Raft ​Foundations

Design and Analysis with a Practical Approach

SHARAT CHANDRA CUPTA

Advisor, Indian Buildings Congress, Former Chief Engineer Central ​Public ​Works Department

PUBLISHING ​FOR ONE ​WORLD ​

NEW AGE INTERNATIONAL (P) LIMITED, PUBLISHERS ​New

Delhi ​- B​ angalore Calcutta Chennai Guwahati Hyderabad Lukhnow ​Mumbai ​. P​ une

Copyright ​O ​1997 New Age International (P) Limited, Publishers

NEW AGE INTERNATIONAL (P) LIMITED, PUBLISHERS

Ballygunge Circular Road, Calcutta-700 019 ​: ​20,
NEW DELHI BANGALORE IInd Main Road Kasthuribai Nagar, Adyar,
Chennai-600 020 ​: ​Pan Bazar, Rani Bari,
CALCUTTA CHENNAI
Guwahati-781 001 ​: ​1-2-4 1219, Gaganmahal, Near
A.V. College, Domalguda,
GUWAHATI HYDERABAD
Hyderabad-500 029 ​: ​18, Madan Mohan Malviya
Marg, Lucknow-226 001 ​: ​1281A. Noorani Building,
LUCKNOW MUMBAI Block No. 3, First Floor.
L.J. Road, Mahim, Mumbai-400 016. ​: ​44, Prashant
PUNE Housing Society, Lane No. 6, Paud Road,
: ​4835124, Ansari Road, Daryaganj, New Delhi-110
Kothrud, Pune-4 ​1 ​1029.
002 :​ ​35, Annapoorna Complex, South End Road,
Basavangudy, Bangalore-560 004 ​: ​4018,

This book cr any-part there of may not be reproduced in any form without the written permission of the publisher

This book is not to be sold outside the country to which ​it ​is consigned by New Age International (P) Limited

ISBN ​: ​81-224-1078-2

Published by H.S. Poplai for New Age International (P) Limited, 4835124, Ansari Road, Daryaganj, New Delhi-
110002. Typeset by EPTECH, and printed ​ai ​Ram Printograph, C-114, Okhla Industrial Area, Phase I, New
Delhi-110020.
Printed in India ​Production ​: ​M.I. Thomas

CONTENTS
Preface ​i ​t ​I 1.
​ ​INTRODUCTION
2. ​NEED OF RAFT FOUNDATION
3. ​TYPES OF RAFT FOUNDATION
4. ​SURVEY OF AVAILABLE LITERATURE ​
I​ 1 Foundation Engineering by Peck, Hansen and

Thornburn ​
I​ 4.2 Foundation Design and Practice by Elwyn. E.S. Seelye ​
1
4.5 Indian Standard Code of Practice
4.3 Foundation Design by Teng 4.4 Foundation of Structures by Dunham ​i ​
for Design and Construction of Raft

Foundation ​- ​IS 2950-1965 Raft Foundation ​- T

​ he Soil Line Method of Design by A.L.L. Baker Indian Standard
Code of Practice for Design and Construction of Raft Foundation 1.S ​: ​2950 (Part-I) 1973 Foundation
Engineering Handbook Edited by' Hans F. Winterkorn ​& ​Hsaiyang Fang Foundation Analysis and Design by
Joseph. E. Bowels Building Code Requirements for Reinforced Concrete (ACI 31 8 ​- ​77) Foundation Design and
Construction by M.J. Tomlinson Design of Combined Footings and Mats ACI Committee 336 Pile Foundation
Analysis and Design by H.G.Poulos and E.H. Davis 1980 Reinforced Concrete Designers Handbook by Charles
E. Reynolds and James C. Steedman ​- ​9th Edition 1981 ​IS ​2950 ​(Part ​I) ​1981 -Code for Design and Construction
of Raft Foundation Part I ~esi~n Eleventh International Conference of Soil Mechanics and Foundation
Engineering San Francisco, August 12 ​- ​16,1985 Foundation Design and Construction by M.J. Tomlinson, 5th
Edition, 1986

CONTENTS ​
i
Handbook of Concrete Engineering -Mark Fintel ​- ​2nd Edition, 1986 Reinforced Concrete Designer Handbook by
Charles E. Reynolds and James Steedman, 10th Edition, 1988 Building Code Requirements in Reinforced
Concrete ​- ​ACI ​- ​3 18 ​- ​1989 Foundation Engineering Hand book by Hsai-Yang-Fang 2nd Edition, 1991 Design
of Combined Footings and Mats ​- ​ACI committee 336 2R ​- 8​ 8 Published in ACI Manual 1993 Foundation
Analysis and Design by Bowles, 4th Edition, 1988 Proceedings of Indian Geo-Technical Conference 1992,
Calcutta, December, 1992 Designs of Foundation Systems ​- ​Principles and Practices by Nainan P. Kwian, 1992
13th International Conference on Soil Mechanics and Foundation Engineering, New Delhi, January, 1994 Soil
Structure Inter-action -The Real Behaviour of Structures, published by the Institution of Structural Engineers,
U.K. The Institution of Civil Engineers, U.K. International Association for Bridge and Structural Engineering in
March, 1989

5. ​DESIGN APPROACH AND CONSIDERATIONS

5.1 Rigid Approach 5.2 Flexible Approach 5.3 Parameters for Raft Design 5.4 Pressure Distribution Under the
Raft 5.5 Rigidity Criteria
5.5.1 Proposed by IS ​: ​2950 (Part I) 1981 5.5.2 ACI Committee, 336 5.5.3 Hetenyi's Criteria 5.6 Modulus of
Sub-Grade Reaction
5.6.1 Recommended by Bowles 5.6.2 IS ​: ​2950 Part I Indian Standard Code of Practice for Design and
 Construction of Raft Foundation 2950 ​- 1​ 981 5.6.3 I.S. 9214-1979 ​- ​Method of Determination of Modulus of
Subgrade
Reaction (k ​value) ​of Soils in Field 5.6.4. IS 8009 ​- ​Part I ​- ​1978. Code of Practice for Calculation of Settlements
of
Foundations ​- ​Part I ​- ​Shallow Foundations. Subjected to Sy_mmetrical Static Vertical Load. 5.6.5
Recommendation by Alpan and Prof. Alarn Singh 5.6.6 Summary

6. STRUCTURAL DESIGNERS DILEMMA

7. ​STUDIES ​CARRIED OUT ON ​EFFECT ​OF VARIOUS PARAMETERS ON DESIGN OF ​RAFT

38 ​7.1 Study ​1 ​40 7.1.1 Examples Selected 41 7.1.2 Raft Size 41
CONTENTS

5 ​7.1.3 Soil Investigation ​! ​7.1.4 Load Considered in Study

7.1.6 ​ Discussions of Results
7.1.5 Analysis ​: ​ '​

7.1.7 Conclusions 7.2 Study 2 -Effect of Horizontal Loads 7.2.1 Example Selected 7.2.2 Discussion of Results
7.2.3 Conclusion 7.3 Study 3: Comparison with Conventional Rigid Methods
7.3.1 Details of Conventional Method: Combined Footing Approach 7.3.2 Examples Selected 7.3.3 Discussion of

Results 7.3.4 Inverted Floor Method 7.3.5 Conclusions ​

1
7.4 Study 4. Another Office Building 7.4.1 Example Details 7.4.2 Comparison of Results ​
t ​ 1 7.4.3
t​ ​ Discussions of

Results ​! ​7.4.4 Conclusions ​

I 8​ . ​STUDIES CARRIED OUT ON ANALYSIS AND DESIGN OF PILED
RAFTS 1​ ​t i​ 8.1
​ Design Procedures being Used ​! ​8.2 Example Selected
8.3 Soil Data 8.4 Methods of Analysis Studied
8.4.1 Conventional Rigid Method with Simplified Models
8.4.1.1 Combined footing approach 8.4.1.2 Continuous beam analogy :inverted floor 8.4.1.3 Comparison of
results 8.4.2 Piled RafPAnalysis Based on Finite Element Approach 8.5 Study of Parameters Influencing the Raft

;​
Behaviour ​: ​8.5.1 Effect of Raft Stiffness on the Pile Loads and Raft Moments ​ 8.5.2 Effect of Superstructure

and Retaining ​Walls ​on Foundation Stiffness ​ 8.5.3 Effect of Earthquake Loads and Moments ​ 8.5.4 Effect of End
I​ I​

Bearing and Friction Piles 8.5.5 Summary of Results ​I ​8.6 ​Discussions 8.7 Conclusions
9. JOINTS IN RAFl'S
10. ​SUMMARY OF STUDIES
11. ​FACTORS AFFECTING CHOICE OF MET,HOD OF ANALYSIS
12. GUIDELINES APENDM ​- ​ILLUSTRATIVE EXAMPLES ​A.l ​Conventional Rigid Method ​- ​Combined
footing approach ​A.2 ​Flexible Raft ​- B
​ eam on elastic foundation ​A.3 ​Piled Raft-Plate on elastic foundation
CONTENTS

INTRODUCTIO
N
I

In ​1957, ​when the author was a student of Civil Engineering at the Indian Institute of Technology,
Kharagpur, the first institute of national importance, one of his professors of Civil Engineering at his first
lecture in the class said:
"Civil Engineering is ​50% ​common sense but common sense is that sense which is quite uncommon. "​ ​After ​34
years of experience in Civil Engineering construction and design, the author only wonders how true the
statement of his Professor was and how much more it is true in case of foundation engineering. ​1.1
Foundation engineering has been practised as an a​ rt, ​without help of science, since time immemorial upto
1920 when it had achieved a considerable amount of refinement. It was in the earlier 1920s that a​ ​concerted

!​ effort was made to study and undentand the physical laks governing the behaviour of sub surface

materials, ​\ i​ .e.. ​soil from which foundations derived their support and on whose behaviour its own
behaviour depends. This
​ is the time when study of soil mechanics was started and it was in 1919 when Karl
Terzaghi, popularly known as 'father of soil mechanics', made successful attempt to explain the
phenomenon of settlement oti a scientific basis. Though study of soil mechanics has provided us with new
techniques for selecting appropriate type of foundation and predicting the behaviour of completed
structures, it has not been able to decrease the importance of the accumulated experience of the ages.
Amount of uncertainty and degree of variation in the properties of soil and number of parameters on which
 performance of a foundation depends, make exact solution impractical, if not impossible. With so much of
advancement in science and computer application, structural design is still defined as:I5
a creation of a structural fonn to satisfy a number of requirements. It is a combination of art and science. As a
rule, there is ​no ​direct procedure leading to the solution of a specific problem. An engineer uses all
his resources of knowledge experience and imagination to produce a trial scheme. He then constructs
a mathematical model of such a solution to assess its adequacy and ifnecessary, modifies the original
concept in the light of analytical results. The process is repeated until the designer is satisfied with
thefinalproduct, taking into account not only structural adequacy but also such non-quantifiable
factors as aesthetics, ease of construction and performance. The design process is characterised by a
complex interaction of parameters a ​ nd t​ he need to arrive at decisions based on incomplete data
Intuitive decisions which have to be taken, appear to be diametrically opposite to the logical nature
of .​ .. ​'
2 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

Foundation design and analysis is, at a stage behind structural analysis and design for
superstructure, and even now continues to be practised more as an ​art ​and will probably continue to be
done so, for many years to come. ​1.2 ​Available textbooks, handbooks, various publications and papers
give widely different approaches to design of raft foundations. A designer, when faced with a task of
designing a raft foundation, finds himself in a precarious position where he has to balance the time
available for design, the cost of design, the need of adequate safety and, above all, acceptance of the
design by the client and the professional c​ ommunity ​in general and decide the method of design to be
followed by him. Generally, it is not practical for any designer to go through the various approaches as
available in engineering literature at a particular time, compare their merits and demerits and select the
most suitable for his purpose. He, therefore, perforce selects a particular textbook and applies the same to
his problem, quite often little realising that the theoretical problem dealt with in the textbook is widely
different from his practical problem relating to an actual building. Resulting solution may not be ​as
​ s ​he ​feels.
satisfactory a
An effort has been made in the following chapters to explain the various approaches suggested in
literature, give their comparative limitations, examine the implications of the so-called more sophisticated
approaches and finally make recommendation for the method which can be followed by a designer till he
accumulates enough experience so ​as ​to select his own method particularly applicable to his problem.
Intention of this publication is not to hinder initiative of an individual in going deeper in any problem, but
to give him a comparative idea of available approaches with sufficient number of references which he can
study during the beginning of his profession and formulate his own opinion in due course but still
continuing to design satisfactory raft foundations.
This publication should, therefore, be studied in this
background.
 NEED OF RAFT
FOUNDATION.

Raft or Mat foundation is a combined footing that covers the entire area beneath a structure and supports
all walls and columns. This raft or mat normally rests directly on soil or rock, but can also be supported
on piles as well.
Raft foundation is generally suggested in the following
situations:
(a) Whenever building loads are so heavy or the allowable pressure on soil so small that individual
footings would cover more than floor area. (b) Whenever soil contains compressible lenses or the soil is
sufficiently erratic and it is difficult to define and assess the extent of each of the weak pockets or
cavities and, thus, estimate the overall and differential settlement. (c) When structures and equipment
to be supported are very sensitive to differential settlement. (d) Where structures naturally lend
themselves for the use of raft foundation such as silos, chimneys,
water towers, ​elc. (​ e) Floating foundation cases wherein soil is having very poor bearing capacity and the weight
of the
super-structure is proposed to ​be ​balanced by the weight of the soil removed. ​(f) ​Buildings where basements are
to be provided or pits located below ground water table. (g) Buildings where individual foundation, if
provided, will be subjected to large widely varying bending moments which may result in differential
rotation and differential settlement of individual footings causing distress in the building. Let us now
examine each of the above situations in greater detail. ​2.1 ​In case of soil having low bearing pressure, use
of raft foundation gives three-fold advantage:
(a) Ultimate bearing capacity increases with increasing width of the foundation bringing deeper soil
layers in the effective zone. (b) Settlement decreases with increased depth. (c) Raft foundation equalises the
differential settlement and bridges over the cavities. Every structure has a limiting differential
settlement which it can undergo without damage. ​The ​amount of differential settlement between
various parts of a structure supported on a mat foundation is much lower than that if the
sarne.structure was supported on individual footings and had undergone the same amount of
maximum settlement. With these considerations, maximum total settlement which
RAFT FOUNDATIONS-DESIGN AND
ANALYSIS
can be allowed for a particular structure on mat foundation is more than what is permitted when the
structure is resting on individual footings. This, therefore, allows a higher bearing capacity for such
situations. It may, however, be noted that if in a case deeper layers of soil are of very poor quality,
increase in width of the foundation may not always lead to higher bearing capacity. In situation where
comparatively shally top layers of soil are underlain with deeper layers of much poorer soils, it may be
advantageous to provide individual footings so that the zone of influence of the footings remains within
the top stronger layer. In such a situation, provision of a mat foundation may ​be ​disadvantageous. ​2.2
Some designers work on the rule that if more than ​50% ​of the area of the structure is occupied by
individual footings, it is necessary to provide an overall raft. This is not true and quite often, the quantity
of reinforcing steel and concrete required to avoid excessive deflection and cracking of a raft carrying
unequal column loads, necessitating carry-over of stresses from one part of the raft to the other part, may
be large and may make raft foundation uneconomical. In such situations, it may be more economical to
excavate the entire site to a level formation, construct individual closed space footings (sometimes
touching each other) and then backfill around them. In these cases, however, one must weigh form work
costs against the extra footing material required by using mat foundation. It should be considered that it is
possible to construct alternate footings by using spacer pads against already laid footings and thus save
form work cost.
Quite often, doubt exists about the structural behaviour of individual footings touching each other. This
problem of interaction of footings has been studied by many researchers. It has been reported that the effect of
adjacent footings may vary considerably with the angle of shearing resistance. For low values, they are

negligible though for high values they appear to be significant, particularly if a footing is surrounded by other ​
I

footings on both sides. It is also stated that these effects are considerably reduced as length over breadth ratio ​I

of the footings approaches unity. There are practically no such effects in the case of punching shear failure. ​1

For these and other reasons, it has been recommended that interference effects need not be considered in
designs. Adesigner should, however, be aware of the possibility of their existence in some special circumstan- ​11

ces ​. ​I 2​ .3 ​Situations exist in practice w h p a soil stratum contains compressible lenses or the soils have a
formation where
​ individual layers of soil are neither parallel nor can be reasonably stratified into different layers
of known properties to enable calculations of settlement to a reasonable accuracy. In such situations, individual
footings, if provided, would undergo widely varying settlements resulting in large differential settlement which

