown in Fig. 1.16.
Different restraint degrees and curing modes can be provided by TSTM, besides,
deformation, restrained stress, and temperature history of concrete can be measured
by TSTM. Shen et al. [9, 30, 42, 62, 139, 280–282] from Hohai University have
investigated the influences of fibers, supplementary cementitious materials, w/b ratio,
and curing temperature on tensile creep and cracking resistance of concrete at early
age using the TSTM. TSTM has been widely used to investigate the behavior of
concrete. For example, Darquennes et al. [283] used TSTM to investigate the behavior
of slag cement concrete under restraint conditions.

2. Internally cured concrete with SAPs

The self-desiccation of HPC at early age will be reduced by the addition of SAPs.
The majorities of available studies concerning the early-age cracking resistance of
concrete at early age internally cured with SAPs do not simultaneously consider
temperature history, autogenous shrinkage, restrained stress development, and tensile
creep under uniaxial restrained condition [41, 119]. The TSTM can be used to eval-
uate the early-age behavior of concrete simultaneously considering these factors [30,
42, 280, 284]. The influence of all relevant parameters must be studied and quanti-
fied to investigate the early-age cracking resistance of ICC with SAPs under uniaxial
restrained condition more accurately. Therefore, investigations on SAPs influence
the cracking resistance of HPC at early age under uniaxial restrained condition are
of great significance.

3. Internally cured concrete with LWAs

ICC with a larger amount of LWAs shows a greater reduction in the rate and volume of
autogenous shrinkage [96], and the use of LWAs with reduced stiffness can improve

Computer

Free end
Shrinkage
Amplifier Free shrinkage specimen

Displacement sensor Fixed end

Gripped end
Controller Shaft

Load cell
Stepper motor
Restrained shrinkage specimen
Universal joint Pulling back

Fig. 1.16 Schematic description of the closed loop instrumented restraining system [278]

@seismicisolation
@seismicisolation
1.2 State-of-the-Art of Internal Curing 35

the shrinkage cracking resistance of concrete [263]. As a result, the cracking resis-
tance is improved in ICC with pre-wetted LWAs. However, the majority of available
studies concerning the cracking resistance of ICC with pre-wetted LWAs focuses on
the mechanical properties, autogenous shrinkage [95, 285], and the failed to evaluate
the early-age cracking behavior of ICC considering different factors simultaneously.
Thus, investigations on whether and how pre-wetted LWAs influence the cracking
resistance of ICC under uniaxial restrained condition are meaningful.

4. Influential factors of cracking resistance of internally cured concrete under

uniaxial restrained condition

(1) Addition of GGBFS

GGBFS is a kind of mineral admixture being widely used in the mixing of ICC. The
use of GGBFS can improve the compressive strength, pore structure, and perme-
ability of the mortars and concretes with time, because the total porosity decrease
with increasing hydration time [286, 287]. Studies on the influence of GGBFS on the
thermal stress, compressive strength, splitting tensile strength, shrinkage, and creep
have been conducted [286, 287]. Investigations on the restrained stress, autogenous
shrinkage, and cracking resistance of ICC with different internal curing materials [14]
or concrete with different mineral admixtures [288] have been conducted. However,
the majority of available studies failed to reveal the influence of GGBFS on early-age
cracking behavior of ICC considering different factors simultaneously. Therefore,
the influence of GGBFS on the cracking resistance of ICC under uniaxial restrained
condition by using TSTM needs to be investigated.

(2) Different w/c ratios

The early-age behavior and cracking resistance of concrete are affected by w/c
ratio, the early-age cracking resistance of concrete with different w/c ratios has been
studied [289–292], however, temperature, restrained stress, autogenous shrinkage,
and tensile creep are not considered simultaneously under restrained condition. Thus,
the influence of w/c ratio on strength, restrained stress, autogenous shrinkage, tensile
creep, and cracking resistance of ICC with pre-wetted LWAs under uniaxial restrained
condition should be further investigated using TSTM.

(3) Addition of PP fiber with different lengths

Many ways are used to improve the cracking resistance of concrete, among which,
adding fibers in concrete mixtures can not only improve the toughness of concrete, but
also inhibit the shrinkage of HPC, thus preventing the propagation of microcracking
[183, 293–295]. Many types of fiber such as steel fiber, carbon fiber, glass fiber and
plastic fiber have been widely used in engineering [296]. As a new type of modified
synthetic macro-fiber, PP fiber has been widely applied in practical engineering due to
its superiorities such as light weight and superior fracture performance [297, 298]. PP
fiber can control the generation and development of micro cracks in cement matrix,
and even can be in the critical state of pull-out and pull-off when the concrete cracks,

@seismicisolation
@seismicisolation
36 1 Introduction

which substantially solves the problem of outcrop of coarse fiber. Meanwhile, internal
curing technology has developed to be a method for HPC curing due to its higher
efficiency than traditional external curing methods. The existing investigations on PP
fiber reinforced concrete are mainly about the mechanical properties, permeability,
and crack width [299]. Therefore, the studies of early-age cracking resistance of
different lengths of PP fiber reinforced ICC with SAPs under uniaxial restrained
condition are needed.

1.3 Objectives and Scope

The authors present systematically their experimental and theoretical researches on

the cracking control on early-age concrete through internal curing in recent years
in this monograph. Six chapters were organized in this monograph, and the main
contents are introduced briefly as follows.
Chapter 2 presents investigations on the IRH of early-age ICC. Firstly, the effect of
IC with SAPs on the IRH of early-age concrete under sealed and unsealed conditions
is investigated, and the models for the IRH of early-age concrete internally cured with
SAPs are proposed considering the effect of IC water content or working w/c ratio
and critical time. Secondly, the prediction models for working w/c ratio, the critical
time, and the IRH variations in early-age ICC with pre-wetted LWAs are investigated.
Besides, the IRH of early-age concrete under different curing humidity conditions is
investigated, and the prediction model for IRH in HSC under the combined effects
of moisture diffusion and self-desiccation considering the effect of curing humidity
is proposed.
Chapter 3 introduces the non-contact measurement method for autogenous
shrinkage of concrete and the ultrasonic velocity measurement method adopted.
Whether and how IC with different amounts of SAPs influences the early-age
expansion, autogenous shrinkage development, autogenous shrinkage rate, and IC
efficiency of concrete are explored. Besides, the prediction model for autogenous
shrinkage of concrete is established based on the results of ultrasonic velocity, which
could be applied in actual projects quickly. Furthermore, whether and how the IC
water entrained in SAPs, w/c ratio, and the proportion of LWAs affect the corre-
lation between autogenous shrinkage and IRH is experimentally studied through
an integrated device that measures the autogenous shrinkage and IRH of HSC
simultaneously.
Chapter 4 introduces the test device to investigate the tensile creep of ICC with
SAPs under constant tensile load. The closed loop restraining system consists of two
specimens one free specimen for free shrinkage and one restrained specimen to apply
constant tensile stress, which can directly test the deformation after concrete pouring
without formwork removal, and can test the tensile creep of very early age. Whether
and how the addition of SAPs affects the early-age basic tensile creep, basic tensile
creep coefficient, and specific tensile creep are investigated. Furthermore, the effect of
initial stress/strength ratio and loading age on the tensile creep of concrete internally

@seismicisolation
@seismicisolation
References 37

cured with SAPs are explored. The prediction models for the early age tensile creep
considering the initial stress/strength ratio or loading age are established which could
be applied to assess the cracking potential and stress in ICC with SAPs. Besides, the
nonlinearity of tensile creep under different initial stress/strength ratios or loading
ages is revealed.
Chapter 5 introduces the restrained ring test adopted in this monograph to inves-
tigate the cracking resistance of ICC by restrained ring test. The methodologies
of restrained ring test combined with free shrinkage are described in details. The
investigations on residual stress, stress rate, stress relation and time to cracking are
conducted for evaluating the cracking resistance of concrete internally cured with
LWAs and SAPs, respectively. Furthermore, cracking resistance of ICC with GGBFS
and PP fiber is also revealed at early age, respectively.
Chapter 6 mainly focus on the temperature process, autogenous shrinkage,
restrained stress, tensile creep, and cracking resistance of ICC under uniaxial
restrained condition at early age by utilizing TSTM, respectively. The integrated crite-
rion of cracking resistance is proposed to evaluate the early-age cracking behavior
of ICC. The early-age behavior and cracking resistance of concrete internally cured
with SAPs and LWAs are analyzed, respectively. In addition, the investigations on
the early-age cracking resistance of ICC considering the influence of GGBFS, w/c
ratio, and PP fiber length are conducted, respectively.

References

1. Miao CW, Mu R, Tian Q et al (2002) Effect of sulfate solution on the frost resistance of
concrete with and without steel fiber reinforcement. Cem Concr Res 32(1):31–34
2. Liu JP, Miao CW, Chen CC et al (2013) Effect and mechanism of controlled permeable
formwork on concrete water adsorption. Constr Build Mater 39:129–133
3. Li QB, Ansari F (2000) High-strength concrete in triaxial compression by different sizes of
specimens. Mater J 97(6):684–689
4. Zhang GX, Luo XY, Liu Y et al (2018) Influence of aggregates on shrinkage-induced damage
in concrete. J Mater Civ Eng 30(11):04018281
5. Shi CJ, Wu ZM, Xiao JF et al (2015) A review on ultra high performance concrete: part I.
Raw materials and mixture design. Constr Build Mater 101:741–751
6. Zhang J, Han YD, Gao Y (2014) Effects of water-binder ratio and coarse aggregate content
on interior humidity, autogenous shrinkage, and drying shrinkage of concrete. J Mater Civ
Eng 26(1):184–189
7. Wei Y, Xiang YP, Zhang QQ (2014) Internal curing efficiency of prewetted LWFAs on concrete
humidity and autogenous shrinkage development. J Mater Civ Eng 26(5):947–954
8. Jiang ZW, Sun ZP, Wang PM (2005) Autogenous relative humidity change and autogenous
shrinkage of high-performance cement pastes. Cem Concr Res 35(8):1539–1545
9. Shen DJ, Jiao Y, Gao Y et al (2020) Influence of ground granulated blast furnace slag on
cracking potential of high performance concrete at early age. Constr Build Mater 241:117839
10. Jensen OM, Hansen PF (2001) Autogenous deformation and RH-change in perspective. Cem
Concr Res 31(12):1859–1865
11. Hu SW, Fan B (2020) Crack extension resistance of concrete at low temperatures 72(16):837–
851

@seismicisolation
@seismicisolation
38 1 Introduction

12. Khan I, Castel A, Gilbert RI (2017) Effects of fly ash on early-age properties and cracking of
concrete. ACI Mater J 114(4):673–681
13. Azenha MÂD (2009) Numerical simulation of the structural behaviour of concrete since its
early ages. University of Porto, Portugal
14. Shen DJ, Feng ZZ, Kang JC et al (2020) Effect of Barchip fiber on stress relaxation and
cracking potential of concrete internally cured with super absorbent polymers. Constr Build
Mater 249:118392
15. Shen DJ, Kang JC, Yi XJ et al (2019) Effect of double hooked-end steel fiber on early-age
cracking potential of high strength concrete in restrained ring specimens. Constr Build Mater
223:1095–1105
16. Bentz DP, Lura P, Roberts JW (2005) Mixture proportioning for internal curing. Concr Int
27(2):35–40
17. Zhutovsky S, Kovler K (2017) Influence of water to cement ratio on the efficiency of internal
curing of high-performance concrete. Constr Build Mater 144:311–316
18. Gagg CR (2014) Cement and concrete as an engineering material: an historic appraisal and
case study analysis. Eng Fail Anal 40:114–140
19. Zhang H, Li L, Wang W et al (2019) Effect of temperature rising inhibitor on expansion
behavior of cement paste containing expansive agent. Constr Build Mater 199:234–243
20. Safiuddin M, Kaish A, Woon CO et al (2018) Early-age cracking in concrete: causes,
consequences, remedial measures, and recommendations. Appl Sci 8(10):1730
21. Bamforth PB (2007) Early-age thermal crack control in concrete. CIRIA C660, London
22. Campbell-Allen D (1979) The reduction of cracking in concrete. The University of Sydney,
Sydney
23. Huang HL, Ye G, Qian CX et al (2016) Self-healing in cementitious materials: materials,
methods and service conditions. Mater Des 92:499–511
24. de Borst R, van den Boogaard A (1994) Finite-element modeling of deformation and cracking
in early-age concrete. J Eng Mech 120:2519–2534
25. Ba HJ, Su AS, Gao XJ et al (2008) Cracking tendency of restrained concrete at early ages. J
Wuhan Univ Technol Mater Sci Edn 23(2):263–267
26. Klemczak B, Knoppik-Wróbel A (2011) Early age thermal and shrinkage cracks in concrete
structures–description of the problem. Archit Civ Eng Environ 4(2):35–48
27. Kumar Sanjeev V, Gangarao Hota VS (1998) Fatigue response of concrete decks reinforced
with FRP rebars. J Struct Eng 124(1):11–16
28. Abel J, Hover K (1998) Effect of water/cement ratio on the early age tensile strength of
concrete. Transp Res Rec 1610(1):33–38
29. Harrison T (1981) Early-age thermal crack control in concrete
30. Shen DJ, Wang WT, Liu JW et al (2018) Influence of Barchip fiber on early-age cracking
potential of high performance concrete under restrained condition. Constr Build Mater
187:118–130
31. Zhu H, Hu Y, Ma R et al (2021) Concrete thermal failure criteria, test method, and mechanism:
a review. Constr Build Mater 283:122762
32. Ng PL, Fung WWS, Chen JJ et al (2011) Adiabatic temperature rise of condensed silica fume
(CSF) concrete. Adv Mater Res 261–263:788–795
33. Wang XZ, Shi MS, Wang XG (2021) Application of hydration heat inhibitor in crack control
of mass concrete of tunnel side wall. E3S Web Conf 283:01032
34. Li L, Dao V, Lura P (2021) Autogenous deformation and coefficient of thermal expansion of
early-age concrete: initial outcomes of a study using a newly-developed temperature stress
testing machine. Cement Concr Compos 119:103997
35. Zhao ZF, Wang KJ, Lange DA et al (2019) Creep and thermal cracking of ultra-high volume
fly ash mass concrete at early age. Cement Concr Compos 99:191–202
36. Weiss WJ (1999) Prediction of early-age shrinkage cracking in concrete elements. North-
western University, Ann Arbor
37. Altoubat SA, Lange DA (2001) Creep, shrinkage, and cracking of restrained concrete at early
age. ACI Mater J 98(4):323–331

@seismicisolation
@seismicisolation
References 39

38. Sule M, van Breugel K (2004) The effect of reinforcement on early-age cracking due to
autogenous shrinkage and thermal effects. Cem Concr Compos 26(5):581–587
39. Igarashi S, Bentur A, Kovler K (2000) Autogenous shrinkage and induced restraining stresses
in high-strength concretes. Cem Concr Res 30(11):1701–1707
40. Yang EI, Morita S, Yi ST (2000) Effect of axial restraint on mechanical behavior of high-
strength concrete beams. Struct J 97(5):751–756
41. Shen DJ, Shi HF, Tang XJ et al (2016) Effect of internal curing with super absorbent polymers
on residual stress development and stress relaxation in restrained concrete ring specimens.
Constr Build Mater 120:309–320
42. Shen DJ, Liu C, Wang M et al (2021) Effect of polyvinyl alcohol fiber on the cracking risk of
high strength concrete under uniaxial restrained condition at early age. Constr Build Mater
300:124206
43. Shen DJ, Liu C, Jiang JL et al (2020) Influence of super absorbent polymers on early-age
behavior and tensile creep of internal curing high strength concrete. Constr Build Mater
258:120068
44. Shen DJ, Wang T, Chen Y et al (2015) Effect of internal curing with super absorbent polymers
on the relative humidity of early-age concrete. Constr Build Mater 99:246–253
45. Shen DJ, Wang ML, Chen Y et al (2017) Prediction model for relative humidity of early-
age internally cured concrete with pre-wetted lightweight aggregates. Constr Build Mater
144:717–727
46. Xin JD, Zhang GX, Liu Y et al (2020) Environmental impact and thermal cracking resistance
of low heat cement (LHC) and moderate heat cement (MHC) concrete at early ages. J Build
Eng 32:101668
47. Li M, Xu W, Wang YJ et al (2020) Shrinkage crack inhibiting of cast in situ tunnel concrete
by double regulation on temperature and deformation of concrete at early age. Constr Build
Mater 240:117834
48. Wang L, He TS, Zhou YX et al (2021) The influence of fiber type and length on the cracking
resistance, durability and pore structure of face slab concrete. Constr Build Mater 282:122706
49. Xin JD, Zhang GX, Liu Y et al (2018) Effect of temperature history and restraint degree on
cracking behavior of early-age concrete. Constr Build Mater 192:381–390
50. Shen DJ, Wang ML, Chen Y et al (2017) Prediction of internal relative humidity in concrete
modified with super absorbent polymers at early age. Constr Build Mater 149:543–552
51. Shen DJ, Zhou BZ, Wang ML et al (2019) Predicting relative humidity of early-age concrete
under sealed and unsealed conditions. Mag Concr Res 71(22):1151–1166
52. Van VTA, Rößler C, Bui DD et al (2014) Rice husk ash as both pozzolanic admixture and
internal curing agent in ultra-high performance concrete. Cem Concr Compos 53:270–278
53. Shen DJ, Liu C, Kang JC et al (2022) Early-age autogenous shrinkage and tensile creep
of hooked-end steel fiber reinforced concrete with different thermal treatment temperatures.
Cem Concr Compos 131:104550
54. Fairbairn EM, Azenha M (2019) Thermal cracking of massive concrete structures. State of
Art Report of the RILEM TC
55. Shen DJ, Liu C, Feng ZZ et al (2019) Influence of ground granulated blast furnace slag
on the early-age anti-cracking property of internally cured concrete. Constr Build Mater
223:233–243
56. Yan Y, Ouzia A, Yu C et al (2020) Effect of a novel starch-based temperature rise inhibitor
on cement hydration and microstructure development. Cem Concr Res 129:105961
57. Fernandes F, Manari S, Aguayo M et al (2014) On the feasibility of using phase change
materials (PCMs) to mitigate thermal cracking in cementitious materials. Cem Concr Compos
51:14–26
58. Shen DJ, Liu C, Li CC et al (2019) Influence of Barchip fiber length on early-age behavior
and cracking resistance of concrete internally cured with super absorbent polymers. Constr
Build Mater 214:219–231
59. Shen P, Lu L, He Y et al (2020) Investigation on expansion effect of the expansive agents in
ultra-high performance concrete. Cem Concr Compos 105:103425

@seismicisolation
@seismicisolation
40 1 Introduction

60. Yang G, Wu YK, Li H et al (2021) Effect of shrinkage-reducing polycarboxylate admixture

on cracking behavior of ultra-high strength mortar. Cem Concr Compos 122:104117
61. Zhang P, Zheng Y, Wang K et al (2020) Combined influence of nano-CaCO3 and polyvinyl
alcohol fibers on fresh and mechanical performance of concrete incorporating fly ash. Struct
Concr 21(2):724–734
62. Shen DJ, Kang JC, Jiao Y, Li M, Li CC (2020) Effects of different silica fume dosages
on early-age behavior and cracking resistance of high strength concrete under restrained
condition. Constr Build Mater 263:120218
63. Bentz DP, Snyder KA (1999) Protected paste volume in concrete: extension to internal curing
using saturated lightweight fine aggregate. Cem Concr Res 29(11):1863–1867
64. Mechtcherine V, Reinhardt HW (2012) Application of super absorbent polymers (SAP) in
concrete construction: state-of-the-art report of RILEM TC 225-SAP. Springer, Dordrecht,
the Netherlands, pp 14–28
65. Bentz DP, Weiss WJ (2011) Internal curing: a 2010 state-of-the-art review. US Department
of Commerce, National Institute of Standards and Technology, Gaithersburg
66. Weiss WJ, Montanari L (2017) Guide specification for internally curing concrete. National
Concrete Pavement Technology Center: Iowa State University, Ames, IA, USA, 2017, 11
67. Yang L, Shi CJ, Liu JH et al (2021) Factors affecting the effectiveness of internal curing: a
review. Constr Build Mater 267:121017
68. Jensen OM, Hansen PF (2001) Water-entrained cement-based materials: I. Principles and
theoretical background. Cem Concr Res 31(4):647–654
69. Fang L, Zhao Y, Tan TW (2006) Preparation and water absorbent behavior of superabsorbent
polyaspartic acid resin. J Polym Res 13(2):145–152
70. Zhang YH, Wang LM, Li XH et al (2011) Salt-resistant superabsorbents from inverse-
suspension polymerization of PEG methacrylate, acryamide and partially neutralized acrylic
acid. J Polym Res 18(2):157–161
71. Liu XW, Luan S, Li W (2019) Utilization of waste hemicelluloses lye for superabsorbent
hydrogel synthesis. Int J Biol Macromol 132:954–962
72. Shi WD, Dumont MJ, Ly EB (2014) Synthesis and properties of canola protein-based
superabsorbent hydrogels. Eur Polymer J 54:172–180
73. Liu ZS, Rempel GL (1997) Preparation of superabsorbent polymers by crosslinking acrylic
acid and acrylamide copolymers. J Appl Polym Sci 64(7):1345–1353
74. Chen Y, Zhang Y, Wang FJ et al (2016) Preparation of porous carboxymethyl chitosan grafted
poly (acrylic acid) superabsorbent by solvent precipitation and its application as a hemostatic
wound dressing. Mater Sci Eng C 63:18–29
75. Fang SX, Wang GJ, Xing R et al (2019) Synthesis of superabsorbent polymers based
on chitosan derivative graft acrylic acid-co-acrylamide and its property testing. Int J Biol
Macromol 132:575–584
76. Yang Z, Peng HD, Wang WZ et al (2010) Crystallization behavior of poly (ε-
caprolactone)/layered double hydroxide nanocomposites. J Appl Polym Sci 116(5):2658–
2667
77. Kabir SM, Sikdar PP, Haque B et al (2018) Cellulose-based hydrogel materials: chemistry,
properties and their prospective applications. Prog Biomater 7(3):153–174
78. Lee JW, Kim SY, Kim SS et al (1999) Synthesis and characteristics of interpenetrating polymer
network hydrogel composed of chitosan and poly (acrylic acid). J Appl Polym Sci 73(1):113–
120
79. Yang YF, Xie YL, Pang LC et al (2013) Preparation of reduced graphene oxide/poly
(acrylamide) nanocomposite and its adsorption of Pb (II) and methylene blue. Langmuir
29(34):10727–10736
80. Fanta GF, Burr RC, Russell CR et al (1967) Graft copolymers of starch. III. Copolymerization
of gelatinized wheat starch with acrylonitrile. Influence of chain modifiers on copolymer
composition. J Appl Polym Sci 11(3):457–463
81. Lura P, Ye G, Cnudde V et al (2008) Preliminary results about 3D distribution of superab-
sorbent polymers in mortars. In: Proceedings of the international conference on microstructure
related durability of cementitious composites, Nanjing, China, pp 13–15

@seismicisolation
@seismicisolation
References 41

82. Wyrzykowski M, Lura P, Pesavento F et al (2012) Modeling of water migration during internal
curing with superabsorbent polymers. J Mater Civ Eng 24(8):1006–1016
83. Boshoff W, Mechtcherine V, Snoeck D et al (2020) The effect of superabsorbent polymers
on the mitigation of plastic shrinkage cracking of conventional concrete, results of an inter-
laboratory test by RILEM TC 260-RSC. Mater Struct 53(4):1–16
84. Schröfl C, Erk KA, Siriwatwechakul W et al (2022) Recent progress in superabsorbent
polymers for concrete. Cem Concr Res 151:106648
85. Ma XW, Liu JH, Shi CJ (2019) A review on the use of LWA as an internal curing agent of
high performance cement-based materials. Constr Build Mater 218:385–393
86. Pickel D, Tighe S, West JS (2017) Assessing benefits of pre-soaked recycled concrete
aggregate on variably cured concrete. Constr Build Mater 141:245–252
87. General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China (2010) Lightweight aggregates and its test methods-part 2: test methods for
lightweight aggregates: GB/T 17431.2-2010. Standard press of China, Beijing. (in Chinese)
88. Zou DH, Li K, Li WD et al (2018) Effects of pore structure and water absorption on internal
curing efficiency of porous aggregates. Constr Build Mater 163:949–959
89. Klieger P (1969) Early high-strength concrete for prestressing. Portland Cement Association
90. Bentur A, Igarashi SI, Kovler K (2001) Prevention of autogenous shrinkage in high-strength
concrete by internal curing using wet lightweight aggregates. Cem Concr Res 31(11):1587–
1591
91. Zhutovsky S, Kovler K, Bentur A (2004) Influence of cement paste matrix properties on the
autogenous curing of high-performance concrete. Cem Concr Compos 26(5):499–507
92. Henkensiefken R, Castro J, Bentz D et al (2009) Water absorption in internally cured mortar
made with water-filled lightweight aggregate. Cem Concr Res 39(10):883–892
93. Lura P, Wyrzykowski M, Tang C et al (2014) Internal curing with lightweight aggregate
produced from biomass-derived waste. Cem Concr Res 59:24–33
94. Han YD, Zhang J, Luosun YM et al (2014) Effect of internal curing on internal relative
humidity and shrinkage of high strength concrete slabs. Constr Build Mater 61:41–49
95. Raoufi K, Schlitter J, Bentz D et al (2011) Parametric assessment of stress development and
cracking in internally cured restrained mortars experiencing autogenous deformations and
thermal loading. Adv Civ Eng 2011:1–16
96. Henkensiefken R, Bentz D, Nantung T et al (2009) Volume change and cracking in inter-
nally cured mixtures made with saturated lightweight aggregate under sealed and unsealed
conditions. Cem Concr Compos 31(7):427–437
97. Kong DY, Lei T, Zheng JJ et al (2010) Effect and mechanism of surface-coating pozza-
lanics materials around aggregate on properties and ITZ microstructure of recycled aggregate
concrete. Constr Build Mater 24(5):701–708
98. Shen DJ, Jiang JL, Shen JX et al (2015) Influence of prewetted lightweight aggregates on
the behavior and cracking potential of internally cured concrete at an early age. Constr Build
Mater 99:260–271
99. Lv Y, Ye G, Schutter GD (2019) Utilization of miscanthus combustion ash as internal
curing agent in cement-based materials: Effect on autogenous shrinkage. Constr Build Mater
207:585–591
100. Kawashima S, Shah SP (2011) Early-age autogenous and drying shrinkage behavior of
cellulose fiber-reinforced cementitious materials. Cem Concr Compos 33(2):201–208
101. Kang SH, Hong SG, Moon J (2017) Absorption kinetics of superabsorbent polymers (SAP)
in various cement-based solutions. Cem Concr Res 97:73–83
102. Shen DJ, Wen CY, Zhu PF et al (2020) Influence of Barchip fiber on early-age autogenous
shrinkage of high-strength concrete internally cured with super absorbent polymers. Constr
Build Mater 264:119983
103. Zhang J, Hou DW, Sun W (2010) Experimental study on the relationship between shrinkage
and interior humidity of concrete at early age. Mag Concr Res 62(3):191–199
104. Paul A, Lopez M (2011) Assessing lightweight aggregate efficiency for maximizing internal
curing performance. ACI Mater J 108(4):385–393

@seismicisolation
@seismicisolation
42 1 Introduction

105. de Sensale GR, Goncalves AF (2014) Effects of fine LWA and SAP as internal water curing
agents. Int J Concr Struct Mater 8(3):229–238
106. Venkateswarlu K, Deo SV, Murmu M (2020) Overview of effects of internal curing agents on
low water to binder concretes. Mater Today Proc 32:752–759
107. Schröfl C, Mechtcherine V, Gorges M (2012) Relation between the molecular structure and the
efficiency of superabsorbent polymers (SAP) as concrete admixture to mitigate autogenous
shrinkage. Cem Concr Res 42(6):865–873
108. Lu JX, Shen PL, Ali HA et al (2021) Development of high performance lightweight concrete
using ultra high performance cementitious composite and different lightweight aggregates.
Cem Concr Compos 124:104277
109. BS EN 206 (2005) Concrete—part 1: specification, performance, production and conformity.
British Standard Institution, Brussels, Belgium
110. Dang JT, Zhao J, Du ZH (2017) Effect of superabsorbent polymer on the properties of concrete.
Polymers 9(12):672
111. Bentz DP (2009) Influence of internal curing using lightweight aggregates on interfacial
transition zone percolation and chloride ingress in mortars. Cem Concr Compos 31(5):285–
289
112. Snoeck D, Steuperaert S, van Tittelboom K et al (2012) Visualization of water penetration in
cementitious materials with superabsorbent polymers by means of neutron radiography. Cem
Concr Res 42(8):1113–1121
113. Montanari L, Suraneni P, Chang MT et al (2018) Absorption and desorption of superabsorbent
polymers for use in internally cured concrete. Adv Civ Eng Mater 7(4):109–128
114. Montanari L, Suraneni P, Weiss WJ (2017) Accounting for water stored in superabsorbent
polymers in increasing the degree of hydration and reducing the shrinkage of internally cured
cementitious mixtures. Adv Civ Eng Mater 6(1):583–599
115. Mechtcherine V, Snoeck D, Schröfl C et al (2018) Testing superabsorbent polymer (SAP)
sorption properties prior to implementation in concrete: results of a RILEM Round-Robin
Test. Mater Struct 51(1):1–16.
116. Esteves LP (2015) Recommended method for measurement of absorbency of superabsorbent
polymers in cement-based materials. Mater Struct 48(8):2397–2401
117. Jensen OM (2011) Water absorption of superabsorbent polymers in a cementitious environ-
ment. In: International conference on advances in construction materials through science and
engineering. RILEM Publications, Hong Kong, China, pp 22–835
118. He ZM, Shen AQ, Guo YC et al (2019) Cement-based materials modified with superabsorbent
polymers: a review. Constr Build Mater 225:569–590
119. Shen DJ, Wang XD, Cheng DB et al (2016) Effect of internal curing with super absorbent
polymers on autogenous shrinkage of concrete at early age. Constr Build Mater 106:512–522
120. Mönning S (2009) Superabsorbing additions in concrete: applications, modeling and
comparison of different internal water sources. Stuttgart University, Germany
121. Justs J, Wyrzykowski M, Bajare D et al (2015) Internal curing by superabsorbent polymers
in ultra-high performance concrete. Cem Concr Res 76:82–90
122. Assmann A, Reinhardt HW (2014) Tensile creep and shrinkage of SAP modified concrete.
Cem Concr Res 58:179–185
123. Wang FZ, Zhou YF, Peng B et al (2009) Autogenous shrinkage of concrete with super-
absorbent polymer. ACI Mater J 106(2):123–127
124. Kong XM, Zhang ZL, Lu ZC (2014) Effect of pre-soaked superabsorbent polymer on
shrinkage of high-strength concrete. Mater Struct 48(9):2741–2758
125. Kong XM, Zhang ZL (2013) Effect of super-absorbent polymer on pore structure of hardened
cement paste in high-strength concrete. J Chin Ceramic Soc 41(11):1474–1480 (in Chinese)
126. Azarijafari H, Kazemian A, Rahimi M et al (2016) Effects of pre-soaked super absorbent
polymers on fresh and hardened properties of self-consolidating lightweight concrete. Constr
Build Mater 113:215–220
127. ASTM International (2019) Standard practice for making and curing concrete test specimens
in the laboratory: ASTM C192/C192M-19. ASTM International, West Conshohocken

@seismicisolation
@seismicisolation
References 43

128. Castro J, Spragg R, Weiss J (2012) Water absorption and electrical conductivity for internally
cured mortars with a w/c between 0.30 and 0.45. J Mater Civ Eng 24(2):223–231
129. Zhang J, Wang J, Han Y (2015) Simulation of moisture field of concrete with pre-soaked
lightweight aggregate addition. Constr Build Mater 96:599–614
130. Schlitter JL, Bentz DP, Weiss WJ (2013) Quantifying stress development and remaining stress
capacity in restrained, internally cured mortars. ACI Mater J 110(1):3–11
131. Dang YD, Shi XM, Mery S et al (2015) Influence of surface sealers on the properties of
internally cured cement mortars containing saturated fine lightweight aggregate. J Mater Civ
Eng 27(12):04015037
132. Lura P, Jensen OM, van Breugel K (2003) Autogenous shrinkage in high-performance cement
paste: an evaluation of basic mechanisms. Cem Concr Res 33(2):223–232
133. Craeye B, Geirnaert M, Schutter GD (2011) Super absorbing polymers as an internal curing
agent for mitigation of early-age cracking of high-performance concrete bridge decks. Constr
Build Mater 25(1):1–13
134. Kim JK, Lee CS (1999) Moisture diffusion of concrete considering self-desiccation at early
ages. Cem Concr Res 29(12):1921–1927
135. Zhang J, Hou DW, Gao Y et al (2011) Determination of moisture diffusion coefficient of
concrete at early age from interior humidity measurements. Drying Technol 29(6):689–696
136. Wang JH, Zhang J, Ding XP et al (2018) Effect of cementitious permanent formwork
on moisture field of internal-cured concrete under drying. Mech Time-Dependent Mater
22(1):95–127
137. Xi Y, Bažant ZP, Jennings HM (1994) Moisture diffusion in cementitious materials adsorption
isotherms. Adv Cem Based Mater 1(6):248–257
138. Zhang J, Han YD, Gao Y et al (2013) Integrative study on the effect of internal curing on
autogenous and drying shrinkage of high-strength concrete. Drying Technol 31(5):565–575
139. Shen DJ, Jiang JL, Wang WT et al (2017) Tensile creep and cracking resistance of concrete
with different water-to-cement ratios at early age. Constr Build Mater 146:410–418
140. Bissonnette B, Pierre P, Pigeon M (1999) Influence of key parameters on drying shrinkage of
cementitious materials. Cem Concr Res 29(10):1655–1662
141. Baroghel-Bouny V, Mainguy M, Lassabatere T et al (1999) Characterization and identifi-
cation of equilibrium and transfer moisture properties for ordinary and high-performance
cementitious materials. Cem Concr Res 29(8):1225–1238
142. Zhang J, Huang Y, Qi K et al (2012) Interior relative humidity of normal- and high-strength
concrete at early age. J Mater Civ Eng 24(6):615–622
143. Zhang J, Qi K, Huang Y (2009) Calculation of moisture distribution in early-age concrete. J
Eng Mech 135(8):871–880
144. Voigt T, Ye G, Sun ZH et al (2005) Early age microstructure of Portland cement mortar
investigated by ultrasonic shear waves and numerical simulation. Cem Concr Res 35(5):858–
866
145. Zhang J, Hou DW, Han YD (2012) Micromechanical modeling on autogenous and drying
shrinkages of concrete. Constr Build Mater 29:230–240
146. Shen DJ, Liu C, Wang ML et al (2020) Prediction model for internal relative humidity in early-
age concrete under different curing humidity conditions. Constr Build Mater 265:119987
147. Chen Y (2015) Research on development of internal relative humidity of internally cured
concrete at early age. Hohai University, Nanjing (in Chinese)
148. Persson B (1998) Experimental studies on shrinkage of high-performance concrete. Cem
Concr Res 28(7):1023–1036
149. Persson B (1997) Self-desiccation and its importance in concrete technology. Mater Struct
30(5):293–305
150. Persson B (1997) Moisture in concrete subjected to different kinds of curing. Mater Struct
30(203):533–544
151. Cusson D, Hoogeveen T (2008) Internal curing of high-performance concrete with pre-soaked
fine lightweight aggregate for prevention of autogenous shrinkage cracking. Cem Concr Res
38(6):757–765

@seismicisolation
@seismicisolation
44 1 Introduction

152. Wang T (2016) Effect of curing temperatures on relative humidity of internally cured concrete
with ceramsite. Hohai University, Nanjing (in Chinese)
153. Wyrzykowski M, Lura P, Pesavento F et al (2011) Modeling of internal curing in maturing
mortar. Cem Concr Res 41(12):1349–1356
154. Wei Y, Wang YQ, Gao X (2015) Effect of internal curing on moisture gradient distribution and
deformation of a concrete pavement slab containing prewetted lightweight fine aggregates.
Drying Technol 33(3):355–364
155. Šelih J, Bremner TW (1996) Drying of saturated lightweight concrete: an experimental
investigation. Mater Struct 29(7):401–405
156. Parrott LJ (1988) Moisture profiles in drying concrete. Adv Cem Res 1(3):164–170
157. McDonald DB, Roper H, Parrott LJ (1991) Discussion: factors influencing relative humidity
in concrete. Mag Concr Res 43(157):305–307
158. Gawin D, Pesavento F, Schrefler BA (2006) Hygro-thermo-chemo-mechanical modelling of
concrete at early ages and beyond. Part I: hydration and hygro-thermal phenomena. Int J
Numerical Methods Eng 67(3):299–331
159. Wang ML (2018) Effect of ambient condition on internal relative humidity in concrete
internally cured with super absorbent polymers. Hohai University, Nanjing (in Chinese)
160. Ding XP, Zhang J, Wang JH (2019) Integrative modeling on self-desiccation and moisture
diffusion in concrete based on variation of water content. Cem Concr Compos 97:322–340
161. Zhang J, Wang JH, Gao Y (2016) Moisture movement in early-age concrete under cement
hydration and environmental drying. Mag Concr Res 68(8):391–408
162. Du MY, Jin XY, Ye HL et al (2016) A coupled hygro-thermal model of early-age concrete
based on micro-pore structure evolution. Constr Build Mater 111:689–698
163. Zhang J, Wang JH, Ding XP (2018) Test and simulation on moisture flow in earlyage concrete
under drying. Drying Technol 36(2):221–233
164. Ishida T, Maekawa K, Kishi T (2007) Enhanced modeling of moisture equilibrium and trans-
port in cementitious materials under arbitrary temperature and relative humidity history. Cem
Concr Res 37(4):565–578
165. Zhou W, Qi TQ, Liu XH et al (2019) A meso-scale analysis of the hygro-thermo-chemical
characteristics of early-age concrete. Int J Heat Mass Transf 129:690–706
166. Bocciarelli M, Ranzi G (2018) An inverse analysis approach for the identification of the
hygro-thermo-chemical model parameters of concrete. Int J Mech Sci 138:368–382
167. Bocciarelli M, Ranzi G (2018) Identification of the hygro-thermo-chemicalmechanical model
parameters of concrete through inverse analysis. Constr Build Mater 162:202–214
168. Akcay B, Tasdemir MA (2008) Internal curing of mortars by lightweight aggregates and its
effects on hydration. Can J Civ Eng 35(11):1276–1284
169. Sant G, Lura P, Weiss J (2006) Measurement of volume change in cementitious materials at
early ages—review of testing protocols and interpretation of results. Transp Res Rec J Transp
Res Board 1979(1):21–29
170. Lura P, Jensen OM (2006) Measuring techniques for autogenous strain of cement paste. Mater
Struct 40(4):431–440
171. de Meyst L, Mannekens E, van Tittelboom K et al (2021) The influence of superabsorbent
polymers (SAPs) on autogenous shrinkage in cement paste, mortar and concrete. Constr Build
Mater 286:122948
172. Chen WW, Li B, Wang J et al (2021) Effects of alkali dosage and silicate modulus on
autogenous shrinkage of alkali-activated slag cement paste. Cem Concr Res 141:106322
173. Abate SY, Park S, Kim HK (2020) Parametric modeling of autogenous shrinkage of sodium
silicate-activated slag. Constr Build Mater 262:120747
174. Jiang CH, Yang Y, Wang Y et al (2014) Autogenous shrinkage of high performance concrete
containing mineral admixtures under different curing temperatures. Constr Build Mater
61:260–269
175. Ren GS, Yao B, Ren M et al (2022) Utilization of natural sisal fibers to manufacture
eco-friendly ultra-high performance concrete with low autogenous shrinkage. J Clean Prod
332:130105

@seismicisolation
@seismicisolation
References 45

176. Jasiczak J, Szymański P, Nowotarski P (2015) Computerised evaluation of the early age of
shrinkage in concrete. Autom Constr 49:40–50
177. Zhutovsky S, Kovler K, Bentur A (2002) Efficiency of lightweight aggregates for internal
curing of high strength concrete to eliminate autogenous shrinkage. Mater Struct 35(246):97–
101
178. Pourjavadi A, Fakoorpoor SM, Khaloo A et al (2012) Improving the performance of cement-
based composites containing superabsorbent polymers by utilization of nano-SiO2 particles.
Mater Des 42:94–101
179. Akcay B, Tasdemir MA (2010) Effects of distribution of lightweight aggregates on internal
curing of concrete. Cem Concr Compos 32(8):611–616
180. Shen DJ, Wen CY, Zhu PF et al (2020) Influence of Barchip fiber on early-age autogenous
shrinkage of high strength concrete. Constr Build Mater 256:119223
181. Shen DJ, Wen CY, Kang JC et al (2020) Early-age stress relaxation and cracking potential of
High-strength concrete reinforced with Barchip fiber. Constr Build Mater 258:119538
182. Snoeck D, Pel L, Belie ND (2019) Comparison of different techniques to study the nanos-
tructure and the microstructure of cementitious materials with and without superabsorbent
polymers. Constr Build Mater 223:244–253
183. Dawood ET, Ramli M (2010) Development of high strength flowable mortar with hybrid fiber.
Constr Build Mater 24(6):1043–1050
184. Sahmaran M, Yurtseven A, Yaman IO (2005) Workability of hybrid fiber reinforced self-
compacting concrete. Build Environ 40(12):1672–1677
185. Kaufmann J, Winnefeld F, Hesselbarth D (2004) Effect of the addition of ultrafine cement
and short fiber reinforcement on shrinkage, rheological and mechanical properties of Portland
cement pastes. Cem Concr Compos 26(5):541–549
186. Huang KJ, Deng M, Mo LW et al (2013) Early age stability of concrete pavement by using
hybrid fiber together with MgO expansion agent in high altitude locality. Constr Build Mater
48:685–690
187. Miao B, Chern JC, Yang CA (2003) Influences of fiber content on properties of self-
compacting steel fiber reinforced concrete. J Chin Inst Eng 26(4):523–530
188. Wang X, Wang YB, Yang LS et al (2013) High-performance high-volume fly ash concrete.
Bull Chin Ceramic Soc 32(3):523–522
189. Yoo SW, Kwon S-J, Jung SH (2012) Analysis technique for autogenous shrinkage in high
performance concrete with mineral and chemical admixtures. Constr Build Mater 34:1–10
190. CEN (2004) Eurocode 2: design of concrete structures: part 1-1: general rules and rules for
buildings: BS EN 1992-1-1. CEN, Brussels
191. fib (2013) fib model code for concrete structures 2010: fib model code 2010. Ernst & Sohn,
Berlin
192. ACI Committee 209 (1982) Prediction of creep, shrinkage, and temperature effects in concrete
structures. American Concrete Institute
193. Tazawa E, Miyazawa S (2001) Prediction model for shrinkage of concrete including autoge-
nous shrinkage. Creep, shrinkage and durability mechanics of concrete and other quasi-brittle
materials. In: Proceedings of sixth international conference. Elsevier Science Ltd., pp 735–746
194. Hashida H, Yamazaki N (2002) Deformation composed of autogenous shrinkage and thermal
expansion due to hydration of high-strength concrete and stress in reinforced structures
13(1):25–32
195. Hua C, Acker P, Ehrlacher A (1995) Analyses and models of the autogenous shrinkage of
hardening cement paste: I. Modelling at macroscopic scale. Cem Concr Res 25(7):1457–1468
196. Bažant ZP, Baweja S (1995) Creep and shrinkage prediction model for analysis and design
of concrete structures: model B3. Mater Struct 28(6):357–365
197. Bažant ZP, Jirásek M, Hubler M et al (2015) RILEM draft recommendation: TC-242-
MDC multi-decade creep and shrinkage of concrete: material model and structural analysis
Model B4 for creep, drying shrinkage and autogenous shrinkage of normal and high-strength
concretes with multi-decade applicability. Mater Struct 48(4):753–770

@seismicisolation
@seismicisolation
46 1 Introduction

198. Wendner R, Hubler MH, Bažant ZP (2015) Statistical justification of model B4 for multi-
decade concrete creep using laboratory and bridge databases and comparisons to other models.
Mater Struct 48(4):815–833
199. Li Y, Bao JL, Guo YL (2010) The relationship between autogenous shrinkage and pore
structure of cement paste with mineral admixtures. Constr Build Mater 24(10):1855–1860
200. Wendner R, Hubler MH, Bažant ZP (2013) The B4 model for multi-decade creep and shrinkage
prediction. In: International conference on creep, pp 429–436
201. Do QH (2013) Modelling properties of cement paste from microstructure: porosity, mechan-
ical properties, creep and shrinkage. École Polytechnique Fédérale De Lausanne, Switzerland
202. Chen H, Wyrzykowsk M, Scrivener K et al (2013) Prediction of self-desiccation in low
water-to-cement ratio pastes based on pore structure evolution. Cem Concr Res 49:38–47
203. Beltzung F, Wittmann FH (2005) Role of disjoining pressure in cement based materials. Cem
Concr Res 35(12):2364–2370
204. Rahimi-Aghdam S, Masoero E, Rasoolinejad M et al (2019) Century-long expansion of
hydrating cement counteracting concrete shrinkage due to humidity drop from selfdesiccation
or external drying. Mater Struct 52(1):11.1–11.21
205. Shen DJ, Li CC, Li M et al (2022) Kang, experimental investigation on correlation between
autogenous shrinkage and internal relative humidity of superabsorbent polymer–modified
concrete. J Mater Civ Eng 34(2):04021456
206. Nguyen DH, Dao V, Lura P (2017) Tensile properties of concrete at very early ages. Constr
Build Mater 134:563–573
207. Briffaut M, Benboudjema F, Torrenti J-M et al (2012) Concrete early age basic creep:
experiments and test of rheological modelling approaches. Constr Build Mater 36:373–380
208. Zhu H, Li QB, Hu Y et al (2018) Double feedback control method for determining early-age
restrained creep of concrete using a temperature stress testing machine. Materials 11(7):1079
209. Cook DJ, Ward MA (1969) The mechanism of tensile creep in concrete. Mag Concr Res
21(68):151–158
210. Domone LP (1974) Uniaxial tensile creep and failure of concrete. Mag Concr Res 26(88):144–
152
211. Brooks JJ, Neville AM (1977) A comparison of creep, elasticity and strength of concrete in
tension and in compression. Mag Concr Res 29(100):131–141
212. Reinhardt HW, Comelissen HAW (1985) Tensile tests under sustained load on concrete.
Baustoffpraxis 85:162–167
213. Aly T, Sanjayan JG (2008) Shrinkage cracking properties of slag concretes with one-day
curing. Mag Concr Res 60(1):41–48
214. Altoubat SA, Lange DA (2001) Tensile basic creep: measurements and behavior at early age.
ACI Mater J 98(5):386–393
215. Kamen A, Denarié E, Sadouki H et al (2008) UHPFRC tensile creep at early age. Mater Struct
42(1):113–122
216. Ranaivomanana N, Multon S, Turatsinze A (2013) Tensile, compressive and flexural basic
creep of concrete at different stress levels. Cem Concr Res 52:1–10
217. Al-Manaseer A, Zayed R (2018) Tensile creep of concrete at early age. ACI Mater J
115(5):769–772
218. Forth JP (2015) Predicting the tensile creep of concrete. Cem Concr Compos 55:70–80
219. Pigeon BM (1995) Tensile creep at early ages of ordinary, silica fume and fiber reinforced
concretes. Cem Concr Res 25(5):1075–1085
220. Yang Y, Xu SF, Ye DY et al (2009) Tensile creep behavior of high strength concrete at early
ages. J Chin Ceramic Soc 37(7):1124–1129 (in Chinese)
221. Shen DJ, Jiang JL, Zhang MY et al (2018) Tensile creep and cracking potential of high
performance concrete internally cured with super absorbent polymers at early age. Constr
Build Mater 165:451–461
222. Li L, Dabarera AGP, Dao V (2022) Basic tensile creep of concrete with and without
superabsorbent polymers at early ages. Constr Build Mater 320:126180

@seismicisolation
@seismicisolation
References 47

223. Østergaard L, Lange DA, Altoubat SA et al (2001) Tensile basic creep of early-age concrete
under constant load. Cem Concr Res 31(12):1895–1899
224. Garas VY, Kahn LF, Kurtis KE (2009) Short-term tensile creep and shrinkage of ultra-high
performance concrete. Cem Concr Compos 31(3):147–152
225. Rossi P, Tailhan J-L, Le Maou F et al (2012) Basic creep behavior of concretes investigation
of the physical mechanisms by using acoustic emission. Cem Concr Res 42(1):61–73
226. Pigeon M, Bissonnette B (1999) Tensile creep and cracking potential. Concr Int 21(11):31–35
227. Atrushi DS (2003) Tensile and compressive creep of young concrete: testing and modelling.
Fakultet for ingeniørvitenskap og teknologi
228. Xu Y, Liu J, Liu J et al (2019) Creep at early ages of ultrahigh-strength concrete: experiment
and modelling. Mag Concr Res 71(15–16):847–859
229. Klausen AE, Kanstad T, Bjøntegaard Ø et al (2017) Comparison of tensile and compressive
creep of fly ash concretes in the hardening phase. Cem Concr Res 95:188–194
230. Ni T, Yang Y, Gu C et al (2019) Early-age tensile basic creep behavioral characteristics of
high-strength concrete containing admixtures. Adv Civ Eng 2019:1–11
231. Jiao YY, Han B, Xie HB et al (2020) Early-age creep behavior of concrete-filled steel tubular
members subjected to axial compression. J Constr Steel Res 166:105939
232. Delsaute B, Boulay C, Staquet S (2016) Creep testing of concrete since setting time by means
of permanent and repeated minute-long loadings. Cem Concr Compos 73:75–88
233. Pane I, Hansen W (2008) Investigation on key properties controlling early-age stress
development of blended cement concrete. Cem Concr Res 38(11):1325–1335
234. Pane I, Hansen W (2002) Early age creep and stress relaxation of concrete containing blended
cements. Mater Struct 35(2):92–96
235. Maou T (2013) Comparison of concrete creep in tension and in compression: influence of
concrete age at loading and drying conditions. Cem Concr Res 51:78–84
236. Gutsch AW (2002) Properties of early age concrete-experiments and modelling. Mater Struct
35(2):76–79
237. Shen DJ, Liu KQ, Ji Y et al (2018) Early-age residual stress and stress relaxation of high-
performance concrete containing fly ash. Mag Concr Res 70(13–14):726–738
238. Carlson RW (1942) Cracking of concrete. Civ Eng 29(2):98–109
239. Malhotra VM, Zoldners NG (1967) Comparison of ring-tensile strength of concrete with
compressive, flexural, and splitting-tensile strengths. J Mater 2(1):160–199
240. Swamy RN, Stavrides H (1979) Influence of fiber reinforcement on restrained shrinkage and
cracking. J Proc 76(3):443–460
241. Grzybowski M, Shah SP (1989) Model to predict cracking in fibre reinforced concrete due to
restrained shrinkage. Mag Concr Res 41(148):125–135
242. AASHTO (1996) Standard practice for estimating the cracking tendency of concrete.
AASHTO Designation, Washington
243. ASTM International (2019) Standard test method for autogenous strain of cement paste and
mortar: ASTM C1581/C1581M-18a[S]. ASTM International, West Conshohocken
244. China Civil Engineering Society (2005) Guide to durability design and construction of
concrete structures: CCES 01–2004. China Architecture & Building Press, Beijing (in
Chinese)
245. Hossain AB, Weiss J (2004) Assessing residual stress development and stress relaxation in
restrained concrete ring specimens. Cem Concr Compos 26(5):531–540
246. Gao Y, Zhang J, Han P (2013) Determination of stress relaxation parameters of concrete in
tension at early-age by ring test. Constr Build Mater 41:152–164
247. Dopko M, Najimi M, Shafei B et al (2020) Strength and crack resistance of carbon microfiber
reinforced concrete. ACI Mater J 117(2):11–23
248. Kanavaris F, Ferreira C, Sousa C et al (2021) Thermo-chemo-hygro-mechanical simulation of
the restrained shrinkage ring test for cement-based materials under distinct drying conditions.
Constr Build Mater 294
249. Kovler K, Sikuler J, Bentur A (1993) Restrained shrinkage tests of fibre-reinforced concrete
ring specimens: effect of core thermal expansion. Mater Struct 26(4):231–237

@seismicisolation
@seismicisolation
48 1 Introduction

250. Sellevold EJ, Bjøntegaard Ø, Justnes H et al (1994) High-performance concrete: early volume
change and cracking tendency. In: Proceeding of thermal cracking at early ages. E. & F.N.
Spon, London
251. Bentur A, Kovler K (2003) Evaluation of early age cracking characteristics in cementitious
systems. Mater Struct 36(3):183–190
252. He Z, Zhou X, Li Z (2004) New experimental method for studying early-age cracking of
cement-based materials. ACI Mater J 101(1):50–56
253. Jensen OM, Hansen PF (2002) Water-entrained cement-based materials: II. Experimental
observations. Cem Concr Res 32(6):973–978
254. Lee HXD, Wong HS, Buenfeld NR (2010) Potential of superabsorbent polymer for self-sealing
cracks in concrete. Adv Appl Ceram 109(5):296–302
255. Chen DP, Liu CL, Qian CX (2011) Study on shrinkage and cracking performance of SAP-
modified concrete. Materials science forum, vol 675. Trans Tech Publications Ltd., pp 697–700
256. Pourjavadi A, Fakoorpoor SM, Hosseini P et al (2013) Interactions between superabsorbent
polymers and cement-based composites incorporating colloidal silica nanoparticles. Cem
Concr Compos 37:196–204
257. Mechtcherine V, Secrieru E, Schröfl C (2015) Effect of superabsorbent polymers (SAPs) on
rheological properties of fresh cement-based mortars—development of yield stress and plastic
viscosity over time. Cem Concr Res 67:52–65
258. Lefever G, Tsangouri E, Snoeck D et al (2020) Combined use of superabsorbent polymers and
nanosilica for reduction of restrained shrinkage and strength compensation in cementitious
mortars. Constr Build Mater 251:118966
259. de la Varga I, Castro J, Bentz D et al (2012) Application of internal curing for mixtures
containing high volumes of fly ash. Cem Concr Compos 34(9):1001–1008
260. Bentz DP, Stutzman PE (2008) Internal curing and microstructure of high performance
mortars. ACI SP-256, internal curing of high performance concretes. Laboratory and Field
Experiences, pp 81–90
261. Henkensiefken R, Sant G, Nantung T et al (2008) Comments on the shrinkage of paste in
mortar containing saturated lightweight aggregate. CONMOD, Delft, The Netherlands
262. Igarashi S, Kubo HR, Kawamura M (2000) Long-term volume changes and microcracks
formation in high strength mortars. Cem Concr Res 30(6):943–951
263. Shin KJ, Bucher B, Weiss J (2011) Role of lightweight synthetic particles on the restrained
shrinkage cracking behavior of mortar. J Mater Civ Eng 23(5):597–605
264. Maghfouri M, Shafigh P, Alimohammadi V et al (2020) Appropriate drying shrinkage predic-
tion models for lightweight concrete containing coarse agro-waste aggregate. J Build Eng
29:101148
265. Zhuang Y, Zheng D, Ng Z et al (2016) Effect of lightweight aggregate type on early-age
autogenous shrinkage of concrete. Constr Build Mater 120:373–381
266. Byard BE, Schindler AK, Barnes RW (2012) Early-age autogenous effects in internally cured
concrete and mortar. Spec Publ 290:1–18
267. Shen D, Jiao Y, Kang J et al (2020) Influence of ground granulated blast furnace slag on
early-age cracking potential of internally cured high performance concrete. Constr Build
Mater 233:117083
268. Smith MA (1975) The economic and environmental benefits of increased use of pfa and
granulated slag. Resour Policy 1(3):154–170
269. Shen DJ, Liu KQ, Wen CY et al (2019) Early-age cracking resistance of ground granulated
blast furnace slag concrete. Constr Build Mater 222:278–287
270. Chi MC, Chi JH, Wu CH (2018) Effect of GGBFS on compressive strength and durability of
concrete. Adv Mater Res 1145:22–26
271. Pal SC, Mukherjee A, Pathak SR (2003) Investigation of hydraulic activity of ground
granulated blast furnace slag in concrete. Cem Concr Res 33(9):1481–1486
272. Altoubat S, Badran D, Junaid MT et al (2016) Restrained shrinkage behavior of self-
compacting concrete containing ground-granulated blast-furnace slag. Constr Build Mater
129:98–105

@seismicisolation
@seismicisolation
References 49

273. Soroushian P, Mirza F, Alhozajiny A (1993) Plastic shrinkage cracking of polypropylene fiber
reinforced concrete. Mater J 92(5):553–560
274. Shen DJ, Zeng X, Zhang JY et al (2019) Behavior of RC box beam strengthened with basalt
FRP using end anchorage with grooving. J Compos Mater 53(23):3307–3324
275. Springenschmid R (1973) Untersuchungen uber die Ursache von Querrissen im jungen Beton.
Beton und Stahlbetonbau 68(9):221–226
276. Springenschmid R (1985) Thermal stresses in mass concrete: a new method and the influence
of different cements. Quinzieme Congres des Grands Barrages, Lausanne
277. Paillere AM, Serrano JJ (1976) Apparatus for studying cracks in concrete. Bull liaison Lab
Ponts Chauss 83:29–38
278. Kovler K (1994) Testing system for determining the mechanical behaviour of early age
concrete under restrained and free uniaxial shrinkage. Mater Struct 27(6):324–330
279. Sule M, van Breugel K (2001) Cracking behaviour of reinforced concrete subjected to early-
age shrinkage. Mater Struct 34(5):284–292
280. Shen DJ, Liu XZ, Li QY et al (2019) Early-age behavior and cracking resistance of high-
strength concrete reinforced with Dramix 3D steel fiber. Constr Build Mater 196:307–316
281. Shen DJ, Wang WT, Li QY et al (2020) Early-age behaviour and cracking potential of fly ash
concrete under restrained condition. Mag Concr Res 72(5):246–261
282. Shen DJ, Jiang JL, Shen JX et al (2016) Influence of curing temperature on autogenous
shrinkage and cracking resistance of high-performance concrete at an early age. Constr Build
Mater 103:67–76
283. Darquennes A, Staquet S, Espion B (2011) Behaviour of slag cement concrete under restraint
conditions. Eur J Environ Civ Eng 15(5):787–798
284. Shen DJ, Liu XZ, Zeng X et al (2020) Effect of polypropylene plastic fibers length on cracking
resistance of high performance concrete at early age. Constr Build Mater 244:117874
285. Barrett TJ, De La Varga I, Weiss WJ (2012) Reducing cracking in concrete structures by
using internal curing with high volumes of fly ash. In: Structures congress: design of concrete
bridges with innovative materials
286. Pane I, Hansen W (2008) Predictions and verifications of early-age stress development in
hydrating blended cement concrete. Cem Concr Res 38(11):1315–1324
287. Lee KM, Lee HK, Lee SH et al (2006) Autogenous shrinkage of concrete containing granulated
blast-furnace slag. Cem Concr Res 36(7):1279–1285
288. Tazawa E, Miyazawa S (1995) Influence of cement and admixture on autogenous shrinkage
of cement paste. Cem Concr Res 25(2):281–287
289. ACI (1995) Effect of restraint, volume change and reinforcement on cracking of mass concrete:
ACI 207.2R-95. American Concrete Institute, Farmington Hills
290. Bentz DP, Peltz MA, Winpigler J (2009) Early-age properties of cement-based materials: II.
Influence of water-to-cement ratio. J Mater Civ Eng 21(9):512–517
291. Kirby DM, Biernacki JJ (2012) The effect of water-to-cement ratio on the hydration kinetics
of tricalcium silicate cements: testing the two-step hydration hypothesis. Cem Concr Res
42(8):1147–1156
292. Zhang MH, Tam CT, Leow MP (2003) Effect of water-to-cementitious materials ratio and
silica fume on the autogenous shrinkage of concrete. Cem Concr Res 33(10):1687–1694
293. Grzybowski M, Shah SP (1990) Shrinkage cracking of fiber reinforced concrete. Mater J
87(2):138–148
294. Balaguru PN, Shah SP (1985) Alternative reinforcing materials for developing countries. J
Int Dev Technol 3:87–105
295. Cyr MF, Ouyang C, Shah SP (2003) Design of hybrid-fiber reinforcement for shrinkage
cracking by crack width prediction. Brittle Matrix Compos 2003:243–252
296. Wongtanakitcharoen T, Naaman AE (2007) Unrestrained early age shrinkage of concrete with
polypropylene, PVA, and carbon fibers. Mater Struct 40(3):289–300
297. Deng Z, Li J (2007) Mechanical behaviors of concrete combined with steel and synthetic
macro-fibers. Comput Concr 4(3):207–220

@seismicisolation
@seismicisolation
50 1 Introduction

298. Behfarnia K, Behravan A (2014) Application of high performance polypropylene fibers in

concrete lining of water tunnels. Mater Des 55:274–279
299. Ramli M, Kwan WH, Abas NF (2013) Application of non-corrosive barchip fibres for high
strength concrete enhancements in aggressive environments. Compos B Eng 53:134–144

@seismicisolation
@seismicisolation
Chapter 2
Internal Relative Humidity of Early-Age
Internally Cured Concrete

Contents

2.1 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.1.1 SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.1.2 LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
2.2 Internal Relative Humidity Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.2.1 Test Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
2.2.2 Calculation of Internal Relative Humidity Decrease Rate . . . . . . . . . . . . . . . . . . . 55
2.3 Internal Relative Humidity of Early-Age Concrete Internally Cured with SAPs . . . . . . . 55
2.3.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.3.2 Change of Internal Relative Humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
2.3.3 Critical Time of Internal Relative Humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
2.3.4 Internal Relative Humidity Decrease Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
2.3.5 Prediction Model for Internal Relative Humidity . . . . . . . . . . . . . . . . . . . . . . . . . . 60
2.3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.4 Internal Relative Humidity of Early-Age Concrete Internally Cured with LWAs . . . . . . 64
2.4.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
2.4.2 Internal Relative Humidity of Concrete with Different Internal Curing w/c
Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2.4.3 Prediction Models for Critical Time of Internal Relative Humidity and Working
w/c Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
2.4.4 Prediction Model for Internal Relative Humidity of Concrete Internally Cured
with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
2.4.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete . . . . . . . . . 77
2.5.1 Influence of Curing Humidity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
2.5.2 Influence of w/c Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
2.5.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101

Modern construction projects increasingly rely on the use of HSC due to its high
strength and reduced permeability. HSC normally has a low w/c ratio, 0.20–0.35 [1].
The self-desiccation is induced when such concrete does not contain enough water
for the hydration of cement [2, 3]. Due to moisture diffusion and this self-desiccation,
IRH decreases in HSC structures and induces shrinkage [4, 5]. As concrete shrinks,

© Science Press 2023 51

D. Shen, Cracking Control on Early-Age Concrete Through Internal Curing,
https://doi.org/10.1007/978-981-19-8398-6_2
@seismicisolation
@seismicisolation
52 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

tensile stresses will develop in the structure due to restraints [6, 7]. The stresses
may overcome the tensile strength and lead HSC to crack [8]. Thus, in order to
avoid early-age cracking in HSC, the decrease of IRH must be mitigated during
cement hydration [9, 10]. IC is an effective technique to reduce early-age cracking
of concrete with a low w/c ratio, which implies distributing IC water reservoirs in
the concrete [3, 8, 9]. SAPs which are a class of polymeric materials with super-
high water absorption capacity or LWAs can serve as IC water reservoirs to provide
curing water to compensate for the moisture loss of hydrating cement paste [11–13],
and thus reduce the effects of self-desiccation and prevent the RH from decreasing
in concrete [14]. Therefore, investigation on the IRH variations in early-age ICC is
fundamental to evaluate its shrinkage and cracking resistance.
In this chapter, firstly, the effect of IC with SAPs on the IRH of early-age concrete
under sealed and unsealed conditions is investigated, and the models for the IRH of
early-age concrete internally cured with SAPs are proposed considering the effect
of IC water content or working w/c ratio and critical time [15, 16]. Secondly, the
prediction models for working w/c ratio, the critical time, and the IRH variations
in early-age ICC with pre-wetted LWAs are investigated [17]. Besides, the IRH of
early-age concrete under different curing humidity conditions is investigated, and the
prediction model for IRH in HSC under the combined effects of moisture diffusion
and self-desiccation considering the effect of curing humidity is proposed [18].

2.1 Materials

2.1.1 SAPs

The SAPs utilized were a suspension-polymerized, covalently cross-linked acry-

lamide/acrylic acid copolymer with a dry-bulk density of 850 kg/m3 . The diameters
of spherical particles varied from 125 to 150 µm in the dry state. Figure 2.1 shows
a photograph of the SAPs used in experiments before and after absorption.

2.1.2 LWAs

The LWAs used in mixtures were manufactured rotary kiln expanded clay [19]. The
LWAs had a dry-bulk density of 1050 kg/m3 and a 24 h absorption value of 12.0%
by mass of dry material. The crushing strength of dry LWAs was 1.18 MPa [20, 21].
The particle size of LWAs used in experiments was 4.7–12.5 mm. Figure 2.2 shows
the photograph and sieving curve of the pre-wetted LWAs used in experiments.

@seismicisolation
@seismicisolation
2.2 Internal Relative Humidity Measurement 53

(a) Before absorption (b) After absorption

Fig. 2.1 SAPs utilized in experiments

100
90
Percent passing by mass (%)

80
70
60
50
40
30
20
10 LWAs
0
0 2 4 6 8 10 12 14 16
Sieve size (mm)

(a) photography (b) sieving curve

Fig. 2.2 LWAs used in experiments

2.2 Internal Relative Humidity Measurement

2.2.1 Test Method

1. Test system and sensors

Figure 2.3 shows the device of measuring the IRH in concrete. A polyvinyl chloride
tube was used to hold the sensor in the center of the specimen and to avoid fresh
concrete contaminating the sensor, as shown in Fig. 2.3. The polyvinyl chloride tube
was inserted with an aluminum bar to avoid the entering of fresh concrete during
the pouring of concrete. Then, the aluminum bar was replaced by the sensor for
measuring the IRH in concrete specimen, as reported in [22, 23]. For the purpose
of facilitating accurate measurement, a small space between the polyvinyl chloride

@seismicisolation
@seismicisolation
54 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

tube and the sensor was isolated by an O-ring. In order to reduce environmental
disturbance, the epoxy resin sealant was utilized to seal the exposed end of the
polyvinyl chloride tube. The position of the polyvinyl chloride tube and the sensor
in the concrete specimen are also shown in Fig. 2.3. The IRH measurement position
was in the center of the concrete specimen, which was 75 mm away from the spec-
imen surface. The IRH in three identical concrete specimens under the same curing
humidity condition was measured simultaneously, and the average value was utilized
for analysis. The concrete specimens were cured in an environmental chamber at a
constant temperature of 20 ± 0.5 °C and a designed constant humidity, respectively.
Figure 2.3 shows the specimen size and arrangements of the temperature and
humidity sensors, which were designed according to the scheme reported in [24,
25]. With the nominal accuracies of ± 1.8% RH and ± 0.3 °C, a digital resistance-
based sensor (DB4850-DB170-N) was utilized to measure the variations of IRH
and temperature in concrete. The test data were collected by the computer system.
Plastic film was utilized to seal six sides of the sealed concrete specimens. For better
preventing moisture diffusion, tinfoil and paraffin wax were also utilized to seal the
surface of the sealed concrete specimens. Meanwhile, only five sides were sealed
with plastic film for unsealed concrete specimens, and the top surface of which was
exposed to external environment directly to ensure that moisture diffused only in
uniaxial direction. Self-desiccation took place inside the concrete, while moisture
diffusion occurred at the exposed surface of the unsealed concrete specimen, as
reported in [26]. Therefore, the variations of IRH in unsealed or sealed HSC were
affected by self-desiccation and moisture diffusion or self-desiccation alone.

2. Calibration method

The salt crystallization and dissolution of salts in the pore water could cause the
changes of IRH and autogenous deformation, as reported in [1]. Additionally, changes
of IRH caused by dissolved salts are difficult to be distinguished from measured IRH
changes [1]. Therefore, the corresponding results and discussion of IRH variations
included the effect of alkali on the IRH in this section. With a RH in the range of
75~100% RH, the saturated salt solution was utilized to calibrate the equipment
before and after every experiment, as reported in [27].

Fig. 2.3 Specimen size and arrangements of the temperature and humidity sensors

@seismicisolation
@seismicisolation
2.3 Internal Relative Humidity of Early-Age Concrete Internally Cured … 55

2.2.2 Calculation of Internal Relative Humidity Decrease

Rate

The decrease of IRH in a given unit of time is defined as the IRH decrease rate [28],
which could be calculated by Eq. (2.1), as reported in [18].

RH(t) − RH(t−1)
DR = (2.1)
Δt

where DR is the IRH decrease rate, in % RH/d; RH (t) and RH (t−1) are the IRH at
time step t and t − 1, respectively, in %; and Δt is the length of time interval between
the time steps, in d.

2.3 Internal Relative Humidity of Early-Age Concrete

Internally Cured with SAPs

2.3.1 Mixture Proportion

Four concrete mixtures with a low w/c ratio were used. The mixture proportions,
which were designated as SAP0, SAP5, SAP15, and SAP25, as shown in Table 2.1.
Mixture SAP0 represented a reference concrete without IC, whereas Mixtures SAP5,
SAP15, and SAP25 underwent IC through SAPs. The concrete compositions were
varied by the addition of SAPs (0.05%, 0.15%, and 0.25% by weight of cement for
Mixtures SAP5, SAP15, and SAP25, respectively). In the reference concrete mixture
(SAP0) without IC, the w/c ratio was fixed as 0.33. The IC water to cement (wic /c)
ratio of Mixtures SAP5, SAP15, and SAP25 varied respectively as 0.01, 0.03, and
0.05. That is to say, the total w/c ratio varied from 0.33 to 0.39.
The theoretical SAPs quantity required to ensure maximum cement hydration
was estimated to be 1539 g/m3 (0.30% by weight of cement) for all mixtures in this
experiment. Given that different levels of SAPs were designed, the masses of SAPs in
Mixtures SAP0, SAP5, SAP15, and SAP25 were lower than the required quantities
for IC. The required SAPs amounts for the mixture designs shown in Table 2.1 are
listed as follows: 0% for Mixture SAP0, 0.05% for Mixture SAP5, 0.15% for Mixture
SAP15, and 0.25% for Mixture SAP25 by weight of cement. IC water contents in
Mixtures SAP0, SAP5, SAP15, and SAP25 were 0, 5.35, 16.06, and 26.78 kg/m3 ,
respectively. Each mixture was applied to both sealed and unsealed specimens to
measure the IRH in concrete. The specimens under sealed condition were denoted as
SAP0-S, SAP5-S, SAP15-S, and SAP25-S; whereas the specimens under unsealed
condition were labeled as SAP0-U, SAP5-U, SAP15-U, and SAP25-U, as indicated
in Table 2.1.

@seismicisolation
@seismicisolation
56 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Table 2.1 Compositions and properties of fresh mixtures

Mixture composition Concrete mixtures
SAP0 SAP5 SAP15 SAP25
Identifications of specimens SAP0-S SAP5-S SAP15-S SAP25-S
SAP0-U SAP5-U SAP15-U SAP25-U
Water (kg/m3 ) 171 171 171 171
Cement (kg/m3 ) 512 512 512 512
Fine aggregate (kg/m3 ) 636 636 636 636
Coarse aggregate (kg/m3 ) 1131 1131 1131 1131
SAPs (kg/m3 ) 0 267.5 803.0 1339.0
Internal curing water (kg/m3 ) 0 5.35 16.06 26.78
Superplasticizer (kg/m3 ) 3.6 3.6 3.6 3.6
Slump (mm) 100 110 105 110

2.3.2 Change of Internal Relative Humidity

The IRH of ICC specimens under sealed conditions is influenced only by self-
desiccation. Figure 2.4 shows the test result regarding the IRH variations of concrete
that was internally cured with SAPs under sealed conditions.
At 14 d, the IRH was 96.5%, 97.5%, 98.5%, and 100% RH given IC water contents
of 0, 5.35, 16.06, and 26.78 kg/m3 for specimens SAP0-S, SAP5-S, SAP15-S, and
SAP25-S under sealed conditions, respectively. At 21 d, the IRH was 94.3%, 95.4%,
97.0%, and 99.7% RH when IC water contents were 0, 5.35, 16.06, and 26.78 kg/m3
for specimens SAP0-S, SAP5-S, SAP15-S, and SAP25-S under sealed condition,
respectively. At 28 d, the IRH was 92.8%, 93.9%, 96.0%, and 98.7%. These humidi-
ties increased by 1.1%, 3.2%, and 5.9% RH in the same period when IC water

Fig. 2.4 Relationship

between IRH and age of 100
sealed specimens

98
IRH (%)

96
SAP0-S
SAP5-S
94
SAP15-S
SAP25-S
92
0 5 10 15 20 25 30
Age of specimens (d)

@seismicisolation
@seismicisolation
2.3 Internal Relative Humidity of Early-Age Concrete Internally Cured … 57

Fig. 2.5 Relationship 100

between IRH and age of
98
unsealed specimens
96
94

IRH (%)
92
90
88 SAP0-U
SAP5-U
86
SAP15-U
84 SAP25-U
82
0 5 10 15 20 25 30
Age of specimens (d)

content increased from 0 to 5.35, 16.06, and 26.78 kg/m3 for specimens SAP5-S,
SAP15-S, and SAP25-S, respectively. Therefore, IC with SAPs strongly affects the
IRH of concrete under sealed conditions; this finding is in accordance with the result
obtained by Kong et al. [29]. The higher the IC water content is, the higher the
IRH becomes. IRH reductions are decreased and delayed considerably for speci-
mens SAP5-S, SAP15-S, and SAP25-S, as shown in Fig. 2.4. Moreover, the water
consumed by cement hydration is complemented by the water released from the SAPs
[30]. Therefore, the IRH of specimens with SAPs under sealed condition decreases
because the IC water supplied by these polymers cannot counteract the loss of water
to cement hydration.
The IRH of concrete specimens that were internally cured under unsealed condi-
tion is influenced by both self-desiccation and drying (moisture diffusion). Figure 2.5
shows the test result regarding IRH variations in IC concrete with SAPs under
unsealed condition.
At 3 d, the IRH was 96.1%, 99.8%, 100%, and 100% RH when IC water contents
were 0, 5.35, 16.06, and 26.78 kg/m3 for specimens SAP0-U, SAP5-U, SAP15-U,
and SAP25-U under unsealed conditions, respectively. At 7 d, the IRH was 92.0%,
96.5%, 97.7%, and 100% RH. At 14 d, the IRH was 88.1%, 93.0%, 94.5%, and 97.5%
RH. At 21 d, the IRH was 86.0%, 90.7%, 92.6%, and 95.6% RH. At 28 d, the IRH was
84.4%, 89.1%, 91.2%, and 94.5% RH. IRH increased by 4.7%, 6.8%, and 10.1%
RH when IC water content increased from 0 to 5.35, 16.06, and 26.78 kg/m3 for
specimens SAP5-U, SAP15-U, and SAP25-U, respectively. The IRH of specimens
with SAPs under unsealed conditions decreases because the IC water in the SAPs
that flows into the cement paste during hydration cannot counteract the IRH loss
caused by self-desiccation and drying completely.
The IRH variations in the specimens were investigated under sealed and unsealed
conditions, as shown in Figs. 2.4 and 2.5, respectively. The IRH of concrete that was
internally cured with SAPs was 8.4%, 5.3%, 5.5%, and 4.7% RH higher under sealed

@seismicisolation
@seismicisolation
58 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

conditions than under unsealed conditions when IC water contents were 0, 5.35,
16.06, and 26.78 kg/m3 for Mixtures SAP0, SAP5, SAP15, and SAP25, respectively.

2.3.3 Critical Time of Internal Relative Humidity

IRH development in IC concrete can be described through a water–vapor saturated

stage with 100% RH (stage I), followed by a stage in which IRH decreases gradually
(stage II). The length of stage I is defined as critical time [8]. Figure 2.4 indicates
that critical times were 7, 8, 10, and 19 d when IC water contents were 0, 5.35,
16.06, and 26.78 kg/m3 for specimens SAP0-S, SAP5-S, SAP15-S, and SAP25-S
under sealed conditions, respectively. These times increased by 1, 3, and 12 d when
IC water contents increased from 0 to 5.35, 16.06, and 26.78 kg/m3 for specimens
SAP5-S, SAP15-S, and SAP25-S, respectively. IRH reductions were decreased and
delayed considerably for specimens SAP5-S, SAP15-S, and SAP25-S, as indicated
in Fig. 2.4. Moreover, the water consumed by cement hydration is complemented by
the water released from SAPs [30]. Therefore, the critical time of IRH increases with
an increase in IC water content. The result suggests that the water–vapor saturated
stage ends early under the effects of both self-desiccation and drying, as shown in
Fig. 2.5.
The critical times were 1, 2, 3, and 7 d when IC water contents were 0, 5.35,
16.06, and 26.78 kg/m3 for specimens SAP0-U, SAP5-U, SAP15-U, and SAP25-U
under unsealed conditions, respectively. These times increased by 1, 2, and 6 d when
IC water contents increased from 0 kg/m3 for specimen SAP0-U to 5.35, 16.06,
and 26.78 kg/m3 for specimens, SAP5-U, SAP15-U, and SAP25-U under unsealed
conditions, respectively. At 7 d, the IRH was 92.0%, 96.5%, 97.7%, and 100% RH for
specimens SAP0-U, SAP5-U, SAP15-U, and SAP25-U under unsealed conditions,
respectively. The IC water released by the SAPs significantly influenced the IRH
of concrete under unsealed conditions, and the IRH was higher than the 95.0% RH
detected in the first seven d after casting.

2.3.4 Internal Relative Humidity Decrease Rate

The IRH is related to autogenous shrinkage [31]. The IRH decrease rate affects
autogenous shrinkage rate of early-age concrete. Furthermore, the strength of this
concrete is lower than the desired strength [32]. A high autogenous shrinkage rate
induces cracking in concrete. Therefore, IRH decrease rate of concrete that was
internally cured with SAPs must be investigated.
Figure 2.6 shows that for specimen SAP0-S, the average IRH decrease rate was
0.408% RH/d when the age of the concrete ranged from 8 to 16 d.
This average IRH decrease rate was 0.29% RH/d when the concrete age increased
from 17 to 24 d. When the concrete age increased from 25 to 28 d, the average

@seismicisolation
@seismicisolation
2.3 Internal Relative Humidity of Early-Age Concrete Internally Cured … 59

Fig. 2.6 Relationship 0.6

between decreasing rate of
IRH and concrete age under
sealed condition

Rate (% RH/d)
0.4

0.2 SAP0-S
SAP5-S
SAP15-S
SAP25-S
0.0
0 5 10 15 20 25 30
Age of specimens (d)

IRH decrease rate was in the range of 0.1–0.2% RH/d. For specimen SAP5-S, this
average rate was 0.35% RH/d when the concrete age increased from 9 to 23 d. When
the concrete age increased from 23 to 28 d, the average IRH decrease rate was in
the range of 0.1–0.2% RH/d. For specimen SAP15-S, this average rate was 0.33%
RH/d when the concrete age increased from 11 to 17 d. The average IRH decrease
rate was in the range of 0.1–0.2% RH/d when the concrete age increased from 18 to
28 d. This rate began to decrease when the concrete age was 20 d. The average IRH
decrease rates were in the range of 0.1–0.2% RH/d when the concrete age increased
from 20 to 28 d.
Figure 2.7 shows the average IRH decrease rate in the concrete with or without
IC water as a result of cement hydration and drying.

Fig. 2.7 Relationship 2.0

between decreasing rate of SAP0-U
IRH and concrete age under
Decrease rate (% RH/d)

1.6 SAP5-U
unsealed condition
SAP15-U
SAP25-U
1.2

0.8

0.4

0.0
0 5 10 15 20 25 30
Age of specimens (d)

@seismicisolation
@seismicisolation
60 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

The average IRH decrease rate for specimen SAP0-U without IC was 0.83% RH/d
in the first 15 d after casting. The average IRH decrease rates for specimens SAP5-
U, SAP15-U, and SAP25-U were 0.58%, 0.48%, and 0.34% RH/d when IC water
contents were 5.35, 16.06, and 26.78 kg/m3 , respectively, in the first 15 d after casting
under the effects of both self-desiccation and drying. These rates ranged from 0.1 to
0.3% RH/d after the first 15 d.
For specimens with SAPs, the IRH decrease rate dropped considerably at early
age because the IC water released from the SAPs complemented the partial water
consumption attributed to self-desiccation and drying. Therefore, IRH decrease rate
in ICC decreased as IC water content increased.

2.3.5 Prediction Model for Internal Relative Humidity

Figure 2.8 indicates that the relationship between IRH and IC water content was
nonlinear.
The IRH of concrete aged 3, 7, 14, 21, and 28 d increased nonlinearly with an
increase in IC water content. This relationship could be expressed by Eq. (2.2).
√
RH (t, Q) = Kt Q + RH (t, 0) (2.2)

where RH (t, Q) is IRH, in %; Q is IC water content, in kg/m3 ; Kt is a parameter

related to IRH and IC water content, in m1.5 /kg0.5 ; and RH (t, 0) is the IRH of concrete
without IC at different ages, in %.
Under sealed conditions, specimens SAP15-S and SAP25-S entered a water–vapor
saturated stage with 100% RH for a total of 10 d after the concrete was cast. Therefore,
only the results for the variations of IRH in concrete that was aged longer than 10 d
were analyzed. The relationship between IRH and IC content was determined with
Eqs. (2.3), (2.4), and (2.5) when the concrete was aged 14, 21, and 28 d, respectively.
√
RH (14, Q) = 0.644 Q + 96.274, R2 = 0.910 (2.3)

√
RH (21, Q) = 0.981 Q + 93.780, R2 = 0.826 (2.4)

√
RH (28, Q) = 1.095 Q + 92.204, R2 = 0.845 (2.5)

The following equations could be obtained by substituting the IRH of the concrete
specimens aged 3, 7, 14, 21, and 28 d under unsealed conditions along with IC water
content into Eq. (2.2).
At 3 d after casting:
√
RH (3, Q) = 0.747 Q + 96.828, R2 = 0.648 (2.6)

@seismicisolation
@seismicisolation
2.3 Internal Relative Humidity of Early-Age Concrete Internally Cured … 61

Fig. 2.8 Relationship

between IRH and IC water 100
content
98

IRH (%)
96
14 d
94 21 d
28 d

92
0 5 10 15 20 25 30
Water (kg/m3)
(a) Sealed specimens

100

96
IRH (%)

92
3d
7d
88 14 d
21 d
84 28 d

0 5 10 15 20 25 30
Water (kg/m3)
(b) Unsealed specimens

At 7 d after casting:
√
RH (7, Q) = 1.475 Q + 92.312, R2 = 0.957 (2.7)

At 14 d after casting:
√
RH (14, Q) = 1.725 Q + 88.318 (2.8)

At 21 d after casting:
√
RH (21, Q) = 1.778 Q + 86.117, R2 = 0.978 (2.9)

@seismicisolation
@seismicisolation
62 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Fig. 2.9 Relationship

between IRH and age of
specimen without internal
curing

At 28 d after casting:
√
RH (28, Q) = 1.868 Q + 84.430, R2 = 0.978 (2.10)

Figure 2.9 shows the variations of the IRH in concrete without IC water. The
relationship between IRH and concrete age under unsealed conditions and without
IC water was determined as follows.

