See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/319999251

Rheological Properties of Wet-Mix Shotcrete Mixtures Made with Crushed

Aggregate

Article in Journal of Materials in Civil Engineering · September 2017

DOI: 10.1061/(ASCE)MT.1943-5533.0002084

CITATIONS READS
3 807

3 authors:

Pangil Choi Kyong Ku Yun

Texas Department of Transportation Kangwon National University
36 PUBLICATIONS 393 CITATIONS 100 PUBLICATIONS 692 CITATIONS

SEE PROFILE SEE PROFILE

Jung Heum Yeon

Texas State University
55 PUBLICATIONS 1,136 CITATIONS

SEE PROFILE

All content following this page was uploaded by Jung Heum Yeon on 10 October 2017.

The user has requested enhancement of the downloaded file.

 Rheological Properties of Wet-Mix Shotcrete Mixtures
Made with Crushed Aggregate
Pangil Choi, Ph.D. 1; Kyong-Ku Yun, Ph.D. 2; and Jung Heum Yeon, Ph.D., A.M.ASCE 3

Abstract: The recent depletion of high-quality and well-graded natural-aggregate sources is one of the factors compromising the production
of good-quality shotcrete. For this reason, the use of various alternative aggregate sources has been receiving increasing attention as one of the
promising solutions for this issue. In this study, the feasibility of using crushed aggregate obtained as a by-product of tunnel-excavation
Downloaded from ascelibrary.org by Jung Heum Yeon on 09/12/17. Copyright ASCE. For personal use only; all rights reserved.

processes in wet-mix shotcrete (WMS) is evaluated based on laboratory tests. To fulfill this aim, efforts have been made to evaluate rheo-
logical properties of WMS mixtures incorporating crushed aggregates because shotcrete rheology is recognized as one of the most significant
components affecting actual shotcrete behavior such as pumpability and shootability. Other tests performed include air content, slump, and
flow. The acceptability of crushed-aggregate WMS mixtures with four standard gradations were evaluated first. Then, allowable gradation
ranges for WMS mixtures made with crushed aggregate were suggested based on modified Dinger-Funk gradations and prescribed WMS
performance criteria. Results indicate that the use of crushed aggregate may be a suitable measure effectively meeting both industrial demands
and shotcrete performance requirements. DOI: 10.1061/(ASCE)MT.1943-5533.0002084. © 2017 American Society of Civil Engineers.
Author keywords: Wet-mix shotcrete mixture; Crushed aggregate; Natural aggregate; Rheological properties; Aggregate gradation.

Introduction One of the possible solutions to this issue is to take advantage

of crushed aggregate generated from tunnel excavations as an alter-
Wet-mix shotcreting is a process to premix portland cement, water, native aggregate source. So far, however, crushed aggregate has
aggregate, and admixtures all together prior to pumping, and then not been a primary option for shotcrete production, particularly
apply the mixture through a hose by means of compressed air because of its angular shape with rough surface textures, which pro-
(Beaupre 1994; Goodier et al. 2008). Wet-mix shotcrete (WMS) vide higher degrees of particle-to-particle interlock and friction,
has been widely accepted for diverse construction and repair appli- thereby reducing the freedom of aggregate particle movements, and
cations, including underground construction, tunnel lining, slope in turn, the workability (Malhotra 1964; Wills 1967; Geiker et al.
stabilization, and structural renovation (Leung et al. 2005). Because 2002; Hu 2005). In addition, aggregate with rougher surface tex-
shotcrete is required to have adequate levels of strength and tures has a larger contact area with the surrounding cement paste
durability, it is of great importance to use sound and high-quality and thus tends to demand more cement, water, and fine aggregate to
aggregate sources, which make up approximately 80% of the total obtain a proper level of workability (Kosmatka et al. 2002; Quiroga
constituents of shotcrete by volume (Jolin and Beaupre 2003). In and Fowler 2004).
Korea, most fine aggregates commonly used for WMS are obtained There have been limited experiences with the use of crushed
from natural sources such as natural land and sea. When these natu- aggregate up to the present. Thalmann-Suter (1999) performed
ral sources are used, however, desired shotcrete quality control petrographic analysis and point-load testing to evaluate the suitabil-
(QC) cannot be strictly achieved because the recent depletion of ity of excavated rock extracted from underground for use as con-
high-quality natural-aggregate sources led to the use of fine aggre- crete aggregate and concluded that the rock excavation materials
gate excessively coated with a dust of microfines with poor grada- can be potentially used as a substitute for gravel. Weiss (2007)
tion and fineness modulus. In addition, the natural-fine-aggregate demonstrated that high-quality concrete can be achieved from
shortage resulted in large variations in their qualities and physical aggregates processed from tunnel-excavation materials. Jolin and
properties from region to region, which made the QC and standardi- Beaupre (2003) developed optimum mixture proportions and ag-
zation of the natural-aggregate sources quite difficult. For these rea- gregate phase for WMS mixtures made with crushed aggregate
sons, contractors are required to screen and scrutinize all incoming and a high dosage of synthetic macrofibers. Laskar (2009) assessed
natural-aggregate sources before acceptance and modify mixture various rheological properties of high-performance concrete using
designs if needed, but this is not well regulated in actual practice. crushed rock as a coarse-aggregate source to facilitate implemen-
tation in shotcrete.
1 In this study, the feasibility of using crushed aggregate obtained
Postdoctoral Research Associate, Dept. of Civil and Environmental
Engineering, Texas Tech Univ., Lubbock, TX 79409. as a by-product of the tunnel-excavation process as an alternative
2
Professor, Dept. of Civil Engineering, Kangwon National Univ., aggregate source for WMS is investigated. The first approach to
Chuncheon 24231, South Korea. achieve this aim was to check whether the existing gradation stan-
3
Assistant Professor, Dept. of Civil and Environmental Engineering, dards and specifications—including Korean Industrial standards
Gachon Univ., Seongnam 13120, South Korea (corresponding author).
(KS) (Korean Industrial Standards 2002), American Concrete Insti-
E-mail: [email protected]
Note. This manuscript was submitted on May 31, 2016; approved on
tute (ACI) specification (ACI 2014, 2016), ASTM specifications
May 24, 2017; published online on September 12, 2017. Discussion per- (ASTM 2008a, b), the European Federation of National Associa-
iod open until February 12, 2018; separate discussions must be submitted tions Representing for Concrete (EFNARC)’s European specifi-
for individual papers. This paper is part of the Journal of Materials in cation for sprayed concrete (EFNARC 1996), and Dinger-Funk
Civil Engineering, © ASCE, ISSN 0899-1561. gradation (Funk and Dinger 1994) can be directly implemented with