I​
cannot ​ be tolerated by the structure. ​I​ 2.4 Situations, as mentioned in (c) and (d) above, are explicit

and do not require further explanation. These are

​ special cases, and adoption of raft foundation is more or less
necessary by the particular nature of the problem involved. ​2.5 ​In cases where soil is very soft and highly
compressible and the buildings cannot be founded on such soils in normal circumstances, it may be possible to
provide the building with a basement in such a manner that weight of the structure is equal to the weight of the
soil removed and, thus, there being no change in the stresses in the soil beneath the basement and, therefore,
little settlement. However, in practice it is rarely possible to balance the loading so that no additional pressure
comes on the soil. However, in such cases still, it is only a part of the total load which comes on the bottom soil
and, thus, it is possible to construct a building inducing a much larger load than the soil would have otherwise
supported. The basement provided, gives additional space in the building for the owner and can be made use of.
However while constructing such foundations,
NEED OF RAFT FOUNDATION ​5

reconsolidation of the soil, which has swelled as a result of removal of over burden pressure in excavating for the
sub-structure, should always be considered and necessary steps be taken to prevent detrimental effects. ​2.6
Basements located below ground water table should use a mat as their base to provide water tight c.onstruction. The
alternative of having individual columns footings connected by thin slabs has not proved to ​be ​successful in most of
the cases; presents difficulties in water proofing; causes concentration of stresses at the junction of the thin slabs and
footings and also at the junction of basement walls and raft causing cracks to develop. This arrangement, therefore,
should not be resorted to unless the economy is of such a magnitude as to outweigh all other considerations.
Even in cases where sub-soil water level is low and basement does not extend below ground water table, long-term
built up of surface water accumulating against basement walls and bottom should be allowed for. This is particularly
so in case of impermeable soils (permeability co-efficient below ​0.1 ​mm per second) or of large surface areas
draining towards the building. ​i.e., ​areas on sloping ground near hillocks. The basement ​ walls should also normally
'​

be designed as self-supporting cantilever retaining walls even though they may ​eventually be strutted by floor
construction. It is inconvenient and often impossible to provide temporary raking struts to support a basement
retaining wall until such time ​as ​strutting given by ground floor or intermediate basement floor is completed. ​2.7
Situations also arise when isolated footings are subjected to very large eccentric loadings, and one is faced with the
possibility of excessive footing rotation, excessive differential settlement or possibility of exceeding the allowable
bearing capacity of the soil at some location. This can happen when the building consists of shear walls and
columns, shear walls sharing most of the horizontal load subjecting its footings to larger settlements and rotation,
decreasing the effectiveness of the shear walls and also creating difficulties by way of large differential settlements.
Raft, if provided, will even out these deformations.
Mats or rafts are supported on piles'in cases where sub-soil conditions warrant provision of piles, but one has to have
the basement. In such situations, raft also helps in making the basement water tight.
It would, therefore, be seen that it is not possible to lay down hard and fast rules defining situations wherein a raft
foundation is required. The author, therefore, opines that every designer should learn all that he can within reason
about the conditions at site, determine the types of foundations that are practical, compare their cost, suitability, ease
of construction, safety and select a type which in his judgement would serve the purpose well. There can always be
differences of opinion about the solution decided by him, but as already mentioned in chapter ​I , ​it cannot be helped
because foundation design still continues to be practised more as an art than an exact science. Two artists seldom
agree.
TYPES OF RAFT FOUNDATION

Raft can be classified into various types on the basis of criteria used for classification. ​3.1 ​Based on the method of
their support, raft can ​be:
(a) Raft supported on soil, (b) Raft supported on piles, and (c) Buoyancy raft. ​3.2 ​On the basis of structural system
adopted for the structure of the raft, these can be classified as:
(a) Plain slab rafts which are flat concrete slabs having a uniform thickness throughout. This can be
with pedestals or without pedestals. (b) Beam and slab raft which can be designed with down stand beam or
upstand beam systems. (c) Cellular raft or framed raft with foundation slab, walls, columns and one of the floor
slabs acting
together to give a very rigid structure. Raft of uniform depth is most popular due to its simplicity of design and
construction. This type is most suitable where the column loads are moderate and the column spacing fairly small
and uniform. Pedestals are utilised to distribute the load on a bigger area in case of heavy column loads.

3.3 ​Slab and beam raft is used as a foundation for heavy buildings where stiffness is the principal requirement to
avoid excessive distortion of the super structure as a-result of variation in the load distribution over the raft or the
compressibility of the supporting soil. These rafts, however, have many obvious difficulties. If the beams are deep,
ribs placed below the basement floor or raft, the bottom of the excavation becomes badly cut up with trenches,
impairing the bearing value of the soil because of its disturbance. Water proofing in case of basements becomes
more complicated arid involved. If the beams are projecting up, usefulness of the basement is destroyed unless the
entire foundation is lowered and the gap filled up or an upper slab is provided supported on these inverted beams to
form the ground floor of the structure. ​3.4 ​Buoyancy raft are necessarily to be provided with a basement so that the
weight of the soil removed balances to a large extent, the imposed load. Cellular raft consisting of foundation
slabs, walls, columns and ground floor slab can ​be ​designed, but it creates considerable amount of uncertainties,
difficulty of construction and quite often even ​in ​such cases, raft is designed as a slab of uniform rhickncss.
TYPES OF RAFT FOUNDATION ​7
Raft, as a slab of uniform thickness, has an additional advantage of providing better water-proofing
treatment ease of reinforcement fabrication and laying of concrete. This type of raft is most commonly used.
Various types of rafts are shown in Fig. ​3.1

RAFT ​
--------------------- SUPPORTED
​ ​ON ​PILE RAFT ​
-- ------- SUPPORTED
​ ​
​ SOIL ​------------- BUOYANCY
------------- ON ​
RAFT FLAT ​
​
--------------- PLATE RAFT ​i ​ FLAT -------------------------
'​
:​
​ ​PLATE WITH PEDESTALS BEAM ------------------
​
 AND SLAB RAFT
Fig. ​3.1 ​Various types of rafts

FRAMED ​
----------- RAFT
​

SURVEY ​OF
AVAILABLE
LITERATURE

Testbooks and design manuals by various authors suggest varying approaches to analysis and design of
raft foundation. Differences of opinion exist in the method of analysis proposed to be adopted while
determining moments, shear forces for the design of raft. Once the bending moments and shear forces
are known, structural design does not present any difficulty and there exists no difference of opinion in
this respect except very minor difference relating to desired thickness of slab and the effectiveness of the
shear reinforcement
Methods suggested by different authors are summarised below. These have been arranged
chronologically with reference to date of publication of the testbooktdesign handbook.

​ eck, Hansen and ​hornb

4.1 ​Foundation Engineering ​by P

burn^
Raft is usually regarded and designed as an inverted continuqus flat slab floor supported without any
upward deflections at the columns and walls. The soil pressure acting against the slab is commonly
assumed to be uniformly distributed and equal to the total of all column loads multiplied by appropriate
load factors and divided by the area of the raft. The moment and shears in the slabs are determined by the
use of appropriate coefficient listed in the specifications for the design of flat slab floors. On account of
erratic variation in compressibility in almost every soil deposit, there are likely to be correspondingly
erratic deviations of the soil pressure from the average value. Since the moment and the shears are
determined on the basis of the average pressure, it is considered good practice to provide this slab with
more than theoretical amount of reinforcement and to use the same percentage of steel at top and bottom.
This method has been widely used, often with complete success. On the other hand, it has also sometimes
led to structural failure not only of the slab but also of the super structure. Therefore, its limitations must
be clearly understood. The analogy follows only if the differential settlement between columns will be
small and if the pattern of the differential settlement will be erratic rather systematic. The method is valid
when the columns ​are ​more or less equally loaded and equally spaced. If the downward loads on some
areas are on the average much heavier than on others, differential settlements may lead to substantial
re-distribution of moments in the slabs resulting in unconservative design. Rafts are sometimes designed
as if they rested on a ​bed ​of closely and equally spaced elastic springs of equal stiffness. The contact
pressure beneath any small ​area ​is then proportional to the deflection of the spring
SURVEY OF AVAILABLE LITERATURE ​9

in that area and thus to the settlement. The constant of proportionality ​'K' ​is called the modulus of sub-grade
reaction. Although, the theory has been well developed but the value of 'K' for real soils is not constant and
depends not only on the stress deformation characteristics of the soil but also in a complex manner on the shape
and size of the loaded area and the magnitude and position of nearby loaded areas. Evaluation of ​'K' ​for design is
difficult and fraught with uncertainty. Whatever method may be adopted for design, there is no guarantee that the
deflections of the raft will actually be unimportant. In case, the structure covers a fairly large area with
possibilities of differential settlements, it is not enough to provide great strength in the slab. It is also necessary to
provide sufficient stiffness. However, a stiff foundation is likely to be subjected to bending moments far in excess
of those corresponding to the flat slabsubgrade modulus analysis.
There appears to.be no further edition of this book after 1954.

4.2 ​Foundation Design and Practice ​by ​Elwyn. ​E.S. s​ eelye9

According to Seelye after determining the soil pressures at various points of raft, shear and moment diagrams can
be constructed for bands assumed from centre of bay to centre of bay. However, 65% of the moment is assumed
to be resjsted by half the width of the band. There has not been any further edition of this book after 1956.

4.3 ​Foundation Design ​by ​en^'

In the conventional method, it is assumed that the mat is infinitely rigid and that the bearing pressure against
bottom of the mat follows the planner distribution. The mat is analysed as a whole in each of two perpendicular
directions. Thus the total shear forces acting on any section cutting across the entire mat is equal to the arithmetic
sum of all forces and reactions (bearing pressure) to the left orright of the section. The total bending moments
acting on such section is equal to the sum of all moments to the left or right of this section.
Although the total shear and moments can be determined by the principles of simple statics, the distribution along
this section is a problem of highly indeterminate nature, the average moment not being indicative of the sign and
the magnitude of the bending moments in the individual strip in either direction. In order to obtain some idea as to
the upper limit of these values, each strip bounded on central line of the column bays, may be analysed as
independent continuous or combined footings. If the column loads are used, the soil reaction under each strip is
determined without reference to the planner distribution determination for the mat as a whole. This method,
undoubtedly, gives very high stress because it ignores the two way action of the mat. Therefore, certain arbitrary
reduction in values (15% to 33113%) is made.
The author gives other method like Finite Difference Method also for the design of the raft. There has not been
any further edition of this book after 1962. The book, however, has been reprinted in 1992.

The recommendation in this book can be summarised in the following words:

A great refinement of calculations is not always justified or practicable in case of raft.foundations because of the
uncertainties of the action of soil and of short thick members that are arranged in complicated and multiple
systems. It is reasonable to assume that the mat is so stiff and the load so constant that plastic soil will compress
and adjust itself so that each column load will spread almost uniformly under the mat in the general vicinity of
that particular column. For example, the total unit pressure under the rectangular area ​D, ​E, ​F, ​G shown in Fig.
4.1 ​may be assumed equal to 114th of the total loads on the columns at D, E ​ , ​F and G divided by the area of D
​ ,
E, ​F, ​G ​plus the weight of the mat per sq m. For the purpose of computing average pressure
10 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

under the slabs, near the walls, the outer column loads are ​treated ​as though they were concentrated at the
columns. For this method, however, the load on adjacent columns should not differ very much and the bays in
either direction should ​be ​reasonably, equal in length, the larger spacing not exceeding 1.2 ​time, ​the smaller one
and the columns should ​be ​arringed in reasonably straight rows.

Fig. ​4.1 ​Plan of assumed columns strips and distribution of loads

One method of making a preliminary analysis of such a mat is on the basis of an assumed supporting system of
columns strips that constitute a grid of beam along the column rows in each direction. The portion of the slabs in
the central areas is taken up to be supported by this grid. The effective width of these strips or shallow beams
has to be assumed and it is normal to take it slightly more than, what is determined by 45 degrees fiom the
pedestal or column, to the lower reinforcement in themat. Technically the top reinforcement of a central panel
may ​be ​less than of the bottom. However, it may be advisable to reinforce both sides equally because any
yielding of end restraint will increase the, tension in the top of the mat above the computed value. Each column
strip may be analysed by moment dishbution if the variation of loading or spans make this desirable, the entire
thing being designed as an inverted floor. The effect of hydrostatic pressure has to be considered wherever it is
present. There has been no further edition of ​this ​book after ​1962.
 ​ oundation ​
4.5 ​Indian ​Standard ​Code of Practice for ​Design ​and ​Construction of ​Raft F - ​IS ​2950-1965'
There ​are ​two approaches for design-conventional method and the elastic method. In the conventional method,
the foundation is considered infinitely rigid and pressure distribution independent of the deflection of the raft.
Soil pressures are also assumed to be planner so that the centroid of the soil pressure coincides with the line of
action of the resulting forces of all the loads acting on the foundation. The method is normally used in design
because of its simplicity ​. ​A ​generous amount of reinforcement is provided to safeguard uncertainties caused

I​ ​ VAILABLE LITERATURE ​11 ​

SURVEY ​OF A
I​ by differential settlement. The raft is anabjsed as a

whole in each of the two perpendicular directions. Thus, total shear forces and total bending moments acting on
any section cutting across the entire raft is equal to the arithmetic sum of all forces and reactions/moments to the

left or right of the section. The actual reinforcement ​

I​ provided shall be twice that worked out theoretically.

Elastic method has two approaches. In one, the soil is replaced by an infinite number of isolated springs. ​
I​ In

the other, the soil is assumed as a continuous elastic medium obeying Hook's Law. These methods are applicable
in case the foundation is comparatively flexible and the loads tend to concentrate over small areas. The actual
reinforcement can be one-and-a-half times that required theoretically. The famous soil line method falls in this

category. ​! A
​ s limitations to applicability of the methods, code mentions that the coda1 provisions:
(1) do not apply to large and heavy industrial construction where special considerations of the base pressure
distribution will be required. ​i ​(2) apply only to fairly uniform soil conditions and for fairly horizontal planes of

separation of layer
(3) ​foundations in seismic area and/or to vibrating load shall ​be ​given special considerations. ​
below. ​I ​ i

This code has been revised in 1973. Kindly see para 4.7. ​
I​ ​ oundation ​
4.6 ​RafL F
- ​The Soil L​ ine

​ .L.L. ​~akeq ​
Method ​of ​Design b​ y A
! ​According to Mr. Baker, the design of raft as a reversed floor is
 dangerous. Engineers being aware of this, who. ​ therefore, normally adopt the second method in which earth
'​
pressure is assumed to be uniform throughout and moments
​ are obtained at any section by statics. He, however,

feels that in the second method also high values ​' ​of moments are obtained, which may or may not ​be ​present,

and it is irrational or wasteful to provide for such moments

​ without investigating the deflections and variation in
soil pressure. Mr. Baker has, therefore, suggested the soil line method which takes into account the variations in
soil pressure and its relation to deflection but in order to simplify the calculations, it is assumed that the earth
pressure varies throughout a beam according to straight line law.
There is no further edition of this book after 1969.

4.7 ​Indian Standard Code of Practice for Design and Construction of ​Raft
Foundation
1.S ​: ​2950 ​(Part-I) ​1973~

In the revised version of the code, following methods of analysis have been
proposed:
(a) Assumption of linearly varying contact pressure (b) Perfectly
rigid structures (c) Perfectly flexible structures (d)
Structures stiffened along one axis (e) Structures
stiffened along both the axis (​ f) ​General methods:
(i) Based on modulus of subgrade reaction, and (ii) Based on modulus of compressibility (half space theory). Method
(a) corresponds to the conventional method in the earlier version of the code and has similar limitations. In
method (b), contact pressure distribution is to be calculated based on Boussineq's Equation for Elastic
Isotropic half space and is applicable when deformations of raft under loads are small as compared to the
mean settlement of the structure.
12 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

Method (c) is applicable for structures which have relatively less stiffening members specially
resting on very stiff foundation soil. In this case, the deflections of the raft are same as the settlements of
the foundation soil under external load.
Method (d) is something in between methods (b) and (c) ​. ​Here in the direction of the stiffened
axis the contact pressure distribution is determined by Boussineq's Equation as in method (b). In
perpendicular direction distribution is determined ​as ​given in ​(f).
Method (e) is same as method ​(b). ​The two methods under ​(f) ​are elastic methods and are used
when simplified methods from (a) to (e) are not applicable. Details given in the codedo not provide
enough guidance to enable the analysis and design ​10 ​be ​completed by the designer. Apart from the
limitations applicable in earlier version of the code it is stated that:
(i) Allowable settlement both total and differential shall satisfy the requirement of the
super-structure (ii) The approximate values of permissible settlements as given in earlier code have
been deleted. This code has further been revised. Please see para 4.15.