RH (t, 0) = 100 − γ × t θ (2.11)

where γ and θ are parameters that can be determined with regression analysis. The
resultant values are γ =2.636, θ =0.543, and R2 =0.989.
Under the effects of drying and self-desiccation for the specimens under unsealed
conditions, the relationship between parameter Kt and concrete age t was shown in
Fig. 2.10 and could be expressed as follows.

Kt = α × (1 − e−βt ) (2.12)

where α and β are parameters that can be determined with regressive analysis. Kt
is a parameter related to concrete age. The resultant values are α=1.852, β=0.197,
and R2 =0.973.
Figure 2.10 indicates that Kt increased with concrete age. This finding suggests
that IC water has an accumulated effect on the IRH of concrete after casting [25].
Moreover, the increase rate of Kt decreased with an increase in concrete age. This
observation implied that the effect of IC water on IRH weakens when concrete age
increased as a result of IC water depletion.
Equation (2.13) was proposed by substituting Eqs. (2.11) and (2.12) into Eq. (2.2).
Then, the relationship between RH and age was determined as follows when IC water
content was considered.

@seismicisolation
@seismicisolation
2.3 Internal Relative Humidity of Early-Age Concrete Internally Cured … 63

Fig. 2.10 Relationship

between Kt and age of
specimens

√
RH (t, Q) = α × (1 − e−βt ) Q + (100 − γ × t θ ) (2.13)

The following equation was generated by substituting the values of parameters α,

β, γ , and θ into Eq. (2.13).
√
RH (t, Q) = 1.852 × (1 − e−0.197t ) Q + (100 − 2.636 × t 0.543 ) (2.14)

Figure 2.11 shows the slight differences between the results from testing and
from Eq. (2.14) with respect to the IRH in concrete that was internally cured under
unsealed conditions.

Fig. 2.11 Fitting curve of 100

relationship between IRH
and age of unsealed 95
specimens with SAPs
90
IRH (%)

85 Fitting curve of SAP0-U

Fitting curve of SAP5-U
Fitting curve of SAP15-U
80 Fitting curve of SAP25-U
Test data of SAP0-U
Test data of SAP5-U
75
Test data of SAP15-U
Test data of SAP25-U
70
0 5 10 15 20 25 30
Age of specimen (d)

@seismicisolation
@seismicisolation
64 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

2.3.6 Summary

The effect of IC with SAPs on the IRH of early-age concrete under sealed and
unsealed conditions was investigated. The following conclusions can be drawn based
on the obtained results.
(1) The IRH of ICC with SAPs increased with IC water content at 28 d under sealed
or unsealed conditions. The decrease degree of the IRH in ICC was lower under
sealed conditions than that under unsealed conditions during this period.
(2) The critical time of the IRH in ICC increased with IC water content under sealed
and unsealed conditions.
(3) The IRH decrease rate of early-age concrete that was internally cured with
SAPs decreased with an increase in IC water content under sealed and unsealed
conditions. This phenomenon occurs because the IC water released from the
SAPs complemented the partial water consumption attributed to self-desiccation
and drying.
(4) The model for the IRH of early-age concrete internally cured with SAPs was
proposed in consideration of IC water content under unsealed conditions.

2.4 Internal Relative Humidity of Early-Age Concrete

Internally Cured with LWAs

2.4.1 Mixture Proportion

Six concrete mixtures with different w/c ratios were used, which were designed in
accordance with results in [33]. The mixture proportions, which were designated as
RC33-0, RC40-0, RC50-0, IC33-10, IC33-30, and IC33-50, as shown in Table 2.2.
Mixtures RC33-0, RC40-0, and RC50-0 were ordinary concrete without pre-wetted
LWAs and the corresponding w/c ratios were 0.33, 0.40, and 0.50, respectively.
Mixtures IC33-10, IC33-30, and IC33-50 were ICC with pre-wetted LWAs and the
corresponding volume of pre-wetted LWAs was 10%, 30%, and 50% of the total
volume of coarse aggregates, respectively.
The ratio of water provided by pre-wetted LWAs to cement is defined as IC w/c
ratio in [8] and the ratio of mixing water (without IC water) to cement in concrete
is defined as the effective w/c ratio in [33]. Although results reported in [14] show
that not all extra water in LWAs is immediately available for IC and some water in
LWAs is only removed at low RH. The ratio of total water (mixing water and IC
water) to cement is defined as the total w/c ratio in [29]. In the reference concrete
mixture (RC33-0), the effective w/c ratio was fixed as 0.33. Therefore the total w/c
ratio [29] is in combination of both the effective w/c ratio and the IC w/c ratio and
could be expressed by Eq. (2.15).
/ / /
wt c = we c + wic c (2.15)

@seismicisolation
@seismicisolation
2.4 Internal Relative Humidity of Early-Age Concrete Internally Cured … 65

Table 2.2 Compositions and properties of fresh mixtures

Mixture composition Concrete mixtures
RC33-0 RC40-0 RC50-0 IC33-10 IC33-30 IC33-50
we /c 0.33 0.40 0.50 0.33 0.33 0.33
Water (kg/m3 ) 171 180 200 171 171 171
Cement (kg/m3 ) 512 450 400 512 512 512
Fine aggregate(sand) (kg/m3 ) 636 655 666 636 636 636
Coarse aggregate(stone) (kg/m3 ) 1131 1165 1184 1018 792 566
LWAs (kg/m3 ) 0 0 0 44.6 133.9 223.2
Internal curing water (kg/m3 ) 0 0 0 5.35 16.06 26.78
wic /c 0 0 0 0.01 0.03 0.05
Superplasticizer (kg/m3 ) 3.6 1.8 0.8 3.6 3.6 3.6
Slump (mm) 100 140 160 110 140 135

/ / /
where wt c is the total w/ c ratio; we c is the effective w/ c ratio; and wic c is the
IC w/c ratio.
Therefore, the IC w/c ratio was 0.01, 0.03, and 0.05 and the total w/c ratios
were 0.34, 0.36, and 0.38 for Mixtures IC33-10, IC33-30, and IC33-50, respectively.
Moreover, the additional w/c ratios (the ratio of additional mixing water [29] to
cement) were 0, 0.07, and 0.17 for RC33-0, RC40-0, and RC50-0, respectively.
The percentage of IC water actually provided into Mixtures to that required in
theory could be designed as follows.

mwat × wic /c
η= (2.16)
Cf

where η is the percentage of the IC water added into Mixtures to the water required
for the maximum degree of hydration.
The theoretical quantities of LWAs required to ensure maximum cement hydration
were estimated at 273.8 kg/m3 for all mixtures in this experiment. Since different
levels of pre-wetted LWAs were designed, thus the mass of LWAs provided in
Mixtures IC33-10, IC33-30, and IC33-50 were lower than required quantities of
LWAs [33]. Therefore, for the mixture designs shown in Table 2.2, the percentage
of IC water actually provided to that required in theory was as follows: η = 0% for
Mixture RC33-0; η = 16.3% for Mixture IC33-10; η = 48.9% for Mixture IC33-30;
and η = 81.5% for Mixture IC33-50, respectively.

@seismicisolation
@seismicisolation
66 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

2.4.2 Internal Relative Humidity of Concrete with Different

Internal Curing w/c Ratios

In concrete specimens exposed to the air, the IRH in ordinary and ICC are influenced
by both self-desiccation and moisture diffusion [34]. Figure 2.12 shows the test
results on IRH variations in concrete.
The IRH at 28 d after casting was 84.2%, 90.0%, 94.1%, 89.2%, 92.0%, and 95.3%
RH when the total w/c ratios were 0.33, 0.40, 0.50, 0.34, 0.36, and 0.38 for Mixtures
RC33-0, RC40-0, RC50-0, IC33-10, IC33-30, and IC33-50, respectively. The IRH
at 28 d after casting increased by 5.0%, 7.8%, and 11.1% RH when the IC w/c ratios
increased from 0 to 0.01, 0.03, and 0.05 for Mixtures RC33-0, IC33-10, IC33-30,
and IC33-50, respectively. The IRH measurement was the average of the results for
three tested specimens. The values of standard deviation of IRH measurements were
0.08%, 0.16%, 0.12%, 0.12%, and 0.05% RH at the age of 3, 7, 14, 21, and 28 d for
Mixture IC33-10, respectively. Therefore, for Mixture IC33-10, the average value of
the standard deviation of IRH measurements was 0.11% RH. And the average values
of the standard deviation of IRH measurements were 0.10%, and 0.09% RH for
Mixtures IC33-30, and IC33-50, respectively. Results reported in [5] also show that
the IRH in ICC with pre-wetted LWAs at 28 d after casting increases with the increase
of IC w/c ratio. The IRH enters a water–vapor saturated stage with 100% RH at the
initial period after casting due to the high amount of liquid water in the pores [35].
Then, the self-desiccation means unhydrated cement in concrete continues to hydrate
accompanying the consumption of the liquid water provided by cement paste and IC
water [36]. Furthermore, the moisture decreases due to moisture diffusion [34, 37].
When the water in concrete pores cannot counteract the water reduction resulting
from cement hydration and drying [8], the IRH in concrete begins to decrease from
the initial water–vapor saturated stage with 100% RH.
The IRH at 28 d after casting was 92.0%, 95.3%, and 90.0% RH when the total
w/c ratios were 0.36, 0.38, and 0.40 for Mixtures IC33-30, IC33-50, and RC40-0,

Fig. 2.12 Test results of 100

relationship between IRH 98
and age of specimens
96
94
IRH (%)

92
90 RC33-0
88 RC40-0
RC50-0
86 IC33-10
84 IC33-30
IC33-50
82
0 5 10 15 20 25 30
Age of specimens (d)

@seismicisolation
@seismicisolation
2.4 Internal Relative Humidity of Early-Age Concrete Internally Cured … 67

respectively. The IRH in ICC at 28 d after casting was 2.0%, and 5.3% RH higher
for Mixtures IC33-30, and IC33-50 with total w/c ratios of 0.36, and 0.38 than that
of Mixture RC40-0 with total w/c ratio of 0.40, respectively. The IC w/c ratios in
Mixtures IC33-30 (with the total w/c ratio of 0.36, which was in combination of the
effective w/c ratio of 0.33 and IC w/c ratio of 0.03), and IC33-50 (with the total w/c
ratio of 0.38, which was in combination of the effective w/c ratio of 0.33 and IC
w/c ratio of 0.05) were more effective in preventing the IRH from decreasing in the
concrete than the additional w/c ratio in Mixture RC40-0 (with the total w/c ratio of
0.40, which was in combination of the effective w/c ratio of 0.33 and additional w/c
ratio of 0.07), respectively, which was in accordance with results in [29]. Results
reported in [29] show that the absorbed water by the pre-soaked SAPs is much
more effective in increasing the RH in concrete than the additional mixing water.
The possible reason for the difference is that the two types of water (IC water and
additional mixing water) result in completely different microstructure of cement
pastes [38], and the IC water in mixture results in smaller average size and porosity
of capillary pores than additional mixing water [29].
The average value of IRH decrease rate with the time interval of three days was
used, which was in accordance with the method used in [28], as shown in Fig. 2.13.
The average value of IRH decrease rate in concrete at the age of 7, 10, 13, 16, and
19 d after casting were 0.54, 0.41, 0.25, 0.43, 0.42, and 0.22% RH/d for Mixtures
RC33-0, RC40-0, RC50-0, IC33-10, IC33-30, and IC33-50, respectively, as shown in
Fig. 2.13. These average values of IRH decrease rate after the first 19 d after casting
ranged from 0.20% RH/d to 0.26% RH/d. The average values of IRH decrease rate
when the age of concrete was from 7 d to 19 d decreased by 0.11%, 0.12%, and 0.32%
RH/d when the IC w/c ratios increased from 0 to 0.01, 0.03, and 0.05 for Mixtures
RC33-0, IC33-10, IC33-30, and IC33-50, respectively. Therefore the average value
of IRH decrease rate in ICC with pre-wetted LWAs decreased nonlinearly with the
increase of IC w/c ratio. The IRH in concrete decreased rapidly during the first 19 d
and then decreased slowly during the following days [28], as shown in Fig. 2.13. The
reason for the changes of IRH decrease rate is also investigated in [28] considering

Fig. 2.13 Relationship 2.0

between IRH decrease rate RC33-0
and age of specimens 1.6 RC40-0
Decrease rate (% RH/d)

RC50-0
1.2 IC33-10
IC33-30
IC33-50
0.8

0.4

0.0

0 5 10 15 20 25 30
Age of specimens (d)

@seismicisolation
@seismicisolation
68 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

the changes of capillary forces. The IRH is high at the beginning, the free liquid
water shown in the pores of the material is subjected to capillary forces, and it moves
due to pressure gradients at a faster rate during the first 19 d than the vapor and bound
water during the following days [28].

2.4.3 Prediction Models for Critical Time of Internal Relative

Humidity and Working w/c Ratio

Figure 2.12 shows the IRH variations in concrete with total w/c ratios of 0.33, 0.40,
0.50, 0.34, 0.36, and 0.38 for Mixtures RC33-0, RC40-0, RC50-0, IC33-10, IC33-
30, and IC33-50, respectively. Many models have been developed to predict the IRH
variations in ordinary concrete, and three models were selected from the literatures
[39–41].
Persson’s model is expressed as follows [39]:

φ(w/ c, t)a50 = α × (1 − β × ln(t)) × (w/ c)δ×(1+γ ×ln(t)) (2.17)

where φ(w/ c, t)a50 is the IRH in concrete with different w/ c ratios and ages, in %;
a50 is the location of IRH which is 50 mm away from the exposed surface under
air condition, in mm; α, β, δ, and γ are the parameters that can be determined with
regression analysis; and t is the age of concrete (28<t<450), in d.
Parrott’s model is expressed as follows [40]:
⎧
⎨ RH = RHA
⎪
/(
+ (100 − RHA) × f (t)
/ )
f (t) = 1 1 + t b (2.18)
⎪
⎩
b = d X (Y − e)(w − Z)/W

where RH is the IRH in concrete, in %; RHA is the ambient RH, in %; t is the

drying time, in d; d is the depth from drying surface, in mm; e is the amount of
Ordinary Portland Cement replaced by pulverized fly ash or GGBFS, e = 0; w is the
water/binder ratio; and W , X , Y , and Z are parameters. The four parameters assigned
by Parrott [40] are W = 8.00, X = 1.35, Y = 70.00, and Z = 0.19, respectively.
Then the parameters in Eq. (2.18) are suggested in [42] to improve Parrott’s original
model, the values of parameters suggested are W =11.40, X =0.85, Y =128.00, and
Z=0.03, respectively.
Based on the experimental data on IRH variations and models in [39, 40], the
model proposed in [41] for IRH in early-age concrete is expressed as follows.
{
( / ) 100 t ≤ Tcri,1
RH w c, t = √ (2.19)
100 + Kt w/ c + bt Tcri,1 < t ≤ 28d

@seismicisolation
@seismicisolation
2.4 Internal Relative Humidity of Early-Age Concrete Internally Cured … 69

where

Tcri,1 = α1 × (w/ c)2 +β1 × (w/ c) + γ1 (2.20)

and

Kt =7.295 + 20.586 × ln(t) (2.21)

and

bt = − 1.822 − 17.052 × ln(t) (2.22)

where RH(w/ c, t) is the IRH in concrete, in %; Tcri,1 is the critical time in ordi-
nary concrete, in d; α1 , β1 , and γ1 are the parameters that can be determined with
regression analysis, and the results are α1 = 31.092, β1 = −16.983, and γ1 = 3.218,
respectively; Kt and bt are parameters related to age of concrete; and t is age of
specimens (1≤t≤28), in d.
Figure 2.14 shows the comparisons of IRH in concrete between the theoretical
results from the models developed by Persson in [39] or Parrott in [40] and test results
of ordinary concrete. The deviations of IRH during the first 5 d after casting between
results obtained from Eq. (2.17) developed by Persson in [39] and test were large.
Therefore, Eq. (2.17) could not be used to accurately predict the IRH in early-age
concrete during the first 5 d. The deviations of IRH in ordinary concrete without IC
between results obtained from Eq. (2.18) using parameters in [40] or [42] and test
were large.
Figure 2.14 shows that Eq. (2.18) could not be used to predict IRH in early-
age concrete with a low w/c ratio [58]. The deviations of IRH in ordinary concrete
between the theoretical results from Eq. (2.19) developed in [41] and test results
were smaller than that from Eq. (2.17) developed by Persson in [39] or Eq. (2.18)
developed by Parrott in [40] and test results. Therefore, Eq. (2.19) developed in [41]
could be used to predict the IRH in ordinary concrete at different ages considering
different w/c ratios.
Furthermore, results reported in [29] show that the IC water in concrete is much
more effective in preventing IRH from decreasing than the corresponding additional
mixing water. Therefore, Eq. (2.19) was not suitable for calculating the IRH variations
in ICC with pre-wetted LWAs. The working w/c ratio (ww /c) was used to accurately
predict the IRH in ICC, as shown in Eq. (2.23).
{
( / ) 100 t ≤ Tcri,2
RH ww c, t = √ (2.23)
100 + Kt w/ c + bt Tcri,3 < t ≤ 28d

where RH(ww /c, t) is the IRH in concrete in consideration of ww /c, in %; and Tcri,2
is the critical time in consideration of ww /c, in d.

@seismicisolation
@seismicisolation
70 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Fig. 2.14 Comparison of 104

RH in ordinary concrete 100
from test results and
theoretical models 96

IRH (%)
92
88
84 Test data of RC33-0
Parrott [40]
80 McDonald [42]
Persson [39]
76 Model in [41]

72
0 5 10 15 20 25 30
Age of specimens (d)
(a) RC33-0

104

100

96
IRH (%)

92
Test data of RC40-0
88 Parrott [40]
McDonald [42]
84 Persson [39]
Model in [41]
80
0 5 10 15 20 25 30
Age of specimens (d)
(b) RC40-0

104

100
IRH (%)

96
Test data of RC50-0
Parrott [40]
92 McDonald [42]
Persson [39]
Model in [41]
88
0 5 10 15 20 25 30
Age of specimens (d)
(c) RC50-0

@seismicisolation
@seismicisolation
2.4 Internal Relative Humidity of Early-Age Concrete Internally Cured … 71

Fig. 2.15 Relationship 0.65

Working w/c ratio (ww /c)

Test data of IC33-10
between working w/c ratio Test data of IC33-30
and age of specimens 0.60 Test data of IC33-50
Fitting curve of IC33-10
Fitting curve of IC33-30
0.55 Fitting curve of IC33-50

0.50

0.45

0.40
0 5 10 15 20 25 30
Age of specimens (d)

Figure 2.15 shows that the working w/c ratio was higher than the total w/c ratio
for each concrete mixture.
Therefore, the influence coefficient f (wic /c, t) was used to evaluate the efficiency
of the IC w/c ratio for better predicting the IRH in ICC. The formula to calculate
working w/c ratio could be expressed as follows.

ww /c = we /c + f (wic /c, t) × wic /c (2.24)

where f (wic /c, t) is the influence coefficient.

The influence coefficient f (wic /c, t) could be determined from the test results and
formula in combination of Eqs. (2.23) and (2.24), as shown in Fig. 2.16. It showed that
f (wic /c, t) was a function of both IC w/c ratio and age of concrete with pre-wetted
LWAs.
Figure 2.17 shows that the relationship between the influence coefficient and IC
w/c ratio was nonlinear and this relationship could be expressed with Eq. (2.25).

Fig. 2.16 Relationship 13

Influence coefficient f (wic /c, t)

Test data of IC33-10

between influence coefficient 12 Test data of IC33-30
f (wic /c, t) and age of 11 Test data of IC33-50
specimens Fitting curve of IC33-10
10 Fitting curve of IC33-30
Fitting curve of IC33-50
9
8
7
6
5
4
3
0 5 10 15 20 25 30
Age of specimens (d)

@seismicisolation
@seismicisolation
72 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

10

Influence coefficient f (wic /c, t)

Fig. 2.17 Relationship
between f (wic /c, t) and Test data of 7 d
wic /c ratio 9 Test data of 14 d
Test data of 21 d
8 Test data of 28 d
Fitting curve of 7 d
7 Fitting curve of 14 d
Fitting curve of 21 d
Fitting curve of 28 d
6
5
4
3
0.01 0.02 0.03 0.04 0.05 0.06
wic / c ratio

f (wic /c, t) = Yt × (wic /c)pt (2.25)

where Yt and pt are parameters related to age of concrete.

Mixture IC33-50 entered a water–vapor saturated stage with 100% RH for a total
of 6.2 d after casting. Therefore, only the results for the variations of the influence
coefficient with the age longer than 6.2 d were analyzed. The following equations
could be obtained by substituting the influence coefficient at the age of 7, 14, 21, and
28 d along with IC w/c ratio (wic /c) into Eq. (2.25).
At 7 d after casting:

f (wic /c, 7) = 1.075 × (wic /c)−0.470 R2 = 0.978 (2.26)

At 14 d after casting:

f (wic /c, 14) = 1.041 × (wic /c)−0.461 R2 = 0.991 (2.27)

At 21 d after casting:

f (wic /c, 21) = 0.839 × (wic /c)−0.490 R2 = 0.971 (2.28)

At 28 d after casting:

f (wic /c, 28) = 0.868 × (wic /c)−0.443 R2 = 0.960 (2.29)

The test results and regression analysis results of influence coefficient are shown
in Fig. 2.16. The deviations of influence coefficient between results from Eq. (2.25)
and test were small. The IC w/c ratios used were between 0 and 0.05, which denoted
that the amounts of IC water were between 0 and 81.5% of the theoretical quantities
of IC water provided by pre-wetted LWAs. The proposed model for calculating the

@seismicisolation
@seismicisolation
2.4 Internal Relative Humidity of Early-Age Concrete Internally Cured … 73

Fig. 2.18 Relationship

between parameter Yt and
age of specimens

influence coefficient of IC w/c ratio was based on the test results and was valid for
w/c ratios and pre-wetted LWAs tested or similar conditions.
As shown in Fig. 2.18 and Eqs. (2.26)–(2.29), the parameter Yt was 1.075, 1.041,
0.839, and 0.868 for concrete at the age of 7, 14, 21, and 28 d after casting, respec-
tively. The value of parameter Yt changed with age of concrete, which was in accor-
dance with results in [25, 43]. As shown in Fig. 2.18, the relationship between Yt
and age of concrete t could be expressed as follows.

Yt = ϕ × t + λ (2.30)

where ϕ and λ are the parameters that can be determined with regression analysis,
and the results were ϕ = −0.0115, λ = 1.156, and R2 = 0.865, respectively.
As shown in Eqs. (2.26)–(2.29), the parameter pt was −0.470, −0.461, −0.490,
and −0.443 for concrete at the age of 7, 14, 21, and 28 d after casting, respectively.
The average value was −0.466. The deviations between the individual value and
average value of parameter pt were 0.8, 1.1, 4.9, and 5.2%.
1. Prediction model for critical time of IRH
Figure 2.19a shows that the critical time was 1.0, 1.4, 2.5, 2.0, 4.3, and 6.2 d when
the total w/c ratios were 0.33, 0.40, 0.50, 0.34, 0.36, and 0.38 for Mixtures RC33-
0, RC40-0, RC50-0, IC33-10, IC33-30, and IC33-50, respectively. The critical time
increased by 1.0, 3.3, and 5.2 d when the IC w/c ratios increased from 0 to 0.01, 0.03,
and 0.05, respectively. Therefore, the critical time of IRH in ICC with pre-wetted
LWAs increased with the increase of IC w/c ratio.
Figure 2.19a shows that there was not obvious relationship between the test results
of critical time and total w/c ratio. Therefore the total w/c ratio could not be used
to obatin the critical time Tcri,2 in Eq. (2.23). The working w/c ratio was influenced
by age of concrete, as shown in Eqs. (2.24)–(2.30). The effect of age on the ww /c
was not considered in proposing the model for the critical time Tcri,2 . The average
value of parameter Yt for the age of 7, 14, 21, and 28 d, which was 0.956, was used

@seismicisolation
@seismicisolation
74 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Fig. 2.19 Relationship

between w/c ratio and
critical time

to obtain the working w/c ratio for the critical time Tcri,2 , ww,T /c. Substituting the
average values of parameter Yt and pt into Eq. (2.24) yields ww,T /c. Figure 2.19b
shows the relationship between critical time Tcri,2 and working w/c ratio for critical
time (ww,T /c) was polynomial, as shown in Eq. (2.31). Therefore the relationship
between critical time Tcri,2 and working w/c ratio for critical time (ww,T /c) could be
expressed as follows.

Tcri,2 = α2 × (ww,T /c)2 +β2 × (ww,T /c) + γ2 (2.31)

where

ww,T /c = we /c + 0.956 × (wic /c)0.534 (2.32)

@seismicisolation
@seismicisolation
2.4 Internal Relative Humidity of Early-Age Concrete Internally Cured … 75

where α2 , β2 , and γ2 are the parameters that can be determined with regression
analysis, and the resultant results are α2 = 119.22, β2 = −79.41, and γ2 = 14.22,
respectively.
The values of critical time Tcri,2 from Eq. (2.31) were 1.1, 1.5, 1.7, 3.5, 4.3, and
5.3 d when the working w/c ratios for critical time ww,T /c were 0.33, 0.40, 0.41,
0.48, 0.50, and 0.52 for Mixtures RC33-0, RC40-0, IC33-10, IC33-30, RC50-0, and
IC33-50, respectively. The deviations of the values between results from models and
test are used to evaluate the accuracy of model. The deviations of values of critical
time between results from Eq. (2.31) and the test were 0.1, 0.1, 0.3, 0.8, 1.8, and 0.9
d, respectively.
2. Prediction model for working w/c ratio
Substituting Eqs. (2.30) and pt into Eq. (2.25), and then substituting Eq. (2.25) into
Eq. (2.24), the relationship between ww /c and age of concrete is shown in Fig. 2.15
and could be expressed as follows.

ww /c = we /c + (ϕ × t + λ) × (wic /c)0.534 (2.33)

/
For Mixture IC33-10, the values of ww c obtained from test results or Eq. (2.33) were
0.414, 0.423, 0.418, 0.411, and 0.397 or 0.426, 0.422, 0.415, 0.408, and 0.401 at the
ages of 3, 7, 14, 21, and 28 d after casting, respectively. The deviations of working w/c
ratio between results from Eq. (2.33) and the test were 2.9%, 0.2%, 0.7%, 0.7%, and
1.0%, respectively. Therefore, for Mixture IC33-10, the average value of deviations
was 1.1%. The average values of deviations were 1.7% and 2.2% for Mixtures IC33-
30 and IC33-50, respectively. The deviations were small. Therefore, Eq. (2.33) could
be used to calculate the working w/c ratio in consideration of both the effective and
IC w/c ratios. This model could be used to predict the critical time of IRH and IRH
variations in early-age concrete internally cured with pre-wetted LWAs.

2.4.4 Prediction Model for Internal Relative Humidity

of Concrete Internally Cured with LWAs

Substituting Eqs. (2.21) and (2.22) into Eq. (2.23) yields Eq. (2.34). Substituting
the values of parameters ϕ and λ into Eq. (2.33) yields Eq. (2.35). Substituting the
values of parameters α2 , β2 , and γ2 into Eq. (2.31) yields Eq. (2.36). Therefore, the
relationship between IRH and age of concrete could be expressed as follows.

t
εicas (t) = − · (7 + 3.73x) (t ≤ t0 ; t0 = 0.35 d) (2.34)
to

where

ww /c = we /c + (−0.0115t + 1.156) × (wic /c)0.534 (2.35)

@seismicisolation
@seismicisolation
76 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Fig. 2.20 Relationship 100

between IRH and age of ICC
96
92

IRH (%)
88 Test data of RC33-0
Test data of IC33-10
84 Test data of IC33-30
Test data of IC33-50
80 Fitting curve of RC33-0
Fitting curve of IC33-10
76 Fitting curve of IC33-30
Fitting curve of IC33-50
72
0 5 10 15 20 25 30
Age of specimens (d)

and
( )2
Tcri,2 = 119.22 × (we /c + 0.956 × (wic /c)0.534
( ) (2.36)
−79.41 × (we /c + 0.956 × (wic /c)0.534 + 14.22

where RH(ww /c, t) is IRH in ICC in consideration of ww /c and critical time, in %.

Figure 2.20 shows the deviations of IRH in concrete between results from Eqs.
(2.34)–(2.36) and test.
For Mixture IC33-10, the IRH obtained from test results or Eqs. (2.34)–(2.36) was
98.7%, 95.7%, 93.0%, 91.1%, and 89.2% RH or 99.0%, 95.8%, 92.9%, 91.0%, and
89.4% RH at the ages of 3, 7, 14, 21, and 28 d after casting, respectively. The devia-
tions of IRH between results from Eqs. (2.34)–(2.36) and the test were 0.3%, 0.1%,
0.1%, 0.1%, and 0.2% RH, respectively. Therefore for Mixture IC33-10, the average
value of deviations was 0.2% RH. The average values of deviations were 0.4%, and
0.4% RH for Mixtures IC33-30, and IC33-50, respectively. Equations (2.34)–(2.36)
could be used to predict the IRH in early-age concrete that was internally cured
with pre-wetted LWAs in consideration of working w/c ratio, critical time, and age.
Furthermore, autogenous shrinkage of early-age concrete could also be predicted
based on IRH from the proposed Eqs. (2.34)–(2.36) for better understanding cracking
resistance of early-age concrete that was internally cured with pre-wetted LWAs.

2.4.5 Summary

The prediction models for working w/c ratio, the critical time, and the IRH variations
in early-age ICC with pre-wetted LWAs were investigated. The following conclusions
can be drawn based on the obtained results.

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 77

(1) The IRH in ICC with pre-wetted LWAs at 28 d after casting increased with the
increase of IC w/c ratio. The average value of IRH decrease rate DR in ICC with
pre-wetted LWAs decreased nonlinearly with the increase of IC w/c ratio.
(2) The critical time of IRH in ICC with pre-wetted LWAs increased with the
increase of IC w/c ratio. A prediction model for the critical time of IRH in
ICC with pre-wetted LWAs was proposed in consideration of the working w/c
ratio.
(3) A prediction model for the working w/c ratio was proposed in consideration of
effective and IC w/c ratio. This model could be used to predict the critical time
of IRH and IRH variations in early-age ICC with pre-wetted LWAs.
(4) In consideration of working w/c ratio, critical time, and age, a prediction model
for the IRH in early-age ICC with pre-wetted LWAs was proposed.

2.5 Influential Factors of the Internal Relative Humidity

of Early-Age Concrete

2.5.1 Influence of Curing Humidity

1. Mixture proportion
For measuring the variations of IRH in early-age concrete under different curing
humidity, the sealed and unsealed concrete specimens were prepared. The sealed
concrete specimens were designated as RCT2S, and the unsealed concrete specimens
were designated as RCT2U-25, RCT2U-50, and RCT2U-75, which represented that
the curing humidity of concrete was 25%, 50%, and 75% RH, respectively. The
curing temperature was 20 °C for all concrete specimens. The mixture proportions
of concrete are shown in Table 2.3.
2. Change of internal relative humidity
The IRH in HSC under the condition of self-desiccation alone and the combined
effects of self-desiccation and moisture diffusion under different curing humidity
conditions is shown in Fig. 2.21.
The critical time was different for the sealed and unsealed HSC. Meanwhile,
the critical time of IRH in unsealed HSC affected by self-desiccation and moisture
diffusion was also different under different curing humidity conditions. The critical
time increased as the curing humidity increased. The critical time increased from
2 d to 3, and 5 d when the curing humidity increased from 25% to 50%, and 75%
RH for specimens RCT2U-25, RCT2U-50, and RCT2U-75, respectively. The crit-
ical time was 9 d for specimen RCT2S, which was longer than that for unsealed

Table 2.3 Mixture proportions of concrete (kg/m3 )

Water Cement Fine aggregate Coarse aggregate Superplasticizer
174 512 636 1131 2.82

@seismicisolation
@seismicisolation
78 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Fig. 2.21 Experimental 100

results of relationship
between IRH and age for 96
concrete specimens under
different curing humidity 92

IRH (%)
conditions
88
RCT2S
84 RCT2U-25
RCT2U-50
80 RCT2U-75

0 4 8 12 16 20 24 28
Age of specimens (d)

HSC. The IRH in HSC was the RH of the pore in which the gas and the interstitial
liquid phases were kept in equilibrium. Most of the pores in concrete were full of
water and showed continuous liquid phase in the initial stage after the pouring of
concrete. At this stage, the liquid water flow was the main factor affecting the loss
of water, and the IRH in HSC was equal to 100% RH. The water content in concrete
decreased gradually with the continuous cement hydration and moisture diffusion
or evaporation to the environment. When the water content in the pores of concrete
decreased to a critical value, the vapor pressure became lower than the saturation
value, and the IRH decreased from 100% RH. From this moment on, the develop-
ment of IRH entered the stage II in which IRH decreased with the growth of age,
as reported in [44]. The curing humidity had an effect on the rate of moisture diffu-
sion. The difference between the IRH in HSC and the curing humidity decreased as
the curing humidity increased at early age, which decreased the moisture gradient.
Therefore, the moisture diffusion decreased and the critical time increased as the
curing humidity increased for unsealed HSC.
Figure 2.21 shows that the IRH in sealed HSC for specimen RCT2S was 100%,
98.30%, 96.81%, and 95.67% RH at the age of 7, 14, 21, and 28 d, respectively.
The internal water was consumed by cement hydration, and the IRH in sealed HSC
decreased constantly with the growth of age. Figure 2.21 also shows that the IRH
in unsealed HSC decreased as the age increased under the same curing humidity.
Meanwhile, the IRH in HSC increased as the curing humidity increased. The IRH
in unsealed HSC was affected by moisture diffusion and self-desiccation, while the
curing humidity had little effect on the self-desiccation of concrete. The increasing
curing humidity decreased the difference between environmental humidity and the
IRH in HSC. The moisture diffusion decreased as the difference between environ-
mental humidity and the IRH in HSC decreased. Therefore, the IRH in unsealed
HSC decreased slowly and the IRH increased as the curing humidity increased.
Figure 2.22 shows the comparison of the IRH in unsealed HSC at the age of 3, 7,
14, 21, and 28 d under different curing humidity conditions.

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 79

Fig. 2.22 Comparison of IRH in the condition of 25% curing humidity

the IRH in unsealed concrete Increased IRH
specimens at different ages 3d 7d 14 d 21 d 28 d
100
under different curing
humidity conditions
80

IRH (%)
60

40

20
25 50 75 25 50 75 25 50 75 25 50 75 25 50 75
Curing humidity (%)

The IRH in HSC under the condition of 50% RH curing humidity for specimen
RCT2U-50 was 0.61%, 3.07%, 4.44%, 5.11%, and 5.52% RH higher than that under
the condition of 25% RH curing humidity for specimen RCT2U-25 at the age of 3, 7,
14, 21, and 28 d, respectively. The IRH in HSC under the condition of 75% RH curing
humidity for specimen RCT2U-75 was 0.61%, 5.49%, 9.18%, 11.29%, and 11.66%
RH higher than that under the condition of 25% RH curing humidity for specimen
RCT2U-25 at the age of 3, 7, 14, 21, and 28 d, respectively. The experimental results
demonstrated that the increase of IRH in unsealed HSC caused by the increasing
curing humidity gradually increased with the growth of age, which indicated that the
effect of curing humidity on the IRH in HSC was a cumulative effect with the growth
of age. The increase of cumulative effect range of curing humidity on the increase of
IRH decreased with the growth of age, which indicated that the difference between
the environmental curing humidity and the IRH in unsealed HSC decreased with the
growth of age. Therefore, the effect of curing humidity on the IRH in unsealed HSC
gradually decreased with the growth of age.

3. Internal relative humidity decrease rate in concrete

The relationship between IRH decrease rate and age for HSC under different curing
humidity conditions is shown in Fig. 2.23.
The IRH decrease rate of sealed HSC decreased as the age increased, and the IRH
began to decrease at the age of 10 d. The average value of IRH decrease rate of sealed
HSC for specimens RCT2S was 0.290%, 0.290%, 0.205%, 0.157%, and 0.156%
RH/d at the age of 13, 16, 20, 23, and 28 d, respectively. Similarly, the IRH decrease
rate of unsealed HSC also decreased as the age increased under the conditions of 25%,
50%, and 75% RH curing humidity. The IRH in unsealed HSC under the condition
of 25% RH curing humidity for specimen RCT2U-25 began to decrease at the age
of 3 d, and the average value of IRH decrease rate was 1.328%, 0.909%, 0.620%,
0.380%, and 0.240% RH/d at the age of 8, 15, 20, 23, and 28 d, respectively. The
IRH in unsealed HSC under the condition of 50% RH curing humidity for specimen

@seismicisolation
@seismicisolation
80 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Fig. 2.23 Experimental 1.4

results of relationship RCT2S
1.2 RCT2U-25
between IRH decrease rate
RCT2U-50

Decrease rate (% RH/d)

and age for concrete RCT2U-75
specimens under different 1.0
curing humidity conditions
0.8

0.6

0.4

0.2

0.0
0 4 8 12 16 20 24 28
Age of specimens (d)

RCT2U-50 began to decrease at the age of 4 d, and the average value of IRH decrease
rate was 0.850%, 0.779%, 0.500%, 0.283%, and 0.200% RH/d at the age of 8, 15,
20, 23, and 28 d, respectively. The IRH in unsealed HSC under the condition of 75%
RH curing humidity for specimen RCT2U-75 began to decrease at the age of 6 d,
and the average value of IRH decrease rate was 0.420%, 0.405%, 0.390%, 0.255%,
and 0.210% RH/d at the age of 9, 13, 17, 23, and 28 d, respectively. Meanwhile,
experimental results generally demonstrated that the IRH decrease rate of unsealed
HSC decreased as curing humidity increased at the same curing age.
The IRH decrease rate of unsealed HSC was relatively higher than that of sealed
HSC, as shown in Fig. 2.23. The difference of the IRH decrease rate between the
sealed and unsealed HSC under the conditions of 25, 50, and 75% RH curing humidity
gradually decreased with the growth of age, which indicated that the effect of moisture
diffusion on the IRH decreased as the age increased. The change of IRH decrease
rate was due to the change of capillary forces, as reported in [28]. At the beginning
of drying, the moisture contents are high, and the free liquid water present in the
pores of concrete moves at a fairly high rate due to capillary forces and pressure
gradients, which is faster than that of bound water and vapour in the late stage of
drying [28]. Meanwhile, the increasing curing humidity decreased the difference
between environmental humidity and the IRH in unsealed HSC, which decreased
the moisture gradient, thereby decreasing the IRH decrease rate.

4. Moisture diffusion in concrete at early age

The decrease of IRH in unsealed concrete was mainly caused by self-desiccation

due to cement hydration and external moisture diffusion. Therefore, the IRH in
unsealed HSC under the condition of moisture diffusion alone could be calculated
by Eq. (2.37), as reported in [36].

He = 100% − (Hs − H ) (2.37)

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 81

Fig. 2.24 Experimental 100

results of relationship
between IRH and age for
concrete specimens under 96
the condition of moisture

IRH (%)
diffusion alone
92

88
RCT2U-25
RCT2U-50
84 RCT2U-75

0 4 8 12 16 20 24 28
Age of specimens (d)

where He is the IRH in HSC under the condition of moisture diffusion alone, in %;
Hs is the IRH in HSC under the condition of self-desiccation alone, in %; and H is the
IRH in HSC under the combined effects of self-desiccation and moisture diffusion,
in %.
Figure 2.24 shows the variations of IRH in HSC under the condition of moisture
diffusion alone.
The IRH in HSC at the age of 28 d after pouring increased by 5.52% and 11.66%
RH when the curing humidity increased from 25 to 50%, and 75% RH for speci-
mens RCT2U-25, RCT2U-50, and RCT2U-75 under the condition of moisture diffu-
sion alone, respectively. The IRH in HSC under the condition of moisture diffusion
decreased fast during the first 20 d and then decreased at a slow rate in the following
days for specimens RCT2U-25 and RCT2U-50. The decreases of IRH in HSC at the
age of 28 d after pouring were 14.72%, 9.20%, and 3.06% RH or 19.05%, 13.53%,
and 7.39% RH when the curing humidity increased from 25 to 50%, and 75% RH for
specimens RCT2U-25, RCT2U-50, and RCT2U-75 under the condition of moisture
diffusion alone or the combined effects of self-desiccation and moisture diffusion,
respectively. The ratios of the decrease of IRH in HSC at the age of 28 d after
pouring under the condition of moisture diffusion alone to that under the combined
effects of self-desiccation and moisture diffusion were 77.3%, 68.0%, and 41.4%
when the curing humidity were 25%, 50%, and 75% RH for specimens RCT2U-25,
RCT2U-50, and RCT2U-75, respectively. The effect of moisture diffusion on the
decrease of IRH in unsealed HSC decreased as the curing humidity increased, and
the decrease of IRH in unsealed HSC was mainly caused by self-desiccation when
the curing humidity was 75% RH. A possible reason for IRH decreasing slowly
for specimens RCT2U-50 and RCT2U-75 after the first 20 d after pouring was that
the cement hydration not only made the microstructure of concrete dense, but also
caused changes of porosity in pastes at early age, as reported in [34].
The calculation of moisture diffusion coefficient is very important for predicting
the IRH in concrete [44]. Experimental results reported in [35] demonstrate that the
effect of temperature variation on the IRH in concrete is relatively small and can be

@seismicisolation
@seismicisolation
82 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

ignored. Meanwhile, the curing temperature was 20 °C for all concrete specimens.
Therefore, Eq. (2.38) could be utilized as the fundamental equation under the condi-
tion of one-dimensional moisture diffusion based on Fick’s second law, as reported
in [37].
( )
∂ρ ∂ ∂ρ
= D (2.38)
∂τ ∂x ∂x

where ρ is the water content in HSC in this section; D is the moisture diffusion
coefficient of HSC, in m2 /s; τ is the diffusion time, in s; and x is the depth of a
position within the concrete from the surface under the condition of one-dimensional
moisture diffusion, in m.
The various phase of water (vapor, capillary water, and absorbed water) in the
pores of concrete can be considered to be in thermodynamic equilibrium at any time
due to the slow cement hydration and moisture diffusion [45]. Therefore, the desorp-
tion or sorption isotherms could be utilized to describe the relationship between the
IRH H and water content ρ, as reported in [45–47], and Eq. (2.39) could be obtained.

∂(H − Hs ) ∂ ∂(H − Hs )
= (D ) (2.39)
∂τ ∂x ∂x
Equation (2.40) could be obtained by substituting Eq. (2.37) into Eq. (2.39).

∂He ∂ ∂He
= (D ) (2.40)
∂τ ∂x ∂x
The moisture diffusion coefficient could be calculated by introducing the
parameter λ through the boltzmann transformation, as given in Eq. (2.41).
x
−λ = √ (2.41)
2 τ

Then, Eq. (2.42) could be obtained as follows.

( )
∂ ∂He
−2λ = D (2.42)
∂He ∂λ

Equation (2.43) could be obtained by integrating Eq. (2.42) from He to 100% RH.

∫
100% | |
∂He || ∂He ||
−2λdHe = D(He ) − D(He ) (2.43)
∂λ |100% ∂λ |He
He

At the beginning of moisture diffusion (He ≈ 100% RH), ∂He /∂λ ≈ 0, and
Eq. (2.43) could be simplified to Eq. (2.44).

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 83
⎛ 100% ⎞ (
∫ | )
∂H |
D(He ) = −2⎝ λdHe ⎠/
e | (2.44)
∂λ |He
He

The values of He under different curing humidity conditions were obtained from
the experimental results, and the parameter λ could be obtained from Eq. (2.41).
Therefore, the relationship between He and λ under different curing humidity
conditions was fitted by Eq. (2.45).
( )
He = 100 × 1 − 2aλ+b (2.45)

When the curing temperature or humidity was 20 °C or 25% RH:

( )
He = 100 × 1 − 2−1.93688λ−1.28332 , R2 = 0.980 (2.46)

When the curing temperature or humidity was 20 °C or 50% RH:

( )
He = 100 × 1 − 2−2.30198λ−1.66774 , R2 = 0.971 (2.47)

When the curing temperature or humidity was 20 °C or 75% RH:

( )
He = 100 × 1 − 2−2.71612λ−3.03040 , R2 = 0.967 (2.48)

Figure 2.25 shows the relationship between the He and λ for unsealed concrete
specimens under different curing humidity conditions.
The moisture diffusion coefficient under different curing humidity conditions
could be calculated with a given He by substituting Eqs. (2.46), (2.47), and (2.48) into
Eq. (2.44). Figure 2.26 shows the relationship between moisture diffusion coefficient
and IRH for unsealed concrete specimens under different curing humidity conditions,

Fig. 2.25 Relationship 100

between the He and λ for
unsealed concrete specimens
under different curing 96
humidity conditions
He (%)

92

88 Test data of RCT2U-25

Test data of RCT2U-50
Test data of RCT2U-75
84 Fitting curve of RCT2U-25
Fitting curve of RCT2U-50
Fitting curve of RCT2U-75
80
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
λ

@seismicisolation
@seismicisolation
84 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Moisture diffusion coefficient ( 10-9m2/s)

Fig. 2.26 Relationship 16
between moisture diffusion
14 RCT2U-25
coefficient and IRH for
RCT2U-50
unsealed concrete specimens RCT2U-75
12
under different curing
humidity conditions 10
8
6
4
2
0
84 88 92 96 100
IRH (%)

which demonstrated that the moisture diffusion coefficient was greatly affected by
the water content in HSC.
The moisture diffusion coefficients were (2.60~14.93) × 10–9 , (2.03~10.33) ×
10 , and (1.56~6.80) × 10–9 m2 /s when the IRH in HSC ranged from 85.28%,
–9

90.80%, and 96.94% to 100% RH for specimens RCT2U-25, RCT2U-50, and

RCT2U-75, respectively. Meanwhile, the variations of IRH with age could be
obtained from the experimental results. Therefore, the relationship between moisture
diffusion coefficient and age for unsealed concrete specimens under different curing
humidity conditions could be obtained, as shown in Fig. 2.27.
The moisture diffusion coefficient gradually decreased as the age increased
when the curing humidity was 25%, 50%, and 75% RH for specimens RCT2U-
25, RCT2U-50, and RCT2U-75, respectively. The moisture diffusion coefficient
Moisture diffusion coefficient ( 10-9m2/s)

Fig. 2.27 Relationship 7

between moisture diffusion Test data of RCT2U-25
coefficient and age for 6 Test data of RCT2U-50
unsealed concrete specimens Test data of RCT2U-75
under different curing Fitting curve of RCT2U-25
humidity conditions 5
Fitting curve of RCT2U-50
Fitting curve of RCT2U-75
4

1
0 4 8 12 16 20 24 28
Age of specimens (d)

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 85

decreased rapidly in the early stage and decreased slowly in the later stage, which
was mainly because the moisture gradient in concrete gradually changed with the
progress of cement hydration. In three stages, the factors dominating the moisture
transport are liquid flow, liquid flow and vapor diffusion, and pure vapor diffusion,
respectively [44]. The variation of moisture diffusion depends on the degree of conti-
nuity in the liquid and vapor phases. The moisture diffusion coefficient varying with
age could be calculated by Eq. (2.49).
a
D(t) = +b (2.49)
t

where D(t) is the moisture diffusion coefficient at the age of t d of specimen, in m2 /s;
and a, b are the parameters determined by fitting the experimental results of moisture
diffusion coefficient and age. The values of parameters a, b, and correlations R2 are
shown in Table 2.4.

5. Prediction model for internal relative humidity in concrete

The decrease of IRH in sealed HSC was mainly due to cement hydration. Therefore,
the variations of IRH was analyzed according to the variations of degree of cement
hydration. Experimental results demonstrated that the IRH in HSC decreased with
the growth of age. The IRH in HSC under the condition of self-desiccation alone
could be calculated by Eq. (2.50) based on degree of cement hydration, as reported
in [35, 48].
⎧
⎨ 100% for α ≤ αc
⎪
( )
Hs = ( ) α − αc β (2.50)
⎪
⎩ 100% + Hs,u − 100% for α > αc
αu − αc

where Hs,u is the IRH in HSC under the condition of self-desiccation alone at the
maximum degree of cement hydration (take the IRH at the age of 28 d in this section),
in %; α is the degree of cement hydration; αc is the critical degree of cement hydration;
αu is the maximum degree of cement hydration (take the degree of cement hydration
at the age of 28 d in this section); and β is the fitting parameter.
The degree of cement hydration which was related to age and temperature could
be calculated by Eq. (2.51), as reported in [49].

Table 2.4 The values of

Concrete specimen a b R2
parameters a, b, and
correlations R2 in Eq. (2.49) RCT2U-25 12.909 2.023 0.986
RCT2U-50 12.348 1.452 0.955
RCT2U-75 9.345 1.191 0.950

@seismicisolation
@seismicisolation
86 2 Internal Relative Humidity of Early-Age Internally Cured Concrete
[ ( ) ]
A B
α = αu exp − (2.51)
t

where t is the age of concrete, in d; and A, B are the fitting parameters, respectively.
The maximum degree of cement hydration with the w/c ratio could be calculated
by Eq. (2.52), as reported in [50].

1.031(w/ c)
αu = (2.52)
0.194 + w/ c

The maximum degree of cement hydration determined by Eqs. (2.52) and (2.51)
were substituted into Eq. (2.50). Meanwhile, the IRH in the sealed concrete obtained
from the experimental results was substituted into Eq. (2.50), and the experimental
results of parameters A, B, and β were 49.88, 0.42, and 0.90, respectively.
The degree of cement hydration over time obtained from Eq. (2.51) was substituted
into Eq. (2.50) to obtain the IRH in the sealed HSC. Figure 2.28 shows the comparison
of IRH in sealed HSC between results obtained from Eq. (2.50) and experiment. The
IRH in sealed HSC at the age of 10, 14, 21, and 27 d under the condition of self-
desiccation alone was 99.48%, 98.30%, 96.81%, and 95.79% RH, respectively, and
the corresponding results calculated by Eq. (2.50) were 99.56%, 98.30%, 96.77%,
and 95.81% RH, respectively. The absolute values of deviations between the results
of IRH in sealed HSC obtained from Eq. (2.50) and experiment were within 0.08%
RH, which were relatively small. Therefore, the decrease of IRH in sealed HSC due
to cement hydration could be predicted by Eq. (2.50).
The boundary condition of IRH node could not be directly defined as environ-
mental humidity when the boundary condition of IRH node was proposed, as reported
in [51]. When it was assumed that the initial IRH in concrete was 100% RH and the
environmental humidity was constant Hw , then the initial conditions of nodes could
be taken as H (x, 0) = 100% RH, and boundary conditions of nodes could be taken
as H (0, t) = γ(Hw ) · Hw and H (∞, t) = 100% RH, as reported in [51, 52].

Fig. 2.28 Comparison of

100
IRH in sealed concrete
specimens between results
obtained from Eq. (2.50) and
experiment under the 98
IRH (%)

condition of self-desiccation
alone

96

Test data of RCT2S

Predicting results of RCT2S
94
0 4 8 12 16 20 24 28
Age of specimens (d)

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 87

The calculation of IRH in HSC could be simplified according to the initial and
boundary conditions of nodes when the moisture diffusion coefficient was consid-
ered to be a constant, as reported in [52]. Meanwhile, the effect coefficient γ(Hw ) of
environmental humidity was introduced to improve the accuracy of node boundary
conditions. Then, the IRH in concrete at any time and location under the condition
of moisture diffusion alone could be calculated by Eq. (2.53).

He = γ(Hw ) · Hw − (γ(Hw ) · Hw − 100%) · erf (ϕ) (2.53)

where γ(Hw ) is the effect coefficient of environmental humidity of the node, which is
a function related to the environmental humidity Hw ; and erf (ϕ) is the error function
and can be calculated by Eq. (2.54).
∫ ϕ
2 ( )
erf (ϕ) = √ exp −ϕ 2 d ϕ (2.54)
π 0

where
x
ϕ= √ (2.55)
2 D(t) · t

The value of the error function erf (ϕ) could be obtained by approximate calcu-
lation with hyperbolic tangent function, as reported in [52, 53] and given in
Eq. (2.56).
( )
erf (ϕ)= tanh × 1.12838ϕ + 0.10277ϕ 3 (2.56)

In order to obtain the variations of IRH under the condition of moisture diffusion
alone, a reasonable constant moisture diffusion coefficient was needed. Therefore,
the concept of equivalent moisture diffusion coefficient of concrete is proposed [52],
as given in Eq. (2.57).
∫ t
1
D= D(t)dt (2.57)
t Tcri

where D is the equivalent moisture diffusion coefficient, which can be substituted

into Eq. (2.55) as a constant for calculation, in m2 /s; and Tcri is the age of concrete
when IRH begins to decrease.
The calculated equivalent moisture diffusion coefficient from 3 d to 28 d was
3.185 × 10–9 m2 /s when the curing humidity was 25% RH for specimen RCT2U-25;
the equivalent moisture diffusion coefficient from 4 d to 28 d was 2.453 × 10–9 m2 /s
when the curing humidity was 50% RH for specimen RCT2U-50; and the equivalent
moisture diffusion coefficient from 6 d to 28 d was 1.845 × 10–9 m2 /s when the
curing humidity was 75% RH for specimen RCT2U-75. Therefore, the relationship
between equivalent moisture diffusion coefficient and curing humidity for unsealed
concrete specimens could be expressed by Eq. (2.58).

@seismicisolation
@seismicisolation
88 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Fig. 2.29 Relationship

between equivalent moisture
diffusion coefficient and
curing humidity for unsealed
concrete specimens

D = m · Hwn (2.58)

where m and n are the parameters determined by fitting the experimental results of
equivalent moisture diffusion coefficient and curing humidity.
Figure 2.29 shows the relationship between equivalent moisture diffusion coef-
ficient and curing humidity for unsealed concrete specimens, which demonstrated
that the equivalent moisture diffusion coefficient of HSC decreased as the curing
humidity increased.
Then, the equivalent moisture diffusion coefficient D under different curing
humidity conditions was substituted into Eq. (2.53), and the effect coefficient γ (Hw )
of environmental humidity was calculated as 2.875, 1.600, and 1.235 when the curing
humidity was 25%, 50%, and 75% RH for specimens RCT2U-25, RCT2U-50, and
RCT2U-75, respectively. Therefore, the relationship between effect coefficient γ(Hw )
of the environmental humidity and curing humidity could be expressed by Eq. (2.59).

γ (Hw ) = h · Hwk (2.59)

where h and k are the parameters determined by fitting the experimental results
of the effect coefficient γ (Hw ) of environmental humidity and curing humidity,
respectively.
Figure 2.30 shows the relationship between effect coefficient of environmental
humidity and curing humidity for unsealed concrete specimens. The effect coefficient
of environmental humidity decreased as the curing humidity increased.
The relationship between IRH in HSC and age considering the effect of curing
humidity under the condition of moisture diffusion alone could be expressed by
substituting Eqs. (2.57), (2.58), and (2.59) into Eq. (2.53), as given in Eq. (2.60).

75
He = γ (Hw ) · Hw − (γ (Hw ) · Hw − 100%) · erf ( √ ) (2.60)
2 D·t

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 89

Fig. 2.30 Relationship

between the effect coefficient
γ (Hw ) of environmental
humidity and curing
humidity for unsealed
concrete specimens

D = 14.257Hw−0.463 (2.61)

γ (Hw ) = 37.312Hw−0.797 (2.62)

The comparison of IRH in unsealed HSC between the experimental results

obtained from Eq. (2.60) and experiment under the condition of moisture diffusion
alone is shown in Fig. 2.31.
The experimental results of IRH at the age of 3, 7, 14, 21, and 28 d were 99.39%,
93.86%, 88.84%, 86.06%, and 85.28% RH when the curing humidity was 25% RH
for specimen RCT2U-25, and the calculation results of the prediction model were
98.15%, 93.53%, 88.83%, 86.22%, and 84.52% RH, respectively. The experimental
results of IRH in HSC at the age of 5, 7, 14, 21, and 28 d were 98.64%, 96.93%,

Fig. 2.31 Comparison of 100

IRH in unsealed concrete
specimens between results 96
obtained from Eq. (2.60) and
experiment under the
condition of moisture 92
IRH (%)

diffusion alone
88
Test data of RCT2U-25
84 Test data of RCT2U-50
Test data of RCT2U-75
Predictting results of RCT2U-25
80
Predictting results of RCT2U-50
Predictting results of RCT2U-75
76
0 4 8 12 16 20 24 28
Age of Specimens (d)

@seismicisolation
@seismicisolation
90 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

93.28%, 91.17%, and 90.80% RH when the curing humidity was 50% RH for spec-
imen RCT2U-50, and the calculation results of the prediction model were 98.36%,
97.25%, 94.46%, 92.77%, and 91.63% RH, respectively. The experimental results
of IRH in HSC at the age of 7, 14, 21, and 28 d were 99.35%, 98.02%, 97.35%,
and 96.94% RH when curing humidity was 75% RH for specimen RCT2U-75, and
the calculation results of the prediction model were 98.75%, 97.18%, 96.16%, and
95.46% RH, respectively. The comparison demonstrated that the deviations between
the results of IRH in unsealed HSC obtained from experiment and Eq. (2.60) were
very small.
The IRH in HSC under the condition of self-desiccation alone could be calculated
by Eq. (2.50), and the IRH in HSC under the condition of moisture diffusion alone
could be calculated by Eq. (2.60). The IRH in early-age HSC was affected by self-
desiccation and moisture diffusion under the condition of unsealing. Therefore, the
IRH in unsealed HSC as the function of age and curing humidity could be calculated
by Eq. (2.63).
⎧
⎪
⎪ H = He + Hs − 100%
⎪
⎪ ⎧
⎪
⎪
⎪
⎪
⎪ ⎨ 100% for α ≤ 0.0843
⎪
⎪
⎪ Hs = ( )
⎪
⎪ ⎪ α − 0.0843 0.90
⎪
⎪ ⎩ 100% − 4.33% for α > 0.0843
⎪
⎪ 0.0992
⎪
⎨ ( ) [( ) ]
He = 37.312Hw−0.797 · Hw − 37.312Hw−0.797 · Hw − 100% · (2.63)
⎪
⎪
⎪
⎪ 75
⎪
⎪ erf ( /( ) )
⎪
⎪ −0.463
⎪
⎪ 2 14.257H w ·t
⎪
⎪ [ (
⎪
⎪ )0.42 ]
⎪
⎪ 49.88
⎪
⎩ α= 0.6564 exp −
⎪
t

Therefore, the IRH in unsealed HSC could be calculated by this prediction model
at early age. The comparison of IRH in unsealed HSC between results obtained from
Eq. (2.63) and experiment is shown in Fig. 2.32.
The experimental results of IRH were 80.95%, 86.47%, and 92.61% RH and the
calculation results of the prediction model were 80.19%, 87.30%, and 91.13% RH at
the age of 28 d when the curing humidity was 25%, 50%, and 75% RH for specimens
RCT2U-25, RCT2U-50, and RCT2U-75, respectively. The absolute values of devi-
ations between the results of IRH in unsealed HSC obtained from experiment and
Eq. (2.63) were 0.76%, 0.83%, and 1.48% RH when the curing humidity was 25%,
50%, and 75% RH for specimens RCT2U-25, RCT2U-50, and RCT2U-75, respec-
tively, which were relatively small. Therefore, the prediction model could well predict
the variations of IRH in HSC at early age under different curing humidity conditions.
Additionally, autogenous shrinkage of HSC could also be predicted based on the
results of IRH obtained from the proposed Eq. (2.63) for better understanding the
early-age cracking resistance of HSC that was cured under different curing humidity
conditions.

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 91

Fig. 2.32 Comparison of 100

IRH in unsealed concrete
specimens between results 96
obtained from Eq. (2.63) and
experiment 92

IRH (%)
88

84 Test data of RCT2U-25

Test data of RCT2U-50
80 Test data of RCT2U-75
Predicting results of RCT2U-25
76 Predicting results of RCT2U-50
Predicting results of RCT2U-75
72
0 4 8 12 16 20 24 28
Age of specimens (d)

2.5.2 Influence of w/c Ratio

1. Mixture proportion
Six concrete mixtures with different w/ c ratios were investigated. The mixture
proportions, which were designated as RC33-0, RC40-0, RC50-0, IC33-5, IC33-
15, and IC33-25, are shown in Table 2.5. Mixtures RC33-0, RC40-0, and RC50-0
represented ordinary concrete without IC. Mixtures IC33-5, IC33-15, and IC33-25
underwent IC through SAPs. The ordinary concrete compositions were varied by
w/ c ratio (0.33, 0.40, and 0.50 for RC33-0, RC40-0, and RC50-0, respectively). The
ICC compositions were varied by the amounts of SAPs (0.05%, 0.15%, and 0.25%
by weight of cement for Mixtures IC33-5, IC33-15, and IC33-25, respectively).

Table 2.5 Compositions and properties of fresh mixtures

Mixture composition Concrete mixtures
RC33-0 RC40-0 RC50-0 IC33-5 IC33-15 IC33-25
we /c 0.33 0.40 0.50 0.33 0.33 0.33
Water (kg/m3 ) 171 180 200 171 171 171
Cement (kg/m3 ) 512 450 400 512 512 512
Fine aggregate (kg/m3 ) 636 655 666 636 636 636
Coarse aggregate (kg/m3 ) 1131 1165 1184 1131 1131 1131
SAPs (kg/m3 ) 0 0 0 267.5 803.0 1339.0
Internal curing water (kg/m3 ) 0 0 0 5.35 16.06 26.78
wic /c 0 0 0 0.01 0.03 0.05
Superplasticizer (kg/m3 ) 3.6 1.8 0.8 3.6 3.6 3.6
Slump (mm) 100 140 160 110 105 110

@seismicisolation
@seismicisolation
92 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

The ratio of IC water provided by SAPs to cement is also defined as IC w/ c

ratio [17]. The formula for calculating the total w/ c ratio could also be expressed by
Eq. (2.15).
In the reference concrete mixture (RC33-0), the effective w/ c ratio was fixed as
0.33. The IC w/ c ratios were 0.01, 0.03, and 0.05 for Mixtures IC33-5, IC33-15, and
IC33-25, respectively. Therefore, the total w/ c ratios were 0.34, 0.36, and 0.38 for
Mixtures IC33-5, IC33-15, and IC33-25, respectively. Moreover, the additional w/ c
ratios were 0, 0.07, and 0.17 for Mixtures RC33-0, RC40-0, and RC50-0, respectively.

2. Effect of w/c ratio on the change of internal relative humidity in early-age concrete
internally cured with SAPs
For concrete specimens under unsealed condition, the IRH in concrete is influenced
by both self-desiccation and external drying [26, 34]. The IRH initially enters a
water–vapor saturated stage with 100% RH due to the high amounts of liquid water
in the pores [35]. The self-desiccation means cement continues to hydrate accompa-
nying the consumption of the mixing water and IC water [36]. The IRH in concrete
internally cured with SAPs will begin to drop from the initial water–vapor saturated
stage with 100% RH when the water in concrete pores cannot counteract the water
reduction resulting from both self-desiccation and external drying [35].
Figures 2.33 and 2.34 show the test results of IRH variations in ordinary and ICC,
respectively.
The IRH at 28 d after casting was 84.2%, 90.0%, 94.1%, 89.1%, 91.2%, and
94.5% RH when the total w/ c ratios were 0.33, 0.40, 0.50, 0.34, 0.36, and 0.38 for
Mixtures RC33-0, RC40-0, RC50-0, IC33-5, IC33-15, and IC33-25, respectively.
The IRH at 28 d after casting increased by 4.9%, 7.0%, and 10.3% RH when the
IC w/ c ratio increased from 0 to 0.01, 0.03, and 0.05 for Mixtures RC33-0, IC33-5,
IC33-15, and IC33-25, respectively. The IRH in concrete internally cured with SAPs
at 28 d after casting increased with the increase of IC w/ c ratio.
The SAPs are supposed to be uniformly distributed in concrete and serve as
IC water reservoirs, which first absorb water during mixing and then release it

Fig. 2.33 Test results of 100

relationship between IRH
and age of specimens for 96
ordinary concrete
IRH (%)

92

88

84 Test data of RC33-0 Uncertainty of RC33-0

Test data of RC40-0 Uncertainty of RC40-0
Test data of RC50-0 Uncertainty of RC50-0
80
0 5 10 15 20 25 30
Age of specimens (d)

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 93

Fig. 2.34 Test results of 100

relationship between IRH
and age of specimens for 96
ICC

IRH (%)
92

88

Test data of RC33-0 Uncertainty of RC33-0

84 Test data of IC33-5 Uncertainty of IC33-5
Test data of IC33-15 Uncertainty of IC33-15
Test data of IC33-25 Uncertainty of IC33-25
80
0 5 10 15 20 25 30
Age of specimens (d)

to surrounding cement paste during cement hydration [54]. The moisture move-
ment between SAPs and cement paste is mainly controlled by osmotic pressure and
humidity gradient [55]. The osmotic pressure emerges resulting from the difference of
solution concentration inside and outside of the SAPs [55, 56]. The solution concen-
tration in cement paste increases gradually during hydration with the consumption
of mixing water. The SAPs will release water when the solution concentration of
cement paste is coincident with that of SAPs or a little higher [56]. Moreover, the
humidity gradient is generated in the concrete between the water in the SAPs and the
pore fluid [55]. Part of the humidity gradient is established by the capillary pressure
developing in the pore fluid as a consequence of emptying of the pores due to cement
hydration or drying [2]. The water movement from SAPs to cement paste will stop
when the capillary pressure in the SAPs is equivalent to that in the hydrating cement
paste.
For concrete with the same effective w/ c ratio of 0.33, the IRH at 28 d after casting
was 1.2%, and 4.5% RH higher for Mixtures IC33-15, and IC33-25 with IC w/ c ratios
of 0.03, and 0.05 than that for Mixture RC40-0 with additional w/ c ratio of 0.07,
respectively, as shown in Figs. 2.33 and 2.34. The IC w/ c ratios in Mixtures IC33-15
(with the total w/ c ratio of 0.36, which was in combination of the effective w/ c ratio
of 0.33 and IC w/ c ratio of 0.03), and IC33-25 (with the total w/ c ratio of 0.38, which
was in combination of the effective w/ c ratio of 0.33 and IC w/ c ratio of 0.05) were
more effective in preventing the IRH from decreasing than the additional w/ c ratio
in Mixture RC40-0 (with the total w/ c ratio of 0.40, which was in combination of
the effective w/ c ratio of 0.33 and additional w/ c ratio of 0.07), respectively. Results
reported in [29] show that the IC water provided by SAPs is much more effective in
increasing the IRH in concrete than the corresponding additional mixing water. The
possible reason for the difference may lie in that the two types of water (IC water
and additional mixing water) result in completely different microstructure of cement
pastes [29, 38].

@seismicisolation
@seismicisolation
94 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

3. Prediction model for critical time of internal relative humidity and working w/c
ratio
Equations (2.19)–(2.22) developed in [41] can be utilized to predict the IRH in
ordinary concrete considering the w/ c ratio and age of concrete at early age. However,
Eqs. (2.19)–(2.22) was not suitable for calculating the IRH variations in early-age
concrete internally cured with SAPs. The working w/ c ratio was also utilized to
accurately predict the IRH in early-age concrete internally cured with SAPs, as
shown in Eq. (2.23).
The working w/ c ratios were obtained by substituting Eqs. (2.21) and (2.22),
as well as the test results of IRH in Mixtures IC33-5, IC33-15, and IC33-25 into
Eq. (2.23). The working w/ c ratios were between 0.40 and 0.55, which were higher
than the total w/ c/ratio for each concrete mixture, as shown in Fig. 2.35. The influence
coefficient f (wic c, t) was also utilized to evaluate the efficiency of the IC w/ c ratio
for better predicting the IRH in early-age concrete internally cured with SAPs. The
working w/ c ratio could be expressed by Eq. (2.24). /
Figure 2.36 shows the relationship between
/ the influence coefficient f (wic c, t)
and age. The influence coefficient f (wic c, t) was influenced by both the IC w/ c
ratio and age. /
The relationship between the influence coefficient f (wic c, t) and IC w/ c ratio
was nonlinear and could be expressed as follows.
/ / pt
f (wic c, t) = Yt × (wic c) + / (2.64)
(wic c)

where Yt and pt are the parameters related to age of concrete, respectively.

The critical time for Mixtures IC33-25 was 6.9 d and only the results on the influ-
ence coefficient of concrete with the age larger /than 6.9 d were analyzed. The rela-
tionships between the influence coefficient f (wic c, t) and IC w/ c ratio of concrete at
the age of 7, 14, 21, and 28 d after casting are shown in Fig. 2.37 and were expressed
as follows.

Fig. 2.35 Relationship 0.65

Working w/c ratio (ww /c)

/ working w/ c ratio
between Test data of IC33-5
(ww c) and age of Test data of IC33-15
0.60 Test data of IC33-25
specimens Fitting curve of IC33-5
0.55 Fitting curve of IC33-15
Fitting curve of IC33-25

0.50

0.45

0.40
0 5 10 15 20 25 30
Age of specimens (d)

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 95

Fig. 2.36 Relationship 14

Influence coefficient f (wic /c, t)

Test data of IC33-5
between
/ influence coefficient Test data of IC33-15
f (wic c, t) and age of 12 Test data of IC33-25
specimens Fitting curve of IC33-5
10 Fitting curve of IC33-15
Fitting curve of IC33-25
8

2
0 5 10 15 20 25 30
Age of specimens (d)

Fig. 2.37 Relationship 12

Influence coefficient f (wic /c, t)

between Test data of 7 d

/ influence coefficient 11
Test data of 14 d
f (wic c, t) and IC w/ c ratio 10 Test data of 21 d
9 Test data of 28 d
Fitting curve of 7 d
8 Fitting curve of 14 d
7 Fitting curve of 21 d
Fitting curve of 28 d
6
5
4
3
2
0.01 0.02 0.03 0.04 0.05 0.06
wic / c

At 7 d after casting:

/ / 0.108 2
f (wic c, 7) = 43.184 × (wic c) + / R = 0.999 (2.65)
(wic c)

At 14 d after casting:

/ / 0.083
f (wic c, 14) = 41.239 × (wic c) + / R2 = 0.999 (2.66)
(wic c)

At 21 d after casting:

/ / 0.070 2
f (wic c, 21) = 40.139 × (wic c) + / R = 0.997 (2.67)
(wic c)

@seismicisolation
@seismicisolation
96 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

At 28 d after casting:

/ / 0.062
f (wic c, 28) = 40.474 × (wic c) + / R2 = 0.996 (2.68)
(wic c)

Figure 2.37 shows that the influence coefficient decreased with the increase of age
of concrete. The deviations on influence coefficient between results from Eq. (2.64)
and test were small.
/ Therefore Eq. (2.64) could be utilized to predict the influence
coefficient f (wic c, t) in consideration of IC w/ c ratio and age of concrete.
The parameters Yt were 43.184, 41.239, 40.139, and 40.474 for concrete at the age
of 7, 14, 21, and 28 d after casting, respectively, and the average value was 41.259.
The deviations on parameter Yt between the individual value and average value were
4.5%, 0.1%, 2.8%, and 1.9%, respectively.
As shown in Eqs. (2.65)–(2.68), the parameters pt were 0.108, 0.083, 0.070, and
0.062 for concrete at the age of 7, 14, 21, and 28 d after casting, respectively. The
relationship between parameter pt and age of concrete t is shown in Fig. 2.38 and
could be expressed as follows.

pt = t ϕ + λ (2.69)

where ϕ and λ are the parameters that can be determined with regression analysis;
and the results were ϕ=−0.0369, λ=−0.823, and R2 =0.997, respectively.
Figure 2.39 shows that the critical times were 1.4, and 2.5 d when the total w/ c
ratios were 0.40, and 0.50 for Mixtures RC40-0, and RC50-0, respectively. The
critical times were 1.0, 2.3, 3.0, and 6.9 d, which increased by 1.3, 2.0, and 5.9 d
when the IC w/ c ratios increased from 0 to 0.01, 0.03, and 0.05 for Mixtures RC33-
0, IC33-5, IC33-15, and IC33-25, respectively. The critical time of IRH in concrete
internally cured with SAPs increased with the increase of IC w/ c ratio.
There was not obvious relationship between critical time and total w/ c ratio, as
shown in Fig. 2.39a. The total w/ c ratio could not be utilized to predict the critical

Fig. 2.38 Relationship

between parameter pt and
age t of specimens

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 97

Fig. 2.39 Relationship

between w/ c ratio and
critical time

time Tcri,2 . The working w/ c ratio was influenced by age of concrete. The effect of
age on the working w/ c ratio was not considered in proposing the model for the
critical time Tcri,2 . The average value of parameter pt for the age of 7, 14, 21, and 28
d after casting was 0.081, which was utilized to obtain the working w/ c ratio for the
critical time Tcri,2 , ww,T /c. Substituting the average values of parameters Yt and pt
into Eq. (2.24) yields. Figure 2.39b shows that the relationship between critical time
Tcri,2 and working w/ c ratio for critical time was polynomial and could be expressed
by Eqs. (2.31) and (2.70), respectively.
/ /
ww,T c = we /c + 41.259 × (wic c)2 + 0.081 (2.70)

The results were α2 = 150.91, β2 = −104.45, γ2 = 19.03, and R2 =0.766,

respectively.

@seismicisolation
@seismicisolation
98 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

Figure 2.39b shows the comparisons of critical time of IRH between results from
Eqs. (2.31) and (2.70) and test. The values of critical time Tcri,2 obtained from Eqs.
(2.31) and (2.70) were 1.0, 1.4, 1.7, 2.5, 4.5, and 5.2 d when the working w/ c
ratios were 0.33, 0.40, 0.42, 0.45, 0.50, and 0.52 for Mixtures RC33-0, RC40-0,
IC33-5, IC33-15, RC50-0, and IC33-25, respectively. The deviations on critical time
between results from Eqs. (2.31) and (2.70) and test were 0, 0, 0.6, 0.5, 2.0, and 1.7 d,
respectively. The deviations were small. Therefore, Eqs. (2.31) and (2.70) could be
utilized to calculate the critical time of IRH in concrete internally cured with SAPs
considering the effective and IC w/ c ratios.
The critical time also depends on more factors, such as the RH sensor that may
has a high uncertainty at these high RH levels, the isothermal condition, and the
capillary pressure that is much earlier than any RH drop that is picked up by RH
sensors [57]. Although only effective and IC w/ c ratios were considered, the models
could be utilized to predict the critical time of IRH in early-age concrete internally
cured with SAPs.
The prediction model for working w/ c ratio was important for accurately
predicting the IRH in early-age concrete internally cured with SAPs. Substituting
Eqs. (2.64) and (2.69) into Eq. (2.24), the working w/ c ratio is shown in Fig. 2.35
and could be expressed as follows.
/ / /
ww c = we c + Yt × (wic c)2 + t ϕ + λ (2.71)

Figure 2.35 shows the comparisons of working w/ c ratio in concrete internally

/ between results from Eq. (2.71) and test. For Mixture IC33-5, the
cured with SAPs
values of ww c obtained from test results or Eq. (2.71) were 0.463, 0.443, 0.418,
0.403, and 0.396 or 0.471, 0.442, 0.418, 0.405, and 0.395 at the / age of 3, 7, 14,
21, and 28 d after casting, respectively. The deviations on ww c between results
from Eq. (2.71) and test were 1.7%, 0.2%, 0%, 0.5%, and 0.3%, respectively, and
the average value of deviations was 0.5%. The average values of deviations were
0.5%, and 0.8% for Mixtures IC33-15, and IC33-25, respectively. The deviations
were small. Therefore, Eq. (2.71) could be utilized to predict the working w/ c ratio
in consideration of effective and IC w/ c ratios, as well as age of concrete. This model
could be utilized to predict the critical time and IRH variations in early-age concrete
internally cured with SAPs.

4. Prediction model for internal relative humidity in concrete internally cured with
SAPs

The prediction model for IRH was necessary for predicting shrinkage of early-
age concrete internally cured with SAPs. Substituting Eqs. (2.21) and (2.22) into
Eq. (2.23) yields Eq. (2.72). Substituting the values of parameters Yt , ϕ, and λ into
Eq. (2.71) yields Eq. (2.73). Substituting the values of parameters α2 , β2 , and γ2 into
Eqs. (2.31) and (2.70) yields Eq. (2.74). The relationship between IRH and age of
concrete could be expressed as follows.

@seismicisolation
@seismicisolation
2.5 Influential Factors of the Internal Relative Humidity of Early-Age Concrete 99
⎧
⎪
⎪ 100 t ≤ Tcri,2
( / ) ⎨ 100 + (7.295 + 20.586 × ln(t))×
RH ww c, t = /
⎪
⎩ w / c + (−1.822 − 17.052 × ln(t)) T
⎪
w cri,2 < t ≤ 28d
(2.72)

where
/ / /
ww c = we c + 41.259 × (wic c)2 + (t −0.0369 − 0.823) (2.73)

and
( / / )2
Tcri,2 = 150.91 × we c + 41.259 × (wic c)2 + 0.081 − 104.45×
( / / ) (2.74)
we c + 41.259 × (wic c)2 + 0.081 + 19.03
/
where RH(ww c, t)/is the IRH in early-age concrete internally cured with SAPs in
consideration of ww c, critical time, and age of concrete, in %.
Figure 2.40 shows the comparisons of IRH in concrete internally cured with
SAPs between results from Eqs. (2.72)–(2.74) and test. For Mixtures IC33-5, the
IRH obtained from test results or Eqs. (2.72)–(2.74) was 99.8%, 96.5%, 93.0%,
90.7%, and 89.1% RH or 99.98%, 96.47%, 93.03%, 90.78%, and 89.08% RH at
the age of 3, 7, 14, 21, and 28 d after casting, respectively. The deviations on IRH
between results from Eqs. (2.72)–(2.74) and test were 0.18%, 0.03%, 0.03%, 0.08%,
and 0.02% RH, respectively, and the average value of deviations was 0.07% RH.
The average values of deviations were 0.07%, and 0.15% RH for Mixtures IC33-15,
and IC33-25, respectively. The deviations were small. Therefore, Eqs. (2.72)–(2.74)
could be utilized to predict the IRH in early-age concrete internally cured with SAPs
in consideration of working w/ c ratio, critical time, and age of concrete.
Results reported in [29] show that the IRH obtained from the test is 98.1%, 97.3%,
and 96.8% RH at the age of 7, 10, and 14 d after casting for concrete HSC-S1,

Fig. 2.40 Relationship 100

between IRH and age of
specimens
96
92
IRH (%)

88 Test data of RC33-0

Test data of IC33-5
84 Test data of IC33-15
Test data of IC33-25
80 Fitting curve of RC33-0
Fitting curve of IC33-5
76 Fitting curve of IC33-15
Fitting curve of IC33-25
72
0 5 10 15 20 25 30
Age of specimens (d)

@seismicisolation
@seismicisolation
100 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

respectively, and the corresponding results obtained from Eqs. (2.72)–(2.74) was
98.5%, 97.1%, and 95.7% RH, respectively. The deviations between results from
Eqs. (2.72)–(2.74) and the test in [29] were 0.4%, 0.2%, and 1.1% RH, respectively.
Furthermore, autogenous shrinkage could also be predicted based on IRH from the
proposed Eqs. (2.72)–(2.74) for better understanding cracking resistance of early-age
concrete internally cured with SAPs.

2.5.3 Summary

The effect of curing humidity on the critical time, IRH, IRH decrease rate, and
moisture diffusion coefficient of early-age HSC under the conditions of sealing and
unsealing, as well as an analysis of prediction model for IRH in HSC considering the
effect of curing humidity were investigated. Besides, the prediction models for the
working w/ c ratio, the critical time, and the IRH variations in early-age concrete inter-
nally cured with SAPs were investigated. The main conclusions could be obtained
as follows.
(1) The IRH in sealed HSC was affected by the self-desiccation alone, and the
critical time of which was 9 d, while the IRH in unsealed HSC was affected by
the self-desiccation and moisture diffusion, and the critical time increased from
2 to 3, and 5 d when the curing humidity increased from 25 to 50%, and 75%
RH, respectively.
(2) The IRH in sealed and unsealed HSC decreased with the growth of age, and
the IRH in unsealed was lower than that in sealed HSC. Meanwhile, the IRH
in unsealed HSC increased as the curing humidity increased at the same curing
age.
(3) The IRH decrease rate of sealed or unsealed HSC gradually decreased with the
growth of age, and the IRH decrease rate of unsealed HSC decreased as the
curing humidity increased at the same curing age.
(4) The moisture diffusion coefficient of HSC was calculated, and the prediction
model for IRH in HSC under the combined effects of moisture diffusion and
self-desiccation considering the effect of curing humidity was proposed.
(5) A prediction model for the working w/ c ratio was proposed in consideration of
age of concrete, effective and IC w/ c ratios, which could be utilized to predict
IRH variations in early-age concrete internally cured with SAPs.
(6) A prediction model for the critical time of IRH in early-age concrete internally
cured with SAPs was proposed in consideration of effective and IC w/ c ratios.
(7) In consideration of working w/ c ratio and critical time, as well as age of concrete,
a prediction model for the IRH in early-age concrete internally cured with SAPs
was proposed.