© ASCE 04017227-1 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2017, 29(11): 04017227

 WMS mixtures made with crushed aggregate. Various laboratory and ASTM C33M-08 limit the maximum coarse-aggregate size to
evaluations, such as a slump test, air-content test, flow test, and 15 mm. When a combined-aggregate gradation is required, it is re-
rheology test, were performed on WMS mixtures with different quired to comply with ACI 506.2-13 or ASTM C1436-08.
standard gradations because all of these properties are recognized European specification for sprayed concrete (EFNARC 1996)
to have substantial effects on the shotcrete behavior. Then, allowable stipulates that the percentage passing a No. 100 sieve (0.125 mm)
gradation ranges that met the prescribed shotcrete performance cri- by weight should be at least 4–5%. No single fraction should
teria were identified using the Dinger-Funk particle-size distribu- exceed 30% of the total aggregate weight. Funk and Dinger (1994)
tion, one of the most commonly accepted gradation standards. suggested an aggregate particle-size distribution that theoretically
This investigation was done by manipulating the percentage passing yields the densest packing of aggregate particles (hereinafter,
through each sieve specified in Dinger-Funk gradations (Funk and Dinger-Funk gradation), which is one class of gradation standards
Dinger 1994) by up to −20 to 20% by weight (hereinafter, modified considered in this study.
Dinger-Funk gradations) when measuring the rheological proper-
ties. Additionally, the rheological characteristics of shotcrete
mixtures made with natural aggregate were evaluated to quantita- Experimental Program
Downloaded from ascelibrary.org by Jung Heum Yeon on 09/12/17. Copyright ASCE. For personal use only; all rights reserved.

tively compare the effect of aggregate type on the shotcrete behav-

ior. The findings of this study are expected to provide data support
and guidance for the extensive use of crushed aggregate in produc- Aggregate Preparation
ing WMS mixtures with enhanced quality and constructability, as Aggregates with various grain sizes were obtained by sieving
well as economic and environmental benefits. raw crushed stone directly collected from a tunnel-excavation site.
To obtain the aggregates complying with gradation standards and
specifications, the weight of aggregate required to be retained in
Review of Aggregate Gradation Standards and
each sieve was calculated, and then aggregate samples meeting
Specifications
each gradation standard were prepared. To compare the aggregate
Most industrial standards and specifications administer gradation type effect on the fresh shotcrete properties, natural aggregate with
criteria for concrete/shotcrete aggregates. KS stipulates to test an identical size distribution was also sampled as shown in Fig. 1.
combined-aggregate gradation for shotcrete as per KS F 2577 Table 1 presents the grading class, specific gravity, and fineness
(Korean Industrial Standards 2002). KS requires maximum aggre- modulus for all four standard gradations adopted in this study;
gate sizes of 9.5 and 12.5 mm, which are consistent with ACI ACI-Fine and EFNARC are gradations for fine aggregate only,
506.2-13 (ACI 2014) and ASTM C1436-08 (ASTM 2008b). In whereas Dinger-Funk and ACI-Combined are gradations for com-
cases when coarse and fine aggregates are graded separately and bined aggregate (i.e., both coarse aggregate and fine aggregate).
mixed during mixing, KS stipulates adoption of KS F 2526 Fig. 2 illustrates the actual aggregate gradations used in the present
(Korean Industrial Standards 2007a). study.
In ACI 506R-16 (ACI 2016) and ASTM C33M-08 (ASTM Fig. 3 depicts the modified Dinger-Funk gradations investigated
2008a), a fineness modulus of 2.5–3.3 is recommended for fine ag- in this study. The specific gravity was consistent with 2.62 for all
gregate. In addition, they require the percentage passing No. 200 the modified gradations. In the plot, Dinger-Funk 0% indicates the
sieve (0.074 mm) by weight not to exceed 2% of the total original Dinger-Funk gradation, and Dinger-Funk þ10% indicates
aggregate weight. Because large coarse aggregate can result in clog- the gradation wherein the percentage passing on each sieve was
ging of a hose and wasteful rebound during shooting, ACI 506R-16 increased (factored) by 10% with respect to Dinger-Funk 0%.