​ dited by Hans ​F. ​Witerkorn ​& ​Hsaiyang

4.8 ​Foundation Engineering Handbook E
an^''
Dr. Joseph ​E. ​Bowles and Wayne ​C. ​Teng are authors of chapters on spread footings, combined and
special footings and mat foundation respectively. Chapter on floating foundation has been written by ​Dr.
H.Q. Golder. This book classifies the method of design of mat foundation according to assumptions
used. The rigid method which is the conventional method assumes that:
(a) Mat is extremely rigid as compared to the sub-soil and, therefore, the flexural deflection of the
mat,
does not alter the contact pressure. (b) The contact pressure or the pile reaction are distributed
in a straight line or a plain surface such that the centroid of the contact pressure coincides with the line of
action of the resultant force of all the loads acting on the mat. When mat foundation is supported on
piles, piles are assumed to ​be ​perfectly elastic. Raft is considered to ​be ​rigid when the column spacing is
less than 1.751h or when the mat is supporting a rigid super-structure. h i s same as defined by Heteny.
The mat is analysed as a whole in each of the two perpehdicular directions. The mat is divided into
perpendicular bands of width between centre lines of adjacent column rows. Each band is assumed to act
as an independent beam subjected to common contact pressure and known column loads. The simplified
elastic method assumes that the soil behaves like an infinite number of individual elastic springs each of
which is not affected by others. This foundation model is also referred to as Winkler foundation.
Analysis procedures have also been developed for the beams on the simplified elastic foundation
concept. The mat is considered as a plate and the effect of each column load is considered in area
surrounding the load. Using the method of super-imposition, effect of all the column loads within the
zone of influence is calculated. Among computer-oriented methods suggested is finite difference method,
based on the assumption that the sub-grade can be substituted by a bed of uniformly distributed elastic
springs with a spring constant equal to coefficient of sub-grade reaction. For this purpose, the mat is
divided into square areas. The deflection at the nodal points of these areas is expressed by a differential
equation in terms of deflection at the adjacent points to the right, left, top and bottom. These
simultaneous equations are solved with an electronic computer and deflection at all the points are
determined. Once deflections are known, the bending moment at any point in each direction is
determined from theory of elasticity.
The finite element method transforms the problem of plates on elastic foundation into a
computer-oriented procedure of matrix structural analysis. The mat is idealised as a mesh of finite
elements inter-connected only
SURVEY ​OF ​AVAILABLE LITERATURE ​13

at the comers and the soil may ​be ​modeled as a set of isolated springs or as an elastic isotropic half space. The
matrix structural analysis can be extended to include the influence of the super-structure as well, thus the
interaction between the super-structure, the foundation and the soil is accounted for.
It is further suggested that in a mat supported on hard rock, the column loads are transmitted to the rock on
relatively small areas directly under the column. ​A ​greater economy may ​be ​achieved by designing the mat by
elastic methods. On very soft soils, the contact pressure against the mat foundation approaches planer distribution
and, therefore, it is commonly justified to design a mat on mud, soft clay, peat, organic soils or even medium clays
by the conventional rigid method. ​A ​generous amount of reinforcement running in both directions at top and
bottom is suggested regardless of method of design used in view of the likelihood that the stresses actually
introduced would bedifferent from those calculated irrespective of the method used foi analysis.
 Second edition of this book is published in ​1991. ​Please see para ​4.21.

4.9 ​Foundation ​Analysis ​and ​Design ​by ​Joseph. ​E. ​~owels' ​"
The mat may be designed as rigid structures thereby soil pressure ​are ​computed as ​Q ​= ​V/A ​in the case where the
resultant of the forces coincide with the centre of the mat area. If resultant has eccentricity with respect to
geometric centre, soil pressure is calculated by the relation

In case, however, if the eccentricity is very large, the resulting internal stresses may be seriously in error. Once the
dimensions of the mat are established, soil pressures at various locations beneath the base may be computed. With
the pressure distribution known, the mat is sub-divided into a series of continuous beams (strips) centred on the
appropriate column lines as shown in Fig. 4​ .2. ​For the series of beams, shear and moment diagram may ​be
established using either combined footinglanalysis or beam moment coefficient. The depth is selected to satisfy
shear stresses and is usually constant but the steel reinforcement vary from strip to strip. The perpendicular
direction is analysed similarly, to complete the design.

Fig. ​4.2
14 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

When the soil bearing pressure is low say 0.5 ​~i~slft2 ​(25 K N I ~ ~ ) or less and if the
deformation of the mat surface can be tolerated, the mat may be designed ​as ​an inverted flat slab, using
heavy beams from column to column. The portion between beams is designed as a conventional one or
two way slabs.
When footings are designed as flexible members, the computation takes some form of the
solution of a beam on an elastic foundation. The experience has indicated that the solution obtained are
generally reliable when the data are satisfactory. Possibly the reasons, as to why the methods have not
been widely used in the past, are ease of making conventional solution, which have been generally
satisfactory and usually not much different from elastic solution. Second reason is that the soil data are
generally obtained using the standard penetration test for which no straight forward conversion to a
value of modulus of sub-grade reaction exists. Various methods for elastic analysis like finite element
and finite differences have also been explained in this book. ​New edition of this book is publisheg in

1988. Kindly see para 4.23

​ 7)18
4~10 ​Building Code Requirements for Reinforced Concrete (ACI ​318 ​- 7

Matters relating to design of footings are included in this code in Chapter 15. paragraph 15.10 relates to
combined footings and mats. This paragraph reads as under: ​15.10.1 Footings supporting more than
one column, pedestal, or wall (combined footings or mats) shall be proportioned to resist the factored
loads and induced reactions, iir accordance with appropriated design requirements of this code. 15.10.2
The Direct Design Method of Chapter 13 shall not be used for design of combined footings and mats.
15.10.3 Distribution of soil pressure under combined footings and mats shall be consistent with
propemees of the soil and the structure and with establishedprinciples ofsoil mechanics.
It would ​be ​seen that this code does not provide for much guidance in design of raft
foundation. This code has been revised several times. Final being in 1989. Please see Para
4.20.

​ onstruction b​ y ​M. J​ . ~
4.11 ​Foundation Design ​and C ​ omlinson'~

Mr. Tomlinson states that it is wrong in principal to assume that araft acts as an inverted floor slab on
unyielding supports and to design the slab on the assumption that its whole area is loaded to the
maximum safe bearing pressure on the soil as this canlead to wasteful and sometimes dangsrous designs.
Allowance must be made for deflection under the most favourable combination of dead and live load
and variation in soil compressibility. Guidance is required from the soil mechanics engineer on the
estimated total and differential settlement for dead and live load considered separately. Some flexibility
is desirable to keep bending moments and shear stresses to a minimum, but the degree of flexibility must
be related to the allowable distortion of the super-structure. Basement rafts carrying heavy building on
weak soils are often founded on piles. The normal function of the piles is to transfer the loading to
stronger and less compressible soil at greater depth or if economically possible, to transfer the load to
bed ​rock or other relatively incompressible strata. The piles also have the effect of stiffening the raft ​and
reducing or eliminating re-consolidation of ground heave, thereby reducing differential settlement or
tilting. In such cases, considerable heave takes place with further upward movement caused by
displacement due to pile driving. After completion of piling, the swelled soil should be trimmed off to
the finished level. The basement walls should generally be designed as self-supportingcantilever
retaining walls even though they may eventually be supported by the floor construction and additional
stability against overturning given by super-structure loading on top of the wall. The basement floor
slabs must be able
SURVEY ​OF ​AVAILABLE LITERATURE ​15

to withstand pressure on the underside of the slab together with stresses caused by differential settlement,

non-uniform column loads, reaction from the retaining walls. If the columns are provided with independent ​
t
bases with only a light slab between them, there would be likelihood of failure of the slab from the pressure of
​
 the underlying soil. ​g ​Fifth edition of this book has been out in 1986. Please see para 4.17.

4.12 ​Design of Combined Footings and Mats ​ACI ​Committee ​33614

The committee observes that no authentic method has been devised that can evaluate all the factors
involved in the problem and allow carrying out determination of contact pressures under combined
footings and mats. Simplifying assumption must, therefore, be made based on the knowledge of the
interaction of the various elements of the system. The following factors should be considered while
examining any problem:
(1) Soil type immediately below the footing ​(2) ​Soil

type at the greater depth ​(3) ​Size of footing ​ (4) Shape

i​
of footing
(5) ​Eccentricity of loading (6) Rigidity of footing (7) Rigidity of the super-structure (8) Modulus of
sub-grade reaction The committee suggests procedure to be followed for design of footings under two
columns: grid foundations and smp footings supporting more than two columns and mat foundation.
Linear soil pressure distribution is suggested for footings which can be considered rigid to the extent that
only very small relative deformations result from the loading. The rigidity may result from the spacing of
the columns on the footing from the rigidity of the footing itself or the rigidity of the super-structure.
Limitations which must be fulfilled to make this assumption valid have been discussed in the report.
Distribution of soil pressure by means of sub-grade reaction has been suggested where sub-soils are of such
character that the deformations are localised in the general vicinity of the loads and when the maximum contact

pressure is smaller than about one and a half times the ultimate bearing capacity. In case of rigid footings, it ​ is
a​

suggested that uniform or linear distribution of soil pressure can be assumed and the design based on statics.
Flexible footing procedure is divided into 2 parts i.e. uniform condition and general condition. Uniform
conditions are considered to be those where the variation in adjacent column loads and spans is not greater than

20%. For cases where supporting columns are at random location with varying intensities of loads a ​1 ​detailed

design procedure based on plate theories has been recommended.

​ avis ​1980~~
​ oundation Analysis ​and ​Design ​by H.G.Poulos and ​E.H. D
4.13 ​Pile F

: ​In this book, Chapter 10 deals with piled raft systems. The author says that, "in design of foundation for a
large building on a deep deposit of clay it may be found that a raft foundation would have an adequate factor of
safety against ultimate bearing capacity failure but the settlement would be excessive; traditional practice would
then be, to pile the foundation and to choose the number of piles to give an adequate factor of safety assuming
the piles take all the load; however it is clearly illogical to design the piles on an ultimate load basis when they
have only been introduced in order to reduce the settlement on other-wise satisfactory raft." According to the
author, once the have been introduced solely for the purpose of reducing the settlement
16 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS
design question becomes not "how many piles are required to carry the weight of the structure" but "how many
piles are required to reduce the settlement to an acceptance level".
However, in Chapter ​5, ​the settlement behaviour of a free standing pile is obtained from the elastic-based
analysis. The pile is divided into number of elements and the expressions for vertical settlement of the pile and
the soil at each element in terms of unknown stresses on the piles are obtained and solved, imposing the vertical
displacement compatibility condition, to arrive at the settlement behaviour of the pile. As a further extension, the
unit consisting of a single pile with an attached cap resting on the soil surface is considered. It is assumed that
purely elastic condition prevails upto the load at which the pile would fail if no cap were present and thereafter
any additional load is taken entirely by ​the ​cap. The book gives charts indicating interaction factor between the
raft and the pile for various values of length of the piles, diameter of the pile, poisson ratio of soil, height of soil
layer over the rigid stratum and the cap diameter.The method is further extended to group of piles upto about ​40
numbers. Curves are drawn which ​are a​ pplicable only for rigid rafts or perfectly flexible rafts. The entire
emphasis is to work out the ratio of the load carried out by the piles and the raft soil system. No details ​are ​given
on &e method to determine the bending moment and shear forces in ​the ​raft. It is only mentioned that none of the
simple methods are satisfactory and a proper analysis of plate on piles and continuum is desirable.

​ harles ​E. ​Reynolds and James C. Steedman ​

4.14 ​Reinforced Concrete Designers Handbook ​by C -
9th Edition 1981"

This book suggests the analysis of a raft foundation supporting a series of symmetrically arranged equal loads on
the assumption of uniformly distributed pressure on the ground considering the structure ​as ​an inverted reinforced
concrete floor acted upon by the load of earth pressure from bottom. It is further suggested that when the columns
on the raft a​ re n​ ot equally loaded or are not symmetrically arranged, the raft should be so designed that the
centroid coincides with the centre of gravity of the loads. If this coincidence of centre of gravity is impracticable
owing to the extent of the raft being limited on one or more sides, the plan of the raft should be made so that the
eccentricity of the total loading is a minimum, though this may produce a raft which is not rectangular in plan.

-​
4.15 IS 2950 (Part ​I) ​1981 ​ Code for Design and Construction of Raft Foundation Part I ~ e s i ~ n ~

In the second revision of the code, two methods of analysis have been suggested depending upon the assumption
involved. Conventional method assuming planner distribution of contact pressure is applicable to foundations
which are rigid relative to supporting soil and the compressible soil layer is relatively shallow. The rigidity of the
foundation is determined with a relative stiffness factor K ​ ​> ​0.5 or columns spacing less than 1.75A. Methods of
determining value of K ​ ​and ​hare ​given in the code. Conventional method is applicable when either of the two
conditions ​are s​ atisfied. The value of ​K ​depends upon the flexural rigidity of the super-structure, modulus of the
compressibility of the foundation soil, thickness of the raft, length of the section in the bending axis and length
perpendicular to the section under investigation. Value of ​h d​ epends upon modulus of sub-grade reaction for the
footing of the width of the raft, modulus of elasticity of concrete and moment of inertia of the raft. In this method,
the ​raft ​is analysed as a whole in each of the two perpendicular directions on the basis of statics.
In case of flexible footings, simplified methods ​are ​applicable when variation in adjacent column load is not
more than 20% of the higher value and the structure (combined action of the super-structure and raft) may be
considered as flexible, ​ie., ​relative stiffness factor ​K ​is greater than 0.5. ​In ​this method, it is assumed that
SURVEY ​OF ​AVAILABLE I-ITERATURE ​17
the sub-grade conslsts of an infinite array of individual elastic springs each of which is not affected by others.
This method is more or less same as the famous soil line method.
When conditions, as mentioned above, for flexible foundations are not satisfied ​, ​a method based on closed form

of solution of elastic plate theory has been suggested. The distribution of deflection and contact pressure ​
​
& on
the raft due to a column load is determined by the plate theory. Since the effect of a column load on the ​ elastic
'​
foundation is damped out rapidly. It is possible to determine the total effect at a point of all column loads
​ with~n
the zone of influence by the method of super-imposition. The computation of the effect at any point is restricted

to columns of two adjoining bays in all directions. ​i :​ T

​ he code also lays down that:
(a) Size and shape of the foundation adopted affects the magnitude of subgrade modulus which should
be taken into consideration. (b) Consideration must be given to the increased contact pressure developed along
the edges of the raft
on cohesive soils and the opposite effect on granular soils. (c) Expansion joint should be provided when the
structure supported by the raft consists of several parts
with varying heights and loads or there is a change in the direction of the raft. (d) This code does not explicitly
provide ​any ​guidance a​ s ​to how factors emphasised in (a) and (b) above should be allowed for. The second ​part
of the code relating to construction aspect is still not printed. There has not been any further revision and this
code was reaffirmed in 1987.
4.16 ​Eleventh Intenationul Conference of Soil Mechanics ​a d ​Foundation Engineering ​San Francisco,

-​
August 12 ​ 16,1985~~
In the conference while two papers were presented on instrumentation of pile raft foundation and cap pile soil
interaction, there was no recommendation or paper on design of raft foundation.
4.17 ​Foundation Design and Construction b ​ y M.J. ​Tomiinson, 5th Edition, 1986"
There is no significant change in this edition from what was recommended in 4th edition

-​ -​
4.18 ​Handbook of Concrete Engineering ​ Mark ​Fintei ​ 2nd Edition, 1986%
This book makes no recommendation about raft foundation.
4.19 ​Reinforced Concrete Designer Handbook ​by Charles E. Reynolds and James Steedman,
10th Edition, 1988~'
There is ​no ​change in recommendations from what was done in the earlier edition published in 1981 ​4.20

Building Code Requirements ​in ​Reinforced Concrete ​ - ​ACI ​- ​318 ​- ​1989~'