@seismicisolation
@seismicisolation
References 101

References

1. Jensen OM, Hansen PF (1999) Influence of temperature on autogenous deformation and relative
humidity change in hardening cement paste. Cem Concr Res 29(4):567–575
2. Lura P, Jensen OM, van Breugel K (2003) Autogenous shrinkage in high-performance cement
paste: an evaluation of basic mechanisms. Cem Concr Res 33(2):223–232
3. Craeye B, Geirnaert M, de Schutter G (2011) Super absorbing polymers as an internal curing
agent for mitigation of early-age cracking of high-performance concrete bridge decks. Constr
Build Mater 25(1):1–13
4. Zhang J, Huang Y, Qi K et al (2012) Interior relative humidity of normal- and high-strength
concrete at early age. J Mater Civ Eng 24(6):615–622
5. Henkensiefken R, Bentz D, Nantung T et al (2009) Volume change and cracking in inter-
nally cured mixtures made with saturated lightweight aggregate under sealed and unsealed
conditions. Cement Concr Compos 31(7):427–437
6. Zhutovsky S, Kovler K (2012) Effect of internal curing on durability-related properties of high
performance concrete. Cem Concr Res 42(1):20–26
7. Mihashi H, Leite JPDB (2004) State-of-the-art report on control of cracking in early age
concrete. J Adv Concr Technol 2(2):141–154
8. Han YD, Zhang J, Luosun YM et al (2014) Effect of internal curing on internal relative humidity
and shrinkage of high strength concrete slabs. Constr Build Mater 61:41–49
9. Bentur A, Igarashi SI, Kovler K (2001) Prevention of autogenous shrinkage in high-strength
concrete by internal curing using wet lightweight aggregates. Cem Concr Res 31(11):1587–
1591
10. Wang T (2016) Effect of curing temperatures on relative humidity of internally cured concrete
with ceramsite. Hohai University, Nanjing (in Chinese)
11. Wyrzykowski M, Lura P, Pesavento F et al (2011) Modeling of internal curing in maturing
mortar. Cem Concr Res 41(12):1349–1356
12. Wei Y, Wang Y, Gao X (2015) Effect of internal curing on moisture gradient distribution and
deformation of a concrete pavement slab containing pre-wetted lightweight fine aggregates.
Drying Technol 33(3):355–364
13. Wang ML (2018) Effect of ambient condition on internal relative humidity in concrete internally
cured with super absorbent polymers. Hohai University, Nanjing (in Chinese)
14. Castro J, Spragg R, Weiss J (2012) Water absorption and electrical conductivity for internally
cured mortars with a W/C between 0.30 and 0.45. J Mater Civ Eng 24(2):223–231
15. Shen DJ, Wang T, Chen Y et al (2015) Effect of internal curing with super absorbent polymers
on the relative humidity of early-age concrete. Constr Build Mater 99:246–253
16. Shen DJ, Wang ML, Chen Y et al (2017) Prediction of internal relative humidity in concrete
modified with super absorbent polymers at early age. Constr Build Mater 149:543–552
17. Shen DJ, Wang ML, Chen Y et al (2017) Prediction model for relative humidity of early-
age internally cured concrete with pre-wetted lightweight aggregates. Constr Build Mater
144:717–727
18. Shen DJ, Liu C, Wang ML et al (2020) Prediction model for internal relative humidity in
early-age concrete under different curing humidity conditions. Constr Build Mater 265:119987
19. Golias M, Castro J, Weiss J (2012) The influence of the initial moisture content of lightweight
aggregate on internal curing. Constr Build Mater 35:52–62
20. General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China. Lightweight aggregates and its test methods-part2: Test methods for
lightweight aggregates: GB/T 17431.2-2010[S]. Standard press of China, 2010, Beijing (in
Chinese)
21. Wei Y, Xiang YP, Zhang QQ (2014) Internal curing efficiency of prewetted LWFAs on concrete
humidity and autogenous shrinkage development. J Mater Civ Eng 26(5):947–954
22. Yang QB, Zhang SQ (2004) Self-desiccation mechanism of high-performance concrete. J
Zhejiang Univ Sci 5(12):1517–1523

@seismicisolation
@seismicisolation
102 2 Internal Relative Humidity of Early-Age Internally Cured Concrete

23. Shen DJ, Zhou BZ, Wang ML et al (2019) Predicting relative humidity of early-age concrete
under sealed and unsealed conditions. Mag Concr Res 71(22):1151–1166
24. El-Dieb AS (2007) Self-curing concrete: water retention, hydration and moisture transport.
Constr Build Mater 21(6):1282–1287
25. Ye JJ, Hu SG, Wang FZ, et al (2006) Effect of pre-wetted light-weight aggregate on internal
relative humidity and autogenous shrinkage of concrete. J Wuhan Univ Technol Mater Sci Ed
21(1):134–137
26. Radlinska A, Rajabipour F, Bucher B et al (2008) Shrinkage mitigation strategies in cemen-
titious systems: a closer look at differences in sealed and unsealed behavior. Transp Res Rec
2070(1):59–67
27. ASTM International. Standard practice for maintaining constant relative humidity by means
of aqueous solutions: ASTM E104-02[S]. West Conshohocken: ASTM International
28. Šelih J, Bremner TW (1996) Drying of saturated lightweight concrete: an experimental
investigation. Mater Struct 29(7):401–405
29. Kong XM, Zhang ZL, Lu ZC (2015) Effect of pre-soaked superabsorbent polymer on shrinkage
of high-strength concrete. Mater Struct 48(9):2741–2758
30. De Sensale GR, Ribeiro AB, Gonçalves A (2008) Effects of RHA on autogenous shrinkage of
Portland cement pastes. Cem Concr Compos 30(10):892–897
31. Pickett G (1956) Effect of aggregate on shrinkage of concrete and a hypothesis concerning
shrinkage. J Proc 581–590
32. Zhang J, Chen Y, Huang E, Chen DS, Zuo YL (2014) Relationship of early estimation on
concrete strength. Chongqing Arch 13(8):61–63 (in Chinese)
33. Cusson D, Hoogeveen T (2008) Internal curing of high-performance concrete with pre-soaked
fine lightweight aggregate for prevention of autogenous shrinkage cracking. Cem Concr Res
38(6):757–765
34. Kim JK, Lee CS (1999) Moisture diffusion of concrete considering self-desiccation at early
ages. Cem Concr Res 29(12):1921–1927
35. Zhang J, Qi K, Huang Y (2009) Calculation of moisture distribution in early-age concrete. J
Eng Mech 135(8):871–880
36. Jiang ZW, Sun ZP, Wang PM (2006) Internal relative humidity distribution in high-performance
cement paste due to moisture diffusion and self-desiccation. Cem Concr Res 36(2):320–325
37. Akita H, Fujiwara T, Ozaka Y (1997) A practical procedure for the analysis of moisture transfer
within concrete due to drying. Mag Concr Res 49(179):129–137
38. Justs J, Wyrzykowski M, Winnefeld F et al (2014) Influence of superabsorbent polymers on
hydration of cement pastes with low water-to-binder ratio. J Therm Anal Calorim 115(1):425–
432
39. Persson B (1997) Moisture in concrete subjected to different kinds of curing. Mater Struct
30(9):533–544
40. Parrott LJ (1988) Moisture profiles in drying concrete. Adv Cem Res 1(3):164–170
41. Chen Y (2015) Research on development of internal relative humidity of internally cured
concrete at early age [D]. Hohai University, Nanjing (in Chinese)
42. McDonald DB, Roper H, Parrott LJ (1991) Discussion: Factors influencing relative humidity
in concrete. Mag Concr Res 43(157):305–307
43. Jiang ZW, Sun ZP, Wang PM (2005) Autogenous relative humidity change and autogenous
shrinkage of high-performance cement pastes. Cem Concr Res 35(8):1539–1545
44. Zhang J, Hou DW, Gao Y et al (2011) Determination of moisture diffusion coefficient of
concrete at early age from interior humidity measurements. Drying Technol 29(6):689–696
45. Bažant ZP, Najjar LJ (1972) Nonlinear water diffusion in nonsaturated concrete. Matériaux et
Construction 5(1):3–20
46. Chen HY, Chen C (2014) Equilibrium relative humidity method used to determine the sorption
isotherm of autoclaved aerated concrete. Build Environ 81:427–435
47. Pei JJ, Zhang JSS (2012) Determination of adsorption isotherm and diffusion coefficient of
toluene on activated carbon at low concentrations. Build Environ 48:66–76

@seismicisolation
@seismicisolation
References 103

48. Oh BH, Cha SW (2003) Nonlinear analysis of temperature and moisture distributions in early-
age concrete structures based on degree of hydration. ACI Mater J 100(5):361–370
49. Pane I, Hansen W (2002) Concrete hydration and mechanical properties under nonisothermal
conditions. ACI Mater J 99(6):534–542
50. Schindler AK, Folliard KJ (2005) Heat of hydration models for cementitious materials. ACI
Mater J 102(1):24–33
51. Wei Y, Liang SM, Gao X (2017) Numerical evaluation of moisture warping and stress in
concrete pavement slabs with different water-to-cement ratio and thickness. J Eng Mech
143(2):04016111
52. Jiang JH, Yuan YS, Wang SL et al (2013) Prediction of response of relative humidity in
concrete under artificial climate environment. J Cent South Univ (Sci Technol) 12:5091–5099
(in Chinese)
53. Tian JZ, Xu NZ, Li FM (2009) Proximate calculation of error function erf(x) and its application
in mining subsidence prediction. Coal Min Technol 14(2):33–35 (in Chinese)
54. Wyrzykowski M, Lura P, Pesavento F et al (2012) Modeling of water migration during internal
curing with superabsorbent polymers. J Mater Civ Eng 24(8):1006–1016
55. Mechtcherine V, Reinhardt HW (2012) Application of super absorbent polymers (SAP) in
concrete construction: state-of-the-Art Report of RILEM TC 225-SAP[M]. Dordrecht, the
Netherlands: Springer, pp 14–28
56. Wang FZ, Jin Y, Hua C et al (2015) Study on mechanism of desorption behavior of saturated
superabsorbent polymers in concrete. ACI Mater J 112(3):463–469
57. Miao CW, Tian Q, Sun W et al (2007) Water consumption of the early-age paste and the
determination of “time-zero” of self-desiccation shrinkage. Cem Concr Res 37(11):1496–1501

@seismicisolation
@seismicisolation
Chapter 3
Autogenous Shrinkage of Early-Age
Internally Cured Concrete

Contents

3.1 Methods and Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

3.1.1 Autogenous Shrinkage Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
3.1.2 Ultrasonic Velocity Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
3.2 Autogenous Shrinkage of Early-Age Concrete Internally Cured with SAPs . . . . . . . . . . 110
3.2.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
3.2.2 Early-Age Expansion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
3.2.3 Autogenous Shrinkage at 28 d . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
3.2.4 Autogenous Shrinkage Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
3.2.5 Prediction Models for Autogenous Shrinkage . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
3.2.6 Internal Curing Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
3.2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
3.3 Autogenous Shrinkage of Early-Age Internally Cured Concrete Reinforced
with Polypropylene Fiber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
3.3.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
3.3.2 Influence of Polypropylene Fiber on Early-Age Expansion . . . . . . . . . . . . . . . . . 124
3.3.3 Influence of Polypropylene Fiber on Autogenous Shrinkage . . . . . . . . . . . . . . . . 127
3.3.4 Influence of Polypropylene Fiber on Ultrasonic Velocity . . . . . . . . . . . . . . . . . . . 130
3.3.5 Prediction Model of Autogenous Shrinkage Based on Ultrasonic Velocity . . . . . 132
3.3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
3.4 Relationship Between Autogenous Shrinkage and Internal Relative Humidity
of Concrete Internally Cured with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
3.4.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
3.4.2 Internal Relative Humidity of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . 137
3.4.3 Early-Age Expansion of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
3.4.4 Autogenous Shrinkage of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
3.4.5 Correlation Between Autogenous Shrinkage and Internal Relative Humidity
of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
3.4.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
3.5 Relationship Between Autogenous Shrinkage and Internal Relative Humidity
of Internally Cured Concrete with Different w/c Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
3.5.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
3.5.2 Internal Relative Humidity of Concrete with Different w/c Ratios . . . . . . . . . . . . 151
3.5.3 Early-Age Expansion of Concrete with Different w/c Ratios . . . . . . . . . . . . . . . . 152
3.5.4 Effect of w/c Ratio on Correlation Between Autogenous Shrinkage and Internal
Relative Humidity of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154

© Science Press 2023 105

D. Shen, Cracking Control on Early-Age Concrete Through Internal Curing,
https://doi.org/10.1007/978-981-19-8398-6_3
@seismicisolation
@seismicisolation
106 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

3.5.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159

3.6 Relationship Between Autogenous Shrinkage and Internal Relative Humidity
of Concrete Internally Cured with Pre-Wetted LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.6.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159
3.6.2 Temperature History of Concrete with Pre-wetted LWAs . . . . . . . . . . . . . . . . . . . 160
3.6.3 Early-Age Expansion of Concrete with Pre-wetted LWAs . . . . . . . . . . . . . . . . . . 161
3.6.4 Shrinkage of Concrete with Pre-wetted LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
3.6.5 Internal Relative Humidity of Concrete with Pre-Wetted LWAs . . . . . . . . . . . . . . 162
3.6.6 Relationship Between Autogenous Shrinkage and Internal Relative Humidity
of Concrete with Pre-wetted LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
3.6.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

Autogenous shrinkage is caused by the self-desiccation of concrete due to the lack of

water during cement hydration [1, 2]. In concrete with a low w/c ratio, the internal
moisture is insufficient for complete cement hydration [3]. Thus, the autogenous
shrinkage is more likely to occur [4]. Tensile stress develops rapidly in structures
when the shrinkage of the concrete is restrained by connected members or adja-
cent materials [5]. Concrete may crack when restrained stresses exceed the tensile
strength [6]. Moreover, the ingress of harmful substances and water through cracks
can result in steel corrosion, internal expansive reactions, and chemical attack; then,
the durability of structures is endangered. Therefore, autogenous shrinkage is a major
concern in early-age cracking of concrete [4]. Many materials have been invented
to mitigate the autogenous shrinkage of concrete such as fibers [7–11], LWAs [12–
15], SAPs [16–20], water-reducing agents [21], shrinkage-reducing agents [22, 23],
and expansive agents [24]. The addition of SAPs to concrete will form water-filled
inclusions that gradually release water to promote further cement hydration and thus
the autogenous shrinkage of ICC is reduced [25]. Besides, the model of early-age
autogenous shrinkage is of great significance in evaluating the cracking potential of
early-age concrete.
In this chapter, firstly, whether and how IC with different amounts of SAPs
influences the early-age expansion, autogenous shrinkage development, autoge-
nous shrinkage rate, and IC efficiency of concrete are explored [26]. The model
for predicting the autogenous shrinkage of ICC considering the expansion of early-
age concrete is established. Secondly, the effect of PP fiber on early-age autogenous
shrinkage of concrete is presented [27]. There is no fast method to obtain early-
age autogenous shrinkage of concrete on site. Non-contact autogenous shrinkage
measurement is adopted in this chapter because it can measure the deformation of
concrete continuously and automatically, and start the measurement after casting.
The prediction model for autogenous shrinkage of ICC reinforced with PP fiber is
established based on test results of ultrasonic velocity, which could be applied in
actual projects quickly. Thirdly, whether and how the IC water entrained in SAPs,
w/c ratio, and the proportion of LWAs affect the correlation between autogenous
shrinkage and IRH was experimentally studied through an integrated device that
measures the autogenous shrinkage and IRH of concrete simultaneously [28, 29].

@seismicisolation
@seismicisolation
3.1 Methods and Calculation 107

The model for predicting the early-age autogenous shrinkage based on test results
of IRH is also proposed.

3.1 Methods and Calculation

3.1.1 Autogenous Shrinkage Measurement

Autogenous shrinkage measurement was conducted using non-contact sensors [30].

This measurement technique accounted for the benefits of using the eddy-current
displacement sensor for measuring autogenous shrinkage [31]. The details of the
testing apparatus are shown in Fig. 3.1. Two identical sensor supports were installed
and fastened on top of a steel prism mold with an interior size of 100 mm ×
100 mm×550 mm. Two eddy-current displacement sensors were fixed on the sensor
supports, and their positions could be regulated previously. Two U-shaped invar
steel target seats were properly positioned on the bottom surface of the steel mold.
After the mixtures were cast, the target seats were embedded in concrete mixtures
and deformed simultaneously with concrete. Standard targets were magnetically
attached to the target seats. Thermocouples were embedded at the center of the
concrete mixture during the casting period to monitor the temperature change within
the concrete for the autogenous shrinkage to be separated from the measured total
deformation.

(a) Test device (all units in mm) (b) Photo of test device

Fig. 3.1 Test setup layout for autogenous shrinkage measurement

@seismicisolation
@seismicisolation
108 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

The standard targets and the corresponding eddy-current displacement sensors

constituted fundamental elements for stirring up eddy current, and any tiny deforma-
tion of the concrete mixture caused by self-desiccation, temperature, and moisture
gradients could be monitored by a signal conversion device that transforms electric
signals into displacements. Intervals of measurement data acquisition could be freely
set. The measurement period lasted 28 days.
To minimize friction between the inner surfaces of the steel mold and concrete
mixture, Teflon sheets were placed on the inner surfaces of the steel mold. The top
surface of the concrete mixture was sealed by two layers of polyethylene sheets for
the measured linear strain to not contain the component owing to exterior drying
(defined as total deformation). Round holes in the target seat ensured its anchorage
effect on concrete and uniformity of mixture. The use of gaskets attached to the
target seat minimized the friction between the target seat and the bottom surface of
the mold.
Temperature variation results in thermal strain, which is fully coupled with auto-
genous shrinkage of concrete. Temperature affects the hydration and self-desiccation
of concrete, which further affects autogenous shrinkage [31]. Positive and negative
strains represent the shrinkage and expansion, respectively. The data measured in the
test included thermal strain caused by temperature variation and the development of
autogenous shrinkage with age was obtained by subtracting the thermal strain [32,
33] as expressed by Eq. (3.1).

εas (t) = εtotal (t) − α(t) · [T (t) − T0 ] (3.1)

where εas (t) is the autogenous shrinkage at t d, in με; εtotal (t) is the total measured
strain at t d, in με; α(t) is the thermal expansion coefficient of mixtures, in με/°C;
T (t) is the temperature of mixtures at t d, in °C; and T0 is the initial temperature of
mixtures, in °C.
One of the main challenges in the study of the autogenous shrinkage in cementi-
tious materials is the determination of time-zero. The technical committee on auto-
genous shrinkage at the Japan Concrete Institute defined the time-zero of autogenous
shrinkage as the initial setting time, whereas ASTM C 1698 defined it as the final
setting time [34, 35]. Darquennes et al. [35] performed a series of experiments for
setting evolution, and the time-zero of classic and slag cement concretes was deter-
mined as equal to the final set and the time of the expansion peak, respectively. Miao
et al. [36] reported that the drop in IRH associated with the self-desiccation results in
an increase in the capillary pressure, and the moment when there is a drop in IRH of
the material is regarded as an indication for the time-zero. In addition, the time-zero
could be verified by measuring the self-stress evolution through restrained shrinkage
tests.

@seismicisolation
@seismicisolation
3.1 Methods and Calculation 109

3.1.2 Ultrasonic Velocity Measurement

Non-destructive testing is a test method that does not affect the mechanical properties
or other functions of the structure or components and is easy to operate [37]. Ultra-
sonic velocity measurement is non-destructive to evaluate the integrity and quality of
the concrete. With a non-metal ultrasonic detector, the ultrasonic velocity of concrete
can be easily measured [38, 39]. Ultrasonic velocity test equipment consisted of two
transducers and a data acquisition system, as shown in Figs. 3.2 and 3.3, respectively.
One transducer was used for transmitting the ultrasonic pulse while the other one
was used for receiving the ultrasonic pulse. Transducer sets having different reso-
nant frequencies are available for special applications: high-frequency transducers
(100 kHz) are used for small-size specimens, relatively short path lengths, or HSC,
whereas low-frequency transducers (below 25 kHz) are used for larger specimens and
relatively longer path lengths, or concrete with larger size aggregates. Each prismatic
steel mold with the dimension of 150 mm × 150 mm × 550 mm was assembled
firmly. A layer of butter was applied inside the steel molds (including the bottom
plate and the side walls). A layer of plastic film was placed inside the steel molds,
while a certain blank area was left on the top and sides to make the transducers
of the ultrasonic tester in close contact with the concrete. The mixed concrete was
poured into the steel mold and vibrated during the casting process with the vibrator.
The ultrasonic velocity measurement was conducted under the same condition as
that of the autogenous shrinkage measurement. Ultrasonic velocity was measured by
the indirect or surface transmission method (ranging distance: 350 mm) before the
specimen was demolded, and measured by the direct transmission method (ranging
distance: 550 mm) after the mold was removed. To keep the data stable during the
test, vaseline was utilized on the surfaces of the specimen and the transducers to
reduce any signal losses owing to air voids, as reported in [40]. Repeat measure-
ments of ultrasonic velocity were conducted at the same location to minimize the
erroneous readings.

Fig. 3.2 Ultrasonic velocity

test equipment

@seismicisolation
@seismicisolation
110 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

150
550

(a) Indirect or surface transmission method

150
550

(b) Direct transmission method

Fig. 3.3 Test setup layout for ultrasonic velocity measurement (all units in mm)

3.2 Autogenous Shrinkage of Early-Age Concrete

Internally Cured with SAPs

3.2.1 Mixture Proportion

Four concrete mixtures with low w/c ratio were used, as reported in [26]. The mixture
proportions, designated as ShS00, ShS05, ShS15, and ShS25, are shown in Table
3.1. Mixture ShS00 represented a reference concrete without IC, whereas Mixtures
ShS05, ShS15, and ShS25 underwent IC through SAP. The concrete compositions
were varied by the addition of SAPs (approximately 0.05%, 0.16%, and 0.26%
by weight of cement for Mixtures ShS05, ShS15, and ShS25, respectively). In the
reference concrete mixture (ShS00) without IC, the w/c ratio was fixed as 0.33. The
ratio of IC water entrained by SAPs to cement of Mixture ShS05, ShS15, and ShS25
varied respectively as 0.01, 0.03, and 0.05. That is to say, the total w/c ratios varied
from 0.33 to 0.38.

@seismicisolation
@seismicisolation
3.2 Autogenous Shrinkage of Early-Age Concrete Internally Cured with SAPs 111

Table 3.1 Concrete mixture proportions

Mixture composition Concrete mixtures
ShS00 ShS05 ShS15 ShS25
Water (kg/m3 ) 171 171 171 171
Cement (kg/m3 ) 512 512 512 512
Fine aggregate (sand) (kg/m3 ) 636 627 610 593
Coarse aggregate (stone) (kg/m3 ) 1131 1131 1131 1131
SAPs (g/m3 ) 0 267.50 803.00 1339.00
IC water (kg/m3 ) 0 5.35 16.06 26.78
Superplasticizer (kg/m3 ) 3.60 3.60 3.60 3.60

3.2.2 Early-Age Expansion

Although the constant curing temperature of 20 °C was imposed on the mixtures,

the temperatures in mixtures were not necessarily constant. The temperature devel-
opments of the four mixtures are shown in Fig. 3.4. The test results showed that
the maximum temperatures of mixtures increased slightly with the increase of SAPs
as the introduced IC water improved the level of hydration and the temperature
increased. The results corroborated a previous study that showed that hydration heat
increased with the increase in the amount of SAPs and IC water provided by SAPs
[41]. In addition, the time at which the maximum temperature occurred was 12 h
after the mixtures were cast with different amounts of SAPs.
The autogenous shrinkage was analyzed from the initial setting time. Many of the
existing studies analyze the autogenous shrinkage of concrete within 28 d to reflect

Fig. 3.4 Relationship 24

between central temperature ShS00
and age of concrete ShS05
ShS15
ShS25
22 Ambient
Temperature

20

18
0 12 24 36 48 60 72
Age (h)

@seismicisolation
@seismicisolation
112 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Autogenous shrinkage (μe) 400 250

200

Autogenous shrinkage (μe)

300
150
200 100

50
100
0
0 -50 ShS00
ShS00
ShS05 ShS05
ShS15 -100 ShS15
-100 ShS25 ShS25
-150
0 5 10 15 20 25 30 0 24 48 72 96 120 144 168
Age (d) Age (h)
(a) 28 days (b) 7 days

Fig. 3.5 Relationship between autogenous shrinkage and age of concrete

its early-age performance [42, 43]. The autogenous shrinkage of each mixture was
determined based on Eq. (3.1), as shown in Fig. 3.5.
As shown in Fig. 3.5, an obvious three-stage pattern of the development of auto-
genous shrinkage versus age starting from the initial setting time was observed and
behaved as continuous expansion before 24 h. The shrinkage stage increased within
a few days, followed by a stable shrinkage stage with a gradually reducing rate.
The expansion of internally cured mixtures with SAPs at early age differed from
those of mixtures without SAP. The ordinary mixture exhibited minimal and rapid
expansion after the initial setting time and soon turned into the second stage with
the fastest shrinkage development, which contributed to large autogenous shrinkage
up to 28 days. The autogenous shrinkage rate then drastically reduced after 7 days,
which marked the onset of the third stage. Conversely, the internally cured mixtures
with SAPs exhibited significant expansion in the first stage. After achieving the
maximum expansion, the mixture turned into the second stage with a lower autoge-
nous shrinkage rate than that in the ordinary mixture. This stage sustained growth
relatively longer, then transited to the third stage about 7 days later. The autogenous
shrinkage development of ordinary mixture and internally cured mixtures with SAPs
were similar at the shrinkage stage.
A substantial reduction in the autogenous shrinkage up to 28 days was observed
in the internally cured mixtures with SAP. This reduction may be explained by
the influence of expansion at early age and the subsequent decrease in autogenous
shrinkage rate. The effect of IC on autogenous shrinkage can be highly advantageous
for controlling early-age cracking because autogenous shrinkage reduces and shifts
to later ages when concrete has developed most of its strength [44]. Through contrast
and analysis, three main differences between ordinary mixtures and internally cured
mixtures can be drawn from the autogenous shrinkage development: (1) the early-age
expansion; (2) the development of autogenous shrinkage rate; and (3) the ultimate
autogenous shrinkage at 28 d.
Early-age expansion that can mitigate autogenous shrinkage [45] appeared after
the setting. In contrast to the ordinary mixture, the early-age expansion of internally

@seismicisolation
@seismicisolation
3.2 Autogenous Shrinkage of Early-Age Concrete Internally Cured with SAPs 113

Fig. 3.6 Relationship 0

between the maximum
expansion and IC water of -20

Maximun expansion (μe)

concrete
-40
2
y= -7-3.73x R =0.896
-60

-80

-100

-120
0 5 10 15 20 25 30
3
IC water (kg/m )

cured mixtures with SAPs was intense. As the production of portlandite formation
is the most common view for the emergence of early-age expansion [45, 46], the
modification of portlandite formation by the presence of SAPs may play a decisive
role in early-age expansion [41]. As shown in Fig. 3.6, the maximum expansion of
mixtures increased with the increase of IC water provided by SAP. The maximum
expansion of mixtures increased from −7 με to −48, −70, and −101 με when the
amounts of IC water increased from 0 kg/m3 to 5.35, 16.06, and 26.78 kg/m3 , which
increased at the rates of 5.86, 9.00, and 13.43 times, respectively.
In addition, the time of maximum expansion of mixtures increased from 7.4 h
to 7.6, 9.2, and 9.4 h for Mixture ShS00, ShS05, ShS15, and ShS25, respectively.
The average time of maximum expansion of mixtures was 8.4 h (0.35 d), and the
deviations between the individual value and the average value were 13.5%, 10.5%,
8.7%, and 10.6%, respectively.
For analysis and establishment of the later computational model, the development
of autogenous shrinkage was divided into the expansion stage and shrinkage stage
[45]. The effect of the amount of SAPs on early expansion can be explained by the
IC water provided by SAPs [41]. The relationship between the maximum expansion
and the amount of IC water provided by SAPs is shown in Fig. 3.6. The maximum
expansion of internally cured mixtures with SAPs increased with the increase of IC
water, as expressed by Eq. (3.2).

εexpmax (x) = εexpmax

0
+k ·x (3.2)

where εexpmax (x) is the maximum expansion for internally cured mixtures, in με;
εexpmax
0
is the maximum expansion for ordinary mixture, in με; and x is the amount
of IC water provided by SAP, in kg/m3 .
The serviceability of Eq. (3.2) is a challenge due to limited data and unde-
fined physical meaning. There is not a formula developed for all compositions. The

@seismicisolation
@seismicisolation
114 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.7 Comparison of the Test data of maximum expansion

0 Test data of [47]
relationship between the Test data of [48]
maximum expansion and IC -50 k=-3.73 Test data of [49](w/c=0.3)

Maximum expansion (μe)

Test data of [49](w/c=0.4)
water of concrete
-100
k=-3.40
-150
k=-4.67 k=-3.64
-200
Fitting curve of Eq. (3-3)
-250 Fitting curve of [47]
Fitting curve of [48]
-300 Fitting curve of [49](w/c=0.3) k=-5.80
Fitting curve of [49](w/c=0.4)
-350
0 10 20 30 40 50 60
IC water (kg/m )
3

coefficient k was determined through regression analysis, as expressed by Eq. (3.3).

εexpmax (x) = −7 − 3.73x (3.3)

Previous studies showed that the maximum expansion increased with the increase
of IC water [47–49], as shown in Fig. 3.7.
The maximum expansion of mixtures increased from −7 με to −48, −70, and
−101 με when the amounts of IC water increased from 0 kg/m3 to 5.35, 16.06,
and 26.78 kg/m3 . The coefficient k was −3.73 for 0.33 w/c. Similarly, coefficient
k was −3.40 for 0.28 w/b in [168], −5.80 for 0.30 w/c in [47], −4.67 and −3.64
for 0.30 and 0.40 w/c in [49], respectively. The coefficient k was −3.73, which is
between −3.40 and −5.80. The early-age expansion of internally cured mixtures
occurred within 24 h after the initial setting time. The expansion increased with the
increase in the age of mixtures, as shown in Fig. 3.8. And the relationship between
the expansion and age of mixtures is linear, which can be written as Eq. (3.4).

t
εicas (t) = ·ε (x) (t ≤ t0 ) (3.4)
to expmax

where εicas (t) is the autogenous shrinkage of internally cured mixtures at t d, in με;
εexpmax (x) is the maximum expansion for internally cured mixtures, in με; t0 is the
time of maximum expansion, in d; and t is the age of concrete, in d.
By substituting Eq. (3.3) into Eqs. (3.4) and (3.5) is obtained as follows.

t
εicas (t) = − · (7 + 3.73x) (t ≤ t0 ; t0 = 0.35 d) (3.5)
to

The expansion of ShS15 and ShS25 at 7 h (0.29 d) calculated from Eq. (3.5) were
−56 and −89 με whereas the test results were −62 and −86 με. The results of the
deviations between Eq. (3.5) and the test were −9.7% and 3.5%.

@seismicisolation
@seismicisolation
3.2 Autogenous Shrinkage of Early-Age Concrete Internally Cured with SAPs 115

Fig. 3.8 Relationship 0

between autogenous
shrinkage and age of -20
concrete (up to t0 hours)

Expansion (μe)
-40

-60

-80
ShS00
ShS05
-100
ShS15
ShS25
-120
0 2 4 6 8 10
Age (h)

3.2.3 Autogenous Shrinkage at 28 d

Figure 3.9 shows the autogenous shrinkage of the tested four mixtures up to 28 days.
The autogenous shrinkage of Mixture ShS00 was larger than that of the internally
cured mixtures. The autogenous shrinkage of internally cured mixtures decreased
by 37.2%, 64.3%, and 100% at 3 d and 22.2%, 41.7%, and 71.8% at 7 d, when the
amounts of IC water increased from 0 to 5.35, 16.06, and 26.78 kg/m3 . At 28 d,
the reductions of autogenous shrinkage were 15.1%, 41.5%, and 56.0%, which were
smaller than that at 3 d and 7 d. Therefore, the results confirmed that the effect of
IC on mitigating autogenous shrinkage was more obvious at early age, coinciding
with the findings of previous studies [50]. The autogenous shrinkage of concrete
decreased with the increase of SAPs significantly at the same age. The decrease in
autogenous shrinkage of ICC was attributed to the availability of water from the
SAPs to fill capillary voids formed by chemical shrinkage [51].
As shown in Fig. 3.10, the relationship between the ratio of the autogenous
shrinkage of internally cured mixtures with SAPs to that of ordinary mixtures and
the amount of IC water can be written as Eq. (3.6).

εicas (28)
= f (x) = 1 − αx (3.6)
εas (28)

where εicas (28) is the autogenous shrinkage of the internally cured mixtures at 28 d,
in με; εas (28) is the autogenous shrinkage of the reference ordinary mixture at 28 d,
in με; f (x) is the ratio of εicas (28) to εas (28) and is defined as the influence coefficient
that describes the effect of IC on autogenous shrinkage; and x is the amount of IC
water provided by SAPs in mixtures, in kg/m3 .
The serviceability of Eq. (3.6) is a challenge due to limited data and unde-
fined physical meaning. There is not a formula developed for all compositions. The
coefficient α was determined through regression analysis expressed by Eq. (3.7).

@seismicisolation
@seismicisolation
116 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.9 Relationship 400

between shrinkage strain and ShS00
ShS05
age of concrete 300 ShS15
ShS25

Shrinkage strain (μe)

200

100
Autogenous shrinkage
0

Net autogenous shrinkage

-100

0 5 10 15 20 25 30
Age (d)

Fig. 3.10 Relationship Test data

1.0
between the influence Fitting curve
coefficient f (x) and IC water
Influence coefficient f ( x)

of concrete at 28 d
0.9

0.8

0.7
y=1-0.0224x R2=0.971
0.6

0.5

0.4

0 5 10 15 20 25 30
IC water (kg/m3 )

εicas (28)
= f (x) = 1 − 0.0224x (3.7)
εas (28)

The values of f (x) calculated from Eq. (3.7) were 1, 0.88, 0.64, and 0.40 whereas
the test results were 1, 0.85, 0.59, and 0.44, when the amounts of IC water increased
from 0 to 5.35, 16.06, and 26.78 kg/m3 . The results of the deviations between Eq. (3.7)
and the test were 0%, 3.5%, 8.5%, and −9.1% when the amounts of IC water increased
from 0 kg/m3 to 5.35, 16.06, and 26.78 kg/m3 , respectively.

@seismicisolation
@seismicisolation
3.2 Autogenous Shrinkage of Early-Age Concrete Internally Cured with SAPs 117

3.2.4 Autogenous Shrinkage Rate

The autogenous shrinkage rate could be utilized to reflect the influence of IC with
SAPs on mitigating the self-desiccation of concrete. Regarding the cracking resis-
tance of early-age concrete, the autogenous shrinkage rate is even of greater impor-
tance than its magnitude [31]. Table 3.2 and Fig. 3.11 show the autogenous shrinkage
rate of four mixtures at different ages.
The autogenous shrinkage rate of concrete was obtained from the relationship
between autogenous shrinkage and the age of concrete [52], calculated from Eq. (3.8)
as follows.
d εicas (t)
R(t) = (3.8)
dt
where R(t) is the autogenous shrinkage rate at t d, in με/d.

Table 3.2 Autogenous shrinkage rate of four mixtures at different ages

Age (d) Autogenous shrinkage rate (με/d)
ShS00 ShS05 ShS15 ShS25
1 62 52 42 42
3 41 31 30 29
5 29 21 17 15
7 12 17 11 11
14 6 6 4 7
28 2 1 1 1

Fig. 3.11 Relationship 70

between autogenous ShS00
ShS05
Autougenous shrinkage rate (μe/d)

shrinkage rate and age of 60

ShS15
concrete
ShS25
50

40

30

20

10

0
0 5 10 15 20 25 30
Age (d)

@seismicisolation
@seismicisolation
118 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Mixture ShS00 maintained the maximum rate before the age of 7 d, followed
by Mixture ShS05 and ShS15, and Mixture ShS25 showed the slowest autogenous
shrinkage rate. The autogenous shrinkage rates of mixtures at 5 d were 29, 21, 17,
and 15 με/d when the amounts of IC water increased from 0 to 5.35, 16.06, and
26.78 kg/m3 , which decreased at the rates of 27.6%, 41.4%, and 48.3%, respectively.
For all the mixtures studied, 70% or more of the autogenous shrinkage up to 28 days
occurred 14 days after the initial setting time. Autogenous shrinkage was relatively
stable at 28 d. The autogenous shrinkage rate was minimal and was approximately
2 με/d at this time. The rate decreased with the increase in the ages of concrete and
the autogenous shrinkage rates of all mixtures were almost the same at 28 d.
The effect of IC with different amounts of SAPs on autogenous shrinkage can be
explained by additional water provided by SAP. The water in the SAPs represents
the water held within pores at the age of batching, which is available to aid IC while
not affecting the initial w/c ratio [53, 54]. Owing to water loss by self-desiccation
during the hydration process, IRH decreases and driving force occurs [55, 56]. In
this case, the IC water is released from SAPs into the surrounding cement paste,
contributing to the continued hydration of cement [50]. Thus, in terms of promoting
hydration, IC has an effect that is similar to a simple increase in the w/c ratio [57]. A
higher SAPs amount indicates that more water is released from SAP; a higher IRH of
concrete indicates that more autogenous shrinkage is reduced. In the later hydration
period, almost all the IC water in the SAPs is released, and the IRH of concrete is
essentially unchanged [55, 56]. Thus, the development of autogenous shrinkage for
all mixtures tends to stabilize at a later age.

3.2.5 Prediction Models for Autogenous Shrinkage

Investigations on the prediction and estimation of the autogenous shrinkage of

concrete [31, 58] based on several classical models [59–61] have been conducted. A
unified prediction model for autogenous shrinkage is impossible to construct because
of differences in concrete materials and conditions [31]. Many theoretical models
have been developed to predict the autogenous shrinkage of concrete, and three
theoretical models were collected from the literature.
CEB-FIP [60]:

εcas (t) = εcas0 (fcm28 ) · βas (t) (3.9)

( )2.5
fcm28 /fcm0
εcas0 (fcm28 ) = −αas × 10−6 (3.10)
6 + fcm28 /fcm0
( ( )0.5 )
t
βas (t) = 1 − exp −0.2 (3.11)
t1

@seismicisolation
@seismicisolation
3.2 Autogenous Shrinkage of Early-Age Concrete Internally Cured with SAPs 119

where εcas (t) is the autogenous shrinkage, in με; εcas0 (fcm28 ) is the notional autoge-
nous shrinkage coefficient, in με; βas (t) is the function describing the time develop-
ment of autogenous shrinkage; fcm28 is the mean compressive strength of concrete at
an age of 28 d, in MPa; fcm0 = 10 MPa; t is the concrete age, in d; t1 is taken as 1 d;
and αas is a coefficient that depends on the type of cement.
EN-1992 [61]:

εca (t) = βas (t) · εca (∞) (3.12)

εca (∞) = 2.5(fck − 10) · 10−6 (3.13)

( )
βas (t) = 1 − exp −0.2t 0.5 (3.14)

where εca (t) is the autogenous shrinkage, in με; εca (∞) is the ultimate autogenous
shrinkage, in με; βas (t) is the function describing the time development of autogenous
shrinkage; t is the concrete age; and fck is the characteristic compressive cylinder
strength of concrete at 28 d.
Tazawa’s model [59]:

ε∞ (t) = γ · ε∞ (w/b) · βas (t) (3.15)

ε∞ (w/b) = 3070 · exp[−7.2(w/b)] for 0.2 ≤ w/b ≤ 0.5 (3.16)

ε∞ (w/b) = 80 for w/b > 0.5 (3.17)

[ ]
β(t) = 1 − exp −a(t − t0 )b (3.18)

where ε∞ (t) is autogenous shrinkage at t d, in με; ε∞ (w/b) is the ultimate autogenous

shrinkage, in με; w/b is the water-to-binder ratio; γ is a cement factor; β(t) is an
age-dependent function of autogenous shrinkage; t is the age of concrete, in d; t0 is
the initial setting time, in d; a and b are constants in terms of numerical fitting of test
data.
Figure 3.12 shows the autogenous shrinkage of ordinary concrete from theoretical
models and tests study. The autogenous shrinkage of concrete at 28 d from the CEB-
FIP model was smaller than 100 με and developed slowly, which was similar to that
from the EN-1992 model. The deviations between the result from the test and that
from the model developed by Tazawa [59] were the smallest, as shown in Fig. 3.12.
Thus, the prediction model for autogenous shrinkage of ICC was shown based on
the model developed by Tazawa.
Equation (3.5) was used to predict the autogenous shrinkage of ICC with SAPs
before the time t0 at which the maximum expansion occurs. The development of
autogenous shrinkage from the time t0 to 28 d was also analyzed. The autogenous

@seismicisolation
@seismicisolation
120 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.12 Comparison of

400 Test data of ShS00
autogenous shrinkage from CEB-FIP [60]

Autogenous shrinkage (μe)

test results and theoretical EN-1992 [61]
models Tazawa [59]
300

200

100

0 5 10 15 20 25 30
Age (d)

shrinkage is the shrinkage strain measured after the initial setting time, whereas
the net autogenous shrinkage εicas (t)net is the difference between the maximum and
minimum autogenous shrinkage [44], as expressed in Eq. (3.19).

εicas net (t) = εicas (t) − εicas (t0 ) (t > t0 ) (3.19)

where εicas (t) and εicas net (t) are autogenous shrinkage and net autogenous shrinkage
of ICC at t d, in με, respectively; εicas (t0 ) is the maximum expansion at t0 d, in με.
The autogenous shrinkage after 28 d is generally believed to be small and the 28-d
autogenous shrinkage is normally considered the ultimate autogenous shrinkage in
practice [33, 49]. Thus, the net autogenous shrinkage εicas net (t) of ICC at t d can be
expressed as Eq. (3.20) based on Tazawa’s model [59].

εicas net (t) = εicas net (28) · βic (t) (3.20)

where εicas net (28) is the net autogenous shrinkage of internally cured mixtures at 28
d, in με; βic (t) is an age-dependent function of autogenous shrinkage of ICC and is
defined as the autogenous shrinkage development rate coefficient of ICC.
The development rate coefficients of autogenous shrinkage of four mixtures are
shown in Fig. 3.13. The development curves of the four mixtures were similar; thus,
the amount of SAPs supposedly had a slight effect on the development rate coefficient
βic (t). Thus, the development rate coefficient βic (t) of ICC can be expressed as
Eq. (3.21) based on the coefficient βas (t) for ordinary concrete in Tazawa’s model.

βic (t) = 1 − e−a·(t−t0 )

b
(3.21)

where t0 is the time of maximum expansion, in d; t is the age of concrete, in d; a and

b are constants in terms of numerical fitting of test data (a = 0.158, b = 0.948).

@seismicisolation
@seismicisolation
3.2 Autogenous Shrinkage of Early-Age Concrete Internally Cured with SAPs 121

Fig. 3.13 Relationship

1.0

Development rate coefficient ic(t)

between the development
rate coefficient of
autogenous shrinkage and 0.8
age of concrete
0.6

0.4
Test data of ShS00
Test data of ShS05
0.2 Test data of ShS15
Test data of ShS25
Fitting curve
0.0
0 5 10 15 20 25 30
Age (d)

Figure 3.13 shows the deviations of the development coefficient βic (t) between
Eq. (3.21) and test results were minimal; thus, Eq. (3.21) can be used to calculate
the development coefficient of autogenous shrinkage of ICC.
Substituting Eq. (3.20) into Eq. (3.19), Eq. (3.22) can be obtained as follows.

εicas (t) = εicas net (28) · βic (t) + εicas (t0 ), (t > t0 ) (3.22)

Substituting t = 28 d into Eq. (3.19), Eq. (3.23) was obtained as follows.

εicas net (28) = εicas (28) − εicas (t0 ), (t0 = 0.35 d) (3.23)

Substituting Eq. (3.7) into Eq. (3.23), Eq. (3.24) was obtained as follows.

εicas net (28) = εas (28) · (1 − 0.0224x) − εicas (t0 ), (t0 = 0.35 d) (3.24)

Substituting Eq. (3.24) into Eq. (3.22), Eq. (3.25) was obtained as follows.

εicas (t) = [εas (28) · (1 − 0.0224x) − εicas (t0 )] · βic (t) + εicas (t0 ) (t > t0 ) (3.25)

Substituting Eq. (3.21) into Eq. (3.25) and combining Eq. (3.5) results in
Eq. (3.26).
⎧
⎪ t
⎪
⎪ − · (7 + 3.73x) (t ≤ t0 ; t0 = 0.35 d)
⎨ t0
εicas (t) =
[εas (28) · (1 − 0.0224x) + 7 + 3.73x)] · (1 − e−0.158·(t−t0 )
0.948
⎪
⎪ )
⎪
⎩
−(7 + 3.73x) (t0 < t ≤ 28 d)
(3.26)

@seismicisolation
@seismicisolation
122 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.14 Fitting results of 400

autogenous shrinkage of
concrete 300

Autogenous shrinkage (μe)

200

100

0
Test data of ShS00 Fitting curve of ShS00
Test data of ShS05 Fitting curve of ShS05
-100 Test data of ShS15 Fitting curve of ShS15
Test data of ShS25 Fitting curve of ShS25

0 5 10 15 20 25 30
Age (d)

where εicas (t) is the autogenous shrinkage of ICC with SAPs at t d, in με; x is
the amount of IC water provided by SAPs in mixtures, in kg/m3 ; t0 is the time of
maximum expansion, in d; and t is the age of concrete, in d.
Figure 3.14 shows that the deviations of autogenous shrinkage between Eq. (3.26)
and test results were minimal. The model can therefore predict the autogenous
shrinkage of ICC with different amounts of SAPs from the initial setting time to
28 d considering the expansion of concrete at an early age.

3.2.6 Internal Curing Efficiency

The IC efficiency is used to evaluate the efficiency of the IC system [49, 62] and the
efficiency of LWAs can be expressed as Eq. (3.27).

Wic
η= · SR (3.27)
S · φ · WLWA

where Wic is the amount of required IC water provided by LWA, in kg/m3 ; S is the
degree of saturation of aggregate, with S = 1 for fully saturated LWA; φ is the water
absorption capacity of LWA; WLWA is the amount of LWAs, in kg/m3 ; and SR is the
percentage of autogenous shrinkage reduction.
f (x) is the ratio of the autogenous shrinkage of internally cured mixtures with
different amounts of IC water to ordinary mixture; thus, the percentage of autogenous
shrinkage reduction SR can be written as Eq. (3.28).

SR = 1 − f (x) (3.28)

@seismicisolation
@seismicisolation
3.2 Autogenous Shrinkage of Early-Age Concrete Internally Cured with SAPs 123

The required amount of water for IC to eliminate self-desiccation based on mix

proportions is 32.85 kg/m3 . Theoretically, f (x) is equal to 1 or 0, when the amount of
IC water is 0 or 32.85 kg/m3 , respectively. The percentages of autogenous shrinkage
reduction SR from the test were 15.1%, 41.5%, and 56.0% for Mixture ShS05, ShS15,
and ShS25 when the amounts of IC water increased from 0 kg/m3 to 5.35, 16.06,
and 26.78 kg/m3 . The theoretical percentages of autogenous shrinkage reduction SR
of ICC were the ratios of the IC water provided by SAPs to the required IC water,
32.85 kg/m3 . Thus, the theoretical percentages of autogenous shrinkage reduction
SR were 16.3%, 48.9%, and 81.5% for Mixture ShS05, ShS15, and ShS25. The
percentages of autogenous shrinkage reduction SR from the test were lower than
the theoretical results. Thus, not all the IC water provided by SAPs can be used to
mitigate the autogenous shrinkage. When the amounts of IC water were 5.35, 16.06,
and 26.78 kg/m3 , the IC efficiency η of SAPs from the test result were approximately
0.93, 0.85, and 0.69 based on Eq. (3.27), respectively.

3.2.7 Summary

The effect of IC on the early-age autogenous shrinkage of ICC with different amounts
of IC water provided by SAPs was analyzed. The results showed that SAPs could
be used successfully to mitigate autogenous shrinkage in low w/ c ratio HPC. The
following conclusions could be drawn based on the obtained results.
(1) The development of the autogenous shrinkage of mixtures with SAPs experi-
enced a stage of early-age expansion. The expansion of concrete was influenced
by the amount of IC water. The maximum expansion and the age at which
the maximum expansion occurred in concrete increased with the increase of IC
water provided by SAPs. The model for the expansion of concrete with different
amounts of IC water was proposed considering the age of concrete.
(2) The ultimate autogenous shrinkage of concrete at 28 d decreased with the
increase of IC water provided by SAPs. The reductions of autogenous shrinkage
at 28 d were 15.1%, 41.5%, and 56.0% when the amounts of IC water increased
from 0 to 5.35, 16.06, and 26.78 kg/m3 , respectively. The model for calculating
the ultimate autogenous shrinkage of ICC with different amounts of IC water
at 28 d was proposed.
(3) The autogenous shrinkage rate of ICC decreased with the increase of IC water
provided by SAP. The autogenous shrinkage rate of concrete at 5 d decreased at
the rates of 27.6%, 41.4%, and 48.3% when the amounts of IC water increased
from 0 kg/m3 to 5.35, 16.06, and 26.78 kg/m3 . For all the mixtures studied, 70%
or more of the autogenous shrinkage up to 28 d occurred 14 days after the initial
setting time.
(4) The model for the autogenous shrinkage of ICC with SAPs was proposed in
consideration of the early-age expansion and amount of IC water.

@seismicisolation
@seismicisolation
124 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Table 3.3 Concrete mixture proportions

Mixture composition Concrete mixtures
BF0 BS0 BS3 BS6 BS9
Water (kg/m3 ) 150 150 150 150 150
Cement (kg/m3 ) 468.75 468.75 468.75 468.75 468.75
Fine aggregate (sand) (kg/m3 ) 657.9 657.9 657.9 657.9 657.9
Coarse aggregate (kg/m3 ) 1073.4 1073.4 1073.4 1073.4 1073.4
SAPs (kg/m3 ) 0 2.917 2.917 2.917 2.917
PP fiber (kg/m3 ) 0 0 2.76 5.52 8.28
Superplasticizer (kg/m3 ) 4.69 4.69 4.69 4.69 4.69

(5) The IC efficiency was used to evaluate the effectiveness of SAPs on autogenous
shrinkage mitigation. The IC efficiency of SAPs decreased with the increase of
IC water in concrete.

3.3 Autogenous Shrinkage of Early-Age Internally Cured

Concrete Reinforced with Polypropylene Fiber

3.3.1 Mixture Proportion

As shown in Table 3.3, a total of five groups of concrete mixtures were designed, as
reported in [27]. A label “XY” is utilized to distinguish each concrete mixture. The
alphabet X in the label means the type of concrete (i.e., BF means concrete with no
addition of SAPs; BS means concrete internally cured with SAPs and reinforced with
PP fiber); and Y means the volume percentage of fiber. For example, BS3 indicates
the concrete internally cured with SAPs reinforced with 0.3% PP fiber by volume.
The basic w/ c ratio was 0.32 for all concrete mixtures.

3.3.2 Influence of Polypropylene Fiber on Early-Age

Expansion

The relationship between internal temperature and age of concrete mixtures is shown
in Fig. 3.15. The temperature rise of Mixture BF0, BS0, BS3, BS6, and BS9 was 5.04,
5.56, 5.37, 5.41, and 5.42 °C, respectively. The temperature rise changed indistinctly
with the volume percentage of PP fiber increasing because the fiber had little influence
on the temperature rise of concrete, as reported in [31, 63]. However, the temperature
rise increased with the addition of SAPs. Besides, the time when the maximum
internal temperature occurred was about 12 h after casting for ICC.

@seismicisolation
@seismicisolation
3.3 Autogenous Shrinkage of Early-Age Internally Cured Concrete … 125

Fig. 3.15 Relationship 28

between internal temperature BF0
27 BS0
and age of concrete mixtures BS3
BS6

Temperature (°C)
26 BS9
Ambient
25

24

23

22

0 12 24 36 48 60 72
Age (h)

The autogenous shrinkage within 28 d was analyzed. Figure 3.16 shows the rela-
tionship between autogenous shrinkage and age of concrete mixtures. A three-stage
pattern of the development of autogenous shrinkage was observed from the initial
setting time. In the first stage, the development exhibited expansion before 24 h for
ICC and showed minimal and rapid expansion for Mixture BF0 after the initial setting
time. The expansion could be explained by the reabsorption of bleeding water, and
the growth of further hydration reaction products inside the network that generated an
internal pressure leading to the expansion, as reported in [64]. The maximum expan-
sion was larger for concrete internally cured with SAPs than Mixture BF0, possibly
due to the additional IC water provided by SAPs. After the expansion period, the
autogenous shrinkage of concrete turned into the second stage and increased rapidly
within a few days. The development of autogenous shrinkage rate then drastically
decreased after 7 d, followed by a relatively stable shrinkage stage with a gradually
declining growth rate. A significant decrease of autogenous shrinkage was observed
up to 28 d in concrete internally cured with SAPs compared with that of Mixture BF0,
resulting from the influence of early-age expansion and the subsequent decrease in
the development of autogenous shrinkage rate.
Early-age expansion that could reduce autogenous shrinkage occurred after the
initial setting time. As shown in Fig. 3.17, the maximum expansion of ICC decreased
with the volume percentage of PP fiber increasing. The maximum expansion of ICC
was −82, −70, −62, and −56 με, the absolute value of which decreased by 14.6%,
24.4%, and 31.7% when the volume percentage of PP fiber increased from 0% to
0.3%, 0.6%, and 0.9%, respectively. The main reason was that the tensile strength
was improved with the volume percentage of PP fiber increasing, thus improving the
ability of ICC to resist expansion. The average time that the maximum expansion
of concrete occurred was 0.5 d, which was utilized to propose the equation when
obtaining the maximum expansion of ICC.
For analysis of the autogenous shrinkage and the establishment of the calculation
model, the development of autogenous shrinkage was separated into two stages:

@seismicisolation
@seismicisolation
126 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

400 250

200
( )

Autogenous shrinkage ( )
300
150
Autogenous shrinkage

200 100

100 50

BF0
0 BF0
BS0
BS0
0 BS3
BS3 -50 BS6
BS6
BS9 BS9
-100 -100
0 5 10 15 20 25 30 0 24 48 72 96 120 144 168
Age (d) Age (h)
(a) 28 d (b) 7 d

Fig. 3.16 Relationship between autogenous shrinkage and age of concrete mixtures

Fig. 3.17 Fitting curve of

the relationship between the
maximum expansion and
volume content of ICC

expansion stage and shrinkage stage. The maximum expansion of ICC could be
obtained with Eq. (3.29).

εicc (t0 ) = l + m · e−θ/n (3.29)

where εicc (t0 ) is the maximum expansion of ICC, in με; t0 is the time when the
maximum expansion occurs, in d; and θ is the volume percentage of PP fiber, in
%. The coefficients l, m and n are obtained from the regression analysis with the
coefficient of determination (R2 ) greater than 0.999 (l = −42.667, m = −39.302,
and n = 0.836), as shown in Fig. 3.17.
The relationship between the expansion and age of concrete could be considered
linear, as shown in Eq. (3.30).

@seismicisolation
@seismicisolation
3.3 Autogenous Shrinkage of Early-Age Internally Cured Concrete … 127

εicc (t0 )
εicc (t) = t(t ≤ t0 ) (3.30)
t0

where t is the age of concrete, in d; and εicc (t) is the expansion of ICC at t d, in με.
The expansion of Mixture BS0, BS3, BS6, and BS9 at 0.42 d calculated by Eqs.
(3.29) and (3.30) was −68, −58, −52, and −47 με, respectively. The deviations
from the above equations and the test data were 12.8%, 14.7%, 7.1%, and 11.3% for
Mixture BS0, BS3, BS6, and BS9, respectively.

3.3.3 Influence of Polypropylene Fiber on Autogenous

Shrinkage

The development of autogenous shrinkage was divided into expansion and shrinkage
stages, and the development of autogenous shrinkage from t0 to 28 d was analyzed
in this section. The autogenous shrinkage of Mixture BF0, BS0, BS3, BS6, and
BS9 at 28 d was 361, 173, 139, 119, and 104 με, respectively. The autogenous
shrinkage decreased significantly with the addition of SAPs at the same age, which
was attributed to the decrease in the self-desiccation when SAPs were utilized, as
reported in [65]. Besides, the autogenous shrinkage decreased by 19.7%, 31.2%, and
39.9% when the volume percentage of PP fiber increased from 0% to 0.3%, 0.6%,
and 0.9%, respectively.
Scanning electron microscope images of Mixture BS0 and BS3 are shown in
Fig. 3.18. For Mixture BS0, obvious initial cracks with relatively large crack width
could be observed in the matrix. The aggregate-paste interface was loose, and thus
microcracks were produced around the interface. For Mixture BS3, the matrix phase
was significantly denser and more homogeneous than that of Mixture BS0 in general.
The fiber-paste interface was dense, which indicated that the bond strength between
PP fiber and the paste was strong so that PP fiber could play the full role in strength-
ening the matrix and reducing the autogenous shrinkage of concrete. With the addi-
tion of PP fiber, the crack width and the number of cracks decreased obviously,
which restrained the propagation of microcracking. The reasons were as follows. On
one hand, the complex three-dimensional random distribution of PP fiber within the
concrete resulted in the reduction of crack formation and propagation, as reported
in [66, 67]. On the other hand, water formed an adsorption film on the surface of
PP fiber with the hardening of concrete, which blocked the passage of water over-
flow and slowed down the water loss, thereby reducing the autogenous shrinkage.
Thus, PP fiber plays an important role in alleviating the formation and propagation
of microcracks and increasing the cracking resistance of concrete. With the addition
of PP fiber, the pore structure was improved, thus reducing the capillary pressure and
autogenous shrinkage of concrete, as reported in [68, 69].
The autogenous shrinkage after 28 d is commonly believed to be small, and the
autogenous shrinkage at 28 d is generally taken as the ultimate autogenous shrinkage
in reality [33]. The net autogenous shrinkage of ICC at t d can be expressed with

@seismicisolation
@seismicisolation
128 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

(a) BS0 (b) BS3

Fig. 3.18 Scanning electron microscope images of the concrete mixtures

Eq. (3.20) and Eqs. (3.31) and (3.32).

εicc (28)net = εicc (28) − εicc (t0 ) (3.31)

εicc (28)
= fic (θ ) = p + q · e−θ/r (3.32)
εicc
0
(28)

where εicc (28)net is the net autogenous shrinkage of ICC at 28 d, in με; εicc (28) is the
autogenous shrinkage of ICC at 28 d, in με, εicc 0
(28) is the autogenous shrinkage of
ICC at 28 d with no addition of fiber, in με, and the influence coefficient fic (θ ) is the
ratio of εicc (28) to εicc
0
(28), which describes the influence of PP fiber on autogenous
shrinkage. The coefficients p, q and r are obtained from the regression analysis with
the coefficient of determination (R2 ) greater than 0.999 (p = 0.512, q = 0.488, and
r = 0.657), as shown in Fig. 3.19.
The development rate coefficients on autogenous shrinkage of four groups of
concrete mixtures are shown in Fig. 3.20. The development trends of four groups of
concrete mixtures were similar; thus, the volume percentage of PP fiber supposedly
had little influence on the development rate coefficient βic (t). Thus, the development
rate coefficient βic (t) of ICC can be obtained with Eq. (3.33) based on the Tazawa-
Miyazawa model [59].

βic (t) = 1 − e−aic ·(t−t0 )

bic
(3.33)

where the coefficients aic and bic are obtained from the regression analysis with the
coefficient of determination (R2 ) greater than 0.993 (aic = 0.196 and bic = 0.796),
as shown in Fig. 3.20. The deviations of the development coefficient βic (t) between
Eq. (3.33) and test data were small; thus, Eq. (3.33) could be utilized to obtain the
development coefficient on the autogenous shrinkage of ICC.

@seismicisolation
@seismicisolation
3.3 Autogenous Shrinkage of Early-Age Internally Cured Concrete … 129

Fig. 3.19 Fitting curve of

the relationship between the
influence coefficient fic (θ )
and volume content of ICC

Fig. 3.20 Fitting curve of

the relationship between the
development rate coefficient
βic (t) and age of ICC

Figure 3.21 shows the comparison of autogenous shrinkage obtained from test
and prediction models. Result reported in [70] shows that the autogenous shrinkage
at 28 d is 139 and 123 με for ICC reinforced with 0% and 0.01% PP fiber by volume,
respectively. The deviations between the above equations and the test data in [70]
were 6.5% and 4.9% for ICC reinforced with 0% and 0.01% PP fiber by volume,
respectively.

@seismicisolation
@seismicisolation
130 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.21 Comparison of 200

autogenous shrinkage
obtained from test and

)
150
prediction models

Autogenous shrinkage strain (

100

50
Test data of BS0
Test data of BS3
0 Test data of BS6
Test data of BS9
Predictio n result of BS0
-50 Predictio n result of BS3
Predictio n result of BS6
Predictio n result of BS9
-100
0 5 10 15 20 25 30
Age (d)

3.3.4 Influence of Polypropylene Fiber on Ultrasonic Velocity

To obtain the ultrasonic velocity at the age corresponding to the age of autogenous
shrinkage, the test of ultrasonic velocity was also conducted from the initial setting
time. The relationship between ultrasonic velocity and age of concrete mixtures is
shown in Fig. 3.22. The development trends of five groups of concrete mixtures were
basically the same. The ultrasonic velocity increased rapidly in the early stage, and
the ultrasonic velocity at 1 d reached 80% of that at 28 d. At the later stage, the
growth rate gradually became slow, and the ultrasonic velocity of concrete became
relatively stable at 28 d. The ultrasonic velocity of Mixture BF0, BS0, BS3, BS6, and
BS9 at 28 d was 5183, 5149, 5105, 5064, and 5035 m/s, respectively. The ultrasonic
velocity decreased with the addition of SAPs at the same age. The reason was that
the addition of SAPs decreased the density of concrete, thus prolonging the propaga-
tion time of the ultrasonic wave in concrete. Also, the ultrasonic velocity decreased
gradually with the volume percentage of PP fiber increasing. The phenomenon could
be explained from the following aspects. On one hand, PP fiber offers resistance to
wave propagation, thus prolonging the propagation time of the ultrasonic wave in
concrete, and decreasing the corresponding wave velocity [71]. On the other hand, the
ultrasonic velocity is directly proportional to the density of the propagation medium
[72]. Adding more fibers to the concrete mixture makes larger capillary pores inside
the concrete, thus decreasing the density and the ultrasonic velocity of concrete.
As shown in Fig. 3.22, the development trend of time-evolution ultrasonic velocity
was similar to that of time-evolution autogenous shrinkage.
A model similar to the prediction model of autogenous shrinkage was proposed
to predict the ultrasonic velocity of ICC reinforced with PP fiber, as shown in Eqs.
(3.34)–(3.35).

νic (t) = νic (28)βν (t) (3.34)

@seismicisolation
@seismicisolation
3.3 Autogenous Shrinkage of Early-Age Internally Cured Concrete … 131

Fig. 3.22 Relationship 5200

between ultrasonic velocity
and age of concrete mixtures 5000

Ultrasonic velocity (m/s)

4800

4600

4400
BF0
4200
BS0
BS3
4000 BS6
BS9
3800
0 5 10 15 20 25 30
Age (d)

vic (28)
= fv (θ ) = u + w · θ (3.35)
vic0 (28)

where νic (t) is the ultrasonic velocity of ICC at t d, in m/s; νic (28) is the ultrasonic
velocity of ICC at 28 d, in m/s; vic0 (28) is the ultrasonic velocity of ICC at 28 d
with no addition of fiber, in m/s, βv (t) is an age-dependent parameter on ultrasonic
velocity, and the influence coefficient fv (θ ) is the ratio of νic (28) to vic0 (28), which
describes the influence of PP fiber on ultrasonic velocity. The coefficients u and w
are obtained from the regression analysis with the coefficient of determination (R2 )
greater than 0.988 (u = 0.999 and w = −0.025), as shown in Fig. 3.23.
The development rate coefficients on ultrasonic velocity of four groups of concrete
mixtures are shown in Fig. 3.24. The development trends of four groups of concrete

Fig. 3.23 Fitting curve of

the relationship between the
influence coefficient fv (θ )
and volume content of ICC

@seismicisolation
@seismicisolation
132 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.24 Fitting curve of

the relationship between the
development rate coefficient
βv (t) and age of ICC

mixtures were similar; thus, the volume percentage of PP fiber supposedly had little
influence on the development rate coefficient βv (t). Thus, the development rate coef-
ficient βv (t) of ICC can be obtained with Eq. (3.36) based on Tazawa-Miyazawa
model [59].

βv (t) = 1 − e−av ·t
bv
(3.36)

where the coefficients av and bv are obtained from the regression analysis with
coefficient of determination (R2 ) greater than 0.994 (av = 1.595 and bv = 0.386).
Figure 3.25 shows the comparison of ultrasonic velocity obtained from test and
prediction models. Result reported in [38] shows that the ultrasonic velocity at 28
d is 4100, 4040, 3940, and 3890 m/s for concrete mixtures reinforced with 0%,
0.25%, 0.50% and 0.75% PP fiber by volume, respectively. The deviations between
the above equations and the test data in [38] were 0.4%, 0.5%, 2.3%, and 3.0% for
concrete mixtures reinforced with 0%, 0.25%, 0.50% and 0.75% PP fiber by volume,
respectively.

3.3.5 Prediction Model of Autogenous Shrinkage Based

on Ultrasonic Velocity

Studies on the prediction of the autogenous shrinkage of concrete with the existing
classical models have been carried out [59–61]. Several theoretical models have been
established taking account of some factors such as cement composition, admixture
type, compressive strength, and the water-cement ratio of concrete. If the prediction
model of early-age autogenous shrinkage based on the results of ultrasonic velocity
is proposed, the autogenous shrinkage of ICC reinforced with PP fiber at various
ages can be predicted easily and quickly in actual projects.

@seismicisolation
@seismicisolation
3.3 Autogenous Shrinkage of Early-Age Internally Cured Concrete … 133

Fig. 3.25 Comparison of 5200

ultrasonic velocity obtained
from test and prediction 5000

Ultrasonic velocity (m/s)

models
4800

4600
Test data of BS0
4400 Test data of BS3
Test data of BS6
4200 Test data of BS9
Prediction result of BS0
Prediction result of BS3
4000 Prediction result of BS6
Prediction result of BS9
3800
0 5 10 15 20 25 30
Age (d)

The relationship between the autogenous shrinkage and ultrasonic velocity of

ICC was analyzed, which exhibited the same trend for all mixtures, as shown in
Fig. 3.26. With ultrasonic velocity increasing, the autogenous shrinkage of ICC
also increased gradually. Owing to the cement hydration and the decrease of the
humidity of the interior ICC, self-desiccation was induced and thus increasing the
autogenous shrinkage of ICC. Besides, the density of concrete gradually increased
with the formation of hydration products, and the corresponding ultrasonic velocity
showed an upward trend. The power function was utilized to predict the autogenous
shrinkage of ICC based on the results of ultrasonic velocity at different ages and at
28 d with no addition of PP fiber, as shown in Eq. (3.37).
[ ( )]β
εicc (t) = α · vic (t)/ vic0 (28)fv (θ ) (3.37)

Fig. 3.26 Relationship 200

)

between autogenous
shrinkage and ultrasonic
Autogenous shrinkage strain (

velocity of ICC 150

100

50
BS0
BS3
0 BS6
BS9

4600 4700 4800 4900 5000 5100 5200

Ultrasonic velocity (m/s)

@seismicisolation
@seismicisolation
134 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Figure 3.27 shows the fitting curve of the relationship between autogenous
shrinkage and ultrasonic velocity of ICC. The deviations of autogenous shrinkage
between Eq. (3.37) and test data were small. The equations were deduced with coef-
ficients of determination (R2 ) all greater than 0.977. The coefficients α and β are
obtained from the regression analysis, as shown in Eqs. (3.38) and (3.39).

α = α0 − 78.232 · θ (3.38)

β = β0 − 9.619 · θ (3.39)

where α0 is taken as 162.643 and β0 is taken as 42.215, which represent the

coefficients α and β for Mixture BS0, respectively.
The autogenous shrinkage of Mixture BS0, BS3, BS6, and BS9 at 28 d calculated
by Eqs. (3.37)–(3.39) was 170, 139, 114, and 97 με, respectively. The deviations
between the above equations and the test data were 1.7%, 0%, 4.2%, and 6.7% for
ICC reinforced with 0%, 0.3%, 0.6%, and 0.9% PP fiber by volume, respectively.
The proposed model for estimating the early-age autogenous shrinkage of ICC was

Fig. 3.27 Fitting curve of the relationship between autogenous shrinkage and ultrasonic velocity
of ICC

@seismicisolation
@seismicisolation
3.3 Autogenous Shrinkage of Early-Age Internally Cured Concrete … 135

simple. The technician could test an existing specimen on a structure, measure its
ultrasonic velocity, and estimate the autogenous shrinkage. However, this model was
only correlated with the ultrasonic velocity. It had limited value since autogenous
shrinkage could not be predicted based on time.
Substituting Eqs. (3.34) and (3.35) into Eq. (3.37) and combining Eqs. (3.36),
(3.38), and (3.39), the prediction model of early-age autogenous shrinkage of
ICC considering the fiber volume percentage and age based on the age-dependent
parameter on ultrasonic velocity was proposed, as shown in Eq. (3.40).
⎧
⎪
⎪ ε (t) = α · βv (t)β
⎪ icc
⎪
⎨
βv (t) = 1 − e−1.595·t
0.386

(3.40)
⎪
⎪ α = α0 − 78.232 · θ
⎪
⎪
⎩
β = β0 − 9.619 · θ

The autogenous shrinkage of Mixture BS0, BS3, BS6, and BS9 at 28 d calcu-
lated by Eq. (3.40) was 143, 123, 103, and 83 με, respectively. The deviations from
Eq. (3.40) and the test data were 17.3%, 11.5%, 13.4%, and 20.2% for ICC reinforced
with 0%, 0.3%, 0.6%, and 0.9% PP fiber by volume, respectively. The deviations
were within the acceptable range, and the proposed model could indicate the devel-
opment trend of autogenous shrinkage. Thus, Eq. (3.40) was rational to predict the
early-age autogenous shrinkage of ICC considering the fiber volume percentage and
age based on the age-dependent parameter on ultrasonic velocity.
Three models for obtaining the autogenous shrinkage of ICC were proposed.
Equations (3.29)–(3.33) could be utilized to predict the early-age autogenous
shrinkage of ICC reinforced with PP fiber considering the fiber volume percentage
and age based on autogenous shrinkage at 28 d. Equations (3.37)–(3.39) could be
utilized to obtain the early-age autogenous shrinkage of ICC based on the results of
ultrasonic velocity at different ages and at 28 d with no addition of PP fiber, which
was relatively simple and could be applied in actual projects quickly. Equation (3.40)
could be utilized to predict the early-age autogenous shrinkage of ICC considering
the fiber volume percentage and age based on the age-dependent parameter on ultra-
sonic velocity, which could be utilized to monitor the early-age autogenous shrinkage
of ICC continuously, thus providing a basis for evaluating the early-age performance
of ICC.

3.3.6 Summary

The autogenous shrinkage of concrete reinforced with PP fiber was investigated. The
relationship between autogenous shrinkage and ultrasonic velocity of ICC reinforced
with PP fiber was also analyzed. Several prediction models were proposed to estimate
the autogenous shrinkage of ICC. The conclusions were listed as follows.

@seismicisolation
@seismicisolation
136 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

(1) The autogenous shrinkage at 28 d was 173, 139, 119, and 104 με for ICC rein-
forced with 0%, 0.3%, 0.6%, and 0.9% PP fiber by volume, respectively; thus,
it decreased with the volume percentage of PP fiber increasing. The predic-
tion model of early-age autogenous shrinkage of ICC reinforced with PP fiber
considering the fiber volume percentage and age based on autogenous shrinkage
at 28 d was proposed.
(2) The ultrasonic velocity at 28 d was 5149, 5105, 5064, and 5035 m/s for ICC
reinforced with 0%, 0.3%, 0.6%, and 0.9% PP fiber by volume, respectively;
thus, it decreased with the volume percentage of PP fiber increasing. The predic-
tion model of early-age ultrasonic velocity of ICC reinforced with PP fiber was
proposed considering the fiber volume percentage and age.
(3) The model of early-age autogenous shrinkage of ICC based on the results of
ultrasonic velocity at different ages and at 28 d with no addition of PP fiber was
proposed, as well as the prediction model of early-age autogenous shrinkage
of ICC considering the fiber volume percentage and age based on the age-
dependent parameter on ultrasonic velocity.
(4) The addition of SAPs may be considered a momentous factor in the investigation
of early-age performance of concrete. Although the addition of SAPs weakened
the mechanical properties of ICC to some extent, SAPs could reduce the early-
age autogenous shrinkage of ICC.
(5) Method of non-destructive testing was adopted to test the ultrasonic velocity,
which provided theoretical support for the prediction of the autogenous
shrinkage of ICC at various ages in actual projects.

3.4 Relationship Between Autogenous Shrinkage

and Internal Relative Humidity of Concrete Internally
Cured with SAPs

3.4.1 Mixture Proportion

As shown in Table 3.4, the coarse aggregate was crushed limestone with a maximum
grain size of 20 mm. The fine aggregate was natural river sand. The superplasticizer
was incorporated to improve the workability of the concrete mixtures. The SAPs
with the grain size in the range of 125–150 μm under dry conditions and a dry-bulk
density of 850 kg/m3 were utilized. The absorption value of SAPs was 13.0 g water
per g SAP.

@seismicisolation
@seismicisolation
3.4 Relationship Between Autogenous Shrinkage and Internal Relative … 137

Table 3.4 Concrete mixture proportions

Mixture composition Concrete mixtures
SAP00 SAP35 SAP70 SAP100
Water (kg/m3 ) 171 171 171 171
Cement (kg/m3 ) 512 512 512 512
Fine aggregate (kg/m3 ) 636 636 636 636
Coarse aggregate (kg/m3 ) 1131 1131 1131 1131
SAPs (kg/m3 ) 0 0.88 1.77 2.53
IC water (kg/m3 ) 0 11.50 23.00 32.85
Superplasticizer (kg/m3 ) 3.3 3.3 3.3 3.3

3.4.2 Internal Relative Humidity of Concrete with SAPs

The IRH changes for four groups of specimens within 28 days are shown in Fig. 3.28.
The IRH of specimens with SAPs was higher than that of the reference specimen
and increased as the dosages of SAPs increased. The IRH at the age of 14 d was
94.9%, 96.8%, 98.7%, and 99.7% RH when the quantity of IC water was 0 (SAP00),
11.50 (SAP35), 23.00 (SAP70), and 32.85 (SAP100) kg/m3 , respectively. The IRH
at 14 d for specimens SAP35, SAP70, and SAP100 increased by 2.0%, 4.0%, and
5.1%, respectively, compared with the reference specimen. The IRH at the age of 21
d was 91.3%, 94.3%, 96.7%, and 98.1% RH, which increased by 3.3%, 5.9%, and
7.4% with the increasing quantity of IC water from 0 kg/m3 to 11.50, 23.00, and
32.85 kg/m3 , respectively. The IRH at the age of 28 d was 88.7%, 92.4%, 95.2%,
and 97.0% RH, which increased by 4.2%, 7.3%, and 9.4% for specimens SAP35,
SAP70, and SAP100, respectively, compared with SAP00. Thus, the incorporation
of SAPs had a strong effect on the IRH of HSC under sealed conditions The SAPs
distributed in the whole cross-section of concrete specimens could provide additional
water to compensate for the early-age moisture loss due to the cement hydration
reaction, whereas the reference specimen had no additional water to complement
the water consumption. Thus, the specimens with SAPs maintained a high level of
IRH as compared with the reference specimen. The IC water desorbed from SAPs is
insufficient to counteract the water consumed by cement hydration [55]. Thus, the
IRH of specimens with SAPs also decreased.
The critical time was 5, 7, 9, and 12 d, which increased by 2, 4, and 7 d with
the increasing quantity of IC water from 0 kg/m3 to 11.50, 23.00, and 32.85 kg/m3 ,
respectively. Compared with SAP00, the reductions of IRH of ICC with SAPs were
significantly reduced and delayed, which indicated that the incorporation of SAPs
could directly affect the development of early-age IRH by releasing the IC water
entrained in SAPs to the surrounding cement paste. As the dosages of SAPs increased,
the IC water needed a prolonged period to release, and the subsequent self-desiccation
was compensated for and delayed. Thus, the critical time of ICC with SAPs increased
as the dosages of SAPs increased. The water desorbing from the pre-soaked LWAs

@seismicisolation
@seismicisolation
138 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.28 Relationship 100

between IRH and age of
concrete 98
SAP100
96

IRH (%)
SAP70
94

92 SAP35

90 SAP00
SAP35
88 SAP70 SAP00
SAP100
86
0 5 10 15 20 25 30
Age (d)

could compensate for the water loss due to water diffusion or cement hydration, and
the internal moisture environment of concrete was improved by IC aggregate.

3.4.3 Early-Age Expansion of Concrete with SAPs

All the concrete specimens were cured at a room temperature of 20 °C. However,
the temperatures inside the specimens were inconsistent. The temperature histories
of the four groups of concrete specimens before 3 d are shown in Fig. 3.29.
The central temperatures in specimens reached the ambient temperature at 3 d.
Thus, the temperature histories after 3 d were not given in the figure. The maximum
temperatures in specimens SAP35, SAP70, and SAP100 were higher than that of the

Fig. 3.29 Relationship 28

between central temperature SAP00
and age of concrete SAP35
SAP70
26 SAP100
Temperature (ºC)

Ambient

24

22

20
0 12 24 36 48 60 72
Age (h)

@seismicisolation
@seismicisolation
3.4 Relationship Between Autogenous Shrinkage and Internal Relative … 139

Fig. 3.30 Relationship 400

between autogenous
shrinkage and age of SAP00
300
concrete

Autogenous shrinkage ( )
200 SAP35

SAP70
100
SAP100

0
SAP00
SAP35
SAP70
-100
SAP100

0 5 10 15 20 25 30
Age (d)

reference specimen and increased as the dosages of SAPs increased. An explanation

may be that the IC water released from SAPs increased the early-age hydration
reaction degree of concrete, which resulted in high hydration heat. Furthermore, the
time when IC mixtures reached the maximum temperature increased as compared
with reference concrete.
The calculated autogenous shrinkages of the four concrete mixtures are shown in
Fig. 3.30.
Compared with SAP00, a larger expansion was observed in specimens with SAPs,
and the expansion peak of ICC with SAPs increased as the dosages of SAPs increased.
The absolute value of expansion peak was 9, 62, 97, and 141 με, which increased at
the rates of 5.89, 9.78, and 14.67 times with the increasing quantity of IC water from
0 kg/m3 to 11.50, 23.00, and 32.85 kg/m3 , respectively. The occurrence of early-age
expansion is mainly due to the formation and growth of large-size Ca(OH)2 crystals
[45, 46], and the incorporation of SAPs may modify the formation of portlandite
[41]. Furthermore, the water entrained in SAPs could participate in the hydration
reaction of the surrounding cement paste, and the volume of hydration products was
larger than the reacted cement. Thus, the early-age expansion deformation occurred
and offset the subsequent volume reduction due to shrinkage. The time when the
expansion peak occurred increased as the dosages of SAPs increased. The expansion
peak occurred at 7.2, 8.4, 9.6, and 10.8 h for specimens SAP00, SAP35, SAP70, and
SAP100, respectively.

3.4.4 Autogenous Shrinkage of Concrete with SAPs

After the early-age expansion stage, the autogenous shrinkage development of

concrete turned into stage II and began to shrink. The length of the first stage was 0.4,

@seismicisolation
@seismicisolation
140 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.31 Relationship 400

between net autogenous
shrinkage and age of SAP00
concrete

Net autogenous shrinkage ( )

300
SAP35

SAP70

200 SAP100

100 SAP00
SAP35
SAP70
SAP100
0
0 5 10 15 20 25 30
Age (d)

2.3, 3.5, and 5.8 d for specimens SAP00, SAP35, SAP70, and SAP100, respectively.
The duration of the expansion stage increased as the dosages of SAPs increased. In
other words, the time for concrete to turn into the shrinkage stage was delayed as the
quantity of IC water entrained in SAPs increased. The net autogenous shrinkage of
the four groups of concrete specimens within 28 days is shown in Fig. 3.31, and the
shrinkage deformations discussed in later sections were net autogenous shrinkage.
The net autogenous shrinkage of ICC with SAPs was lower than that of reference
concrete and decreased as the dosages of SAPs increased. As shown in Table 3.5,
the net autogenous shrinkage at the age of 3 d was 147, 132, 126, and 111 με, which
decreased by 10.2%, 14.3%, and 24.5% with the increasing quantity of IC water
from 0 kg/m3 to 11.50, 23.00, and 32.85 kg/m3 , respectively. The net autogenous
shrinkage at the age of 28 d was 362, 305, 257, and 223 με, which decreased by
15.7%, 29.0%, and 38.4% with the increasing quantity of IC water from 0 kg/m3 to
11.50, 23.00, and 32.85 kg/m3 , respectively. The net autogenous shrinkage decreased
significantly as the dosages of SAPs increased at the same age, and the extent of
reduction was more obvious at later ages. An explanation for the reduction of net
autogenous shrinkage in ICC with SAPs may be that the IC water entrained in
SAPs fills the capillary pores formed by chemical shrinkage [51]. The chemical
shrinkage reduces concrete volume during cement hydration and is the main reason
for autogenous deformation. Furthermore, the supplement of additional IC water
alleviated the shortage of water consumed by hydration reaction in concrete, which
reduced the effect of self-desiccation on the early-age concrete. Then the drops of
IRH were postponed, and the driving force of autogenous shrinkage was reduced.
The autogenous shrinkage of HSC was much higher than that of normal concrete at
the same age. The IRH of concrete under sealed condition decreases due to the effect
of self-desiccation [47]. Significant self-desiccation occurred because the amount of
water is insufficient to sustain the cement hydration for HSC. Therefore, the reduction
of the IRH was more significant in HSC than that in normal strength concrete. The

@seismicisolation
@seismicisolation
3.4 Relationship Between Autogenous Shrinkage and Internal Relative … 141

Table 3.5 Net autogenous shrinkage of four mixtures at different ages

Net autogenous shrinkage (με) Concrete mixtures
SAP00 SAP35 SAP70 SAP100
3d 147 132 126 111
7d 223 212 181 159
14 d 282 254 218 193
21 d 330 289 244 215
28 d 362 305 257 223

shrinkage strain is inversely proportional to the pore humidity [73]. Therefore, the
autogenous shrinkage of HSC was much higher than that of normal concrete at the
same age.
The autogenous shrinkage rates of four concrete mixtures are shown in Fig. 3.32.
Before the age of 7 d, the autogenous shrinkage rate of ICC with SAPs was lower
than that of the reference mixture and decreased as the dosages of SAPs increased.
The autogenous shrinkage rate at the age of 3 d was 37, 34, 29, and 25 με/d, which
decreased by 8.1%, 21.6%, and 32.4% with the increasing quantity of IC water from
0 kg/m3 to 11.50, 23.00, and 32.85 kg/m3 , respectively. The autogenous shrinkage
rates of all the concrete specimens were lower than 10 με/d at 14 d and tended
to be stabilized at basically 1–2 με/d at 28 d. Moreover, the autogenous shrinkage
at 14 d reached 77.9%, 83.3%, 84.8%, and 86.5% of the shrinkage value at 28 d
for specimens SAP00, SAP35, SAP70, and SAP100, respectively. Thus, it could be
concluded that most of the autogenous shrinkage deformations developed 14 d after
the initial setting time.
The hydration reaction rate of cement-based materials was very fast at early age.
A large amount of mixing water was consumed during the hydration process, which

Fig. 3.32 Relationship 60

between autogenous SAP00
shrinkage rate and age of SAP35
Autogenous shrinkage rate ( /d)

concrete SAP70
SAP100
40

20

0
0 5 10 15 20 25 30
Age (d)

@seismicisolation
@seismicisolation
142 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Table 3.6 Autogenous shrinkage rate of four mixtures at different ages

Autogenous shrinkage rate (με/d) Concrete mixtures
SAP00 SAP35 SAP70 SAP100
1d 51 47 44 40
3d 37 34 29 25
5d 18 18 14 11
7d 14 10 8 6
14 d 5 6 6 4
21 d 9 5 3 3
28 d 2 1 1 1

increased the self-desiccation and decreased the IRH in concrete. Under the driving
force of moisture loss, the water absorbed in SAPs is released to the nearby cement
paste to promote cement hydration [74]. The quantity of the IC water increased as
the dosages of SAPs increased, hence for specimens with larger contents of SAP, the
IRH in concrete was higher, and more autogenous shrinkage was reduced. Results
reported in [74] show that the SAPs particles can desorb most of the IC water after
7 d of cement hydration. Furthermore, in the later period of cement hydration, the
variation of the IRH in concrete is low [55, 56]. Accordingly, as shown in Table
3.6, the autogenous shrinkage rates of four concrete mixtures were low after 7 d and
essentially unchanged at later ages.