Fig. 1. Sampled aggregates: (a) crushed aggregate; (b) natural aggregate

Table 1. Grading Class and Physical Properties of Aggregates Used in This Study
Specific Fineness
Gradation standard Gradation identifier Grading class gravity modulus
ACI 506-66/ASTM C33M-08/KS F 2526 ACI-fine Fine aggregate 2.62 2.77
ACI 506.2-95/ASTM C1436-08/KS F 2577 (Grading no. 1) ACI-combined Combined aggregate 2.62 3.75
European specification for sprayed concrete EFNARC Fine aggregate 2.62 3.04
Dinger-Funk size distribution Dinger-Funk Combined aggregate 2.62 3.98

© ASCE 04017227-2 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2017, 29(11): 04017227

 100
Percentage passed by weight [%] ACI-Fine
moving through a hose under pressure (Jolin et al. 2009)] and shoot-
ACI-Combined ability [ability of a mixture to build up on or adhere to vertical and/or
80 EFNARC overhead surfaces (Beaupre 1994; Choi 2008)], along with a 28-day
Dinger-Funk compressive strength of 48 MPa, slump of 220 mm, initial air content
60 of 17% before shooting, and air content of 5% after shooting. The
mixtures had a consistent water-to-cementitious ratio (w=cm) of
0.41, a fine aggregate-to-total aggregate fraction (S/a) of 0.71, and
40
a maximum aggregate size of 10 mm.

20
Methods
0
0.01 0.1 1 10 Air Content and Slump
Sieve size [mm] The air contents of fresh WMS mixtures were measured as per
the test procedures specified in KS F 2421 (Korean Industrial
Downloaded from ascelibrary.org by Jung Heum Yeon on 09/12/17. Copyright ASCE. For personal use only; all rights reserved.

Fig. 2. Standard aggregate gradations used in this study Standards 2006). The slump, the most common indicator to evalu-
ate the workability of fresh concrete mixtures, was determined in
accordance with KS F 2402 (Korean Industrial Standards 2007b)
100 using fresh WMS mixture samples prepared in accordance with KS
Percentage passed by weight [%]

Dinger-Funk -20% F 2401 (Korean Industrial Standards 2007c).

80 Dinger-Funk -10%
Dinger-Funk 0% Flow
Dinger-Funk +10% KS L 5111 (Korean Industrial Standards 2007d) was used to mea-
60 sure the flow characteristics of fresh WMS mixtures. Once the sur-
Dinger-Funk +20%
face of a flow table was cleaned and dried, a 25-mm layer of fresh
40 concrete was filled in a cylindrical steel mold placed at the center
of the flow table. Subsequently, the concrete layer was tamped
20 20 times using a tamping rod with a consistent pressure. The rest
of the mold was filled with concrete and tamped again, followed by
0 final surfacing was done to keep the top surface flat. After 1 min,
0 1 10 the mold was slowly lifted, allowing the concrete to spread. The
Sieve size [mm] table was then lifted upward by 12.7 mm and then freely dropped
25 times for the next 15 s to spread the concrete sample in a circular
Fig. 3. Modified Dinger-Funk gradations used in this study shape. The flow can be simply determined as the percentage in-
crease in average diameter of the concrete spread over the base
diameter of the mold. The diameter of the concrete spread was mea-
Materials and Mixture Proportions sured in at least four different locations to obtain enough accuracy.
This study used Type I portland cement conforming to ASTM Rheology
C150 (ASTM 2015), whose fineness and specific gravity were The rheological measurements were conducted using an IBB rhe-
3,200 cm2 =g and 3.15, respectively. The chemical compositions ometer (IBB Rheology, Quebec, Canada), which has an H-shaped
of the cement used were 20.8% SiO2 , 6.3% Al2 O3 , 3.2% impeller vertically mounted onto planetary gears extending from
Fe2 O3 , 61.2% CaO, 3.3% MgO, and 2.3% SO3 . The silica fume the motor (Fig. 4), because this apparatus is known to be well-
used in this study was composed of over 97% SiO2 , with a specific suited for monitoring rheological properties of cement-based
surface area of 150,000–300,000 cm2 =g and a specific gravity of
2.22. Powder-type polycarboxylate-based superplasticizer and
sulfonate silica-based air-entraining agent were added as chemical
admixtures.
Table 2 provides the mixture proportions for WMS mixtures
used in this study. The mixture design used in this study was
inspired by one developed by Beaupre (1994), which is considered
a representative mixture design for WMS because it provides
adequate pumpability [degree of mobility and stability of WMS

Table 2. Mixture Proportions for All WMS Mixtures Examined in This

Study
Material Weight per unit volume (kg=m3 )
Cement 400
Water 180
Fine aggregate 1,100
Coarse aggregate 460
Silica fume 40
Air-entraining agent 0.044
Superplasticizer 4.84 Fig. 4. (a) IBB rheometer; (b) H-shaped impeller

© ASCE 04017227-3 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2017, 29(11): 04017227

 160
materials with inclusion of coarse aggregate (Beaupre 1994). The
rheometer test was initiated from placing torque readings of the y = 2.0909x - 383.09
2
built-in strain gauge to a null position and placing fresh concrete 140
R = 0.7344
into the mixing bowl up to 200 mm in height. Subsequently, the