Building code requirements since their second edition ​in ​1977 have gone in for further revision 1983, 1989 and
1992. ​In the latest revision there is no change in the code requirements for design of combir.ed footings and mats,
but in commentary a reference has been made to 'design procedure for combined footings and mat i​ s ​per report
prepared by ACI committee 336'and also to a paper 'simplified design of footings by' Kramrisch, Fritz and Rpgers
Paul published in American Society of Civil Engineers Proceeding, ​V. ​87, ​NOSM ​5, ​October 1961, p. 19.
​ AFT FOUNDKTIONS-DESIGN AND ANALYSIS
18 R

​ y Hsai-Yang-Fang 2nd Edition, 1 9 9 1 ~ ~

4.21 ​Foundation Engineering Handbook b
This edition has omitted the chapter on mat foundation which was originally'included in first edition. ​4.22 Design

of Combined Footings and Mats ​ - ​ACI committee 336 2R ​- ​88 ​Published in ACI
Manual 1 ~ 3 ~ ~

1966 report mentioned in para ​4.12 ​above was reaffirmed in 1980 but has been completely revised and elaborated
in 1988. This report suggests that:
(a) ​Maximum unfactored design contact pressure should not exceed the available soil pressure deter- mined by
geotechnical engineer. Where wind or earthquake forces form a part of the load combination, the allowable soil
pressure may ​be ​increased as allowed by the local code and in consultation with geo-technical engineer. (b)
Combined footings and mats are sensitive to time dependent sub surface response. Many structural engineers
analyse and design mat foundations by computer using the finite element method. Soil response can be estimated
by modelling with coupled or uncoupled "Soil springs". The spring properties are usually calculated using a
modulus of subgrade reaction, adjusted for footing size, tributary area to the node, effective depth, and change of
modulus with depth. The use of uncoupled springs in the model is a simplified approximation. The time dependent
characteristics of the soil response, consolidation settlement or partial consolidation settlement, often can
significantly influence the subgrade reaction values. Thus, the use of a single constant modulus of subgrade
reaction can lead to misleading results. (c) Caution should be exercised when using finite element analysis for
soils. Without good empirical results, soil springs derived form values of subgrade reaction may only ​be ​a rough
approximation of the actual response of soils. Some designers perform several finite element analyses with soil
springs calculated from a range of subgrade moduli to obtain an adequate design. (d) The response of a footing is a
complex interaction of the footing itself, the superstructure above, and the soil. That interaction may continue for a
long time until final equilibrium is established between the superimpos&l loads and the supporting soil reactions.
Moments, shears, and deflections can only be computed if these soil reactions can ​be ​determined. (e) No analytical
method has been devised that can evaluate all of the various factors involved in the problem of soil-structure
interaction and allow the accurate determination of the contact pressures and associated subgrade response. ​(f) ​For
mat foundations modulus of subgrade reaction cannot be reliably estimated on the basis of field
plate load tests because the scale effects are too severe. ​ (g) Mats may ​be ​designed and analysed as either rigid
I​

bodies or as flexible plates supported by elastic ​ I​ foundation. A combination analysis is common in current

​ as plate on an elastic foundation can ​be ​made. However a number of

practice. An exact theoretical design of ​\ mat

factors like, difficulty in ​1 ​I p​ rojecting subgrade responses, variation in soil properties both horizontal and

vertical, mat shape, ​ ; ​variety of superstructure loads and assumption in their development and effect of
*​

superstructure stiffness
​ on mat rapidly reduce exactness to a combination of approximations. The design is further
affected by excavation heave. (h) After propottioning the mat size, compute the minimum mat thickness based on
punching shear at critical columns based on column load and shear perimeter. It is common practice not to use
shear reinforcement so that mat depth is maximum.
i​ SURVEY OF ​AVAILABLE LITERATURE ​19 ​
(i) In case column spacing is less than 1.75 divided by ​h ​or

the mat is very thick and variation of column loads and spacing is not over 2096, mat may be designed by treating
it a​ s ​a rigid body and considering ​I ​strips both ways. These strips are analysed ​as ​combined footings with multiple
column loads and loaded with the soil pressure on the strip and column reactions equal to loads obtained from the
superstructure analysis. Since a mat transfers load honzontally, any given strip may not ​satisfy ​vertical load
summation. ​Q) ​In case the criteria is not met with an approximate analysis can be made using the method
suggested
by ACI Committee 336 in 1966. (k) Computer aided finite differences, finite grid or finite element methods can
be used where computers are available. The report gives details of these ​3 ​methods. ​In ​any of these ​3 ​methods
node pressure should not exceed the safe bearing pressure value recommended by the geotechnical engineer. (1)
A mat analysis is only as good as the soil parameters. Since it is very difficult for the geotechnlcal engineer to
provide accurate vdues of moGulus of subgrade reaction, the structural designer may do the parametric study,
varying the value of ​K ​over range of one half the furnished value to ​5 ​or 10 times the furnished value. (m) The
analysis and design of combined footings and mats is a soil-structure interaction effort in which there is no unique
method to determine mat deflection. The determination of mat deflection extends far beyond the analysis of a
beam or finite element model to the prediction of subgrade response. The prediction of subgrade response, though
part of the structural analysis of the mat, is more elusive than designers wish to admit. Experience with extensive
measurements of both foundation loadings and subgrade response are needed to develop a high degree of
confidence in the method selected. A very close working relationship must exist between the geotechnical and
structural engineers to properly analyse comb~ned footings and mats.

4.23 ​Foundation Analysis ​and D

​ esign ​by Bowles, ​4th ​Edition, 1 9 8 8 ~ ~

In this edition analysis of mat foundation has further been elaborated considerably. Among the design methods
included are conventional or rigid methods as explained in earlier edition stating that this method is not
recommended at present because of substantial amount of approximations and the wide availability of computer
programmes which are relatively easy to use and mat being generally too expensive and important not to use most
refined analytical method available.
The approximate flexible procedure suggested by ACI Committee ​436 ​(1966) has been retained and elaborated.
Further details have been given for finite difference method, finite element method and finite grid method
applicable with computer.

4.24 ​Proceedings of ​Indian ​Geo-Technical Conference 1992, Calcutta, December, 1 9 9 2 ~ ~

This conference does not have papers relating to design and analysis of raft foundation.

-​
4.25 ​Designs of Foundation Systems ​ Principles a​ nd ​Prrictices ​by Nainan P. Kurian, 1 9 9 2 ~ ~

The book details conventional approach to raft design ​as ​a flat slab and beam and slab raft, following the Indian
Standard Code of Practice, more on the inverted floor approach. The book only mentions that an integrated
analysis of the beam and slab on the computer by the finite element method using package programmes such as
SAP ​IV ​which will give exact results based on the actual behaviour of the system can be carried out. This book
also mentions about the design of raft foundation by the Soil line method stating that this method has

20 ​RAW FOilNDATlONSDESlGN AND ANALYSIS ​ 1 ​I rather

​ become obsolete in the wake of possibility of
using more refined flexible methods with the aid of computer.
4.26 ​13th International Conference on Soil Mechanics and Foundation Engineering, ​New ​Delhi

I​
January, 1 9 9 4 ~ ~ ​ I ​A ​paper by M.F. Randolph was presented a​ s ​a special lecture on design methods for

Pile Groups and Piled ​i ​Rafts.

The paper recalls that in majority of the cases where piles form part of the foundation for a building or other
structures, the primary reason for inclusion of ​the ​piles is to reduce settlements. However, once the decision has
been made that piles are required t​ he t​ raditional design approach has been to ensure that the total structural load
can be carried out by the piles, with adequate factor of safety against bearing failure. However, there is elastic

interaction'between the raft and soil below, between piles and piles ​as ​the performance of a pile within ​i a​ group

is affected by the presence of other piles. The key question that arises in the design of pile rafts concerns ​I ​the

,​
relative proportion of load carried out by raft and the piles and the effect of additional pile support on ​ absolute

and differential settlements. ,The paper suggests that this distribution of load between the raft and piles ​ 1 b​ e
taken into account. The paper also gives methods by which this proportion of load between the two components
​

are carried out. ​

​ ​Soil Structure Inter-action ​- ​The Real Behaviour of Structures, published by the
I 4.27
Institution of
Structure Engineers, U.K. The Institution of ​Civil ​Engineers, U.K. International Association for Bridge
and Structural Engineering ​in ​March, 1 9 ~ 9 ~ ~
The above institutions constituted a joint committee under ​Dr. ​Sam Thornborn which prepared this report.
Pointing out that,
(i) Red behaviour of structures in contact with ground involves an inter-active process beginning with the
construction phase and ending with a state of balance after a period of adjustment of stresses and strains within
the structure and within the ground influenced by the structure. (ii) Actual behaviour of the structure relates to
the inherent spatial variations in the ground and it should be appreciated that these variations are not always
readily identifiable by occasional and local boring, sampling and testing. The report deals with the question of
soil structure interaction in ​2 ​parts. Pari ​I ​relates to structures supported by ground and Part I1 for ground
supported by structures.
(a) Under structures supported by ground, the report points out that engineers could estimate the settlements for
a perfectly flexible load or they could estimate the avenge settlement of a rigid load but in between these limits,
the engineers could say nothing. (b) Analytical methods have been developing so rapidly over the last few years
that it is now possible to obtain solution to many complex problems which a few years ago would have been
quite out of reach. If used sensibly and with discernment, these powerful analytical methods can be of consid-
erable assistance enabling a designer to gain a feel for the behaviour of soil structure system. However, if used
blindly, such methods cause menace and can be extremely misleading. The key to successful use is to gain a
clear understanding of the idealisations that are being made and to be aware of, how far they may be, from
reality. (c) For a framed building founded on a raft, during excavation some heave of the soil will occur. The raft
will then be constructed and will be influenced by the differential settlement there after. As the
SURVEY ​OF ​AVAILABLE LITERATURE ​i
distorts
structural load is applied short term settlements take place, the part of the structure in existence ​t ​
and the overall stiffness gradually increases. The cladding is then added and may substan- ​
tially increase the
stiffness of the building. Finally, the imposed load is applied. Not all the components of the buildings are subject
to the same relative deflection. The relative deflections experienced by the raft ​will ​be the largest. Those
experienced by the structural members will vary with location and elevation in the building. The likelihood of
damage will diminish, the larger the proportion of medium and long-term settlements, the smaller the ratio of
imposedldead loads and later the stage at which the finishes are applied. (d) The report has an appendix which has
reviewed currently available techniques for the analysis of the total soil structure system. More readily available
computer packages that utilise these techni- ques, have been listed in the appendix. (e) The manner in which and
the limitations with which super-structure can be modelled have been singled out. For soil model, it is pointed out
that commonly known approach of treating the soil as a set of liner unconnected springs cannot be recommended
for the analysis of rafts and continuous footings although this model has the advantage of being easily included in
standard computer programmes for structural analysis. It is a poor physical model. The results of analysis based on
use of this model may be excessively sensitive to the pattern of applied load. (​ f) ​The half space continuum using
elastic theory for both stresses and strains has severe limitations because it does not take into account, the soil
layering or the variation of soil modulus with depth within a given layer. In an extension of this method where
elastic theory is used for strains only and then stresses are calculated using the various deformation moduli of the
soil is better approximation. In a further improvement of a layered coniinuum the exact stresses and strains in a
layered soil mass are calculated. (g) Super structure stiffness has a marked influence on the behaviour of the raft
and should not be ignored although the quantitative assessment of all but the simplest of the wall system
connected to the raft may prove difficult. However, often the raft is itself a major contributor to the overall
stiffness of the building. Since the raft is in intimate contact with the supporting soil, the inter-active effects are
perhaps most marked in consideration of its own behaviour. In the design of raft foundation, it is totally unrealistic
to ignore deformation and rely on moment and shears obtained from the analysis of the conventional flat slab
method. It is equally unrealistic to compute deformation without consideration of the structural stiffness and then
to design on the basis of the corresponding stress resultants. Rational design approach must be based on the results
of an interactive analysis.
 DESIGN APPROACH AND
CONSIDERATIONS

Summary of methods suggested by various authors discussed in Chapter ​4 ​would indicate that basically two
approaches have been suggested for analysing the behaviour of raft foundation:
A. Rigid foundation approach ​B. ​Flexible foundation approach

5.1 ​Rigid ​Approach

In rigid foundation approach, it is presumed that raft is rigid enough to bridge over non-uniformities of soil
structure. Pressure distribution is considered to be either uniform or varying linearly. Design of rigid raft follows
convkntional methods where again following two approaches have been suggested:
(a) Inverted floor system (b) Combined footing approach In rigid rafts, differential settlements are comparatively
low but bending moment and shear forces to which raft is subjected are considerably high.

5.2 ​Flexible ​Approach

In ​flexible foundation approach, raft is considered to distribute load in the area immediately surrounding the
column depending upon the soil characteristics. In this approach differential settlements are comparatively larger
but bending moments and shear forces to which the raft is subjected are comparatively low. Analysis is suggested
basically on two theories
(a) Flexible plate supported on elastic foundation, ​i.e., H
​ etenyi's Theory (b) Foundation supported on bed of
uniformly distributed elastic springs with a spring constant determined using coefficient of sub-grade reaction.
Each spring is presumed to behave inde- pendently, ​i.e., W ​ inklers's foundation Based on these two basic
approaches, methods suggested include simplified methods subject to certain limitations which can be carried out
by manual computation. Also now available are computer based methods
DESIGN APPROACH AND CONSIDERATIONS ​23

like finite element and finite differences methods. Finite differences method is based on the second approach uf
uniformly distributed elastic springs and can consider one value of sub-grade modulus for the entire area.
Finite element method transforms the problem of plates on elastic foundation into a computer oriented
method of matrix structural analysis. In this method, plate is idealised as a mesh of finite elements
inter-connected only at the nodes (corners), and the soil may be modelled as a set of isolated springs or as
 an elastic isotropic half space. The matrix structural analysis can be extended to include the influence of the
super-structure ​as ​well. Thus, the interaction between the super-structure, the foundation and the soil can be
accounted for. It is possible to consider different values of sub-grade modulus in different areas of the raft
foundation.
In case of piled rafts against the usual assumption of entire load being carried by piles alone, emphasis is now
being laid on sharing of load between raft supported on soil, ​i.e., r​ aft soil system and raft pile system.
Sufficiently accurate methods for practical distribution of these loads are not yet available.
As a simplification of treating the entire raft ​as ​a plate, concept of beam on elastic foundation is also being used. For this

purpose raft is considered to consist of beams in both the directions. Each of these beams is ​
-, t​ reated as
supported on springs having spring constant calculated using modulus of subgrade reaction and carrying
​ column
loads. The beam is then analysed as a bean1 on elastic foundation.

5.3 ​Parameters ​for Raft ​Design

In all these methods, however, three basic parameters, ​i.e., r​ igidity of the raft, pressure distribution under
the raft and value of sub-grade modulus become important in addition to whatever other info&ation'is
received from soil investigation report. These three parameters and method of their determination are
discussed in subsequent paragraphs.