3.4.5 Correlation Between Autogenous Shrinkage

and Internal Relative Humidity of Concrete with SAPs

The internal free water in concrete was consumed because of the progressive cement
hydration reaction, which would lead to a decrease in IRH and increase the autoge-
nous shrinkage. The correlations between the IRH and autogenous shrinkage of ICC
with SAPs are shown in Fig. 3.33 and could be estimated by Eq. (3.41).

εas = ε0 + k(100 − RH) RH < 100% (3.41)

where ε0 is the autogenous shrinkage developed during the moisture saturated stage,
which is mainly related to the w/c ratio, in με; k is autogenous shrinkage generated
by one-unit RH reduction.
The correlation between the IRH and the autogenous shrinkage of concrete spec-
imens with four groups of IC water contents could be determined by the regressive
analysis and could be expressed by Eqs. (3.42)–(3.45).
For reference concrete specimen SAP00 with no IC water:

@seismicisolation
@seismicisolation
3.4 Relationship Between Autogenous Shrinkage and Internal Relative … 143

Fig. 3.33 Relationship between autogenous shrinkage and IRH

εas = 211 + 13.8(100 − RH) R2 = 0.986 (3.42)

For concrete specimen SAP35 with IC water of 11.50 kg/m3 :

εas = 209 + 13.5(100 − RH) R2 = 0.988 (3.43)

For concrete specimen SAP70 with IC water of 23.00 kg/m3 :

εas = 198 + 13.3(100 − RH) R2 = 0.976 (3.44)

For concrete specimen SAP100 with IC water of 32.85 kg/m3 :

εas = 187 + 12.8(100 − RH) R2 = 0.978 (3.45)

The R2 values obtained from Eqs. (3.42)–(3.45) were 0.986, 0.988, 0.976, and
0.978 for concrete specimens SAP00, SAP35, SAP70, and SAP100, respectively,
which indicated that these equations had a good correlation with the test results,
and the autogenous shrinkage generated in stage II was well correlated with IRH of

@seismicisolation
@seismicisolation
144 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

concrete. Besides, the parameter k of ICC with SAPs decreased as the dosages of
SAPs increased. An explanation may be that the incorporation of SAP, which was
much softer than the hydrated matrix, made the elastic modulus of ICC with SAPs
low compared with reference concrete. The ε0 decreased as the quantity of IC water
increased. Thus, Eq. (3.41) could be utilized to predict the autogenous shrinkage of
ICC with SAPs if the IRH distribution of concrete was determined.
During the moisture saturated stage, the deformation of concrete was mainly
related to the chemical shrinkage induced during this period. Because the stiffness
of concrete at this age is relatively low, the macroscopic shrinkage should be closely
related to chemical shrinkage [75]. The internal skeleton structure of concrete grad-
ually formed as the hydration reaction of cement-based materials continued, the
capillary pores between cement particles occurred. After the connectivity of capil-
lary pores was gradually limited and destroyed by the skeleton structure, the IRH of
concrete began to decline, and the corresponding capillary pore force was induced.
The autogenous shrinkage of concrete after the moisture saturated stage is mainly
caused by the capillary pore force inside the concrete [76]. When the IRH of concrete
was known, the autogenous shrinkage generated in stage II could be calculated
through Eq. (3.41). However, the autogenous shrinkage generated in stage I cannot
be predicted. Accordingly, a two-stage autogenous shrinkage model considering the
quantity of IC water and concrete RH was developed.
Following the assumption that the shape of the capillary pores is nearly circular
cylindrical, a model for predicting the autogenous shrinkage induced by capillary
tensile stress through the IRH of cement paste in the humidity reduction stage is
proposed in [77] and could be expressed by Eq. (3.46).

SρRT 1 1
εas = ( − ) ln(RH) (3.46)
3M Ks K

where S is the saturation degree in the pore space, ranging from 0 to 1; ρ is the
density of water, in kg/m3 , and calculated as 1 × 103 ; R is the molar gas constant,
in J/mol·K and calculated as 8.314; T is the absolute temperature, in °C; M is the
molar mass of water, in kg/mol and calculated as 0.01802; Ks is the bulk modulus of
the solid skeleton; and K is the bulk modulus of porous media.
The stiffness of concrete at early age was relatively low. The deformation that
occurred during the moisture saturated stage was mainly related to the chemical
shrinkage generated during this period. Based on the microscopic mechanism of the
autogenous shrinkage in the moisture saturated stage [78] and the model developed
for the autogenous shrinkage in the humidity reduction stage [3, 77], the two-stage
autogenous shrinkage model considering concrete RH is developed in [78] and could
be expressed by Eq. (3.47).
⎧ ( √ )
⎪
⎨ η 1 − 1 − Vcs
3
RH = 100%
εas = Sνp ρRT 1 (3.47)
⎪
⎩ ε0 +
1
( − ) ln(RH) RH < 100%
3M Ks K

@seismicisolation
@seismicisolation
3.4 Relationship Between Autogenous Shrinkage and Internal Relative … 145

where η is the influential factor of stiffness and can be determined through regression
analysis; Vcs is the chemical shrinkage induced after the initial setting time of concrete
(in volume); and νp is the pore structure influential factor.
Following the assumption that the autogenous shrinkage should be linearly propor-
tional to ln(RH) [77], the two-stage autogenous shrinkage model proposed in [78]
was simplified to predict the autogenous shrinkage of ICC with SAPs and could be
expressed as the following equation.
⎧ ( √ )
⎨ η 1 − 3 1 − Vcs RH = 100%
εas = (3.48)
⎩ ε − k ln(RH) + k RH < 100%
0 1 2

where k1 and k2 are the influential factors of concrete RH.

Following the assumption that the total volume of water and cement particles is
1, the hydration degree of cement is α, the equation for calculating the chemical
shrinkage Vcs can be expressed by Eq. (3.49) [55, 78].

Vcs = 0.2(1 − p)α (3.49)

where p is the influential factor of w/c ratio, and p = w/c+ρ w/c

w /ρc
, ρw and ρc are the
3
density of water and cement, respectively, in kg/m .
The pre-soaked SAPs particles will release almost all the water after 7 days of
cement hydration, and only a few large particles retain some water [74]. The crit-
ical time was 7, 9, and 12 d when the quantity of IC water was 11.50, 23.00, and
32.85 kg/m3 , respectively. It was approximately thought that all the water absorbed
in SAPs had been released since the critical time of all the ICC with SAPs was greater
than or equal to 7 d. Therefore, the w/c ratio utilized for calculating p should be the
sum of the basic w/c ratio and the IC w/c ratio.
According to the models developed in [79, 80], the cement hydration degree can
be calculated through Eq. (3.50).
( )q
tk
α = αmax exp − (3.50)
t
/ /
where αmax is calculated as the ratio/ of wt c to 0.36 when the wt c is lower than
0.36 and calculated as 1 when the wt c is higher than 0.36; tk and q are two empirical
constants which can be determined from experiments [79].
According to the experimental results, three parameters tk , q, and η were deter-
mined by regression analysis. The fitting results of the three parameters are shown
in Table 3.7.
The three parameters were related to the quantity of IC water entrained in SAPs
and could be expressed as Eqs. (3.51)–(3.53).

tk = 3.0806e−0.01263x R2 = 0.902 (3.51)

@seismicisolation
@seismicisolation
146 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Table 3.7 Parameters of concrete with SAPs determined by regression analysis

Parameters SAP00 SAP35 SAP70 SAP100
tk 3.000 2.840 2.224 2.018
q 0.669 0.736 0.749 0.823
η 0.0129 0.0112 0.0088 0.0075

q = 0.67336e−0.00579x R2 = 0.904 (3.52)

η = 0.01308e−0.01659x R2 = 0.982 (3.53)

where x is the quantity of IC water entrained in SAP, in kg/m3 .

As the hydration reaction of concrete specimens proceeded, the humidity reduc-
tion stage occurred. According to the Kelvin-Laplace equation [77], the development
of autogenous shrinkage due to the reduction in IRH should be linearly related to
the logarithm of the IRH. Thus, the correlation between the IRH (RH < 100%) and
the autogenous shrinkage of concrete specimens with four IC water contents could
be determined by Eqs. (3.54)–(3.57).
For reference concrete specimen SAP00 with no IC water:

εas = ε0 − 1300 ln(RH) + 6008R2 = 0.985 (3.54)

For concrete specimen SAP35 with IC water of 11.50 kg/m3 :

εas = ε0 − 1298 ln(RH) + 5975R2 = 0.987 (3.55)

For concrete specimen SAP70 with IC water of 23.00 kg/m3 :

εas = ε0 − 1292 ln(RH) + 5952R2 = 0.975 (3.56)

For concrete specimen SAP100 with IC water of 32.85 kg/m3 :

εas = ε0 − 1265 ln(RH) + 5827R2 = 0.977 (3.57)

The R2 values obtained from Eqs. (3.54)–(3.57) were 0.985, 0.987, 0.975, and
0.977 for concrete specimens SAP00, SAP35, SAP70, and SAP100, respectively,
which indicated that these equations were well in line with the experimental results,
as shown in Fig. 3.34. The slope of the fitting curves varies with IC water contents.
Specifically, the absolute values of the fitting slope were 1300, 1298, 1292, and
1265 with the increasing quantity of IC water from 0 kg/m3 to 11.50, 23.00, and
32.85 kg/m3 , respectively. This result also illustrated that the introduction of IC water
reduced the increase of autogenous shrinkage induced by one-unit RH reduction in
concrete. The test results fitting with Eq. (3.41) showed a slightly higher value of R2 ,

@seismicisolation
@seismicisolation
3.4 Relationship Between Autogenous Shrinkage and Internal Relative … 147

Fig. 3.34 Relationship between autogenous shrinkage and ln(RH)

which meant that the data fitting with the linear function result in a better correlation.
The R2 of linear and nonlinear fitting is very close to 1. Therefore, the autogenous
shrinkage generated in stage II could be calculated with Eq. (3.41).
Based on the regression analysis, the fitting results of the parameters k1 and k2
considering the IC water contents of the concrete specimens could be expressed as
Eqs. (3.58)–(3.59).

k1 = −0.988x + 1305 (3.58)

k2 = −5.058x + 6026 (3.59)

Substituting Eq. (3.48) with Eqs. (3.49), (3.53), (3.58), and (3.59), and combining
Eqs. (3.50)–(3.52), the model for the correlation between the autogenous shrinkage
and the IRH of ICC with SAPs could be expressed by Eqs. (3.60)–(3.61).
{ ( / )
0.01308e−0.01659x 1 − 3 1 − 0.2(1 − w/c
)α RH = 100%
εas = w/c+ρw /ρc
ε0 − (−0.988x + 1305) ln(RH) − 5.058x + 6026 RH < 100%
(3.60)

@seismicisolation
@seismicisolation
148 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.35 Comparisons of 400

autogenous shrinkage
SAP00

)
obtained from the proposed
model and test data for 300

Autogenous shrinkage (
SAP35
specimens with different
SAPs dosages SAP70

200 SAP100

100 Test data of SAP00 Fitting curve of SAP00

Test data of SAP35 Fitting curve of SAP35
Test data of SAP70 Fitting curve of SAP70
Test data of SAP100 Fitting curve of SAP100
0
0 5 10 15 20 25 30
Age (d)

⎧ ( )0.67336e−0.00579x
⎪
⎪ 3.0806e−0.01263x
⎪
⎪ α = αmax exp −
⎨ t
{ / / / (3.61)
⎪
⎪ (wt c) 0.36 wt c < 0.36
⎪
⎪ /
⎩ αmax =
1 wt c ≥ 0.36

The fitting curves, according to the modified two-stage model developed, are
shown in Fig. 3.35. The modified two-stage model can effectively predict the
autogenous shrinkage of concrete modified with SAP.
To check the applicability of the proposed model, the test results of the IRH and
autogenous shrinkage in [74] were chosen. It should be noted that for simplification,
one group of the specimens in [74] with no IC water and two groups of the specimens
done with the quantity of IC water x equal to 37.5 and 59.3 kg/m3 was finally
compared for each case. The comparisons between the autogenous shrinkage from
different experimental cases and the predicted values obtained from the proposed
autogenous shrinkage model are shown in Fig. 3.36. Generally, the predicted results
were well in line with the test results besides some discrepancies observed at the
initial period of cement hydration. The main characteristic that autogenous shrinkage
showed a high autogenous shrinkage rate at early age and then tended to be stabilized
with age was captured. The underestimation of the test results done before the critical
time observed in Fig. 3.36 may be mainly attributed to the fact that the proposed model
in stage I had a close correlation with chemical shrinkage Vcs which was calculated
by Eq. (3.62).

w/c
Vcs = 0.2(1 − )α (3.62)
w/c + ρw /ρc

@seismicisolation
@seismicisolation
3.4 Relationship Between Autogenous Shrinkage and Internal Relative … 149

Fig. 3.36 Comparisons

between the test results from
different references and
analytical predictions
obtained from the proposed
model

It should be noted that the above equation can be used only in the shrinkage
calculation for concrete without silica fume application [75]. For the concrete with
silica fume addition, the Vcs is calculated by Eq. (3.63).

1 ( s) w/c
Vcs = 0.2 + 0.7 (1 − )α (3.63)
1 + 1.4(s/ c) c w/c + (ρw /ρc )(1 + s/ c)

where s and c are the weight of silica fume and cement, respectively.
Obviously, the chemical shrinkage of concrete with silica fume was higher. There-
fore, the test results done were high than the analytical prediction in stage I. Generally,
the proposed model can well capture the development trend of autogenous shrinkage
in concrete, and can be used to predict the autogenous shrinkage simply by the
IRH. However, it also had some limitations. For example, in the humidity saturation
period, it will underestimate the autogenous shrinkage of concrete with silica fume.
Furthermore, the physical meanings of the influential factors of the RH of concrete
k1 and k2 needed to be further investigated.

3.4.6 Summary

The influence of SAPs dosages on the IRH, critical time, early-age expansion, auto-
genous shrinkage at 28 d, autogenous shrinkage rate, and the correlation between
autogenous shrinkage and IRH of ICC with SAPs was investigated. Through the
experimental findings, the conclusions were drawn as follows.
(1) The IRH at 28 d and the critical time of four concrete mixtures increased as the
dosages of SAPs increased. The IRH at 28 d or the critical time of the IRH was
88.7%, 92.4%, 95.2%, and 97.0% RH or 5, 7, 9, and 12 d, which increased by

@seismicisolation
@seismicisolation
150 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

4.2%, 7.3%, and 9.4% or 40.0%, 80.0%, and 140.0% when the IC water content
increased from 0 kg/m3 to 11.50, 23.00, and 32.85 kg/m3 , respectively.
(2) The development of autogenous shrinkage experienced an obvious expansion
stage at early age. The expansion peak increased as the dosages of SAPs
increased. The absolute value of expansion peak was 9, 62, 97, and 141 με,
which increased at the rates of 5.89, 9.78, and 14.67 times with the increasing
IC water contents from 0 kg/m3 to 11.50, 23.00, and 32.85 kg/m3 , respectively.
Furthermore, the time when the expansion peak occurred also increased as the
dosages of SAPs increased.
(3) The ultimate autogenous shrinkage at 28 d decreased as the dosages of SAPs
increased. The ultimate autogenous shrinkage at 28 d was 362, 305, 257, and 223
με, which decreased by 15.7%, 29.0%, and 38.4% with the increasing quantity
of IC water from 0 kg/m3 to 11.50, 23.00, and 32.85 kg/m3 , respectively. The
autogenous shrinkage at the same age decreased significantly as the dosages of
SAPs increased.
(4) The autogenous shrinkage rate decreased as the dosages of SAPs increased. The
autogenous shrinkage rate at 3 d was 37, 34, 29, and 25 με/d, which decreased by
8.1%, 21.6%, and 32.4% with the increasing quantity of IC water from 0 kg/m3
to 11.50, 23.00, and 32.85 kg/m3 , respectively. For four mixtures investigated,
most of the autogenous shrinkage deformations were developed 14 d after the
initial setting time.
(5) The correlation between autogenous shrinkage and IRH could be divided into
two stages. A model for predicting the autogenous shrinkage of ICC with SAPs
was proposed considering the quantity of IC water and concrete RH.

3.5 Relationship Between Autogenous Shrinkage

and Internal Relative Humidity of Internally Cured
Concrete with Different w/c Ratios

3.5.1 Mixture Proportion

Mixtures internally cured with SAPs with different w/c ratios (0.33, 0.40, and 0.50)
were prepared as Mixture IC33-04, IC40-04, and IC50-04, respectively. To reveal
the IC effect of the SAPs used on the IRH and autogenous shrinkage of concrete, an
ordinary concrete mixture with w/c ratio of 0.33 was prepared as RC33-00. Table
3.8 depicts the compositions and properties of mixtures.

@seismicisolation
@seismicisolation
3.5 Relationship Between Autogenous Shrinkage and Internal Relative … 151

Table 3.8 Concrete mixture proportions

Mixture composition RC33-00 IC33-04 IC40-04 IC50-04
Water (kg/m3 ) 171 171 205 256
Cement (kg/m3 ) 512 512 512 512
Fine aggregate (kg/m3 ) 636 636 624 606
Coarse aggregate (kg/m3 ) 1131 1131 1109 1076
SAPs (kg/m3 ) 0 1.77 1.77 1.77
IC water (kg/m3 ) 0 23.00 23.00 23.00
Superplasticizer (kg/m3 ) 3.328 3.328 1.800 0.800
Slump (mm) 100 108 110 111

3.5.2 Internal Relative Humidity of Concrete with Different

w/c Ratios

The IRH of ICC was mainly affected by self-desiccation when the concrete specimen
was under sealed condition. Figure 3.37 depicts the correlation between IRH and age
of mixtures. The IRH of concrete increased with the addition of SAPs, and the IRH of
ICC increased with the increase of w/c ratio. Figure 3.37 depicts that IRH decreased
with age. The 21-d IRH of concrete was 91.3%, 96.7%, 98.3%, and 99.1% RH, and the
28-d IRH of concrete was 88.7%, 95.2%, 97.1%, and 98.1% RH for Mixture RC33-
00, IC33-04, IC40-04, and IC50-04, respectively. Compared to Mixture RC33-00,
the addition of SAPs significantly increased the IRH in the Mixture IC33-04. This
suggested that the pre-soaked SAPs released water into the concrete, which increased
the IRH in ICC.
As a multiphase system, the pores in concrete formed by aggregates and cement
particles are full of liquid water, gas mixture, and water vapor. The particles of cement

Fig. 3.37 Relationship 100

between IRH and age of ICC
98

96
IRH (%)

94

92
RC33-00
90 IC33-04
IC40-04
88 IC50-04

0 4 8 12 16 20 24 28
Age (d)

@seismicisolation
@seismicisolation
152 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

are dispersed in the water inside the concrete after pouring, and the space occupied by
liquid water is gradually filled with hydration products with high volume during the
process of hydration. Therefore, the liquid water in the pores of concrete is reduced,
and the concrete becomes dense gradually. The size, as well as the amount of pores,
is decided by the w/c ratio. Concrete mainly includes four kinds of pores: gel pores,
transitional pores, capillary pores, and macro pores. More pores with a large diameter
as well as higher porosity can be found in the concrete with a higher w/c ratio, and
more liquid water can be found in the pores of concrete with a higher w/c ratio.
Therefore, compared with concrete with a lower w/c ratio, the IRH of concrete with
a higher w/c ratio was higher at the same age. Besides, water loss rate was considered
to be the same for the mixtures because the concrete specimens were all sealed and
cured under the same environmental condition, thus, the reduction of IRH was lower
in the concrete with a higher w/c ratio.
The pores in concrete were mostly filled with liquid water at the very early age
of concrete after pouring, which made the IRH of concrete 100%. The duration of
stage I with 100% RH was defined as the critical time, which lasted for a long time
due to the slow rate of water consumption. When the dosage of water decreased to a
critical value due to cement hydration, the vapor pressure decreased, which caused
the reduction of IRH in concrete, and meant that concrete went into stage II. The
critical time was 6.3, 10.5, 14.4, and 18.2 d for Mixture RC33-00, IC33-04, IC40-04,
and IC50-04, respectively. Results reveal that the critical time of ICC increased by
37.14%, and 73.33% as increase of w/c ratio ranging from 0.33 to 0.40, and 0.50,
respectively. The initial dosage of water in pores and loss rate affect the development
of IRH. The initial size and amount of pores in the concrete are affected by w/c
ratio. More large pores filled with liquid water and more initial dosage of water can
be found in the concrete with a higher w/c ratio. Therefore, for the same water loss
rate, the critical time of concrete increased as increase of w/c ratio.

3.5.3 Early-Age Expansion of Concrete with Different w/c

Ratios

The maximum temperature was 26.01, 26.37, 25.41, and 24.93 °C for Mixture RC33-
00, IC33-04, IC40-04, and IC50-04 at 12, 12, 14, and 15 h, respectively, as depicted
in Fig. 3.38. Compared with Mixture RC33-00, the maximum temperature of IC33-
04 increased slightly with the addition of SAPs (0.35% by weight of cement). The
addition of SAPs has contributed to enhancing the degree of hydration of the cement,
which results in higher maximum temperature in concrete. Results also reveal that
the maximum temperature of ICC decreased as increase of w/c ratio. A possible
explanation for this might be that increase of w/c ratio slows down the increasing rate
of Ca2+ concentration in the cement pore solution, which may increase the induction
period of hydration and slightly decrease the maximum hydration. Another possible
explanation for this is that the degree of dilution of cement in the cement–water

@seismicisolation
@seismicisolation
3.5 Relationship Between Autogenous Shrinkage and Internal Relative … 153

solution increases as increase of w/c ratio, which suppresses the hydration rate, and
reduces hydration heat.
Figure 3.39 depicts autogenous shrinkage of four mixtures.
Mixture RC33-00 slightly expanded by 9 με after the initial setting time, and the
addition of SAPs significantly increased the expansion peak of Mixture IC33-04.
Results also reveal that the expansion peak increased as increase of w/c ratio. The
absolute value of expansion peak was 97, 106, and 109 με for Mixture IC33-04,
IC40-04, and IC50-04, which increased by 9.28%, and 12.37% as the increase of
w/c ratio ranging from 0.33 to 0.40, and 0.50, respectively. There are several possible
explanations for the results. Firstly, the expansion is probably caused by the forma-
tion of Ca(OH)2 with a large size during the hydration of cement. A crystallization
pressure on the walls of pores is created due to the development of crystal, which

Fig. 3.38 Temperature

history of mixtures in the 26 RC33-00
first 3 days IC33-04
IC40-04
)

IC50-04
Temperature (

24 Ambient

22

20
0 12 24 36 48 60 72
Age h

Fig. 3.39 Correlation 400

RC33-00
between autogenous IC33-04
shrinkage and age of 300 IC40-04
mixtures IC50-04
Autogenous shrinkage ( )

200

100

0
80
Autogenous shrinkage ( )

40
-100
0

-40
-200
-80
expansion
-120
-300 0.0 0.5 1.0 1.5 2.0
Age (d)

0 5 10 15 20 25 30
Age (d)

@seismicisolation
@seismicisolation
154 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

can cause the expansion of the microstructure by increasing porosity in the concrete.
The addition of SAPs and the increase of w/c ratio increase size of crystal, which
increases the expansion of concrete. Secondly, C–S–H rims develop surrounding the
residual unreacted cement cores in the concrete due to hydration reaction, and the
portions of anhydrous grains in concrete are replaced by C–S–H rims with higher
volume. Thickness of C–S–H rims increases as increase of w/c ratio, which results in
higher expansion peak of concrete with higher w/c ratio. Thirdly, addition of SAPs
promotes the formation of hydroxide in concrete, and makes the ICC have a larger
expansion value than reference concrete, and the existence of expansion stage in
ICC is important in decreasing autogenous shrinkage compared with the reference
concrete.

3.5.4 Effect of w/c Ratio on Correlation Between Autogenous

Shrinkage and Internal Relative Humidity of Concrete
with SAPs

The concrete entered the shrinkage stage after expansion, and the time for Mixture
RC33-00, IC33-04, IC40-04, and IC50-04 to enter the shrinkage stage was 0.4, 3.5,
4.6, and 7.2 d, respectively. The main reason was that the absolute value of expansion
peak increased and the autogenous shrinkage rate decreased with the addition of SAPs
and the increase of w/c ratio, as depicted in Fig. 3.39. The net shrinkage was taken to
evaluate the autogenous deformation of concrete. Figure 3.40 depicts the correlation
between net shrinkage strain and age of mixtures.
The autogenous shrinkage of ICC decreased with the addition of SAPs and the
increase of w/c ratio. The net autogenous shrinkage of mixtures at 3, 7, 14, 21, and
28 d is depicted in Table 3.9. For instance, the 28-d net autogenous shrinkage was

Fig. 3.40 Correlation 350

between net shrinkage strain
and age of mixtures 300
Net shrinkage strain ( )

250

200

150

100 RC33-00
IC33-04
50 IC40-04
IC50-04
0
0 5 10 15 20 25 30
Age (d)

@seismicisolation
@seismicisolation
3.5 Relationship Between Autogenous Shrinkage and Internal Relative … 155

Table 3.9 Net autogenous shrinkage of mixtures at different ages

Concrete mixtures w/ c Autogenous shrinkage (με)
3d 7d 14 d 21 d 28 d
RC33-00 0.33 147 223 282 330 362
IC33-04 0.33 126 181 218 244 257
IC40-04 0.40 103 157 181 198 205
IC50-04 0.50 82 117 138 150 155

362, 257, 205, and 155 με for Mixture RC33-00, IC33-04, IC40-04, and IC50-04,
respectively. Results reveal that the 28-d net autogenous shrinkage of Mixture RC33-
00 decreased by 29.01% with the addition of SAPs with a dosage of 0.35% by weight
of cement. However, the autogenous shrinkage was not eliminated when the dosage
of SAPs reached 0.35% by weight of cement.
The efficiency of IC is mainly affected by the availability of IC water, which
is influenced by the particle size of SAPs as well as the spacing between SAPs
particles, and the actual efficiency of many types of SAPs was not complete. A
possible explanation for this might be that the water fails to move freely in the
concrete, which makes that the required amount of IC water calculated in theory was
underestimated than the actual amount in need.
Besides, the 28-d net autogenous shrinkage of ICC decreased by 20.23%, and
39.69% as increase of w/c ratio ranging from 0.33 to 0.40, and 0.50, respectively.
There are several possible explanations for the result of AS. Firstly, more water is
supplied by concrete with higher w/c ratio, and additional water is released by SAPs
in concrete, both of which increase the IRH of concrete, and decrease the capillary
pressure in concrete. Secondly, porosity as well as mean pore radius in concrete
increases as increase of w/c ratio, and lower capillary pressure is created in the
pore water of concrete. Therefore, the autogenous shrinkage of ICC decreases as
increase of w/c ratio. Results also reveal that autogenous shrinkage of ICC increases
with age. The main reason for the results is that the IRH of concrete decreases due
to the hydration reaction, which induces a great amount of pores in the hardened
concrete and decreases the dosage of water in pores. Capillary tension increases
when the saturation state of capillary pore changes from saturated to unsaturated,
which induces the autogenous shrinkage in concrete.
The autogenous shrinkage rate of Mixture RC33-00, IC33-04, IC40-04, and IC50-
04 is depicted in Table 3.10, respectively.
The results reveal that the addition of SAPs with a dosage of 0.35% by weight
of cement decreased the autogenous shrinkage rate of Mixture IC33-04 compared to
Mixture RC33-00, and the autogenous shrinkage rate of Mixture IC33-04, IC40-04,
and IC50-04 decreased as increase of w/c ratio. Autogenous shrinkage rate of all
mixtures was less than 10 με/d after the age of 14 d, which decreased significantly
compared with that at an earlier age. The values of autogenous shrinkage rate of
four mixtures were small at the age of 28 d, and the development of autogenous
shrinkage tended to be stable. The main reason was that when the ICC was at early

@seismicisolation
@seismicisolation
156 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Table 3.10 Autogenous shrinkage rate of mixtures at different ages

Age (d) Autogenous shrinkage rate (με/d)
RC33-00 IC33-04 IC40-04 IC50-04
1 51 44 39 31
3 37 29 23 17
5 18 14 13 8
7 14 8 6 5
14 5 6 5 3
21 9 3 2 2
28 2 1 1 1

age, the mechanical properties of concrete were still at the development stage, elastic
modulus was low, and creep coefficient was high. However, the development rate of
internal hydration was fast, which caused great self-desiccation and made that the
autogenous shrinkage develops faster than that of later stage.
Considering that the development of IRH in concrete experiences two stages, a
two-stage prediction model considering the effect of w/c ratio is necessary for better
predicting autogenous shrinkage of concrete based on the results of IRH. Based on
experiment results and prediction model in previous study, the autogenous shrinkage
on early-age ICC was calculated with Eqs. (3.48)–(3.50).
According to the experiment results, the regression analysis of the influential
factors for four mixtures was carried out, and the results are depicted in Table 3.11.
The influential factors depicted in Table 3.11 were all related to w/c ratio, and
results of regression analysis of influential factors and w/c ratio at stage I are given
in Eq. (3.64).
⎧ −0.073(w/c)
⎨ tk = 1.5472e
⎪
q = 1.1890e−0.128(w/c) (3.64)
⎪
⎩
η = 0.0123e−1.633(w/c)

The IRH decreased gradually due to the development of hydration in ICC with
SAPs, which meant that the concrete entered stage II. The capillary pressure is the
main cause of autogenous shrinkage in concrete at stage II. The autogenous shrinkage
and logarithm of RH were in a linear relationship according to the Kelvin-Laplace
equation, and the fitting results of the data are depicted in Fig. 3.41.

Table 3.11 Influential

Influential factor IC33-04 IC40-04 IC50-04
factors of mixtures
tk 1.511 1.501 1.492
q 1.241 1.250 1.268
η 0.0073 0.0062 0.0055

@seismicisolation
@seismicisolation
3.5 Relationship Between Autogenous Shrinkage and Internal Relative … 157

Fig. 3.41 Correlation

between autogenous
shrinkage and ln(RH) of
mixtures

The absolute value of the fitting slope of IC33-04, IC40-04, and IC50-04 decreased
from 1292 to 924, and 595 as increase of w/c ratio ranging from 0.33 to 0.40, and
0.50, respectively. Results reveal that when reduction of IRH was the same, the
increase of autogenous shrinkage decreased as increase of w/c ratio.
At stage II, results of regression analysis on influential factors and w/c ratio are
given in Eq. (3.65).

@seismicisolation
@seismicisolation
158 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete
{
k = −4045(w/c) + 2595
(3.65)
b = −18641(w/c) + 11960

The prediction model of autogenous shrinkage based on the results of IRH of

ICC considering the effect of w/c ratio was proposed, as depicted in Eqs. (3.66) and
(3.67).
⎧ ( / )
⎪ ( ( )q )
⎪
⎨ η 1 − 3 1 − 0.2(1 − w/c t
− kt
)αmax e , RH = 100%
εas = w/c + ρw /ρc
⎪
⎪
⎩
εcs − k ln(RH) + b, RH < 100%
(3.66)
⎧ { /
⎪
⎪ [w/c] 0.36,w/c < 0.36
⎪
⎪ αmax =
⎪
⎪ 1,w/c ≥ 0.36
⎪
⎪
⎪
⎪
⎪
⎪ tk = 1.5472e−0.073(w/c)
⎨
⎪ q = 1.1890e−0.128(w/c) (3.67)
⎪
⎪
⎪
⎪ η = 0.0123e−1.633(w/c)
⎪
⎪
⎪
⎪
⎪ k = −4045(w/c) + 2595
⎪
⎪
⎩
b = −18641(w/c) + 11960

The comparison between the results of the prediction model and the experiment
is depicted in Fig. 3.42.
Results reveal that a two-stage prediction model was able to effectively predict
autogenous shrinkage of ICC with different w/c ratios ranging from 0.33 to 0.50
based on the results of IRH. Although the effect of w/c ratio on autogenous shrinkage
and IRH of the concrete IC by SAPs, and the IC effect of SAPs on the concrete with

Fig. 3.42 Comparison 350

Test data of IC33-04
between results of prediction Test data of IC40-04
model and experiment 300
Test data of IC50-04
Fitting curves
250
Autogenous shrinkage

200

150

100

50

0
0 5 10 15 20 25 30
Age (d)

@seismicisolation
@seismicisolation
3.6 Relationship Between Autogenous Shrinkage and Internal Relative … 159

w/c ratio of 0.33 were investigated, further studies on the IC effect of SAPs on the
concrete with high w/c ratios are necessary for better understanding the correlation
between autogenous shrinkage and IRH of ICC with different w/c ratios.

3.5.5 Summary

Effect of w/c ratio on the IRH and autogenous shrinkage of ICC was investigated.
Tests and analysis on IRH, temperature, expansion, and shrinkage of concrete for
Mixture IC33-04, IC40-04, and IC50-04 were conducted, respectively. The properties
of an ordinary concrete mixture with w/c ratio of 0.33 (Mixture RC33-00) were tested
simultaneously. Based on the experimental findings, the following conclusions were
obtained.
(1) The IRH of ICC increased as increase of w/c ratio, and the IRH of concrete
increased with the addition of SAPs. The development of IRH experienced
two stages: a water–vapor saturated stage with 100% RH (stage I), and a stage
in which IRH decreased gradually (stage II). A two-stage prediction model of
autogenous shrinkage based on the results of IRH for ICC considering the effect
of w/c ratio was proposed.
(2) The critical time increased as increase of w/c ratio, and the critical time of
concrete increased by adding SAPs.
(3) The expansion peak of ICC increased as increase of w/c ratio. The addition of
SAPs significantly increased the expansion peak of concrete.
(4) The autogenous shrinkage and autogenous shrinkage rate of ICC decreased as
increase of w/c ratio. The autogenous shrinkage and autogenous shrinkage rate
of concrete decreased by adding SAPs.

3.6 Relationship Between Autogenous Shrinkage

and Internal Relative Humidity of Concrete Internally
Cured with Pre-Wetted LWAs

3.6.1 Mixture Proportion

Conforming to Chinese Standard GB 175 [81], P·O 52.5R Portland cement was
utilized. The fine or coarse aggregate was natural river sand or the crushed limestone
with a size not exceeding 20 mm. The superplasticizer was polycarboxylate-based.
LWAs were pre-wetted to avoid the absorption of effective water from the mixture
in case of changing concrete properties. The dry-bulk density of the utilized LWAs
was equal to 1050 kg/m3 . The crushing strength of LWAs in a dry state was equal
to 1.7 MPa as per Chinese Standard GB/T 17431.2 [82]. The 72-h absorption value
was equal to 12% by the mass of LWAs. As exhibited in Table 3.12, four proportions

@seismicisolation
@seismicisolation
160 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Table 3.12 Concrete mixture proportions

Mixture composition Concrete mixtures
LWA00 LWA10 LWA30 LWA50
Water (kg/m3 ) 171 171 171 171
Cement (kg/m3 ) 512 512 512 512
Fine aggregate (kg/m3 ) 636 636 636 636
Coarse aggregate (kg/m3 ) 1131 1018 792 566
LWA (kg/m3 ) 0 44.6 133.9 223.2
Internal curing water (kg/m3 ) 0 5.35 16.06 26.78
Superplasticizer (kg/m3 ) 3.33 3.33 3.33 3.33

of pre-wetted LWAs were designed, namely Mixture LWA00, LWA10, LWA30, and
LWA50 with the volume replacement ratio of 0%, 10%, 30%, and 50%, respectively.

3.6.2 Temperature History of Concrete with Pre-wetted LWAs

Figure 3.43 exhibits the internal temperature versus age of early-age concrete.
The maximum temperature rise of Mixture LWA00, LWA10, LWA30, and LWA50
was 4.8, 4.5, 4.2, and 3.8 °C, respectively. Compared with Mixture LWA00, the
maximum temperature rise decreased by 6.3%, 12.5%, and 20.8% for Mixture
LWA10, LWA30, and LWA50, respectively. In addition, the time when the peak
value of internal temperature occurred was 13, 13.3, 13.6, and 14 h after casting
for Mixture LWA00, LWA10, LWA30, and LWA50, respectively. The addition of
pre-wetted LWAs delayed the time when the peak temperature was reached.

Fig. 3.43 Internal 27

temperature versus age of LWA00
26 LWA10
concrete mixtures LWA30
LWA50
25
Ambient
Temperature(°C)

24

23

22

21

20
0 12 24 36 48 60 72
Age (h)

@seismicisolation
@seismicisolation
3.6 Relationship Between Autogenous Shrinkage and Internal Relative … 161

3.6.3 Early-Age Expansion of Concrete with Pre-wetted

LWAs

Figure 3.44 exhibits the evolution of autogenous shrinkage with age of concrete
mixtures. The maximum expansion of mixtures or the time to reach the maximum
expansion was 0, -20, -50, and -96 με or 0, 14.4, 19.2, and 24.0 h for Mixture LWA00,
LWA10, LWA30, and LWA50, respectively. The absolute value of the maximum
expansion was larger and the time to reach the maximum expansion of mixtures was
longer when a larger quantity of IC water was provided (i.e., increasing proportions
of pre-wetted LWAs).
Figure 3.45 exhibits the fitting curve of the maximum expansion versus the
quantity of IC water.

Fig. 3.44 Autogenous 400

shrinkage versus age of
concrete mixtures 300
Autogenous shrinkage ( )

200

100

0
LWA00
LWA10
-100 LWA30
LWA50
0 5 10 15 20 25 30
Age (d)

Fig. 3.45 Fitting curve of

the relationship between the
maximum expansion and the
quantity of IC water

@seismicisolation
@seismicisolation
162 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

The maximum expansion could be calculated with a linear function, as given in

Eq. (3.68):

εexp max (x) = εexp

0
max + a · x (3.68)

where εexp max (x) and εexp

0
max denote the maximum expansion of concrete mixtures
and Mixture LWA00, respectively, in με; a denotes the fitting parameter, taken as
3.47 with the coefficient of determination (R2 ) greater than 0.988, and x denotes the
quantity of IC water supplied by pre-wetted LWAs, in kg/m3 .

3.6.4 Shrinkage of Concrete with Pre-wetted LWAs

The 14-d or 28-d autogenous shrinkage of concrete with pre-wetted LWAs was 273,
223, 157, and 65 με or 353, 301, 225, and 132 με for Mixture LWA00, LWA10,
LWA30, and LWA50, respectively. When compared with Mixture LWA00, the auto-
genous shrinkage at 14 or 28 d decreased by 18.3%, 42.5%, and 76.2% or 14.7%,
36.3%, and 62.6% for Mixture LWA10, LWA30, and LWA50, respectively. The
autogenous shrinkage decreased dramatically when more proportions of pre-wetted
LWAs were added based on the same age. The occurrence of autogenous shrinkage
was postponed owing to the incorporation of pre-wetted LWAs and concrete mixtures
had a relatively high strength at that time, which was beneficial to controlling the
cracking of early-age concrete. Besides, the porous LWAs have high water absorption,
and the absorbed water could offer a decent IC environment during concrete hard-
ening, as reported in previous work [83, 84]. The 14-d autogenous shrinkage reached
77.3%, 74.1%, 69.8%, and 49.2% of the 28-d autogenous shrinkage, suggesting that
the autogenous shrinkage of concrete rose dramatically before 14 days and then
became relatively steadily.
The evolution of the autogenous shrinkage rate of concrete mixtures with age is
exhibited in Fig. 3.46. The autogenous shrinkage rate at 3 d was 35, 31, 16, and 18
με/d for Mixture LWA00, LWA10, LWA30, and LWA50, respectively. Compared
with Mixture LWA00, the autogenous shrinkage rate decreased by 11.4%, 54.3%,
and 48.6% for Mixture LWA10, LWA30, and LWA50, respectively. The autogenous
shrinkage rate at 28 d was about 2 με/d for all concrete mixtures.

3.6.5 Internal Relative Humidity of Concrete with Pre-Wetted

LWAs

Figure 3.47 exhibits IRH versus age of concrete.

The critical time of IRH was 5, 6, 8, and 10 d for Mixture LWA00, LWA10,
LWA30, and LWA50, respectively. The critical time of IRH increased when more

@seismicisolation
@seismicisolation
3.6 Relationship Between Autogenous Shrinkage and Internal Relative … 163

Fig. 3.46 Autogenous 60

shrinkage rate versus age of LWA00
concrete mixtures 50 LWA10

Autogenous shrinkage rate ( /d)

LWA30
LWA50
40

30

20

10

0
0 5 10 15 20 25 30
Age (d)

Fig. 3.47 IRH versus age of

100
concrete mixtures
98

96
IRH (%)

94

92
LWA00
90 LWA10
LWA30
LWA50
88
0 5 10 15 20 25 30
Age (d)

proportions of pre-wetted LWAs were added. The critical time of IRH versus the
quantity of IC water is exhibited in Fig. 3.48.
The linear correlation between the critical time of IRH and the quantity of IC
water supplied by pre-wetted LWAs was described as Eq. (3.69):

t0 = 5.0 + 0.19 · x (3.69)

where t0 denotes the critical time of IRH, in d.

The IRH at the age of 14 or 28 d was 94.9%, 96.0%, 97.6% and 99.4% RH
or 88.7%, 90.1%, 93.2%, and 95.9% RH for Mixture LWA00, LWA10, LWA30,
and LWA50, respectively. The IRH was larger and the humidity-saturated stage was

@seismicisolation
@seismicisolation
164 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.48 Fitting curve of

the relationship between the
critical time of IRH and the
quantity of IC water

longer with more proportions of pre-wetted LWAs at the same age. Large quanti-
ties of pre-absorbed water in LWAs migrated to cement pastes and was provided
for unhydrated cement particles under capillary tension, thereby increasing water
saturation in the capillary pores, and maintaining the gradually decreasing IRH of
concrete.

3.6.6 Relationship Between Autogenous Shrinkage

and Internal Relative Humidity of Concrete
with Pre-wetted LWAs

Figure 3.49 exhibits the autogenous shrinkage versus IRH for concrete with various
quantities of IC water.
For Mixture LWA00 without IC water:

εias = −13.8RH + 1583.1 R2 = 0.986 (3.70)

For Mixture LWA10 with IC water of 5.35 kg/m3 :

εias = −17.5RH + 1892.3 R2 = 0.976 (3.71)

For Mixture LWA30 with IC water of 16.06 kg/m3 :

εias = −19.9RH + 2087.7 R2 = 0.963 (3.72)

For Mixture LWA50 with IC water of 26.78 kg/m3 :

εias = −21.6RH + 2212.5 R2 = 0.967 (3.73)

@seismicisolation
@seismicisolation
3.6 Relationship Between Autogenous Shrinkage and Internal Relative … 165

Fig. 3.49 Autogenous shrinkage versus IRH of concrete: a LWA00; b LWA10; c LWA30; and d
LWA50

The R2 values obtained from Eqs. (3.70)–(3.73) were in the range of 0.963–0.986.
The absolute values of the slopes of the fitting curves increased with a larger quantity
of IC water.
The expansion of concrete could be considered linearly proportional to the age
of concrete. Based on the autogenous shrinkage mechanism, the model proposed in
previous work [77, 78] could be modified as Eq. (3.74) to predict the autogenous
shrinkage considering the quantity of IC water.
⎧ ( )
⎪ t 0
⎪
⎪ ε + k · x t ≤ t0 , RH = 100%
⎪
⎪
expmax
⎨ t0 √
εias = η(1 − 3 1 − Vics ) − εexpmax t > t0 , RH = 100% (3.74)
⎪
⎪
⎪
⎪ SV ρRT 1 1
⎪
⎩ εics −
p
( − ) ln(RH) RH < 100%
3M K Ks

@seismicisolation
@seismicisolation
166 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Table 3.13 Parameters of concrete with LWA determined by regression analysis

Parameters LWA00 LWA10 LWA30 LWA50
η 0.0141 0.0087 0.0061 0.0054
tk 3.963 3.880 3.674 3.341
q 0.829 0.922 1.186 1.260

where η denotes the correction factor of chemical shrinkage; εics denotes the auto-
genous shrinkage at the critical time of IRH, in με; k denotes the parameter that can
be obtained from the regression analysis; and Vics denotes the chemical shrinkage.
The parameters η, tk , and q in Eq. (3.50) were relevant to the quantity of IC water.
Table 3.13 exhibits the results of the regression analysis of η, tk , and q.
The calculation of the above three parameters was described as Eqs. (3.75)–(3.77):

tk = 3.9959e−0.0062x (3.75)

q = 0.8607e0.0154·x (3.76)

η = 0.0130e−0.042·x (3.77)

Autogenous shrinkage versus ln(RH) for concrete with various quantities of IC

water is exhibited in Fig. 3.50.
For Mixture LWA00 without IC water:

εias = εics − 1300.1 ln(RH) + 6003.5 R2 = 0.985 (3.78)

For Mixture LWA10 with IC water of 5.35 kg/m3 :

εias = εics − 1659.8 ln(RH) + 7666.9 R2 = 0.972 (3.79)

For Mixture LWA30 with IC water of 16.06 kg/m3 :

εias = εics − 1914.7 ln(RH) + 8839.1 R2 = 0.960 (3.80)

For Mixture LWA50 with IC water of 26.78 kg/m3 :

εias = εics − 2118.1 ln(RH) + 9770.3 R2 = 0.966 (3.81)

The R2 values obtained from Eqs. (3.78)–(3.81) were in the range of 0.960 to
0.985. The fitting parameters of the slope m and the intercept n were obtained from
the regression analysis. The linear correlations between the fitting parameters and
the quantity of IC water could be described as Eqs. (3.82)–(3.83), respectively:

@seismicisolation
@seismicisolation
3.6 Relationship Between Autogenous Shrinkage and Internal Relative … 167

Fig. 3.50 Autogenous shrinkage versus ln(RH): a LWA00; b LWA10; c LWA30; and d LWA50

m = −28.6x − 1403.4 (3.82)

n = 131.7x + 6483.2 (3.83)

Substituting the calculation of the parameters η, tk , q, m, and n into Eq. (3.74),

the prediction model for estimating the autogenous shrinkage of concrete containing
pre-wetted LWAs based on IRH could be obtained considering the quantity of IC
water, as given in Eq. (3.84):
⎧ t ( )
⎪
⎪ ε0 + k · x t ≤ t0 , RH = 100%
⎪
⎪ t0 expmax
⎪
⎪
⎪
⎪ /
⎪
⎨ ( ) 0.0154·x
−0.042·x 3 (w/c)t 3.9959e−0.0062x 0.8607e
εias = 0.0130e (1 − 1 − 0.2(1 − )αmax · exp − ) − εexpmax
⎪
⎪ (w/c)t + ρw /ρc t
⎪
⎪
⎪
⎪
⎪ t > t0 , RH = 100%
⎪
⎪
⎩
εics − (28.6x + 1403.4) ln(RH) + 131.7x + 6483.2 RH < 100%
(3.84)

@seismicisolation
@seismicisolation
168 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

Fig. 3.51 Fitting curve of 400

autogenous shrinkage strain
of concrete mixtures 300

Autogenous shrinkage ( )
200

100

0 LWA00
LWA10
LWA30
-100 LWA50
Fitting curve

0 5 10 15 20 25 30
Age (d)

The fitting curve of the autogenous shrinkage is exhibited in Fig. 3.51. The predic-
tion results coincided with the test results. The modified model was valuable in
estimating the autogenous shrinkage of concrete based on IRH, thus providing the
theoretical foundation for assessing the early-age performance of pre-wetted LWAs
concrete and calculating the shrinkage-induced stress in structures.

3.6.7 Summary

An effective approach for reducing the autogenous shrinkage of concrete through IC.
IRH and autogenous shrinkage measurement were adopted to investigate whether and
how pre-wetted LWAs proportioning work on the early-age performance of concrete
mixtures. The relationship between autogenous shrinkage and IRH of early-age
concrete with pre-wetted LWAs was also analyzed. The findings were summarized
as follows.
(1) The development of autogenous shrinkage experienced a stage of expansion
stage at the very early age for concrete incorporated with pre-wetted LWAs. The
maximum expansion increased with a larger amount of IC water. The maximum
expansion was 0, −20, −50, and −96 με for Mixture LWA00, LWA10, LWA30,
and LWA50, respectively.
(2) The autogenous shrinkage decreased significantly with a larger pre-wetted
LWAs replacement ratio. The 28-d autogenous shrinkage of internally cured
concrete was 353, 301, 225, and 132 με, which decreased by 14.7%, 36.3%,
and 62.6% for Mixture LWA00, LWA10, LWA30, and LWA50, respectively.
(3) The IRH and the critical time of IRH increased with a larger pre-wetted LWAs
replacement ratio. The 28-d IRH or the critical time of IRH was 88.7%, 90.1%,

@seismicisolation
@seismicisolation
References 169

93.2%, and 95.9% RH or 5, 6, 8, and 10 d for Mixture LWA00, LWA10, LWA30,

and LWA50, respectively.
(4) The established model of autogenous shrinkage was divided into the humidity-
saturated stage and IRH drop stage, which was based on IRH of concrete
and considered the quantity of IC water. The predicted autogenous shrinkage
coincided with the test results.
(5) The proportion of pre-wetted LWAs was crucial when evaluating the early-age
performance of concrete because there existed a balance between minimizing the
autogenous shrinkage and not sacrificing the mechanical properties of concrete
with pre-wetted LWAs.

References

1. Shen DJ, Liu KQ, Ji Y et al (2018) Early-age residual stress and stress relaxation of high-
performance concrete containing fly ash. Mag Concr Res 70(14):726–738
2. Shen DJ, Liu C, Li CC et al (2019) Influence of Barchip fiber length on early-age behavior and
cracking resistance of concrete internally cured with super absorbent polymers. Constr Build
Mater 214:219–231
3. Wu LM, Farzadnia N, Shi CJ et al (2017) Autogenous shrinkage of high performance concrete:
a review. Constr Build Mater 149:62–75
4. Shen DJ, Liu C, Feng ZZ et al (2019) Influence of ground granulated blast furnace slag on the
early-age anti-cracking property of internally cured concrete. Constr Build Mater 223:233–243
5. Han YD, Zhang J, Luosun YM et al (2014) Effect of internal curing on internal relative humidity
and shrinkage of high strength concrete slabs. Constr Build Mater 61:41–49
6. Liu YL, Schindler AK (2020) Finite-element modeling of early-age concrete stress develop-
ment. J Mater Civ Eng 32(1):04019338
7. Snoeck D, Pel L, de Belie N (2019) Comparison of different techniques to study the nanos-
tructure and the microstructure of cementitious materials with and without superabsorbent
polymers. Constr Build Mater 223:244–253
8. Dawood ET, Ramli M (2010) Development of high strength flowable mortar with hybrid fiber.
Constr Build Mater 24:1043–1050
9. Sahmaran M, Yurtseven A, Yaman IO (2005) Workability of hybrid fiber reinforced self-
compacting concrete. Build Environ 40(12):1672–1677
10. Shen DJ, Wen CY, Kang JC et al (2020) Early-age stress relaxation and cracking potential of
High-strength concrete reinforced with Barchip fiber. Constr Build Mater 258:119538
11. Wu YH (2017) Experimental study on effect of polypropylene fiber on autogenous shrinkage
of concrete. Hohai University, Nanjing (in Chinese)
12. Shen DJ, Jiang JL, Shen JX et al (2015) Influence of prewetted lightweight aggregates on the
behavior and cracking potential of internally cured concrete at an early age. Constr Build Mater
99:260–271
13. Shen DJ, Jiang JL, Jiao Y et al (2017) Early-age tensile creep and cracking potential of concrete
internally cured with pre-wetted lightweight aggregate. Constr Build Mater 135:420–429
14. Shen DJ, Feng ZZ, Zhu PF et al (2020) Effect of pre-wetted lightweight aggregates on residual
stress development and stress relaxation in restrained concrete ring specimens. Constr Build
Mater 258:119151
15. Deng SC (2016) Research on relationship between relative humidity and autogenous shrinkage
of internally cured concrete with lightweight aggregates. Hohai University, Nanjing (in
Chinese)

@seismicisolation
@seismicisolation
170 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

16. Shen DJ, Shi HF, Tang XJ et al (2016) Effect of internal curing with super absorbent polymers
on residual stress development and stress relaxation in restrained concrete ring specimens.
Constr Build Mater 120:309–320
17. Shen DJ, Liu C, Jiang JL et al (2020) Influence of super absorbent polymers on early-age
behavior and tensile creep of internal curing high strength concrete. Constr Build Mater
258:120068
18. Shen DJ, Jiang JL, Zhang MY et al (2018) Tensile creep and cracking potential of high perfor-
mance concrete internally cured with super absorbent polymers at early age. Constr Build
Mater 165:451–461
19. Cheng DB (2015) Research on early-age autogenous shrinkage development of internally cured
concrete. Hohai University, Nanjing (in Chinese)
20. Wang XD (2016) Research on relationship of relative humidity and autogenous shrinkage of
concrete internally cured with SAP. Hohai University, Nanjing (in Chinese)
21. Holt E (2005) Contribution of mixture design to chemical and autogenous shrinkage of concrete
at early ages. Cem Concr Res 35(3):464–472
22. Yoo DY, Banthia N, Yoon YS (2015) Effectiveness of shrinkage-reducing admixture in
reducing autogenous shrinkage stress of ultra-high-performance fiber-reinforced concrete.
Cement Concr Compos 64:27–36
23. Yoo DY, Kang ST, Lee JH et al (2013) Effect of shrinkage reducing admixture on tensile and
flexural behaviors of UHPFRC considering fiber distribution characteristics. Cem Concr Res
54:180–190
24. Liu JP, Tian Q, Tang MS (2006) Influence of expansion agent and shrinkage reducing agent on
shrinkage cracking of high performance concrete. J Southeast Univ 36:195–199 (in Chinese)
25. He ZM, Shen AQ, Guo YC et al (2019) Cement-based materials modified with superabsorbent
polymers: a review. Constr Build Mater 225:569–590
26. Shen DJ, Wang XD, Cheng DB et al (2016) Effect of internal curing with super absorbent
polymers on autogenous shrinkage of concrete at early age. Constr Build Mater 106:512–522
27. Shen DJ, Wen CY, Zhu PF et al (2020) Influence of Barchip fiber on early-age autogenous
shrinkage of high-strength concrete internally cured with super absorbent polymers. Constr
Build Mater 264:119983
28. Kang JC, Shen DJ, Li CC et al (2022) Effect of water-to-cement ratio on internal relative
humidity and autogenous shrinkage of early-age concrete internally cured by superabsorbent
polymers. Struct Concr 1–15
29. Shen DJ, Li CC, Li M et al (2022) Experimental investigation on correlation between autoge-
nous shrinkage and internal relative humidity of superabsorbent polymer–modified concrete.
J Mater Civ Eng 34(2):04021456
30. Ministry of Housing and Urban Rural Development of the People’s Republic of China. Standard
for test methods of long-term performance and durability of ordinary concrete: GB/T 50082-
2009[S]. China Architecture and Building Press, Beijing, 2009 (in Chinese)
31. Jiang CH, Yang Y, Wang Y et al (2014) Autogenous shrinkage of high performance concrete
containing mineral admixtures under different curing temperatures. Constr Build Mater
61:260–269
32. Chu I, Kwon SH, Amin MN et al (2012) Estimation of temperature effects on autogenous
shrinkage of concrete by a new prediction model. Constr Build Mater 35:171–182
33. Amin MN, Kim JS, Dat TT et al (2010) Improving test methods to measure early age autogenous
shrinkage in concrete based on air cooling. IES J Part A Civ Struct Eng 3(4):244–256
34. ASTM International (2019) Standard test method for autogenous strain of cement paste and
mortar: ASTM C1581/C1581M-18a[S]. ASTM International, West Conshohocken
35. Darquennes A, Staquet S, Espion B (2011) Determination of time-zero and its effect on
autogenous deformation evolution. Eur J Environ Civ Eng 15(7):1017–1029
36. Miao CW, Tian Q, Sun W et al (2007) Water consumption of the early-age paste and the
determination of “time-zero” of self-desiccation shrinkage. Cem Concr Res 37(11):1496–1501
37. Tsioulou O, Lampropoulos A, Paschalis S (2017) Combined non-destructive testing (NDT)
method for the evaluation of the mechanical characteristics of ultra high performance fibre
reinforced concrete (UHPFRC). Constr Build Mater 131:66–77

@seismicisolation
@seismicisolation
References 171

38. Yap SP, Alengaram UJ, Jumaat MZ (2013) Enhancement of mechanical properties in
polypropylene-and nylon-fibre reinforced oil palm shell concrete. Mater Des 49:1034–1041
39. Lee KM, Lee HK, Lee SH et al (2006) Autogenous shrinkage of concrete containing granulated
blast-furnace slag. Cem Concr Res 36(7):1279–1285
40. Petro JT, Kim J (2012) Detection of delamination in concrete using ultrasonic pulse velocity
test. Constr Build Mater 26(1):574–582
41. Schröfl C, Mechtcherine V, Gorges M (2012) Relation between the molecular structure and
the efficiency of superabsorbent polymers (SAP) as concrete admixture to mitigate autogenous
shrinkage. Cem Concr Res 42(6):865–873
42. Darquennes A, Staquet S, Delplancke-Ogletree MP, et al (2011) Effect of autogenous
deformation on the cracking risk of slag cement concretes. Cem Concr Compos 33(3):368–379
43. Zhang H, She W, Li L et al (2019) Effect of temperature rising inhibitor on autogenous shrinkage
of cement pastes. Constr Build Mater 220:329–339
44. Paul A, Lopez M (2011) Assessing lightweight aggregate efficiency for maximizing internal
curing performance. ACI Mater J 108(4):385–393
45. Sant G, Lothenbach B, Juilland P et al (2011) The origin of early age expansions induced in
cementitious materials containing shrinkage reducing admixtures. Cem Concr Res 41(3):218–
229
46. Baroghel-Bouny V, Mounanga P, Khelidj A et al (2006) Autogenous deformations of cement
pastes: Part II. W/C effects, micro-macro correlations, and threshold values. Cem Concr Res
36(1):123–136
47. Akcay B, Tasdemir MA (2008) Internal curing of mortars by lightweight aggregates and its
effects on hydration. Can J Civ Eng 35(11):1276–1284
48. Henkensiefken R, Bentz D, Nantung T et al (2009) Volume change and cracking in inter-
nally cured mixtures made with saturated lightweight aggregate under sealed and unsealed
conditions. Cement Concr Compos 31(7):427–437
49. Wei Y, Xiang YP, Zhang QQ (2014) Internal curing efficiency of prewetted LWFAs on concrete
humidity and autogenous shrinkage development. J Mater Civ Eng 26(5):947–954
50. Igarashi S, Watanabe A (2006) Experimental study on prevention of autogenous deformation by
internal curing using super-absorbent polymer particles. In: International RILEM conference
on volume changes of hardening concrete: testing and mitigation. RILEM Publications SARL,
Paris, France, pp 77–86
51. Henkensiefken R (2008) Internal curing in cementitious systems made using saturated
lightweight aggregate. Purdue University, Indiana
52. See HT, Attiogbe EK, Miltenberger MA (2004) Potential for restrained shrinkage cracking of
concrete and mortar. Cem Concr Aggreg 26(2):123–130
53. Browning J, Darwin D, Reynolds D et al (2011) Lightweight aggregate as internal curing agent
to limit concrete shrinkage. ACI Mater J 108(6):638–644
54. Wang FZ, Zhou YF, Peng B et al (2009) Autogenous shrinkage of concrete with super-absorbent
polymer. ACI Mater J 106(2):123–127
55. Jensen OM, Hansen PF (2001) Water-entrained cement-based materials: I. principles and
theoretical background. Cem Concr Res 31(4):647–654
56. Jensen OM, Hansen PF (2002) Water-entrained cement-based materials: II. experimental
observations. Cem Concr Res 32(6):973–978
57. Wyrzykowski M, Lura P (2013) Controlling the coefficient of thermal expansion of cemen-
titious materials—a new application for superabsorbent polymers. Cem Concr Compos
35(1):49–58
58. Yoo SW, Kwon SJ, Jung SH (2012) Analysis technique for autogenous shrinkage in high
performance concrete with mineral and chemical admixtures. Constr Build Mater 34:1–10
59. Tazawa E, Miyazawa S (2001) Prediction model for shrinkage of concrete including autoge-
nous shrinkage. Creep, shrinkage and durability mechanics of concrete and other quasi-brittle
materials. In: Proceedings of sixth international conference, Elsevier Science Ltd., pp 735–746
60. fib.fib Model Code for Concrete Structures 2010: fib Model Code 2010. Ernst and Sohn, Berlin,
2013

@seismicisolation
@seismicisolation
172 3 Autogenous Shrinkage of Early-Age Internally Cured Concrete

61. CEN (2004) Eurocode 2: design of concrete structures: Part 1–1: general rules and rules for
buildings: BS EN 1992-1-1. CEN, Brussels
62. Zhutovsky S, Kovler K, Bentur A (2004) Influence of cement paste matrix properties on the
autogenous curing of high-performance concrete. Cem Concr Compos 26(5):499–507
63. Wu SX, Wang X, Shen DJ et al (2020) Simulation analysis on hydration kinetics and microstruc-
ture development of tricalcium silicate considering dissolution mechanisms. Constr Build
Mater 249:118535
64. Craeye B, Geirnaert M, Schutter GD (2011) Super absorbing polymers as an internal curing
agent for mitigation of early-age cracking of high-performance concrete bridge decks. Constr
Build Mater 25(1):1–13
65. Snoeck D, Jensen OM, de Belie N (2015) The influence of superabsorbent polymers on the
autogenous shrinkage properties of cement pastes with supplementary cementitious materials.
Cem Concr Res 74:59–67
66. Shen DJ, Feng ZZ, Kang JC et al (2020) Effect of Barchip fiber on stress relaxation and
cracking potential of concrete internally cured with super absorbent polymers. Constr Build
Mater 249:118392
67. Shen DJ, Wang WT, Liu JW et al (2018) Influence of Barchip fiber on early-age cracking
potential of high performance concrete under restrained condition. Constr Build Mater
187:118–130
68. Shen DJ, Liu XZ, Zeng X et al (2020) Effect of polypropylene plastic fibers length on cracking
resistance of high performance concrete at early age. Constr Build Mater 244:117874
69. Wang KJ, Shah SP (2001) Plastic shrinkage cracking in concrete materials—influence of fly
ash and fibers. ACI Mater J 98(6):458–464
70. Zhao YH, Xu BH, Chang JM (2019) Addition of pre-wetted lightweight aggregate and
steel/polypropylene fibers in high-performance concrete to mitigate autogenous shrinkage.
Struct Concr 21(3):1134–1143
71. Hesami S, Salehi Hikouei I, Emadi SAA (2016) Mechanical behavior of self-compacting
concrete pavements incorporating recycled tire rubber crumb and reinforced with polypropy-
lene fiber. J Clean Prod 133:228–234
72. Sukontasukkul P, Pomchiengpin W, Songpiriyakij S (2010) Post-crack (or post-peak) flexural
response and toughness of fiber reinforced concrete after exposure to high temperature. Constr
Build Mater 24(10):1967–1974
73. Zhang J, Wang JH, Gao Y (2016) Moisture movement in early-age concrete under cement
hydration and environmental drying. Mag Concr Res 68(8):391–408
74. Wang FZ, Yang J, Hu SG et al (2016) Influence of superabsorbent polymers on the surrounding
cement paste. Cem Concr Res 81:112–121
75. Zhang J, Hou DW, Sun W (2010) Experimental study on the relationship between shrinkage
and interior humidity of concrete at early age. Mag Concr Res 62(3):191–199
76. Grasley ZC, Lange DA, D’Ambrosia MD (2006) Internal relative humidity and drying stress
gradients in concrete. Mater Struct 39(9):901–909
77. Lura P, Jensen OM, van Breugel K (2003) Autogenous shrinkage in high-performance cement
paste: an evaluation of basic mechanisms. Cem Concr Res 33(2):223–232
78. Zhang J, Hou DW, Han YD (2012) Micromechanical modeling on autogenous and drying
shrinkages of concrete. Constr Build Mater 29:230–240
79. Pane I, Hansen W (2002) Concrete hydration and mechanical properties under nonisothermal
conditions. ACI Mater J 99(6):534–542
80. Zhang J, Qi K, Huang Y (2009) Calculation of moisture distribution in early-age concrete. J
Eng Mech 135(8):871–880
81. General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China. Common Portland cement: GB 175–2007/XG1–2018[S]. Standard press
of China, Beijing, 2018 (in Chinese)
82. General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China. Lightweight aggregates and its test methods-part2: Test methods for
lightweight aggregates: GB/T 17431.2-2010[S]. Standard press of China, Beijing, 2010 (in
Chinese)

@seismicisolation
@seismicisolation
References 173

83. Kohno K, Okamoto T, Isikawa Y et al (1999) Effects of artificial lightweight aggregate on

autogenous shrinkage of concrete. Cem Concr Res 29(4):611–614
84. Bentur A, Igarashi SI, Kovler K (2001) Prevention of autogenous shrinkage in high-strength
concrete by internal curing using wet lightweight aggregates. Cem Concr Res 31(11):1587–
1591

@seismicisolation
@seismicisolation
Chapter 4
Tensile Creep of Early-Age Internally
Cured Concrete with SAPs

Contents

4.1 Test and Calculation Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176

4.1.1 Test Device . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
4.1.2 Testing Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
4.1.3 Calculation of Tensile Creep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
4.2 Tensile Creep of Early-Age Concrete with Different Amount of SAPs . . . . . . . . . . . . . . 182
4.2.1 Mixture Proportions and Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
4.2.2 Mechanical Properties of Concrete Internally Cured with SAPs . . . . . . . . . . . . . 182
4.2.3 Basic Tensile Creep of Concrete Internally Cured with SAPs . . . . . . . . . . . . . . . . 183
4.2.4 Basic Tensile Creep Coefficient of Concrete Internally Cured with SAPs . . . . . . 185
4.2.5 Specific Tensile Creep of Concrete Internally Cured with SAPs . . . . . . . . . . . . . 186
4.2.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
4.3 Tensile Creep of Early-Age Internally Cured Concrete with Different Initial
Stress/strength Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
4.3.1 Mixture Proportions and Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
4.3.2 Basic Tensile Creep of Concrete with Different Initial Stress/strength
Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
4.3.3 Basic Tensile Creep Coefficient of Concrete with Different Initial Stress/strength
Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191
4.3.4 Basic Tensile Creep Rate of Concrete with Different Initial Stress/strength
Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
4.3.5 Specific Basic Tensile Creep of Concrete with Different Initial Stress/strength
Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
4.3.6 Nonlinear Basic Tensile Creep of Concrete with Different Initial Stress/strength
Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194
4.3.7 Prediction Model for Early Age Tensile Creep Considering the Initial
Stress/strength Ratio . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
4.3.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
4.4 Tensile Creep of Early-Age Internally Cured Concrete with Different Loading
Ages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
4.4.1 Mixture Proportions and Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199
4.4.2 Basic Tensile Creep of Concrete with Different Loading Ages . . . . . . . . . . . . . . . 199
4.4.3 Basic Tensile Creep Coefficient of Concrete with Different Loading Ages . . . . . 202
4.4.4 Specific Basic Tensile Creep of Concrete with Different Loading Ages . . . . . . . 203
4.4.5 Specific Basic Tensile Creep Rate of Concrete with Different Loading
Ages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204

© Science Press 2023 175

D. Shen, Cracking Control on Early-Age Concrete Through Internal Curing,
https://doi.org/10.1007/978-981-19-8398-6_4
@seismicisolation
@seismicisolation
176 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

4.4.6Nonlinear Specific Basic Tensile Creep of Concrete with Different Loading

Ages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 205
4.4.7 Modeling Tensile Creep of Concrete with Different Loading Ages . . . . . . . . . . . 207
4.4.8 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211

The tensile creep is an inherent viscoelastic property of concrete, and can relieve
approximately 50% of the restrained stress [1]. The tensile creep property of concrete
is one of the essential parameters for evaluating the shrinkage stress and cracking
resistance of concrete accurately [2–4]. The tensile creep is influenced by many
factors, such as w/c ratio [5], initial strength/stress ratio [6], loading age [7], elevated
temperature, ambient relative humidity, etc. The early-age tensile creep can be
measured by lever system loading method [8], spring system loading method [9],
and TSTM method [10]. The TSTM system consists of two specimens one free
specimen for free shrinkage and one restrained specimen to apply constant tensile
stress [11], which can directly test the deformation after concrete pouring without
formwork removal, and can test the tensile creep of very early age. The tensile creep
of ordinary concrete increases as the w/c ratio increases [12]. Early-age tensile creep
of normal concrete increases as the initial stress/strength ratio increases [13]. The
tensile creep of ordinary concrete decreases as the initial loading ages increase [14].
The majority of past experimental investigations dealt with the tensile creep of ordi-
nary concrete. Studies on the effect of SAPs additions, initial strength/stress ratios
and loading ages on the tensile creep of ICC are meaningful to determine the specific
loading age or initial strength/stress ratio at which the nonlinear tensile creep occurs
and to propose models to predict creep of ICC.
In this chapter, the TSTM system is applied to investigate the tensile creep of HSC
modified with SAPs under constant tensile load [15, 16]. The methodology of TSTM
system is described in details. Whether and how the addition of SAPs affects the early-
age tensile creep is investigated [17]. Furthermore, the effect of initial stress/strength
ratio [18] and loading age [19] on the tensile creep of concrete internally cured with
SAPs are explored to reveal the nonlinearity of tensile creep. The prediction models
for the early age tensile creep considering the initial stress/strength ratio or loading
age are established which could be applied to assess the cracking resistance and
stress in ICC with SAPs.

4.1 Test and Calculation Methods

4.1.1 Test Device

The TSTM system characterized by a fully automated closed-loop control is devel-

oped as a uniaxial test apparatus for the concrete at early age [20, 21]. The appa-
ratus was mainly equipped with controller, load cell, two molds, temperature control

@seismicisolation
@seismicisolation
4.1 Test and Calculation Methods 177

system, temperature sensors, and displacement sensors, as shown in Figs. 4.1 and
4.2. The isothermal condition for loaded specimens at 20 °C could be realized by
the temperature control system, as reported in [21]. The tests of free specimens were
conducted in a controlled temperature of 20 °C. Therefore, the curing conditions for
the loaded and free specimens were maintained the same.
The gripped end of the loaded specimen was applied with a constant tensile load
through the actuator, while the other end was fixed. Both ends of the free specimen
could move freely. Two displacement sensors were mounted on both sides of the
specimens at the time of concrete pouring to record the corresponding deformation
of the free and loaded specimens. There are the detailed dimensions of the free and
loaded specimens, as shown in Fig. 4.3a, b. Dog-bone shaped specimen with 150 mm

Computer

Displacement sensors Free specimen

Controller

Amplifier
Gripped end Fixed end
Shaft

Load cell
Stepper motor
Pulling Tensile specimen

Fig. 4.1 Schematic description of TSTM utilized in the experiment [20]

Fig. 4.2 Photo of TSTM utilized in the experiment

@seismicisolation
@seismicisolation
178 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Temperature sensors Displacement sensors

150
315 450 315

1080

(a) free specimen (mm)

Displacement sensors
Concrete Temperature sensors
Insulation mold
Moving End

Fixed End
250
150

315 450 315

1080

(b) loaded specimen (mm)

Fig. 4.3 Schematic diagram of specimens

× 150 mm linear cross section in the central part and 150 mm × 280 mm cross
section at the end was designed for the loaded specimen. The length of the central
part was 1080 mm. Both enlarged ends of the loaded specimen were designed for
the uniform distribution of stress in the working length, as reported in [22, 23]. The
grip was a special shape characterized by gradual widening of the central part to
exclude premature failure of specimen and eliminate any stress concentration. The
free specimen was a prism with a dimension of 150 mm × 150 mm × 1080 mm. The
gauge length of displacement sensors for the loaded and free specimens was designed
as 450 mm for the purpose of excluding the influence of geometry on deformation,
as shown in Fig. 4.3.

4.1.2 Testing Procedure

After complete mixing, the fresh ICC with SAPs mixtures were poured into the free
and loaded molds, respectively. A thin plastic film sheet was placed in the mold to
decrease the friction between specimen and mold. The surface of free and loaded
specimens was covered by plastic film immediately after pouring to provide sealed
conditions, as reported in [8]. Before the test, the poles holding the displacement

@seismicisolation
@seismicisolation
4.1 Test and Calculation Methods 179

sensors were inserted into the concrete through the external aluminum mould and
plastic film sheet. Test began immediately after concrete pouring, and all the data
were automatically collected and recorded by the computer system until the end of
test.
The mechanical properties tests of the concrete mixture were simultaneously
conducted according to Chinese Standard GB/T 50081-2019 [24]. The tests included
the compressive strength, splitting tensile strength, and elastic modulus at 1, 3, 7, and
28 d. The concrete samples were sealed with plastic film immediately after pouring to
provide sealed conditions, demolded at 1 d, and then kept in a controlled environment
of > 95% relative humidity and 20 ± 2 °C until the test age, as reported in [8].
For better studying the early-age behavior of HSC with SAPs, the prediction model
of mechanical properties with increasing age was established. The time-dependent
compressive strength of ICC with SAPs could be obtained from Eq. (4.1).
 
f c (t) = f c,28 exp −λ1 [ln(1 + (t − t0 ))]−k1 (4.1)

where f c (t) is the time-dependent compressive strength of ICC with SAPs, in MPa; t
is the age of ICC with SAPs after pouring, in d; f c,28 is the 28-d compressive strength
of ICC with SAPs, in MPa; and t0 is the initial setting time, in d.
The time-dependent axial tensile strength of ICC with SAPs could be obtained
from Eqs. (4.2) and (4.3), as reported in [25, 26].

f t = 0.77 · f spl + 0.21 (4.2)

 
f t (t) = f t,28 exp −λ2 [ln(1 + (t − t0 ))]−k2 (4.3)

where f t and f spl are the axial and splitting tensile strength of ICC with SAPs, in
MPa; f t (t) is the time-dependent axial tensile strength of ICC with SAPs, in MPa;
and f t,28 is the 28-d axial tensile strength of ICC with SAPs, in MPa.
The values of secant tensile and compressive elastic modulus are approximately
the same at early age [27]. Therefore, the compressive elastic modulus obtained
from tests on prisms could be utilized as tensile elastic modulus at early age, and
the time-dependent elastic modulus E t (t) of ICC with SAPs could be obtained from
Eq. (4.4).
 
E t (t) = E t,28 exp −λ3 [ln(1 + (t − t0 ))]−k3 (4.4)

where E t (t) is the time-dependent elastic modulus of ICC with SAPs, in GPa; and
E t,28 is the 28-d elastic modulus of ICC with SAPs, in GPa.
The Davis-Granville rule shows that the creep and the stress have a linear rela-
tionship when the strength/stress ratio is less than 0.4, and the specific creep curve
coincides at arbitrary loading age. Meanwhile, the tensile creep behavior of early-age
concrete is studied in [28], and the stress to splitting tensile strength ratios are chosen
as 0.2, 0.3, and 0.4. The specific tensile creep of concrete is basically the same when

@seismicisolation
@seismicisolation
180 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

the stress to splitting tensile strength ratio is less than 0.4 and initial loading age is
1.0 d. For studying the tensile creep of concrete in the range of linear characteristics,
the stress to the splitting tensile strength ratio was chosen as 0.3. Meanwhile, the
stress was equal to 35% of the calculated axial tensile strength by Eq. (4.3).

4.1.3 Calculation of Tensile Creep

1. Total strain
The free and loaded specimens were sealed after pouring into molds. Moisture
exchange from the specimens to external environment at early age was not significant,
as reported in [27]. Therefore, drying creep was not obvious and could be ignored,
as reported in [29]. Accordingly, only basic tensile creep was considered, as reported
in [30]. The measured total strain) of the loaded specimens consisted of three strain
components: shrinkage (εsh ), basic tensile creep, and elastic (or instantaneous) strain
(εel ), as reported in [31]. The total strain could be expressed by Eq. (4.5).

εtotal = εsh + εcr + εel (4.5)

where εsh and εtotal are directly measured total strain from displacement sensors
mounted in the free and loaded specimens, respectively, in με; and εel is the elastic
strain of the loaded specimens during the process of loading, in με.
2. Basic tensile creep
The basic tensile creep could be obtained from Eq. (4.6).

εcr =εtotal − εsh − εel (4.6)

3. Basic tensile creep rate

The basic tensile creep rate reflected the change of development rate of tensile creep
of ICC with SAPs under different initial stress/strength ratios and could be obtained
from Eq. (4.7).
 
•   dεcr t, t 
εcr t, t  = (4.7)
dt
•  
where εcr t, t  is the basic tensile creep rate in concrete at age t induced by an initial
constant tensile stress applied at age t  , in με/h.
4. Basic tensile creep coefficient
Due to the different development of mechanical properties of ICC with SAPs utilizing
different SAPs contents, basic tensile creep could not fully reflect the tensile creep

@seismicisolation
@seismicisolation
4.1 Test and Calculation Methods 181

behavior of ICC with SAPs. Therefore, the basic tensile creep coefficient was utilized
to eliminate the influence of the tensile elastic modulus and applied tensile stress on
the tensile creep of ICC with SAPs and could be obtained from Eq. (4.8), as reported
in [31].
 
   εcr t, t 
φ t, t = (4.8)
εe (t  )
   
where φ t, t  is the basic tensile creep coefficient of ICC with SAPs; εcr t, t  is
the basic tensile creep of ICC with SAPs at t h induced by an initial constant tensile
  
   applied at t h, in με; t is the loading age of concrete, in h (t = 24 h); and
stress  
εel t is the initial elastic strain at the time of loading, in με. The value of εel t 
could be taken as the same as that of εel .

5. Specific tensile creep

The tensile creep with different tensile stresses was normalized by the specific tensile
creep, which represented the basic tensile creep of concrete at t h induced by an initial
constant tensile stress applied at t  h to the applied tensile stress ratio, as reported in
[20]. The specific tensile creep of ICC with SAPs could be obtained from Eq. (4.9),
as reported in [18].
 
   εcr t, t 
C t, t = (4.9)
σ
 
where C t, t  is the specific tensile creep of ICC with SAPs, in με/MPa; and σ is
the initial constant tensile stress, in MPa.

6. Specific basic tensile creep rate

The specific basic tensile creep rate ν J represented the variation in the velocity of
tensile creep development at different loading ages and could be calculated utilizing
Eq. (4.10).

∂C(t, t  )
v J (t, t  ) = (4.10)
∂t

7. Tensile creep compliance function

The tensile creep compliance function was defined as the strain (elastic strain plus
basic tensile creep) at the age of t caused by a unit uniaxial constant tensile stress
exerted at the loading age of t  could be calculated by Eq. (4.11).
   
 
 εcr t, t  + εe t, t 
J t, t = (4.11)
σ

@seismicisolation
@seismicisolation
182 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Table 4.1 Concrete mixture proportions

Mixture composition Concrete mixtures
IC0035 IC0635 IC0935 IC1235
Water (kg/m3 ) 144 144 144 144
Cement (kg/m3 ) 480 480 480 480
Fine aggregate (kg/m3) 690 690 690 690
Coarse aggregate (kg/m3 ) 1127 1127 1127 1127
SAPs (kg/m3 ) 0 2.74 4.11 5.49
IC water (kg/m3 ) 0 28.8 43.2 57.6
Superplasticizer (%) 0.70 0.70 0.70 0.70

4.2 Tensile Creep of Early-Age Concrete with Different

Amount of SAPs

4.2.1 Mixture Proportions and Materials

Portland cement (P·II 52.5R) was utilized. The loss on ignition and the Blaine fineness
of the cement was 3.11% and 375 m2 /kg, respectively. The crushed limestone with
grain size below 25 mm was utilized as the coarse aggregate. The fine aggregate
utilized was natural river sand, which had a maximum size and fineness modulus
of 2.36 mm and 2.35, respectively. The SAPs utilized changed from collapsed to
swollen state when contacting water. The SAPs with the grain size between 100 and
120 μm in the dry state and a dry-bulk density of 700 kg/m3 were utilized.
The basic w/c ratio of 0.30 was taken for all mixtures, and the concrete mixture
proportions are shown in Table 4.1. The mortar sieved from the mixtures were utilized
to determine the setting time by the penetration test, as reported in Chinese Standard
GB 175-2007 [32]. The initial setting time was 0.17, 0.24, 0.30, and 0.40 d for
mixtures IC0035, IC0635, IC0935, and IC1235, respectively.