Flow [%]
torque exerted on the impeller as it stirred the mixture with a speci-
fied speed was recorded in real time. Based on the rotational speed 120
and torque measured, torque viscosity and flow resistance of the y = 0.5091x - 5.9091
2
R = 0.2945
mixture were estimated. These torque viscosity and flow resistance
obtained can be converted to the rheological parameters in the 100
Immediately after mixing
Bingham model (i.e., plastic viscosity and yield stress) based on 20 minutes after mixing
the following relationships (Ferraris and de Larrard 1998; de 80
Larrard et al. 1998): 220 230 240 250 260
2π 3 Slump [mm]
Γ0 ¼ ðR − R31 Þτ 0 ð1Þ
3 2 Fig. 5. Relationship between slump and flow values for crushed-
Downloaded from ascelibrary.org by Jung Heum Yeon on 09/12/17. Copyright ASCE. For personal use only; all rights reserved.

aggregate WMS with different gradation standards

2πbþ1
A¼ ðRbþ3 − Rbþ3
1 Þμ ð2Þ
ðb þ 3Þhb 2
where Γ0 = flow resistance (N · m); R2 and R1 = inner and outer standards. In this figure, the y-intercept indicates the flow resis-
radiuses of the sheared concrete sample (mm); τ 0 = yield stress tance, whereas the slope of the linear regression line represents the
(Pa); A = torque viscosity (N · m · s); h = height of the sheared torque viscosity. It can be seen that the data for fine-aggregate mix-
concrete sample (mm); b ¼ 1 = numerical parameter; and μ = tures (i.e., ACI-Fine and EFNARC) were all below those for
plastic viscosity (Pa · s). combined-aggregate mixtures (i.e., Dinger-Funk and ACI-Combined),
indicating that the fine-aggregate mixtures had a smaller flow re-
sistance than the combined-aggregate mixtures. Furthermore, the
Results and Discussion ACI-Fine mixture was positioned at the bottommost of the plot
(lowest flow resistance), and the regression line for ACI-combined
mixture was positioned at the topmost of the plot (greatest flow
Slump, Air Content, and Flow of Crushed-Aggregate
resistance). This finding demonstrates that the flow resistance of
WMS Mixtures with Standard Gradations
a given mixture can be largely dependent on the grading class
Table 3 summarizes the results of slump, air-content, and flow tests (i.e., fine aggregate and combined aggregate) used, even for mix-
for crushed-aggregate WMS mixtures with four different gradation tures with the same specification.
standards. The results have shown that the initial air content for all Fig. 6 also demonstrates that the torque viscosity was overall
the mixtures fell within a target range of 17–19%, indicating a non- similar among the mixtures. When the fine-aggregate mixtures
significant effect of aggregate gradation on the air content of WMS. were considered only, ACI-Fine mixture had a slightly steeper
The slump tended to vary in a narrow range between 230 and slope, and in turn a higher torque viscosity, than EFNARC mixture.
250 mm, with a slightly greater value for the ACI-Fine mixture. Contrary to the torque viscosity, the flow resistance was higher for
It is also interesting that although ACI-Combined and Dinger-Funk the EFNARC mixture than ACI-Fine mixture. Of the combined-
mixtures had an identical slump of 230 mm, the flow of the ACI- aggregate mixtures, the Dinger-Funk mixture was found to have a
Combined mixture was approximately 10–14% greater than that of slightly higher torque viscosity than the ACI-combined mixture,
Dinger-Funk mixture. To investigate the correlation between flow whereas the ACI-combined mixture exhibited a flow resistance
and slump in more detail, two relationships, one between slump almost twice as high as that of the Dinger-Funk mixture.
and flow immediately after mixing and the other between slump To examine the reasons that gave rise to the differences in
and flow 20 min after mixing, are plotted in Fig. 5. As can be seen rheological behavior among the mixtures, each mixture’s fineness
in the results, as the slump increased, both flows at mixing and at modulus was compared with its flow resistance and torque viscos-
20 min tended to increase; however, the identified relationships ity (evaluated based on the data shown in Fig. 6), and the results are
were not sufficiently clear to be generalized.
5
Rheological Behavior of Crushed-Aggregate WMS ACI-Fine
Mixtures with Standard Gradations 4
EFNARC
Applied torque [N·m]

Dinger-Funk
Fig. 6 illustrates the rheometer test results measured from four ACI-Combined
crushed-aggregate WMS mixtures with different gradation 3

2
Table 3. Results of Slump, Air-Content, and Flow Tests for Crushed-
Aggregate WMS Mixtures with Different Gradation Standards
1
Flow (%)
Gradation Slump Initial air Immediately 20 min
0
standard (mm) content (%) after mixing after mixing 0.0 0.3 0.6 0.9 1.2
ACI-fine 250 17 125 148 Speed of rotation [rev/s]
EFNARC 240 18 109 102
Dinger-Funk 230 18 106 97 Fig. 6. Applied torque versus speed of rotation for crushed-aggregate
ACI-combined 230 19 120 107 WMS with different gradation standards

© ASCE 04017227-4 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2017, 29(11): 04017227

 Flow resistance 260 150
5 5 Slump
Torque viscosity
Flow resistance limit: 4.3 N·m Flow at mixting

Torque viscosity [N·m·s]

Flow resistance [N·m]

4 4 250 Flow 20 min after mixing 135

Slump [mm]

Flow [%]
Torque viscosity limit: 2.9 N·m·s
3 3
240 120

2 Dinger- 2
EFNARC ACI- Funk
ACI-Fine 230 105
Combined
1 1

0 0 220 90
0 0.5 1 1.5 2
2.6 2.9 3.2 3.5 3.8 4.1
(a) Flow resistance [N·m]
Fineness modulus [-]
Downloaded from ascelibrary.org by Jung Heum Yeon on 09/12/17. Copyright ASCE. For personal use only; all rights reserved.