A ​problem which has to be solved while designing a raft foundation is to evaluate the actual contact
pressure of the soil against the raft. This problem has occupied many researchers theoretically and a lesser
number experimentally with no exact values being known. Contact pressure, settlement of foundation, soil
charac- teristics and its behaviour are so much inter-related and their relationship so complex, that soil
foundation -​ ​structure interaction is not clear even now. Considering all these aspects it can be said that the
contact pressure distribution under the raft depends upon:
(1) ​The nature of the soil below the raft, ​i.e., a​ single homogenous mass or a layered formation,
thicknesses of various layers and their relative locations ​(2) ​Properties of the soil (3) The nature of ​the

foundation, ​i.e., ​whether rigid, flexible or soft ​(4) ​Rigidity of the super-structure (5)
​ ​The quantum of
​ resence of adjoining foundation ​(7) ​Size of raft ​(8) ​Time at
loads and their relative magnitude ​(6) P
which pressure measurements are taken The total settlement under the raft foundation can be considered
to be made up of three components, ​i.e.,

S ​= ​Sd+Sc+Ss w
​ here ​Sd ​is the immediate or distortion settlement, ​Sc ​the
consolidation settlement and ​Ss ​is the secondary compression settlement. The immediate component is that
portion of the settlement which occurs simul-
24 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

taneously with the load application, primarily as aresultof distortion within the foundation soils.
Thesettlement is generally not elastic although it is calculated using elastic theory. The remaining
components result from the gradual expulsion of water from the void and corresponding compression of
the soil skeleton. The distinction between the consolidation and secondary compression settlement is
made on the basis of physical process which control the time rate of settlement. Consolidation settlements
are largely due to primary consolidation in which the time rate of settlement is controlled by the rate at
which water can be expelled horn the void spaces in the soil. The secondary compression settlement, the
speed of settlement is controlled largely by the rate at which the soil skeleton itself yields and compresses.
The time rate and the relative magnitude of the ​3 ​components differ for different soil types. Water flows
so readily through most clean granular soil that the expulsion of water from the pores for all practical
purposes is instantaneous and thus ​foundation ​settles almost simultaneously with the application of load.
In cohesive soil, it takes considerable time for water to escape and thus settlement in cohesive soils
continue much longer. In fact, it has been reported that the pressure under a mat foundation on clay may
vary from time to time.
It is usual to assume that the soil below the foundation is an isotropic homogeneous material
for its entire depth. But normally this is not the situation and we get different layers in varying thickness,
having different properties below foundation. If the thickness of the upper most layer is large relative to
the dimension of the loaded area, it would probably be sufficient if the soils were considered as a
homogenous layer of indefinite depth. However, if the upper stratum is relatively thin ignoring theeffect of
layering, it may have an appreciable influence on the contact pressure distribution and consequently
settlements. This is likely to be of special importance when a compressive stratum is underlain by rock or
a very hard or dense soil. Such presence decreases the settlement considerably. It is very significant when
this occurs within a depth equal to width of the footings. Incase, there is a stiff stratum underlain by a soft
stratum like layer of sand over soft clay layer, effect is negligible if depth is greater or equal to ​3.5 ​b2.1n
case of raft, dimensions of raft are generally such that the possibilities of encountering a different soil
layer within the significant depth are quite large and as such it would be necessary to account for the
different soil layers within the significant depth. Moreover it is to ​be ​remembered that properties of soil
constituting each layer which determine the shear strength charac- teristics and settlement characteristics
of the soil become more important as rafts are generally adopted in areas where soils of poorer types
are'~ncountered and which some years ago might have not been taken up for construction at all.
Effect of groundwater table is appreciable on the load carrying capacity of the soil and
consequently settlements. It is, therefore, necessary to consider the expected ground water table in life
time of the structure including the temporary rises as during floods. Even in areas where sub-soil water
table is not present, it is necessary to consider long term built up water for design of basement and raft
foundation. If permeability coefficient of the soil is below ​0.1 ​mm per second, soil is cohesive and
probability of surface water accumulated against basement walls exist'. In such situations, it may be
necessary to design raft foundations of basement for water uplift also.
The conventional analysis of footings, in general, uses the concept of a rigid fcotings and with
rigid footing ​are a​ ssociated the concept of uniform soil pressure. Actually to have a uniform soil pressure
distribution, we require a very flexible footing. If simultaneously we accept the concept of soil being
elastic (modulus of elasticity or coefficient of sub-grade modulus), settlement of rigid footing will be
uniform and that for a flexible footing the settlement would be non-uniform and but if this is the case then
how c​ an ​the contact pressure be uniform (under a rigid footing). In reality we have a soil snucture
interaction problem and there is a non-uniform soil pressure and differential settlements within the
footings. It has been suggested that in case of square footing resting on clay on average contact pressure of
0.6 ​PIA with additional ​0.1 ​PIAalong edges would be reasonable
.D ​ ESIGN APPROACH AND CONSIDERATIONS ​25
 pressure distribution. For a rectangular footing of large length it is suggested that it would be reasonable to have
​ .1 ​PIB ​for the edges. Here ​P i​ s total load, ​A, a​ rea and ​B, l​ ength of
an average pressure equal to 0.8 ​P a​ verage ​+ 0

the footing. ​
1​ For footings on sands a pressure distribution of uniform soil pressure is reasonable. ​ I

Rigidity of foundation gets modified by the rigidity of super-structure. Arigid super-structure will not allow ​
1
differential settlement to take place in foundation. Situation can arise when a particular column of the building ​
:
may ​be ​hanging from the super-structure and even transmitting the weight of attached soil mass to the super
structure rather than transmitting any load from the super-structure to the foundation soil. In fact, a rigid
foundation with a rigid super structure means less differential settlement, large variation of contact pressure ​i- ​i

and high bending and shear stress in foundation members. ​A ​flexible foundation with flexible super structure :​
means large differential settlements, uniform contact pressure and lower values of bending and shear stresses
in foundation members.
Quantum of loads and their relative magnitude affect the contact pressure. When the loads are so high that bearing
pressures are increased to the point of shear failure in the soil, the contact pressure is changed leading to an
increase in pressure over the centre of the loaded area in all cases.
and
The consolidation pressure involves expulsion of water from the soil being compressed. This takes time ​1 ​
at any time between the application of the load producing consolidation and the time at which essentially ​
:
ultimate or 100 per cent consolidation has occurred, the measured settlements and consequently contact
pressure distribution would ​be ​different. Many times it may take several years to achieve final
settlement.
There are situations in engineering practice where footings are placed so close to each other that their zones of
influence overlap. Studies have shown that effect of adjacent footings may vary considerably with angle of
shearing resistance. For low values they are negligible. For higher values they appear to be significant
particularly if footing is surrounded by others on all sides. There are practically no effects in case of
punching shear failure. It is generally recommended that interference effect may be neglected. ​ ,

In view of various factors affecting the pressure distribution under a raft foundation and difficulties in determining
affect of each, it is generally believed that contact pressure distribution under a raft could be of the

following type as shown in Fig. 5.1. ​(c) S

​ OFT SOIL ​-------
Fig. 5.1 Contact pressure distribution under a
raft
26 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

Fig. ​5.1 ​(a) is applicable when the mat is supported on hard rock and column loads are transmitted to the rock on
areas of relatively small size directly under the columns. If the raft rests on a stiff dense soil, then loads are
distributed to the sub-soil in relatively large areas, ​as ​shown in Fig. ​5.1 ​(b). It is only on very soft soils that the
contact pressure against the mat foundation approaches linear distribution as shown in Fig. ​5.1 ​(c). Therefore, it is
commonly justified to design a mat on mud, soft clay, peet or organic soil by the conventional rigid method using
uniform pressure. In fact assumption of rigid footings with uniform soil pressure results in designing the raft for
assumed bending moments which are larger than the actual bending moments. The resulting design is conservative
generally but may not be economical. A greater economy can, perhaps, be achieved by designing the mat with
elastic methods, but at what risk and is it really so ?​ ​Actual pressure distribution under the raft, therefore, remains
unanswered.

5.5 ​Rigidity Criteria

Whether a structure behaves as rigid or flexible, it depends on the relative stiffness of the structure and the
foundation soil.The behaviour of the foundation as rigid or flexible will also depend upon the rigidity of the
super-structure above and properties of soil below. In physical terms, a rigid foundation would mean a foundation
which is capable of bridging over pockets of soil with different properties and thus ​try ​to even out the settlements
at various points. A rigid foundation would, therefore, have comparatively lower values of differential settlement
but higher values of stresses. A rigid foundation with a rigid super-structure on a comparatively compressible soil
will result in uniform settlements of structure.
A flexible foundation with a flexible super-structures and a comparatively rigid soil below will behave as a flexible
foundation and would result in large differential settlements and low stresses. Thus:
(i) Arigid member is characterised by high bending moments and relatively small, uniform deflections. Over all

differential settlements are small. ​

- ​(ii) An intermediate member, ​as ​the term implies, has intermediate bending
and deflection values. (iii)
​ The flexible member has comparatively smaller bending moments and deflection is
maximum in vicinity of the loads and small values else where. Overall differential settlement would be of higher
orders. Rigidity criteria proposed by various authorities are discussed below:

5.5.1 Proposed b ​ y ​IS :​ ​2950 (Part I) 19813 Appendix C

​ ​of this standard gives the method of deciding rigidity of
super-structure and foundation. This is reproduced below:

Rigidity of Superstructure and Foundation

​ etermination of the Rigidity of the Structure ​C-1.1 ​Theflexural rigidity E

C-1 D ​ l ​of the structure of any section
may b​ e estimatcdaccotding to the relation given below (see also Fig. 5​ .2):
DESIGN APPROACH AND
CONSIDERATIONS
 Fig. ​5.2 ​Determination of rigidity of a
structure

​ ​modulus
where ​El = of elasticity of the infilling material (wall material) in kg/crn2,
​ ​Moment of,inertia of the infilling in cm4, ​b ​= ​length or breadth of the
I, =

s ructure in the direction of bending. ​

J H​ =​ ​total~height of the infill In
cm, E,
​ =​ modulus of elasticity of frame material in kg/cm2 ​Ib =
​ ​moment of
inertia of the beam in cm4

where ​/. ​1 =
​ ​Spacing of columns in cm,
h, ​= ​Length of upper column in
cm, ​hl ​= ​Length of lower column
in cm,

​ ​Moment of inertia of upper column in cm ​, Il

4 ​I,,
= ​ =​
Moment of inertia of lower column in cm4 ​If ​= ​hioment
of inertia of foundation beam or raft in cm4,
28 ​RAFT ​FOUNDATIONS-DESIGN AND ANALYSIS

Note ​: ​The summation is to ​be ​done over all the storeys, including the foundation beam of ​raft. In ​the
case of the' foundation I;replaces ​Pb a​ nd ​1, b​ ecomes zero, whereas for the topmost ​beam ​1'" ​become
zero

C-2 Relative Sti#hess Factor ​K: ​C-2.1 Whether a structure behave a ​ s ​rigid orflexible depends on the
relative s t i m s s ofthe structure and the foundation soil. This relation i​ s e​ xpressed by the relative
stimess factor ​K ​given below:
​ or the whole structure
(a) F

​ or rectangular rafts or beams

(b) F

(c) ​For circular rafts

where ​El ​= ​Flexible rigidity of the structure over the length

(a) in kg/cm2 ​E, = ​ ​Modulus of compressibility of the foundation
soil in kg/cm2 ​b =​ ​Length of the section in the bending axis in
​ ​Length perpendicular to the section under investigation
cm, ​a =
in cm, ​d ​= ​Thickness of the raft or beam in cm, ​R = ​ ​Radius of
the raft in cm ​C-2.1.1 For K ​ >​ ​0.5, t​ he foundation may be
considered as rigid

C-3 Determination of Critical Column Spacing C-3.1

Evaluation of the characteristics ​h i​ s m
​ ade ​as ​follows:

where ​k = ​Modulus of sub-grade reaction in kg/cm3 for footing of width

B ​in cm ​B ​= ​Width of raft in cm, ​E, =
​ ​Modulus of elasticity of concrete
in kgf/cm2 ​1 ​=Moment of inertia of the raft in cm4
Modulus of compressibility of the soil is the additional property required in this particular
case.

5.5.2 A
​ CI Committee, 4 ​ uggested design procedure for combined footings and mats ​- ​American
​ 36 S
Concrete Institute Journal, ​October, 196614
Relevant extracts from this paper are given
below:
 I​ DESIGN APPROACH AND CONSIDERATIONS ​
29 ​
a ​1 ​
Footings supportingjield structures Continuous
​

strip footings supporting structures which because of their rigidity will not allow the individual ​
I​ columns to

This
settle differentially should ​be ​designed ​as ​rigid footings with a linear distribution of soil pressure. ​g ​
distribution can be determined on the basis of simple statics.

To determine the approximate rigidity of the structure, an analysis must be made comparing the combined
stiffness of the footings, super-structure framing members, and shear walls with the stiffness of the soil. The
relative stiffness will determine whether the footing should be considered rigid or flexible.
The following formulas may ​be ​used in this analysis :​
where ​E ​= ​Modulus of elasticity of the materials used in the structure, kips per sq.ft (metric tons per sq.m)
I, =Moment of inertia of the structure per unit length, ft3 (m3)
IF ​= ​Moment of inertia of the footing per unit length, ft3 (m3) ​Es= ​Modulus of elasticity of the soil, kips per
sq.ft (metric tons per sq.m)
b ​=Width of footings, ft (m)
An approximate value of ​ElIC ​per unit length of building can bedetermined by summing the flexural rigidity of
the footing ​(E'L,) ​the flexural rigidity of the each framed member (FIB) and the flexural rigidity of any shear
walls (F3112) where a and h ​ a​ re the thickness and height of the wall, respectively.
Computations indicates that as the relative stiffness K ​ , ​increases, the differential settlement decreases rapidly.
For ​K, = ​ ​0 ​, ​the ratio of differential to total settlement is 0.5 for long footing and 0.35 for a square one. For ​K, ​=
0.5 ​, ​the ratio of differential to total settlement is about 0.1. If the analysis of the relative stiffness of the footing
yields a value above 0.5, the footing can be considered rigid and the variation of soil pressure determined on the
basis of simple statistics.
If the relative stiffness factor is found to b​ e ​less than 0.5, the footing shall b​ e ​designed as a flexible member using
the foundation modulus approach ​as d​ escribed under section ​6.4 ​of the report.
Columns Spacing
The column spacing on continuous footings is important in determining the variation in soil pressure distribution.
If the average of two adjacent spans in a continuous strip having adjacent loads and column spacings that vary by
not more than 20 per cent of the greater value or is less than 1.75/h, the footing can be considered rigid and the

variation of soil pressure determined on the basis of simple statics. ​- ​If the average of two adjacent span, as

​ ​governed by subgrade modulus theories.

limited above, is greater than 1.75/h, the design of the footing shall be
For general cases falling outside the limitation stated above, the critical spacing i t which the subgrade modulus
theory becomes effective has to ​be ​determined individually.
Evaluation of the factor can be made on the basis of the following formulae:
'​

30 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

Where ​K, ​= ​SR,r

K, ​= ​Coefficient of vertical subgrade reaction, Kips per cu ft (metric tons per cu m) ​K',= b​ asic value
of coefficient of vertical subgrade reaction for a square area with width ​b ​= ​1 ft ​(0.3 ​m).
Kips per cu ft (metric tons per cu m) ​b ​= ​Width of footings, ft
(m) S ​= ​Size or shape factor for a footing on a particular type of soil ​E, ​=
Modulus of elasticity of concrete, Kips per sq ​ft ​(metric tons per sq m) ​I
= ​Moment of inertia of footings ft4 (m4

For sandy soils the size factor S can be determined from the following
formula:

with a limiting value of 0.25 for large

footings.
As for clay soils, the shape factor S can be determined from the following
formula:

When ​n i​ s the ratio of the longer side to the shorter side of the footing. As for extremely long
footings, where ​n ​approaches infinity, S can be assumed as 0.67.
Values for ​Kt,, c​ an ​be ​determined from the results of field tests performed on the subgrade of the
proposed structure or can be estimated on the basis of empirical values in "Evaluation of coefficients of
Subgrade Reaction" by Terzaghi.

​ etenyi's Criteria ​From theory of beams on elastic foundation, Hetenyi proposed rigidity criteria on
5.5.3 H
the basis of ​hL t​ erm which considers width, length and elastic properties of the media. This term is

(K, .​ ​L~)"~ ​hL ​= ​4

El ​where ​K, =​ ​KB = ​Modulus of sub grade reaction ​X ​Width of footing ​-

units of psf. L ​= ​Total length of foundation member ​E = ​ ​Modulus of
elasticity of footing material ​I ​= ​Moment of inertia of footing
If ​1, ​c ​W4 ​footing can be considered ​as ​rigid. For value between ​W4 ​and ​l3 ​semi rigid, and
elastic, if ​> ​I7 ​~ o w l e s ' ~ found this criteria of very limited application.

5.6 ​Modulus ​of ​Sub-Grade ​Reaction

One of the important terms required in analysing foundation on the basis of flexible footings is value of
modulus of sub-grade reaction also called coefficient of sub-grade reaction for the particular soil in the
foundation of the buildings. Mathematically, this can be axpressed ​as ​intensity of soil pressure required to
create a unit
DESIGN APPROACH AND CONSIDERATIONS ​31
 deflection. Theoretically, it can be determined by performing a plate load test and plotting a curve of soil
pressure versus deflection. In actual practice, however, many other factors enter and actual value in field is
different from what can be determined by a simple plate load'test. Major problems associated are: (a) Soil is

not perfectly elastic and results are effected by the magnitudes of soil pressure and deflection ​
3 ​(b) Footing
size affects the value (c) Footing shape also affects
(d) Depth at which footing is located also affects (e) Soil stratification and other changes with depth

which may not show when testing with a small plate ​I ​(f) ​In methods where soil modulus is determined in

laboratory, site condition can not be exactly

duplicated in field laboratory (g) Various authors have suggested different factors to take these problems into
account On the other hand, certain authors have suggested very simple values for modulus of sub-grade
reaction which can be determined from bearing capacity factors used in Terzaghi bearing capacity
equation.