4.2.2 Mechanical Properties of Concrete Internally Cured

with SAPs

The compressive strength decreased by 1.7%, 17.1%, and 31.1% at 1 d, 11.1%,

14.8%, and 37.3% at 3 d, 4.4%, 9.2%, and 31.3% at 7 d, and 6.7%, 12.8%, and
26.1% at 28 d when the SAPs content increased from 0% to 0.57%, 0.86%, and
1.14% for mixtures IC0035, IC0635, IC0935, and IC1235, respectively, as shown in
Table 4.2. The splitting tensile strength decreased by 9.1%, 10.5%, and 13.8% at 1 d,
9.4%, 18.3%, and 22.3% at 3 d, 9.4%, 21.1%, and 23.6% at 7 d, and 10.8%, 20.3%,
and 24.9% at 28 d when the SAPs content increased from 0% to 0.57%, 0.86%, and

@seismicisolation
@seismicisolation
4.2 Tensile Creep of Early-Age Concrete with Different Amount of SAPs 183

Table 4.2 Compressive strength of concrete mixtures

Compressive strength (MPa) (d) Concrete mixtures
IC0035 IC0635 IC0935 IC1235
1 41.22 40.50 34.17 28.41
3 57.93 51.48 49.37 36.32
7 61.77 59.03 56.06 42.44
28 66.71 62.26 58.16 49.28

Table 4.3 Splitting tensile strength of concrete mixtures

Splitting tensile strength (MPa) (d) Concrete mixtures
IC0035 IC0635 IC0935 IC1235
1 2.76 2.51 2.47 2.38
3 3.72 3.37 3.04 2.89
7 4.36 3.95 3.44 3.33
28 4.74 4.23 3.78 3.56

Table 4.4 Elastic modulus of concrete mixtures

Elastic modulus (GPa) (d) Concrete mixtures
IC0035 IC0635 IC0935 IC1235
1 32.24 29.24 26.70 25.05
3 47.74 41.18 38.08 32.56
7 55.45 46.93 44.56 38.20
28 58.57 51.62 48.72 42.14

1.14% for mixtures IC0035, IC0635, IC0935, and IC1235, respectively, as shown in
Table 4.3. The elastic modulus decreased by 9.3%, 17.2%, and 22.3% at 1 d, 13.7%,
20.2%, and 31.8% at 3 d, 15.4%, 19.6%, and 31.1% at 7 d, and 11.9%, 16.8%, and
28.1% at 28 d when the SAPs content increased from 0% to 0.57%, 0.86%, and
1.14% for mixtures IC0035, IC0635, IC0935, and IC1235, respectively, as shown in
Table 4.4.

4.2.3 Basic Tensile Creep of Concrete Internally Cured

with SAPs

The calculated results of basic tensile creep of ICC with different SAPs contents, as
shown in Fig. 4.4.
The basic tensile creep of ICC with SAPs increased rapidly in the first 3 days
after loading, and then increased slowly. The introduction of IC water contributed to

@seismicisolation
@seismicisolation
184 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Fig. 4.4 Relationship 90

between basic tensile creep
and age for different 75

Basic tensile creep (με)

concrete mixtures
60

45

30 IC0035
IC0635
15 IC0935
IC1235
0

0 24 48 72 96 120 144 168 192

Age (h)

more moving gel water and lower elastic modulus of ICC with SAPs, which would
be conducive to the development of basic tensile creep at early age. The cement
hydration degree increased as time increased, the internal structure of concrete
became dense, and the increasing rate of basic tensile creep decreased gradually.
Meanwhile, the basic tensile creep of ICC with SAPs increased gradually as SAPs
content increased, which was consistent with the result of elastic modulus of ICC
with SAPs. The basic tensile creep of ICC with SAPs after loading for 7 days was
41, 46, 62, and 81 με, which increased by 12.2%, 51.2%, and 97.6% when the
SAPs content increased from 0% to 0.57%, 0.86%, and 1.14% for mixtures IC0035,
IC0635, IC0935, and IC1235, respectively. Experimental results showed that the
basic tensile creep increased nonlinearly as SAPs content increased. The basic tensile
creep increased only a little when the SAPs content increased from 0% to 0.57%,
while the basic tensile creep increased significantly when the SAPs content increased
from 0.57% to 0.86% and 1.14%. During the tensile creep tests, the stress equiva-
lent to 35% of the calculated axial tensile strength at 1 d after pouring was applied
by the step motor of TSTM to the loaded specimen. However, the actual applied
strength/stress ratio decreased as the strength of ICC with SAPs increased during the
creep test. The actual stress to the calculated axial tensile strength ratio was 0.267,
0.268, 0.290, and 0.293, or 0.229, 0.230, 0.258, and 0.258, which increased by 0.4%,
8.6%, and 9.7%, or 0.4%, 12.7%, and 12.7% when the SAPs content increased from
0% to 0.57%, 0.86%, and 1.14% after loading for 2 or 6 days for mixtures IC0035,
IC0635, IC0935, and IC1235, respectively. Part of increased tensile creep of ICC
with SAPs is related to the increase of actual strength/stress ratio as SAPs content
increased, but the influence of actual strength/stress ratio was relatively small. The
results of mechanical properties showed that both the elastic modulus and compres-
sive strength decreased as SAPs content increased, which could increase the basic
tensile creep of ICC with SAPs. The results of the influence of SAPs on the basic
tensile creep coefficient and specific tensile creep were also provided for further
discuss the mechanism of SAPs on the tensile creep of concrete.

@seismicisolation
@seismicisolation
4.2 Tensile Creep of Early-Age Concrete with Different Amount of SAPs 185

4.2.4 Basic Tensile Creep Coefficient of Concrete Internally

Cured with SAPs

There are the calculated results of basic tensile creep coefficient of ICC with SAPs
utilizing different SAPs contents, as shown in Fig. 4.5.
The basic tensile creep coefficient of ICC increased as SAPs content increased,
which was consistent with the results of basic tensile creep. The development rate
of tensile creep coefficient of ICC with SAPs gradually slowed down as the age
increased after loading. The basic tensile creep coefficients of ICC with SAPs after
loading for 7 days were 1.63, 1.81, 2.25, and 2.84, which increased by 11.0%,
38.0%, and 74.2% when the SAPs content increased from 0% to 0.57%, 0.86%,
and 1.14% for mixtures IC0035, IC0635, IC0935, and IC1235, respectively. Exper-
imental results showed that the basic tensile creep coefficient also increased a little
when the SAPs content increased from 0% to 0.57%, while the basic tensile creep
coefficient increased significantly when the SAPs content increased from 0.57% to
0.86%, and 1.14%, respectively.
The test results showed that the initial elastic strain increased by 0.8%, 9.1%, and
12.2% when the SAPs content increased from 0% to 0.57%, 0.86%, and 1.14% for
mixtures IC0035, IC0635, IC0935, and IC1235, respectively. The increase degree of
basic tensile creep of ICC with SAPs was larger than that of elastic strain as SAPs
content increased. Therefore, the basic tensile creep coefficient of ICC also increased
as SAPs content increased according to Eq. (4.8). This phenomenon showed that the
tensile creep behavior of the ICC with SAPs was not only related to the elastic
modulus of the concrete when it was loaded, but also to the change of microstructure
with the introduction of IC water. The positive influence of IC water on concrete is
to promote the cement hydration and make the internal structure of concrete dense,
while negative influence of IC water is to increase the total w/c ratio of concrete

Fig. 4.5 Relationship 3

between basic tensile creep
Basic tensile creep coefficient

coefficient and age for

different concrete mixtures
2

1
IC0035
IC0635
IC0935
IC1235
0

0 24 48 72 96 120 144 168 192

Age (h)

@seismicisolation
@seismicisolation
186 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

and the number of macro pores inside the matrix [33]. The positive influence of IC
water on concrete might be slightly lower than the negative influence when the SAPs
content increased from 0% to 0.57%, and the tensile creep coefficient increased a
little. The increase of SAPs content and additional IC water would contribute to the
larger amount of voids, which resulted in the large deformability and increasing creep
and relaxation, as reported in [33]. Tensile creep characteristics of HPC utilizing w/c
ratios of 0.29, 0.33, and 0.37 are studied in [34], which show that the tensile creep
coefficient of concrete increases as w/c increases.

4.2.5 Specific Tensile Creep of Concrete Internally Cured

with SAPs

There are the calculated results of early-age specific tensile creep of ICC utilizing
different SAPs contents as shown in Fig. 4.6. The specific tensile creep of ICC
with SAPs after loading for 7 days were 50.4, 61.8, 84.3, and 113.3 με/MPa, which
increased by 22.6%, 67.3%, and 124.8% when the SAPs content increased from
0% to 0.57%, 0.86%, and 1.14% for mixtures IC0035, IC0635, IC0935, and IC1235,
respectively. The applied stress to the loaded specimens was 0.819, 0.749, 0.739, and
0.714 MPa, which decreased by 8.5%, 9.8%, 12.8% when the SAPs content increased
from 0% to 0.57%, 0.86%, and 1.14% for mixtures IC0035, IC0635, IC0935, and
IC1235, respectively. Therefore, similar to the results of basic tensile creep, the
specific tensile creep of ICC with SAPs also increased as SAPs content increased.
All the tensile creep, basic tensile creep coefficient, and specific tensile creep of
ICC increased as SAPs content increased, which showed that the addition of SAPs
increased the tensile creep of ICC with SAPs not only by reducing the elastic modulus
and strength of concrete, but also by influencing the microstructure of concrete. The

Fig. 4.6 Relationship 120

between specific tensile
Specific tensile creep (με/MPa)

creep and age for different 100

concrete mixtures
80

60

40
IC0035
IC0635
20
IC0935
IC1235
0
0 24 48 72 96 120 144 168 192
Age (h)

@seismicisolation
@seismicisolation
4.2 Tensile Creep of Early-Age Concrete with Different Amount of SAPs 187

total w/c ratio increased as the IC water increased, which made the internal structure
of concrete loose and increased the capillary aperture, as reported in [28]. Meanwhile,
the addition of SAPs could promote the cement hydration, but the internal structure of
cement-based materials would become more porous as a consequence of the SAPs
addition, as reported in [35]. SAPs swelled after absorbing water, collapsed after
releasing water, and left a certain amount of large pores in the hardened concrete,
as reported in [36]. Results of mercury intrusion porosimetry (MIP) reported in [37]
show that the addition of pre-soaked SAPs certainly increases the total porosity of
the cement paste when compared to the reference concrete HSC-0, because of the
extra IC water entrained by SAPs. Results of 3D volume analysis reported in [38]
show that the voids in the concrete matrix increases as the SAPs content increases.
Some theories have been developed to explain the mechanism of creep, and the main
assumptions are the seepage theory, viscous flow hypothesis, viscoelastic theory, and
microcrack theory [39]. The viscous flow hypothesis assumes that creep is caused
by shearing or sliding of calcium silicate hydrate gel under load. The addition of
SAPs would accelerate the formation of hydration products such as calcium silicate
hydrate gel, as reported in [40]. Therefore, the viscous flow generated by the calcium
silicate hydrate gel in cement paste was easier to migrate into the capillary due to
the large porosity and loose pore structure, which resulted in the increase of tensile
creep, as reported in [28]. The sensitivity of impact factors of early-age tensile creep
of concrete is studied in [41], which shows that the tensile creep and the w/c ratios are
basically positively related, and the specific tensile creep increases as the w/c ratio
increases. However, the influence of IC water introduced by SAPs was not exactly
the same as the influence of directly adding water. Therefore, the mechanism of SAPs
on tensile creep of ICC needs to be further investigated in the following studies.
The development rate of tensile creep decreased until it was stable as the age
increased after loading. The development of tensile creep could be roughly divided
into three stages, as reported in [19]. Stage I (initial loading age to 3 d) was the
zone where the tensile creep developed quickly. The internal hydration degree of
concrete was low and the pore structure was loose when the concrete was loaded in
the initial stage. Therefore, the elastic modulus and strength of concrete were low,
and a large creep phenomenon was found in Stage I. Stage II (loading age 3–7 d)
was the zone where the basic tensile creep developed slowly. The cement hydration
degree gradually increased as the age increased, and the internal structure gradually
became dense, so the development rate of tensile creep gradually decreased. Stage
III (loading age 7 d to unloading age) was the stable development zone, and the
development of tensile creep gradually stabilized as the concrete strength and degree
of cement hydration increased. Therefore, the development of tensile creep of ICC
with SAPs within 7 days belonged to stage I and II. Similar tensile creep development
results are reported in [28], which show that the tensile creep velocities have a law
of exponential decay.

@seismicisolation
@seismicisolation
188 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

4.2.6 Summary

Experimental studies on the influence of SAPs on the early-age behavior and tensile
creep of ICC under a constant initial stress/strength ratio were conducted. The devel-
opment of mechanical properties, autogenous shrinkage, and tensile creep of ICC
with SAPs with four mixture proportions on the basis of test results were investigated.
The conclusions obtained from this research were as follows.
(1) The compressive strength, splitting tensile strength, and elastic modulus of ICC
with the same basic w/c ratio decreased as SAPs content increased, and the
decrease of early-age compressive strength of ICC with SAPs decreased as the
age increased from 3 to 7 days at the same time.
(2) The basic tensile creep of ICC increased nonlinearly as SAPs content increased.
The basic tensile creep increased only a little when the SAPs content increased
from 0% to 0.57%, while the basic tensile creep increased significantly when
the SAPs content increased from 0.57% to 0.86%, and 1.14%, respectively.
(3) The basic tensile creep coefficient and specific tensile creep of ICC utilizing
SAPs were higher than that of reference concrete with the same basic w/c ratio.
The possible explanation for the results was that the internal structure of ICC
would become more porous as a consequence of the SAPs addition, which
contributed to the increase of viscous flow.

4.3 Tensile Creep of Early-Age Internally Cured Concrete

with Different Initial Stress/strength Ratios

4.3.1 Mixture Proportions and Materials

The basic w/c ratio was 0.30. The IC w/c ratio was designed as 0.12 and corresponding
IC water required was 57.6 kg/m3 . Therefore, the quantity of IC water was adjusted
by adding the amount of dry SAPs to meet the required total w/c ratio of 0.42.
The proportions of concrete mixture are shown in Table 4.5. Four groups of initial
stress/strength ratios were designed. The nomenclature utilized was related to the
type of concrete utilized (i.e., IC means ICC with SAPs) and initial stress/strength
ratio (i.e., 20, 35, 50, and 65 means that the initial stress/strength ratio was 20%,
35%, 50%, and 65%, respectively). Four groups of initial stress/strength ratios were
designated as IC-20, IC-35, IC-50, and IC-65, respectively.

@seismicisolation
@seismicisolation
4.3 Tensile Creep of Early-Age Internally Cured Concrete with Different … 189

Table 4.5 Concrete mixture proportions

Water (kg/m3 ) Cement Fine aggregate Coarse SAPs (kg/m3 ) IC water
(kg/m3 ) (kg/m3 ) aggregate (kg/m3 )
(kg/m3 )
144 480 690 1127 5.49 57.6

4.3.2 Basic Tensile Creep of Concrete with Different Initial

Stress/strength Ratios

The free and restrained shrinkage specimens were sealed after pouring into respective
molds in TSTM system, moisture exchange from the internal concrete to surrounding
environment at early age was not significant. In this way, drying creep is not obvious
and can be ignored [29, 42]. Therefore, only basic tensile creep was taken into
consideration, as reported in [43].
During the tensile creep tests, the initial stress/strength ratios were applied in real
time. That is to say, the tensile stress was controlled at a constant value and applied
on the restrained shrinkage specimen at the age of 1 d, which was based on the
method proposed in [3, 8]. The actual applied stress level was decreased because
of the increased concrete strength. If the initial stress/strength ratio was calculated
as the ratio of applied constant tensile stress to the concrete strength at 28 d, the
applied tensile stress for all specimens at 1 d was relatively at a high level (29%,
51%, 74%, and 96% for IC-20, IC-35, IC-50, and IC-65, respectively), which can
lead to very different results in terms of creep, as reported in [44]. Then, the effect
of initial stress/strength ratios on the tensile creep could not be reflected actually.
There are the total strains measured for restrained shrinkage specimens under
different initial stress/strength ratios, as shown in Fig. 4.7.
The total strain increased as the initial stress/strength ratio increased and was
above zero lines for IC-65. The basic tensile creep increased greatly and nonlinear

Fig. 4.7 Total strain of 40

restrained shrinkage
specimen 20

0 IC-20
IC-35
Total strain

-20 IC-50
IC-65
-40

-60

-80
1 2 3 4 5 6 7 8
Age (d)

@seismicisolation
@seismicisolation
190 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Fig. 4.8 Basic tensile creep 160

of specimens under different 140 IC-20
initial stress/strength ratios
IC-35
120 IC-50

Basic tensile creep (

100 IC-65

80

60
40

20

1 2 3 4 5 6 7 8
Age (d)

basic tensile creep occurred when the total strain was above zero line. As shown in
Fig. 4.8, the early-age basic tensile creep of ICC with SAPs increased gradually as
the initial stress/strength ratio increased.
The basic tensile creep at 7 d of loading was 37, 66, 95, and 148 με, which
increased by 78.4%, 156.8%, and 300.0% with the increasing initial stress/strength
ratio from 20 to 35%, 50%, and 65%, respectively. Similar results are reported in [45,
46]. Result reported in [45] shows that early-age tensile creep of concrete increases
as the tensile stress increases from 0.50 to 1.00 and 2.00 MPa. Result reported in
[46] shows that early-age tensile creep of concrete increases linearly with the applied
tensile stress when the initial stress/strength ratio is lower than 50%.
Under the similar total w/c ratio, the tensile creep of ordinary concrete at 7 d of
loading is about 85 με when the initial stress/strength ratio is 19% [47], which is large
than that of IC-20. According to the assumption that basic tensile creep is mainly
caused by the creation of the micro-cracks during the static loading, the possible
explanation may be that the addition of SAPs can reduce the micro-cracks under
constant tensile loading. The interfacial transition zones between the cement paste
and aggregate are often regarded as initial micro-cracks in concrete which will prop-
agate even under small constant loading [48]. The addition of SAPs improves the
microstructure in concrete [49, 50]. The denser matrix of concrete with SAPs may
enhance the interface adhesion of interface transition zones and restrain the prop-
agation of micro-cracks under loading, which resulted in the reduced basic tensile
creep of ICC with SAPs.

@seismicisolation
@seismicisolation
4.3 Tensile Creep of Early-Age Internally Cured Concrete with Different … 191

4.3.3 Basic Tensile Creep Coefficient of Concrete

with Different Initial Stress/strength Ratios

The splitting tensile strength was 2.38, 2.89, 3.33, and 3.56 MPa at the age of 1, 3,
7, and 28 d, respectively. Axial tensile strength at different ages could be obtained
from Eqs. (4.2) and (4.3). Coefficients λ1 = 0.171 and k1 = 1.069 with an R 2 value
of 0.99 were obtained from the regression analysis. The calculated value of axial
tensile strength at the age of 1 d was 2.01 MPa. Therefore, the tensile stress applied
by TSTM system was 0.40, 0.70, 1.00, and 1.31 MPa for IC-20, IC-35, IC-50, and
IC-65, respectively.
The test results of compression elastic modulus were 25.05, 32.56, 38.20, and
42.14 GPa at the age of 1, 3, 7, and 28 d, respectively. Similar test method for elastic
modulus calculated from the slope of the most linear unloading branch can also be
found in [51–53]. Although the elastic modulus according to Chinese standard GB/T
50081-2019 [24] is measured by the slope of the loading branch, the test results
of elastic modulus obtained from the initial loading portion and the unloading part
are approximately the same [54]. The time-dependent development elastic tensile
modulus E t (t) of the concrete at age t could be obtained from Eq. (4.4). Coefficients
λ2 = 0.240 and k2 = 1.083 with an R 2 value of 0.98 were obtained from the regression
analysis. The compression elastic modulus obtained from tests on prisms could be
utilized as elastic tensile modulus in Eq. (4.8) because the values of secant tensile
and compressive elastic modulus were approximately the same at early age. The
calculated value of elastic tensilemodulus
 at the age of 1 d was 24.47 GPa. Therefore,
the initial elastic deformation εe t  at the age of 1 d (loading age) was 16.35, 28.60,
40.87, and 53.53 με for IC-20, IC-35, IC-50, and IC-65, respectively.
As shown in Fig. 4.9, the basic tensile creep coefficients of ICC with SAPs at 7
d of loading were 2.27, 2.30, 2.33, and 2.76, which increased by 1.3%, 2.6%, and
21.6% with the increasing initial stress/strength ratio from 20 to 35%, 50%, and 65%,
respectively.

Fig. 4.9 Basic tensile creep 3.0

coefficient of specimens
Basic tensile creep coefficient

under different initial 2.5

stress/strength ratios
2.0

1.5
IC-20
1.0 IC-35
IC-50
IC-65
0.5

0.0

1 2 3 4 5 6 7 8
Age (d)

@seismicisolation
@seismicisolation
192 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

The basic tensile creep coefficient increased as the initial stress/strength ratio
increased, which was in accordance with the results reported in [13, 55]. Results
reported in [13] show that increasing the initial stress/strength ratio from 40% to 60%
increases the tensile creep coefficient of ultra-high performance concrete increases
by 44% at 14 d of loading. Results reported in [55] also show that the early-age tensile
creep coefficient of concrete increases with the increase of initial stress/strength ratio
(40%–50%). The early-age basic tensile creep coefficient was basically unchanged
at 20%–50% of the initial stress/strength ratio, however, a sudden change of the
basic tensile creep coefficient occurred at 50%–65% of the initial stress/strength
ratio, which illustrated that the basic tensile creep showed a nonlinear increase when
a certain range of the initial stress/strength ratio was exceeded. Furthermore, the
development rate of early-age basic tensile creep coefficient decreased gradually as
the time after loading increased.
Generally, the additional water entrained by SAPs may increase the number of
capillary pores in the cementitious matrix after hardening, and a considerable number
of voids are left in the hardened cementitious matrix when the water is desorbed from
SAPs, which diminishes the mechanical properties of cement-based materials. Since
the basic tensile creep coefficient was the ratio of basic tensile creep strain to initial
elastic strain, the elastic modulus of ICC with SAPs should be low the initial elastic
strain was large. Therefore, the incorporation of SAPs was supposed to result in a
low basic tensile creep coefficient. However, under the same total w/c ratio of 0.42,
the basic tensile creep coefficient of ordinary concrete with strength/stress ratio of
60% loaded at 28 d in [56] is lower than that of IC-65. A possible reason may be that
early-age tensile creep of concrete decreased as the loading age increases.

4.3.4 Basic Tensile Creep Rate of Concrete with Different

Initial Stress/strength Ratios

The basic tensile creep rate reflected the change of development rate of tensile creep
of ICC with SAPs under different initial stress/strength ratios and could be obtained
from Eq. (4.7).
As shown in Fig. 4.10, the basic tensile creep rate of ICC with SAPs decreased as
the time after loading increased. The basic tensile creep rate at 1 d of loading was 0.57,
1.01, 1.70, and 3.04 με/h, which increased by 77.2%, 198.2%, and 433.3% with the
increasing initial stress/strength ratio from 20 to 35%, 50%, and 65%, respectively.
Therefore, the basic tensile creep rate of ICC with SAPs increased as the initial
stress/strength ratio increased.
The trend of the relationship curves between basic tensile creep rate and time
after loading could be roughly divided into three stages. Stage I was the zone where
the basic tensile creep rate developed quickly and showed an obvious decrease. At
initial time after loading, the degree of hydration of cement was low. Therefore,
the basic tensile creep increased rapidly because of the low tensile strength and

@seismicisolation
@seismicisolation
4.3 Tensile Creep of Early-Age Internally Cured Concrete with Different … 193

Fig. 4.10 Basic tensile

creep rate of specimens
under different initial
stress/strength ratios

elastic modulus of concrete. Stage II was the zone where the basic tensile creep rate
developed slowly. The degree of hydration of cement increased gradually and the
internal structure become dense which induced a low reduction of the basic tensile
creep rate. Stage III was the zone where the value of basic tensile creep rate was almost
unchanged. The basic tensile creep rate under different initial stress/strength ratios
gradually stabilized and tended to be constant with the further enhanced strength
and dense internal structure of concrete, which correlated with the results reported
in [57, 58]. Results reported in [57] show that the early-age basic tensile creep rate is
pronounced due to that the majority of the basic tensile creep deformation occurs in
the first hours after loading and subsequently the development of basic tensile creep
stabilizes. Results reported in [58] show that the early-age basic tensile creep rate
is significant in the first 10–20 h after loading and then achieves a constant value
gradually.

4.3.5 Specific Basic Tensile Creep of Concrete with Different

Initial Stress/strength Ratios

The specific basic tensile creep of HSC could be obtained from Eq. (4.9). As shown
in Fig. 4.11, the specific basic tensile creep at 7 d of loading was 52, 53, 54, and
72 με/MPa, which increased by 1.9%, 3.8%, and 38.5% with the increasing initial
stress/strength ratio from 20% to 35%, 50%, and 65%, respectively.
The specific basic tensile creep at 7 d of loading increased as the initial
stress/strength ratio increased, which correlated with the results reported in [47,
59]. Results reported in [47] show that specific basic tensile creep is larger when
the applied load increases from 25% to 45% of the axial tensile strength. Results
reported in [59] show that the specific basic tensile creep of concrete at 28 d of
loading increases as the initial stress/strength ratio increases when the ratio ranges

@seismicisolation
@seismicisolation
194 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

80
Fig. 4.11 Specific basic
tensile creep of specimens
under different initial

Specific basic tensile creep

60
stress/strength ratios

40
IC-20
IC-35
20 IC-50
IC-65

1 2 3 4 5 6 7 8
Age (d)

from 30% to 50%. The relationship curves between specific basic tensile creep and
the time after loading were very close to each other for IC-20, IC-35, and IC-50.
However, the specific basic tensile creep showed a significant increase at 50% to
65% of the initial stress/strength ratio. Furthermore, the specific basic tensile creep
increased as the initial stress/strength ratio increased at the initial time after loading,
which indicated that the effect of initial stress/strength ratio on the early-age basic
tensile creep was significant.
Under the same basic w/c ratio of 0.30, the specific basic tensile creep of ordinary
concrete loaded at 1 d under strength/stress ratio of 0.3 in [8] is larger than that of IC-
35. Micro-cracks cause serious moisture imbalance in concrete, and vacuums (cracks
when they are formed) produce local humidity shocks, resulting in concentration and
pressure gradients of water molecules. Therefore, the existence of these two gradients
would cause water vapor (according to Fick’s law) and liquid water (according to
Darcy’s law) to move from the capillaries around the micro-cracks towards the micro-
cracks. Then, new micro-cracks were formed by the additional tensile stress induced
by the two kinds of movements. The water entrained in SAPs may compensate
moisture loss and then alleviate these movements. Thus the specific basic tensile
creep of ICC with SAPs was reduced as compared with that of ordinary concrete.

4.3.6 Nonlinear Basic Tensile Creep of Concrete

with Different Initial Stress/strength Ratios

The basic tensile creep at 1, 3, 5, and 7 d of loading showed a linear increase with
the initial stress/strength ratio for IC-20, IC-35, and IC-50, as shown in Fig. 4.12.
However, the basic tensile creep increased significantly when the initial
stress/strength ratio was larger than 50%. The amplification of creep deformation
was much higher than that of the initial stress/strength ratio, which meant that the
nonlinear basic tensile creep occurred. The results were in accordance with the results

@seismicisolation
@seismicisolation
4.3 Tensile Creep of Early-Age Internally Cured Concrete with Different … 195

Fig. 4.12 Relationship

between basic tensile creep
and initial stress/strength
ratio

reported in [57, 60]. Results reported in [60] show that the nonlinear tensile creep of
HPC occurs at 60% to 80% of the initial stress/strength ratio. Results reported in [57]
show that the basic tensile creep of ultra-high performance fiber reinforced concrete
presents a viscoplastic response, and the nonlinear basic tensile creep induced by
the simultaneous interaction of micro-cracks occurs when the initial stress/strength
ratio exceeds 35%. Results reported in [61] also show that the nonlinear basic tensile
creep occurs at an uncertain initial stress/strength ratio ranging from 40% to 50%.
The nonlinearity of the basic tensile creep, basic tensile creep coefficient, and
specific basic tensile creep occurred when the initial stress/strength ratio ranged from
50% to 65%. Therefore, it could be considered that the nonlinear basic tensile creep
behavior occurred at 50% to 65% of the initial stress/strength ratio. The nonlinear
tensile creep of ordinary concrete occurs when the initial stress/strength ratio is larger
than 39% [62], 35% [57], and 30% [3, 59]. The incorporation of SAPs increased the
proportionality limit of tensile creep to 50% because of the appearance and propa-
gation of cracks in ICC with SAPs was different from that of ordinary concrete. The
creep, physically sourcing from progressive micro-cracks, presents nonlinearly under
high stress levels [62]. Micro-cracks in concrete include the bond cracks that occur at
the interface between aggregates and mortar matrix and the cracks can penetrate the
mortar or aggregates, as well as the bond cracks that exist in concrete before loading
[63]. The bond cracks at the interface between aggregates and mortar matrix develop
first when the load increases gradually, however, the effect of the bond cracks can be
ignored when the applied initial stress/strength ratio is relatively small, as reported in
[64]. The relationship between initial stress/strength ratio and the creep deformation
was approximately linear at this stage. The bond cracks increased rapidly in quantity
and the total length when the initial stress/strength ratio exceeded a certain range.
Moreover, the mortar cracks occurred resulting in the aggregates to slipping along
a split surface of the mortar. Thus the nonlinear basic tensile creep occurred with
the unrecoverable creep deformation increased. However, the improved microstruc-
ture of ICC with SAPs may strengthen the interface adhesion of interface zones and

@seismicisolation
@seismicisolation
196 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

hinder the micro-cracks to propagate during loading. Therefore, for ICC with SAPs
with the same stress level, the number and total length of the micro-cracks were less
than that of ordinary concrete, and as a result the proportionality limit of tensile creep
was increased.

4.3.7 Prediction Model for Early Age Tensile Creep

Considering the Initial Stress/strength Ratio

The B3 model based on solidification theory is well-known for characterizing the

early-age tensile creep behavior [6]. The B4 model is improved according to the B3
model in two aspects: one improvement is introducing equivalent age to consider the
effect of temperature; the second is that drying and autogenous shrinkage are split
into separate functions [65]. All the specimens were cured at isothermal 20 °C and
the drying shrinkage was not considered. Therefore, the basic tensile creep could be
predicted by B3 model and could be obtained from Eq. (4.12).

 
  n
 t
J (t, t ) = q1 + q2 Q(t, t ) + q3 ln 1 + (t − t ) + q4 ln  (4.12)
t

where J (t, t ) is the tensile creep compliance function (basic tensile creep/ initial
constant tensile stress); q1 , q2 , q3 , and q4 are empirical constants, which repre-
sent the instantaneous compliance, the ageing viscoelastic compliance, non-ageing
viscoelastic compliance, and flow compliance, respectively, q1 =0.6 × 106 /E 28 ,
q2 =185.4c0.5 f c−0.9 , q3 =0.29(w / c)4 q2 , q4 =20.3(a / c)−0.7 , n = 0.1, E 28 is the elastic
modulus at the age of 28 d, in MPa, f c is the cylinder compressive strength at the age
of 28 days, and could be calculated as 0.79 times of the cubic compressive strength at
the age of 28 d, w / c is water-to-cement ratio, and a / c is aggregate-to-cement ratio;
and Q(t, t  ) could be obtained from Eqs. (4.13)–(4.16).

 r (t  ) −1/r (t  )
Q f (t  )
Q(t, t  ) = Q f (t  ) 1 + (4.13)
Z (t, t  )

Q f (t  ) = [0.086(t  )2/9 + 1.21(t  )4/9 ]−1 (4.14)

Z (t, t  ) = (t  )−m ln[1 + (t − t  )n ] (4.15)

 
r t  = 1.7(t  )0.12 + 8 (4.16)

where m = 0.5.

@seismicisolation
@seismicisolation
4.3 Tensile Creep of Early-Age Internally Cured Concrete with Different … 197

A model by Østergaard et al. [47] is proposed to seize the very early-age high
viscoelastic behavior under different initial stress/strength ratios through modifying
q4 in B3 Model. Therefore, the tensile creep compliance function J (t, t  ) of ICC
with SAPs under different initial stress/strength ratios was fitted based on Model in
[47] and could be expressed as Eq. (4.17).

 
  n
 t
J (t, t ) = q1 + q2 Q(t, t ) + q3ln 1 + (t − t ) + q4 q5 ln  (4.17)
t

where q5 is the influence coefficient of the initial stress/strength ratio.

The influence coefficient of the initial stress/strength ratio q5 was fitted according
to the test results of the tensile creep compliance function J (t, t  ) (basic tensile creep/
initial constant tensile stress) at 7 d of loading, that is to say, after obtaining the calcu-
lation results of q1 , q2 , q3 , and q4 , the fitted values q5 under four groups of initial
stress/strength ratios were adjusted until the J (t, t  ) predicted by Model in [47] at 7
d of loading reached the test results of J (t, t  ) at 7 d of loading. The coefficients q1 ,
q2 , q3 are fitted as empirical constants which are 6.7, 40.49, and 47.1, respectively
[47]. According to the fitted values q4 for different initial stress/strength ratios in
[47], the influence coefficient of the initial stress/strength ratio q5 should be 2.20,
2.50, 2.80, and 4.60 for specimens IC-20, IC-35, IC-50, IC-65, respectively. It can be
seen from the fitting results that q5 increased linearly when the initial stress/strength
ratio was 20%–50%, and when the ratio was more than 50%, q5 increased signifi-
cantly, showing a nonlinear change. However, the nonlinear relationship between q5
and the initial stress/strength ratio can not be fitted because there was no test data
between 50% and 65% of the initial stress/strength ratio. Comparisons of tensile
creep compliance function of specimens under different initial stress/strength ratios
from test results and Eq. (4.17) are shown in Fig. 4.13.
The deviations of the results obtained from test and Eq. (4.17) were small. There-
fore, the proposed model can be utilized to predict the tensile creep of ICC with
SAPs considering initial stress/strength ratios.

4.3.8 Summary

The effect of initial stress/strength ratio on the early-age basic tensile creep behavior
of ICC with SAPs was investigated. Analysis on the basic tensile creep, basic
tensile creep coefficient, basic tensile creep rate, and specific basic tensile creep
was conducted by utilizing TSTM. The following conclusions can be drawn.
(1) The early-age basic tensile creep of ICC with SAPs increased as the initial
stress/strength ratio increased and decreased as the loading ages increased. The
basic tensile creep at 7 d of loading was 37, 66, 95, and 148 με, which increased
by 78.4%, 156.8%, and 300.0% with the increasing initial stress/strength ratio
from 20% to 35%, 50%, and 65%, respectively. The basic tensile creep of ICC

@seismicisolation
@seismicisolation
198 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Fig. 4.13 Comparisons of tensile creep compliance function of specimens under different initial
stress/strength ratios from test results and proposed model

with SAPs at 7 d after loading was 116, 106, 97, 81, 55, 47, and 36 με when
the loading age was 0.50, 0.67, 0.83, 1, 3, 5, and 7 d, respectively.
(2) The early-age basic tensile creep coefficient of ICC with SAPs increased as the
initial stress/strength ratio increased and decreased as the loading ages increased.
The basic tensile creep coefficient at 7 d of loading was 2.27, 2.30, 2.33, and
2.76, which increased by 1.3%, 2.6%, and 21.6% with the increasing initial
stress/strength ratio from 20% to 35%, 50%, and 65%, respectively. The basic
tensile creep coefficient of ICC with SAPs at 7 d after loading was 4.45, 4.27,
4.05, 3.46, 2.39, 2.04, and 1.55 when the loading age was 0.50, 0.67, 0.83, 1, 3,
5, and 7 d, respectively.
(3) The early-age basic tensile creep rate of ICC with SAPs increased as the initial
stress/strength ratio increased. The basic tensile creep rate at 1 d of loading
was 0.57, 1.01, 1.70, and 3.04 με/h, which increased by 77.2%, 198.2%, and
433.3% with the increasing initial stress/strength ratio from 20% to 35%, 50%,
and 65%, respectively.
(4) The specific basic tensile creep of ICC with SAPs increased as the initial
stress/strength ratio increased and decreased as the loading ages increased. The
specific basic tensile creep at 7 d of loading was 52, 53, 54, and 72 με/MPa, which
increased by 1.9%, 3.8%, and 38.5% with the increasing initial stress/strength

@seismicisolation
@seismicisolation
4.4 Tensile Creep of Early-Age Internally Cured Concrete with Different … 199

ratio from 20% to 35%, 50%, and 65%, respectively. The specific basic tensile
creep of ICC with SAPs at 7 d after loading was 611, 286, 198, 140, 63, 51,
and 38 με/MPa when the loading age was 0.50, 0.67, 0.83, 1, 3, 5, and 7 d,
respectively.
(5) The relationship between the basic tensile creep and the initial stress/strength
ratio of the specimens at 1, 3, 5, and 7 d of loading was basically linear for
IC-20, IC-35, and IC-50. However, a nonlinear relationship occurred for IC-65
with a significant increase of the basic tensile creep. Therefore, the nonlinear
basic tensile creep of ICC with SAPs occurred when the initial stress/strength
ratio ranged from 50% to 65%.
(6) A model for predicting the tensile creep of ICC with SAPs in considera-
tion of initial stress/strength ratios was proposed based on Østergaard model.
Further experimental investigations on the nonlinear tensile creep with the initial
stress/strength ratio ranging from 50% to 65% were needed to build the nonlinear
model for early-age tensile creep of ICC with SAPs considering the initial
stress/strength ratio.

4.4 Tensile Creep of Early-Age Internally Cured Concrete

with Different Loading Ages

4.4.1 Mixture Proportions and Materials

The proportions of concrete mixture are shown in Table 4.5.

4.4.2 Basic Tensile Creep of Concrete with Different Loading

Ages

The surfaces of the concrete specimens were wrapped with plastic film as soon as
the concrete was cast into molds to minimize the exchange of moisture between
the concrete inside the molds and the surrounding environment. That means drying
creep of sealed HSC specimens is not obvious and can be ignored [29, 42]. The
total deformations measured for restrained shrinkage specimens are illustrated in
Fig. 4.14.
The total deformations of restrained shrinkage specimens decreased with the
increasing loading age. The εcr was calculated by subtracting the deformation of the
free shrinkage specimen and the instantaneous elastic strain of restrained shrinkage
specimen from the measured total strain of restrained shrinkage specimen and could
be determined by Eq. (4.5).
The instantaneous elastic strain could be formulated as the ratio of exerted tensile
stress to the elastic modulus. The tensile stress was 35% of the axial tensile strength

@seismicisolation
@seismicisolation
200 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Fig. 4.14 Total deformation 0.50 d

80
of restrained shrinkage 0.67 d
specimens at different 0.83 d
60 1d
loading ages

Total deformation (με)

3d
40 5d
7d
20

-20

-40

0 2 4 6 8 10 12 14
Age (d)

at the age of 0.50, 0.67, 0.83, 1, 3, 5, and 7 d, respectively. The axial tensile strength
f t was calculated by Eqs. (4.2) and (4.3).
According to the equations, the splitting tensile strength of ICC with SAPs was
1.80, 2.82, 3.40, and 3.65 MPa at the age of 1, 3, 7, and 28 d, respectively. The
coefficients for time-dependent axial tensile strength were modified based on the
splitting tensile strength of ICC with SAPs obtained from the test. Based on the
regression analysis of the test results for splitting tensile strength, the coefficients
λ1 = 0.309 and k1 = 1.431 with an R 2 value of 0.99 were obtained. The fitted
axial tensile strength and the calculated value of constant tensile stress exerted on
restrained shrinkage specimens was 0.55, 1.05, 1.40, 1.66, 2.50, 2.66, and 2.73 MPa
and 0.19, 0.37, 0.49, 0.58, 0.88, 0.93, and 0.96 MPa when the loading age was 0.50,
0.67, 0.83, 1, 3, 5, and 7 d, respectively.
The compressive elastic modulus was 24.98, 36.85, 42.51, and 44.08 GPa at the
age of 1, 3, 7, and 28 d, respectively. The coefficients for time-dependent tensile
elastic modulus E t in Eq. (4.4) were modified based on the experimental results of
elastic modulus.
Based on the regression analysis of the test results for elastic modulus, the coeffi-
cients λ2 = 0.264 and k2 = 1.939 with an R 2 value of 0.99 were obtained. The elastic
modulus of early-age concrete under tensile state was regarded to be equal to that
under compressive state, as reported in [48]. The fitted tensile elastic modulus and
the calculated value of instantaneous elastic strain were 7.29, 14.89, 20.45, 24.74,
38.22, 40.45, and 41.34 GPa and 26.06, 24.85, 23.96, 23.44, 23.02, 22.99, and 23.22
με when the loading age was 0.50, 0.67, 0.83, 1, 3, 5, and 7 d, respectively.
After obtaining the instantaneous elastic strains of ICC with SAPs at different
loading ages, the basic tensile creep can be calculated. The basic tensile creep of
ICC with SAPs at 7 d after loading was 116, 106, 97, 81, 55, 47, and 36 με when the
loading age was 0.50, 0.67, 0.83, 1, 3, 5, and 7 d, respectively, as shown in Fig. 4.15.
The aging effect on the basic tensile creep was significant within 1 d. The basic
tensile creep at 7 d after loading significantly increased when the concrete was loaded

@seismicisolation
@seismicisolation
4.4 Tensile Creep of Early-Age Internally Cured Concrete with Different … 201

Fig. 4.15 Basic tensile 120

creep of specimens at 0.50 d
different loading ages 100 0.67 d

Basic tensile creep (με )

0.83 d
1d
80
3d
5d
60 7d

40

20

0 2 4 6 8 10 12 14
Age (d)

within 1 d and decreased linearly with the increasing loading age ranging from 3 to
7 d. A similar result can also be found in [56]. The investigation in [56] indicates
that the effect of loading age on the early-age tensile creep of HSC with admixtures
becomes small when the concrete is loaded after 3 d. This phenomenon may be
explained by the short-term micro-diffusion of water between the capillary pores.
At the microcosmic scale, the tensile stress in the cement paste is transferred to the
hydrates, which results in the diffusion of water molecules from the free adsorbed
water layer to free zones, and a deformation of the solid skeleton in concrete occurs
[66]. The degree of cement hydration was low at early loading ages, and the hydrates
were not sufficient, which led to large deformations of the solid skeleton in concrete.
Accordingly, the basic tensile creep was large at early loading ages.
Under a similar total w/c ratio, the basic tensile creep strain of ordinary concrete
loaded at 1 d in the study done by [47] is about 91 με which is 12.3% larger than that
of SAPs modified concrete. Results in [47] suggests that the basic tensile creep is
assumed to be mainly induced by the development of the microcracks under sustained
loading. Regarding the concrete mixtures under equal total w/c ratio, the incorpo-
ration of SAPs can reduce the voids in the cement-based matrix and strengthen the
interface adhesion between the cement paste and aggregate, and the denser matrix
enhanced by SAPs might restrict the development of microcracks [49, 50]. Accord-
ingly, the basic tensile creep of ICC with SAPs was reduced compared to that of
ordinary concrete.
The normalized basic tensile creep, which was defined as the ratio of basic tensile
creep at an arbitrary time to that at 7 d after loading, is illustrated in Fig. 4.16.
A similar definition can also be found in [67]. The normalized basic tensile creep
at the loading age within 1 d was greater than that at the loading age ranging from
3 to 7 d. The normalized basic tensile creep decreased with the increasing loading
age from 0.50 d to 1 d and was approximately equal when the concrete was loaded
at 3, 5, and 7 d. The tensile strength and elastic modulus of concrete are low at early
loading ages, and the internal microstructure is not dense. The creep is inversely

@seismicisolation
@seismicisolation
202 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Fig. 4.16 Normalized basic 100

tensile creep of specimens at

Normalized basic tensile creep (%)

different loading ages
80

60 0.50 d
0.67 d
0.83 d
40 1d
3d
5d
20
7d

0 1 2 3 4 5 6 7
Time after loading (d)

proportional to the strength of concrete. Accordingly, the basic tensile creep at early
ages was large and increased rapidly. This phenomenon was significant when the
concrete was loaded within 1 d. A similar phenomenon can also be found in [47].
With the increase of the loading age, the hydration reaction continued to proceed,
the strength of concrete was continuously increased, and the growth rate of tensile
creep was gradually slowed down. The improvement of the strength was relatively
slow when the concrete was loaded at 3, 5, and 7 d, and the differences of normalized
basic tensile creep at different loading ages were relatively small.

4.4.3 Basic Tensile Creep Coefficient of Concrete

with Different Loading Ages

In the structural analysis of reinforced and pre-stressed concrete structures, the creep
phenomenon was mainly treated by the creep coefficient which was calculated as
the ratio of the creep strain to the instantaneous elastic strain at loading. The basic
tensile creep coefficient could be determined by Eq. (4.8), as reported in [68].
The basic tensile creep coefficient of ICC with SAPs at 7 d after loading was 4.45,
4.27, 4.05, 3.46, 2.39, 2.04, and 1.55 when the loading age was 0.50, 0.67, 0.83, 1,
3, 5, and 7 d, respectively, as illustrated in Fig. 4.17.
The basic tensile creep coefficient at the first few hours after loading showed
a strong age dependency and was high at the early loading age. Furthermore, the
growth rate of the creep coefficient decreased with the increasing loading age. This
suggested that the creep coefficient had a strong dependency on the loading ages.
Normally, the external water absorbed in SAPs will increase the capillary pores
in the solidified cement matrix. After the entrained water in SAPs is released, a large
number of voids remain in the cement-based materials, thereby reducing the elastic

@seismicisolation
@seismicisolation
4.4 Tensile Creep of Early-Age Internally Cured Concrete with Different … 203

Fig. 4.17 Basic tensile 0.50 d

creep coefficient of 4 0.67 d
specimens at different 0.83 d

Basic tensile creep coefficient

loading ages 1d
3 3d
5d
7d
2

0 2 4 6 8 10 12 14
Age (d)

modulus of concrete. The elastic modulus of ICC with SAPs was supposed to be rela-
tively small, and then the instantaneous elastic strain should be large. Accordingly,
the inclusion of SAPs will lead to a reduced basic tensile creep coefficient. Under
the same basic w/c ratio, the creep coefficient of SAPs modified concrete loaded at
28 d is about 0.80 which is 20% lower than that of reference concrete.

4.4.4 Specific Basic Tensile Creep of Concrete with Different

Loading Ages

The restrained shrinkage specimens were under different constant tensile stresses.
The ratio of the cumulative basic tensile creep of concrete at the age of t caused by
constant tensile stress exerted at the age of t  to the exerted tensile stress was defined
as the specific basic tensile creep which could normalize the tensile creep under
different stress levels. The specific basic tensile creep of HSC could be calculated
utilizing Eq. (4.9).
Among then, σ is the exerted constant tensile stress (0.19, 0.37, 0.49, 0.58, 0.88,
0.93, and 0.96 MPa when the loading age was 0.50, 0.67, 0.83, 1, 3, 5, and 7 d,
respectively), in MPa.
The specific basic tensile creep of ICC with SAPs at 7 d after loading was 611,
286, 198, 140, 63, 51, and 38 με/MPa when the loading age was 0.50, 0.67, 0.83, 1,
3, 5, and 7 d, respectively, as illustrated in Fig. 4.18.
The amplification of the specific creep became large when the loading age changed
from 3 to 1 d. The specific creep at the very early loading age was extremely large,
and the changes of concrete loaded at 3, 5, and 7 d were relatively small. The specific
creep of ICC with SAPs decreased with the increasing loading age, which agreed well
with the investigation done by [8]. The investigation in [8] indicates that the specific

@seismicisolation
@seismicisolation
204 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Fig. 4.18 Specific basic 600 0.50 d

tensile creep of specimens at 0.67 d

Specific basic tensile creep

different loading ages 500 0.83 d
1d
400 3d

(με/MPa)
5d
300 7d

200

100

0 2 4 6 8 10 12 14
Age (d)

tensile creep of concrete with mineral additives decreases with the increasing loading
age and the aging effect is less pronounced when the loading age is later than 3 d.
Regarding concrete mixtures under equal basic w/c ratio of 0.30, the specific basic
tensile creep of ordinary concrete loaded at 1 d is about 150 με/MPa in the study done
by [8], which is 7.1% larger than that of SAPs modified concrete at the same loading
age, respectively. The imbalance of moisture may occur when the microcracks form,
leading to the pressure and concentration gradients of water molecules. Based on
Darcy’s law and Fick’s law, liquid water and water vapor under local humidity shocks
may move from the capillaries around the microcracks towards the microcracks.
These movements could cause additional tensile stress and promote the formation of
new microcracks. The inclusion of SAPs could reduce the moisture loss by releasing
the IC water, thus alleviating the two kinds of movements. Accordingly, the specific
basic tensile creep of ICC with SAPs with additional IC water was normally low
compared to that of ordinary concrete.

4.4.5 Specific Basic Tensile Creep Rate of Concrete

with Different Loading Ages

According to the Eq. (4.10), the specific basic tensile creep rates of ICC with SAPs
at 0.50 and 3 d after loading are shown in Fig. 4.19.
The nonlinear specific basic tensile creep rate occurred when the loading age
changed from 1 to 3 d. The specific basic tensile creep rate could reflect the rela-
tionship between the strength of concrete and the loading age. The creep rate was
relatively large at an early loading age, then decreased and tended to be stable. More-
over, the earlier the loading age was, the greater the variation between the initial creep
rate and the later creep rate at the same loading age was. For different loading ages,
the differences of the creep rate at 3 d after loading were small, which indicated
that the aging effect on the creep rate became small at 3 d after loading. When the

@seismicisolation
@seismicisolation
4.4 Tensile Creep of Early-Age Internally Cured Concrete with Different … 205

180
Fig. 4.19 Relationship
between specific basic tensile 160
creep rate and loading age

Specific basic tensile creep rate

140

120 0.50 d after loading

(με/(MPa·d))
3 d after loading
100 Fitting curve of 0.50 d after loading
Fitting curve of 3 d after loading
80

60

40

20

0
0.5 1 3 5 7
Loading age (d)

concrete was loaded at an early age, the hydration reaction was not sufficient and
the internal structure was not yet dense, thereby it showed great creep capacity. With
the growth of the age, the cement was constantly hydrated and the structure became
denser. Thus the creep capacity of concrete loaded at later ages was weakened. A
similar phenomenon is also found in [28]. The investigation in [28] indicates that
the creep rate reaches its maximum when the loading age is 0.50 d and then declines
exponentially with the increase of the loading ages.

4.4.6 Nonlinear Specific Basic Tensile Creep of Concrete

with Different Loading Ages

The relative specific basic tensile creep, which represented the ratio of the specific
basic tensile creep at 7 d after loading at arbitrary loading ages to that at 1 d after
loading, is illustrated in Fig. 4.20.
The relative specific basic tensile creep reduced exponentially with the increasing
loading age. When the loading age increased from 0.50 d to 0.67, 0.83, and 1 d,
the decreased degree of the relative specific basic tensile creep became low, and
the relationship between the relative specific basic tensile creep and loading age
was nonlinear. The relative specific basic tensile creep decreased linearly when the
loading age was later than 3 d. Hence the nonlinear basic tensile creep occurred
when the loading age changed from 1 to 3 d. It was assumed that the relative specific
basic tensile creep decreased exponentially when the loading age was within 1 d and
decreased linearly when the loading age was later than 3 d. The relative specific basic
tensile creep of ICC with SAPs could be calculated utilizing Eq. (4.18).

 β1 [exp(−β2 (t  − 1)) − 1] + 1 t  ≤ 1
ϕ(t )= (4.18)
−β3 t  + β4 t  ≥ 3

@seismicisolation
@seismicisolation
206 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Fig. 4.20 Relationship 5.0

Relative specific basic tensile creep

between relative specific 4.5
basic tensile creep and 4.0
loading age
3.5
Test result
3.0 Nonlinear stage
Linear stage
2.5
2.0
1.5
1.0
0.5
0.0
0.5 1 1.6 3 5 7
Loading age (d)

 
where ϕ t  is the relative specific basic tensile creep at the loading age t  ; and β1 ,
β2 , β3 , and β4 are the regression coefficients.
By the regression analysis of Eq. (4.18), the proportionality limit of specific basic
tensile creep of ICC with SAPs occurred at the loading age of 1.60 d. The results of
β1 , β2 , β3 , and β4 were 0.540, 3.849, 0.045, and 0.587, respectively. Thus the relative
specific basic tensile creep of ICC with SAPs can be obtained from Eq. (4.19).

 0.540[exp(−3.849(t  − 1)) − 1] + 1 t  < 1.60
ϕ(t )= (4.19)
−0.045t  + 0.587 t  ≥ 1.60

The correlation between loading age and the specific basic tensile creep changed
from nonlinear to linear after 1.60 d. The earlier the loading age was, the greater
the increase of the specific basic tensile creep. This phenomenon was related to the
mechanical properties of concrete, as reported in [7, 57]. The investigation in [7]
indicates that high strength could reduce the creep deformation of concrete. The
investigation in [57] indicates that the tensile creep response is more sensitive to
the loading age and negatively correlated with the loading age because of the low
material stiffness at early age. When the concrete was loaded within 1 d, the strength
and elastic modulus of concrete were low, and consequently, the concrete showed a
great creep capacity. With the development of cement hydration reaction, the strength
and elastic modulus of concrete increased rapidly. Accordingly, the tensile creep of
concrete loaded at later ages was low. As illustrated in Fig. 4.21, the compressive
strength, the splitting tensile strength, and the compressive elastic modulus of ICC
with SAPs at the age of 1.60 d reached 60%, 69%, and 75% of the corresponding
strength at 28 d, respectively.
The strength of concrete was relatively high at 1.60 d and changed from the rapid
development phase to the slow growth phase after 1.60 d. Thus specific basic tensile
creep curves of ICC with SAPs loaded at 3, 5, and 7 d were closed to each other and
exhibited linear characteristics.

@seismicisolation
@seismicisolation
4.4 Tensile Creep of Early-Age Internally Cured Concrete with Different … 207

1.0

Relative mechanical properties

0.8 0.75
0.69

0.6 0.60

Compressive strength
Splitting tensile strength
0.4 Elastic modulus

0.2

0.0
0 0.5 1 1.6 3 5 7
Age (d)

Fig. 4.21 Relationship between relative mechanical properties of concrete and loading age

The stress-induced dissolution is an important mechanism for the time-dependent

viscoelastic/viscoplastic behavior of cementitious materials [69]. The gradual disso-
lution of cement grains induces a redistribution of the stress from dissolving, load-
bearing cement grains to precipitate hydration products, increasing the elastic defor-
mation of the remaining solid constituents, and yielding viscoelastic/viscoplastic
behavior of cement paste on the bulk scale [70]. As the hydrated phases normally
carry most of the stress at later ages, the dissolution of hydrated phases is expected
to cause significant stress redistribution to other parts of the load-bearing network,
manifested as substantial creep in bulk cement composites [70]. The mechanism
of stress-induced dissolution of solid constituents can increase the overall creep of
cement paste. During cement hydration, the water entrained in SAPs is gradually
released within the first 2 days [71], which may increase the moisture in the mortar
and reduce solid volume, thereby accelerating the stress-induced dissolution of solid
constituents. Therefore, large tensile creep of ICC with SAPs loaded before 2 d was
observed.

4.4.7 Modeling Tensile Creep of Concrete with Different

Loading Ages

The solidification theory-based B3 model is well-known for characterizing the creep

performance of concrete [6]. The solidification theory combining elastic theory,
viscoelastic theory, and rheological theory could be utilized to simulate the time-
dependent macroscopic physical and mechanical properties of concrete. According
to this theory, the aging aspect is because of the increased volume fraction of the
load-carrying part of solidified substances, the properties of which are independent

@seismicisolation
@seismicisolation
208 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

of age [47]. The B3 model satisfies the condition that the creep behavior at different
loading ages does not diverge during loading. Therefore, the tensile creep of ICC with
SAPs at different loading ages could be described by the B3 model, the mathematical
expression of which could be expressed as Eq. (4.12).
According to the Eq. (4.11), the tensile creep compliance function of ICC with
SAPs at 7 d after loading was 752, 362, 247, 181, 90, 75, and 62 με/MPa when the
loading age was 0.50, 0.67, 0.83, 1, 3, 5, and 7 d, respectively.
The creep magnitude of concrete loaded at early ages increases significantly, espe-
cially for concrete loaded earlier than 3 d [72]. A study done by Østergaard indicates
that the great creep behavior
   at a very early age can be captured by redefining the

parameter q2 (q2 = q2 t  −qt
6
) in the B3 Model [47]. The long-term creep is prin-
cipally the result of the relaxation of micro-prestress in the micro-pores of cement-
based materials, resulting in the concrete flowing viscously [73]. For concrete under
constant tensile loading, the relaxation of micro-prestress will increase with the
generation of new hydrates. At the macroscopic scale, such locomotion that occurs
inside the micro-pores of concrete will lead to the flow creep strain. A study done
by Wei suggests a modified micro-prestress solidification model that could seize the
long-term large creep induced as a result of the flow properties of concrete through
modifying the parameter q4 (q4 =q4 q7 ) in the B3 Model [74]. Furthermore, the param-
eter q5 was introduced to consider the effect of SAPs on the aging viscoelastic compli-
ance. Thus, the tensile creep compliance function J (t, t  ) of ICC with SAPs loaded
at different ages was modified following the Østergaard model and the Wei model,
as shown in Eq. (4.20).
  
 t 
  n
 t
J (t, t ) = q1 + q2 q5 Q(t, t ) + q3 ln 1 + (t − t ) + q4 q7 ln 
t  − q6 t
(4.20)

where q5 is the influence factor of SAPs; q6 is regarded as the time for the internal
structure of concrete to transform from a fluid viscoelastic state to a solid viscoelastic
state (i.e., structural setting time), and the value of q6 should be lower than the earliest
(valid) physical test time; and q7 is assumed to be more than 1 to explain the long-term
flow creep.
The calculation results of q1 , q2 , q3 , and q4 were 13.61, 147.25, 0.35, and 7.99,
respectively. Afterthe values
 of empirical material constitutive coefficients q1 to q4
t
were obtained, q5 t  −q6 and q7 were fitted in accordance with the tensile creep
compliance function of ICC with SAPsat different loading ages. Based on the
t
regression analysis, the values of q5 t  −q 6
and q7 at different loading ages could be
obtained, as illustrated in Table 4.6.   
t
The coefficients q5 and q6 could be obtained by regression analysis of q5 t  −q 6
at different loading ages, as illustrated in Fig. 4.22.
The influence coefficient of SAPs was calculated as 0.381. The best fit value of the
structural setting time was 0.444 d which was less than the valid physical test time

@seismicisolation
@seismicisolation
4.4 Tensile Creep of Early-Age Internally Cured Concrete with Different … 209

Table 4.6 Parameters for

Loading age (d) q5 q6 q7
creep compliance
0.50 0.381 0.444 11.67
0.67 9.46
0.83 9.44
1 7.63
3 4.85
5 5.39
7 5.76

Fig. 4.22 Fitting of

influence coefficient of SAPs
and structural setting time

of 0.50 d. The fitting results of q7 decreased nonlinearly when the loading ages were
within the proportionality limit of 1.60 d and then increased linearly. Comparisons
on tensile creep compliance function obtained from test results for specimens at
different loading ages and Eq. (4.20) are illustrated in Fig. 4.23.
The analytical predictions of specimens at different loading ages were well in
line with the test results. The proposed model could be used for the prediction of the
tensile creep of HSC considering loading ages and the effect of IC with SAPs.

4.4.8 Summary

The effect of loading age on the early-age basic tensile creep behavior of ICC with
SAPs was investigated. Analysis on the basic tensile creep, basic tensile creep coeffi-
cient, specific basic tensile creep, and specific basic tensile creep rate was conducted
by utilizing TSTM. The following conclusions can be drawn.

@seismicisolation
@seismicisolation
210 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

Fig. 4.23 Comparisons on

tensile creep compliance
function obtained from test
results and analytical
predictions for specimens at
different loading ages

(1) The basic tensile creep of ICC with SAPs at 7 d after loading was 116, 106, 97,
81, 55, 47, and 36 με when the loading age was 0.50, 0.67, 0.83, 1, 3, 5, and 7
d, respectively.
(2) The basic tensile creep coefficient of ICC with SAPs at 7 d after loading was
4.45, 4.27, 4.05, 3.46, 2.39, 2.04, and 1.55 when the loading age was 0.50, 0.67,
0.83, 1, 3, 5, and 7 d, respectively.
(3) The specific basic tensile creep of ICC with SAPs at 7 d after loading was 611,
286, 198, 140, 63, 51, and 38 με/MPa when the loading age was 0.50, 0.67,
0.83, 1, 3, 5, and 7 d, respectively.
(4) The specific basic tensile creep rate was relatively large at an early loading age,
then decreased and tended to be stable. Furthermore, the earlier the loading age
was, the greater the variation between the initial creep rate and the later creep
rate at the same loading age was.
(5) The relative specific basic tensile creep decreased exponentially when the
concrete was loaded within 1 d and decreased linearly when the concrete was
loaded after 3 d. The aging effect was much significant when the loading age
was at 0.5, 0.67, 0.83, and 1 d. The proportionality limit of specific basic tensile
creep of ICC with SAPs, determined by linear and nonlinear regression analysis,
occurred at the loading age of 1.60 d.
(6) A model was proposed by introducing the influence coefficient of SAPs to the
Wei model and the Østergaard model and was well in line with the test results.
The proposed model could be applied for the prediction of the tensile creep of
ICC with SAPs in consideration of loading ages.

@seismicisolation
@seismicisolation
References 211

References

1. Alexander S (2006) Creep, shrinkage and cracking of restrained concrete at early age. Concrete
40(2):38–39
2. Nguyen DH, Dao V, Lura P (2017) Tensile properties of concrete at very early ages. Constr
Build Mater 134:563–573
3. Briffaut M, Benboudjema F, Torrenti JM et al (2012) Concrete early age basic creep:
experiments and test of rheological modelling approaches. Constr Build Mater 36:373–380
4. Zhu H, Li Q, Hu Y et al (2018) Double feedback control method for determining early-age
restrained creep of concrete using a temperature stress testing machine. Materials 11:1079
5. Bentz DP, Snyder KA (1999) Protected paste volume in concrete: extension to internal curing
using saturated lightweight fine aggregate. Cem Concr Res 29(11):1863–1867
6. Bažant ZP, Baweja S (1995) Creep and shrinkage prediction model for analysis and design of
concrete structures: model B3. Mater Struct 28(6):357–365
7. Xu Y, Liu J, Liu J et al (2019) Creep at early ages of ultrahigh-strength concrete: experiment
and modelling. Mag Concr Res 71(16):847–859
8. Gu CP, Wang YC, Gao F et al (2019) Early age tensile creep of high performance concrete
containing mineral admixtures: experiments and modeling. Constr Build Mater 197:766–777
9. Aly T, Sanjayan JG (2008) Shrinkage cracking properties of slag concretes with one-day curing.
Mag Concr Res 60(1):41–48
10. Kovler K (1994) Testing system for determining the mechanical behaviour of early age concrete
under restrained and free uniaxial shrinkage. Mater Struct 27(6):324–330
11. Shen DJ, Liu KQ, Ji Y et al (2018) Early-age residual stress and stress relaxation of high-
performance concrete containing fly ash. Mag Concr Res 70(13–14):726–738
12. Pigeon BM (1995) Tensile creep at early ages of ordinary, silica fume and fiber reinforced
concretes. Cem Concr Res 25(5):1075–1085
13. Garas VY, Kahn LF, Kurtis KE (2009) Short-term tensile creep and shrinkage of ultra-high
performance concrete. Cem Concr Compos 31(3):147–152
14. Ni TY, Yang Y, Gu CP et al (2019) Early-age tensile basic creep behavioral characteristics of
high-strength concrete containing admixtures. Adv Civ Eng 2019(7):1–11
15. Jiang JL (2017) Experimental study on tensile creep of concrete internally cured with super
absorbent polymers at early age. Hohai University, Nanjing
16. Liu KQ (2017) Experimental study on tensile creep of concrete internally cured with super
absorbent polymers at early age. Hohai University, Nanjing
17. Shen DJ, Liu C, Jiang JL et al (2020) Influence of super absorbent polymers on early-age
behavior and tensile creep of internal curing high strength concrete. Constr Build Mater
258:120068
18. Shen DJ, Li CC, Liu C et al (2020) Experimental investigations on early-age tensile creep
of internally cured high strength concrete under different initial stress/strength ratios. Constr
Build Mater 265(5):120312
19. Shen DJ, Li CC, Kang JC et al (2022) Influence of loading ages on the early age tensile
creep of high-strength concrete modified with superabsorbent polymers. J Mater Civ Eng
34(5):04022064
20. Kolver K, Igarashi S, Bentur A (1999) Tensile creep behavior of high strength concretes at
early ages. Mater Struct 32(5):383–387
21. Shen DJ, Jiang JL, Shen JX et al (2015) Influence of prewetted lightweight aggregates on the
behavior and cracking potential of internally cured concrete at an early age. Constr Build Mater
99:260–271
22. Shen DJ, Liu XZ, Li Q et al (2019) Early-age behavior and cracking resistance of high-strength
concrete reinforced with Dramix 3D steel fiber. Constr Build Mater 196:307–316
23. Shen DJ, Jiao Y, Gao Y et al (2020) Influence of ground granulated blast furnace slag on
cracking potential of high performance concrete at early age. Constr Build Mater 241:117839

@seismicisolation
@seismicisolation
212 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

24. Ministry of Housing and Urban Rural Development of the People’s Republic of China (2019)
Standard for test methods of concrete physical and mechanical properties: GB/T 50081-2019.
China Architecture & Building Press, Beijing. (in Chinese)
25. Kanstad T, Hammer TA, Bjontegaard O et al (2003) Mechanical properties of young concrete:
Part I: experimental results related to test methods and temperature effects. Mater Struct
36(4):218–225
26. Yoo DY, Park JJ, Kim SW et al (2014) Influence of ring size on the restrained shrinkage behavior
of ultra high performance fiber reinforced concrete. Mater Struct 47(7):1161–1174
27. Shen DJ, Liu XZ, Zeng X et al (2020) Effect of polypropylene plastic fibers length on cracking
resistance of high performance concrete at early age. Constr Build Mater 244:117874
28. Yang Y, Xu S, Ye D et al (2009) Tensile creep behavior of high strength concrete at early ages.
J Chin Ceramic Soc 37(7):1124–1129 (in Chinese)
29. Zhang T, Qin WZ (2006) Tensile creep due to restraining stresses in high-strength concrete at
early ages. Cem Concr Res 36(3):584–591
30. Shen DJ, Jiang JL, Zhang MY et al (2018) Tensile creep and cracking potential of high perfor-
mance concrete internally cured with super absorbent polymers at early age. Constr Build
Mater 165:451–461
31. Khan I, Xu T, Castel A et al (2018) Early-age tensile creep and shrinkage induce cracking in
internally restrained concrete members. Mag Concr Res 71(22):1–41
32. General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China (2018) Common portland cement: GB 175-2007/XG1-2018. Standard press
of China, Beijing. (in Chinese)
33. Beushausen H, Gillmer M (2014) The use of superabsorbent polymers to reduce cracking of
bonded mortar overlays. Cem Concr Compos 52:1–8
34. Li SC, Zhang YS, Zhang GR (2013) Study on tensile creep characteristics of high performance
concrete. J Civ Archit Environ Eng 35(2):40–44 (in Chinese)
35. Kang SH, Hong SG, Moon J (2018) The effect of superabsorbent polymer on various scale of
pore structure in ultra-high performance concrete. Constr Build Mater 172:29–40
36. Ma XW, Zhang JK, Liu JH (2015) Review on superabsorbent polymer as internal curing agent
of high performance cement-based material. J Chin Ceramic Soc 43(8):1099–1110 (in Chinese)
37. Kong XM, Zhang ZL, Lu ZC (2015) Effect of pre-soaked superabsorbent polymer on shrinkage
of high-strength concrete. Mater Struct 48(9):2741–2758
38. Olawuyi BJ, Boshoff WP (2017) Influence of SAP content and curing age on air void distribution
of high performance concrete using 3D volume analysis. Constr Build Mater 135:580–589
39. Maou T (2013) Comparison of concrete creep in tension and in compression: influence of
concrete age at loading and drying conditions. Cem Concr Res 51:78–84
40. Esteves LP, Lukošiūtė I, Čėsnienė J (2014) Hydration of cement with superabsorbent polymers.
J Therm Anal Calorim 118(2):1385–1393
41. Wang LA, Guo L, Guo LX (2014) Sensitivity analysis of impact factors of concrete tensile
creep at early ages. J Build Mater 17(5):896–900
42. Sellevold EJ, Bjntegaard Ø (2006) Coefficient of thermal expansion of cement paste and
concrete: mechanisms of moisture interaction. Mater Struct 39(9):809–815
43. Sant G (2012) The influence of temperature on autogenous volume changes in cementitious
materials containing shrinkage reducing admixtures. Cem Concr Compos 34(7):855–865
44. Reinhardt HW, Rinder T (2006) Tensile creep of high-strength concrete. J Adv Concr Technol
4(2):277–283
45. Forth JP (2015) Predicting the tensile creep of concrete. Cem Concr Compos 55:70–80
46. Bissonnette B, Pigeon M, Vaysburd AM (2007) Tensile creep of concrete: study of its sensitivity
to basic parameters. ACI Mater J 104(4):360–368
47. Østergaard L, Lange DA, Altoubat SA et al (2001) Tensile basic creep of early-age concrete
under constant load. Cem Concr Res 31(12):1895–1899
48. Zhao ZE, Wang KJ, Lange DA et al (2019) Creep and thermal cracking of ultra-high volume
fly ash mass concrete at early age. Cem Concr Compos 99:191–202

@seismicisolation
@seismicisolation
References 213

49. Snoeck D, Schaubroeck D, Dubruel P et al (2014) Effect of high amounts of superabsorbent

polymers and additional water on the workability, microstructure and strength of mortars with
a water-to-cement ratio of 0.50. Constr Build Mater 72:148–157
50. Klemm AJ, Sikora KS (2013) The effect of superabsorbent polymers (SAP) on microstructure
and mechanical properties of fly ash cementitious mortars. Constr Build Mater 49:134–143
51. Luczynski KW, Brynk T, Ostrowska B et al (2013) Consistent quasistatic and acoustic elasticity
determination of poly-l-lactide-based rapid-prototyped tissue engineering scaffolds. J Biomed
Mater Res Part A 101:138–144
52. Luczynski KW, Steiger-Thirsfeld A, Bernardi J et al (2015) Extracellular bone matrix exhibits
hardening elastoplasticity and more than double cortical strength: Evidence from homogeneous
compression of non-tapered single micron-sized pillars welded to a rigid substrate. J Mech
Behav Biomed Mater 52:51–62
53. Karte P, Hlobil M, Reihsner R et al (2015) Unloading-based stiffness characterisation of cement
pastes during the second, third and fourth day after production. Strain 51(2):156–169
54. Abdel-Wahab AA, Alam K, Silberschmidt VV (2011) Analysis of anisotropic viscoelastoplastic
properties of cortical bone tissues. J Mech Behav Biomed Mater 4(5):807–820
55. Yang CY (2015) Investigation on tensile creep of high strength/high performance concrete
containing blast furnace slag at early ages. Zhejiang University of Technology, Hangzhou
56. Assmann A, Reinhardt HW (2014) Tensile creep and shrinkage of SAP modified concrete.
Cem Concr Res 58:179–185
57. Kamen A, Denarié E, Sadouki H et al (2008) UHPFRC tensile creep at early age. Mater Struct
42(1):113–122
58. Altoubat SA, Lange DA (2001) Tensile basic creep: measurements and behavior at early age.
ACI Mater J 98(5):386–393
59. Ranaivomanana N, Multon S, Turatsinze A (2013) Tensile, compressive and flexural basic
creep of concrete at different stress levels. Cem Concr Res 52:1–10
60. Atrushi DS (2003) Tensile and compressive creep of young concrete: testing and modelling.
The Norwegian University of Science and Technology
61. Neville AM, Brooks JJ (1987) Concrete technology. Longman Scientific & Technical, England
62. Han B, Xie HB, Zhu L et al (2017) Nonlinear model for early age creep of concrete under
compression strains. Constr Build Mater 147:203–211
63. Hsu T, Slate FO, Sturman GM et al (1963) Microcracking of plain concrete and the shape of
the stress-strain curve. J ACI 60(2):09–224
64. Ngab AS, Slate FO, Nilson AH (1981) Microcracking and time-dependent strains in high
strength concrete. ACI Struct J 78(4):262–268
65. Bažant ZP, Hubler MH, Wendner R et al (2015) Model B4 for creep, drying shrinkage and
autogenous shrinkage of normal and high-strength concretes with multi-decade applicability.
TC-242-MDC multidecade creep and shrinkage of concrete: material model and structural
analysis. Mater Struct 48:753–770
66. Hilaire A, Benboudjema F, Darquennes A et al (2014) Modeling basic creep in concrete at
early-age under compressive and tensile loading. Nucl Eng Des 269:222–230
67. Zheng XY, Ji T, Easa SM et al (2019) Tensile basic creep behavior of lightweight aggregate
concrete reinforced with steel fiber. Constr Build Mater 200:356–367
68. Delsaute B, Boulay C, Staquet S (2016) Creep testing of concrete since setting time by means
of permanent and repeated minute-long loadings. Cem Concr Compos 73:75–88
69. Moradian M, Ley MT, Grasley ZC (2018) Stress induced dissolution and time-dependent
deformation of portland cement paste. Mater Des 157:314–325
70. Li XD, Grasley ZC, Bullard JW et al (2018) Creep and relaxation of cement paste caused by
stress-induced dissolution of hydrated solid components. J Am Ceram Soc 101(9):4237–4255
71. Nestle N, Kühn A, Friedemann K et al (2009) Water balance and pore structure development
in cementitious materials in internal curing with modified superabsorbent polymer studied by
NMR. Microporous Mesoporous Mater 125(1–2):51–57
72. Wei Y, Liang SM, Guo WQ et al (2017) Stress prediction in very early-age concrete subject to
restraint under varying temperature histories. Cem Concr Compos 83:45–56

@seismicisolation
@seismicisolation
214 4 Tensile Creep of Early-Age Internally Cured Concrete with SAPs

73. Bažant ZP, Hauggaard A, Baweja S et al (1997) Microprestress solidification theory for concrete
creep. I: aging and drying effects. J Eng Mech 123(11):1188–1194
74. Wei Y, Hansen W (2013) Tensile creep behavior of concrete subject to constant restraint at
very early ages. J Mater Civ Eng:1277–1284

@seismicisolation
@seismicisolation
Chapter 5
Cracking Resistance of Internally Cured
Concrete by Restrained Ring Test
at Early-Age

Contents

5.1 Experimental Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216

5.1.1 Free Shrinkage Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
5.1.2 Mechanical Properties Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
5.1.3 Restrained Ring Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
5.2 Cacking Resistance of Early-Age Concrete Internally Cured with SAPs . . . . . . . . . . . . . 223
5.2.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
5.2.2 Strain in Restrained Steel Ring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
5.2.3 Residual Stress of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
5.2.4 Stress Rate of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
5.2.5 Stress Relaxation of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
5.2.6 Time to Cracking of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
5.2.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
5.3 Cracking Resistance of Early-Age Concrete Internally Cured with LWAs . . . . . . . . . . . . 234
5.3.1 Mixture Proportion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 234
5.3.2 Strain in Restrained Steel Ring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
5.3.3 Residual Stress of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
5.3.4 Stress Rate of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
5.3.5 Stress Relaxation of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
5.3.6 Time to Cracking of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241
5.3.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally Cured Concrete
Using Ring Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
5.4.1 Addition of Ground Granulated Blast Furnace Slag . . . . . . . . . . . . . . . . . . . . . . . . 244
5.4.2 Addition of Polypropylene Fibers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
5.4.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263

IC is always employed in HSC or HPC that exhibits a low w/c ratio [1, 2]. With
the decrease of w/c ratio, the drawback of early-age cracking is gradually exposed,
which affects the performance of concrete [3, 4]. The early-age cracking may be
observed resulting from the restraint on volume changes [5]. When shrinkage occurs
in concrete under the restrained condition, the related residual stress will increase,
and the cracking may take place at the time when the residual stress exceeds the

© Science Press 2023 215

D. Shen, Cracking Control on Early-Age Concrete Through Internal Curing,
https://doi.org/10.1007/978-981-19-8398-6_5
@seismicisolation
@seismicisolation
216 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

tensile strength of concrete [6–8]. The estimation of residual stress is significant to

accurately evaluate the cracking resistance of concrete at early age [9]. Restrained
ring test is used to investigate the cracking resistance of cementitious materials [10].
The effect of shrinkage, tensile creep, and continuous restraint can be considered in
restrained ring test, which vividly simulates the side wall of basement in practical
engineering [11]. Since firstly proposed by Carlson in 1942 [12], the restrained ring
test has been continuously developed and improved [13]. Chinese Standard CCES01
[14] as well as AASHTO [15] and ASTM C1581 [16] have been provided to guide
the use of this test method. The restrained ring test has been extensively utilized in
evaluating the cracking resistance of concrete [17]. The cracking resistance under
continuous restraint is significant for performance of ICC at early age. Therefore,
cracking resistance of ICC under continuous restraint needs to be further studied.
In this chapter, the methodologies of restrained ring test combined with free
shrinkage are described in details. Theoretical calculation method for quantifying
the stress relaxation is introduced. The cracking resistance of ICC with SAPs and
LWAs are presented by restrained ring test, including the free shrinkage, residual
stress, stress rate, stress relaxation, time to cracking and cracking potential [18–20],
respectively. Moreover, early-age cracking resistance of ICC with GGBFS and PP
fibers is also included for better understanding [21–24], respectively.

5.1 Experimental Program

5.1.1 Free Shrinkage Measurement

1. Direct method
The free shrinkage referred to the superposition of autogenous and drying shrinkage.
The free shrinkage of concrete was measured with 100 mm × 100 mm × 515 mm
prism specimen [25].
The specimens for measuring free shrinkage were moved into a constant temper-
ature and humidity environment with a temperature of 20 ± 2 °C and a relative
humidity of 60% ± 5% after 1.5 d of curing in a standard curing chamber which
was also utilized for mechanical properties test. In order to make the test accurate,
it was necessary to keep the same ratio of the volume of concrete specimen to the
surface area for dispersing water between concrete ring and prism specimen. The
free shrinkage in concrete could be calculated through two dial gauges utilized to
record in free shrinkage measurement. The average value of free shrinkage measured
in three prism specimens was considered as the final result. Figure 5.1 illustrates the
schematic description of the machine for measuring the free shrinkage.
2. Indirect method
Autogenous shrinkage measurement was conducted with non-contact sensors. This
measurement technique takes into account the benefits of using the eddy-current

@seismicisolation
@seismicisolation
5.1 Experimental Program 217

Dial gauge Sealed 1 Sealed

Metal nail

100 mm
Concrete
Magnetic support Fixed steel plate 1 Floor 100 mm
81 mm 353 mm 81 mm
515 mm
1-1

Fig. 5.1 Schematic description of shrinkage measurement machine

displacement sensor for measuring autogenous shrinkage, as proposed by [26, 27].

A thermocouple was embedded at the center of concrete specimen during the casting
period to monitor the change in temperature within the concrete, so that autogenous
shrinkage was separated from the measured total deformation. The specimens were
completely covered in polythene sheet in order to ensure that no transport of moisture
occurs between specimen and atmosphere. Free shrinkage measurements at the age
of t from the concrete was cast, ε(t)sh , are calculated by using Eq. (5.1), which is
proposed in [28].
√
εsh (tnet ) = C1 · (t)C2 + C3 · t − t0 (5.1)

where C1 , C2 and C3 represents material constants, respectively; t is the age of the

specimen, in d; t0 represents the age of the specimen at initial setting (time-zero), in
d; and td is the time after initiation of drying (i.e., 2 d). The form of Eq. (5.1) is based
on the approximation that total shrinkage is the sum of an autogenous component
(first term) and a drying component (second term), though it should be noted that this
is not absolutely true. To obtain the material constants C1 and C1 (which corresponds
to autogenous shrinkage), the shrinkage measured in sealed specimens, are used to
fit results. C3 was determined through the strain of autogenous shrinkage specimens
and the ratio of autogenous to drying shrinkage is 2, which is proposed in [29].

5.1.2 Mechanical Properties Test

Mechanical properties tests were simultaneously conducted on the specimens made

of the four kinds of concrete mixtures at the age of 28 d after casting in terms of
Chinese Standard GB/T 50081-2019, as reported in [30]. For obtaining the elastic
modulus, three 150 mm × 150 mm × 300 mm prisms in each group would be made
for the results when the age was 28 d after casting. Meanwhile, for obtaining the
splitting tensile strength of concrete mixtures, three 150 mm cubes in each group
would be made for the results when the age was 28 d after casting. The result of the
elastic modulus and splitting tensile strength of concrete mixture was the average

@seismicisolation
@seismicisolation
218 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

value of the measurements on three identical specimens in each group, which was in
terms to the Chinese Standard GB/T 50081-2019 [30].

5.1.3 Restrained Ring Test

1. Restrained ring specimen and procedure

The specimen geometric sizes and notations are shown in Fig. 5.2 The specimens
are shown in Fig. 5.3
In steel ring, the inner and outer radius were expressed as R I S and R O S , respec-
tively. In concrete ring, the outer and inner radius were expressed as R OC and R I C ,
respectively. The inner edge of concrete ring coincided with the outer edge of steel
ring. The steel ring thickness was the difference between outer radius and inner

Strain
gauge t Sealed
Drying
Steel

150 mm
ring

2RIS
Strain
gauge 2ROS or 2RIC
Concrete 2ROC
ring

(a) Top view (b) Cross view

Fig. 5.2 Geometry of the ring specimen

(a) test set-up (b) placing concrete

Fig. 5.3 Restrained ring test specimens

@seismicisolation
@seismicisolation
5.1 Experimental Program 219

radius of steel ring, expressed as t. The dimensions were as follows: ROC = 203 mm,
R O S = RIC = 162.5 mm, R I S = 148.5 mm, t = 14 mm. The height of the spec-
imen placed horizontally as 150 mm tall. The strain gauges in the circumferen-
tial and vertical direction were posted at each sampling point by half-bridge self-
compensation configuration. All apparatuses utilized in the restrained ring test were
in terms to ASTM C1581, as reported in [16]. The cracking in specimens occurred
when the strain of restraint steel ring had a sudden change (the sudden change value
exceeded 30 με). The values of strain collected were revised for eliminating the effect
of temperature change and considered as the actual circumferential strain of steel
ring. In each group, two restrained ring specimens which had the same proportion
were cast for test to obtain the average value of the strain of steel ring.
The circumference of concrete ring was exposed to the atmosphere to dry, and
the top and bottom surfaces of the concrete ring were sealed with an aluminum
adhesive tape immediately after demolding to prevent water evaporation. It could be
supposed that the loss of moisture along the height of the restrained ring specimen
was uniform. The experiment was carried out at the relative humidity of 50 ± 4%
and the temperature of 23 ± 2 °C [16]. Data would be collected every 15 min.
Although the restrained conditions of prismatic specimen and ring specimen are
surely different, the free shrinkage of prismatic specimens and ring specimens can
be assumed to be equal when their volume-to-surface ratios are the same (Vr /Sr =
[π(2032 − 162.52 ) × 150]/(2π × 203 × 150) = 36.46 mm) [31]. To obtain the same
volume-to-surface area ratio as the ring specimen, 81 mm lengths from the ends
of the prismatic specimen were sealed with vinyl after demolding. These prismatic
specimens were placed in the same environment as that of the ring specimens.