260 150
Fig. 7. Variations of rheological parameters for crushed-aggregate
Slump
WMS with different gradation standards (fineness modulus)
Flow at mixting
250 Flow 20 min after mixing 135

Slump [mm]

Flow [%]
presented in Fig. 7. As the results show, however, there was no
apparent relationship identified between mixture’s fineness modu- 240 120
lus and rheological parameters. The findings from this investigation
imply that the fineness modulus of aggregate may not be a direct
indicator characterizing shotcrete rheology. 230 105
In addition, Fig. 7 compares the measured rheological param-
eters with their threshold values suggested by Beaupre (1994). 220 90
It was found that the measured rheological parameters were all 2 2.2 2.4 2.6 2.8
below their upper limits, indicating that the aggregate gradations (b) Torque viscosity [N·m·s]
complying with the described standards and specifications can
be implemented to shotcrete with minimal issues, given that mix- Fig. 8. Relationships between rheological parameters and slump/flow
ture proportioning is adequately provided. In particular, the ACI- for crushed WMS with different gradation standards: (a) flow resis-
Combined mixture is expected to provide good pumpability and tance versus slump and flows measured at mixing and 20 min after
shootability because it had high flow resistance but low torque mixing; (b) torque viscosity versus slump and flows measured at mix-
viscosity (Yun et al. 2015). ing and 20 min after mixing
Variations in rheological parameters with different slump and
flow values are shown in Fig. 8. Both slump and flow gradually
diminished as flow resistance increased. This is because flow Table 4. Results of Fineness Modulus, Slump, Air-Content, and Flow
resistance is proportionally related to the stress level to initiate Tests for Crushed-Aggregate WMS Mixtures with Modified Dinger-
or maintain flow behavior (i.e., yield stress); in other words, if Funk Gradations
the flow resistance increases, the stress required to initiate flow Flow (%)
becomes higher, leading to reductions in slump and flow. The Initial
findings of this study agree well with those of previous studies on air Immediately 20 min
rheology and flow of fresh concrete (Murata and Kikukawa 1992; Fineness Slump content after after
Aggregate gradation modulus (mm) (%) mixing mixing
Laskar 2009; Koehler and Fowler 2010). Furthermore, the mea-
sured slump and flow were compared with the torque viscosity Dinger-Funk −20% (Ca) 4.58 230 19 89 85
estimated by the rheometer tests. The findings revealed that both Dinger-Funk −10% (C) 4.28 220 19 87 84
flows measured immediately after mixing and measured after Dinger-Funk 0% (C) 3.98 230 18 86 84
Dinger-Funk þ10% (C) 3.77 215 18 83 81
20 min increased almost linearly with increased torque viscosity,
Dinger-Funk þ20% (C) 3.57 175 19 61 60
which is contrary to the finding of a previous work (Laskar 2009).
a
The probable reason for this behavior is that torque viscosity gen- Crushed aggregate.
erally affects the flow speed more than the flow diameter itself
(EFNARC 2006). Further in-depth investigations are required in
measurements of WMS mixtures with varying percentages passing
this regard.
on each sieve specified in the Dinger-Funk gradation by up to −20
to 20% by weight. The allowable gradation ranges were suggested
Determination of Allowable Gradation Ranges Based for both crushed aggregate and natural aggregate to compare the
on Modified Dinger-Funk Gradations impact of aggregate type.
Table 4 summarizes the fineness modulus, slump, air content,
Slump, Air Content, and Flow of WMS Mixtures with and flow measured for crushed-aggregate WMS mixtures with
Modified Dinger-Funk Gradations modified Dinger-Funk gradations. The results were all fairly accept-
Allowable gradation ranges resulting in ideal shotcrete behavior able, although the flow values were slightly lower than the standard
were determined on the basis of Dinger-Funk gradation because mixtures detailed in Table 3. The air content was found to be nearly
of its ease of gradation modifications and appropriate levels of similar for all the mixtures with a maximum 1% difference. Also,
flow resistance and torque viscosity. As described previously, there was a general tendency that the slump and flow values (both
the allowable gradation ranges were assessed based on rheological measured immediately after mixing and 20 min after mixing)

© ASCE 04017227-5 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2017, 29(11): 04017227

 Table 5. Results of Fineness Modulus, Slump, Air-Content, and Flow 6
Tests for Natural-Aggregate WMS Mixtures with Modified Dinger-Funk
5
Gradations

Applied torque [N·m]

Flow (%) 4
Initial
air Immediately 20 min
3
Fineness Slump content after after Dinger-Funk -20% (C)
Aggregate gradation modulus (mm) (%) mixing mixing Dinger-Funk -10% (C)
2
Dinger-Funk −20% (Ra) 4.58 245 16 127 100 Dinger-Funk 0% (C)
Dinger-Funk −10% (R) 4.28 225 17 119 89 1 Dinger-Funk +10% (C)
Dinger-Funk 0% (R) 3.98 220 19 92 90 Dinger-Funk +20% (C)
Dinger-Funk þ10% (R) 3.77 210 20 80 79 0
Dinger-Funk þ20% (R) 3.57 220 20 95 87 0 0.3 0.6 0.9 1.2
a (a) Speed of rotation [rev/s]
Natural river gravel.
Downloaded from ascelibrary.org by Jung Heum Yeon on 09/12/17. Copyright ASCE. For personal use only; all rights reserved.