5.6.1 ​Recommended ​by ​~ o w l e s '

~

​ ​36 ​qa
Has related value of modulus of sub-grade reaction with safe bearing capacity by the relation ​Ks =
where qa is the allowable bearing capacity in Kips per sq ​ft. ​A ​slightly improved values are also
suggested by the equation.

where ​c ​is cohesion, ​Nc ​and ​Nq ​are bearing capacity factors, ​Sc ​and ​Sq a​ re shape factors for particular
soil in foot units ​. ​Moreover:

​ =​ ​I+-
NcxB Sc

NcxL

General values suggested by Bowles are given

below:

​ ange of Ks. Kef

Soil R

~ o o s e sand ​30 ​- ​100 ​Medium sand ​60 ​-

500 ​~ e n s e sand ​100-800 ​clayey sand
(Medium) ​200 ​- 5​ 00 ​Silty sand (Medium)
150 ​- ​300 ​Clayey soil ​: ​qu ​5 ​4 ​Ksf ​75 ​- 1​ 50

4<qu18Ksf 1​ 50 ​- 3​ 00
32 ​RAFT FOLINDATIONS-DESIGN AND ANALYSIS

5.6.2 I​ S 2​ 950 :​ ​2950 ​ - ​1981'

Part I​ Indian Standard C
​ ode o​ f Practice for Design a​ nd C ​ afC ​Foundation
​ onstruction of R
Provision relating to determination of modulus of sub-grade reaction are included in ​Appendix B. ​This is
reproduced below. Figures given in bracket in Tables ​I ​and I1 are in Kipdc ​ft. ​units.
B-1 ​General - 1 The modulus of subgrade reaction ​(k) ​as applicable to the case of ​load t​ hrough a plate of size 30
x ​30 cm or between 30 cm wide on the soil ​is ​given in Table 1 ​ f​ or cohesionless soils and in Table 2 for cohesive
soils. Unless more specific determination of ​K ​is done (see B-2 and B-3) these values ​may ​be used for design of
rafl foundation in cases where the depth of the soil affected by the width of the footing m ​ ay b​ e considered
isotropic and the extrapolation of plate load test results is valid.
​ odulus of Subgrade Reaction (​ K) ​for Cohesionless Soils
Table I​ M
Soil Characteristic
Modulus ​Of ​Subgrade Reactions (K) in kg/cm3.
Relative Density
Standard Penetration test
​ oist state
​ rm
For dry o
For submerged state value ​(N)
(1)
(2)
(3)
(4)
Loose
< ​10
1.5 (95)
0.9 (57)
Medium
10 ​to 3​ 0
1.5 t​ o 4​ .7
0.9 t​ o 2​ .9
(95 ​to 3​ 00)
(57 ​to 1​ 85)
Dense
30 ​and over
4.7 t​ o 1​ 8.0
2.9 t​ o 1​ 0.8
(300 ​to 1​ 146) ​( ​185 ​to 6​ 87)
* ​The above values apply to a square plate 3​ 0 X​ ​30 ​cm or beams ​30 ​cm wide
Table I​ I ​Modulus of Subgrade Reaction ​(K) ​for Cohesive Soils
Soil Characteristic

Modulus of Subgrade Reaction (K) in

K~/C~~
Consistency
Unconfined compressive strength, kg/cm2
(1)
(2)
(3)
Stiff
l ​to 2​
2.7 (1 72)
Very stiff
2 t​ o4
2.7 t​ o 5​ .4 (172 ​to 3​ 44)
Hard
4 a​ nd o​ ver
5.4 t​ o 1​ 0.8 (344 t​ o 6​ 88)
* ​The values apply to a square plate ​30 ​x ​30 ​cm. The above values are bared on the assumption that the average loading
intensity does not exceed half the ultimate bearing capacity.
DESIGN APPROACH AND CONSIDERATIONS ​33

B-2 Field Determination In cases where the depth of the soil affected by the width of the footing may be
considered as isotropic the value of K ​ ay be determined in accordance with IS ​: ​9214 ​- ​1979". The test shall be
​ m
carried out with a plate of size not less than 30 cm. B-2.2 The average value of K shall be based on a number of
plate load tests carried out over t​ he ​area, the number and location of the tests depending upon the extent and
importance of the structure. NOTE ​IS:9214 ​- ​1979 lays down that ​Ks c​ an be determined ​as ​slope of the secant
drawn between the points corresponding to zero settlement and point corresponding to ​1.25 ​mm settlement of a
load settlement curve obtained from a plate load test on the soil using a 75 cm dia plate or smaller dia with
corrections for size of the plate used.

B-3 Laboratory Determination B-3.1 For stratifed deposits or deposits with lenses of different materials,
​ e b​ ased on laboratory tests (see
evaluation of Kfrom plate load will be unrealistic and its determination shall b
IS: 2720 (Part ​XI)- ​1972" and IS: 2720 (Part X1I)- 1 9 8 1 ) ~ ~ B-3.2 In carrying out the test, the continuing
​ ay b​ e so selected a
cell pressure m ​ s ​to be-representative of the depth of average stress influence zone (about ​0.5
B to B) B-3.3 The value of K shall be determined from the following relationship

where ​Es ​= ​Modulus of elasticity of soil (see Appendix A) ​E ​= ​Young's modulus of foundation material ​p ​=
​ ​Moment of inertia of structure if determined or of the foundation
Poisson's ratio of soil ​( ​see Appendix A) and ​I =
B-3.4 In the absence of laboratory test ​data, a​ ppropriate values of ​Es a​ nd p
​ ​may b​ e determined in accordance
with Appendix A a​ nd u​ sed in B-3.3 for evaluation of K.

8-4 Calculations B-4. I​ W

​ hen the structure is rigid (see Appendix C) the average modulus of sub grade reaction
may ​also be determined as follows:

​ =
Average contact pressure Ks ​ ​Average settlement of the raft Appendix
​ ​C ​lays down the method of determining
the rigidity of superstructure and foundation and has +n dealt with in ​para ​5.5 above.
Appendix A lays down mettrod of ​determination ​of modulus of elasticity of soil by field tests or laboratory tests.

Equation in ​B-3.3 ​above is based on work carried out by ~ e s i c ~ ~ . ~ o w l e s ' ~ ​has ​observed that the 12th

root of
​ any value will be close to 1 and equation can be considered to be equivalent to
​ AFT FOUNDATIONS-DESIGN AND ANALYSIS
34 R
 ​ ​36 qa where qa is allowable bearing
and suggested that value of ​Ks ​can be calculated by the equation ​Ks =
capacity in kips per sq. ft.

-​
5.6.3 1.S. 9214-1979 ​ Method of Determination of Modulus of ​Subgrade ​Reaction (​ k ​value) of Soils in
Fiedo

Modulus of sub-grade reaction is defined as a ratio of load per unit area (applied through a centrally loaded rigid
body) of a horizontal surface of a mass of soil to corresponding setdement of the surface. It is determined as the
slope of secant drawn between the point corresponding to zero settlement and the points of ​1.25 ​rnm ​settlement,
of a load settlement curve obtained on a soil using ​75 ​cm dia or smaller loading plates with corrections for size
of the plate.
The value of modulus of subgrade reaction so determined is required to be corrected for
(a) when using plates smaller than ​75 ​cm in ​dia (​ b) correction for bending of the plate. (c) correction for
saturation. Average value of k is to be based on a number of plate load tests carried out over the ​area, ​the
number and location depending upon the extent and importance of the structure.

Final correction is required to be applied for the size of actual raft being different from plate. ​5.6.4. IS 8009 ​ -
-​
Part I ​ 1978. Code of Practice for Cirlculation of SettCements of Foundations ​ Part I ​-​ -
Shullow Foundations, Subjected to Symmetrical Static Vertical ​Load ​IS 6403 ​ - ​1981 ​- ​Code of pmctice for
determination of bearing capacity of shallow foundations.

Another method of arriving at the value of modulus of subgrade reactions would be to determine the bearing
capacity of soil for the contemplated raft foundation and the settlement for the same raft foundation in
accordance with the two codes referred above and utilize the same. This value should be more realistic as it is
usual in case of all foundations to fix their dimensions in plan for ​full ​bearing capacity.
However, determination of bearing capacity of soils is not an exact mathematical exercise leading to accurate
results. Large number of approximations and engineering judgements are involved. Two types of failure, ​i.e.,
general sheer failureand local sheer failures have been recognised.
Settlement calculations in the present state of knowledge ​are ​considered to be at best estimate of the most
probable magnitude of settlement. Calculations in this code are based on the assumption that the loads
transmitted to the foundation are static and vertical. The soil mass below is considered to consist of horizontal
soil layers having known properties determined on the basis of base log data from several bores. In practice,
however, no two base log data is similar, soil layers ​are ​not horizontal and it is quite difficult to idealise the soil
below foundation in the manner contemplated in the code.
Different memods of calculating settlement ​are ​applicable for cohesionless and cohesive soils. Because of
difficulty in sampling of cohesionless soil and consequent inability for determining their compressibility
characteristics, settlement calculations are based on semi empirical methods utilising results of either static cone
penetration tests, standard penetration test or plate load test Plate load tests being getting out of fashion, it will
normally be worked on the basis of ​'K v​ alues from standard penetration test.
In case of cohesive soils, settlement is considered to be built up to two components; immediate settlement plus
primary consolidation settlement Procedure for estimation of immediate and consolidation settlement differ for
different types of soil profile, ​i.e., ​nature and location of various soil layers below the foundation. These also
depend even on the fact whether the cohesive soil layer is pre-consolidated or normally loaded clay.
DESIGN APPROACH AND CONSIDERATIONS ​35

Settlements as calculated are required to be corrected for the effect of depth of the foundation and effect of rigidity
of raft. Correction due to depth of foundation is applied as a depth factor. For rigidity it is assumed that the
deflection at the centre of rigid foundation is equal to 0.8 times that for a flexible foundation. TO apply this factor,
one has to decide whether the foundation is rigid or flexible. As already discussed in para 5.5 this itself is full of
uncertainties and approximations.
Further settlements of an actual structure would depend upon the time rate of loading. Methods have been
suggested to take this into account, but these methods again are based on number of assumptions and neglecting
the effect of loading and unloading cycles which undergo during the construction process.
Having determined bearing capacity and settlement modulus of subgrade reaction can be determined by the basic
definition.

5.6.5 Recommendation ​by A ​ mf. ​Alam Singh ​~ l ~ a n ' ~ determined settlement curves of plate loads
​ lpan ​and P
tests already reported by Terzaghi, Peck and Thornburn. He correlated the values of ​np ​(reciprocal of the modulus
of subgrade reaction) with SPT blows which were also available for the tests. Alam singhZ5 re-plotted this
correlation in SI units. Alam Singh has also developed a correction chart for overburden pressure. S.P.T. value
determined in field is corrected for overburden pressure from these charts. He has further suggested that the value
so determined should not be more than 3 ​ ​times the original value of ​N. ​When ​N ​is greater than 15, it should be
further corrected as per relationship.
N = 1​ 5 +0.5 ​(WT- 1​ 5) This ​N ​corr is used in the curve to find out reciprocal of modulus of subgrade reactions.
These values are for plates and have to be corrected for size of the raft foundation. Alam Singh has suggested use
of curves presented by Bjermm and Eggested which are based on a plate size of 0.32 m size sq plate. In this plot
curve 1 and 2 represent the extreme boundaries. Average curve is suggested for ​N ​values between 10 to 20, curve 2
​ ​50, curve 1 for ​N ​< 10 and Terzaghi, Peck curve for ​N ​values between 30 and 50.
for ​N ​value >
As is apparent there are number of limitations to this method. All suggestions are for cohesionless soil. Original
test reported by Terzaghi, Peck and Thornburn cannot be considered to have universal application.

5.6.6 S
​ ummary
It would thus ​be ​seen that even for the same soil data values of modulus of sub grade reaction determined by
methods suggested in the different text books and codes will be different. Which value to ​be ​adopted for the correct
design, is a million dollar question.

STRUCTURAL ​DESIGNER^
DILEMMA

Structural design of raft foundation is being carried out by structural engineers individually, or while working in
any consultancy organisation which could be in public or private sector. Competence of these structural designers
varies widely. On one extreme are those who have some knowledge of structural design, but do not have much of
guidance from their seniors. Such engineers when faced with problem of undertaking design of raft foundation will
pick up a text book or manual on reinforced concrete design and follow the procedure laid down which in most

cases would be conventional combined footing and would normally be safe ut expensive. ​
e​ On the other

extreme are top class engineers who have wide experience and knowledge of structural designs with ​ ability to
analyse the problem, carry out alternative analysis and design a raft foundation which would not only be safe, but
would be economical. Such designers are, however, very few and seldom undertake design for normal buildings.
Majority of the designers have knowledge in between those two extremes. They have read moderately, have some
experience of design and would normally ​try ​to make a design which is not only safe but should also be
economical. These designers will study more than one book or manual on concrete and structural design and will
normally find that the opinions and methods of design recommended in various text books and manuals differ
widely. They will also find that the examples considered in the text books and manuals mostly are very simple and
regular-shaped uniformly loaded rafts which satisfy number of assumptions made in the text book, whereas their
problem is much different. They may also have a feeling that conventional methods of raft design are old fashion
and may lead to high thickness and high quantity of reinforcement; Flexible methods give low thickness and low
values of bending moments and accordingly low cost of reinforcements. They f&e the dilemma as to which of the
methods they should adopt. Quite often since no straightaway guidance is available in the books or manuals for
practical design of raft, they also finally take up one or two text books and work out a design. Quite often analysis
is carried out on computer using flexible approach utilising the value of modulus of subgrade reaction suggested by
the soil consultant or worked out by them ​as ​per method given in any book and simultaneously design the raft for
vertical loads neglecting various other factors which affect the design of raft. Such designs are seldom satisfactory,
though structures designed by them, do not show any distress to start with. Structures seldom get loaded to the
design loads. How will these structures behave when subjected to full designed loads including seismic effects can
be judged only by number of otherwise standing structures failings in such circumstances.

​ ​:: K
STRUCTURAL DESIGNERS'S DILEMMA ​$ E ​ eeping such designers in view, it was felt that enough

material should ​be ​made available to them to allow them

​ to appreciate the effect on raft foundation of variation
in the values of the various parameters adopted by them in design and also the effect of neglecting base moments
of the column and horizontal loads.
Studies have, therefore, been carried out on the effect of various parameters on the values of bending moments,
shear force, contact pressure and deflection in the rafts already constructed for actual buildings. These
 studies are presented in Chapters ​7 ​and ​8.
Finally, keeping in view the various constraints under which a designer has to cany out the design, suggestions have
been made for methods to be adopted in Chapter ​12 ​for various situations.

STUDIES CARRIED OUT ON

EFFECT OF VARIOUS
PARAMETERS ON DESIGN OF
RAFT

The usual practice of design being followed is to work out preliminary sizes of the raft, ​i.e., t​ hickness of
the slabs, if it is uniformly thick raft or beam size and slab thickness in case it is beam and slab system on
the basis of shear and analyse the raft for vertical loads alone. As an improvement where computer
facilities and greater expertise are available, raft is analysed as flexible raft selecting one particular value
of modulus of subgrade reaction, one assumed size of the raft and vertical loads alone. Values of bending
moments thus obtained are used. In both these designs unless the preliminary sizes selected are found to be
structurally unsafe in resisting moments and shears, even after addition of permissible reinforcement, the
design is completed and finalised. As already pointed out in previous chapters the real position is not so
simple. Different designers may select different preliminary sizes, different values of modulus of subgrade
reaction even for the same soil, and pattern of pressure distribution under the raft. In actual buildings,
columns have base moments which are resisted by the junction of the raft and the columns. Buildings
subjected to earthquake forces have not only increased column base moments but also undergo cyclic
effect in which vertical loads in different groups of columns decrease and increase. Studies have, therefore,
been carried out to consider on the design of raft foundation the effect of neglecting some of these aspects
and making assumptions which in fact are not true. These studies have been carried out in four parts.

7.1 ​Study ​1

In sophisticated flexible analysis, utilising computer, it is soil properties which matter to a large extent. In
exact analysis all soil properties matter, but in commonly adopted analysis where soil-raft interaction is
idealised as a.spring of known rigidity most important soil property is modulus of sub-grade reaction. The
rigidity of raft which is determined by the size of the raft and effect of super-structure on the same, is
another vital parameter which comes into play in any analysis. The effect of variation in values of both
these parameters on the value of bending moments and shear forces, one gets on an analysis, has been
studied in this study. Efforts have been
STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS ​39

made to present results, in numerical values and show the large variation which, one can get .for the same
structure, having a particular loading pattern founded on the same soil when different sizes of raft or
values of modulus of sub grade reacrion determined by various methods available in literature are
adopted. While carrying out this study, only vertical loads have been considered. Contribution made by
super structure in the rigidity of raft has been neglected.