2. Residual stress of restrained ring specimen

With the restraint imposed by steel ring on shrinkage, tensile stress in concrete ring
begins to develop correspondingly. Measuring the strain obtained in inner steel ring
can indirectly calculate the actual circumferential tensile stress in concrete on the
basis of assuming that concrete is the elastic body.
In order to satisfy the displacement compatibility between steel ring and concrete
ring, the corresponding stress in concrete ring is generated, which is defined as the
actual residual interfacial stress (Pr es. ) [28], as expressed in Eq. (5.2).

(R 2O S − R 2I S )
Pr es. = −εst · E s · (5.2)
2R 2O S

where ROS and R I S represent geometric dimensions of steel ring, in mm; εst represents
the strain outputted by static strain gauge, in με, which can be equivalent to the
superposition of free shrinkage stress, tensile creep strain and elastic strain in concrete
ring; and E s represents the elastic modulus of steel, in MPa.
The strain obtained by the strain gauges in steel ring can be utilized for calcu-
lating the actual residual interface pressure. In addition, steel properties and geomet-
rical dimensions will also be applied to express the residual stress. Equation (5.3)

@seismicisolation
@seismicisolation
220 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

gives an expression of the circumferential tensile stress (σr es. (r )) which can be
considered as the value of strain varying with radius [28].
( )
R 2O S R 2OC
σr es. (r ) = Pr es. · 2 · 1+ 2 (5.3)
R OC − R 2O S r

where r in the range from R I C to R OC (i.e., geometric dimensions of inner and outer
edges of concrete ring) represents the random value at the radius direction, in mm.
According to Eq. (5.3), the maximum circumferential tensile stress appears at the
junction of steel ring and concrete ring. Equation (5.2) is introduced into Eq. (5.3)
to get Eq. (5.4) in order to obtain the maximum residual tensile stress (σr es.max ) at
the inner boundary of concrete ring [28].

R 2O S − R 2I S R 2O S + R 2OC
σr es.max = −εst · E s · · 2 (5.4)
2R 2O S R OC − R 2O S

3. Stress relaxation of restrained ring specimen

Stress relaxation of concrete has a close relationship with creep. PP fibers will
improve the resistance to creep in concrete [32, 33]. In this section, the effect of
stress relaxation was mainly expressed by comparing theoretical elastic stress with
residual stress. Since maximum residual stress had been emphatically studied above,
the following main study was how to obtain the theoretical elastic stress of concrete
at early age.
Previous studies were based on the high stiffness of steel rings, as reported in
[34]. Thus, the outer radius of the steel would not generate any deformation under
the pressuring force in order to make calculations simpler. However, the stiffness of
steel was much lower and steel ring would deform due to load. This section narrated
a new method “shrinkage-fit” expected to calculate elastic stress on the condition
that steel ring would be deformed under the loading of concrete, as reported in [28].
Initially, no gap was between steel ring and concrete ring (i.e., R O S = R I C ). In order
to solve the theoretical elastic stress resulting from shrinkage, the restraint imposed
by steel ring should be temporarily eliminated so as to make concrete shrink freely.
The shrinkage in concrete ring shorten the inner radius of concrete and the change
of radial deformation (ΔUsh ) is directly calculated in Eq. (5.5) [28].

ΔUsh = R I C · Δεsh (5.5)

where Δεsh represents the incremental free shrinkage, in με.

For solving the problem that the radius of concrete after shrinkage mismatches
geometric dimensions of the original steel, it can be considered that the pressurizing
stress (ΔPel. ) is applied on the interface where maximum residual stress occurs
[34]. This pressure acts on compressing steel ring and expanding concrete ring
until displacement compatibility is met. Equation (5.6) gives the expression of the

@seismicisolation
@seismicisolation
5.1 Experimental Program 221

displacement (Ust ) resulting from the pressurizing stress applied on steel ring. Mean-
while, Eq. (5.7) gives the expression of the displacement (Uc ) of concrete under the
equal pressurizing stress [28].
[ ]
R 2O S (1 + νs )R 2I S + (1 − νs )R 2O S
Ust = −ΔPel. · ( ) (5.6)
E s R O S R 2O S − R 2I S
[ ]
ΔPel. R 2I C (1 + νc )R 2OC + (1 − νc )R 2I C
Uc = · ( ) (5.7)
Ec R I C R 2OC − R 2I C

where νs represents the Poisson’s ratio of steel; νc represents the Poisson’s ratio of
concrete; E c represents the time-dependent elastic modulus of concrete, in GPa; and
ΔPel. represents the incremental pressurizing stress, in MPa. The Poisson’s ratio
was 0.27 and 0.20 for steel and concrete, respectively. For meeting the displacement
compatibility, Eq. (5.8) expresses the relationship between the radial displacement
(ΔUsh ) and the displacements in steel ring (Ust ) and concrete ring (Uc ) [28].

Ust − Uc = ΔUsh (5.8)

Considering that R O S and R I C are equal, the incremental pressurizing stress is

derived by Eq. (5.9) [28].

Δεsh E c
ΔPel. = − (5.9)
E c [(1+νs )R I S +(1−νs )R O S ] (1−ν )R 2 +(1+ν )R 2
+ [ c R 2O S −R 2 c OC ]
2 2

Es ( R 2O S −R 2I S ) ( OC O S )

Equation (5.10) describes the increment of circumferential tensile stress

(Δσel. (r )), which changes with the distance to the geometric center and is calculated
by utilizing the elastic theory similar to acquiring residual stress [28].
( )
R 2O S R 2OC
Δσel. (r ) = ΔPel. · · 1 + (5.10)
R 2OC − R 2O S r2

According to Eq. (5.10), the maximum circumferential tensile stress will take
place at the junction of steel ring and concrete ring. This allows Eq. (5.10) to calculate
the maximum theoretical elastic stress based on the incremental pressurizing stress
derived from Eq. (5.9) and replacing r with R O S or R I C applied in the actual geometry
size of specimens, thereby obtaining Eq. (5.11) [28].

4.57Δεsh E c
Δσel. max = − (5.11)
10.88 EEcs + 4.77

This makes the calculation of theoretical elastic stress be the sum of incre-
mental shrinkage stress utilizing elastic modulus of concrete and steel as well as

@seismicisolation
@seismicisolation
222 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

the geometric dimensions of specimens. In order to express and quantify the incre-
ment of stress relaxation (Δσr el ), it is an effectual method to compare the increment of
theoretical elastic stress (Δσel. max )with the increment of residual stress (Δσr es. max ),
as expressed in Eq. (5.12) [28].

Δσr el = Δσel. max − Δσr es. max (5.12)

Substituting maximum residual stress at net age and Eq. (5.11) into Eq. (5.12)
can calculate the increment of relaxed stress, as expressed by Eq. (5.13) [28].

4.57Δεsh E c R 2O S − R 2I S R 2O S + R 2OC
Δσr el = − + Δεst · E S · · 2 (5.13)
10.88 EEcs + 4.77 2R 2O S R OC − R 2O S

The ratio of relaxed stress to theoretical elastic stress, RA, is utilized to assess the
degree of stress relaxation of specimens, as expressed in Eq. (5.14). The higher the
ratio RA is, the higher the level of stress relaxation will be [28].
σr el
RA = (5.14)
σel. max

4. Stress rate of restrained ring specimen

Stress rate related to the humidity gradient of concrete is one of the possible reasons
for the change of the cracking potential at early age. At present, the calculation of
concrete stress rate has become one of the methods of ASTM to assess the cracking
potential [16]. Generally, the cracking potential and stress rate are closely related. The
increase of the stress rate increases the cracking potential. The method of calculating
stress rate is introduced in [16], as expressed in Eqs. (5.15) and (5.16).

E s R I C (R O S − R I S )|α|
S(te ) = √ (5.15)
2R I S (R OC − R I C ) te
√
εnet (te ) = α te + b (5.16)

where te represents the net age, in d; εnet (te ) represents the difference between the
strain measured in steel ring and that before the beginning of drying, in με; and α
√ as b represent the constants to express the linear function containing εnet
as well
and te . According to the stress rate classification method, stress rate and net age at
cracking were the factors to evaluate the cracking potential, as shown in Table 5.1
[16].

5. Cracking potential of restrained ring specimen

Comparing the time-dependent splitting tensile strength with residual stress is an

effectual method to estimate the cracking potential. That is the ratio of residual
stress to splitting tensile strength, as expressed in Eq. (5.17) [28].

@seismicisolation
@seismicisolation
5.2 Cacking Resistance of Early-Age Concrete Internally Cured with SAPs 223

Table 5.1 Classification of

Net time to Stress rate, S Cracking potential
cracking potential
cracking, te (d) (MPa/d)
0 < te ≤ 7 S ≥ 0.34 High
7 < te ≤ 14 0.17 ≤ S < 0.34 Moderated-high
14 < te ≤ 28 0.10 ≤ S < 0.17 Moderated-low
te > 28 S < 0.10 Low

σr es. max (t)

φC R (t) = (5.17)
f sp (t)

where φC R (t) represents the cracking potential at age of t d; σr es. max (t) represents the
maximum residual stress in concrete at the age of t d, in MPa; and f sp (t) represents
the splitting tensile strength at the age of t d, in MPa.

5.2 Cacking Resistance of Early-Age Concrete Internally

Cured with SAPs

5.2.1 Mixture Proportion

1. Materials
Portland Cement (Cement II 52.5R) in accordance with Chinese Standard GB 175-
2007 was used with a Blaine fineness of 375 m2 /kg [35]. The initial setting time of
cement is 167 min while the final is 221 min, and its compressive strength of 3 d is
35.5 MPa while that of 28 d is 66.9 MPa. Normal weight river sand was used with
a fineness modulus of 2.05. Crushed limestone (maximum size of 16 mm) was used
as coarse aggregates.
A suspension-polymerized, covalently cross-linked acrylamide/acrylic acid
copolymer as SAPs with a dry-bulk density of 850 kg/m3 were used. The spherical
particles have diameters varying from 125 to 150 μm in the dry state. The absorption
capacity of SAPs was tested. A liquid polycarboxylate-based superplasticizer with
the type of PCA- I was utilized to adjust the workability of different mixtures. Tap
water was used as mixture and IC water.
2. Mixture proportions
Four low w/c concrete mixtures were used. Mixture proportions designated as SAP0,
SAP05, SAP15, and SAP25 are shown in Table 5.2 [18]. Mixture SAP0 was one
reference concrete with no IC, while Mixture SAP05, SAP15, and SAP25 used IC
with SAPs. The concrete compositions were varied by the addition of SAP (0.05%,
0.16%, and 0.26% by weight of cement for Mixtures SAP05, SAP15, and SAP25,
respectively). All mixtures had a basic w/c ratio of 0.33 [18].

@seismicisolation
@seismicisolation
224 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Table 5.2 Mix proportions of concrete

Mixture composition Concrete mixtures
SAP0 SAP05 SAP15 SAP25
Water (kg/m3 ) 171 171 171 171
Cement (kg/m3 ) 512 512 512 512
Fine aggregate(sand) (kg/m3 ) 636 636 636 636
Coarse aggregate (kg/m3 ) 1131 1131 1131 1131
SAP (g/m3 ) 0 267.5 803.0 1339.0
IC water (kg/m3 ) 0 5.35 16.06 26.78
Superplasticizer (kg/m3 ) 3.6 3.6 3.6 3.6
Slump (mm) 81 83 86 89

5.2.2 Strain in Restrained Steel Ring

The strain in the restrained steel ring was monitored throughout the test. Figure 5.4
shows the relationship between the strain in the restrained steel ring and age of
mixtures. All results were the average value of two ring specimens. The strain in the
restrained steel ring increased with the increase of concrete age after the concrete
was cast. The outer surface drying was clearly reflected on the curves corresponding
to the increased rate of strain development in the restrained steel ring. After that,
rate of strain development in the restrained steel ring gradually decreased. The strain
achieved a maximum value finally at a certain age then it could be seen that an abrupt
change in the strain readings corresponds to the development of visible crack in the
specimen [36].
The internally cured mixtures measured a lower strain in the restrained steel ring
at the time to cracking. The strain in the restrained steel ring for Mixture SAP25

Fig. 5.4 Strain in the 10 0

restrained steel ring 0 Initiation of drying
-10
-20
Steel strain (με)

-30
-40
-50
-60
-70 SAP0
-80 SAP05
-90 SAP15
(2 days)
SAP25
-100
0 3 6 9 12 15 18 21
Age (d)

@seismicisolation
@seismicisolation
5.2 Cacking Resistance of Early-Age Concrete Internally Cured with SAPs 225

showed it shrunk at a slower rate and shrunk later than the reference Mixture SAP0.
Both of the features improve cracking resistance. Shrinkage at a slower rate will allow
relaxation to reduce tensile stress and shrinkage at a later age allows the concrete to
develop more tensile strength before being loaded with residual tension [31].
The strain in the restrained steel ring decreased with the increase of amount of
SAPs, in spite of applying an identical size of concrete ring and an exposed condition.
For example, the strain in the restrained steel ring at the age of 7 d was − 73, − 62,
− 50, and − 35 με which decreased by 15.1%, 31.5%, and 52.1% when the amount
of SAPs increased from 0% to 0.05%, 0.16%, and 0.26% for Mixture SAP0, SAP05,
SAP15, and SAP25, respectively. The mixture with a higher amount of SAPs showed
a much lower strain in the restrained steel ring at the age that the concrete of Mixture
SAP05 cracked. The strain was − 89 − 78, − 68, and − 58 με which decreased by
12.4%, 23.6%, and 34.8% when the amount of SAPs increased from 0% to 0.05%,
0.16%, and 0.26% for Mixture SAP0, SAP05, SAP15, and SAP25, respectively. The
effect of IC on the strain in steel ring is also investigated in [37–40]. Result in [37]
shows that testing of the mixture with 23.7% replacement amount of saturated LWAs
is stopped after the age of 13 d and has not developed any residual strain. Result in
[38] shows that the strain at the age of 3 d decreases by 64.3% for plain Mixture when
using SAPs to provide 33% of the chemical shrinkage volume. Result in [39] shows
that autogenous shrinkage at the age of 7 d decreases by 10.0%, and 94.8%, when the
replacement amount of IC material increases from 0 to 6%, and 20%, respectively.
Result in [40] shows that the autogenous shrinkage at the age of 7 d decreases by
100% when the amount of porous ceramic coarse aggregate (an IC material) reaches
40%.

5.2.3 Residual Stress of Concrete with SAPs

The actual maximum residual tensile stresses for the mixtures tested were obtained
using Eq. (5.4), as shown in Fig. 5.5.
Several interests could be noted from these curves. First, the residual stress of
concrete ring decreased with the increase of amount of SAPs. The maximum residual
stress was 4.60, 4.38, 3.81, and 3.28 MPa which decreased by 4.8%, 17.2%, and
28.7% when the amount of SAPs increased from 0% to 0.05%, 0.16%, and 0.26%
for Mixture SAP0, SAP05, SAP15, and SAP25, respectively. The maximum residual
stress at the age of 7 d was 4.10, 3.49, 2.84, and 1.98 MPa which decreased by 14.9%,
30.7%, and 51.7% when the amount of SAPs increased from 0% to 0.05%, 0.16%, and
0.26% for Mixture SAP0, SAP05, SAP15, and SAP25, respectively. The addition of
SAPs to concrete with a low w/c ratio can reduce autogenous shrinkage and cracking
potential considerably. The results on residual stress were in accordance with the
results in [38, 39]. Result in [38] shows that the residual stress of the mixture which
provides 33% of the chemical shrinkage volume using SAPs is reduced by 30%
when compared with the plain concrete. The residual stress decreases by 83.3%, and
40% when the chemical shrinkage volume using SAPs is reduced from 0 to 33%,

@seismicisolation
@seismicisolation
226 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.5 Maximum residual 5.5

SAP0
stress development in 5.0
SAP05
concrete ring specimens with 4.5 SAP15
SPAs 4.0 SAP25

Residual stress (MPa)

3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
0 3 6 9 12 15 18 21
Age (d)

and 50%, respectively. Result in [39] shows that the residual stress at the age of 3
d decreases by 10%, and 55% when the amount of IC material increases from 0 to
6%, and 20%, respectively.

5.2.4 Stress Rate of Concrete with SAPs

Fitting curves were calculated by Eq. (5.16), as shown √ in Fig. 5.6. The constant α
was − 34.582, − 29.056, − 22.870, and − 11.117 με/ d with an R 2 value of 0.986,
0.979, 0.980, and 0.975 for Mixture SAP0, SAP05, SAP15, and SAP25, repectively.
The values of stress rate for the test mixtures from initiation of drying are shown in
Fig. 5.7.

Fig. 5.6 Relationship 0

between net strain and
square root of net age
-20
Net strain (με)

-40 Test data of SAP0

Test data of SAP05
Test data of SAP15
-60 Test data of SAP25
Fitting curve of SAP0
Fitting curve of SAP05
-80 Fitting curve of SAP15
Fitting curve of SAP25

0.5 1.0 1.5 2.0 2.5 3.0

Square root of elapsed time (days1/2)

@seismicisolation
@seismicisolation
5.2 Cacking Resistance of Early-Age Concrete Internally Cured with SAPs 227

Fig. 5.7 Development of 1.8

SAP0
stress rate of different
SAP05
concrete mixtures with SPAs 1.5 SAP15
SAP25

Stress rate (MPa/d)

1.2

0.9

0.6

0.3

0 1 2 3 4 5 6 7 8 9
Elapsed time (d)

The general trend of the results showed that the stress rate decreased with the
increase of amount of SAPs. The stress rate at the age of 4 d or 7 d (te ) was 0.49,
0.41, 0.32, and 0.24 MPa/d or 0.37, 0.31, 0.25, and 0.18 MPa/d which decreased
by 16.3%, 34.7%, and 51.0% or 16.2%, 32.4%, and 51.4% when the amount of
SAPs increased from 0% to 0.05%, 0.16%, and 0.26% for Mixture SAP0, SAP05,
SAP15, and SAP25, respectively. The cracking potential was in “Moderate-High”
zone or “Moderate-Low” zone for Mixtures SAP0 and SAP05 or Mixtures SAP15
and SAP25, respectively. Thus, the concrete with a high amount of SAPs showed a
low cracking potential based on results of stress rate.

5.2.5 Stress Relaxation of Concrete with SAPs

Autogenous shrinkage measured in sealed specimens, was used to fit by Eq. (5.1),
and the results are shown in Fig. 5.8. The values of C1 , C2 , and C3 are shown in
Table 5.3.
The compressive elastic modulus at the age of 28 d was 3.45 × 104 , 3.39
× 104 , 3.30 × 104 , and 3.13 × 104 MPa for Mixture SAP0, SAP05, SAP15,
and SAP25, respectively. The time-dependent elastic modulus of concrete was
determined according to CEB 90 Code Model [41], as shown in Eq. (5.18).
√
E c (t) = E c28 · {ex p[0.5S(1 − 28/t)]} (5.18)

where E c (t) represents the compressive elastic modulus of concrete at the age of t,
in GPa; E c28 represents the compressive elastic modulus of concrete at the age of 28
d, in GPa; and S = 0.25 for Ordinary Portland Cement. The relationships between
elastic modulus and age of concrete for Mixture SAP0, SAP05, SAP15, and SAP25

@seismicisolation
@seismicisolation
228 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.8 Relationship 50

Test result of SAP0
between autogenous 0 Test result of SAP05
shrinkage and age of Test result of SAP15

Autogenous shrinkage (με)

concrete with SAPs -50 Test result of SAP25

-100
-150
-200
-250
-300
Fitting curve of SAP0
-350 Fitting curve of SAP05
Fitting curve of SAP15
-400 Fitting curve of SAP25

0 5 10 15 20 25 30
Age (d)

Table 5.3 Regression coefficients

Mixture C1 C2 C3 R2
SAP0 -87.01 0.44 -38.12 0.989
SAP05 -91.25 0.42 -37.68 0.982
SAP15 -81.89 0.39 -32.37 0.964
SAP25 -69.04 0.42 -28.46 0.976

are shown in Fig. 5.9. Thus, the maximum theoretical elastic stress was shown in
Fig. 5.10.
The relaxed stress was influenced by the strain in the restrained steel ring, the free
shrinkage and elastic modulus of concrete, as shown in Eq. (5.14). The strain in the
restrained steel ring, the free shrinkage and elastic modulus of concrete decreased

Fig. 5.9 Relationship

between elastic modulus and
age of concrete with SPAs

@seismicisolation
@seismicisolation
5.2 Cacking Resistance of Early-Age Concrete Internally Cured with SAPs 229

10 10
Residual stress Mixture SAP0 Residual stress Mixture SAP15
9 9
Theoretical elastic stress Theoretical elastic stress
8 8
7 7
Stress (MPa)

6 6

Stress (MPa)
5 5
4 4
3 3
2 2
1 1
0 0
-1 -1
0 3 6 9 12 15 18 21 0 3 6 9 12 15 18 21
Age (d) Age (d)
(a) SAP0 (c) SAP15

10 10
Residual stress Mixture SAP05 Residual stress Mixture SAP25
9 9
Theoretical elastic stress Theoretical elastic stress
8 8
7 7
Stress (MPa)

6
Stress (MPa)

6
5 5
4 4
3 3
2 2
1 1
0 0
-1 -1
0 3 6 9 12 15 18 21 0 3 6 9 12 15 18 21
Age (d) Age (d)
(b) SAP05 (d) SAP25

Fig. 5.10 Comparison on stress and age of concrete with SPAs

with the increase of amount of SAPs. The decrease degree of theoretical elastic
stress was lower than that of corresponding residual stress when the amount of SAPs
increased from 0% to 0.05%, 0.16%, and 0.26%, respectively, as shown in Fig. 5.11.
The theoretical elastic stress or the corresponding residual stress at the age of 6 d
after the concrete was cast was 4.58, 4.53, 3.93, and 3.23 MPa or 3.73, 3.36, 2.60,
and 1.77 MPa which decreased by 1.1%, 14.2%, and 29.5% or 0.10%, 30.3%, and
52.5% when the amount of SAPs increased from 0% to 0.05%, 0.16%, and 0.26% for
Mixture SAP0, SAP05, SAP15, and SAP25, respectively. Thus, the relaxed stress
increased with the increase of amount of SAPs after the initiation of drying. The
relaxed stress at the age of 6 d or 9 d after the concrete was cast was 0.85, 1.17, 1.33,
and 1.46 MPa or 0.90, 1.52, 1.62, and 1.69 MPa which increased by 37.6%, 56.5%,
and 71.8% or 68.9%, 80.0%, and 87.8% when the amount of SAPs increased from
0% to 0.05%, 0.16%, and 0.26%, respectively. The result that relaxed stress increased
with the increase of amount of SAPs was in accordance with the results in [42]. Two
prismatic test pieces 150 mm cube cross section and 500 mm in height are used for
the determination of the basic creep behavior, and the results show that the creep
strain increases by 16.7%, 50.0%, and 51.7% when the SAPs amount increase from

@seismicisolation
@seismicisolation
230 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

0 to 1.11, 1.56, and 2.00 kg/m3 , respectively. And the relaxed stress is proportional
to the creep strain [43].
The relationship between the ratio RA and concrete age was obtained by using
Eq. (5.14), as shown in Fig. 5.12. The mixture with SAPs relaxed more than 60% of
elastic stress at most, after which the relaxation behavior was less obvious. The ratio
of relaxed stress to theoretical elastic stress increased with the increase of amount
of SAPs. The ratio RA at the age of 3 d or 5 d was 0.530, 0.602, 0.648, and 0.732
or 0.276, 0.349, 0.405, and 0.532 which increased by 13.6%, 22.3%, and 38.1% or
26.4%, 46.7%, and 92.8% when the amount of SAPs increased from 0% to 0.05%,
0.16%, and 0.26%, for Mixture SAP0, SAP05, SAP15, and SAP25, respectively.
Thus, the ratio of relaxed stress to theoretical elastic stress increased with the increase
of amount of SAPs.

Fig. 5.11 Estimated relaxed 2.5

stress in concrete ring
specimens with SAPs 2.0
Relaxed stress (MPa)

1.5

1.0

0.5 SAP0
SAP05
0.0 SAP15
SAP25

0 2 4 6 8 10
Age (d)

1.0
Ratio of relaxed to theoretical elastic stress

Fig. 5.12 Relationship

between ratios of relaxed to 0.9
SAP0
theoretical elastic stress and SAP05
age of concrete with SAPs 0.8 SAP15
0.7 SAP25
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1 2 3 4 5 6 7
Age (d)

@seismicisolation
@seismicisolation
5.2 Cacking Resistance of Early-Age Concrete Internally Cured with SAPs 231

5.2.6 Time to Cracking of Concrete with SAPs

The time to cracking for concrete increased with the increase of amount of SAPs, and
cracking behaviors were illustrated in Fig. 5.13. The time to cracking was 10.3, 11.1,
15.2, and 18.1d after the concrete was cast which increased by 7.8, 47.6, and 75.7%
when the amount of SAPs increased from 0 to 0.05, 0.16, and 0.26% for Mixture
SAP0, SAP05, SAP15, and SAP25. The result was in accordance with the results
in [37, 44]. The time to cracking for the concrete under unsealed curing conditions
increases by 300%, and 600% when the replacement volume of LWAs increases
from 0% to 23.7%, and 29.3%, respectively [37]. The cracking potential decreases
by 50% when the replacement volume of LWAs increases from 0 to 10% [44].
The axial tensile strength of concrete may be determined either directly by a
uniaxial tensile test or indirectly by a splitting tensile test [45]. The splitting tensile
test is widely used because of its convenience. The splitting tensile strength of Mixture
SAP0, SAP05, SAP15, and SAP25 was 6.27, 5.81, 5.01, and 4.28 MPa, respectively.
Equation (5.19) was used to obtain axial tensile stress from splitting tensile stress
[46]. Axial tensile strength of the concrete at different ages, f t (t), was calculated by
using Eqs. (5.19) and (5.20) [34].

f t28 = 0.77 × f ts28 + 0.21 (5.19)

(a) SAP0 (b) SAP05

(c) SAP15 (d) SAP25

Fig. 5.13 Concrete cracking

@seismicisolation
@seismicisolation
232 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …
{ }
f t (t) = f t28 exp −λ1 [ln(1 + (t − t0 ))]−k1 (5.20)

where f t (t) is the axial tensile strength of concrete at time t, in MPa; f ts28 and f t28
are the splitting tensile strength and axial tensile strength at 28 d, in MPa; t0 is the
time when penetration resistance is 1.5 MPa, in d; and the regression coefficients
were obtained as λ1 = 0.204, k1 = 1.292.
Thus, the time dependent axial tensile strength of ICC was determined from
splitting tensile strength. The residual tensile stresses at any age of concrete internally
cured with SAPs were obtained before. To better understand the effect of the amount
of SAPs on the time to cracking for ICC, comparisons on the time dependent axial
tensile strength with residual stress were conducted, as shown in Fig. 5.14.
The cracking potential can be assessed by comparing the values of residual stress
with tensile strength [47, 48]. The specimen will crack when these two lines for
residual stress and axial tensile strength intersect, as shown in Fig. 5.14. Similarly,
when the axial tensile strength of the concrete was always larger than the corre-
sponding residual stress, no cracking will occur. Therefore, the theoretical value of
the time to cracking for ICC defined was the time when the residual stress reached
the axial tensile strength of concrete. Table 5.4 shows that the theoretical value of the
time to cracking was 9.4, 10.9, 14.7, and 17.4 d for Mixture SAP0, SAP05, SAP15,

5 5 5 5
Mixture SAP0 Mixture SAP15

Axial tensile strength (MPa)

theoretical time to cracking
Residual stress (MPa)

Axial tensile strength (MPa)

(14.7 d)
Residual stress (MPa)

4 theoretical time to cracking 4 4 4

(9.4 d)

3 3 3 3

2 Residual stress 2 2 2
Axial tensile Residual stress
strength Axial tensile
1 1 1 strength 1

0 0 0 0
0 4 8 12 16 20 0 4 8 12 16 20
Age (d) Age (d)
(a) SAP0 (c) SAP15
5 5 5 5
Mixture SAP05 Mixture SAP25
Residual stress
Axial tensile strength (MPa)

Axial tensile strength (MPa)

Axial tensile
Residual stress (MPa)

4 4 4 4
Residual stress (MPa)

strength
theoretical time to cracking
(10.9 d)

3 3 3 3

Residual stress
2 Axial tensile 2 2 2
strength theoretical time to cracking
(17.4 d)

1 1 1 1

0 0 0 0
0 4 8 12 16 20 0 4 8 12 16 20
Age (d) Age (d)
(b) SAP05 (d) SAP25

Fig. 5.14 Comparison on of tensile strength and residual stress

@seismicisolation
@seismicisolation
5.2 Cacking Resistance of Early-Age Concrete Internally Cured with SAPs 233

Table 5.4 Properties of lightweight aggregates (%)

Particle size ≤ 2.36 mm ≤ 4.75 mm ≤ 9.5 mm ≤ 12.5 mm
Volumetric percentage 4.18 35.88 99.04 100

and SAP25, respectively. The theoretical value of the time to cracking was lower than
the practical value of the time to cracking for Mixture SAP0, SAP05, SAP15, and
SAP25, respectively. The results of the deviations for the time to cracking obtained
from theoretical value and the test were 8.7%, 1.8%, 3.3%, and 3.9% for Mixture
SAP0, SAP05, SAP15, and SAP25, respectively.

5.2.7 Summary

Experimental findings on the effect of the IC with SAPs on the strain of steel ring,
residual stress of concrete, stress rate of concrete, stress relaxation of concrete and
time to cracking of concrete were drawn. When the amount of SAPs used was lower
than the required quantities for IC, the following conclusions can be drawn based on
the obtained results.
(1) The strain in the restrained steel ring decreased with the increase of amount of
SAPs. The strain in the restrained steel ring at the age of 7 d was − 73, − 62,
− 50, and − 35 με which decreased by 15.1%, 31.5%, and 52.1% when the
amount of SAPs increased from 0% to 0.05%, 0.16%, and 0.26% for Mixture
SAP0, SAP05, SAP15, and SAP25, respectively.
(2) The residual stress of concrete ring decreased with the increase of amount of
SAPs. The maximum residual stress was 4.60, 4.38, 3.81, and 3.28 MPa which
decreased by 4.8%, 17.2%, and 28.7% when the amount of SAPs increased
from 0% to 0.05%, 0.16%, and 0.26% for Mixture SAP0, SAP05, SAP15, and
SAP25, respectively.
(3) The stress rate decreased with the increase of amount of SAPs. The stress rate at
the age of 7 d after the initiation of drying was 0.37, 0.31, 0.25, and 0.18 MPa/d
which decreased by 16.2%, 32.4%, and 51.4%, when the amount of SAPs
increased from 0% to 0.05%, 0.16%, and 0.26% for Mixture SAP0, SAP05,
SAP15, and SAP25, respectively. The cracking potentials were in “Moderate-
High” or in “Low” zone for Mixtures SAP0 and SAP05 or Mixtures SAP15 and
SAP25, respectively.
(4) The relaxed stress increased with the increase of age in 2 d after the concrete
was cast. And the relaxed stress increased with the increase of amount of SAPs.
The ratio of relaxed stress to theoretical elastic stress increased with the increase
of amount of SAPs. The ratio of relaxed stress to theoretical elastic stress of
concrete at the age of 3 d was 0.530, 0.602, 0.648, and 0.732 which increased
by 13.6%, 22.3%, and 38.1%, when the amount of SAPs increased from 0% to
0.05%, 0.16%, and 0.26%, respectively.

@seismicisolation
@seismicisolation
234 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

(5) The time to cracking for concrete increased with the increase of the amount
of SAPs. The time to cracking obtained from test was 10.3, 11.1, 15.2, and
18.1 d which increased by 7.8%, 47.6%, and 75.7% when the amount of SAPs
increased from 0% to 0.05%, 0.16%, and 0.26% for Mixture SAP0, SAP05,
SAP15, and SAP25, respectively.

5.3 Cracking Resistance of Early-Age Concrete Internally

Cured with LWAs

5.3.1 Mixture Proportion

1. Materials

Portland cement (P·II 52.5R) with the Blaine fineness of 375 m2 /kg was utilized.
The results of penetration test reflected that the initial setting time and final setting
time of cement paste were 167 and 221 min, respectively. The normal river sand
(maximum size of 2.6 mm) was employed as the fine aggregates. As the normal
weight coarse aggregates, crushed limestones (maximum size of 20 mm) were
employed. LWAs employed in Mixture LW10, LW30, and LW50 were manufac-
tured rotary kiln expanded clays. The grading curves of LWAs, fine aggregates, and
normal weight coarse aggregates are shown in Fig. 5.15.
The dry-bulk density of LWAs employed was 1050 kg/m3 . The particle size of
LWAs was from 4.7 to 12.5 mm. Table 5.4 shows the properties of the LWAs. The
crushing strength of dry LWAs was 1.7 MPa in terms of Chinese Standard GB/T
17431.2-2010 [49]. For improving the workability of concrete mixtures and reducing
consumption of tap water, PCA-I superplasticizer was employed during the mixing
process.

2. Mixture proportions
All the concrete mixtures had the same w/c ratio of 0.33. According to the different
proportions of pre-wetted LWAs, four kinds of concrete mixtures named as Mixture
LW0, LW10, LW30, and LW50 are shown in Table 5.5 [19]. The reference concrete
mixture without pre-wetted LWAs was named as Mixture LWA0. Part of the normal
weight coarse aggregates in ICC mixtures were replaced by LWAs, and the volume
replacement ratio was 10%, 30%, and 50% for Mixture LW10, LW30, and LW50,
respectively.

@seismicisolation
@seismicisolation
5.3 Cracking Resistance of Early-Age Concrete Internally Cured with LWAs 235

Fig. 5.15 Aggregates 120

110 LWAs
utilized in experiment Lower limit (ASTM C1761)
100 Upper limit (ASTM C1761)

Percent passing (%)

90
80
70
60
50
40
30
20
10
0
0 2 4 6 8 10 12
Diameter (mm)
(a) grading curve of LWAs
120
Fine aggregate
110
Lower limit (ASTM C1761)
100 Upper limit (ASTM C1761)
Percent passing (%)

90
80
70
60
50
40
30
20
10
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Diameter (mm)
(b) grading curve of fine aggregates
120
Normal weight coarse aggregate
110 Lower limit (ASTM C1761)
100 Upper limit (ASTM C1761)
90
Percent passing (%)

80
70
60
50
40
30
20
10
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26
Diameter (mm)
(c) grading curve of normal weight coarse aggregates

@seismicisolation
@seismicisolation
236 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Table 5.5 Mix proportions of concrete

Mixture composition Concrete mixtures
LW0 LW10 LW30 LW50
Water (kg/m3 ) 171 171 171 171
Cement (kg/m3 ) 512 512 512 512
Fine aggregates (kg/m3 ) 636 636 636 636
Normal weight coarse aggregates (kg/m3 ) 1131 1018 792 566
Lightweight aggregates (kg/m3 ) 0 44.6 133.9 223.2
IC water (kg/m3 ) 0 5.35 16.06 26.78
Superplasticizer (kg/m3 ) 3.58 3.58 3.58 3.58
Slump (mm) 82 83 87 89

5.3.2 Strain in Restrained Steel Ring

The strain that developed in the restrained steel rings was monitored throughout the
restrained ring test. Figure 5.16 shows the relationship between strain of steel ring
and age for Mixture LW0, LW10, LW30, and LW50.
Overall, the strain increased with age, and decreased with the increase of the
proportion of pre-wetted LWAs. The absolute value of strain of steel ring at the
age of 7 d was 73, 59, 44, and 36 με for Mixture LW0, LW10, LW30, and LW50,
respectively. Compared with Mixture LW0, the absolute value of strain at the age
of 7 d decreased by 19%, 40%, and 51% for Mixture LW10, LW30, and LW50,
respectively. The investigation in [37] indicates that the development strain of the
mixture ceases to increase at 13 d when the proportion of LWAs reaches 23.7%.
The investigation in [39] indicates the absolute value of strain decreases by 10.0%
and 94.8% at 7 d when the proportion of IC agent increases from 0 to 6%, and 20%,

Fig. 5.16 Relationship

between strain of steel ring 0
and age of different concrete
mixtures
Strain of steel ring (με)

-20

-40

-60
LW0
LW10
-80
LW30
LW50
-100
0 2 4 6 8 10 12 14 16 18
Age (d)

@seismicisolation
@seismicisolation
5.3 Cracking Resistance of Early-Age Concrete Internally Cured with LWAs 237

Fig. 5.17 Maximum 5.0

LW0
residual stress development 4.5 LW10
in concrete ring

Maximum residual stress (MPa)

4.0 LW30
specimens with LWAs LW50
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
0 3 6 9 12 15 18
Age (d)

respectively. The investigation in [40] indicates the absolute value of strain decreases
by 100% at 7 d when the proportion of IC agent increases from 0 to 40%.

5.3.3 Residual Stress of Concrete with LWAs

Figure 5.17 gives the expression of the relationship between the maximum residual
stresses.
The maximum residual stress of concrete ring decreased significantly with the
increase of the proportion of pre-wetted LWAs. The maximum residual stress at the
age of 8.3 d was 4.56, 3.57, 2.85, and 2.17 MPa for Mixture LW0, LW10, LW30,
and LW50, respectively. Compared with Mixture LW0, the maximum residual stress
at the age of 8.3 d decreased by 21.71%, 37.50%, and 52.41% for Mixture LW10,
LW30, and LW50, respectively. The investigation in [39] indicates that the residual
stress at the age of 3 d decreases by 10% and 55% when the proportion of IC agent
increases from 0 to 6%, and 20%, respectively.

5.3.4 Stress Rate of Concrete with LWAs

The√constant α in Eq. (5.16) was − 33.893, − 21.661, − 19.709, and − 16.354

με/ d, and the coefficient of determination R2 was 0.976, 0.951, 0.974, and 0.954
for Mixture LW0, LW10, LW30, and LW50, repectively, as shown in Fig. 5.18.
Figure 5.19 shows that stress rate of the concrete mixture decreases with the
increase of the proportion of pre-wetted LWAs. When concrete mixtures were at
the net age of 7 d (te ), the stress rate was 0.363, 0.232, 0.211, and 0.175 MPa/d for
Mixture LW0, LW10, LW30, and LW50, respectively. Compared with Mixture LW0,

@seismicisolation
@seismicisolation
238 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.18 Relationship 20

between net strain and 10
square root of net age 0
-10

Net strain (με)

-20
-30
-40 Test data of LW0
Test data of LW10
-50 Test data of LW30
Test data of LW50
-60 Fiting curve of LW0
-70 Fiting curve of LW10
Fiting curve of LW30
-80 Fiting curve of LW50

-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

Square root of net age (d1/2)

the stress rate at the net age of 7 d (te ) decreased by 36.088%, 41.873%, and 51.791%
for Mixture LW10, LW30, and LW50, respectively. The stress rate classification
method was utilized to assess cracking of the concrete mixture, as shown in Table
5.5. The time to cracking of Mixture LW0 was at the age of 10.4 d and that of Mixture
LW10 was at the age of 12.7 d. The cracking phenomenon of LW50 was observed at
the age of 8.7 d. Mixture LW30 did not crack during the test time. Therefore, Mixture
LW0 was between the “High” zone and the “Moderated-High” zone. Mixture LW10
and LW50 belonged to the “Moderated-High” zone. Mixture LW30 belonged to the
“Low” zone. According to the results of stress rate, the concrete mixture with the
higher proportion of pre-wetted LWAs has lower stress rate.

Fig. 5.19 Development of 1.8

stress rate of different LW0
LW10
concrete mixtures with 1.5
LW30
LWAs
Stress rate (MPa/d)

LW50
1.2

0.9

0.6

0.3

0.0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Net age (d)

@seismicisolation
@seismicisolation
5.3 Cracking Resistance of Early-Age Concrete Internally Cured with LWAs 239

5.3.5 Stress Relaxation of Concrete with LWAs

The autogenous shrinkage obtained from the measurement was utilized to fit the
function, as shown in Fig. 5.20. Constants C 1 , C 2 and C 3 as well as the coefficient
of determination R2 are shown in Table 5.6.
The compressive elastic modulus was 34.5, 33.5, 31.0, and 28.6 GPa for Mixture
LW0, LW10, LW30, and LW50 at the age of 28 d, respectively. Increasing the propor-
tion of pre-wetted LWAs caused the elastic modulus of concrete mixture to decrease.
The time-dependent elastic modulus was obtained in terms to [41], as expressed in
Eq. (5.18). The relationships between the elastic modulus of Mixture LW0, LW10,
LW30, and LW50 and the age of concrete mixtures are shown in Fig. 5.21.
Figure 5.22 shows that the maximum theoretical elastic stress decreases when the
proportion of pre-wetted LWAs increases from 0 to 10, 30, and 50%.
Figure 5.23 shows the relationship between relaxed stress and age.
Generally, the relaxed stress increased, and after the certain degree of decrease
it began to stabilize. As far as the proportion change was concerned, the relaxed
stress of the concrete mixture decreased when the proportion of pre-wetted LWAs
increased from 0 to 10, 30, and 50%. The relaxed stress at the age of 2 or 6 d was 1.52,
0.91, 0.66, and 0.46 MPa or 0.19, 0.14, 0.09, and 0.05 MPa for Mixture LW0, LW10,
LW30, and LW50, respectively. Compared with Mixture LW0, the relaxed stress at

50
Fig. 5.20 Relationship
Test result of LW0
between autogenous 0 Test result of LW10
Autogenous shrinkage (µε)

shrinkage and age of Test result of LW30

-50 Test result of LW50
different concrete
Fitting curve of LW0
mixtures with LWAs -100 Fitting curve of LW10
Fitting curve of LW30
-150 Fitting curve of LW50

-200

-250

-300

-350

-400
0 5 10 15 20 25 30
Age (d)

Table 5.6 Regression coefficients

Mixture C1 C2 C3 R2
LW0 − 85.53 0.45 − 37.22 0.989
LW10 − 56.47 0.56 − 33.69 0.982
LW30 − 40.35 0.61 − 27.62 0.964
LW50 − 33.77 0.62 − 23.51 0.979

@seismicisolation
@seismicisolation
240 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.21 Relationship between elastic modulus and age of different concrete mixtures with LWAs

10 10
Maximum theoretical elastic stress (MPa)

10 10

Maximum theoretical elastic stress (MPa)

Maximum residual stress LW0 Maximum residual stress LW30
9 9 9 9
Maximum residual stress (MPa)

Maximum residual stress (MPa)

Maximum theoretical elastic stress Maximum theoretical elastic stress

8 8 8 8
7 7 7 7
6 6 6 6
5 5 5 5
4 4 4 4
3 3 3 3
2 2 2 2
1 1 1 1
0 0 0 0
-1 -1 -1 -1
0 3 6 9 12 15 18 0 3 6 9 12 15 18
Age (d) Age (d)

(a) LW0 (c) LW30

10 10
Maximum theoretical elastic stress (MPa)

10 10
Maximum theoretical elastic stress (MPa)

Maximum residual stress LW10

9 9
Maximum residual stress (MPa)

Maximum residual stress LW50

Maximum theoretical elastic stress 9 9
Maximum residual stress (MPa)

8 8 Maximum theoretical elastic stress

8 8
7 7 7 7
6 6 6 6
5 5 5 5
4 4 4 4
3 3 3 3
2 2 2 2
1 1 1 1
0 0 0 0

-1 -1 -1 -1
0 3 6 9 12 15 18 0 3 6 9 12 15 18
Age (d) Age (d)

(b) LW10 (d) LW50

Fig. 5.22 Comparison on stress and age of concrete with LWAs

@seismicisolation
@seismicisolation
5.3 Cracking Resistance of Early-Age Concrete Internally Cured with LWAs 241

Fig. 5.23 Estimated relaxed LW0

stress of different concrete 1.5
LW10
mixtures with LWAs LW30

Relaxed stress (MPa)

LW50
1.0

0.5

0.0

0 2 4 6 8 10
Age (d)

the age of 2 or 6 d decreased by 40.13%, 56.58%, and 69.74% or 26.33%, 52.64%,

and 73.68% for Mixture LW10, LW30, and LW50, respectively. The investigations
in [50] indicate that when the loading level is 60% of tensile strength at the age of 28
d, the creep strain decreases by 22.6% when the proportion of the IC agent increases
from 0 to 1.13 kg/m3 .

5.3.6 Time to Cracking of Concrete with LWAs

The time to cracking was postponed with the increasing proportion of pre-wetted
LWAs when the proportion was lower than 30%, as illustrated in Fig. 5.16. Mixture
LW0 cured traditionally cracked at 10.4 d. The time to cracking for LW10 with 10%
proportion of pre-wetted LWAs was 12.7 d. Mixture LW30 was not observed the
phenomenon of cracking during the test time. Compared with the other concrete
mixtures, Mixture LW50 exhibited the earliest cracking at the age of 8.7 d. The
cracking behaviors of concrete mixtures with the different proportions of pre-
wetted LWAs are shown in Fig. 5.24. When the proportion of pre-wetted LWAs
increased within the certain range, the time to cracking could be effectively post-
poned. However, the further increase in the proportion of LWAs resulted in the
adverse effect on the time to cracking.
The axial tensile strength could be obtained from the uniaxial tensile test or the
splitting tensile test. The splitting tensile strength at the age of 28 d was 6.27, 5.27,
4.53, and 2.65 MPa for Mixture LW0, LW10, LW30, and LW50, respectively. Equa-
tion (5.19) gives an expression to obtain the axial tensile strength ( f t ) at 28 d utilizing
the splitting tensile strength ( f ts ). Equation (5.20) is expressed to determine the axial
tensile strength at different age [51]. Figure 5.25 shows the comparison of axial tensile
strength development with maximum residual stresses.

@seismicisolation
@seismicisolation
242 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

(a) LW0 (b) LW10

(c) LW30 (d) LW50

Fig. 5.24 Concrete cracking of concrete with LWAs

5 5 5 5
LW0 LW30 Maximum residual stress
Axial tensile strength
Maximum residual stress (MPa)

Maximum residual stress (MPa)

4 Theoretical time to cracking 4 4 4

Axial tensile strength (MPa)

(9.4 d)

Axial tensile strength (MPa)

Maximum residual stress
Axial tensile strength
3 3 3 3

2 2 2 2

1 1 1 1

0 0 0 0
0 3 6 9 12 15 18 0 3 6 9 12 15 18
Age (d) Age (d)
(a) LW0 (c) LW30
5 5 5 5
LW10 LW50 Maximum residual stress
Axial tensile strength
Maximum residual stress (MPa)

Maximum residual stress (MPa)

4 4 4 4
Axial tensile strength (MPa)

Axial tensile strength (MPa)

3 3 3 Theoretical time to cracking 3

Theoretical time to cracking (7.9 d)
(12.1 d)
Maximum residual stress
2 Axial tensile strength 2 2 2

1 1 1 1

0 0 0 0
0 3 6 9 12 15 18 0 3 6 9 12 15 18
Age (days) Age (d)
(b) LW10 (d) LW50

Fig. 5.25 Comparison on of tensile strength and residual stress of concrete with LWAs

@seismicisolation
@seismicisolation
5.3 Cracking Resistance of Early-Age Concrete Internally Cured with LWAs 243

Table 5.7 Time to cracking

Mixture Theoretical value Practical value Deviation (%)
of concrete
(d) (d)
LW0 9.4 10.4 9.6
LW10 12.1 12.7 4.7
LW30 – – –
LW50 7.9 8.7 9.2

Moreover, the concrete mixtures would probably crack when these two lines inter-
sected, as shown in Fig. 5.25. Therefore, when the maximum residual stress exceeded
the axial tensile strength, the cracking occurred and this time was considered as the
theoretical time to cracking. The theoretical time to cracking was 9.4, 12.1, and 7.9 d
for Mixture LW0, LW10, and LW50, respectively. The practical time to cracking was
slightly postponed than the theoretical time to cracking for LW0, LW10, and LW50.
The results of the deviation for the time to cracking between theoretical value and the
test value was 9.6%, 4.7%, and 9.2% for Mixture LW0, LW10, and LW50, respec-
tively, as shown in Table 5.7. The deviations between practical time and theoretical
time were small.
Cement or mortar with lower density is floated in concrete, while normal coarse
aggregates are submerged [32]. When concrete is mixed with pre-wetted LWAs, pre-
wetted LWAs are floated due to the lower density [52]. The floatation of the pre-wetted
LWAs not only reduces the effective evaporation area, but also reduces bleeding [37].
The fraction of the paste within pre-wetted LWAs also increases with the increase in
the replacement rate, which makes the paste obtain water supply more effectively. As
a result, within the certain pre-wetted LWAs proportion range, the microstructure of
the paste became denser and stronger to resist cracking, thus postponing the time to
cracking, as reported in [37, 52]. However, when the proportion of pre-wetted LWAs
exceeded the limit value, due to the low strength of pre-wetted LWAs, the axial tensile
strength of concrete decreased, and the time to cracking was not postponed. Although
the high proportion of pre-wetted LWAs had an adverse effect on the performance
of concrete, pre-wetted LWAs could still be considered as the advantageous material
in IC to reduce the shrinkage and postpone the time to cracking of concrete.

5.3.7 Summary

The conclusions of the effect of pre-wetted LWAs on concrete mixtures with the
experimental data analysis were reported. The following conclusions would be
summarized as follows.
(1) The strain of steel ring decreased with the increase of the proportion of pre-
wetted LWAs. The absolute value of strain of steel ring at 7 d was 73, 59, 44,

@seismicisolation
@seismicisolation
244 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

and 36 με which decreased by 19%, 40%, and 51% when the proportion of pre-
wetted LWAs increased from 0 to 10%, 30%, 50% for Mixture LW0, LW10,
LW30, and LW50, respectively.
(2) The maximum residual stress of concrete mixture decreased with the increase
of the proportion of pre-wetted LWAs. The maximum residual stress at the age
of 8.3 d was 4.56, 3.57, 2.85, and 2.17 MPa for Mixture LW0, LW10, LW30,
and LW50, respectively. Compared with Mixture LW0, the maximum residual
stress at the age of 8.3 d decreased by 21.71%, 37.50%, and 52.41% for Mixture
LW10, LW30, and LW50, respectively.
(3) The relaxed stress of concrete mixture decreased with the increase of the propor-
tion of pre-wetted LWAs on the general trend. The relaxed stress at 2 d or 6 d was
1.52, 0.91, 0.66, and 0.46 MPa or 0.19, 0.14, 0.09, and 0.05 MPa for Mixture
LW0, LW10, LW30, and LW50, respectively. Compared with Mixture LW0, the
relaxed stress at 2 d or 6 d decreased by 40.13%, 56.58%, and 69.74% or 26.33%,
52.64%, and 73.78% for Mixture LW10, LW30, and LW50, respectively.
(4) The stress rate of concrete mixture decreased with the increase of the proportion
of pre-wetted LWAs. The stress rate at the net age of 3 d was 0.363, 0.232, 0.211,
and 0.175 MPa/d for Mixture LW0, LW10, LW30, and LW50, respectively.
Compared with Mixture LW0, the stress rate at the net age of 3 d decreased by
36.1%, 41.9%, and 51.8% for Mixture LW10, LW30, and LW50, respectively.
According to the classification method of stress rate, Mixture LW0 was between
the “High” zone and the “Moderated-High” zone. Mixture LW10 and LW50
belonged to the “Moderated-High” zone. Mixture LW30 belonged to the “Low”
zone.
(5) The time to cracking increased with the increase of the proportion of pre-wetted
LWAs when the proportion of pre-wetted LWAs was less than 30%. The practical
time to cracking was slightly postponed than the theoretical time to cracking
for Mixture LW0, LW10, and LW50. The deviation for the time to cracking
between theoretical and test value was 9.6%, 4.7%, and 9.2% for Mixture LW0,
LW10, and LW50, respectively. No cracking was observed in Mixture LW30.

5.4 Influential Factors of the Cracking Resistance

of Early-Age Internally Cured Concrete Using Ring
Test

5.4.1 Addition of Ground Granulated Blast Furnace Slag

1. Mixture proportion of internal cured concrete with ground granulated blast

furnace slag

(1) Materials

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 245

According to Chinese Standard GB 175-2007 [35], Portland Cement with a cement

grade of 52.5R was used. The Blaine fineness of cement was 375 m2 /kg. Table
5.8 shows the strength characteristics of cement. Normal river sand with a fineness
modulus of 2.2 was used. Crushed limestone with the maximum size of 13 mm was
used as coarse aggregates. The GGBFS (S95) was used as a kind of mineral mixture. A
kind of liquid polycarboxylate-based superplasticizer was mixed in Mixture IGS00,
IGS20, IGS35, and IGS50 to adjust the workability, and tap water was applied as the
mixture and IC water. A kind of SAPs based on cross-linked poly acrylic acid was
used as an IC material. SAPs produced by the technical process are spherical, which
can be single spherical particles or raspberry like agglomerates of smaller spherical
particles. The diameters of the spherical particles varied from 150 to 200 μm, and
the bulk density was 750 kg/m3 in the dry state. The absorption capacity of SAPs
was tested by using the slump flow measurement [53], and was 13 g of water/g.
The slump for Mixture IGS00, IGS20, IGS35, and IGS50 was 165, 171, 177, and
189 mm, which increased by 3.64%, 7.27%, and 14.55%, respectively. In addition,
the slump is also used to indicate the workability of concrete. The workability for
different mixtures increased with the increase of GGBFS content [54]. The study in
[54] shows that the addition of GGBFS improves the workability of concrete, and
reduces the bleeding of fresh concrete or mortar.
(2) Mixture proportions

Four different concrete mixtures with a w/b ratio of 0.32 were designed [21]. The
mixtures were designated as IGS00, IGS20, IGS35, and IGS50, as shown in Table
5.9.

Table 5.8 Strength characteristics of cement

Setting time (min) Flexural strength (MPa) Compressive strength (MPa)
Initial Final 3d 28 d 3d 28 d
159 213 6.6 9.3 35.6 66.9

Table 5.9 Mixture proportions of concrete

Mixture composition (kg/m3 ) Concrete mixtures
IGS00 IGS20 IGS35 IGS50
Water 173 173 173 173
Cement 541.0 432.8 351.6 270.5
GGBFS 0 108.2 189.4 270.5
Fine aggregate 625 625 625 625
Coarse aggregate 1111 1111 1111 1111
Superplasticizer 4.328 4.599 4.599 4.599
Super absorbent polymers 2.589 2.589 2.589 2.589

@seismicisolation
@seismicisolation
246 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.26 Relationship 0

IGS00
between free shrinkage and IGS20
net age of ICC with GGBFS

Free shrinkage strain (με)

IGS35
-50 IGS50

-100

-150

-200
0 5 10 15 20 25 30
Net age (d)

Four mixtures in which the cement was replaced with the same weight of GGBFS,
and the replacement ratios were 0%, 20%, 35%, and 50% for Mixture IGS00, IGS20,
IGS35, and IGS50, respectively. Mixture IGS00 was set as the reference concrete
without GGBFS.

2. Free shrinkage of concrete

Figure 5.26 shows that the development of free shrinkage (εsh ) for Mixture IGS00,
IGS20, IGS35, and IGS50, respectively.
The elapsed time after 1.5 d after casting was defined as the net age. The free
shrinkage at the net age of 21 d for Mixture IGS00, IGS20, IGS35, and IGS50 was −
185, − 166, − 147, and − 129 με, respectively. The absolute value of free shrinkage
decreased by 10.27%, 20.54%, and 30.27% when the GGBFS content increased
from 0%, to 20%, 35%, and 50%, respectively. The absolute value of free shrinkage
decreased with the increase of GGBFS content, which was in accordance with the
results in [55, 56]. The study in [55] shows that the free shrinkage of concrete with
GGBFS at the age of 50 d decreases by 5% and 14% when the GGBFS content
increases from 20 to 40%, and 60%, respectively. The study in [56] shows that
the free shrinkage of concrete decreases greatly with the addition of GGBFS. The
shrinkage reduction was due to the GGBFS content because the SAPs content for
mixtures was the same. An alternative mechanism is that the free shrinkage is directly
associated with the water held in the small pores, the capillary pore refinement of
concrete and the negative pressure of capillary pores increases with the increase of
GGBFS content [57].

3. Strain in the steel ring

Relationship between strain in steel ring and age of concrete was illustrated in
Fig. 5.27.
The residual tensile stress was determined based on the shrinkage in concrete.
However, this strain could not be measured directly. Therefore, the shrinkage in

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 247

Fig. 5.27 Relationship 0

IGS00
between strain in steel ring IGS20
and age of concrete with -10
IGS35
GGBFS -20 IGS50

Actual strain (με)

-30

-40

-50

-60

-70
0 5 10 15 20 25 30
Age (d)
(a) Actual strain

0 IGS00
IGS20
-10 IGS35
IGS50
-20
Net strain (με)

-30

-40

-50

-60
0 5 10 15 20 25 30
Net Age (d)
(b) Net strain

concrete could be determined based on the strain in the steel ring. Figure 5.27 shows
that the actual strain on the inner surface of the steel ring increased with the age of
concrete. The actual strain in the steel ring was calculated based on the test results,
as shown in Fig. 5.27a. The net strain (εnet ) in the steel ring was defined as the
increment of the strain between the actual age and the age of 1.5 d after casting, as
shown in Fig. 5.27b. The actual strain in the steel ring at the age of 28 d after casting
for Mixture IGS00, IGS20, IGS35, and IGS50 was − 65, − 56, − 48, and − 40
με, respectively. The absolute value of actual strain in the steel ring decreased by
13.85%, 26.15%, and 38.46% when the GGBFS content increased from 0%, to 20%,
35%, and 50%, respectively. The absolute value of actual strain decreased with the
increase of GGBFS content, which was in accordance with results in [58]. The study
in [58] shows that the actual strain at the age of 14.25 d after casting was − 83, −
67, − 55, and − 49 με when the GGBFS content increases from 0%, to 20%, 35%,
and 50%, respectively.

@seismicisolation
@seismicisolation
248 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

4.0

Maximum residual tensile stress (MPa)

Fig. 5.28 Relationship
between actual maximum
3.5
residual tensile stress in
concrete ring and age 3.0
2.5
2.0
1.5
1.0 IGS00
IGS20
0.5 IGS35
0.0 IGS50
0 5 10 15 20 25 30
Age (d)

For Mixture IGS20, IGS35, and IGS50, the increase rates of actual strain in the
steel ring were slower than that for Mixture IGS00, which indicated that the addition
of GGBFS reduced the restrained shrinkage of concrete specimens. Results on the
shrinkage were in accordance with the results in [59]. The study in [59] shows that
the shrinkage decreases when the GGBFS content increases from 35%, to 50%, and
70%, respectively.

4. Residual stress of concrete

Figure 5.28 shows the maximum residual tensile stress for Mixture IGS00, IGS20,
IGS35, and IGS50, respectively.
The development rate of residual tensile stress for Mixture IGS00 was consider-
ably higher than that for Mixture IGS20, IGS35, and IGS50 at a very early age. The
maximum residual tensile stress at the age of 28 d after casting for Mixture IGS00,
IGS20, IGS35, and IGS50 was 3.65, 3.16, 2.70, and 2.27 MPa, respectively. The
value of maximum residual tensile stress decreased by 13.42%, 26.03%, and 37.81%
when the GGBFS content increases from 0%, to 20%, 35%, and 50%, respectively.
The results based on the value of maximum residual tensile stress were in accor-
dance with the results in [60]. The study in [60] shows that slag cement decreases
the value of tensile stress. Therefore, the addition of GGBFS decreased the value
of maximum residual tensile stress of ICC with SAPs. The reason may be that the
addition of GGBFS reduces the early-age thermal effect compared with the concrete
with Portland cement [60].

5. Stress relaxation of concrete

Figure 5.29 shows that the relationship between relaxed stress and net age for different
mixtures.
The relaxed stress at the net age of 7, 14, and 21 d for Mixture IGS20 was 1.39,
2.88, and 3.97 MPa, respectively. The relaxed stress for Mixture IGS20 increased
with the net age of ICC, which indicated that the cracking potential for Mixture

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 249

Fig. 5.29 Relationship 5.0

between relaxed stress and IGS00
4.5 IGS20
net age of ICC with GGBFS

Relaxed stress (MPa)

4.0 IGS35
3.5 IGS50
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0 5 10 15 20 25 30
Net age (d)

IGS20 decreased with the net age of ICC. The relaxed stress at the net age of 21 d
for Mixture IGS00, IGS20, IGS35, and IGS50 was 4.27, 3.97, 2.90, and 2.38 MPa,
which decreased by 7.03%, 32.08%, and 44.26% when the GGBFS content increased
from 0 to 20%, 35%, and 50%, respectively. The relaxed stress of ICC with SAPs
decreased with the increase of GGBFS content. The reason may be that the addition
of GGBFS reduces the creep of ICC [56, 61]. The study in [56] shows that the creep
decreases by 51.8% when the GGBFS content increases from 0 to 30%. The study in
[61] shows the results on the creep behavior of concrete with three different GGBFS
contents, and the creep decreases with the increase of GGBFS content.

6. Stress rate of concrete

The strain rate factor (α) in Eq. (5.16) was determined with the linear regression
analysis results about the net steel ring strain and the square root of net age, as shown
in Fig. 5.30. The fitting result of α or R 2 for Mixture IGS00, √ IGS20, IGS35, and
IGS50 was − 16.098, − 14.739, − 13.913, and − 13.263 με/ d or 0.966, 0.960,
0.985, and 0.979, respectively.
For Mixture IGS00, IGS20, IGS35, and IGS50, no cracking was observed when
the test was terminated at the age of 28 d after casting. The stress rate at the
net age of 7 d for Mixture IGS00, IGS20, IGS35, and IGS50 was 0.173, 0.158,
0.149, and 0.142 MPa/d, which decreased by 8.67%, 13.87%, and 17.92% when
the GGBFS content increased from 0 to 20%, 35%, and 50%, respectively. The
stress rate decreased with the increase of GGBFS content, which indicated that the
cracking potential decreased with the increase of GGBFS content. The results based
on the stress rate were in accordance with the results in [62]. The cracking potential
decreases with the decrease of stress rate [62]. In addition, the magnitudes of stress
rate can be compared to evaluate the relative cracking potential of concrete. Cracking
potential of concrete can be classified into four levels depending on cracking time
and stress rate. The cracking potential was in “Moderate-High” level for Mixture

@seismicisolation
@seismicisolation
250 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.30 Relationship

0
between actual strain and
square root of net age of
ICC with GGBFS -10

Net strain (με)

-20

-30
Test data of IGS00
Test data of IGS20
Test data of IGS35
-40 Test data of IGS50
Fitting curve of IGS00
Fitting curve of IGS20
-50 Fitting curve of IGS35
Fitting curve of IGS50

0 1 2 3
Square root of net age (d1/2)

IGS00, and “Moderate-Low” level for Mixture IGS20, IGS35, and IGS50, respec-
tively. Therefore, the cracking potential of ICC with SAPs decreased with the increase
of GGBFS content.
7. Cracking potential of concrete
The tensile stress of concrete is an important factor to evaluate cracking potential
of concrete when the concrete ring is under restrained condition. The cracking may
occur when the axial tensile stress of concrete exceeds the axial tensile strength.
The time-dependent development model of axial tensile strength of concrete can be
calculated by Eqs. (5.21) and (5.22) [46, 63].

f sp,28
f t,28 = (5.21)
[(0.1983 − 0.0017 f cu,28 ) ln t + 0.2282 + 0.009 f cu,28 ]
[ ( / )]
28
f t (t) = f t,28 exp S3 1 − (5.22)
t

where f cu,28 is the compressive strength of concrete at the age of 28 d after casting,
in MPa; f t (t) is the time-dependent axial tensile strength at different ages, in MPa;
and S3 is the fitting parameter obtained by the regression analysis on the test results
of axial tensile strength at the age of 3, 7, and 28 d after casting, respectively.
The cracking potential parameter decreased with the increase of GGBFS content,
as shown in Fig. 5.31.
The cracking potential parameter at the age of 28 d after casting for Mixture
IGS00, IGS20, IGS35, and IGS50 was 0.91, 0.87, 0.81, and 0.73, which decreased by
4.40%, 10.99%, and 19.78% when the GGBFS content increased from 0 to 20%, 35%,
and 50%, respectively. The study in [64] shows the cracking potential of concrete
with slag is lower than that for concrete without slag due to their slower hydration,
expansion of cement matrix, and stress relaxation at early age. Therefore, the cracking

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 251

Fig. 5.31 Relationship 1.0

between cracking potential

Cracking potential parameter

parameter and age of 0.8
ICC with GGBFS

0.6

IGS00
0.4 IGS20
IGS35
IGS50
0.2

0.0
0 5 10 15 20 25 30
Age (days)

potential of ICC with SAPs decreased with the increase of GGBFS content based on
cracking potential parameter.

5.4.2 Addition of Polypropylene Fibers

1. Mixture proportion of internal cured concrete with PP fiber

(1) Materials

According to Chinese Standard, Portland Cement (Cement II 52.5R) with a Blaine

fineness of 375 m2 /kg was utilized, as reported in [35]. The mortar sieved from the
mixtures with w/c of 0.32 were utilized to determine the initial and final setting time.
The setting time started from the time when the water was added in the mixtures.
According to the properties of the mixture, the penetration test was carried out every
half an hour, and the interval of test should be shortened when approaching the initial
and final setting time. When the unit area penetration resistance was 3.5 MPa and
28 MPa, the corresponding time was initial and final setting time respectively. The
results of penetration test reflected that the initial setting time and final setting time
of cement paste were 167 and 221 min. The fineness modulus of normal river sand
(maximum size of 2.6 mm) employed was 2.2. As the coarse aggregates, the crushed
limestones (maximum size of 19 mm) utilized were evenly distributed in the range of
5–16 mm. The coarse aggregates utilized were washed on polythene sheet by water
pump and stored in dry conditions for reducing the influence of the water in coarse
aggregate and surface floating dust on the performance of concrete.
Barchip fiber, a new type of modified PP fiber, was the latest generation of building
fiber from the future fiber research project of Elasto Plastic Concrete. The length
of PP fibers was 42 mm, as shown in Fig. 5.32. The tensile strength of PP fibers
was 550 MPa. The surface texture of PP fibers was continuous embossed. The

@seismicisolation
@seismicisolation
252 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.32 The diagram of

the length of PP fibers

elastic modulus of PP fibers is 8 GPa. A kind of covalently cross-linked acid acry-

lamide/acrylic copolymers as SAPs with the spherical particle diameters ranging
from 125 to 150 μm in the dry state was utilized. The dry-bulk density of SAPs
utilized was 850 kg/m3 . For improving the workability of concrete and reducing
consumption of tap water, PCA-I superplasticizer was utilized in the mixing process.

(2) Mixture proportions

Five concrete mixtures were designed with the same w/c ratio of 0.32 which repre-
sented the ratio of the water excluding the IC water to cement, as shown in Table
5.10 [22]. According to the different proportions of PP fibers and different curing
condition, the concrete mixtures were designated as Mixture BS00, BS03, BA03,
BS06, and BS09. The alphabet BS in the label of mixture meant that the specimen
reinforced with PP fibers was internally cured with SAPs. Meanwhile, the alphabet
BA in the label of mixture meant that the specimen reinforced with PP fibers was
cured traditionally. Mixture BS00, as the reference mixture, had the same curing
condition but did not contain the component of fibers, while other specimens were
diverse due to the proportion of fibers (0.3%, 0.6%, and 0.9% by volume fraction of
concrete for Mixture BS03, BS06, and BS09, respectively).
2. Free shrinkage of concrete

The free shrinkage (εsh ) was calculated from the net age considered equivalent to the
elapsed time after the age of 1.5 d, as shown in Fig. 5.33. The absolute value of free
shrinkage in the concrete mixture was 170, 142, 121, and 116 με for Mixture BS00,
BS03, BS06, and BS09 at the net age of 3 d, respectively. Compared with the reference
Mixture BS00, the absolute value of free shrinkage in Mixture BS03, BS06, and BS09
decreased by 16.5%, 28.8%, and 31.8%, respectively. This showed that the increasing
addition of PP fibers could effectually make the early-age free shrinkage of concrete
decrease to a certain extent when PP fibers were less than or equal to 0.9%. The
result in [65] shows that PP fibers are highly effective in controlling shrinkage. The
results in [66, 67] shows that when the PP fiber addition increases from 0.1 to 0.5%,
the shrinkage decreases by approximately 15%. Meanwhile, the result in [68] shows

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 253

Table 5.10 Mix proportions of concrete

Mixture composition (kg/m3 ) Concrete mixtures
BA03 BS00 BS03 BS06 BS09
Water 150 150 150 150 150
Cement 468.75 468.75 468.75 468.75 468.75
Fine aggregate (sand) 657.90 657.90 657.90 657.90 657.90
Coarse aggregate 1073.4 1073.4 1073.4 1073.4 1073.4
SAP 0 2917 2917 2917 2917
IC water 0 29.167 29.167 29.167 29.167
Superplasticizer 4.69 4.69 4.69 4.69 4.69
Fiber 2.76 0 2.76 5.52 8.25

that the incorporation of fibers reduces the shrinkage evidently. The transfer and loss
of internal moisture caused the shrinkage of concrete. Moreover, the addition of PP
fibers would hinder the sinking of aggregates. Fibers not only improved the uniform
distribution of concrete, but also blocked the channel of water overflow, as reported
in [69]. When PP fibers were mixed into the concrete, a complex three-dimensional
random system was formed, and water formed an adsorption film on the surface of PP
fibers with the hardening of concrete, which hindered the passage of water overflow
and slowed down the water loss. Meanwhile, large diameter holes will be formed in
the concrete with the addition of PP fibers. The number of capillary holes increased
with the increase of fiber proportion, which would result in reducing the capillary
pore pressure and water loss in concrete, as reported in [69]. Therefore, the addition
of PP fibers could improve the uniformity and microstructure of concrete, so as to
restrain shrinkage.

Fig. 5.33 Relationship

between free shrinkage and
net age of ICC with PP fibers

@seismicisolation
@seismicisolation
254 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Under the same condition of external environment and ring size, the free shrinkage
of fiber reinforce concrete internally cured with SAPs was smaller than that of corre-
sponding traditionally cured fiber reinforce concrete, as shown in Fig. 5.33. The
absolute value of free shrinkage in fiber reinforce concrete with traditional curing
and IC at the net age of 6.6 d was 297 and 190 με for Mixture BA03 and BS03,
respectively. Compared with BA03, the addition of SAPs utilized to IC made the
free shrinkage in BS03 decrease by 36.0%. In the scanning electron microscopy test
conducted in [70], SAPs can produce holes in the microstructure of concrete while
releasing water. Some of the cement hydration products will enter the holes later to
make space for the cement hydration reaction, while the unfilled holes make up for
the shrinkage of concrete. SAPs release the inner curing water into the surrounding
paste, which effectually alleviate the phenomenon of self-drying, thus reducing the
autogenous shrinkage in concrete. The result in [71] shows that the autogenous
shrinkage in concrete can be significantly decreased by the addition of SAPs and
additional water.
3. Strain in steel ring
The strain measured could indirectly reflect the shrinkage of concrete under the
restraint imposed by steel ring and the average strain outputted by static resistance
strain gauge was conducted as the test result, as shown in Fig. 5.34. In general,
strain inevitably increased with the growth of age under the identical curing and
geometric dimensions condition. However, the incremental strain decreased with the
increase in fiber proportion at the same age. For instance, the absolute value of strain
measured at 3 d was 45, 36, 32, and 29 με. Compared with the reference Mixture
BS00, the absolute value of strain for Mixture BS03, BS06, and BS09 decreased
by 20%, 28.9%, and 35.6% with the increasing fiber addition from 0% to 0.3%,
0.6%, and 0.9%, respectively. The result in [67] shows that the strain in concrete
will decrease when the fiber proportion is below 0.06%. The absolute value of strain
in steel ring decreases from 30 to 25 με in 10 d and the absolute value of strain
of concrete decreases by 16.7% [68]. When the concrete mixtures with the same
PP fiber proportion (0.3%) under two different curing conditions, the absolute value
of strain in steel ring at the net age of 4.2 d was 61 and 42 με for Mixture BA03
and BS03, respectively. The strain in BS03 which was internally cured with SAPs
decreased by 31.1% when compared with that in BA03.
Strain changed rapidly at early age, but the curve gradually began to flatten out.
Tensile stress in concrete ring is induced from the restraint imposed by steel ring. The
outer steel surface is deformed by the pressure of concrete ring, so the compressive
strain increased when the shrinkage of concrete developed with the elapsed time. A
slow rate of strain development in steel ring provides the time for increasing stress
relaxation and tensile strength, thus contributing to decreasing the cracking potential
[72].
4. Residual stress of concrete
Figure 5.35 shows the relationship between the actual maximum residual tensile
stress and age.

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 255

Fig. 5.34 Relationship

between strain in steel ring
and age

Fig. 5.35 Maximum 5.0 BS00

Maximum residual stress (MPa)

residual stress development BS03

4.5 BS06
in concrete ring specimens
4.0 BS09
BA03
3.5
3.0
2.5
2.0
Age at cracking of BA03
1.5 (12.3 days)
Age at cracking of BS00
1.0 (15.5 days)

0.5 Age at cracking of BS03

(17.7 d)
0.0
0 3 6 9 12 15 18 21
Age(d

Under the same external environment and ring size, when the proportion of
PP fibers was less than 0.9%, increasing the proportion of fibers can prominently
decrease the residual stress of concrete ring. The residual stress of concrete at 3 d
was 2.54, 2.05, 1.82, and 1.65 MPa for Mixture BS00, BS03, BS06, and BS09, which
decreased by 19.3%, 28.3%, and 35.0% with the increase of fiber proportion from
0% to 0.3%, 0.6%, and 0.9%, respectively. The residual stress at 7 d was 3.42, 2.98,
2.58, and 2.18 MPa. The value compared with the reference Mixture BS00 decreased
by 12.9%, 24.6%, and 36.3% for Mixture BS03, BS06, and BS09, respectively. The
results derived were consistent with the investigations that showed the decrease of
residual stress when the fiber proportion was lower than 0.06%, as reported in [67].
When the concrete with the same fiber proportion (0.3%) under two different curing
conditions, residual stress at the net age of 3 d was 2.74 and 2.05 MPa for Mixture
BA03 and BS03, respectively. The residual stress in BS03 which was internally cured
with SAPs decreased by 25.2% when compared with that in BA03.

@seismicisolation
@seismicisolation
256 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.36 Relationship

between elastic modulus and
age of ICC with PP fibers

5. Stress relaxation of concrete

For Mixture BS00, BS03, BS06, BS09, and BA03, the compressive elastic modulus
of concrete at 28 d was 42.5, 45.3, 46.6, 47.1, and 47.4 GPa, respectively. The
compressive elastic modulus of concrete was the variable over time and calculated
in accordance with CEB 90 Code Model given by Eq. (5.18), as reported in [41].
Figure 5.36 shows the curves of elastic modulus and age for four concrete mixtures
with the different proportions of PP fibers from 0 to 0.9%. Therefore, the maximum
theoretical elastic stress could be yielded.
From the above equation, it can be seen that the relaxation stress was affected by
the strain measured in steel ring, the free shrinkage, and the time-dependent elastic
modulus of concrete. The theoretical elastic stress and residual stress decreased
together, and the theoretical elastic stress decreased slower than the corresponding
residual stress when the fiber proportion increased, which led to the improvement of
stress relaxation with the increase of the proportion of PP fibers, as shown in Fig. 5.37.
Under the identical curing condition, the theoretical elastic stress at the net age of
4 d was 3.41, 2.86, 2.65, and 2.43 MPa for Mixture BS00, BS03, BS06, and BS09,
respectively. The corresponding residual stress was 2.31, 1.67, 1.39, and 1.09 MPa,
respectively. Compared with the Mixture BS00, increasing the proportion of fibers
from 0% to 0.3%, 0.6%, and 0.9% made the theoretical elastic stress or residual
stress decrease by 16.05%, 22.21%, and 28.63% or 27.93%, 39.67%, and 52.75%,
respectively. When the concrete with the same fiber proportion (0.3%) under two
different curing conditions, the theoretical elastic stress in concrete with traditional
curing and IC at the net age of 6.6 d was 5.87 and 3.68 MPa for Mixture BA03 and
BS03, respectively. The theoretical elastic stress in BS03 which was internally cured
with SAPs decreased by 37.3% when compared with that in BA03. Results indicated
that increasing fiber proportion could decrease the theoretical elastic stress obviously
and SAPs had a significant influence on decreasing the theoretical elastic stress in
fiber reinforce concrete.

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 257

7 7
Theoretical elastic stress Mixture BS00 Mixture BS06
Theoretical elastic stress
6 Residual stress 6 Residual stress
5 5

Stress (MPa)
Stress (MPa)

4 4

3 3

2 Net age of cracking of BS00 2

(14.0 d)

1 1

0 0
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
Net age (d) Net age (d)

(a) BS00 (c) BS06

7 7
Theoretical elastic stress Mixture BS03 Theoretical elastic stress Mixture BS09
6 Residual stress 6 Residual stress
5 5
Stress (MPa)

Stress (MPa)

4 4
3 3
2 2
Net age of cracking of BS03
(16.2 days)
1 1
0 0
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
Net age (d) Net age (d)
(b) BS03 (d) BS09
7
Mixture BA03
6

5
Theoretical elastic stress
Stress (MPa)

4 Residual stress

2
Net age of cracking of BA03
1 (10.8 d)

0
0 2 4 6 8 10 12 14 16 18 20
Net age (d)

(e) BA03

Fig. 5.37 Comparison on stress and net age of ICC with PP fibers

@seismicisolation
@seismicisolation
258 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

The phenomenon that relaxed stress increased with the age demonstrated that the
stress relaxation in fiber reinforce concrete was developing under the continuous
restrained condition at early age, as shown in Fig. 5.38. From the analysis of two
kinds of stress above, the relaxed stress increased when the proportion of PP fibers
proportion increased from 0 to 0.9% with the addition of SAPs utilized for IC. At
the net age of 4 d, the relaxed stress in Mixture BS00, BS03, BS06, and BS09 was
1.10, 1.20, 1.26, and 1.34 MPa, respectively. Compared with the reference mixture
without PP fibes, the relaxation stress of concrete with 0.3%, 0.6%, and 0.9% PP
fibers increased by 9.1%, 14.5%, and 21.8%, respectively. The addition of SAPs
decreased the theoretical elastic stress and residual stress in fiber reinforced concrete.
Furthermore, under the combined action with PP fibers, the reduction of theoretical
elastic stress by SAPs was larger than that of residual stress, which directly led to
the reduction of relaxed stress. The relaxed stress at the net age of 0.7 d was 2.06
and 0.98 MPa for Mixture BA03 and BS03, respectively. The relaxed stress in BS03
which was internally cured with SAPs decreased by 52.4% when compared with that
in BA03. The larger the creep was, the larger the relaxation stress was, as reported
in [43]. The analysis of relaxation stress could be transformed into the analysis of
creep. When the elastic modulus of fiber (8.20 GPa) was less than that of concrete
(47.72 GPa), the addition of fiber was not conducive to the resistance of concrete
to creep. At the same time, the incorporation of fiber would cause some defects in
concrete, including the defects and uneven distribution at the interface between fiber
and cement paste, as reported in [67]. Once the concrete bore the load, it would cause
the migration and exudation of the adsorbed water and interlayer water of the gel,
which weakened the ability of concrete to resist creep as reported in [69].
The change of stress relaxation degree with age at early age of internally cured
concrete was roughly separated into two stages: in the first stage (rapid decline stage),
which occurred at 2 d (net age), the degree of stress relaxation was high but rapidly
declines, as shown in Fig. 5.39.

Fig. 5.38 Estimated relaxed 4.0

BS00
stress in concrete ring
3.5 BS03
specimens with PP fibers BS06
Relaxed stress (MPa)

3.0 BS09
BA03
2.5
2.0
1.5
1.0
0.5
0.0
0 2 4 6 8 10
Net age (d)

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 259

1.0

Ratio of relaxed to theoretical elastic stress

Fig. 5.39 Relationship
between ratios of relaxed to 0.9
theoretical elastic stress and
0.8
net age of ICC with PP fibers
0.7
0.6
0.5
0.4
0.3 BS00
BS03
0.2 BS06
0.1 BS09
BA03
0.0
1 2 3 4 5 6 7 8 9
Net age (d)

Although SAPs employed to curing had some negative influences on stress relax-
ation, with the increase of the proportion of PP fibers, the stress relaxation of concrete
ring with IC still had been improved to the varying degree generally, which showed
that the addition of PP fibers played a certain role in increasing the stress relaxation,
alleviating the tensile stress, and decreasing the cracking potential at early age.
6. Stress rate of concrete

Figure 5.40 shows the fitting curves of net strain. Figure 5.41 shows an obvious rule
that increasing the PP fibers proportion can make the stress rate of concrete decrease.
When mixtures were at 7 d (te ), stress rate for Mixture BS00, BS03, BS06, and
BS09 was 0.260, 0.236, 0.209, and 0.181 MPa/d, respectively. Compared with the
reference mixture without PP fibers, the stress rate of mixtures with 0.3%, 0.6%, and
0.9% PP fibers decreased by 9.2%, 19.6%, and 30.4%, respectively. The addition of

Fig. 5.40 Relationship

between net strain and 0
square root of net age of ICC
with PP fibers -20
Net strain (με)

Test data of BS00

-40
Test data of BS03
Test data of BS06
Test data of BS09
-60 Test data of BA03
Fitting curve of BS00
Fitting curve of BS03
-80 Fitting curve of BS06
Fitting curve of BS09
Fitting curve of BA03
-100
0 1 2 3 4
1/2
Square root of net age (d )

@seismicisolation
@seismicisolation
260 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

Fig. 5.41 Development of 1.5 BA03

stress rate of ICC with PP
BS00
fibers
BS03

Stress rate (MPa/d)

1.2 BS06
BS09
0.9

0.6

0.3

0 1 2 3 4 5 6 7 8
Net age (d)

PP fibers significantly decreased the stress rate of ICC. The stress rate in mixtures
with traditional curing and IC at the net age of 7 d was 0.291 and 0.236 MPa/d for
Mixture BA03 and BS03, respectively. Compared with Mixture BA03, the stress
rate of Mixture BS03 which was cured internally with SAPs decreased by 18.8%.
The stress rate classification method was utilized to assess the cracking potential of
mixtures, as shown in Table 5.1. The cracking time of Mixture BS00 was the net age
of 14.0 d and that of Mixture BS03 was the net age of 16.2 d, while the cracking
phenomena of BS06 and BS09 were not observed within 28 d test time. Mixture
BA03 which was cured traditionally cracked at the net age of 10.8 d. According
to the stress rate classification method, the Mixture BS00 belonged to “Moderated-
High” zone, and Mixture BS03 was between the “Moderated-High” zone and the
“Moderated-Low” zone. Both Mixture BS06 and BS09 belonged to the “Low” zone.
Therefore, according to the results of stress rate, concrete with higher fiber proportion
has the lower cracking potential. The addition of SAPs could significantly decrease
the stress rate and delay the appearance time of cracks in concrete.