Flow resistance
5 5
Torque viscosity
gradually increased as the fineness modulus increased. This is be- Flow resistance limit: 4.3 N·m

Torque viscosity [N·m·s]

Flow resistance [N·m]
cause mixtures with a higher fineness modulus included a higher 4 4
amount of coarser aggregate (decreased S/a). This in turn resulted Torque viscosity limit: 2.9 N·m·s
in a smaller aggregate specific surface area, which eventually 3 3
reduced the particle-to-particle friction (Mindess et al. 2003). How-
2 2
ever, the effect of fineness modulus decreased to an almost negli-
gible level above a slump of 220 mm and flow of 85 mm.
1 Allowable range: -15% to +20% 1
Fineness modulus, slump, air-content, and flow tests were also
conducted for natural-aggregate WMS mixtures with modified
0 0
Dinger-Funk gradations, and the test results are presented in Dinger- Dinger- Dinger- Dinger- Dinger-
Table 5. It was found that the air contents tended to increase with Funk -20% Funk -10% Funk 0% (C) Funk +10% Funk +20%
decreased fineness modulus, which was contrary to conventional (b) (C) (C) (C) (C)
air-content behavior. The slump and flow values measured immedi-
ately after mixing and 20 min after mixing dramatically increased Fig. 9. Results of rheology tests for crushed-aggregate WMS with
almost linearly with the aggregate’s fineness modulus, which was modified Dinger-Funk gradations: (a) applied torque versus speed of
similar to the behavior observed in the crushed-aggregate WMS rotation; (b) variations in rheological parameters
mixtures.
The results confirm that the slump and flow values were
greater for the natural-aggregate mixtures than crushed-aggregate
result implies that, except for the Dinger-Funk −20% (C) mixture,
mixtures. This is because, as mentioned earlier, the aggregate with
whose torque viscosity surpassed a prescribed limit of 2.9 N · m · s,
smoother surface textures (i.e., natural aggregate) has a smaller sur-
all other mixtures resulted in satisfactory rheological performance.
face contact area covered with the cement paste, forming a rela-
In particular, both of the parameters were kept almost consistent,
tively thicker lubricating layer (Kosmatka et al. 2002; Quiroga and
irrespective of the gradation, within a gradation range of −10 and
Fowler 2004; Hu 2005). On the other hand, the initial air content
þ10%. This is a quite favorable result because the crushed aggre-
had no consistent relationship with aggregate type.
gate may be accepted by a wide range of gradations with minimal
Rheological Behavior of WMS Mixtures with Modified issues on pumpability and shootability.
Dinger-Funk Gradations To quantitatively identify how effectively the crushed aggregate
Fig. 9(a) presents the relationships between applied torque and can be used in shotcrete, the rheological properties of natural-
rotational speed for crushed-aggregate WMS mixtures with modi- aggregate shotcrete mixtures with identical gradations were also
fied Dinger-Funk gradations. Here, 20% differences with re- evaluated. Fig. 10(a) illustrates the applied torque and impeller’s
spect to the original Dinger-Funk gradation [i.e., Dinger-Funk revolving speed relationships for natural-aggregate WMS mixtures
−20% (C) and Dinger-Funk þ20% (C)] resulted in substantial with modified Dinger-Funk gradations. It can be clearly seen that
changes in the rheological measurements, whereas 10% differences the torque viscosity became greater as the content of coarser aggre-
[i.e., Dinger-Funk −10% (C) and Dinger-Funk þ10% (C)] exhib- gate increased (increased fineness modulus). In addition, the flow
ited only slight variations in their rheology; when the original resistance significantly increased as the content of finer aggregate
gradation was factored by þ20% by weight, the flow resistance increased (decreased fineness modulus). The flow resistance of
increased by 64% along with a reduction in torque viscosity by Dinger-Funk þ10% and Dinger-Funk þ20% mixtures was approx-
31%, and when factored by −20% by weight, the torque viscosity imately twice as high as that of the original Dinger-Funk gradation
increased by approximately 16% while keeping an almost identical mixture, even though the difference between those two was quite
flow resistance. These findings indicate that a 20% increase in small. When compared to the crushed-aggregate WMS mixtures,
the coarser-aggregate portion affected the torque viscosity only, small differences were observed but some similarities could be
whereas a 20% decrease in the coarser-aggregate portion affected found in their general rheological behavior patterns—a decrease
both flow resistance and torque viscosity. in fineness modulus mainly affected the flow resistance, whereas
Variations in flow resistance and torque viscosity evaluated an increase in fineness modulus mainly affected the torque viscosity.
based on Fig. 9(a) are shown in Fig. 9(b), along with their threshold Fig. 10(b) plots the variations in flow resistance and torque vis-
values proposed by previous research (Beaupre 1994). In the plot, cosity of natural-aggregate WMS mixtures with different degrees
the shaded area represents an allowable gradation range meeting of gradation modification. The upper limits for each rheological
both performance-related rheological criteria simultaneously. The parameter are also presented. Again, the shaded area refers to an