​ xamples Selected ​Most of the text-books on structural engineering and reinforced concrete design,while
7.1 .I E
dealing with examples on raft analysis, generally consider a simple symmetrical shape with more or less
symmetricaVunifom loading. But in practice this never happens. Even when the shape may be
symmetrical, the loading is not. TO make the study realistic, raft foundations for actual buildings have
been considered in this study.
One eight-storeyed block of residential flats consisting of four flats on each floor with a central core having
staircase, lift and other service areas, which has already been constructed few years ago, has been
considered in this study. The central core goes beyond eight storeys to provide staircase mumty,
machineroom and water tank. Ground floor has got part parking. The central core is separate from other
blocks. The raft foundation consists of a slab having uniform thickness, one for each side block and other
for central core. The central core is symmetrical in shape about one axis. The side block is not
symmetrical about any axis. Loading on these blocks are as per actual loads obtained during design
process. Example 1 relates to the side block, and Example 2 to central core.
The third example considered is the front block of another six-storeyed institutional building, which consists of a
front block and rear block separated by expansion joints. Its front block isrectangular in shape but has
unsymmetrical loading. The raft consists of beams in both direction and a slab monolithic with the
b%ams.
In two out of these three examples, raft dimensions have been so adjusted that the centre of gravity of vertical loads
and centre of gravity of raft area coincide. The loading can, therefore, be considered to be symmetrical
and it is this aspect which is very important. In practical examples, it may generally be possible to
coincide CG of raft and load. It is, however, not possible to have simultaneously a symmetrical shape in
plan also.

7.1.2 ​Rafr Size ​The raft thckness actually provided for Example 1 (Fig. 7.1) is one metre. In this study, it was
considered that this thickness could vary from 80 cm to 1.2 m. For Example ​2 ​(Fig. 7.2) actual thickness
provided is 1.2 m. This was considered to vary from ​1 ​M ​to 1.4 M ​. ​In Example 3 (Fig. 7.3) the slab
thickness is 50 cm and the transverse beams are 80 cm ​x ​150 cm (including slab). Longitudinal beams are
85 cm ​x ​110 cm. While the beam sizes are considered to remain the same, the slab thickness is taken to
vary from 50 cm to.90 cm ​. All ​these variations are considered in steps of 10 cm each. The effect of
rigidity of the super structure has not been taken into account.

7.1.3 ​Soil ​Investigation .​ ​Soil investigations to determine the safe bearing capacity of soil for purposes of design were
done through specialised consultants and their reports obtained. Since in conventional design, properties like
modulus of sub-grade reaction are not utilised, these consultants were not requested to intimate value of
modulus of sub-grade reactions and they did not do so. Values of modulus of sub grade reaction were,
therefore, calculated by variousmethods described in ​5.6 ​above. Details of soil investigations indicating soil
strata at various depths, 'N' values from standard penetration tests, location of water table and values of
modulus of sub-grade reaction calculated are indicated in Figs. 7.4 and 7.5 for Examples l , 2 , and 3.
RAFT FOUNDATIONS-DESIGN AND
ANALYSIS

T ​16985 ​t

RAFT ​SLAB
THICKNESS4OOoMM

Fig. ​7.1 ​- 1 0 6 6 0 ~ ​. .​ t 2 2 8 0 t 6090

22904
 + l 5 8 O b ​7500 ​-tlW+ ​-10660

~t
Fig. ​7.2
41
STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS ​

f​ I i~nn
​ ​JKN ​1 ​-N ​I ​6213 ​4m.4 ​1 ​SRI ​~ K N ​UN ​1767 ​TKN ​~9- BEAM
​ SIZES ​
5 ​1 B1=650X1500mm
​ ​B2=600X1500mm ​g

B3=800X 1 1 00mm

COLUMN SIZE

=400X800mm

T h ​i ​ckness

PLAN OF RAFT AT FRONT BLOCK

​ RE ​IN
<ALL ​DIMENSIONS A MM)
 Fig. ​7.3

In soil investigations for Examples 1 and 2, four standard penetration tests have been carried out extending to a
depth of 9 m. ​'N' ​value are varying from 9 to ​50, ​water table in different bore holes is varying between 3.1 8 to 4 m.
below existing groupd level. In bore hole No.3 top 30 cm or so had brick bats mixed with silty clay. Four dynamic
cone penetra\tion tests were also carried out on an interval of 30 cm each with depths extending upto 14 m. Values
of K,in Kipp per ft' units have been calculated by the four methods given in para 5.6. Soil consultant had
recommended a safe bearing capacity of 150 ICbI/m2 at a depth of 2.5 m. It would be seen that there is large
variation in the values of K,determined by various methods and for various bore holes. The minimum being 70 and
maximum being 476. Average value for the four bore holes by each methods are also indicated. For the purpose of
study, 4 values, ​i.e., 7​ 0, 156, 389 and 415 kipps per ft. have been adopted for studying the effect of variation in
value of modulus of subgrade reaction. Any designer could have adopted any of the 20 values in Fig. 7.4. These
values in metric units would be 10996,24506,61108,65203 K N / ~ ~ . For soil investigation in case of Example 3,
three standard penetration tests and four dynamic cone penetration tests extending upto a depth of 10 m. were
carried out. The depth of the water table varied form 1.5 to 1.84 m. below ground level. Fig. 7.5 would show that
value of K,varies from bore hole to bore hole and from one method to another method. Out of three bor holes one

bore hole was located within the raft dimensions of front block, another one was very near to ​ i​ is raft and 3rd

one was under the rear block. Two values selected are 1 17.42,90.14 Kip ft. or 18402 and I4956 ICbI/m3 ​. ​It is
usual practice to assume one value of K,y for entire raft, but this is not the reality ​. ​Ifmil values under the raft
changes from one bore hole to other, soil parameters would also change. To take into account the variation of K,
under the same raft from one portion
42 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

EXAMPLE ​1 ​AND ​2
 Average

100 70 295 160 156

Ks ​By IS CODE ​IN Kcf ​

229 202 ​
​ cf ​
Ks ​By BOVLES MD ​I ​a K 254 264 237.25

344 347.0 311

Ks ​BY BOVLES MD ​11 ​tn ​K c f 268 285 ​

ALPAN'S CURVE In ​Kcf 393 348 476 476 389

Average Value ​of ​N ​22.71 20.03 25.2 26.21 23.54

Fig.
7.4

to anofher, the raft was considered to have different values of ​K, ​in the ​3 ​portion which had been demarcated
on a rational basis. These values are two combinations of ​224.65,270.33,372.14 ​and ​114.07 ​, ​131.23,118.19
Kip/c ft.in metric units these values would be ​34976, 42466, 58460 ​and ​17920, 20615,18567 ​in KN/~~,
respectively. ​A ​kafe bearing capacity of ​1.5 ​kg/crn2 ​was ​suggested by soil consultants for ​this ​raft.

7.1.4 L ​ onsidered ​in ​Study I​ n accordance with the common practice, only vertical load ​as
​ wd C
being transferred to raft ​from ​columns, neglecting base moment for columns are considered.
STUDIES CARRIED OUT ON EFFECT OF VARIOUS
PARAMETERS

EXAMPLE ​-3
 SILTY W D

SMDY SOU

average

KsBy BOWLES IN ​MD ​I ​in K

​ cf ​118.19 90.14 144.27 117.47

KS ​BY ​BOWLES ​MD ​I1 in ​Kcf ​372.14 435.45 ​270.33 327.89

​ cf ​127 106.1 115.74 ​115.74

Ks By IS CODE ​IN ​Kcf ​274.17 113.32 312.57 ​274.19 ​Ks By ALPAN'S CURVE ​m K

ACRAGE VALUE ​OF ​N ​11.8 9 14.4 11.73

Fig.
7.5

​ he analysis has been camed out on an electronic computer utilising ​SAP ​-1V c​ omputer programme based on
​ nalysis T
7.1.5 A
finite element methodj~inite element method of structural analysis is based upon the general principle of
going from part to whole. The elastic continuum whole forming the structure is discretised with a number
of finite elements just as the way a building is discretised as elementary beams, columns and slabs for
purpose of analysis and design. For a two-dimensional continuum considered for raft analysis, such finite
elements are triangular, group of triangles or quadrilaterals. Each of the node has six degrees of freedom,
three each for displacement and rotation. Raft is taken as a thin plate on elastic foundation ​. ​The soil is
idealised as a spring
44 ​RAFT FOUNDATIONSDESIGN AND ANALYSIS

under each node on the concept of Winkler's model. Spring constant is a function of soil sub-grade
modulus and the area under the node. Suchan analysis suffers from the disadvantage that when the raft
gets separated ​from ​the soil supporting it, physically, under such condition the soil will no longer act as
support. But in the method adopted for analysis the spring which is representing the soil continuum as a
support takes tension. However, it has been felt that this limitation is not of much significance because
such tension would be exceptional.
Fig. ​7.6 ​indicates the finite elements in which raft of Fig. ​7.1 ​has been divided. Six elements in
this figure have been selected and used for comparison of values of bending moments in ​X a​ nd ​Y
directions and.deflection in the.2 direction. Comparative values for Example ​1 ​are indicated in Table ​7.1,
for a constant thickness of 1 M and varying values of modulus of sub-grade reaction. Ratio of maximum
and minimum values of ​M, a​ nd ​My i​ n ​X ​and ​Y ​directions for particular elements have been worked out.
These ratios vary considerably. Maximum value among all the elemental ratios have been indicated at the
bottom of the same Table. This ratio has also been worked out for deflection and indicated at the bottom
of the same Table. Table 7.2indicates similar values for a constant value of ​K, t​ aken as ​70 ​KipsJc ft
(10996 ​KN/m3) and varying values of thickness from ​80 ​to 120 cm. Another set of ratios has been
worked out for bending moments and deflection selecting the maximum and minimum values for the
same element irrespective of the thickness or modulus of sub-grade reaction. Maximum values of these
ratios have been shown for each element in the last column of Table 7.2 and-their maximum have been
indicated at its bottom.

(ALL DIMENSIONS ARE IN MM>

Fig.
7.6
STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS
-, ​10720 ​-1-

<ALL DIMENSIONS ARE IN MH) ​Fig. ​7.7

<ALL DIMENSIONS ARE IN MM) ​Fig. 7.8

RAFT FOUNDA-HONS-DESIGN AND ANALYSIS
EXAMPLEI
​ niform Average Thickness ​1.0 ​Metre
Table ​7.1 U
Plate Element No.
K, Value In
Average ​'6' i​ n
BM, in
BM, in
​ f and ​Node No. K N / ~ ~
Ratio o
mm
KNdm
KNdm
MadMin

--
65203

& ​---- 18.0

​ 30,81,36,31
10996

- 1​ 6.08 ​154.6 24506

- 0​ 7.94
130.7 28
61 108
94.3 65203
98.5
80,33,32,37
130.5 106.5 29
58.9 62.7
Maximum Ratio ​in ​any Element

​ M,
, Mm ​

Fig. 7.7, Tables 7.3 and 7.4 indicate similar information for Example ​2. ​Fig. 7.8, Table 7.5 and 7.6 indicate
similar information for Example 3.
7.1.6 ​Discussions o​ fR
​ esults
Table ​7.1 ​indicates that for 1 ​M ​thickness of raft in example ​1, ​bending moment in ​X ​direction could be ​as
much ​as ​8.43 times more depending upon the value of modulus of sub-grade reaction selected in the analysis.
Bending moment in ​Y ​direction could ​be ​7.97 times more. Deflection value could increase upto ​6.66 ​times. ​A
higher value of modulus of sub-grade reaction means lesser values of deflection directly proportional to its
value.

- ​ 03.06 ​02.86
-​
10996 24506

- ​ 05.28 11.80 61108 - 02.10 65203

-​ ​
- 01.95
11.0
135.0 11 5.2 95.7

M, ​
4 ​6 ​1.41 1.64 5.62 97.9
67.5 65.9 59.5

MU Mfl 6
​

1.13 2.22 ​- ​6.05 61 .O

8.43 ​I ​7.97
I
STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS
Plate Element No.

(​
Thickness in ​ A v​ eroge ​ '6' ​ BM, ​ in ​
I​ I​
BM, in ​
I​
Table ​7.2 ​ Uniform ​ Modulus ​ of ​ Sub-grade ​ Reaction ​ 10996 K N ​ /

Ratio ​
~​ ~

of ​
I​ Over All ​
and ​Node No.

tnetre in mm K N d m K N d m Max./Min. Max./Min

18,24,25, 19
0.80
- ​10.08
49.4 0.90
- ​10.31
64.5 18
1 ​.OO
- ​10.52
79.0 1.10
- ​10.70
92.8 1.20
- ​10.85
105.5
19,25,26,20
0.80
- ​10.35
53.3 0.90
- ​10.53
70.0 19
1 ​.OO
- ​10.70
86.0 1.10
- ​10.84
101 .O 1.20
- ​10.96
1 ​14.9
30,81,36,31
0.80
- ​17.07
122.7 0.90
- ​16.57
129.2 28
1 ​.OO
- ​16.08
135.0 1.10
- ​15.63
140.2 1.20
- ​15.21
145.0
80,33,32,37
0.80
- ​11.53
66.5 0.90
- ​1 ​1.66
67.2 29
1 .​ oO ​1.10
Maximum Ratio in ​any ​Element
Table 7.2 shows that for a constant value of modulus of sub-grade reaction taken ​as ​70 Kipslc ft (10996 ​Km3), ​the
bending moment in ​X ​direction could be ​as ​much as 2.16 times and can even change siin depending upon the
thickness of raft selected by the designer. This values for bending moment ​in ​Y ​direction is 3.53. The value of
deflection does not change much and becomes only a maximum of 1.21 time depending upon the thickness of the

raft. ​- ​11.80 ​- ​11.94

53.6 69.7 84.5
67.5 67.6 ​1 ​.20
- ​12.07
67.5

​ 6
Mu Mv ​ M​ y" 6
​ M, ​ 2.14
​
2.04 ​,
8.79
1.08 97.9 109.5
44.3
2.16 53.4
1.71 61.7
1.059 69.1 75.6
130.7
1.18 143.0
1.34 154.6
1.12 165.2 174.9
100.5
1.02 115.9
1.55 130.5
1.047 143.9 155.9
2.16 3.53 1.12
10.33 6.87

MU ​My? ​6 11.27
​ 4.20 6.60

​ 1.52
MU M~? 6 ​ 1.86 5.98

M,
1.14

My, ​
6.
2.65 6.19
11.27 10.33 06.87
RAFT FOUNDATIONS-DESIGN AND ANALYSIS
EXAMPLE ​I1
Table ​7 3 ​Uniform Average Thickness 1.20 Metre
Ks Value ​in
Average ​'6'
BM, ​in
BMvv i​ n
Ratio of K N / ~ ~
in mm
KNdm
KNdm
Mar/Min.
10996
8.83
144.6
344.6
1.49 24506
3.91
130.0
291.6
1.56 61108
1.58
99.7
233.1
6.05 65203
1.46
97.2
220.6
10996
9.86
140.6
461.6
1.32 24506
4.34
137.0
402.8
1.44 61 108
1.82
106.6
326.1
5.87 65203
1.68
109.9
321.5
10996
9.76
73.3
280.1
1.67 24506
4.33
64.3
100.7
2.99 61 108
1.53
44.0
128.1
4.38 65203
1.55
48.5
93.6 10996
11.49
343.8
397.4
1.56 24506
05.60
300.8
262.8
2.52 61 108
02.43
240.8
157.4
4.83 65203
02.38
220.6
165.8
10996
7.7 ​1
130.7
152.3 24506
3.38
113.8
74.5 61 108
1.23
93.4
31.7
Table 7.3 shows that ​the ​variation of bending moment in ​X ​direction for the same thickness of raft and varying
modulus of sub-grade reactionis 1.67. This value for bending moment in ​Y ​direction is 7.46 and for deflection is
6.93. For the fixed value of modulus of sub-grade reaction and varying values of thickness these ratios in ​X
​ ​direction and deflection are 1.5888,2.149, 1.105, respectively. The overall variation values for this
direction, Y
raft of bending moment in ​X ​direction, bending moment in ​Y ​direction and deflection are 2.24, 10.01 and 7.26,
respectively.
On comparison of values' for Examples 1 and 2, it would be apparent that
(a). The variation in bending moments in X ​ ​direction is more for Example 1 ​as ​compared to Example 2. (b) The
variation in value of bending moment in ​Y ​direction for Example 2 is lower ​as ​compared to
values of variation in this direction for Example 1. ​Plate Element No. and Node No.
3.4, 10,9