7. Cracking potential of concrete

The splitting tensile strength at the age of 28 d ( f sp,28d ) obtained by the splitting tensile
strength test was 4.5, 4.9, 5.3, 5.6, and 5.2 MPa for Mixture BS00, BS03, BS06,
BS09, and BA03, respectively. Splitting tensile stress at any age can be expressed by
Eq. (5.23) [63].
[ ( / )]
28
f sp (t) = f sp,28d · exp S1 1 − (5.23)
t

where f sp (t) represents the splitting tensile strength at the age of t d, in MPa; f sp,28d
represents the splitting tensile strength at 28 d, in MPa; and the regression coefficients
S1 for Mixture BS00, BS03, BS06, BS09, and BA03 was 0.029, 0.041, 0.050, 0.044,

@seismicisolation
@seismicisolation
5.4 Influential Factors of the Cracking Resistance of Early-Age Internally … 261

Fig. 5.42 Relationship 1.0

between cracking potential
and age of ICC with PP fibers 0.8

Cracking potential
0.6

0.4
Age at cracking of BA03
(12.3 days) BS00
0.2 Age at cracking of BS00 BS03
(15.5 days)
Age at cracking of BS03
BS06
(17.7 days) BS09
0.0 BA03

0 3 6 9 12 15 18 21
Age (d)

and 0.058, respectively. The coefficient of determination R 2 for Mixture BS00, BS03,
BS06, BS09, and BA03 was 0.968, 0.998, 0.978, and 0.961, respectively.
In order to investigate the influence of the different proportion of PP fibers on the
cracking potential in early-age concrete thoroughly, Fig. 5.42 intuitively shows the
cracking potential calculated by Eq. (5.17).
Increasing the proportion of PP fibers would decrease the cracking potential signif-
icantly. The value of cracking potential of Mixture BS00, BS03, BS06, and BS09 rein-
forced with 0.3%, 0.6%, and 0.9% PP fibers at the age of 2 d was 0.433, 0.352, 0.319,
and 0.243, which decreased by 18.7%, 26.3%, and 43.9%, respectively. Adding PP
fibers in concrete will not only decrease the residual stress, but also increase the
splitting tensile strength. Therefore, adding PP fibers can effectually decrease the
cracking potential of concrete. The compressive strength of Mixture BA03 was 39.0,
47.6, 50.2, and 56.1 MPa at 3, 7, 14, and 28 d, respectively. Meanwhile, the compres-
sive strength of Mixture BS03 was 37.3, 45.1, 47.8, and 53.2 MPa at 3, 7, 14, and
28 d, respectively. Compared with the traditional curing, IC made the compressive
strength decrease by 4.36%, 5.25%, 4.78%, and 5.17% at 3, 7, 14, and 28 d, respec-
tively. The compressive strength of concrete with addition of SAPs decreased when
the IC water was added as the extra water. The reason for this phenomenon was that
the additional amount of IC water made the w/c ratio increase and thus decreasing
the compressive strength of mixtures [42]. Although SPAs decreased the compres-
sive strength to some extent, the addition of SAPs decreased the cracking potential
due to the increase of tensile strength. At 5.3 d, the value of cracking potential in
Mixture BA03 and BS03 was 0.770 and 0.585, respectively. Compared with BA03,
the value of cracking potential of BS03 decreased by 24.0%. The reason for this
test result was that the restraint strain of steel ring in concrete could be effectually
decreased by adding SAPs, and the residual stress of concrete could also be decreased
accordingly. Therefore, utilizing SAPs for IC could decrease the early-age cracking
potential significantly.

@seismicisolation
@seismicisolation
262 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

The value of cracking potential of concrete can reach 1 in theory, but the result
in [28] show that cracking occurs when the value of the cracking potential caused
by shrinkage does not reach the theoretical value. The value of cracking potential
of BS00 was 0.96, and that of BS03 was 0.93 when it actually cracked. Although
some deviations exist between theory and practice, it will be generally considered
that the cracking potential can well reflect the relationship between cracking and
age. Stress relaxation will cause micro-cracks in the structure, and the additional
effects such as stress concentration caused by these micro-cracks will then promote
the cracking of concrete. Usually, when the internal stress reaches 60–85% of the
tensile resistance, creep failure is likely to occur [73]. The result in [65] shows that
the cracking potential would decrease when the incorporation of fiber increases from
0.1 to 0.7%. The addition of fiber could reduce the shrinkage in concrete and led to
the corresponding reduction of tensile stress, as reported in [74]. When the tensile
strength of concrete increased with the increase of fiber content, the initial crack time
of concrete would be delayed, as reported in [28]. Meanwhile, from the micro point
of view, there were inevitably micro cracks in the concrete. The three-dimensional
random distribution system could effectively prevent the development of cracks.
When the length of microcracks was larger than the spacing of the fibers, the fibers
would play a role in transferring the load across the cracks, making the stress field
in the concrete more continuous and uniform, so as to restrict the extension of the
cracks, as reported in [75]. When the length of the crack was less than the spacing of
fiber, the existence of the fiber would change the direction of the crack to form the
subtler crack and increased the energy needed for crack propagation, as reported in
[76]. Based on the mechanism and the results of relevant literatures, it was reasonable
that the incorporation of PP fibers could effectively reduce the cracking potential.

5.4.3 Summary

The influence of GGBFS and PP fibers on the free shrinkage, residual tensile
stress, stress relaxation, cracking potential, and stress rate of ICC was investigated.
Conclusions of the influences of GGBFS and PP fibers on ICC were drawn on the
basis of experimental data analysis. The following conclusions were summarized in
accordance with the obtained results.
(1) The absolute value of free shrinkage or relaxed stress of ICC with SAPs
decreased with the increase of GGBFS content. Increasing the proportion of PP
fibers caused the decrease of free shrinkage in fiber reinforce concrete internally
cured with SAPs. Increasing the proportion of PP fibers caused the decrease of
the strain in steel ring.
(2) The maximum residual tensile stress decreased with the increase of GGBFS
content. T Increasing the proportion of PP fibers caused the decrease of residual
stress in fiber reinforce concrete internally cured with SAPs.

@seismicisolation
@seismicisolation
References 263

(3) The stress rate of ICC with SAPs decreased with the increase of GGBFS content.
The cracking potential of ICC with SAPs decreased with the increase of GGBFS
content, as determined from the stress rate. Increasing the proportion of PP fibers
made stress rate reduce in fiber reinforce concrete internally cured with SAPs.
According to the classification method of stress rate, Mixture BS00 belonged
to “Moderated-High” zone, and Mixture BS03 was between the “Moderated-
High” zone and the “Moderated-Low” zone. Both Mixture BS06 and BS09
belonged to the “Low” zone.
(4) The cracking potential parameter of ICC with SAPs decreased with the increase
of GGBFS content. Increasing the proportion of PP fibers caused the increase of
relaxed stress in fiber reinforce concrete internally cured with SAPs. Increasing
the proportion of PP fibers delayed the time when cracks occurred in concrete
and made the cracking potential of concrete decrease significantly.
(5) Increasing the proportion of PP fibers caused the increase of relaxed stress
in fiber reinforce concrete internally cured with SAPs. The relaxed stress of
Mixture BS00, BS03, BS06, and BS09 at the net age of 4 d was 1.10, 1.20,
1.26, and 1.34 MPa, which increased by 9.1%, 14.5%, and 21.8% with the
increase of fiber proportion, respectively. The ratio of relaxed stress to theoretical
elastic stress in concrete was 0.310, 0.415, 0.497, and 0.543, which increased
by 33.77%, 60.14%, and 75.02% when the PP fibers proportion increased from
0% to 0.3%, 0.6%, and 0.9%, respectively.
(6) SAPs utilized in IC had an effectual improvement on performance of concrete
reinforced with PP fibers. Although the addition of SAPs weakened the mechan-
ical properties of fiber reinforce concrete to some extent, SAPs could decrease
free shrinkage, strain, residual stress, theoretical elastic stress, cracking poten-
tial and stress rate of fiber reinforce concrete at early age. The addition of SAPs
may be considered as a momentous factor in the investigation on early-age
performance of concrete.

References

1. Lura P, Ye G, Cnudde V et al (2008) Preliminary results about 3D distribution of superabsorbent

polymers in mortars. In: Proceedings of the international conference on microstructure related
durability of cementitious composites, Nanjing, China, pp 13–15
2. Bentz DP, Snyder KA (1999) Protected paste volume in concrete: extension to internal curing
using saturated lightweight fine aggregate. Cem Concr Res 29(11):1863–1867
3. Messan A, Ienny P, Nectoux D (2011) Free and restrained early-age shrinkage of mortar:
influence of glass fiber, cellulose ether and EVA (ethylene-vinyl acetate). Cem Concr Compos
33(3):402–410
4. Shen DJ, Feng ZZ, Zhang TT et al (2022) Estimating stress relaxation and cracking potential
of high-strength concrete reinforced with polyvinyl alcohol fiber at early age. J Mater Civ Eng
34(9):1–14
5. Weiss WJ, Shah SP (2002) Restrained shrinkage cracking: the role of shrinkage reducing
admixtures and specimen geometry. Mater Struct/Materiaux et Constructions 34(246):85–91

@seismicisolation
@seismicisolation
264 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

6. Xia Q, Li H, Yao T et al (2017) Cracking behaviour of restrained cementitious materials with

expansive agent by comprehensive analysis of residual stress and acoustic emission signals.
Adv Cem Res 29(2):81–90
7. Shen DJ, Kang JC, Yi X et al (2019) Effect of double hooked-end steel fiber on early-age
cracking potential of high strength concrete in restrained ring specimens. Constr Build Mater
223:1095–1105
8. Shen D, Jiao Y, Liu C et al (2022) Effect of water-to-cement ratio on cracking resistance of
concrete at early age in restrained ring specimens. Can J Civ Eng 49(5):706–715
9. See HT, Attiogbe EK, Miltenberger MA (2003) Shrinkage cracking characteristics of concrete
using ring specimens. ACI Mater J 100(3):239–245
10. Shen DJ, Shi HF, Liu KQ et al (2018) Early-age residual stress and stress relaxation of high-
performance concrete containing fly ash. Mag Concr Res 70(14):726–738
11. Kwak HG, Ha SJ, Kim JK (2006) Non-structural cracking in RC walls: part I. Finite element
formulation. Cem Concr Res 36(4):749–760
12. Carlson RW (1942) Cracking of concrete. Boston Soc Civ Eng 29(2):98–109
13. Kanavaris F, Azenha M, Soutsos M et al (2019) Assessment of behaviour and cracking suscep-
tibility of cementitious systems under restrained conditions through ring tests: a critical review.
Cem Concr Compos 95:137–153
14. China Civil Engineering Society (2005) Guide to durability design and construction of concrete
structures: CCES 01–2004. China Architecture & Building Press, Beijing (in Chinese)
15. AASHTO (1996) Standard practice for estimating the cracking tendency of concrete. AASHTO
Designation, Washington
16. STM International (2019) Standard test method for autogenous strain of cement paste and
mortar: ASTM C1581/C1581M-18a. ASTM International, West Conshohocken
17. Radlinska A (2006) The ring test: a review of recent developments. In: International
RILEM conference on volume changes of hardening concrete: testing and mitigation. RILEM
Publications SARL, Lyngby, Denmark, pp 205–214
18. Shen DJ, Shi HF, Tang XJ et al (2016) Effect of internal curing with super absorbent polymers
on residual stress development and stress relaxation in restrained concrete ring specimens.
Constr Build Mater 120:309–320
19. Shen DJ, Feng ZZ, Zhu PF et al (2020) Effect of pre-wetted lightweight aggregates on residual
stress development and stress relaxation in restrained concrete ring specimens. Constr Build
Mater 258:119151
20. Tang XJ (2015) Research on early-age cracking resistance of internally cured concrete. Hohai
University, Nanjing (in Chinese)
21. Shen DJ, Jiao Y, Kang JJ et al (2020) Influence of ground granulated blast furnace slag on
early-age cracking potential of internally cured high performance concrete. Constr Build Mater
233:117083
22. Shen DJ, Feng ZZ, Kang JC et al (2020) Effect of Barchip fiber on stress relaxation and
cracking potential of concrete internally cured with super absorbent polymers. Constr Build
Mater 249:118392
23. Zhu SS (2015) Research on early-age cracking resistance of internally cured concrete. Hohai
University, Nanjing )
24. Shi HF (2017) Research on early age stress relaxation of concrete reinforced with plastic fibers.
Hohai University, Nanjing (in Chinese)
25. Ministry of Housing and Urban Rural Development of the People’s Republic of China (2009)
Standard for test methods of long-term performance and durability of ordinary concrete: GB/T
50082-2009. China Architecture & Building Press, Beijing. (in Chinese)
26. Shen DJ, Wang XD, Cheng DB et al (2016) Effect of internal curing with super absorbent
polymers on autogenous shrinkage of concrete at early age. Constr Build Mater 106(1):512–522
27. Jiang CH, Yang Y, Wang Y et al (2014) Autogenous shrinkage of high performance concrete
containing mineral admixtures under different curing temperatures. Constr Build Mater
61:260–269

@seismicisolation
@seismicisolation
References 265

28. Hossain AB, Weiss J (2004) Assessing residual stress development and stress relaxation in
restrained concrete ring specimens. Cem Concr Compos 26(5):531–540
29. Gao Y, Zhang J, Han P (2013) Determination of stress relaxation parameters of concrete in
tension at early-age by ring test. Constr Build Mater 41:152–164
30. Ministry of Housing and Urban Rural Development of the People’s Republic of China (2019)
Standard for test methods of concrete physical and mechanical properties: GB/T 50081-2019.
China Architecture & Building Press, Beijing. (in Chinese)
31. Hossain AB, Weiss J (2006) The role of specimen geometry and boundary conditions on stress
development and cracking in the restrained ring test. Cem Concr Res 36(1):189–199
32. Zhao QX, Yu JC, Geng GQ et al (2016) Effect of fiber types on creep behavior of concrete.
Constr Build Mater 105:416–422
33. Zhang J (2003) Modeling of the influence of fibers on creep of fiber reinforced cementitious
composite. Compos Sci Technol 63(13):1877–1884
34. Weiss WJ, Yang W, Shah SP (2000) Influence of specimen size/geometry on shrinkage cracking
of rings. J Eng Mech 126(1):93–101
35. General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China (2018) Common Portland cement: GB 175-2007/XG1-2018. Standard Press
of China, Beijing. (in Chinese)
36. Moon JH, Weiss J (2006) Estimating residual stress in the restrained ring test under
circumferential drying. Cem Concr Compos 28(5):486–496
37. Henkensiefken R, Bentz D, Nantung T et al (2009) Volume change and cracking in inter-
nally cured mixtures made with saturated lightweight aggregate under sealed and unsealed
conditions. Cem Concr Compos 31(7):427–437
38. Schlitter J (2010) New methods to quantify the cracking performance of cementitious systems
made with internal curing. Purdue University, Indiana
39. Cusson D, Hoogeveen T (2005) Internally-cured high-performance concrete under restrained
shrinkage and creep. Concrete 12–14:579–584
40. Suzuki M, Seddik Meddah M, Sato R (2009) Use of porous ceramic waste aggregates for
internal curing of high-performance concrete. Cem Concr Res 39(5):373–381
41. fib (2013) fib model code for concrete structures 2010: fib model code 2010. Ernst & Sohn,
Berlin
42. Craeye B, Geirnaert M, Schutter GD (2011) Super absorbing polymers as an internal curing
agent for mitigation of early-age cracking of high-performance concrete bridge decks. Constr
Build Mater 25(1):1–13
43. Brooks JJ, Neville AM (1976) Relaxation of stress in concrete and its relation to creep. ACI
Struct J 73(4):227–232
44. Zou DH, Weiss WJ (2014) Early age cracking behavior of internally cured mortar restrained
by dual rings with different thickness. Constr Build Mater 66:146–153
45. Hansen EA (1991) Time dependent tensile fracture of concrete. The Norwegian University of
Science and Technology, Trondheim
46. Kanstad T, Hammer TA, Bjontegaard Ø et al (2003) Mechanical properties of young concrete:
part I: experimental results related to test methods and temperature effects. Mater Struct
36(258):218–225
47. Wang KJ, Jansen DC, Shah SP et al (1997) Permeability study of cracked concrete. Cem Concr
Res 27(3):381–393
48. Hossain AB, Pease B, Weiss J (2003) Quantifying early-age stress development and cracking
in low water-to-cement concrete: restrained ring test with acoustic emission. In: Transportation
research record. Purdue University, West Lafayette, pp 24–32
49. General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China (2010) Lightweight aggregates and its test methods-part2: test methods for
lightweight aggregates: GB/T 17431.2-2010. Standard Press of China, Beijing. (in Chinese)
50. Assmann A, Reinhardt HW (2014) Tensile creep and shrinkage of SAP modified concrete.
Cem Concr Res 58:179–185

@seismicisolation
@seismicisolation
266 5 Cracking Resistance of Internally Cured Concrete by Restrained Ring …

51. Yoo DY, Park JJ, Kim SW et al (2013) Early age setting, shrinkage and tensile characteristics
of ultra high performance fiber reinforced concrete. Constr Build Mater 41:427–438
52. Kong D, Lei T, Zheng JJ et al (2010) Effect and mechanism of surface-coating pozzalanics mate-
rials around aggregate on properties and ITZ microstructure of recycled aggregate concrete.
Constr Build Mater 24(5):701–708
53. Ma XW, Yuan Q, Liu JH et al (2019) Effect of water absorption of SAP on the rheological
properties of cement-based materials with ultra-low w/b ratio. Constr Build Mater 195:66–74
54. Duran Atiş C, Bilim C (2007) Wet and dry cured compressive strength of concrete containing
ground granulated blast-furnace slag. Build Environ 42(8):3060–3065
55. Güneyisi E, Gesolu M, Özbay E (2010) Strength and drying shrinkage properties of self-
compacting concretes incorporating multi-system blended mineral admixtures. Constr Build
Mater 24(10):1878–1887
56. Li J, Yao Y (2001) A study on creep and drying shrinkage of high performance concrete. Cem
Concr Res 31(8):1203–1206
57. Mokarem DW, Weyers RE, Lane DS (2005) Development of a shrinkage performance specifica-
tions and prediction model analysis for supplemental cementitious material concrete mixtures.
Cem Concr Res 35(5):918–925
58. Shen DJ, Liu KQ, Wen CY et al (2019) Early-age cracking resistance of ground granulated
blast furnace slag concrete. Constr Build Mater 222:278–287
59. Altoubat S, Badran D, Junaid MT et al (2016) Restrained shrinkage behavior of self-compacting
concrete containing ground-granulated blast-furnace slag. Constr Build Mater 129:98–105
60. Wei Y, Hansen W (2013) Early-age strain-stress relationship and cracking behavior of slag
cement mixtures subject to constant uniaxial restraint. Constr Build Mater 49:635–642
61. Khatri RP, Sirivivatnanon V (1995) Effect of different supplementary cementitious materials
on mechanical. Cem Concr Res 25(1):209–220
62. Shen DJ, Shi X, Zhang H et al (2016) Experimental study of early-age bond behavior between
high strength concrete and steel bars using a pull-out test. Constr Build Mater 113:653–663
63. CEN (2004) Eurocode 2: design of concrete structures: part 1-1: general rules and rules for
buildings: BS EN 1992-1-1. CEN, Brussels
64. Darquennes A, Staquet S, Delplancke-Ogletree MP et al (2011) Effect of autogenous
deformation on the cracking risk of slag cement concretes. Cem Concr Compos 33(3):368–379
65. Banthia N, Gupta R (2006) Influence of polypropylene fiber geometry on plastic shrinkage
cracking in concrete. Cem Concr Res 36(7):1263–1267
66. Mesbah HA, Buyle-Bodin F (1999) Efficiency of polypropylene and metallic fibres on control
of shrinkage and cracking of recycled aggregate mortars. Constr Build Mater 13(8):103–110
67. Passuello A, Moriconi G, Shah SP (2009) Cracking behavior of concrete with shrinkage
reducing admixtures and PVA fibers. Cem Concr Compos 31(10):699–704
68. Shah HR, Weiss J (2006) Quantifying shrinkage cracking in fiber reinforced concrete using
the ring test. Mater Struct 39(9):887–899
69. Wang KJ, Shah SP, Phuaksuk P (2001) Plastic shrinkage cracking in concrete materials—
influence of fly ash and fibers. ACI Mater J 98(6):458–464
70. Justs J, Wyrzykowski M, Bajare D et al (2015) Internal curing by superabsorbent polymers in
ultra-high performance concrete. Cem Concr Res 76:82–90
71. Jensen OM, Hansen PF (2001) Water-entrained cement-based materials—I. Principles and
theoretical background. Cem Concr Res 31(4):647–654
72. Schlitter JL, Senter AH, Bentz DP et al (2010) A dual concentric ring test for evaluating residual
stress development due to restrained volume change. J ASTM Int 7(9):103–110
73. Wittmann FH, Roelfstra PE, Mihashi H et al (1987) Influence of age of loading, water-cement
ratio and rate of loading on fracture energy of concrete. Mater Struct 20(2):103–110
74. Gao SL, Wang Z, Wang WC et al (2018) Effect of shrinkage-reducing admixture and expansive
agent on mechanical properties and drying shrinkage of engineered cementitious composite
(ECC). Constr Build Mater 179:172–185

@seismicisolation
@seismicisolation
References 267

75. Soroushian P (1997) Secondary reinforcement-adding cellulose fibers. Concr Int 19(6):28–34
76. Romualdi JP, Mandel JA (1964) Tensile strength of concrete affected by unigormlydistributed
and closely spaced short lengths of wire reinforcement. ACI J Proc 61(6):657–672

@seismicisolation
@seismicisolation
Chapter 6
Cracking Resistance of Internally Cured
Concrete Under Uniaxial Restrained
Condition at Early-Age

Contents

6.1 Methods and Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270

6.1.1 Test Details of Temperature Stress Test Machine . . . . . . . . . . . . . . . . . . . . . . . . . . 270
6.1.2 Testing Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
6.1.3 Calculation of Equivalent Age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 272
6.1.4 Calculation of Axial Tensile Strength and Elastic Modulus . . . . . . . . . . . . . . . . . 273
6.1.5 Calculation of Temperature Rise and Drop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
6.1.6 Calculation of Autogenous Shrinkage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
6.1.7 Calculation of Creep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
6.1.8 Calculation of Ratio of Cracking Stress to Axial Tensile Strength . . . . . . . . . . . . 277
6.1.9 Calculation of Cracking Reserve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
6.2 Integrated Criterion of Cracking Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
6.3 Cracking Resistance of Concrete Internally Cured with SAPs Under Uniaxial Restrained
Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
6.3.1 Mixture Proportions and Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
6.3.2 Temperature of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
6.3.3 Autogenous Shrinkage and Restrained Stress of Concrete with SAPs . . . . . . . . . 280
6.3.4 Creep of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
6.3.5 Cracking Resistance of Concrete with SAPs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
6.3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
6.4 Cracking Resistance of Concrete Internally Cured with LWAs Under Uniaxial
Restrained Condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
6.4.1 Mixture Proportions and Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
6.4.2 Temperature of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
6.4.3 Free Deformation of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
6.4.4 Restrained Stress of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
6.4.5 Creep of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
6.4.6 Cracking Resistance of Concrete with LWAs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
6.4.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete . . . . . . . . . . . . . 304
6.5.1 Addition of Ground Granulated Blast Furnace Slag . . . . . . . . . . . . . . . . . . . . . . . . 304
6.5.2 w/c Ratios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
6.5.3 Addition of Polypropylene Fiber with Different Lengths . . . . . . . . . . . . . . . . . . . 325
6.5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341

© Science Press 2023 269

D. Shen, Cracking Control on Early-Age Concrete Through Internal Curing,
https://doi.org/10.1007/978-981-19-8398-6_6
@seismicisolation
@seismicisolation
270 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

In the real environment, as the temperature of early-age concrete first increases due to
hydration heat and decreases, rapid thermal shrinkage may lead to cracking. In addi-
tion, in concrete with low w/c ratio, hydration of cement may cause self-desiccation
and autogenous shrinkage. The cracking resistance of concrete in the real environ-
ment is related to the temperature history, autogenous shrinkage, creep, and restraint
condition. For many practical concrete structures, such as mass concrete, there is no
significant moisture exchange between concrete and environment, and temperature
rise is more obvious. Temperature process [1], autogenous shrinkage [2], restrained
stress [3], and tensile creep [4] of concrete under uniaxial restrained condition all can
be investigated by utilizing TSTM. In addition, TSTM provides different degrees of
restraint and an adiabatic condition to concrete specimens for better simulating the
actual working conditions of the mass concrete. Uniaxial constant restraint degree can
be achieved by free and restrained shrinkage test simultaneously using TSTM. Many
single criteria, such as temperature drop, cracking stress, cracking time, and ratio of
cracking stress to axial tensile strength are used to evaluate the cracking behavior of
concrete. However, integrated criterion considering the influence of different factors
is limited. Therefore, investigations on the criterion of cracking resistance and early-
age cracking resistance of ICC under uniaxial restrained condition using TSTM are
needed.
In this chapter, the temperature process, autogenous shrinkage, restrained stress,
tensile creep, and cracking resistance of ICC at early-age are simultaneously studied
using TSTM. The integrated criterion of cracking resistance is proposed to evaluate
the early-age cracking behavior of ICC [5]. The early-age behavior and cracking
resistance of ICC containing SAPs and LWAs are also presented. Moreover, the
investigations on the cracking resistance of ICC at early-age considering the influence
of GGBFS, w/c ratio, and PP fiber length are conducted. Furthermore, the prediction
model for the tensile creep and autogenous shrinkage of ICC is proposed based on
the results [1, 3, 5–10].

6.1 Methods and Calculation

6.1.1 Test Details of Temperature Stress Test Machine

Restrained and free shrinkage tests of sealed specimens were conducted with TSTM,
which was developed combining the principles of Kovler [11] as uniaxial restrained
shrinkage closed-loop computer-controlled setup, and of Springenschmid [12] as
TSTM machine, as shown in Fig. 6.1. There were two molds in one TSTM system
called free and restrained specimens for calculating the creep and elastic deformation.
In order to obtain the temperature history, the temperature sensors were anchored in
concrete. Adiabatic condition was realized by circulating a specific liquid, of which
the temperature was controlled to be the same as the specimen, along the sides of
molds.

@seismicisolation
@seismicisolation
6.1 Methods and Calculation 271

Fig. 6.1 TSTM used in experiment

The restrained mold was a horizontal steel frame. One end was a fixed steel head,
the other was a moving end, and the center was a unidimensional part. The moving
head was connected to a stepper motor. The free mold had same geometry as the
restrained mold. However, both of the ends of free mold were free. The restrained
stress caused by restrained deformation was also obtained in TSTM system. Also,
creep behavior, not only in compressive but in tensile condition, can be calculated in
TSTM system [13].
The shape of the molds was a dog-bone and the dimensions of the cross section
were 150 mm × 150 mm in the center part and 150 mm × 280 mm at the ends; there-
fore, two ends of the specimens were enlarged to ensure uniform stress distribution at
the central part. The length of center part was 1500 mm. The transition area between
the ends and the central part was characterized by a rounded shape, which meant the
gradual widening of the cross-section of the grip, in order to minimize a possible
stress concentration and the risk of premature cracking in this zone. Therefore, it is
possible to realize tests for concrete with a maximum aggregate size of 28 mm with
these dimensions. The restraint degree can be achieved by controlling the deforma-
tion of restrained specimen. When shrinkage or expansion strain reaches 2 με (i.e.,
3 μm shrinkage for a 1500 mm long specimen), the stepper motor automatically
pulls or pushes the specimen back to the initial position; so the full restraint degree
can be reached [14]. The load cell would record the load which reflected the stress
development induced in this motion. The TSTM in Hohai University is shown in
Fig. 6.2.

@seismicisolation
@seismicisolation
272 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Fig. 6.2 TSTM in Hohai University

6.1.2 Testing Procedure

The concrete was cast into the free and restrained molds of TSTM system directly.
In order to reduce the frictional resistance, a thin vinyl sheet was placed between
the molds and concrete. A plastic sheet was used to cover the surface of concrete to
prevent evaporation. In practice, the temperature of interior concrete in mass concrete
construction increases firstly, and keeps constant for a while, then decreases until it
reaches environmental condition during early-age [15]. This temperature history
is close to adiabatic temperature rise profile. Therefore, the adiabatic temperature
rise profile was used to study the performance of concrete at early-age. The adiabatic
temperature rise of concrete is related to the following parameters: the proportion and
chemical composition of the cement (of blended cement); the fineness and particle
size distribution of the cement; the w/c ratio; the initial reaction temperature; the
presence and type of admixtures, etc. Adiabatic temperature rise is the inherent
characteristics of concrete. The concrete was cured at the highest temperature for
several hours, and then the temperature of concrete reduced at a constant rate until
restrained specimen cracked. If the tensile stress caused by a further temperature
drop exceeded the tensile strength of a specimen in the TSTM test, cracking would
occur. Cracking of the specimen was marked by a sharp drop in the stress curve.
When the cooling rate of temperature is reasonable, the viscoelastic effect could be
minimized and thermal gradients could be prevented [16]. If the cooling rate was
high, the thermal stresses develop faster and the stress gradient through the cross
section would occur due to different temperature between the center and boundary
of the concrete.

6.1.3 Calculation of Equivalent Age

To compare the cracking stress and axial tensile strength at the same equivalent age,
the real age of concrete in experimental results was converted to equivalent maturity

@seismicisolation
@seismicisolation
6.1 Methods and Calculation 273

age at 20 °C by the temperature history. For arbitrary temperature history, concrete

age can be converted to equivalent age t e through Arrhenius’ law [17].

t Ea (T ) 1 1
te = ∫ exp[ ( − )]dt (6.1)
0 R Tref + 273 T (t) + 273

where Ea (T ) is the activation energy, in kJ/mol [T (t) ≥ 20 °C, Ea (T ) kJ/mol;

T (t) ≥ 20 °C, Ea (T ) = 33.5 + 1.47 × (20 − T (t)) kJ/mol]; R is the ideal gas
constant (8.315 J mol−1 K−1 ); Tref is the reference temperature, in °C (20 °C); and
T (t) is the real temperature, in °C.

6.1.4 Calculation of Axial Tensile Strength and Elastic

Modulus

There was a certain relationship between splitting tensile strength and axial tensile
strength, the splitting tensile strength of the specimen could be measured by test,
and then the results could be converted to the axial tensile strength. The specific
relationship is shown in Eq. (6.2) [18].

ft,28 = 0.77 × fspl,28 + 0.21 (6.2)

where ft,28 and fspl,28 are the splitting tensile strength and axial tensile strength at 28
d, in MPa.
Yoo et al. [18] proposed and experimentally validated the development model of
concrete axial tensile strength and elastic modulus with age based on the hydration
degree model of concrete, as shown in Eq. (6.3) and Eq. (6.4).
{ [ ]}
ft (t) = ft,28 exp −λ1 ln(1 + (t − t0 ))−k1 (6.3)

{ [ ]}
Et (t) = Et,28 exp −λ2 ln(1 + (t − t0 ))−k2 (6.4)

where ft (t) is the axial tensile strength at age t, in MPa; Et (t) and Et,28 are the elastic
modulus of internally cured HPC at age t and 28 d, respectively, in GPa; and t0 is the
initial setting time, in d.

6.1.5 Calculation of Temperature Rise and Drop

In practical engineering, the concrete releases a great amount of heat due to the hydra-
tion of cement at early-age, which contributes to an increasing interior temperature
of the concrete. The restrained concrete may crack when the temperature cools down.

@seismicisolation
@seismicisolation
274 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Actually, the temperature history affects the development of the concrete performance
at early-age, which directly affects the cracking resistance. Therefore, the related
temperature parameters were used to analyze the early-age cracking resistance of
concrete.
The interior temperature of the mass concrete would increase at early-age which
results from the reaction of cement with water. Equation (6.5) was used to calculate
the adiabatic temperature rise of HPC.

Ttr = Tht − Tct (6.5)

where Ttr is the adiabatic temperature rise of HPC, in °C; Tht is the highest temperature
of HPC before cracking, in °C; and Tct is the initial temperature of HPC, in °C.
The highest temperature was forced to cool down after 36 h constant temperature
curing. Under the same temperature history, the temperature drop of concrete could
directly reflect the cracking resistance during the forced cooling stage. The specific
process of calculation is shown in Eq. (6.6).

Ttd = Tht − Tck (6.6)

where Ttd and Tck are the temperature drop and temperature of concrete at the cracking
time, in °C.
The temperature at which the corresponding restrained stress switched from
compression to tension was denoted as the zero-stress temperature or Tzero−stress ,
as shown in Fig. 6.3.

3.0 60
2.5
50
2.0
1.5 Tzero-stress
)

Stress 40
Stress (MPa)

Temperature (

1.0 Temperature
0.5 30
0.0
20
-0.5
-1.0
10
-1.5
-2.0 0
0 24 48 72 96 120
Age (h)

Fig. 6.3 Typical curves of temperature and stress

@seismicisolation
@seismicisolation
6.1 Methods and Calculation 275

6.1.6 Calculation of Autogenous Shrinkage

The autogenous shrinkage of concrete was the main cause of restrained stress under
restrained conditions, which led to the early-age cracking of concrete. Therefore,
it was of great significance to analyze the early-age autogenous shrinkage in order
to further understand the cracking resistance of concrete. Usually, the measurement
of free deformation of concrete at early-age was not conducted immediately after
pouring, some scholars would select a time as the point of zero-deformation for
starting to measure the free deformation of concrete specimens. Actually, the selec-
tion of the time to start the measurement will have a great influence on the results of
free deformation and autogenous shrinkage. Bentur et al. [19] took the moment when
the stress began to appear in the restrained specimen as the zero-deformation point
to study the early shrinkage and cracking of cementitious materials; Similarly, Lura
et al. [20] chose the moment of stress generation as the starting point to measure
free deformation when studying the influence of temperature on the shrinkage of
concrete. The moment of stress appearance in the restrained specimen was selected
as the starting measurement point of free deformation because only deformation
caused by stress was evaluated. No moisture diffusion occurred as the reason that
the concrete specimens were sealed by polyethylene film and aluminum foil. Auto-
genous shrinkage is the volume change of concrete under constant temperature.
Thus, besides autogenous shrinkage, the thermal deformation was also included in
the free deformation of free shrinkage specimen. However, separating autogenous
shrinkage from free deformation was very difficult due to the changing CTE of HSC
at early-age.
As the reason of that was generally not under a constant temperature condition for
an actual concrete structure, the autogenous shrinkage needed to be separated from
the free total deformation. In combination with the measured temperature history and
then the autogenous shrinkage deformation was further separated. In fact, the CTE
changes drastically and the development of autogenous shrinkage is also extremely
rapid in the first day of concrete hardening. Using a simple constant CTE to separate
the temperature deformation and autogenous shrinkage deformation would cause
greater errors in the case of a large range of temperature changes. Therefore, the
calculation of concrete CTE is shown in Eq. (6.7).
The CTE was calculated by using Eq. (6.7), as described in [5].

αT (t) = αk × (1 + 41 × t −m ) (6.7)

where αT (t) is CTE of HSC at t h, in με/°C; αk is 28-d CTE of HSC, in με/°C; and
the value of m is 2.0.
The autogenous shrinkage of concrete mostly occurs in the first 24 h after pouring
[6], and the forced cooling started a few days after pouring and lasted for a short
time. Thus, the autogenous shrinkage of concrete was small in the forced cooling
phase, and the deformation was mainly caused by temperature changes. The CTE
of concrete was obtained by regression analysis of the time history and the free

@seismicisolation
@seismicisolation
276 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

deformation in the forced cooling phase. The CTE decreases to a constant value
within 24 h. Thus, the 28-d CTE of concrete could be replaced by the CTE in the
forced cooling phase.
Equation (6.8) was applied to separate the autogenous shrinkage from the free
deformation of HSC, as described in [21].
[ ]
εas (t) = εsh (t) − αT (t) · T (t) − Ttime−zero (6.8)

where εas (t) is the autogenous shrinkage at t h, in με; εsh (t) is the free deformation at
t h, in με; T (t) is the temperature of HSC at t h, in °C; and Ttime−zero is the time-zero
temperature of HSC, in °C.

6.1.7 Calculation of Creep

The total creep of concrete is composed of basic creep and dry creep. Since all
concrete specimens were sealed with polyethylene film and aluminum foil after
pouring, almost no drying creep occurs. Thus, the relationship between the total
deformation, free shrinkage, basic creep, and elastic strain of the restrained shrinkage
specimen could be expressed by Eq. (6.9), as described in [22].

εtotal = εel + εsh + εcr = 0 (6.9)

where εtotal is the total deformation of the restrained shrinkage specimen, in με;
εel is the accumulation of elastic strain, in με; εsh is the free shrinkage of the
restrained shrinkage specimen, which is consistent with the free deformation of
the free shrinkage specimen, in με; and εcr is the basic creep, in με.
Thus, the basic creep of the restrained shrinkage specimen could be calculated by
Eq. (6.10).

εcr = −εel − εsh (6.10)

The relationship between elastic strain, creep, and free shrinkage is shown in
Fig. 6.4. The creep of concrete consists of basic creep and drying creep. Only basic
creep was considered in this research because all the concrete specimens were sealed
immediately after pouring.
For many practical concrete structures, especially in mass concrete, there is no
significant moisture exchange between concrete and environment, so only basic
tensile creep was investigated in TSTM. The tensile stress zero, which was used
to calculate the specific basic tensile creep, was defined as the tensile stress at the
time when the cooling period began, since the tension was applied at this instant.
The visco-elastic response of early-age concrete under fully restraint condition was
expressed in terms of basic tensile creep and specific basic tensile creep (basic tensile
creep/restrained stress). The basic tensile creep/shrinkage, which reflected reduction

@seismicisolation
@seismicisolation
6.1 Methods and Calculation 277

Free shrinkage

Shrinkage+creep
Creep (Cumulative curve)
Strain Stress

Restrained stress

Elastic strain

} Compensation
cycles

Age

Fig. 6.4 Relationship between elastic strain, creep, and free shrinkage [11]

of the development of tensile strain in the restrained concrete, was also utilized to
account for the creep behaviors.
Specific creep is normally defined as the creep strain per unit stress (με/MPa) of
concrete in a constant load creep test [23]. Therefore, the creep tests for different
stress levels can be normalized. The restrained tensile stress develops with time in
the restrained shrinkage test. Thus, the specific basic tensile creep was defined as the
ratio of cumulative basic tensile creep to restrained tensile stress at any time, which
reflected the history dependence of tensile creep properties.

6.1.8 Calculation of Ratio of Cracking Stress to Axial Tensile

Strength

The ratio of cracking stress to axial tensile strength was utilized to evaluate the
cracking resistance of ICC [24], which was obtained on the basis of restrained
stress and the axial tensile strength obtained from the prediction model. Axial tensile
strength is approximately determined by Eq. (6.11).
σcracking
R= (6.11)
ft

where ft is axial tensile strength, in MPa; and σcracking is cracking stress, in MPa.

@seismicisolation
@seismicisolation
278 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

6.1.9 Calculation of Cracking Reserve

Cracking temperature drop, cracking stress/axial tensile strength, cracking age, and
elastic strain/strain capacity have been applied as single criteria to estimate the
cracking resistance of concrete. Additionally, the stress reserve taking the cracking
stress and stress at room temperature into consideration was applied to estimate the
cracking resistance of concrete, as given in Eq. (6.12).

σc − σt
φ= × 100% (6.12)
σc

where φ is the stress reserve; σc is the cracking stress, in MPa; and σt is the stress at
room temperature, in MPa.

6.2 Integrated Criterion of Cracking Resistance

Cracking resistance of concrete at early-age is influenced by strength, elastic

modulus, temperature rise, autogenous shrinkage, rate of stress development, and
tensile creep. So a comprehensive evaluation on the cracking resistance of concrete at
early-age is necessary. Quantitative criteria related to TSTM are proposed to evaluate
cracking potential of concrete under different test conditions, such as temperature
drop, cracking age, cracking stress, and stress reserve.
Cracking potential is classified based on two criteria, namely, the net time of
cracking and the stress rate at cracking, according to the ASTM standard [25]. An
integrated criterion for estimating cracking potential was suggested by Kovler and
Bentur [26], as shown in Eq. (6.13).

S
ϕ= (6.13)
tcr

where ϕ is the integrated criterion reflecting the cracking potential, S is the stress
rate at cracking, and t cr is the net time of cracking in the ring test (time from the
initiation of drying).
The integrated criterion of cracking resistance based on the results of tensile
stress rate at cracking age and net age of cracking was first proposed to evaluate the
early-age cracking behavior of concrete, as given in Eq. (6.14) [27].

S
ϕN = (6.14)
ttcr

where ϕN is the integrated criterion of cracking resistance, in MPa/(h·day); and ttcr

is the net age of cracking from starting point of the tensile state to cracking age, in

@seismicisolation
@seismicisolation
6.3 Cracking Resistance of Concrete Internally Cured with SAPs Under … 279

h. Integrated criterion ϕN was obtained by dividing the stress rate by the net time to
cracking. The integrated criterion of cracking resistance has been adopeted in many
studies [1, 3, 5, 6, 28, 29].

6.3 Cracking Resistance of Concrete Internally Cured

with SAPs Under Uniaxial Restrained Condition

6.3.1 Mixture Proportions and Materials

Four HPC mix designs were evaluated. Each mixture was identified by a label X–
Y. The first part of the specimen name X referred to type of concrete (i.e., SAP
= concrete internally cured with SAPs), Y indicated the percentages of amount of
SAPs to amount of cement (i.e., 17, 35, and 49 represented that the amount of SAPs
was about equal to 0.1728%, 0.3501%, and 0.4935% by mass of cement). Four mix
designs were named as SAP-0, SAP-17, SAP-35, and SAP-49. Mixture SAP-0 was
the reference concrete without SAPs and the basic w/c ratio was 0.33. Mixtures
SAP-17, SAP-35, and SAP-49 were the HPC internally cured with SAPs.
The theoretical amount of water required ensuring maximum cement hydration
was estimated at 32.85 kg/m3 for all mixtures in the experiment. For the mixture
designs shown in Table 6.1 the percentages of SAPs actually provided to that required
in theory were as follows: 0% for mixture SAP-0, 35% for mixture SAP-17, 71% for
mixture SAP-35, and 100% for mixture SAP-49, respectively. The internal curing
water was 11.51, 23.30, and 32.85 kg/m3 for mixtures SAP-17, SAP-35, and SAP-
49, respectively. Therefore, the internal curing w/c ratio or total w/c ratio was 0.022,
0.046, and 0.064 or 0.352, 0.376, and 0.394 for mixtures SAP-17, SAP-35, and SAP-
49, respectively. Portland Cement (Cement II 52.5R) was employed in accordance
with Chinese National Standard GB 175-2007 [30]. The Blaine fineness of the cement
was 375 m2 /kg and the loss on ignition was 3.11%. The fineness modulus of river
sand used was 2.35, and the maximum size of river sand was 2.0 mm. The coarse
aggregates used were crushed limestone and the maximum size was 26 mm. A
suspension-polymerized, covalently cross-linked acrylamide/acrylic acid copolymer
as SAPs with a dry-bulk density of 700 kg/m3 was used. The spherical particles had
diameters varying from 200 μm to 250 μm in the dry state. The absorption capacity
of SAPs used was tested using slump flow measurement. A liquid polycarboxylate-
based superplasticizer was added to mixtures SAP-0, SAP-17, SAP-35, and SAP-49.
Tap water was used as mixing water. The concrete mixtures were prepared according
to the following mixing procedure: cement powder, oven-dried aggregates, and dry
SAPs were premixed for 30 s before adding the first half of the water. The other
half of the water was added together with superplasticizer after 120 s of mixing. The
resultant total mixing time was 5 min. Both cement and water were conditioned at
room temperature for at least 24 ± 1 h. Once the concrete was fully mixed, the fresh
concrete was cast into the mold and consolidated with a needle vibrator.

@seismicisolation
@seismicisolation
280 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Table 6.1 Concrete mixture proportions

Mixture composition Concrete mixtures
SAP-0 SAP-17 SAP-35 SAP-49
Water (kg/m3 ) 171 171 171 171
Cement (kg/m3 ) 512 512 512 512
Fine aggregate(sand) (kg/m3 ) 636 636 636 636
Coarse aggregate (kg/m3 ) 1131 1131 1131 1131
SAPs (kg/m3 ) 0 0.885 1.792 2.527
Superplasticizer (%) 0.65 0.65 0.65 0.65

6.3.2 Temperature of Concrete with SAPs

Figure 6.5 shows that the adiabatic temperature rise of HPC was 27.6, 29.3, 31.0,
and 34.9 °C, which increased by 6.2%, 12.3%, and 26.4% when the amount of SAPs
increased from 0% to 0.17%, 0.35%, and 0.49%, respectively. This may be due to
the higher degree of reaction in HPC internally cured with higher amount of SAPs.
Equation (6.15) was used to calculate the degree of hydration, as reported in [31].

1.031wt /c
α28 = (6.15)
0.194 + wt /c

where α28 is the degree of hydration of concrete at 28 d; and w/c is the total w/c ratio
of concrete.
Therefore, the degree of hydration was 0.65, 0.66, 0.68, and 0.69 at 28 d for
mixtures SAP-0, SAP-17, SAP-35, and SAP-49, respectively. The adiabatic tempera-
ture rise increased with increase of the amount of SAPs. Therefore, the water released
from SAPs had a function to continue the hydration of cement further in sealed
condition.

6.3.3 Autogenous Shrinkage and Restrained Stress

of Concrete with SAPs

The total deformation is shown in Fig. 6.6. The most early-age autogenous shrinkage
occurred within 24 h after casting. Therefore, in the cooling period, most of the defor-
mation was caused by the temperature change, and the CTE of concrete decreased
to a constant value within 1 day. Therefore, the average CTE of HPC in the cooling
period could be considered as the CTE of HPC at the age of 28 d of concrete and
calculated by dividing the deformation change by the temperature change and was
6.5, 5.9, 5.8, and 4.6 με/°C for mixtures SAP-0, SAP-17, SAP-35, and SAP-49,
respectively.

@seismicisolation
@seismicisolation
6.3 Cracking Resistance of Concrete Internally Cured with SAPs Under … 281

Fig. 6.5 Temperature 60

measured in free shrinkage SAP-0
concrete specimens 50 SAP-17
SAP-35

Temperature (°C)
40 SAP-49

30

20

10

0
0 20 40 60 80 100
Real age (hour)
(a) Actual temperature rise and drop

40

30 SAP-0
SAP-17
SAP-35
Temperature (°C)

20 SAP-49

10

-10

-20
0 20 40 60 80 100
Real age (hour)
(b) Adiabatic temperature rise and drop

The absolute value of autogenous shrinkage of HPC at the age of 70 h after casting
was 121, 90, 41, and 26 με and the absolute value of which decreased by 25.6%,
66.1%, and 78.5% when the amount of SAPs increased from 0% to 0.17%, 0.35%,
and 0.49%, respectively, as shown in Fig. 6.7. The HPC internally cured with SAPs
showed a lower autogenous shrinkage than that without SAPs, which was due to the
decreased self-desiccation of HPC internally cured with SAPs.
The equivalent maturity age when the restrained specimens cracked was 124,
131, 134, and 147 h when the amount of SAPs was 0%, 0.17%, 0.35%, and 0.49%,
respectively. The cracking stress or axial tensile strength at the same equivalent
maturity age was 2.9, 2.8, 2.6, and 2.5 MPa or 3.7, 3.5, 3.2, and 3.0 MPa when
the amount of SAPs was 0%, 0.17%, 0.35%, and 0.49%, respectively. Test results

@seismicisolation
@seismicisolation
282 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

200
SAP-0
150 SAP-17
SAP-35
100 SAP-49
(
Total deformation ( με

50

-50

-100

-150
0 20 40 60 80 100
Real age (hour)

Fig. 6.6 Total deformation measured in free shrinkage concrete specimens

20 SAP-0
SAP-17
0 SAP-35
(
Autogenous shrinkage (με

-20 SAP-49

-40

-60

-80

-100

-120

-140
0 20 40 60 80
Real age (hour)

Fig. 6.7 Autogenous shrinkage measured in free shrinkage concrete specimens

showed the ratio of cracking stress to axial tensile strength was 0.78, 0.80, 0.81, and
0.83 when the amount of SAPs was 0%, 0.17%, 0.35%, and 0.49%, respectively.
The ratios of cracking stress to axial tensile strength of all specimens were less than
1.0. Rapid loading rate of the splitting tensile strength test, damage due to coupling
between creep and damage, and internal microcracking may be responsible for this

@seismicisolation
@seismicisolation
6.3 Cracking Resistance of Concrete Internally Cured with SAPs Under … 283

3.0
SAP-0
2.5
SAP-17
Restrained stress (MPa)
2.0 SAP-35
SAP-49
1.5

1.0

0.5

0.0

-0.5

-1.0
0 20 40 60 80 100
Real age (hour)

Fig. 6.8 Restrained stress measured in restrained shrinkage concrete specimens

phenomenon [1]. The restrained tensile stress and axial tensile strength decreased
with the increase of amount of SAPs, however, the decreasing degree of cracking
stress was less than that of axial tensile strength due to the less internal micro-cracking
in concrete. Therefore, the ratio of cracking stress to axial tensile strength increased
with the increase of amount of SAPs.
The restrained compressive and tensile stress both decreased with the increase of
the amount of the SAPs, as shown in Fig. 6.8.
The rate of restrained tensile stress was 1.7, 1.5, 1.4, and 1.2 MPa/day, which
decreased by 11.8%, 17.6%, and 29.4% when the amount of SAPs increased from
0% to 0.17%, 0.35%, and 0.49%, respectively.

6.3.4 Creep of Concrete with SAPs

Results in [32] show that the effect of temperature on tensile creep and relaxation
could be dealt with the maturity method. The maturity expresses the equivalent
hydration period, normally at 20 °C, which gives the current value of a material
property. Therefore, Eq. (6.1) was utilized to compare the behavior of HPC with
different amounts of SAPs under different temperature histories.
The relationships between basic tensile creep and real age or equivalent maturity
age are shown in Fig. 6.9, respectively.
Specific basic tensile creep which is defined as the basic tensile creep strain per
unit stress (με/MPa) of concrete in a constant load creep test is used to normalize the

@seismicisolation
@seismicisolation
284 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

160
SAP-0
140 SAP-17
Basic tensile creep (με) SAP-35
120
SAP-49
100
80
60
40
20
0
35 40 45 50 55 60 65 70 75 80 85
Real age (hour)
(a) real age

180
160 SAP-0
SAP-17
140 SAP-35
Basic tensile creep (με)

120 SAP-49

100
80
60
40
20
0
60 70 80 90 100 110 120 130 140 150 160
Equivalent maturity age (hour)
(b) equivalent maturity age

Fig. 6.9 Basic tensile creep calculated from restrained and free shrinkage

tensile creep tests for different stress development. The restrained tensile stress of
HPC increased with the increase of age in the restrained specimens. Therefore, the
specific basic tensile creep was defined as the ratio of cumulative basic tensile creep
strain to restrained tensile stress at any time. The age when the restrained tensile
stress occurred differed from each other due to different adiabatic temperature rises.
Therefore, the initial age of basic tensile creep for mixtures SAP-0, SAP-17, SAP-35,
and SAP-49 differed from each other. The values of restrained tensile stress of HPC

@seismicisolation
@seismicisolation
6.3 Cracking Resistance of Concrete Internally Cured with SAPs Under … 285

at the age when the restrained specimen of mixture SAP-0 cracked were 3.0, 2.7, 2.3,
and 2.2 MPa, which decreased by 10.0%, 23.3%, and 26.7% when the amount of
SAPs increased from 0% to 0.17%, 0.35%, and 0.49%, respectively. And the values
of basic tensile creep of HPC at the age when the restrained specimen of mixture
SAP-0 cracked were 148, 60, 31, and 15 με, which decreased by 59.5%, 79.1%, and
89.9% when the amount of SAPs increased from 0% to 0.17%, 0.35%, and 0.49%,
respectively.
As shown in Fig. 6.10, the specific basic tensile creep of HPC decreased with the
increase of amount of SAPs. Mixture SAP-0 showed the largest specific basic tensile
creep among the four HPC mixtures. The values of specific basic tensile creep of
HPC at the age when the restrained specimen of mixture SAP-0 cracked were 45, 23,
13, and 7 με/MPa, which decreased by 48.9%, 71.1%, and 84.4% when the amount
of SAPs increased from 0% to 0.17%, 0.35%, and 0.49%, respectively. This tendency
indicated that the addition of SAPs would decrease the specific basic tensile creep
behavior of HPC.
The B3 model is attractive for describing early-age tensile creep because it is based
on a phenomenological approach for aging of concrete [33], as shown in Eq. (6.16).

, , , t
J (t, t ) = q1 + q2 Q(t, t ) + q3 ln[1 + (t − t )n ] + q4 ln( , ) (6.16)
t
,
where J (t, t ) is the tensile creep compliance function ((tensile creep plus elastic
strain)/restrained stress); the individual terms q1 − q4 are the empirical material
constitutive parameters based on concrete strength and composition, which represent
physically distinct components of creep, q1 = 0.6 × 106 /E28 , q2 = 185.4c0.5 fc−0.9 ,

70
SAP-0
Specific basic tensile creep (με/MPa)

60 SAP-17
SAP-35
50 SAP-49
40

30

20

10

0
60 70 80 90 100 110 120 130 140 150 160
Equivalent maturity age (hour)

Fig. 6.10 Specific basic tensile creep calculated from restrained and free shrinkage

@seismicisolation
@seismicisolation
286 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

q3 = 0.29(w/c)4 q2 , q4 = 20.3(a/c)−0.7 is elastic modulus of concrete at 28 d, f c

is compressive strength of concrete cylinders at 28 d, c is cement content, and a/c
,
( , )while t denotes the age of
is aggregate to cement ratio; t is the age of concrete
loading in d; and n is taken as 0.1. The function Q t, t is given by [33], as shown
in Eq. (6.17).

⎡
r(t , ) ⎤−1/r(t )
,

,
Qf (t )
Q(t, t ) = Qf (t )⎣1 + ⎦
, ,
(6.17)
Z(t, t , )

, ,
where Qf (t ), Z(t, t ), and r are given by [33].
, , ,
Qf (t ) = [0.086(t )2/9 + 1.21(t )4/9 ]−1 (6.18)

, , ,
Z(t, t ) = (t )−m ln[1 + (t − t )n ] (6.19)

, ,
r(t ) = 1.7(t )0.12 + 8 (6.20)

where m is taken as 0.5.

In order to obtain the tensile creep compliance function of HPC internally cured
with SAPs, a modification based on B3 model was conducted by comparing the
difference of tensile creep compliance function of HPC with different amounts of
SAPs, as shown in Eq. (6.21).
, , ,
J (t, t , α) = J0c (tc , t ) × ψ × β(t, t ) × f (α) (6.21)
,
where J (t, t , α) is the tensile creep compliance function at the equivalent maturity
age t when the amount of SAPs is α; ψ is defined as the ratio of the tensile creep
compliance function ( of) mixture SAP-0 obtained from test to that from B3 model
and was 0.68; J0c tc , t , is the tensile
( ) creep compliance function of mixture SAP-
0 at the age when it cracked; β t, t , is the development rate coefficient of tensile
creep compliance function, as shown in Fig. 6.11; f (α) is the influence coefficient
of amount of SAPs; α is the amount of SAPs in %; t is the equivalent maturity age
of HPC while t , denotes the equivalent maturity age of loading in hours; tc is the
equivalent maturity age when the restrained specimen of mixture SAP-0 cracked. ( )
Figure 6.11 shows the relationship between development rate coefficient β t, t ,
of tensile creep compliance function and equivalent maturity age since t , of HPC for
mixtures SAP-0, SAP-17, SAP-35, and SAP-49, respectively, as shown in Eq. (6.22).
,
,
β(t, t ) = 1.0 − le−
t−t
b (6.22)

@seismicisolation
@seismicisolation
6.3 Cracking Resistance of Concrete Internally Cured with SAPs Under … 287

( )
Fig. 6.11 Relationship between the development rate coefficient β t, t , of tensile creep compli-
ance function and equivalent maturity age since t , of HPC

( )
where β t, t , are calculated by dividing the tensile creep compliance function at
equivalent maturity age since t , of HPC by that at the age when the restrained spec-
imen of mixture SAP-0 cracked for mixtures SAP-0, SAP-17, SAP-35, and SAP-49,
respectively.
Coefficients l and b could be determined by regression analysis, as shown in
Eq. (6.23).
,
,
β(t, t ) = 1.0 − 1.0e− 4.16
t−t
(6.23)

Figure 6.12 shows the relationship between the influence coefficient f (α) of
HPC at the equivalent maturity age when the restrained specimen of mixture SAP-0
cracked and the amount of SAPs. The relationship between the ratio of the tensile
creep compliance function of HPC with different amounts of SAPs to that without
SAPs and amount of SAPs was linear and could be shown as follows.

f (α) = 1.0 + s × α (6.24)

where f (α) is calculated by dividing the tensile creep compliance function of mixture
SAP-0, SAP-17, SAP-35, and SAP-49 by that of mixture SAP-0 at the equivalent
maturity age when the restrained specimen of mixture SAP-0 cracked, respectively,
and are 1.00, 0.81, 0.68, and 0.51 for mixtures SAP-0, SAP-17, SAP-35, and SAP-49,
respectively.
Coefficient s was determined by regression analysis, as shown in Eq. (6.25).

@seismicisolation
@seismicisolation
288 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Fig. 6.12 Relationship

between the influence
coefficient f (α) and amount
of SAPs α

f (α) = 1.0 − 0.98α (6.25)

The model for tensile creep compliance function of HPC internally cured with
SAPs was shown in Eq. (6.26).
⎧  ,

⎪
⎪
,
J (t, t , α) = J0c (tc , t) × 0.68 × 1.0 − 1.0e − t−t
× (1.0 − 0.98α)
⎪
⎪
4.16
⎪
⎪
⎪
⎪ , ,
J0c (tc , t) = q1 + q2 Q(tc , t ) + q3 ln[1 + (tc − t ) ] + q4 ln( ttc, )
0.1
⎪
⎪
⎪
⎪  −1/r(t , )
⎨  , r(t )
,
, , Qf (t )
⎪ Q(tc , t ) = Qf (t ) 1 + Z(t ,t , ) (6.26)
⎪
⎪
c
⎪
⎪
⎪
⎪
, ,
Qf (t ) = [0.086(t ) + 1.21(t )4/9 ]−1
2/9 ,

⎪
⎪
⎪ Z(tc , t ) = (t )−0.5 ln[1 + (t − t , )0.1 ]
⎪
, ,
⎪
⎩ , ,
r(t ) = 1.7(t )0.12 + 8

where the results of q1 , q2 , q3 , and q4 were 12.6, 120.4, 0.41, and 8.53, respectively.
The B3 model is also utilized in [33] to predict the tensile creep of concrete. The
q3 /q2 was 0.0034.
Figure 6.13 shows the tensile creep compliance function of HPC with different
amounts of SAPs obtained from the test and Eq. (6.26).
For instance, the values of tensile creep compliance function at the age when the
restrained specimen of mixture SAP-0 cracked obtained from Eq. (6.26) or test were
85, 71, 55, and 44 με/MPa or 85, 69, 58, and 43 με/MPa when the amount of SAPs
was 0%, 0.17%, 0.35%, and 0.49%, respectively. The deviations between results
obtained from Eq. (6.26) and test were 0%, 2.9%, -5.2%, and 2.3%, respectively.

@seismicisolation
@seismicisolation
6.3 Cracking Resistance of Concrete Internally Cured with SAPs Under … 289

Fig. 6.13 Comparison of

tensile creep compliance
function of HPC internally
cured with SAPs from test
results and proposed model

@seismicisolation
@seismicisolation
290 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Fig. 6.13 (continued)

6.3.5 Cracking Resistance of Concrete with SAPs

Integrated criterion ϕN was obtained by dividing the rate of restrained tensile stress
by the net time of cracking. The net time of cracking was 78, 81, 82, and 84 h when
the amount of SAPs was 0%, 0.17%, 0.35%, and 0.49% for mixtures SAP-0, SAP-
17, SAP-35, and SAP-49, respectively. The results of the evaluation of the cracking
resistance of the HPC mixtures are shown in Fig. 6.14. Integrated criterion ϕN was
0.523, 0.444, 0.410, and 0.343 MPa/day2 , which decreased by 15.4%, 21.6%, and
34.4% when the amount of SAPs increased from 0% to 0.17%, 0.35%, and 0.49%,
respectively.

6.3.6 Summary

The experimental findings on the influence of SAPs on the behavior and cracking
resistance of HPC at early-age under adiabatic condition were investigated. Test and
analysis on temperature, autogenous shrinkage, restrained stress, and tensile creep
of HPC with four groups of mixture proportions were conducted. The amount of
SAPs on behavior and cracking resistance of HPC at early-age was quantified using
TSTM. The following conclusions were drawn from the obtained results.
(1) The adiabatic temperature rise of HPC increased with the increase of amount of
SAPs. The adiabatic temperature rise was 27.6, 29.3, 31.0, and 34.9 °C, which
increased by 6.2%, 12.3%, and 26.4% when the amount of SAPs increased from
0% to 0.17%, 0.35%, and 0.49%, respectively.
(2) The autogenous shrinkage of HPC at the age of 70 h after casting decreased
with the increase of amount of SAPs. The autogenous shrinkage of HPC at the
age of 70 h after casting was -121, -90, -41, and -26 με and the absolute value

@seismicisolation
@seismicisolation
6.3 Cracking Resistance of Concrete Internally Cured with SAPs Under … 291

1.2
Cracking Resistance Using Integrated Criterion

Low
1.0
Integrated criterion (MPa/day2)

0.8
High
0.6

0.4

0.2

0.0
SAP-0 SAP-17 SAP-35 SAP-49

Fig. 6.14 Estimation of cracking resistance using integrated criterion

of which decreased by 25.6%, 66.1%, and 78.5% when the amount of SAPs
increased from 0% to 0.17%, 0.35%, and 0.49%, respectively.
(3) The rate of restrained tensile stress of HPC decreased with the increase of
amount of SAPs. The rate of restrained tensile stress was 1.7, 1.5, 1.4, and
1.2 MPa/day, which decreased by 11.8%, 17.6%, and 29.4% when the amount
of SAPs increased from 0% to 0.17%, 0.35%, and 0.49%, respectively.
(4) The specific basic tensile creep of HPC decreased with the increase of amount
of SAPs. The specific basic tensile creep of HPC at the age when the restrained
specimen of mixture SAP-0 cracked was 45, 23, 13, and 7 με/MPa, which
decreased by 48.9%, 71.1%, and 84.4% when the amount of SAPs increased
from 0% to 0.17%, 0.35%, and 0.49%, respectively.
(5) A model for predicting the tensile creep compliance function of HPC was
proposed considering the influence of amount of SAPs.
(6) The cracking resistance which was based on the integrated criterion increased
with the increase of amount of SAPs. The integrated criterion was 0.977, 0.800,
0.730, and 0.600 MPa/day2 , which decreased by 17.6%, 24.8%, and 38.2%
when the amount of SAPs increased from 0% to 0.17%, 0.35%, and 0.49%,
respectively.

@seismicisolation
@seismicisolation
292 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

6.4 Cracking Resistance of Concrete Internally Cured

with LWAs Under Uniaxial Restrained Condition

6.4.1 Mixture Proportions and Materials

Four concrete mixtures with low w/c ratio were used. The mixture proportions,
designated as LWA-0, LWA-10, LWA-30, and LWA-50, are presented in Table 6.2.
Mixture LWA-0 is the reference concrete with no internal curing, and Mixtures
LWA-10, LWA-30, and LWA-50 use internal curing with pre-wetted LWA. For the
ICC in which part of the normal-weight coarse aggregate is replaced with LWAs, the
replacement ratios of coarse aggregates are 10%, 30%, and 50% of the total volume
of coarse aggregates for Mixtures LWA-10, LWA-30, and LWA-50, respectively. All
mixtures have a w/c ratio of 0.33.
Portland Cement (Cement II 52.5R) with a Blaine fineness of 375 m2 /kg was
employed in accordance with Chinese National Standard GB 175–2007 [30]. Normal
weight river sand with a fineness modulus of 2.05 was used. Crushed limestone (with
a maximum size of 26 mm) was employed as coarse aggregates. The LWAs used in
Mixtures LWA-10, LWA-30, and LWA-50 were manufactured rotary kiln expanded
clay. The LWAs had a dry-bulk density of 1050 kg/m3 and a 24 h absorption value
of 12% by mass of the dry material. The LWAs were oven dried, air cooled, and then
submerged in water for 24 h. After 24 h, the water was decanted and the surface of
the aggregate was patted dry. The LWAs were spread out using paper towel method
and the paper towel was placed across the surface of the aggregates. This process was
repeated up to the surface drying condition. Once it appeared that the paper towel
was no longer picking up moisture from the aggregate, it was assumed that a surface
dry condition has been reached and the absorption of LWAs could be determined.
The crushing strength of dry LWAs is 1.2 MPa, which is significantly lower than
that of normal aggregates (typically between 100 and 300 MPa). The lightweight
sand was used as internal curing material to replace the fine aggregates in concrete
in the conventional approaches of internal curing. The particle size of the LWAs

Table 6.2 Concrete mixture proportions

Mixture composition Concrete mixtures
LWA-0 LWA-10 LWA-30 LWA-50
Water (kg/m3 ) 171 171 171 171
Cement (kg/m3 ) 512 512 512 512
Fine aggregate(sand) (kg/m3 ) 636 636 636 636
Coarse aggregate (kg/m3 ) 1131 1018 792 566
LWA (kg/m3 , dry) 0 45 134 223
Internal curing water (kg/m3 ) 0 5.4 16.1 26.7
Superplasticizer (%) 0.6 0.6 0.6 0.6

@seismicisolation
@seismicisolation
6.4 Cracking Resistance of Concrete Internally Cured with LWAs Under … 293

used in the experiment is from 4.7 mm to 9.5 mm. A liquid polycarboxylate-based

superplasticizer was employed to adjust the workability of different mixtures. Tap
water was used as internal curing water.

6.4.2 Temperature of Concrete with LWAs

The reaction of cement with water is exothermic and results in a temperature rise
in massive concrete members [34]. When the structure cools down, the concrete
contracts and may crack if restrained. Figure 6.15 shows that the temperature rises
of Mixture LWA-0, LWA-10, LWA-30, and LWA-50 are 31, 33, 33, and 37 °C,
respectively.
A high temperature rise means a high degree of reaction in ICC with pre-wetted
LWAs. The temperature rise of Mixture LWA-50 was higher than that of the other
three Mixtures, which indicated the former’s higher hydration degree and higher
hydration heat with pre-wetted LWAs. However, the temperature rises of Mixture
LWA-10 and LWA-30 were almost equal to that of normal concrete maybe because
a small amount of internal curing water does not work well in concrete with low w/c
ratio.
Only temperature rise has been considered in most measures. However, cracking
does not occur in concrete members by temperature rise itself. The temperature drop
approach was also employed to investigate concrete cracking; the use of pre-wetted
LWAs has been shown to be beneficial in improving the thermal cracking performance
of concrete [35]. In the period of cooling in the TSTM test, the magnitude of the
temperature drop was used to quantify the thermal cracking resistance of a specimen
at that age [36].
The temperature drops of concrete at the age of cracking were 37, 43, 43, and
47 °C for Mixture LWA-0, LWA-10, LWA-30, and LWA-50, respectively. Mixture
LWA-50 (which had the largest amount of internal curing water) required the largest
amount of temperature change to induce cracking; this condition demonstrates higher
cracking resistance with a high mixture of pre-wetted LWAs, as shown in Fig. 6.15.
The cracking temperature drop increased for autogenous and drying shrinkage were
reduced or delayed in the ICC [37].

6.4.3 Free Deformation of Concrete with LWAs

Figure 6.16 shows the total strain measured for a period of testing in each unrestrained
concrete specimen.
The extent of expansion increased with the increase in the amount of pre-wetted
LWAs used in the concrete. The occurrence of expansion in the sealed unrestrained
specimens was measured in mortars and concrete containing LWAs. Figure 6.16
also shows that the maximum values of free total strain were 201, 209, 229, and

@seismicisolation
@seismicisolation
294 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

60

50
Temperature (°C)

40
LWA-0
30 LWA-10
LWA-30
LWA-50
20

10

0
0 24 48 72 96 120
Age (h)
(a) Actual temperature rise and drop

40

30
Temperature (°C)

20
LWA-0
10 LWA-10
LWA-30
LWA-50
0

-10

-20
0 24 48 72 96 120
Age (h)
(b) Absolute temperature rise and drop

Fig. 6.15 Temperature measured in free shrinkage concrete specimens

288 με when the amounts of pre-wetted LWAs were 0, 10%, 30%, and 50% for
Mixture LWA-0, LWA-10, LWA-30, and LWA-50, respectively. The total strain of
ICC increased with the increase in the amount of pre-wetted LWAs.
The autogenous shrinkage of concrete at an age of 60 h was − 116, − 101,
− 43, and 56 με when the amounts of pre-wetted LWAs were 0, 10%, 30%, and
50% for Mixture LWA-0, LWA-10, LWA-30, and LWA-50, respectively. Autogenous

@seismicisolation
@seismicisolation
6.4 Cracking Resistance of Concrete Internally Cured with LWAs Under … 295

Strain (με)

LWA-0
LWA-10
LWA-30
LWA-50

Age (h)

Fig. 6.16 Total strain measured in free shrinkage concrete specimens

shrinkage decreased with the increase in the amount of pre-wetted LWAs; this result is
in accordance with the results of Henkensiefken [38]. The effect of pre-wetted LWAs
on the reduction in autogenous shrinkage in ICC was a clear manifestation of their
effectiveness as a means of internal curing. The pores of the cement matrix became
empty as hydration occurred from the microstructure. The pores of the cement matrix
were considerably smaller than those of pre-wetted LWAs; thus, capillary suction
may occur. Internal curing water was transported from the pre-wetted LWAs into the
pores of the cement matrix and weakened self-desiccation, which in turn decreased
the autogenous shrinkage of the concrete. When a small amount of LWAs was added,
only a small beneficial effect was observed. This phenomenon may be due to the fact
that the water within LWAs can only move approximately 1.8 mm into the paste
surrounding the aggregate particles; LWAs replaced a part of coarse normal-weight
aggregates; therefore the spacing between neighboring particles containing water
was relatively far, which resulted in large distances of water migration from the
LWAs into surrounding cement paste matrix. Consequently, the LWAs were spaced
too far apart to effectively supply water to all the paste, or the internal curing water
supply was depleted too early in the hydration process [38].
Figure 6.17 shows the autogenous shrinkage development of the four mixtures.
Mixture LWA-10 had a smaller improvement in autogenous shrinkage than
Mixture LWA-0. Mixture LWA-50 exhibited a drastic improvement with initial
expansion, which was maintained thereafter. While the mechanisms of pre-wetted
LWAs contributing to a reduction in autogenous shrinkage are relatively well known,
the mechanisms leading to early-age expansion are not fully understood and seem to
occur with concretes of different w/c ratios [19]. A possible reason for this expansion
could be the reabsorption of bleeding water; removing bleeding water reduces the

@seismicisolation
@seismicisolation
296 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

100

50
Autogenous shrinkage (με)
0

-50

-100

-150 LWA-0
LWA-10
-200 LWA-30
LWA-50
-250
0 12 24 36 48 60
Age (h)

Fig. 6.17 Autogenous shrinkage strain measured in free shrinkage concrete specimens

expansion but does not eliminate it totally. The residual expansion could be due to
internal bleeding in the mixture. Another possible reason for this phenomenon can
be found at the scale of the hydration products. Although the reaction products have
a lower volume than the reagents, they form a spatial network because of their shape.
Growth of further reaction products inside the network generates an internal pressure
that may cause moderate expansion of the system [39].
Figure 6.17 also shows that most of the autogenous shrinkage of concrete occurred
within 24 h, with only limited changes thereafter. An important requirement to miti-
gate concrete autogenous shrinkage is that shrinkage prevention measures must be
in effect 24 h after casting concrete. The thermal cracking performance of the spec-
imens improved with a reduction in their autogenous shrinkage [36]; thus, thermal
cracking performance increased with the increase in the amount of pre-wetted LWAs.

6.4.4 Restrained Stress of Concrete with LWAs

The development of restrained stress was measured in TSTM. Figure 6.18 shows the
tensile stress development over time in all the restrained concrete specimens.
The concrete was in compression when the peak temperature was reached because
the expansion of the concrete caused by the temperature rise was restrained. When
the peak temperature was reached, the hydrating paste was still a developing struc-
ture, and its strength was low. Early-age relaxation may occur when the concrete is
subjected to high compressive stress. The phenomenon of gradual decrease in stress
when a material is subjected to sustained strain is called stress relaxation. Figure 6.19

@seismicisolation
@seismicisolation
6.4 Cracking Resistance of Concrete Internally Cured with LWAs Under … 297

3.0

2.5 LWA-0
2.0 LWA-10
1.5
LWA-30
LWA-50
Stress (MPa)

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0
0 24 48 72 96 120
Age (h)

Fig. 6.18 Concrete stress measured in restrained concrete specimens

shows that restrained stress was gradually relaxed when the concrete temperature was
subsequently kept constant for 48 h.
The zero-stress temperature was not equal to the initial temperature because the
high relaxation property of young concrete causes most of the restrained stress to be
relaxed, as shown in Fig. 6.19. After the specimens were cured at the peak temperature
for 48 h, they were cooled at 1.0 °C/h, and Tzero−stress was reached. After zero-
stress temperature Tzero−stress , any temperature drop and shrinkage deformations, if
restrained, will generate tensile stress.
The cracking stresses of concrete were 2.8, 2.4, 2.3, and 1.8 MPa when the amounts
of pre-wetted LWAs were 0, 10%, 30%, and 50% for Mixture LWA-0, LWA-10,
LWA-30, and LWA-50, respectively. The cracking stress of ICC decreased with the
increase in the amount of pre-wetted LWAs. The shrinkage stress also decreased
in the ICC with pre-wetted LWAs [40]. For subsequent analysis, the experimental
results of concrete cracking stress were converted from actual time to equivalent age
at 20 °C by the temperature condition in the TSTM test.
The test result showed that the compressive elastic modulus of concrete at 28 d
were 34.5, 33.5, 31.0, and 28.6 GPa for Mixture LWA-0, LWA-10, LWA-30, and
LWA-50, respectively. The tensile elastic modulus should be used in the case of
shrinkage, which causes tensile stress in the restrained specimen in the TSTM test;
the compressive elastic modulus should be used in the case of expansion, which
induces compression in TSTM [20]. The compressive elastic modulus measured on
prisms was employed as tensile and compressive elastic modulus because early-age
secant elastic moduli in compression and tension do not differ significantly.
Adding pre-wetted LWAs to concrete influences its elastic modulus in two ways.
First, the addition of low-stiffness LWAs reduces the elastic modulus of the composite

@seismicisolation
@seismicisolation
298 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Fig. 6.19 Development of

free strain and stress of
concrete with different
mixtures

@seismicisolation
@seismicisolation
6.4 Cracking Resistance of Concrete Internally Cured with LWAs Under … 299

Fig. 6.19 (continued)

[41]. Second, the internal curing water stored inside LWAs increases cement hydra-
tion and thereby increases the elastic modulus of the concrete. The net effect generally
reduced the elastic modulus. The test result showed that the splitting tensile strengths
of concrete at 28 d were 6.3, 5.3, 4.5, and 2.7 MPa for Mixture LWA-0, LWA-10,
LWA-30, and LWA-50, respectively.
The test result showed that the ratios of cracking stress to tensile strength were
0.60, 0.62, 0.66, and 0.85 when the amounts of the pre-wetted LWAs were 0, 10%,
30%, and 50% for Mixture LWA-0, LWA-10, LWA-30, and LWA-50, respectively.
All the specimens cracked at a ratio of cracking stress to tensile strength less than
1.0; this result may be due to the fact that the rapid loading rate of the splitting
tensile test provides a higher measured tensile strength than the true concrete tensile
strength when loaded slowly, as was the case with thermal stresses. In addition,
internal microcracking may occur because of the restraint of TSTM offered to the
shrinking paste by the non-shrinking aggregates. And the restrained specimen in
TSTM would be pulled and pressed frequently to keep the restraint degree 100%.
Thus the microcracking was also induced by pulling or pressing the specimen and
thermal loading. The ratio of cracking stress to tensile strength increased with the
increase in the amount of pre-wetted LWAs. The elastic modulus of LWAs was lower
than that of normal aggregates. The low modulus of the LWAs helped reduce the
probability of internal microcracking. Thus, the probability of internal microcracking
decreased with the increase in the amount of pre-wetted LWAs. Stress homogeneity
in ICC with pre-wetted LWAs caused by the limited elastic mismatch between the
LWAs and the cement paste might reduce the occurrence of internal microcracking.
Stress homogeneity was improved with the increase in the amount of pre-wetted
LWAs.
The rate of tensile stress was utilized to evaluate the cracking resistance of ICC
with pre-wetted LWAs [36]. When the cooling began, thermal deformation was
restrained in the fully restrained TSTM test, resulting in the development of tensile
stress in the specimen. The rate of stress development decreased with the increase in
the amount of pre-wetted LWAs.

@seismicisolation
@seismicisolation
300 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

0.09

Average rate of stress development

0.08

(MPa/h) 0.07

0.06

0.05

0.04

0.03
0 10 20 30 40 50
Percent lightweight aggregate(%)

Fig. 6.20 Relationship between average rate of tensile stress and the amount of pre-wetted LWAs

Figure 6.20 shows that the average rates of tensile stress were 0.08, 0.06, 0.06,
and 0.05 MPa/h when the amounts of pre-wetted LWAs were 0, 10%, 30%, and 50%
for Mixtures LWA-0, LWA-10, LWA-30, and LWA-50, respectively.
The rate of tensile stress decreased when the amount of pre-wetted LWAs increased
from 0 to 10%, 30%, and 50%. The differences in the average rate of tensile
stress development of ICC with pre-wetted LWAs were attributed to different age-
dependent elastic modulus, stress relaxation, and autogenous shrinkage development
of concrete. The elastic modulus decreased with the increase in the amount of pre-
wetted LWAs. The reduced elastic modulus had a significant influence on stress devel-
opment; therefore, the observed average rate of tensile stress development decreased
with a reduction in the elastic modulus of ICC.
The cracking age of ICC was also utilized to evaluate the effect of pre-wetted
LWAs on cracking resistance. The cracking ages of concrete were 109, 115, 111,
and 120 h for Mixture LWA-0, LWA-10, LWA-30, and LWA-50, respectively. The
cracking age of ICC with pre-wetted LWAs was longer than that of concrete without
pre-wetted LWAs. The longest cracking age indicated the highest cracking resistance.
Thus, the cracking resistance of Mixture LWA-50 was the highest based on the result
of cracking age. The cracking age of ICC increased with the increase in the amount
of pre-wetted LWAs. Although the amount of LWAs for Mixture LWA-30 was larger
than that for Mixture LWA-10, the cracking age for Mixture LWA-30 was shorter
than that for Mixture LWA-10. Pre-wetted LWAs reduced the shrinkage and tensile
strength of ICC. A possible reason is that the effect of pre-wetted LWAs on the
decrease in the tensile strength of concrete was larger than that on the decrease in
autogenous shrinkage.

@seismicisolation
@seismicisolation
6.4 Cracking Resistance of Concrete Internally Cured with LWAs Under … 301

6.4.5 Creep of Concrete with LWAs

In this experiment, the specimens displayed initial expansion that resulted in

compressive stress. The maximum restrained compressive stresses of Mixture LWA-
0, LWA-10, LWA-30, and LWA-50 were 0.81, 0.84, 1.1, and 1.8 MPa, respectively.
The maximum restrained compressive stress of ICC increased with the increase in
the amount of pre-wetted LWAs.
Creep strain developed differently depending on the extent of restrained expansion
occurring in the early-age ICC specimens in TSTM. The maximum compressive
creeps for Mixture LWA-0, LWA0-10, LWA-30, and LWA-50 were 192, 215, 238,
and 251 με, corresponding to 83%, 105%, 106%, and 117% of the highest free
expansive strains, respectively. The maximum compressive creep increased with the
increase in the amount of pre-wetted LWAs in the specimens in the compressive
period in the TSTM test. When an increasing amount of pre-wetted LWAs was used
in the ICC, a large compression creep strain developed in the compression period
because of the significant restrained expansion and reduced compressive stress.
Figure 6.21 shows that all creep strains were negative, with negative increments
in creep strain obtained during restrained expansion and positive increments in
creep strain obtained during the restrained shrinkage mode in the cooling period for
Mixtures LWA-10, LWA-30, and LWA-50 with pre-wetted LWAs. All creep strains
were negative for ICC with 20% pre-wetted LWAs of the total sand content.
In the cooling period, the specific tensile creeps of ICC at the cracking time were
71, 44, 33, and 50 με/MPa for Mixture LWA-0, LWA-10, LWA-30, and LWA-50,
respectively. The specimens with pre-wetted LWAs offered an obvious reduction in
specific tensile creep. The ICC with pre-wetted LWAs exhibited less specific creep

50

-50
Creep strain(με)

LWA-0
-100 LWA-10
LWA-30
-150 LWA-50

-200

-250
increase in
contant temperature (48 h) increase in tensile creep
compressive creep
-300
0 24 48 72 96 120
Age (h)

Fig. 6.21 Creep strain calculated from the data of restrained autogenous shrinkage tests

@seismicisolation
@seismicisolation
302 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

than the Mixture LWA-0 without pre-wetted LWAs, and the specific creep of LWAs
remained less than that of normal weight aggregate.
The reason pre-wetted LWAs reduced the specific tensile creep of ICC was that
hydration was enhanced. Internal curing water for hydration and high-water activity
lead to increased cement hydration of concrete with a low w/c ratio. In contrast
to specimens without internal curing water, the microstructures of ICC contained
fewer and smaller unhydrated cement particles (indicating enhanced hydration) and
a denser and more homogeneous interfacial transition zone between the LWAs and
hydrating cement paste, as observed in previous studies on lightweight concrete.
Thus, the cementitious paste microstructure could have become denser, stronger,
and better able to resist creep. The tensile stress of ICC decreased with the increase
in the amount of pre-wetted LWAs, which also leads to small tensile creep in the
tensile period.
However, previous experimental studies have established opposite conclusions.
Cusson and Hoogeveen measured a moderate increase in the tensile creep coefficient
of w/c = 0.34 concrete mixtures with pre-wetted LWAs measured at 7 d versus a
controlled mixture [42].
The creep of mortar materials containing LWAs is generally expected to increase
with the increase in the amount of pre-wetted LWAs because of the low stiffness.
However, an increase in the cement hydration of mortar materials caused by internal
curing water results in a reduction in creep. As such, the net effect of pre-wetted
LWAs on the creep of mortar materials depends on both the amount of low-stiffness
LWAs and the beneficial contribution of internal curing water to cement hydration.