© ASCE 04017227-6 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2017, 29(11): 04017227

 6
• The fine-aggregate gradation mixtures had a smaller flow re-
Applied torque [N·m]
5 sistance than the combined-aggregate gradation mixtures, but
the torque viscosity was nearly consistent for all the mixtures.
4 This finding indicates that the torque viscosity is less sensitive to
aggregate gradation standards compared to the flow resistance;
3 • Regarding shotcreting performance, the gradation standards and
Dinger-Funk -20% (R)
specifications described in this study could be adopted for the
2 Dinger-Funk -10% (R)
production of crushed-aggregate WMS mixtures with minimal
Dinger-Funk 0% (R)
Dinger-Funk +10% (R)
issues. In particular, the ACI-combined gradation is anticipated
1
Dinger-Funk +20% (R)
to provide stable pumping and shooting performances;
0 • Investigations on the modified Dinger-Funk mixtures revealed
0 0.3 0.6 0.9 1.2 that a reduction in fineness modulus mainly increased the flow
(a) Speed of rotation [rev/s] resistance, whereas an increase in fineness modulus mainly
increased the torque viscosity;
• Allowable gradation ranges for crushed-aggregate WMS and
Downloaded from ascelibrary.org by Jung Heum Yeon on 09/12/17. Copyright ASCE. For personal use only; all rights reserved.

Flow resistance
5
Torque viscosity
5 natural-aggregate WMS were between −15 and þ20% and be-
Flow resistance limit: 4.3 N·m tween 0 and þ20%, respectively. These findings demonstrate

Torque viscosity [N·m·s]

Flow resistance [N·m]

4 4 that crushed aggregate may provide a greater benefit in shotcrete

Torque viscosity limit: 2.9 N·m·s production than natural aggregate; and
3 3 • Crushed aggregate can be an effective alternative aggregate
source for shotcrete in that it suitably meets both industrial
2 2
demands and shotcreting performance criteria.
The results outlined in this study may serve as a preliminary and
1 1
Allowable range: 0% to +20% indirect option to check the applicability of crushed aggregates in
shotcrete when actual shotcrete data are not available for a given
0 0
Dinger- Dinger- Dinger- Dinger- Dinger- project. Further research is required to identify actual shotcrete
Funk -20% Funk -10% Funk 0% (R) Funk +10% Funk +20% behavior such as pumpability and shootability with varying aggre-
(b) (R) (R) (R) (R)
gate gradations and aggregate types.
Fig. 10. Results of rheology tests for natural-aggregate WMS with
modified Dinger-Funk gradations: (a) applied torque versus speed of
rotation; (b) variations in rheological parameters Acknowledgments

This research was supported by Grants 13RDRP-B066780 and

15TLRP-B079261-02, funded by Ministry of Land, Infrastructure
allowable gradation range that satisfies both of the rheological and Transport of the Korean government and was performed using
performance criteria at the same time. The results indicate that the facilities of the Institute for Advanced Construction Materials at
so long as natural aggregate is used for WMS, there should be Kangwon National University, Chuncheon, South Korea.
no issues with shootability because the measured flow resistance
was consistently below the threshold value with sufficient margins
throughout a given range. However, the torque viscosity, which is References
associated with shootability of WMS, tended to far exceed the ACI (American Concrete Institute). (2014). “Specification for shotcrete.”
threshold level of 4.3 N · m when the fineness modulus increased ACI 506.2-13, Farmington Hills, MI.
from its original value. Based on these findings, the optimum gra- ACI (American Concrete Institute). (2016). “Guide to shotcrete.”
dation range for natural-aggregate WMS meeting both pumping ACI 506R-16, Farmington Hills, MI.
and shooting performance was found to be between 0 and þ20%. ASTM. (2008a). “Standard specification for concrete aggregates.”
In sum, the allowable gradation range for crushed-aggregate ASTM C33/C33M-08, West Conshohocken, PA.
WMS mixtures was found to be approximately twice as wide as ASTM. (2008b). “Standard specification for materials for shotcrete.”
that for natural-aggregate WMS mixtures. The findings confirm ASTM C1436-08, West Conshohocken, PA.
ASTM. (2015). “Standard specification for portland cement.”
that crushed aggregate has a rather wider acceptability than natural
ASTM C150/C150M-15, West Conshohocken, PA.
aggregate, providing a greater benefit in shotcrete production. Beaupre, D. (1994). “Rheology of high performance shotcrete.” Ph.D.
dissertation, Univ. of British Columbia, Vancouver, BC, Canada.
Choi, S. Y. (2008). “Rheology and air void structures of wet-mix shotcrete.”
Conclusions and Recommendations Ph.D. dissertation, Kangwon National Univ., Chuncheon, Korea.
de Larrard, F., Ferraris, C. F., and Sedran, T. (1998). “Fresh concrete:
The research described in this study was intended to evaluate the A HerscheI-Bulkley material.” Mater. Struct., 31(7), 494–498.
viability of using crushed aggregate, the most common by-product EFNARC (European Federation of National Associations Representing
of tunnel-excavation processes, as an alternative aggregate source for Concrete). (1996). European specification for sprayed concrete,
for concrete production, especially for implementation in wet-mix Brussels, Belgium.
EFNARC (European Federation of Producers and Applicators of Specialist
shotcrete. The key findings obtained from this study can be sum-
Products for Structures). (2006). Guidelines for viscosity modifying
marized as follows: admixtures for concrete, Brussels, Belgium.
• Results obtained from the crushed-aggregate WMS mixtures Ferraris, C. F., and de Larrard, F. (1998). “Testing and modelling of fresh
with four different standard gradations showed that no consis- concrete rheology.” NISTIR 6094, Building and Fire Research Labora-
tent relationship exists between fineness modulus and rheologi- tory, National Institute of Standards and Technology, Gaithersburg,
cal properties; MD.

© ASCE 04017227-7 J. Mater. Civ. Eng.

J. Mater. Civ. Eng., 2017, 29(11): 04017227

 Funk, J. E., and Dinger, D. R. (1994). Predictive process control of Kosmatka, S. H., Kerkhoff, B., and Panarese, W. C. (2002). Design
crowded particulate suspensions: Applied to ceramic manufacturing, and control of concrete mixture, 14th Ed., Portland Cement Associa-
Kluwer Academic, Boston. tion, Skokie, IL.
Geiker, M. R., Brandi, M., Thrane, L. N., and Nielsen, L. F. (2002). “On Laskar, A. I. (2009). “Correlating slump, slump flow, Vebe and flow tests
the effect of coarse aggregate fraction and shape on the rheological to rheological parameters of high-performance concrete.” Mater. Res.,
properties of self-compacting concrete.” Cem. Concr. Aggregate, 12(1), 75–81.
24(1), 3–6. Leung, C. K. Y., Lai, R., and Lee, A. Y. F. (2005). “Properties of wet-mixed
Goodier, C. I., Austin, S. A., and Robins, P. J. (2008). “Low-volume wet- fiber reinforced shotcrete and fiber reinforced concrete with similar
process sprayed concrete: Hardened properties.” Mater. Struct., 41(1), composition.” Cem. Concr. Res., 35(4), 788–795.
99–111. Malhotra, V. M. (1964). “Correlation between particle shape and surface
Hu, J. (2005). “A study of effects of aggregate on concrete rheology.” texture of fine aggregates and their water requirement.” Mater. Res.
Ph.D. dissertation, Iowa State Univ., Ames, IA. Stand., 4(12), 656–658.
Jolin, M., and Beaupre, D. (2003). “Understanding wet-mix shotcrete: Mindess, S., Young, J. F., and Darwin, D. (2003). Concrete, 2nd Ed.,
Mix design, specifications, and placement.” Shotcrete, 6–12. Prentice Hall, Upper Saddle River, NJ.
Murata, J., and Kukawa, H. (1992). “Viscosity equation for fresh
Jolin, M., Burns, D., Bissonnette, B., Gagnon, F., and Bolduc, L.-S. (2009).
concrete.” ACI Mater. J., 89(3), 230–237.
“Understanding the pumpability of concrete.” Proc., Shotcrete for
Downloaded from ascelibrary.org by Jung Heum Yeon on 09/12/17. Copyright ASCE. For personal use only; all rights reserved.

Quiroga, P. N., and Fowler, D. W. (2004). “The effect of aggregates char-

Underground Support XI, Davos, Switzerland.
acteristics on the performance of portland cement concrete.” ICAR Rep.
Koehler, E. P., and Fowler, D. W. (2010). “Comparison of workability
104-IF, International Center for Aggregates Research, Univ. of Texas at
test methods for self-consolidating concrete.” J. ASTM Int., 7(2),
Austin, Austin, TX.
1–19. Thalmann-Suter, C. N. (1999). “Concrete aggregate production with TBM
Korean Industrial Standards. (2002). “Materials for shotcrete.” KS F 2577, muck explained on the AlpTransit tunnel projects.” Utilizing Ready Mix
Seoul. Concr. Mortar, 131–144.
Korean Industrial Standards. (2006). “Standard test method for air content Weiss, R. (2007). “Concrete and shotcrete with crushed aggregates: Expe-
of fresh concrete by the pressure method.” KS F 2421, Seoul. rience from the AlpTransit construction site Mitholz, underground
Korean Industrial Standards. (2007a). “Aggregate for concrete.” KS F 2526, space–The 4th dimension of metropolises.” Proc., World Tunnel
Seoul. Congress 2007 and 33rd ITA/AITES Annual General Assembly, Taylor &
Korean Industrial Standards. (2007b). “Method of test for slump of Francis, London.
concrete.” KS F 2402, Seoul. Wills, M. H., Jr. (1967). “How aggregate particle shape influences concrete
Korean Industrial Standards. (2007c). “Sampling method for fresh mixing water requirement and strength.” J. Mater., 2(4), 843–865.
concrete.” KS F 2401, Seoul. Yun, K.-K., Choi, P., and Yeon, J. H. (2015). “Correlating rheological
Korean Industrial Standards. (2007d). “Flow table for use in tests of properties to the pumpability and shootability of wet-mix shotcrete
hydraulic cement.” KS L 5111, Seoul. mixtures.” Constr. Build. Mater., 98, 884–891.

© ASCE 04017227-8 J. Mater. Civ. Eng.

View publication stats J. Mater. Civ. Eng., 2017, 29(11): 04017227