6.4.6 Cracking Resistance of Concrete with LWAs

The results of the evaluation of the cracking resistance of the concrete mixtures are
shown in Fig. 6.22.
Integrated criterion ϕN was 0.060, 0.040, 0.040, and 0.025 MPa/day2 for Mixture
LWA-0, LWA-10, LWA-30, and LWA-50, respectively. When the amount of LWAs
increased from 0 to 10%, 30%, and 50%, the integrated criterion decreased at rates
of 33%, 33%, and 58%, respectively. Concrete with a large amount of pre-wetted
LWAs exhibited a high cracking resistance. The increment in cracking resistance is
related to the reduced elastic modulus of ICC.

6.4.7 Summary

The experimental findings on the effect of pre-wetted LWAs on the behavior and
cracking resistance of ICC under adiabatic conditions were investigated. Analysis
was based on temperature difference, strain difference, restrained stress induced
under a restrained condition, tensile and compressive creeps, and induced stress

@seismicisolation
@seismicisolation
6.4 Cracking Resistance of Concrete Internally Cured with LWAs Under … 303

0.07
Cracking Resistance by
Integrated Criterion

Integrated criterion (MPa/day2)

0.06
Low
0.05

0.04

0.03

0.02
High

0.01

0.00
LWA-0 LWA-10 LWA-30 LWA-50

Fig. 6.22 Estimation of cracking resistance using integrated criterion

relaxation with four groups of different pre-wetted LWAs mixture proportions of ICC.
This research quantified the amount of pre-wetted LWAs on stress development and
cracking behavior in the restrained specimen test. The following conclusions were
drawn from the obtained results.
(1) Both the temperature rise and drop of ICC were enhanced when the amount of
pre-wetted LWAs increased from 0 to 10%, 30%, and 50%. The temperature rises
of concrete were 31.0, 33.0, 33.0, and 37.0 °C when the amounts of pre-wetted
LWAs were 0, 10%, 30%, and 50% for mixtures LWA-0, LWA-10, LWA-30,
and LWA-50, respectively. The reason was that the hydration degree increased
with the increase in the amount of pre-wetted LWAs. The temperature drops of
concrete at the age of cracking were 37.0, 43.0, 43.0, and 47.0 °C for Mixture
LWA-0, LWA-10, LWA-30, and LWA-50, respectively.
(2) The autogenous shrinkage of ICC decreased with the increase in the amount
of pre-wetted LWAs. The autogenous shrinkage of early-age concrete at an age
of 60 h was − 116, − 101, − 43, and 56 με when the amounts of pre-wetted
LWAs were 0, 10%, 30%, and 50% for Mixture LWA-0, LWA-10, LWA-30, and
LWA-50, respectively. The reason was that internal curing water transported to
the emptied pore of cement matrix weakened self-desiccation.
(3) The cracking stress of ICC decreased with the increase in the amount of pre-
wetted LWAs. The cracking stresses of concrete were 2.8, 2.4, 2.3, and 1.8 MPa
when the amounts of pre-wetted LWAs were 0, 10%, 30%, and 50% for Mixture
LWA-0, LWA-10, LWA-30, and LWA-50, respectively.
(4) Tensile stress rate decreased when the amount of pre-wetted LWAs increased
from 0 to 10%, 30%, and 50%. The tensile stress rates were 0.08, 0.06, 0.06,
and 0.05 MPa/h when the amounts of pre-wetted LWAs were 0, 10%, 30%, and
50% for Mixture LWA-0, LWA-10, LWA-30, and LWA-50, respectively.

@seismicisolation
@seismicisolation
304 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

(5) The cracking age of ICC increased when the amount of pre-wetted LWAs
increased from 0 to 10%, 30%, and 50%. The cracking ages were 109, 115,
111, and 120 h when the amounts of pre-wetted LWAs were 0, 10%, 30%, and
50% for Mixture LWA-0, LWA-10, LWA-30, and LWA-50, respectively. The
cracking age of Mixture LWA-30 was 111 h, which was lower than that of
Mixture LWA-10. This result may be due to the influence of low tensile strength
and high internal curing effect.
(6) The compressive creep of ICC increased with the increase in the amount of
pre-wetted LWAs. The highest compressive creeps were 192, 215, 238, and
251 με for Mixture LWA-0, LWA-10, LWA-30, and LWA-50, respectively. The
specific tensile creep of the concrete with pre-wetted LWAs decreased compared
with that of the concrete without pre-wetted LWAs. The specific tensile creeps
were 71, 44, 33, and 50 με/MPa for mixtures LWA-0, LWA-10, LWA-30, and
LWA-50, respectively. The compressive and tensile creeps of concrete occurred
because of the restrained condition.
(7) The ratio of cracking stress to tensile strength in ICC increased with the increase
in the amount of pre-wetted LWAs because of the low elastic modulus. The ratios
of cracking stress to tensile strength were 0.60, 0.62, 0.66, and 0.85 when the
amounts of pre-wetted LWAs were 0, 10%, 30%, and 50% for Mixtures LWA-
0, LWA-10, LWA-30, and LWA-50, respectively. The ratios of cracking stress
to tensile strength were all lower than 1.0 because of the higher loading rate
in the tensile strength test than the development of thermal stress and internal
microcracking.
(8) The cracking resistance of ICC with pre-wetted LWAs was improved, as
determined from the integrated criterion for the assessment of cracking
resistance.

6.5 Influential Factors of Cracking Resistance of Internally

Cured Concrete

6.5.1 Addition of Ground Granulated Blast Furnace Slag

1. Mixture proportions and materials

Five groups of concrete mixtures were studied. Table 6.3 shows the concrete mixture
proportions which were designed as GS00, IGS00, IGS20, IGS35, and IGS50.
Mixture GS00 was the reference concrete without SAPs and GGBFS. Mixtures
IGS00, IGS20, IGS35, and IGS50 referred to the HPC internally cured with SAPs
and the replacement of GGBFS was 0%, 20%, 35%, and 50% by weight of cement,
respectively. The basic w/b ratio of all mixtures was 0.32. The fine aggregate, coarse
aggregate used in all mixtures remained constant.
The calculated theoretical amount of internal curing water required to realize the
maximum degree of cement hydration was 33.66 kg/m3 and the theoretical amount

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 305

Table 6.3 Concrete mixture proportions

Mixture Composition Concrete Mixtures
GS00 IGS00 IGS20 IGS35 IGS50
Water (kg/m3 ) 173 173 173 173 173
Cement (kg/m3 ) 541.0 541.0 432.8 351.6 270.5
GGBFS (kg/m3 ) 0 0 108.2 189.4 270.5
Fine aggregate(sand) (kg/m3 ) 625 625 625 625 625
Coarse aggregate (kg/m3 ) 1111 1111 1111 1111 1111
SAPs (kg/m3 ) 0 2.589 2.589 2.589 2.589
Superplasticizer (%) 0.80 0.80 0.75 0.70 0.70
Slump (mm) 90 110 115 120 125

of SAPs was 2.589 kg/m3 for all ICC mixtures. Therefore, the internal curing w/b
ratio of all ICC mixtures was 0.06 and the total w/b ratio was 0.38.
2. Modulus and strength
The tests results showed that the cubic compressive strength at 28 d was 75.6,
61.9, 56.1, 50.1, and 45.5 MPa for mixtures GS00, IGS00, IGS20, IGS35, and
IGS50, respectively. The compressive strength for mixture IGS00 with the addition
of SAPs decreased by 18.1% compared with that for mixture GS00 without SAPs.
The compressive strength decreased by 9.4%, 19.1%, and 26.5% when the GGBFS
content increased from 0 to 20%, 35%, and 50%, respectively. The tensile strength
and modulus of concrete also decreased with the increase of GGBFS content, as
shown in Fig. 6.23.
The elastic modulus and splitting tensile strength of ICC decreased with the
increase of GGBFS content. The decrease of strength was mainly due to that the
hydration products would decrease greatly with the increasing partial replacement
of Portland cement with GGBFS.
3. Temperature process
Figure 6.24 shows the temperature history of mixtures GS00, IGS00, IGS20, IGS35,
and IGS50, respectively. The temperature of concrete rose rapidly with the cement
hydration under adiabatic condition. The temperature rise time or the highest temper-
ature were 15.2, 26.9, 28.1, 38.0, and 38.6 h, or 51.18, 51.04, 48.79, 43.73, and
42.69 °C for mixtures GS00, IGS00, IGS20, IGS35, and IGS50, respectively. There-
fore, the calculated temperature rise was 31.67, 35.44, 33.50, 31.44, and 29.19 °C
for mixtures GS00, IGS00, IGS20, IGS35, and IGS50, respectively. The tempera-
ture rise for mixture IGS00 with the addition of SAPs increased by 11.9% compared
with that for mixture GS00 without SAPs. The temperature rise decreased by 5.5%,
11.3%, and 17.6% when the GGBFS content increased from 0 to 20%, 35%, and
50% for mixtures IGS00, IGS20, IGS35, and IGS50, respectively. In addition, the
rate of temperature rise decreased gradually with the increase of GGBFS content at
early-age.

@seismicisolation
@seismicisolation
306 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Fig. 6.23 Experimental results and predicting models of elastic modulus and axial tensile strength
for different concrete mixtures

Higher adiabatic temperature rise than reference concrete mixture GS00 was
found in mixture IGS00. The temperature rise of ICC decreased with the increase of
GGBFS content. The heat development in concrete with GGBFS is slower than that
of ordinary concrete. The results on the temperature rise also indicated that the total
cement hydration heat decreased with the increase of GGBFS content.

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 307

50 GS00
IGS00
IGS20
40 IGS35
IGS50
)
Temperature (

30

20

10

0
0 20 40 60 80 100 120
Age (h)
(a) Actual temperature history

40
GS00
IGS00
30
IGS20
IGS35
)

20 IGS50
(

10

-10

0 20 40 60 80 100 120
Age (h)
(b) Absolute temperature history

Fig. 6.24 Adiabatic temperature history of restrained shrinkage specimens

Under the same temperature history, the temperature drop could directly reflect
the cracking resistance of concrete in the forced cooling period. The temperature
drop of concrete under adiabatic condition could be calculated by Eq. (6.6).
Figure 6.24 shows the temperature of mixtures GS00, IGS00, IGS20, IGS35,
and IGS50 at the cracking age was 9.62, 7.53, 14.04, 21.19, and 25.44 °C, respec-
tively. Therefore, the calculated temperature drop was 41.56, 43.51, 34.75, 22.54,

@seismicisolation
@seismicisolation
308 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

and 17.25 °C for mixtures GS00, IGS00, IGS20, IGS35, and IGS50, respectively.
The temperature drop for mixture IGS00 with the addition of SAPs increased by
4.7% compared with that for mixture GS00 without SAPs. The temperature drop
decreased by 20.1%, 48.2%, and 60.4% when the GGBFS content increased from 0
to 20%, 35%, and 50% for mixtures IGS00, IGS20, IGS35, and IGS50, respectively.
The temperature drops of ICC with GGBFS decreased greatly, which indicated that
the cracking resistance of ICC decreased with the increase of GGBFS content.
4. Free deformation
Figure 6.25 shows the total deformation of the free shrinkage specimens under adia-
batic condition. The total deformation of concrete continuously increased with the
increase of time after pouring. The total deformation measured in the free shrinkage
specimen for mixture IGS00 with the addition of SAPs increased by 19.2% compared
with that for mixture GS00 without SAPs. The total deformation measured in the
free shrinkage specimen decreased by 7.7%, 16.1%, and 18.2% when the GGBFS
content increased from 0 to 20%, 35%, and 50% for mixtures IGS00, IGS20, IGS35,
and IGS50, respectively. The results on the rate of deformation growth correlated
with that of temperature rise. The temperature deformation remained unchanged in
the constant temperature period. Therefore, the total deformation of concrete without
GGBFS for mixtures IGS00 remained unchanged due to the slow development of
autogenous shrinkage. The autogenous shrinkage of ICC with GGBFS developed
rapidly, and the total deformation decreased obviously in the constant temperature
period. The negative sign indicated that the total deformation of the free shrinkage
specimen was shrinkage.

150
GS00
IGS00
100
IGS20
Total deformation (με)

IGS35
50 IGS50

-50

-100
0 20 40 60 80 100 120
Age (h)

Fig. 6.25 Evolution of total deformation measured in the free shrinkage specimens

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 309

GS00
0 IGS00
IGS20
IGS35
Autogenous shrinkage (με) -50 IGS50

-100

-150

-200

-250

-300
0 20 40 60 80 100 120
Age (h)

Fig. 6.26 Evolution of autogenous shrinkage measured in the free shrinkage specimens

Figure 6.26 shows the development of autogenous shrinkage of mixtures GS00,

IGS00, IGS20, IGS35, and IGS50 under adiabatic condition. The negative sign indi-
cated that the autogenous shrinkage of the free shrinkage specimen was shrinkage.
The experimental results showed that the autogenous shrinkage of concrete decreased
with the addition of SAPs. Meanwhile, the growth rate of autogenous shrinkage of
ICC with GGBFS was lower than that of mixture GS00 during the first day. The auto-
genous shrinkage of mixture IGS20 was lower than that of mixture GS00 at early-age.
The absolute value of autogenous shrinkage for mixture IGS00 with the addition of
SAPs decreased by 63.2% compared with that for mixture GS00 without SAPs at
the age of 90 h. The absolute value of autogenous shrinkage increased by 141.6%,
222.1%, and 246.8% at the age of 90 h when the GGBFS content increased from 0
to 20%, 35%, and 50% for mixture IGS00, IGS20, IGS35, and IGS50, respectively.
The addition of SAPs significantly decreased the autogenous shrinkage of HPC.
The early-age autogenous shrinkage of ICC increased with the increase of GGBFS
content. The autogenous shrinkage of concrete with GGBFS is higher than that of
ordinary Portland cement concrete, which may be due to that the chemical shrinkage
of concrete with GGBFS is larger than that of ordinary Portland cement concrete.
Therefore, the high chemical shrinkage contributes to the large self-desiccation which
leads to the large autogenous shrinkage. In addition, the test of permeability confirms
that the addition of GGBFS contributes to the fine pore structure of the cement paste,
which helps to reduce the relative humidity and increases the self-desiccation degree
inside the cement paste.
5. Restrained stress
When the early-age shrinkage deformation of concrete is restrained, tensile stress
will occur inside the concrete. Cracking may occur when the restrained tensile

@seismicisolation
@seismicisolation
310 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

3.5

3.0
GS00
Restrained stress (MPa) 2.5 IGS00
IGS20
2.0
IGS35
1.5 IGS50

1.0

0.5

0.0

-0.5
Stress relaxation
-1.0
0 20 40 60 80 100 120
Age (h)

Fig. 6.27 Development of restrained stress of concrete measured in restrained shrinkage specimens

stress reaches to or exceeds the tensile strength of restrained shrinkage specimen.

Figure 6.27 shows the restrained stress development of all restrained shrinkage
specimens.
The results showed that the maximum compressive stress induced by thermal
expansion was 0.84, 0.91, 0.82, 0.65, and 0.61 MPa, and the appearance time was
12.0, 17.2, 18.4, 27.1, and 27.6 h for mixtures GS00, IGS00, IGS20, IGS35, and
IGS50, respectively. Experimental results on compressive stress correlated with the
temperature history. The rate of temperature rise and the highest temperature of
ICC decreased with increase of GGBFS content, which resulted in the decrease of
thermal expansion. Therefore, the corresponding thermal expansion stress of ICC
also decreased with the increase of GGBFS content. The restrained tensile stress rate
could be used to evaluate the cracking resistance of concrete. The high restrained
tensile stress rate contributed to the cracking of concrete and decreased the cracking
resistance of concrete. In the forced cooling period, the corresponding tensile stress
rate of mixtures GS00, IGS00, IGS20, IGS35, and IGS50 was 1.64, 1.32, 1.37, 1.43,
and 1.53 MPa/day, respectively. The tensile stress rate for mixture IGS00 with the
addition of SAPs decreased by 19.5% compared with that for mixture GS00 without
SAPs. The tensile stress rate increased by 3.8%, 8.3%, and 15.9% when the GGBFS
content increased from 0 to 20%, 35%, and 50% for mixtures IGS00, IGS20, IGS35,
and IGS50, respectively.
The expansion due to temperature rise led to the occurrence of compressive stress
in the restrained shrinkage specimens, as shown in Fig. 6.27. When the highest
temperature was reached, the strength of concrete which was in a state of flux was
still low. Therefore, stress relaxation of concrete may occur due to high restrained
compressive stress and creep. The compressive stress of specimen IGS00 decreased

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 311

from 0.91 to 0.07 MPa, and then kept constant. Tensile stress occurred due to the
increase of the autogenous shrinkage in the constant temperature period. At the
end of constant temperature period, the maximum tensile stress was 0.18, 0.54, and
0.59 MPa when the GGBFS content was 20%, 35%, and 50% for mixtures IGS20,
IGS35, IGS50, respectively. The maximum tensile stress in the constant temperature
period increased with the increase of GGBFS content.
In the forced cooling period, the stress at the temperature when the concrete is
poured is called the stress at room temperature. Figure 6.24 shows that the cracking
temperature of the mixtures GS00, IGS00 and IGS20 fell below the room temper-
ature, the cracking temperature of mixtures IGS35 and IGS50 did not fall to the
room temperature. The stress at room temperature was 2.81, 2.54, and 2.89 MPa
for mixtures GS00, IGS00, and IGS20, respectively. The cracking age was 107.0,
117.0, 109.0, 104.0, and 96.0 h for mixtures GS00, IGS00, IGS20, IGS35, and
IGS50, respectively. The experimental result of cracking age could also show that
the cracking resistance of ICC decreased with the increase of GGBFS content.
The cracking resistance of ICC with GGBFS could also be evaluated by using the
cracking stress, tensile strength, and the ratio of the cracking stress to tensile strength.
The equivalent age of the restrained shrinkage specimens was 260.0, 279.0, 263.0,
212.0, and 206.0 h at the cracking age of 107.0, 118.0, 111.0, 102.0, and 96.0 h for
mixtures GS00, IGS00, IGS20, IGS35, and IGS50, respectively. Figure 6.28 shows
that the cracking stress or axial tensile strength at the equivalent age was 3.21, 2.99,
2.93, 2.45, and 2.15 MPa, or 4.06, 3.34, 3.12, 2.85, and 2.44 MPa for mixtures GS00,
IGS00, IGS20, IGS35, and IGS50, respectively. The experimental result showed that
the cracking resistance of ICC decreased slightly when the GGBFS content increased
from 0 to 20%, the cracking resistance of ICC decreased severely once the GGBFS
content exceeded 20%. The ratio of the cracking stress to tensile strength was 0.79,
0.90, 0.94, 0.86, and 0.88 for mixtures GS00, IGS00, IGS20, IGS35, and IGS50,
respectively.
The increase of restrained tensile stress rate may due to the increase of autogenous
shrinkage, which indicated that the early-age cracking resistance of ICC decreased
with the increase of GGBFS content. The cracking stress of the restrained shrinkage
specimens are generally less than the tensile strength at the same equivalent age. Two
reasons can account for the results. One is that the loading speed of the splitting test
is too fast, which results in high tensile strength of concrete. Another is that for the
purpose of achieving 100% degree of restraint, the stepper motor should be started
frequently to push or pull the specimen back to the initial position when it produces
a small deformation during the TSTM test. This repeated compression and tension
process may cause varying degrees of damage on the internal structure of concrete.
6. Cracking resistance
The restrained tensile stress rate or net time of cracking was 1.64, 1.32, 1.37, 1.43,
and 1.53 MPa/day or 2.35, 2.38, 1.97, 1.33, and 0.98 days for mixtures GS00, IGS00,
IGS20, IGS35, and IGS50, respectively. Therefore, the integrated criterion calculated
by Eq. (6.14).was 0.698, 0.555, 0.694, 1.079, and 1.563 MPa/day2 for mixtures GS00,
IGS00, IGS20, IGS35, and IGS50, respectively, as shown in Fig. 6.29. The low

@seismicisolation
@seismicisolation
312 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

5 5
Fig. 6.28 Development of
Tensile strength GS00
tensile strength and Restrained stress
restrained stress of different 4 4
concrete mixtures

Restrained stress (MPa)

Tensile strength (MPa)

3 3

2 2

1 1

0 0

-1 -1
0 20 40 60 80 100 120
Age (h)
(a) GS00
5 5
Tensile strength IGS00
4 Restrained stress 4
Restrained stress (MPa)

Tensile strength (MPa)

3 3

2 2

1 1

0 0

-1 -1
0 20 40 60 80 100 120
Age (h)
(b) IGS00
5 5
Tensile strength IGS20
Restrained stress
4 4
Restrained stress (MPa)

Tensile strength (MPa)

3 3

2 2

1 1

0 0

-1 -1
0 20 40 60 80 100 120
Age (h)
(c) IGS20

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 313

Fig. 6.28 (continued) 5 5

Tensile strength IGS35
4
Restrained stress 4

Restrained stress (MPa)

Tensile strength (MPa)

3 3

2 2

1 1

0 0

-1 -1
0 20 40 60 80 100 120
Age (h)
(d) IGS35
5 5
Tensile strength IGS50
Restrained stress
4 4
Restrained stress (MPa)

Tensile strength (MPa)

3 3

2 2

1 1

0 0

-1 -1
0 20 40 60 80 100 120
Age (h)
(e) IGS50

integrated criterion indicates the high cracking resistance of concrete. The integrated
criterion also indicated that the cracking resistance of HPC increased obviously with
the addition of SAPs. The cracking resistance of ICC based on the integrated criterion
decreased by 25.0%, 94.4%, and 181.6% when the GGBFS content increased from 0
to 20%, 35%, and 50% for mixtures IGS00, IGS20, IGS35, and IGS50, respectively.
Meanwhile, the cracking resistance of mixture IGS20 with 20% of GGBFS content
was better than that for mixture GS00 without GGBFS and SAPs. Therefore, the
excessive addition of GGBFS would decrease the cracking resistance of ICC.
The addition of SAPs or GGBFS increases or decreases the cracking resistance
of HPC due to the decrease or increase of autogenous shrinkage. When the content
of GGBFS was larger than 20%, the influence of GGBFS on the decreasing degree
was larger than that of SAPs on the increasing degree of cracking resistance of HPC.
Therefore, the cracking resistance of ICC decreased with the increase of GGBFS
content when the GGBFS content exceeded 20%.

@seismicisolation
@seismicisolation
314 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

2.0
Cracking Resistance Using Integrated Criterion

Integrated criterion (MPa/day2) 1.5

1.0

0.5

0.0
GS00 IGS00 IGS20 IGS35 IGS50

Fig. 6.29 Estimation of cracking resistance using integrated criterion of different concrete mixtures

6.5.2 w/c Ratios

1. Mixture proportions and materials

Six concrete mixture designs were evaluated. Each mixture was identified by a label
X–Y. The first part of the specimen name X referred to type of concrete (i.e., NC =
normal weight concrete; IC = internal curing concrete), Y indicated the basic w/c
ratio of concrete (i.e., 33 = 0.33 (basic w/c ratio); 40 = 0.40 (basic w/c ratio); 50
= 0.50 (basic w/c ratio)). Mixtures NC-33, NC-40, and NC-50 were the reference
concrete without internal curing and the basic w/c ratio was 0.33, 0.40, and 0.50,
respectively. Mixtures IC-33, IC-40, and IC-50 were the ICC with pre-wetted LWAs
and the basic w/c ratio was 0.33, 0.40, and 0.50, respectively. The LWAs had a 24 h
absorption value of 12% by mass of the dry material. The water absorbed by LWAs
is defined as entrained water, as reported in [43]. The entrained water was 16.1, 16.6,
and 16.8 kg/m3 for mixtures IC-33, IC-40, and IC-50, respectively, as shown in Table
6.4.
Therefore, the entrained w/c ratio was 0.031, 0.037, and 0.042 for mixtures IC-33,
IC-40, and IC-50, respectively. The total w/c ratio is defined as the sum of basic w/c
ratio and entrained w/c ratio. Therefore, the total w/c ratio wt /c was 0.361, 0.437,
and 0.542 for mixtures IC-33, IC-40, and IC-50, respectively. Mixture designs are
shown in Table 6.4. The normal weight coarse aggregates were replaced by pre-
wetted LWAs of the same diameter by volume, and the replacement ratios were all
30% of the total volume of coarse aggregates for the ICC. The cement contents were
not kept constant in concrete mixtures.
Portland Cement (Cement II 52.5R) was employed in accordance with Chinese
National Standard GB 175–2007 [30]. The Blaine fineness of the cement was

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 315

Table 6.4 Concrete mixture proportions and fresh concrete properties

Mixture composition Concrete mixtures
NC-33 NC-40 NC-50 IC-33 IC-40 IC-50
Water (kg/m3 ) 171 180 200 171 180 200
Cement (kg/m3 ) 512 450 400 512 450 400
Fine aggregate(sand) (kg/m3 ) 636 655 666 636 655 666
Coarse aggregate (kg/m3 ) 1131 1165 1184 792 816 829
LWA (kg/m3 , dry) 0 0 0 134 138 140
Internal curing water (kg/m3 ) 0 0 0 16.1 16.6 16.8
Superplasticizer (%) 0.6 0.4 0 0.6 0.4 0

375 m2 /kg and the loss on ignition was 3.11%. The fineness modulus of river sand
used was 2.05 and the maximum size was 1.8 mm. The coarse aggregates used were
crushed limestone and the maximum size was 26 mm. The LWAs used in mixtures
IC-33, IC-40, and IC-50 were manufactured rotary kiln expanded clay. The LWAs
had a dry-bulk density of 1050 kg/m3 . A liquid polycarboxylate-based superplasti-
cizer was added to mixtures NC-33, NC-40, IC-33, and IC-40. Tap water was used
as mixture water. The concrete was mixed for about 2 min.
2. Compressive strength
When internal curing technology is applied to concrete in real structures, it is impor-
tant to know the change of strength [44]. Therefore, predicting the compressive
strength of concrete with both different w/c ratios and amounts of internal curing
materials is necessary. The prediction models have been presented for concrete
without internal curing. However, investigations on the model for predicting the
compressive strength of concrete with internal curing, such as pre-wetted LWAs, are
still lacking. Results in [45] show that the compressive strength of cement paste is
the function of gel/space ratio of cement paste, as shown in Eq. (6.27).

fc = AX n (6.27)

where fc is the cylinder compressive strength of cement paste at 28 d, in MPa; A is

the constant determined by regression analysis and represents the intrinsic strength
of cement gel; n is the constant determined by regression analysis which depends on
the properties of cement; X is the gel/space ratio of cement paste.
Without internal curing agent, the calculation model of gel/space ratio X is defined
in [46], as shown in Eq. (6.28).
⎧
⎪ Vg
⎪
⎨X =
⎪
Vg + Pc
(6.28)
⎪
⎪ Vg = 0.68α
⎪
⎩
Pc = w/c − 0.36α

@seismicisolation
@seismicisolation
316 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

where Vg is the total volume of cement gel and gel pore; α is the degree of hydration;
Pc is the volume of capillary porosity.
To predict the ultimate degree of hydration α∞ of cement paste, Eq. (6.29) is
presented in [47], as shown in Eq. (6.29).

1.031w0
α∞ = (6.29)
0.194 + w0

where w0 is the w/c ratio of cement paste.

The degree of hydration at 28 d is close to the ultimate degree of hydration.
Therefore, to predict the cubic compressive strength of concrete at 28 d, the ultimate
degree of hydration α∞ was replaced by the degree of hydration at 28 d α28 , and
the w/c ratio of cement paste was replaced by the total w/c ratio wt /c, as shown in
Eq. (6.30).

1.031wt /c
α28 = (6.30)
0.194 + wt /c

where α28 is the degree of hydration of concrete at 28 d; wt /c is the total w/c ratio
of concrete.
To predict cubic compressive strength of concrete at 28 d with different w/c ratios,
as shown in Eq. (6.31).

fcu = BX r (6.31)

where fcu is the cubic compressive strength of concrete at 28 d, in MPa; B and r are
the constant determined by regression analysis which depends on the properties of
concrete; X is the gel/space ratio.
Test result of the cubic compressive strength of concrete was 64.2, 55.3, and
44.1 MPa for mixtures NC-33, NC-40, and NC-50, respectively. Equation (6.31)
was utilized to predict cubic compressive strength of normal concrete. Thus, B and
r were determined with regression analysis as 95.0, and 2.00, respectively. Result of
the cubic compressive strength of concrete obtained from Eq. (6.31) was 63.9, 54.9,
and 43.9 MPa for mixtures NC-33, NC-40, and NC-50, respectively. The deviations
between the results from Eq. (6.31) and the test were −0.5%, −0.7%, and −0.5%.
The crushing strength of dry LWAs utilized was 1.2 MPa due to its highly porous
structure. A prediction model for the cubic compressive strength of concrete was
presented based on Eq. (6.31) considering the amounts of pre-wetted LWAs, as
shown in Eq. (6.32).

ficu = fcu · β (6.32)

A decrease of the compressive strength dependent on the amounts of LWAs content

is found in [36] due to the lower strength and modulus of the LWAs. Therefore

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 317

parameter β is defined as the impact factor of the replaced LWAs on the strength of
concrete, which can be determined with regression analysis.
For ICC with pre-wetted LWAs, the total w/c ratio wt /c is shown in Eq. (6.33).

wt /c = wm /c + we /c (6.33)

where wm /c is the basic w/c ratio of concrete; we /c is the entrained w/c ratio of
concrete.
Substituting the w/c ratio with the total w/c ratio wt /c in Eq. (6.33) and the degree
of hydration α with the degree of hydration at 28 d α28 , Eq. (6.34) is obtained.

0.68α28
X = (6.34)
0.32α28 + wm /c + we /c

Test result of the compressive strength of ICC was 51.3, 44.7, and 35.2 MPa for
mixtures IC-33, IC-40, and IC-50, respectively. Parameter β was 0.86 which was
determined with regression analysis, as shown in Fig. 6.30.
Thus, Eq. (6.35) can be obtained and utilized to predict cubic compressive strength
of concrete with or without LWAs, which is shown as follows.
⎧ ⎡   ⎤2.00
⎪
⎪ 0.68 × 1.031(wm /c+we /c)
⎪
⎨ 0.194+wm /c+we /c
fcu = 95.0 × ⎣   ⎦
1.031(wm /c+we /c) (6.35)
⎪ 0.32 × + wm /c + we /c
⎪
⎪
0.194+wm /c+we /c
⎩
ficu = 0.86 · fcu

Fig. 6.30 Relationship between compressive strength and gel/space ratio of concrete

@seismicisolation
@seismicisolation
318 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Fig. 6.31 Total deformation measured in free shrinkage concrete specimens

The values of cubic compressive strength of concrete were 51.0, 43.5, and
36.7 MPa for mixtures IC-33, IC-40, and IC-50, respectively. The deviations between
the results and the test were -0.6%, -2.7%, and 4.3%.
3. Free deformation
The total deformations of free specimens are shown in Fig. 6.31.
The autogenous shrinkage of concrete at early-age is shown in Fig. 6.32, and the
value at the age of 60 h after casting was − 116, − 63, − 44, − 42, − 19, and 15 με
for mixtures NC-33, NC-40, NC-50, IC-33, IC-40, and IC-50, respectively.
The absolute value of autogenous shrinkage of concrete increased by 84.1% or
121.1% when the basic w/c ratio decreased from 0.40 to 0.33 for mixtures NC-40 and
NC-33 or IC-40 and IC-33, respectively. The absolute value of autogenous shrinkage
of ICC with pre-wetted LWAs decreased by 69.8%, and 63.8% when the basic w/c
ratio decreased from 0.40 to 0.33 for mixtures IC-40 and IC-33, compared to that
for mixtures NC-40 and NC-33, respectively. The autogenous shrinkage of mixture
IC-50 even was above 0 lines. Early-age expansion occurred when the autogenous
shrinkage of concrete was above 0 lines. It may be not an important issue from
the view point of engineering, the compressive stress could be easily resisted by the
concrete. However, it is important to know the reason of expansion for understanding
autogenous shrinkage more exactly. Formation of ettringite, reabsorption of bleeding
water, and formation of large crystals of calcium hydroxide may be responsible for
the expansion of concrete at early-age [48].
When external curing water is not available, self-desiccation will occur, and a set
of empty pores will be created within the microstructure. The correlation between
w/c ratio and autogenous shrinkage may be explained by the self-desiccation which

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 319

Fig. 6.32 Autogenous shrinkage measured in free shrinkage concrete specimens

dominates the autogenous shrinkage after the formation of a solid skeleton of the
cement paste. Self-desiccation can be reduced by the use of pre-wetted LWAs which
could provide an entrained water to replace that consumed during hydration. The
entrained water will move from the larger pores in the LWAs into the smaller pores in
the cement paste during hydration [49]. Therefore, the autogenous shrinkage could
be reduced. Furthermore, the effect of the entrained water introduced by the pre-
wetted LWAs was compared with that of the more water of higher basic w/c ratio.
The autogenous shrinkage of concrete with a lower total w/c ratio for mixtures IC-
33 or IC-40 was smaller than that for mixtures NC-40 or NC-50, respectively. This
could clearly show the shrinkage reduction in ICC with pre-wetted LWAs resulted
from the effective internal curing w/c ratio which was due to the water absorption of
LWAs and not from the possible increase in total w/c ratio. The postponed drop of
the internal humidity may be responsible for the reduced shrinkage due to the release
of internal curing water from the pre-wetted LWAs, which reduces the driving force
of autogenuous shrinkage [50].
4. Tensile creep
The basic tensile creep increases with the cement hydration process when the cement
paste is subjected to a sustained stress [51]. As shown in Fig. 6.33, the basic tensile
creep of normal concrete was − 161, − 103, and − 45 με or − 90, − 72, and −
42 με for mixtures NC-33, NC-40, and NC-50 or mixtures IC-33, IC-40, and IC-50,
respectively.
The absolute value of basic tensile creep increased with the decrease of the basic
w/c ratio for both normal concrete and ICC. The absolute value of basic tensile
creep of ICC with pre-wetted LWAs decreased by 6.7%, 30.1%, and 44.1% when

@seismicisolation
@seismicisolation
320 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

0
-20
Basic tensile creep (με) -40
-60
-80
-100 NC-33
-120 NC-40
NC-50
-140 IC-33
-160 IC-40
IC-50
-180
70 80 90 100 110 120 130 140
Age (h)

Fig. 6.33 Basic tensile creep calculated from the results of restrained and free shrinkage

the basic w/c ratio decreased from 0.50 to 0.40, 0.33 for mixtures IC-50, IC-40, and
IC-33, compared to that for mixtures NC-50, NC-40, and NC-33, respectively. And
the decreased degree of the absolute value of basic tensile creep of ICC increased
with the decrease of basic w/c ratio.
The restrained tensile stress or the specific basic tensile creep of concrete at the age
of cracking was 2.7, 2.7, and 2.4 MPa or −60, −38, and −19 με/MPa for mixtures
NC-33, NC-40, and NC-50, respectively. And the specific basic tensile creep for
mixture NC-33 was approximately 1.6 and 3.2 times of that for mixtures NC-40 and
NC-50 at the age of cracking, respectively, as shown in Fig. 6.34.
This tendency indicated that the early-age tensile visco-elastic response in
concrete increased with the decrease of basic w/c ratio under restrained shrinkage
condition. The restrained tensile stress or the specific basic tensile creep of concrete
at the age of cracking was 2.7, 2.4, and 1.8 MPa or −33, −30, and −23 με/MPa
for mixtures IC-33, IC-40, and IC-50, respectively. The absolute value of specific
basic tensile creep increased by 100.0%, and 215.8% or 30.4%, and 43.5%, when
the w/c ratio decreased from 0.50 to 0.40, and 0.33 for mixtures NC-50, NC-40, and
NC-33 or IC-50, IC-40, and IC-33, respectively. The absolute value of specific basic
tensile creep for mixture IC-50 was a little higher than that for mixture NC-50, which
needed to be investigated furthermore. The absolute value of specific basic tensile
creep of ICC with pre-wetted LWAs decreased by 21.1%, and 45.0% when the basic
w/c ratio decreased from 0.40 to 0.33 for mixtures IC-40 and IC-33, compared to
that for mixtures NC-40 and NC-33, respectively.
The total deformation in the cooling period was defined as the shrinkage of
concrete in the cooling period, which consisted of thermal deformation and autoge-
nous shrinkage. Therefore, the shrinkage of concrete in the cooling period was −251,

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 321

Fig. 6.34 Specific basic tensile creep calculated from the results of restrained and free shrinkage

−213, and −160 με or −229, −238, and −219 με for mixtures NC-33, NC-40, and
NC-50 or IC-33, IC-40, and IC-50, respectively. The basic tensile creep/shrinkage of
normal and ICC at the age of cracking both increased with the decrease of basic w/c
ratio, as shown in Fig. 6.8. The basic tensile creep/shrinkage of concrete was 0.28,
0.48, and 0.64 or 0.19, 0.30, and 0.39, which increased by 71.4%, and 128.6% or
57.9%, and 105.3% when the basic w/c ratio decreased from 0.50 to 0.40, and 0.33
for mixtures NC-50, NC-40, and NC-33 or IC-50, IC-40, and IC-33, respectively.
The basic tensile creep/shrinkage of ICC with pre-wetted LWAs decreased by 32.1%,
37.5%, and 39.1% when the basic w/c ratio decreased from 0.50 to 0.40, 0.33 for
mixtures IC-50, IC-40, and IC-33, compared to that for mixtures NC-50, NC-40,
and NC-33, respectively. The decreased degree of basic tensile creep/shrinkage of
ICC increased with the decrease of basic w/c ratio. Both the specific basic tensile
creep and basic tensile creep/shrinkage of ICC with pre-wetted LWAs were lower
than that of the normal concrete when the basic w/c ratios were the same, as shown
in Figs. 6.34 and 6.35.
The enhanced cement hydration of ICC with pre-wetted LWAs may be respon-
sible for the decreased specific basic tensile creep and basic tensile creep/shrinkage.
In contrast to normal concrete, the microstructures of ICC with pre-wetted LWAs
contain fewer and smaller unhydrated cement particles and a denser and more
homogeneous interfacial transition zone between the LWAs and hydrating cement
paste.
As shown in Figs. 6.34 and 6.35, the absolute value of specific basic tensile creep
and basic tensile creep/shrinkage of six mixtures increased firstly, and decreased
thereafter, which were due to the strong effect of aging. For example, the absolute
value of specific basic tensile creep of mixture NC-33 increased before 84 h after

@seismicisolation
@seismicisolation
322 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

1.0
NC-33
0.8 NC-40
Basic tensile creep/shrinkage
NC-50
IC-33
0.6 IC-40
IC-50
0.4

0.2

0.0

70 80 90 100 110 120 130 140

Age (h)

Fig. 6.35 Basic tensile creep/shrinkage calculated from the results of restrained and free shrinkage

casting and thereafter it decreased. The solidification theory is a possible explanation

for this phenomenon. The effective load carrying volume of concrete at early-age
increases rapidly due to hydration. Therefore, the internal stress of concrete at early-
age under constant external load decreases, as hydration is proceeding, because
of the deposition of new solid that participates in carrying the stress. Thus, the
material experiences partial internal unloading due to the stress transferred to the
newly deposited solid structure. This unloading may explain the reversal in creep
strain of mixtures.
Comparing the strength and basic tensile creep behavior of these mixtures, the
higher strength did not necessarily mean lower basic tensile creep. The concrete at
early-age with lower w/c ratio and higher strength showed higher absolute value of
specific basic tensile creep and basic tensile creep/shrinkage. The absolute value of
specific basic tensile creep or the basic tensile creep/shrinkage of mature concrete in
[52] decreases from about 150 to about 100 με/MPa at the age of 40 d or decreases
from 0.29 to 0.17 at the age of 28 d when the w/c ratio decreases from 0.55 to 0.35,
respectively. This may suggest that the basic tensile creep behavior of concrete at
early-age is governed by different factors than in mature concrete [53].
5. Cracking resistance
The development of restrained stress is shown in Fig. 6.36.
The rate of tensile stress became lower and lower with the increase of basic w/c
ratio or the adding of pre-wetted LWAs. Figure 6.37 depicts the relationship between
average rate of tensile stress and the mixtures of concrete. The rate of tensile stress
is used to evaluate the cracking resistance of concrete at early-age [36]. The average
rate of tensile stress was 0.08, 0.07, 0.06, 0.06, 0.05, and 0.04 MPa/h for mixtures
NC-33, NC-40, NC-50, IC-33, IC-40, and IC-50, respectively, as shown in [37].

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 323

3
NC-33
NC-40
2 NC-50
IC-33
IC-40
Stress (MPa)

1 IC-50

-1

-2
0 20 40 60 80 100 120 140
Age (h)

Fig. 6.36 Concrete stress measured in restrained concrete specimens

0.09
0.08
Average rate of tensile stress

0.07
0.06
(MPa/h)

0.05
0.04
0.03
0.02
0.01
0.00
NC-33 NC-40 NC-50 IC-33 IC-40 IC-50

Fig. 6.37 Relationship between average rate of tensile stress and the mixtures of concrete

The average rate of tensile stress increased by 16.7%, and 33.3% or 25.0%, and
50.0% when the basic w/c ratio decreased from 0.50 to 0.40, and 0.33 for mixtures
NC-50, NC-40, and NC-33 or IC-50, IC-40, and IC-33, respectively. The average
rate of tensile stress of ICC with pre-wetted LWAs decreased by 33.3%, 28.6%,
and 25.0% when the basic w/c ratio decreased from 0.50 to 0.40, 0.33 for mixtures
IC-50, IC-40, and IC-33, compared to that for mixtures NC-50, NC-40, and NC-33,
respectively.

@seismicisolation
@seismicisolation
324 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

0.07 Cracking Resistance by

Integrated Criterion Low
0.06
Integrated criterion (MPa/(h⋅day))
0.05

0.04

0.03 High

0.02

0.01

0.00
NC-33 NC-40 NC-50 IC-33 IC-40 IC-50

Fig. 6.38 Estimation of cracking resistance using integrated criterion

Higher integrated criterion means lower cracking resistance. The integrated crite-
rion could also be utilized to investigate the cracking resistance in the TSTM
test.
The net time of cracking or integrated criterion was 1.3, 1.2, 1.1, 1.5, 1.7, and 2.2
d or 0.062, 0.057, 0.054, 0.040, 0.030, and 0.018 MPa/(h · day) for mixtures NC-33,
NC-40, NC-50, IC-33, IC-40, and IC-50, respectively, as shown in Figs. 6.36 and
6.38.
Figure 6.38 shows that concrete with lower basic w/c ratio exhibited a lower
cracking resistance and the cracking resistance decreased by 5.6%, and 14.8% or
66.7%, and 122.2% when basic w/c ratio decreased from 0.50 to 0.40, and 0.33 for
mixtures NC-50, NC-40, and NC-33 or mixtures IC-50, IC-40, and NC-33, respec-
tively. The cracking resistance increased by 66.7%, 47.4%, and 35.5% when the basic
w/c ratio decreased from 0.50 to 0.40, and 0.33 for mixtures IC-50, IC-40, and IC-33,
compared to that for mixtures NC-50, NC-40, and NC-33, respectively. Although the
total w/c ratio of mixtures IC-33 or IC-40 was lower than that of mixtures NC-40
or NC-50, the cracking resistance increased by 29.8% or 44.4% for mixtures IC-33
or IC-40, compared to that for mixtures NC-40 or NC-50, respectively. This could
also clearly show the cracking resistance improvement in ICC with pre-wetted LWAs
resulted from the effective internal curing w/c ratio.

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 325

6.5.3 Addition of Polypropylene Fiber with Different Lengths

1. Mixture proportions and materials

Five groups of concrete mixtures were studied. The mixture proportions were char-
acterized by NC-L00, BF-L00, BF-L42, BF-L54, and BF-L60, respectively, which
is shown in Table 6.5.
Mixture NC-L00 was ordinary concrete mixture with no PP fiber or internal
curing, and mixture BF-L00 was the ICC mixture with no PP fiber. The lengths of
PP fibers were 42, 54, and 60 mm, and the contents were 8 kg/m3 for mixtures BF-
L42, BF-L54, and BF-L60, respectively. The basic w/b ratio of all concrete mixtures
were 0.33.
For all ICC mixtures, the calculated theoretical amount of internally curing water
needed to achieve the maximum degree of cement hydration was 32.85 kg/m3 . The
water absorption of SAPs was taken to be 11 g of water/g of SAPs, which was the
72 h saturated absorption value, thus the calculated weight of SAPs was 2.99 kg/m3 .
Therefore, the internally curing w/b ratio was 0.06, and the total w/b ratio was 0.39
for all ICC mixtures.
2. Temperature
The adiabatic temperature history of mixture NC-L00, BF-L00, BF-L42, BF-L54,
and BF-L60 is shown in Fig. 6.39.
As the result of the great amount of heat generated by cement hydration, the
temperature of the concrete specimen rose rapidly in the initial stage after pouring.
The initial temperature was 15.91, 20.43, 20.76, 20.81, and 19.79 °C. The temperature
rise time was 21, 24, 22, 20, and 21 h, then reached to the highest temperatures of
44.33, 50.47, 50.39, 50.36, and 50.26 °C. Therefore, the temperature rise was 28.42,
32.41, 31.67, 31.81, and 31.92 °C for mixture BF-L00, BF-L42, BF-L54, and BF-
L60, respectively. Higher adiabatic temperature rise than ordinary concrete mixture
was found in internally cured HPC mixture, which indicated that the internally cured
HPC has higher cement hydration degree. There was no obvious difference between

Table 6.5 Concrete mixture proportions

Mixture composition Concrete Mixtures
NC-L00 BF-L00 BF-L42 BF-L54 BF-L60
Water (kg/m3 ) 171 171 171 171 171
Cement (kg/m3 ) 512 512 512 512 512
Fine aggregate(sand) (kg/m3 ) 636 636 636 636 636
Coarse aggregate (kg/m3 ) 1131 1131 1131 1131 1131
SAPs (kg/m3 ) 2.99 2.99 2.99 2.99 2.99
Superplasticizer (kg/m3 ) 0 3.07 3.07 3.07 3.07
PP fiber content (kg/m3 ) 0 0 8 8 8
PP fiber length (mm) 0 0 42 54 60

@seismicisolation
@seismicisolation
326 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

60
NC-L00
50 BF-L00
BF-L42
BF-L54
Temperature (°C)

40 BF-L60

30

20

10

0
0 20 40 60 80 100
Age (h)
(a) Actual temperature rise and drop

40
NC-L00
BF-L00
30
BF-L42
BF-L54
Temperature (°C)

20 BF-L60

10

-10

-20
0 20 40 60 80 100
Age (h)
(b) Absolute temperature rise and drop

Fig. 6.39 Adiabatic temperature history of free shrinkage concrete specimens

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 327

the temperature rise of PP fiber reinforced internally cured HPC mixtures, which
indicated that PP fibers did not participate in the cement hydration reaction.
The highest temperature was forced to cool down after 36 h constant temperature
curing. Under the same temperature history, the temperature drop of concrete could
directly reflect the cracking resistance during the forced cooling stage.
The calculated temperature drop was 34.7, 35.9, 37.5, 37.6, and 39.7 °C for
mixture NC-L00, BF-L00, BF-L42, BF-L54, and BF-L60, respectively. Compared
with mixture NC-L00, the temperature drops of mixture BF-L00, BF-L42, BF-L54,
and BF-L60 increased by 3.5%, 8.1%, 8.4%, and 14.4%, respectively, which indi-
cated that with the use of SAPs and increasing length of PP fibers, it required larger
amount of temperature change to induce cracking for the PP fiber reinforced concrete.
The increasing length of the PP fiber had a significant improvement on concrete
resistance to the temperature change. Due to the three-dimensional distribution in
the specimen, the fiber can inhibit the production and development of initial microc-
racking, and play the role of bridging effect after initial cracking of concrete, reducing
the internal defects and improving the early-age cracking resistance of concrete.
Besides, due to the dense dispersion of the fine fiber in the concrete, the dense micro
crack system replace the large single crack, and the opening and expansion of the
micro crack are all controlled. Additionally, the bonding area between the concrete
matrix and the fiber increased as the length of PP fibers increased. Therefore, the
bonding strength and cracking resistance of fiber net was enhanced and the cracking
of concrete required more temperature change with the increasing length of PP fibers.
3. Free deformation
Figure 6.40 shows that the deformation of all specimens increased with the increase
of the temperature.
The maximum free total deformation of the mixture NC-L00, BF-L00, BF-L42,
BF-L54, and BF-L60 was 112, 154, 194, 180, and 171 με, respectively. Subse-
quently, the specimens entered into the constant temperature period. The autoge-
nous shrinkage deformation increased slightly and the temperature deformation kept
almost unchanged. Thus, the free total deformation of all specimens would have a
slight decrease under the condition of constant temperature curing stage. After the
end of the constant temperature period, the temperature of specimens was forced to
cool down. The free total deformation of the specimens gradually decreased until
the restrained specimens cracked finally during the cooling period.
The CTE basically stabilized within 1 day according to Eq. (6.7). As the cooling
time was very short and negligible, the autogenous shrinkage of concrete increased
very slowly, and the free deformation of concrete in the cooling period could be
considered as the temperature-strain curve and the CTE of the concrete was calculated
to replace the CTE of concrete at 28 d. The CTE of each group of concrete was 4.77,
4.54, 4.46, 4.32, and 4.18 με/°C for mixture NC-L00, BF-L00, BF-L42, BF-L54,
and BF-L60, respectively, which was the result of regression analysis during cooling
period according to the deformation-temperature curve of free shrinkage concrete
specimens. However, the result of CTE was higher than actual CTE due to the free

@seismicisolation
@seismicisolation
328 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

200 NC-L00
BF-L00
160 BF-L42
BF-L54
Total deformation (με)
120 BF-L60
80

40

-40

-80

0 20 40 60 80 100
Age (h)

Fig. 6.40 Total deformation of free shrinkage concrete specimens

total deformation was composed of temperature shrinkage and autogenous shrinkage

deformation.
Figure 6.41 shows that the autogenous shrinkage of HPC was significantly reduced
with the application of SAPs, which was because that the self-desiccation decreased
when the SAPs was used.

40
Autogenous shrinkage (με)

-40
NC-L00
BF-L00
BF-L42
-80
BF-L54
BF-L60

-120

0 20 40 60 80 100
Age (h)

Fig. 6.41 Autogenous shrinkage of free shrinkage concrete specimens

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 329

Also, the autogenous shrinkage of PP fiber reinforced internally cured HPC

decreased as the fiber length increased. The autogenous shrinkage of mixture BF-
L00, BF-L42, BF-L54, and BF-L60 at the time when the restrained concrete NC-L00
cracked initially was − 31, − 17, 8, and 16 με, respectively, which decreased by
73.9%, 85.7%, 106.7%, and 113.4%. The phenomenon could be explained by the
following three aspects: Firstly, the pore structure of concrete has undergone signif-
icant changes with the addition of fiber from the microscopic point of view. There
are more pores with larger diameters in the pore structure, and the longer the fiber
length is, the larger number of pores with larger diameters will be produced. The
larger pore size reduced the capillary pressure and thereby the autogenous shrinkage
of the concrete was reduced according to the Kelvin-Laplace equation. Secondly, the
bonding strength between the fiber and the concrete matrix increased as the strength
of concrete increased and the greater the bonding strength was, the stronger the ability
of the fiber to inhibit the autogenous shrinkage of the concrete would be. Therefore,
the longer the fiber length was, the greater the bonding surface between the fiber
and the cement paste would be. Thus, the stronger the degree of inhibition of the
autogenous shrinkage can result in a smaller autogenous shrinkage of the concrete.
Thirdly, the surface of the fiber was continuously embossed and not a smooth surface,
which made the adhesion between the fiber and the cement paste more compact. The
bonding stress between the fiber and the cement paste increased and the ability to
inhibit the autogenous shrinkage was stronger when the fiber length increased in
concrete. The mixture BF-L54 and BF-L60 were even characterized by expansion
deformation. The reabsorption of bleeding water, the formation of ettringite, and the
larger spatial network of reaction products may account for the early-age expansion
deformation of internally cured HPC.
4. Restrained stress
Figure 6.42 shows the early-age stress development of five groups of concrete
specimens under the condition of restraint.
The test results showed that the expansion of the test specimens was restrained
to generate a large compressive stress as the temperature of the specimen increased
after pouring. The maximum compressive stress of the restrained specimens was
0.78, 0.79, 0.77, 0.74, and 0.75 MPa, and the stress occurred at 14.06, 17.25, 16.33,
16.17, and 16.25 h when the temperature reached to 39.82, 48.72, 49.46, 49.00, and
48.69 °C for mixture NC-L00, BF-L00, BF-L42, BF-L54, and BF-L60, respectively.
As shown in Fig. 6.43, stress relaxation occurred when the early-age concrete with
small strength and elastic modulus was subjected to high compressive stress.
The reduction of the compressive stress of the restrained specimen in the constant
temperature stage gradually decreased as the strength of concrete increased. The
compressive stress of the specimen was 0.04, 0.11, 0.03, 0.10, and 0.10 MPa for
mixture NC-L00, BF-L00, BF-L42, BF-L54, and BF-L60 at the beginning of the
forced cooling. Therefore, the stress relaxation was 0.68, 0.74, 0.64, and 0.65 MPa,
which decreased by 8.1%, 0%, 13.5%, and 12.2% when compared with 0.74 MPa of
mixture NC-L00, which also indicated that the compressive creep of HPC decreased
with the use of SAPs and the increasing length of PP fibers.

@seismicisolation
@seismicisolation
330 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

4
NC-L00
BF-L00
Restrained stress (MPa) 3 BF-L42
BF-L54
BF-L60
2

-1
0 20 40 60 80 100
Age (h)

Fig. 6.42 Development of restrained stress measured in restrained shrinkage specimens

The cracking stress and cracking age of concrete could be applied to assess the
cracking resistance. The stress curve was linear when the specimens were forced
to cool in the forced cooling stage. Then, the sudden drop of stress curve indicated
the initial cracking of concrete. The cracking stress of mixture NC-L00, BF-L00,
BF-L42, BF-L54, and BF-L60 was 3.08, 2.24, 2.32, 2.44, and 2.51 MPa when
the restrained specimens cracked at the time of 103, 102, 105, 106, and 106 h,
respectively, as shown in Fig. 6.44.
The cracking stress decreased by 27.3% with the application of SAPs. This was
because that the application of SAPs contributed to an increasing number larger pores
that were short of the filling of cement hydration products, which contributed to the
premature cracking of concrete. Increasing the length of PP fibers from 0 to 42, 54,
and 60 mm, the cracking stress of internally cured HPC increased by 3.6%, 8.9%,
and 12.1%, respectively. The results of cracking stress and cracking age indicated
that the cracking resistance of internally cured HPC increased as the PP fiber length
increased.
Meanwhile, the development of the restrained tensile stress rate of the test spec-
imens in the cooling period could be obtained according to Fig. 6.43, which could
be used to analyze the cracking resistance of concrete. Figure 6.45 shows that the
restrained tensile stress rate of mixture NC-L00, BF-L00, BF-L42, BF-L54, and BF-
L60 was 0.074, 0.061, 0.054, 0.055, and 0.060 MPa/h, respectively, which decreased
by 11.3%, 10.7%, and 7.1%.
The tensile stress rates of the restrained specimens decreased when the SAPs and
PP fibers were applied, but gradually increased in the cooling period as the PP fiber
length increased. The development of tensile stress rate of restrained specimen is

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 331

Fig. 6.43 Adiabatic 3.5 60

NC-L00
temperature history and 3.0
restrained stress for different 50
2.5

Restrained stress (MPa)

concrete mixtures Tzero-stress

)
2.0 40

Temperature (
Stress
1.5
Temperature 30
1.0
0.5 20
0.0
10
-0.5
Stress relaxation
-1.0 0
0 20 40 60 80 100
Age (h)
(a) NC-L00
3.5 60
BF-L00
3.0
Tzero-stress 50
2.5
Restrained stress (MPa)

)
2.0 40

Temperature (
Stress
1.5
Temperature 30
1.0
0.5 20
0.0
10
-0.5
Stress relaxation
-1.0 0
0 20 40 60 80 100
Age (h)
(b) BF-L00
3.5 60
BF-L42
3.0
Tzero-stress 50
2.5
Restrained stress (MPa)

2.0 40
)

Stress
Temperature (

1.5
Temperature 30
1.0
0.5 20
0.0
10
-0.5
Stress relaxation
-1.0 0
0 20 40 60 80 100
Age (h)
(c) BF-L42

@seismicisolation
@seismicisolation
332 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

Fig. 6.43 (continued) 3.5 60

BF-L54
3.0
Tzero-stress 50
2.5

Restrained stress (MPa)

)
2.0 40

Temperature (
Stress
1.5
Temperature 30
1.0
0.5 20
0.0
10
-0.5
Stress relaxation
-1.0 0
0 20 40 60 80 100
Age (h)
(d) BF-L54
3.5 60
BF-L60
3.0
50
2.5 Tzero-stress
Restrained stress (MPa)

)
2.0 40

Temperature (
Stress
1.5
Temperature 30
1.0
0.5 20
0.0
10
-0.5
Stress relaxation
-1.0 0
0 20 40 60 80 100
Age (h)
(e) BF-L60

Fig. 6.44 Cracking age, Cracking age (h)

Cracking stress and axial
Cracking stress (MPa)
tensile strength for different
Axial tensile strength
concrete mixtures
at the age of cracking (MPa)
103 102 105 106 106

5.12 5.18 5.45

4.84
4.42
3.08
2.24 2.32 2.44 2.51

NC-L00 BF-L00 BF-L42 BF-L54 BF-L60

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 333

0.08

Average rate of restrained tensile stress

0.06
(MPa/h)

0.04

0.02

0.00
NC-L00 BF-L00 BF-L42 BF-L54 BF-L60

Fig. 6.45 Average rate of restrained tensile stress for different concrete mixtures

related to the elastic modulus, stress relaxation, and degree of autogenous shrinkage.
The increasing fiber length increased the elastic modulus of concrete, which inhibited
the creep deformation. Therefore, the tensile stresses rate of the concrete would
increase gradually as the smaller creep caused less stress relaxation, which was
negative to the cracking resistance.
For the purpose of further understanding the development of cracking resistance of
PP fiber reinforced internally cured HPC, the splitting tensile strength and compres-
sive elastic modulus of different lengths of PP fiber reinforced internally cured HPC
with SAPs at 3, 7, and 28 d were tested. Table 6.6 shows the test results, which indi-
cated that the splitting tensile strength of all mixtures showed an increasing trend
with the increase of age and the splitting tensile strength of fiber reinforced internally
cured HPC gradually increased when the length of the PP fibers increased.
The splitting tensile strength of mixture BF-L42, BF-L54, and BF-L60 was 6.42,
6.87, and 7.29 MPa, respectively, which was 8.1%, 15.7%, and 22.7% higher than the
5.94 MPa of mixture BF-L00 at 28 d. While the splitting tensile strength of mixture

Table 6.6 Mechanical properties and regression coefficients of concrete mixtures

Test No. Splitting tensile strength (MPa) Elastic Modulus (GPa)
3 7 28 λ1 k1 3 7 28 λ2 k2
NC-L00 5.54 6.29 6.52 0.417 3.549 41.01 44.56 47.72 0.296 2.335
BF-L00 4.42 5.01 5.94 0.518 2.043 28.59 31.05 33.51 0.304 2.247
BF-L42 4.55 5.60 6.42 0.683 2.566 35.44 35.74 36.99 0.071 1.551
BF-L54 5.46 6.05 6.87 0.408 2.075 35.45 37.28 38.15 0.167 2.922
BF-L60 5.96 6.37 7.29 0.333 1.798 38.26 38.88 41.81 0.148 1.550

@seismicisolation
@seismicisolation
334 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

BF-L00 was 8.9% lower than mixture NC-L00 of 6.52 MPa, which had no internal
curing.
Table 6.6 also shows the compressive elastic modulus of HPC, which indicated
that the elastic modulus of all concrete mixtures showed an increasing trend with
the increase of age, and the compressive elastic modulus of the PP fiber reinforced
concrete gradually increased with the increasing length of fiber at the same age.
The elastic modulus of mixture BF-L42, BF-L54, and BF-L60 was 36.99, 38.15,
and 41.81 GPa, respectively, which increased by 10.4%, 13.9%, and 24.8% when
compared with 33.51 GPa of mixture BF-L00 at 28 d. The compressive elastic
modulus of mixture BF-L00 was 30.7% lower than mixture NC-L00 of 47.72 GPa.
Since there is no difference between the tensile elastic modulus and compressive
elastic modulus at early-age, the elastic modulus of concrete in the compression
state could also be used as tensile elastic modulus in this research.
Figure 6.46 shows the development models of axial tensile strength and elastic
modulus over time for all concrete mixtures.
Figure 6.46 shows the elastic modulus and tensile strength of the concrete
increased rapidly at the initial stage after the concrete was poured, which was
due to the continuous hydration of the cement, but cement hydration was basically
completed and the growth of tensile strength and elastic modulus gradually became
flat over time.
The mechanical properties test of concrete was conducted under the standard
curing condition and the restrained test specimens in the TSTM test underwent a
complicated temperature history. Obviously, the tensile strength of restrained spec-
imens and mechanical properties could not be compared directly. In order to elimi-
nate the effects of different temperature histories, the age of the restrained specimen
under actual temperature history was converted to equivalent age at 20 °C by using
the equivalent age method.
The equivalent cracking age of the restrained specimen was 223.0, 280.3, 283.8,
283.3, 261.0 h, the cracking stress was 3.08, 2.24, 2.32, 2.44, and 2.51 MPa, and the
tensile strength at the equivalent cracking age of the restrained specimen was 5.12,
4.42, 4.84, 5.18, and 5.45 MPa for mixture NC-L00, BF-L00, BF-L42, BF-L54, and
BF-L60, respectively. Therefore, the ratio of cracking stress to axial tensile strength
was 0.60, 0.51, 0.48, 0.47, and 0.46, respectively, which decreased by 15.0%, 20.0%,
21.7% and 23.3% when the SAPs was applied and the length of the PP fibers increased
from 0 to 42, 54, and 60 mm for mixture BF-L00, BF-L42, BF-L54, and BF-L60.
This was due to that the creep deformation of the concrete was inhibited and the
restrained stress cannot be reduced in time as the PP fiber length increased, which
led to the increase in tensile stress rate and premature cracking of concrete. The axial
tensile strength of the concrete also increased as the the PP fiber length increased
at the same time, so the ratio of cracking stress to axial tensile strength gradually
decreased. The tensile strength was higher than the cracking stress of the restrained
specimen at the same time, which was due to that for the purpose of achieving a
100% degree of restraint in the TSTM test, the stepper motor was frequently started
and the test specimen was pulled or pushed back to the initial position. This repeated
tension and compression process caused different degrees of damage to the internal

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 335

Fig. 6.46 Test results and time-dependent development models of axial tensile strength and elastic
modulus for different concrete mixtures

@seismicisolation
@seismicisolation
336 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

80
NC-L00
70
BF-L00
Basic tensile creep (με)
60 BF-L42
BF-L54
50 BF-L60

40

30

20

10

0
60 70 80 90 100 110
Age (h)

Fig. 6.47 Basic tensile creep for different concrete mixture

structure of the concrete. In addition to these, the loading speed of the splitting test
is too fast, which result in the higher axial tensile strength of concrete than actual
tensile strength.
5. Tensile creep behaviour
As shown in Fig. 6.47, the basic tensile creep was 74, 51, 68, 55, and 55 με for mixture
NC-L00, BF-L00, BF-L42, BF-L54, and BF-L60 when the restrained specimen NC-
L00 cracked initially.
The results indicated that the basic tensile creep was reduced when the SAPs
were applied. The results also indicated that the basic tensile creep of internally
cured HPC increased when the PP fibers were applied, but decreased as the PP fiber
length increased. The main factors affecting the creep of PP fiber reinforced internally
cured HPC mainly include the following aspects. Firstly, it can improve the ability
of concrete to resist creep when the elastic modulus of the fiber exceeds that of the
ordinary concrete, on the contrary, it is not conducive to the resistance of concrete to
the creep. The elastic modulus of the plastic fiber used in this working condition is
lower than that of the ordinary concrete. Therefore, the PP fibers was not conducive to
improving the ability of concrete to resist creep. Secondly, some defects were brought
about in addition to enhancing the properties of concrete when the fiber materials
were added. These defects included the weakness of the bonding interface between
the fiber and the concrete matrix and the uneven distribution of the PP fibers. Under
the effect of the external loads, the migration and exudation of the gel adsorbed water
and interlayer water would occur, thereby weakening the ability of concrete to resist
creep. Moreover, the tensile strength and elastic modulus of internally cured HPC
increased gradually at early-age as the PP fiber length increased, and the increasing

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 337

tensile strength and elastic modulus would restrain the creep of concrete. Therefore,
the basic tensile creep of internally cured HPC decreased when the length of PP
fibers increased.
In order to compare the creep under different stress, some scholars have proposed
the concept of specific creep, which is the creep behavior under a constant load. Since
the restrained stress was constantly changed in this experiment, the ratio of cumula-
tive basic tensile creep to restrained stress was used to represent the creep behavior.
Figure 6.48 shows that the specific basic tensile creep of mixture NC-L00, BF-L00,
BF-L42, BF-L54, and BF-L60 was 24.6, 22.9, 30.3, 24.1, and 23.0 με/MPa when
the restrained HPC NC-L00 cracked initially in the cooling stage, which indicated
that the specific basic tensile creep of the specimen increased obviously when the
PP fibers were applied, but decreased as the PP fiber length increased.
At the same time, in order to better reflect the deformation reduction of concrete
in the tensile state, the ratio of the basic tensile creep to shrinkage was used to
explain the creep characteristics of concrete. Figure 6.49 shows the ratio of the basic
tensile creep to shrinkage of internally cured HPC specimen was the same as the
development trend of specific basic tensile creep, which increased firstly and then
decreased with the increase of age.
The shrinkage was 140, 121, 129, 117, and 114 με when the restrained HPC
NC-L00 cracked in the cooling period. Therefore, the ratio of basic tensile creep to
shrinkage was 0.53, 0.42, 0.52, 0.48, and 0.48 for mixture NC-L00, BF-L00, BF-
L42, BF-L54, and BF-L60, respectively. The basic tensile creep of internally cured
HPC increased obviously when the PP fibers were applied, but decreased gradually
as the PP fiber length increased, which indicated that the tensile creep behavior is
related not only to PP fiber incorporation, but also to fiber length at early-age.

100
NC-L00
Specific basic tensile creep (με/MPa)

BF-L00
80 BF-L42
BF-L54
BF-L60
60

40

20

0
60 70 80 90 100 110
Age (h)

Fig. 6.48 Specific basic tensile creep for different concrete mixture

@seismicisolation
@seismicisolation
338 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

1.0
NC-L00
BF-L00
Basic tensile creep/shrinkage 0.8
BF-L42
BF-L54
BF-L60
0.6

0.4

0.2

0.0
60 70 80 90 100 110
Age (h)

Fig. 6.49 Basic tensile creep/shrinkage for different concrete mixture

6. Cracking resistance
Figure 6.50 shows the calculated cracking resistance evaluation integrated criterion
for mixture NC-L00, BF-L00, BF-L42, BF-L54, and BF-L60 was 0.038, 0.034,
0.029, 0.025, and 0.028 MPa/(h·day), respectively, which indicated that the early-
age cracking resistance of HPC was significantly improved when SAPs and PP fibers
were applied.
Test results also showed that the cracking resistance of internally cured HPC
increased as the PP fiber length increased. This was not only because the PP fibers
were added to the concrete to form a three-dimensional irregular distribution support
system, the bonding between the concrete and the fiber was also enhanced by the
high tensile strength and the continuous embossing shape of PP fibers, both of which
could increase the ability of energy absorption when the concrete was subjected to
load. Additionally, the theory of fiber spacing shows that when the microcrack length
is greater than the fiber spacing in the hardened state of the concrete, the fiber will
transmit the load across the crack. Meanwhile, the fiber can passivate the internal
stress field, making it more continuous and uniform, weakening the internal stress
of the concrete, and then the expansion of the crack will be restrained. On the other
hand, when the length of the microcrack is less than the fiber spacing, the fiber
will force the cracking to change its development direction or generate a finer crack
field, thereby increasing the energy required for crack propagation and preventing
the development of the cracking significantly.
However, increasing the length of PP fibers from 54 to 60 mm, the cracking
resistance of PP fiber reinforced internally cured HPC decreased slightly. The effect
of PP fiber length on cracking resistance of internally cured HPC mainly includes two
aspects. Firstly, the boundary effect causes the PP fibers to rely on the surrounding

@seismicisolation
@seismicisolation
6.5 Influential Factors of Cracking Resistance of Internally Cured Concrete 339

0.04
Cracking Resistance by
Integrated Criterion
Integrated criterion (MPa/(h⋅day))
0.03

0.02

0.01

0.00
NC-L00 BF-L00 BF-L42 BF-L54 BF-L60

Fig. 6.50 Estimation of cracking resistance using integrated criterion for different concrete
mixtures

of coarse aggregate when the fiber length is too short, which greatly affects the
encapsulation and anchorage effect of concrete slurry on fiber, and then leads to the
weakening of fiber reinforcement effect. Meanwhile, shorter fibers are more likely to
be distributed in sequence, which is difficult to form a supporting net. The fiber even
becomes the inducer of the cracking development when the cracking is parallel to the
fiber. Secondly, when the length of fiber is too long, the fiber dispersion decreases,
which will lead to fiber clumping, crimping and winding easily, thereby weakening
the adhesion between the fiber and the concrete matrix. In addition, when the content
of fiber is the same, the longer the length of fiber is, the less the number of fiber is,
and the more cracking are difficult to suppress effectively. Therefore, the cracking
resistance of PP fiber reinforced internally cured HPC is effectively improved with
the increasing length of fiber, but the effect of cracking resistance will be reduced
when the fiber length increased from 50 to 60 mm.

6.5.4 Summary

The influences of GGBFS, w/c ratio, and PP fiber length on the performance
and cracking resistance of ICC under uniaxial constant restraint were investigated.
Analysis was conducted on the development of mechanical properties, temperature
history, deformation, and restrained stress of ICC based on the results of TSTM test.
The main conclusions drawn from the study are as follows.
(1) The compressive strength, splitting tensile strength, and elastic modulus of ICC
decreased with the increase of GGBFS content. A model for predicting the

@seismicisolation
@seismicisolation
340 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

compressive strength of concrete was presented in consideration of the influence

of pre-wetted LWAs. The compressive strength, splitting tensile strength, and
elastic modulus of early-age internally cured HPC increased when length of PP
fibers increased (up to 60 mm).
(2) The temperature rise of concrete increased with the decrease of w/c ratio under
adiabatic condition. The temperature rise was 17.8, 23.6, and 31.0 °C, which
increased by 32.6%, and 74.2% when w/c ratio decreased from 0.50 to 0.40,
and 0.33, respectively.
(3) Restrained tensile stress rate increased with the decrease of w/c ratio under
adiabatic condition and at full restraint degree when the cooling rate was
1.0 °C/h. The restrained tensile stress rate was 0.06, 0.07, and 0.08 MPa/h,
which increased by 16.7%, and 33.3% when w/c ratio decreased from 0.50 to
0.40, and 0.33, respectively.
(4) The addition of GGBFS increased the autogenous shrinkage, restrained tensile
stress rate and decreased the cracking age, cracking stress of ICC. The abso-
lute value of autogenous shrinkage of normal and ICC both increased with
the decrease of w/c ratio and the adding of pre-wetted LWAs reduced the
autogenous shrinkage. The autogenous shrinkage and restrained tensile stress
rate decreased, and the temperature drop, cracking age, and cracking stress of
early-age internally cured HPC increased when the PP fibers were applied.
(5) The absolute value of basic tensile creep of normal and ICC both increased
with the decrease of basic w/c ratio. The addition of PP fibers increased the
basic tensile creep, specific tensile creep, and basic tensile creep/shrinkage of
internally cured HPC, while the tensile creep behavior decreased as the PP fiber
length increased.
(6) GGBFS had a negative effect on the early-age behavior and cracking resistance
of ICC. The addition of 20% of GGBFS content in ICC showed better cracking
resistance than that for mixtures with 35% and 50% of GGBFS content. The
cracking resistance of normal and ICC both decreased with the decrease of basic
w/c ratio and the adding of pre-wetted LWAs improved the cracking resistance,
as determined from the integrated criterion, which was based on combining the
criteria of the net time of cracking and the average rate of tensile stress.
(7) The cracking resistance of ICC increased first and then decreased with the
increase of PP fiber length. Increasing the length of PP fiber from 0 to 42, 54,
and 60 mm, the integrated criterion for assessing the cracking resistance of ICC
decreased by 14.7%, 26.5%, and 17.6%, respectively. This was mainly because
that the PP fiber contributed to improving the load transfer between the fibers
and concrete; while the poor dispersion of PP fibers may weaken the bonding
interface between concrete matrix and fiber when the length was too long.

@seismicisolation
@seismicisolation
References 341

References

1. Shen DJ, Jiang JL, Shen JX et al (2015) Influence of prewetted lightweight aggregates on the
behavior and cracking potential of internally cured concrete at an early-age[J]. Constr Build
Mater 99:260–271
2. Igarashi S, Bentur A, Kovler K (2000) Autogenous shrinkage and induced restraining stresses
in high-strength concretes[J]. Cem Concr Res 30(11):1701–1707
3. Shen DJ, Jiang JL, Zhang MY et al (2018) Tensile creep and cracking potential of high perfor-
mance concrete internally cured with super absorbent polymers at early-age[J]. Constr Build
Mater 165:451–461
4. Aly T, Sanjayan JG (2008) Shrinkage cracking properties of slag concretes with one-day
curing[J]. Mag Concr Res 60(1):41–48
5. Shen DJ, Liu C, Li CC et al (2019) Influence of Barchip fiber length on early-age behavior
and cracking resistance of concrete internally cured with super absorbent polymers[J]. Constr
Build Mater 214:219–231
6. Shen DJ, Liu C, Feng ZZ et al (2019) Influence of ground granulated blast furnace slag on the
early-age anti-cracking property of internally cured concrete[J]. Constr Build Mater 223:233–
243
7. Yao PP (2016) Study on cracking resistance of internally cured concrete with super absorbent
polymers[D]. Hohai University, Nanjing (in Chinese)
8. Shen JX (2015) Evaluation of crack resistance of internal cured concrete at early-age[D]. Hohai
University, Nanjing (in Chinese)
9. Zhu SS (2016) Effect of ground granulated blast furnace slag on crack resistance of high
strength concrete at early-age[D]. Hohai University, Nanjing (in Chinese)
10. Zhao XG (2016) Experimental study on evaluation of early-age cracking resistance of concrete
reinforced with plastic fibers[D]. Hohai University, Nanjing (in Chinese)
11. Kovler K (1994) Testing system for determining the mechanical behaviour of early-age concrete
under restrained and free uniaxial shrinkage[J]. Mater Struct 27(6):324–330
12. Springenschmid R (1985) Thermal stresses in mass concrete: a new method and the influence
of different cements[C]. Quinzieme Congres des Grands Barrages, Lausanne
13. Cusson D, Hoogeveen T (2007) An experimental approach for the analysis of early-age
behaviour of high-performance concrete structures under restrained shrinkage[J]. Cem Concr
Res 37(2):200–209
14. Wei Y, Hansen W (2013) Tensile creep behavior of concrete subject to constant restraint at
very early-ages[J]. J Mater Civ Eng 25(9):1277–1284
15. Zhang T, Qin WZ (2006) Tensile creep due to restraining stresses in high-strength concrete at
early-ages[J]. Cem Concr Res 36(3):584–591
16. Schlitter J, Bentz DP, Weiss WJ (2013) Quantifying stress development and remaining stress
capacity in restrained, internally cured mortars[J]. ACI Mater J 110(1):3–11
17. Krauß M, Gutsch A-W, Rostásy FS (2001) Modelling of degree of hydration on basis of adia-
batic heat release[M]. University of Technology, Department of Civil & Mining Engineering,
Division of Structural Engineering
18. Yoo DY, Park J-J, Kim SW et al (2014) Influence of ring size on the restrained shrinkage
behavior of ultra high performance fiber reinforced concrete[J]. Mater Struct 47(7): 1161-1174
19. Bentur A (2000) Early-ages shrinkage and cracking in cementitious systems[C]. International
RILEM workshop on shrinkage of concrete Shrinkage 2000. RILEM publications, Paris
20. Lura P (2003) Autogenous deformation and internal curing of concrete[D]. Technical
University Delft, Delft
21. Shen DJ, Wang XD, Cheng DB et al (2016) Effect of internal curing with super absorbent
polymers on autogenous shrinkage of concrete at early-age[J]. Constr Build Mater 106:512–522
22. Ni TY, Yang Y, Gu CP et al (2019) Early-age tensile basic creep behavioral characteristics of
high-strength concrete containing admixtures[J]. Advan Civil Eng 7:1–11
23. Kolver K, Igarashi S, Bentur A (1999) Tensile creep behavior of high strength concretes at
early-ages[J]. Mater Struct 32(5):383–387

@seismicisolation
@seismicisolation
342 6 Cracking Resistance of Internally Cured Concrete Under Uniaxial …

24. Riding KA, Poole JL, Schindler AK et al (2009) Effects of construction time and coarse
aggregate on bridge deck cracking[J]. ACI Mater J 106(5):448–454
25. ASTM International (2019) Standard test method for autogenous strain of cement paste and
mortar: ASTM C1581/C1581M-18a[S]. ASTM International, West Conshohocken
26. Kovler K, Bentur A (2009) Cracking sensitivity of normal- and high-strength concretes[J]. ACI
Mater J 106(6):537–542
27. Shen DJ, Kang JC, Jiao Y et al (2020) Effects of different silica fume dosages on early-age
behavior and cracking resistance of high strength concrete under restrained condition[J]. Constr
Build Mater 263:120218
28. Xin JD, Zhang GX, Liu Y et al (2020) Numerical analysis of effect of temperature history and
restraint degree on cracking behavior of early-age concrete[J]. J Eng Res 8(2):24–43
29. Zhu H, Li QB, Ma R et al (2020) Water-repellent additive that increases concrete cracking
resistance in dry curing environments[J]. Constr Build Mater 249:118704
30. General Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China. Common Portland cement: GB 175-2007/XG1-2018[S]. Standard press of
China, Beijing (2018) (in Chinese)
31. Shen DJ, Jiang JL, Jiao Y et al (2017) Early-age tensile creep and cracking potential of concrete
internally cured with pre-wetted lightweight aggregate[J]. Constr Build Mater 135:420–429
32. Gutsch AW (2002) Properties of early-age concrete-experiments and modelling[J]. Materials
Structures 35(2):76–79
33. Bažant ZP, Murphy W (1995) Creep and shrinkage prediction model for analysis and design
of concrete structures-model B3[J]. Matériaux et Constructions 28(180):357–365
34. Lura P, Bisschop J (2004) On the origin of eigenstresses in lightweight aggregate concrete[J].
Cement Concr Compos 26(5):445–452
35. Schlitter J, Senter A, Bentz D et al (2010) A dual concentric ring test for evaluating residual
stress development due to restrained volume change[J]. J ASTM Int 7(9):1–13
36. Raoufi K, Schlitter J, Bentz D et al (2011) Parametric assessment of stress development and
cracking in internally cured restrained mortars experiencing autogenous deformations and
thermal loading[J]. Advan Civil Eng 2011:1–16
37. Weiss J, Bentz D, Schindler A et al (2012) Internal curing[J]. Structure 12:10–14
38. Henkensiefken R, Bentz D, Nantung T et al (2009) Volume change and cracking in inter-
nally cured mixtures made with saturated lightweight aggregate under sealed and unsealed
conditions[J]. Cement Concr Compos 31(7):427–437
39. Bažant ZP, Wittmann FH (1982) Creep and shrinkage in concrete structures[M]. Wiley
40. Barrett TJ, de la Varga I, Weiss WJ (2012) Reducing cracking in concrete structures by using
internal curing with high volumes of fly ash[C]. Structures Congress: Design of Concrete
Bridges with Innovative Materials
41. Hobbs DW (1971) The dependence of the bulk modulus, Young’s modulus, creep, shrinkage and
thermal expansion of concrete upon aggregate volume concentration[J]. Matériaux et Constr
4(2):107–114
42. Cusson D, Lounis Z, Daigle L (2010) Benefits of internal curing on service life and life-cycle
cost of high-performance concrete bridge decks—a case study[J]. Cement Concr Compos
32(5):339–350
43. Lura P, Wyrzykowski M, Tang C et al (2014) Internal curing with lightweight aggregate
produced from biomass-derived waste[J]. Cement Concr Res 59:24–33
44. Hasholt MT, Jensen OM, Kovler K et al (2012) Can superabsorent polymers mitigate auto-
genous shrinkage of internally cured concrete without compromising the strength?[J]. Constr
Build Mater 31:226–230
45. Powers TC (1958) Structure and physical properties of hardened Portland cement paste[J]. J
Am Ceram Soc 41(1):1–6
46. Powers TC (1958) The physical structure and engineering properties of concrete[R]
47. Pantazopoulou SJ, Mills R (1995) Microstructural aspects of the mechanical response of plain
concrete[J]. Mater J 92(6):605–616

@seismicisolation
@seismicisolation
References 343

48. Bentz DP, Stutzman PE (1994) Evolution of porosity and calcium hydroxide in laboratory
concretes containing silica fume[J]. Cement Concr Res 24(6):1044–1050
49. Weber S, Reinhardt HW (1997) A new generation of high performance concrete: concrete with
autogenous curing[J]. Adv Cem Based Mater 6(2):59–68
50. Kong XM, Zhang ZL, Lu ZC (2014) Effect of pre-soaked superabsorbent polymer on shrinkage
of high-strength concrete[J]. Mater Struct 48(9):2741–2758
51. Zhuang YZ, Chen CY, Ji T (2013) Effect of shale ceramsite type on the tensile creep of
lightweight aggregate concrete[J]. Constr Build Mater 46:13–18
52. Bissonnette B, Pigeon M (1995) Tensile creep at early-ages of ordinary, silica fume and fiber
reinforced concretes[J]. Cem Concr Res 25(5):1075–1085
53. Altoubat SA, Lange DA (2001) Tensile basic creep: measurements and behavior at early-age[J].
ACI Mater J 98(5):386–393

@seismicisolation
@seismicisolation