3 ​Mn ​
'7 ​ 5,6,7,8 ​4 ​4 ​Mw ​6 ​17.18.23.24

15 ​M, ​
4 6​ 21,22,28,27
​

18 ​Mu 4
​ 6​ 34,35,39,38
​
28

​ 6 1.47
M, 4 ​​ 7.46 6.27 ​ 1​ 65203

1.28
89.2
20.4
36,37,41,40
29
10996
6.72 24506
2.79 61 108
I .09
60.10
148.9 59.6
119.8 53.8
96.5

Mu M
​ ​ ,​ 1.12
S ​ 1.55 6.93 65203
0.97
53.5
96.0

Mu ​Maximum Ratio in any Element

6
M,~n ​

1.67 7.46 6.93

I​ STUDIES CARRIED OUT ​ON ​EFFECT ​OF ​VPiRlOUS PARAMETERS ​49

Table ​7.4 ​Uniform Modulus of Sub-grade Reaction 10996 ​K N I ~ ~

Plate Element No.
Thickness in
Average ​'6'
EM, ​in
EM, in and No& No.
Metre
in mm
KNmh
KN&
(c) Deflection of raft is more or less independent of the thickness of the raft and varies only very slightly, whereas
it is more or less directly proportional to the value of modulus of sub-grade reaction. Higher the value of modulus
of sub-grade reaction selected lower are the values of deflection and higher the values of bending moments.
Ratio of Max./Min.
Max./Min
3,4,10,9
1 .OO
8.72
136.5
308.8 1.10
8.77
141.5
327.8
1.077
1.47 1.206
1.69 3
1.20
8.83
144.6
344.6 1.03 1.30
8.94
146.3
359.4 ​ 1.40
'​

8.98
147.0
372.5
6.15
P

9.74 1.10
9.79 4 1.20
9.86 1.30
9.92 10.01
9.74 9.77 9.76 9.73 9.69
11.13 11.82 11.49 1 1.20 10.95
7.62 7.66 7.70 7.75 7.80 ​. 6​ .37 6.55 6.72 6.88 7.04
MU

: ​My). 6
​
​ ! ​i
5.6.7.8
1 ​.OO
139.2
421.1
1.029 ​
Mu ​ 140.8
442.4 140.6
461.6

M~~ ​6 ​1.173 1.027 139.1

479.0 1.40
136.8
494.7
1 .OO 1.10 1.20
68.2
208.6 ​M* ​71.1 244.6 73.3
280.1

My?, 6
​

5.96
17,18,23,24
1.116
1.73 1.664
3.7 ​1 ​15
1.009
6.39 1.30
74.9
314.5 1.40
76.1
21,22,28,27
1 .OO
3 17.8
1.122 1.10
333.0
1.588 18
1.20
343.8
1.08 1.30
351.4 1.40
356.6
34,35,39,38
1 ​.OO
199.3
1.588 1.10
125.5
2.149 28
1.20
130.7
1.024 1.30
135.1 1.40
138.9
36,37,41,40
1 ​.OO
60.1

1.027 ​1 ​29
1.10 1.20
60.3 60.1
347.2'
305.4 351.4 397.4

Mu M
​ ~~
6​ ​1.62 3.08 4.97 442.3 485.1
96.2 124.0 152.3

​ ~~ ​6 ​2.24 10.01 6.35 1 80.1 206.8

MU M
127.9 138.4 148.9

MU 4 ​
​ 6
1.13 1.317
1.76 1.105
7.26 1.30
59.5
158.9 1.40
58.7
168.5
Maximum Ratio in any Element

M u ​M)? ​6
1.588
2.24 2.149
:0.01 1.105
7.26
RAFT FOUNDATIONSDESIGN AND ANALYSIS

Table ​7.5 ​Uniform Average Thickness ​0.50 ​Metre ​1

Note: ​Set-I ​- ​For Combination Set-I1 -For Combination of of Modulus Modulus of of ​Plate Element No. K, Value in
Average ​'6'
BM, in
BM, in
Ratio of and Node No.
KN/~~
in ​mm
KNm/m
KN;~/~
MdMin.
31,32,41,40
1 8402
0.52
31.0
198.0
4.48 141 56
0.18
39.4 251.4
3.67 28
SET-1
0.56

8.8
68.2
3.71 0.65
29,30,39,38

​0.32 26
- 0.32
​ ​
0.52 0.45
34,35,44,43
0.70 0.28 3 1
0.5 1 0.63
47,48,57,56
-01.10
42

- ​ 04.46 12.70
-​ ​ ​ 9.15 14156 ​- 1​ 1.60 10 ​SET-I ​04.12
-09.10 11,12,21,20 1 8402 ​- 0
SET-I1 ​- ​08.85 ​- ​14.0 18.0
Maximum Ratio in ​any ​Element
Subgrade Subgrade Reaction Reaction ​- - ​34976,42466,58460 22731,2061 5, ​KN/m3 ​18567 ​KN/m3
Observation (a) above is justified because Example 1 is unsymmetrical in plan in respect of both axis as compared
to Example 2 which is symmetrical abput ​Y ​axis. This symmetry results in lower variation in the value of bending
moment in ​X ​direction. It is, therefore, apparent that while the thickness and the value of modulus of sub-grade
reaction selected for a particular raft which is symmetrical in plan and symmetridal in loading may not be of much
consequence, but this will not hold good ​as ​soon as there is deviation from symmetry and the values may go
several times more.
Tables 7.5 and 7.6relating to Example 3 which is raft with beam and slab construction indicate that bending
moment in ​X ​direction for the same thickness of raft, but varying values of modulus of sub-grade reaction can be
as ​much ​as ​4.6 times. The corresponding value in Ydirection and deflection are 4.71 and 3.71, respectively. For
and the 8.61, same respectively, value of modulus for ​direction o​ f sub-grade bending reaction moments, and ​Y
varying direction thickness bending values, moments these and ratios deflection. are 4.31.6.55 Values of these
ratios for overallrariation are 20, 12.17 and 16.4, respectively.

​ 6​ SET-11
44 ​
29.0
18402

13.0 14156 18.3 ​SET- I ​1.8 SET-I1

​
11.0
1 8402

16.0 14156 20.7 ​SET-I ​4.5 SET-I1

​
15.0
18402
- ​11.0 14156
- ​11.9 ​SET-I SET-I1
- ​ 11.0 7.6
​ 176.0 216.0 269.5 80.6 188.0
-​
196.0 248.9 52.9 187.0
52.0 60.2 42.9 50.0

M, ​M.v 6
​ 10.16
​ 03.34 01.42

Mn

4.60 ​M~ ​6 4.7

​ 1 2.50
Mu

6.
My, ​

1.67 1.40 2.85

STUDIES ​I
i​
​ CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS ​Table ​7.6 ​Uniform ​Modulus ​of ​Sub-grade

51 ​
Reaction ​- ​18402 K N I ~ ~ ​ I ​t
P!ate Elenent No.
Thickness in
Average ​'6'
BM, in
BM,, in and Node NO.
Metre
in ​mm
KNmh
KNmh
31,32,41,40
0.50
0.52
31.0
198.0
Rafio of
Over All Max./Min.
MaxJMin
0.60
- ​0.34
50.7
331.1 28
Mu
0.70
- ​1.28
69.2
451.6

My! ​
6 3.18
​
11.19 3.32
09.64 5.56
16.45 0.80
- ​2.14

85.5
562.1 0.90
- ​2.89
98.5
657.2
29,30,39,38
26
0.50
0.32
13.0
2 16.0 0.60
- ​0.43
22.1
347.7 0.70
- ​1.30
28.4
464.4

MLV ​Mu ​6 ​2.77

20.00 3.07
08.21 8.61
08.61 0.80
- ​2.09

33.0
570.6 0.90
- ​2.78

36.0
662.0
34,35,44,43
31
0.50
0.70
16.0
196.0 0.60
- ​0.02
24.6
316.5 0.70
- ​0.09
30.9
436.3

​ 2.46
Mxr ​M\,' 6 ​
08.73 3.28
12.17 3.33
08.32 0.80
- ​1.63
35.9
547.2 0.90
- ​2.33
'​39.3

643.6 ​
I​ 47,48,57,56 ​42

0.50
- ​10.10
- ​11.0
52.0 0.60
- ​10.78

- ​15.3
101.4 0.70
- ​10.01
- ​20.3
137.9

MAT ​2.80

3.94 1.24 ​1​ 03.90 04.78 02.85 0.80 0.90

11, ​12.21.20
03.77 10.67 10
02.82
3.18
20.00

My? ​
6
6.55 8.6 ​1
​ onclusions
7.1.7 C
The study shows that values of bending moments in a raft may vary several times depending upon the raft size
selected For the design and the soil properties under the raft. This variation increases further ​as ​the deviation ​from ​the
symmetry of the shape or loading of the raft increases. In the study under consideration, the values
-09.32 ​- ​08.73

- -25.0
​ ​30.0
173.1 205.0
0.50
-09.15
- ​15.0
21 ​.O ​0.60
- ​08.78
- ​21.6
42.3 0.70
- ​08.27
- ​28.0
71.4 0.80
 -07.84
- ​33.9
104.3 0.90
- ​07.44
- ​39.2
137.5

MY,. ​6'
M,

My? ​
6 ​M,
2.61 6.55 1.23
Maximum ​Ratio in ​any ​Element
1 2.17 14.45
A ​comparison of values in Tables ​7.5 ​and ​7.6 ​with those given in Tables ​7.3, 7.4 ​will show that ratio of bending
moment in ​X ​and ​Y ​directions for raft in Example ​3 ​is higher as compared to raft in Example ​2. ​Comparison is not
made with Example ​1 ​which is very unsymmetrical in plans.This indicates that variation ​in ​values of bending moment
would be more in case of rafts with beam and slab construction as compared to raft of uniform thickness.
52 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

could be as much as 20 times. In these examples effect of base moment of columns due to vertical or
horizontal loads or variation in wind ward and leeward column loads due to horizontal forces has not
been-considered. It is felt that consideration of these values would increase the variation further. Adoption
of an accurate method of analysis like finite element method, solved on an electronic computer using
vertical loads alone, with a tentative raft size and value of sub-grade modulus determined in a routine
manner in itself is, therefore, not enough in determining realistically the values of bending moment and
deflection for a particular raft. Raft designed in such a manner may not be able to resist the forces
properly and finally lead to failure.

The results of the above study were presented by the author in a 'Seminar on Concrete in
Foundation Systems' organised by the Indian Concrete Institute at Madras on December 29-30, 1988.

7.2 ​Study 2 -Effect ​of ​Horizontal Loads

Seismic loads are basically inertia loads which can act on the building in any direction. For the purpose of
analysis and design these are assumed to be acting in ​X ​and ​Y ​directions. Though these are basically
dynamic loads, it is common practice to treat them as static horizontal loads determined by seismic
co-efficient method. For more accurate analysis, the response spectrum method or detailed dynamic
analysis using design accellerograrn as input is required to be canied out. It is usual to adopt seismic
co-efficient method for buildings upto 12 to 13-storeyed height. In this study, the effect of horizontal
loads has accordingly been restricted to seismic co-efficient method. It is expected that the effect would
not be much different in case any of the other methods was used.

The effect of horizontal loads on the building structure is to alter the moments at the column bases
and also to increase the vertical load in some columns and decrease it in the others. Examples 1 and 2 in
Study 1 were selected to study the effect on raft moments where horizontal load are also acting. The
super-structure in both these examples consist of frames in both the directions. The super-structure for
both these examples was analysed on three dimensional analysis for buildings programme utilising
earthquake force on seismic co-efficient method with earthquake acting in ​+x d​ irection, ​-x d​ irection, ​+y
and -​ y ​directions. The output obtained from the ​TAB ​analysis was used as an input for the analysis of raft
foundation for both the examples. Further details and the results achieved are discussed below:

7.2.1 E ​ elected
​ xample S

Examples 1 relates to the side block. For the purpose of analysis a thickness of 1 metre for the raft slab and
​ ​10776. KN/m3 has been assumed. There are 6 loads cases which have been
value of soil modulus ​Ks =
analysed; load case 1 vertical loads only; case No.2 vertical ​+ c​ olumn base moments; case No.3 vertical

loads ​
+ b​ ase moments ​+ s​ eismic force in ​+ '​ x' d​ irection; case No.4 ​: ​vertical load ​+ b​ ase moments ​+
​ ​direction; case No.5 vertical load ​+ b​ ase moments ​+ s​ eismic force in ​+y ​direction; case
seismic forces -'x'

No. 6 vertical loads ​

+ ​base moments ​+ ​seismic force in ​-y ​direction. For the purpose of comparison 6
plate elements as originally ​selected for this example in Study 1 have been considered. Comparative values
of Delta, ​i.e., ​settlements, bending moments ​in ​x d​ irection, bending moments in ​y ​direction have been
tabulated in Table ​7.7. ​The range over which each of these three values vary from one load case to another
are indicated in column 7. Column 8 indicates minimum and maximum ratio these values have with
reference to value when vertical loads only are acting. These ratios, thus, indicate the number of times
each of these value can vary if horizontal loads are acting.
STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS ​53 ​ able ​7.7
​g T
P ​Plate Element No.
Load

Average
BM, i​ n
BM,, i​ n ​& ​Node No.
Case
%'in ​mm
KNdm
KNdm
Range of
Ratio Variation
(W.R.T Case-I) ​Min. Max.
Min. Max.
3
1 - 11.62
83.3
96.7
tiAvG
-11.05 ​+ -​ 15.83 ​2 -​ 12.82
41.6 1,7,8,2

3 -​ ​15.83
42.7 - 11.18
-9.7

​ ​
BM, 3 1​ - ​ 15.26 11.05
-​ ​ -9.97 -9.15 - 7.42

Similar study is done for Example ​2, ​i.e., C​ entral core the values are represented in Table 7.8. Thickness
selected is that as actually provided, ​i.e., ​1.0 m with soil modulus ​= ​10776 ​KNI~~. ​Mark (S.C.) in these
Tables mean sign change.
36.7 65.2
9.7 ​+ 8​ 3.3 ​4 5 ​41.5 ​6
49.9
10.5 19.4 50.4
BM,
10.5 ​+ 9​ 6.7 ​(S.C) ​0.1 08 ​+ 0​ .67
9
-1.2 ​~AVG

14.7 ​13, 14.6.79

28.4 ​1 ​.O
BM,
23.2 57.3
BM,,
L
18
79.0 ​ti, ​136.1 18,24,25,1?
174.3 115.2 227.1
3.3
7.0 ​-+ 1​ 0.92
0.74 ​+ 1​ .095 - 24.5 -16.7
23.2 ​+ 5​ 7.3

19.33 ​+ 4​ 7.75 4 ​5 ​- 30.8 - 23.4

30.8 ​-+ ​3.3
(S.C) ​5.06 ​+ 9​ .33 ​6 -​ 26.2
(S.C) ​(S.C)

1 ​84.5
​ ​1.03
10.52 ​+ 1

1.016 ​+ 1​ .048 ​2 3 4 5 ​129.4 94.0 181.1 138.5

'
BM, BM,
​
84.5 79.0 ​-+ -+ ​181.1 227.1
1.112 1.46 ​+ + 2​ .87
2.143
6 138.1
​
137.2
19

1 86.0
​
61.7
ti, ​9.55+11.26
0.89 ​+ 1​ .05 ​2 19,25,26,20
​

3 4 5 6 132.7
​ 170.9 11'3.5 238.8 128.1
83.6 56.9 121.0

BM,
117.1 71.6
BM,,
86.0 ​+ 2​ 38.8
56.3 ​-+ 1​ 21.0
1.32 ​+ 2​ .78
0.92 ​+ 1​ .96
28

1 ​135.0
157.6
tiAvG
12.1 ​1 ​-+ ​20.58
0.75 ​+ 1​ .28 ​2 30,81,36,3
​ 1

3 4 5 6 ​135.3
157.7 132.6 143.7 220.8 132.0
149.7 172.8
BM,
135.0 204.2
BM,
132.0 ​-+ ​220.8
135.0 ​+ 2​ 04.2
0.98 ​+ 1​ .64
0.87 ​-+ 1​ .32
29

1 67.5
​
130.5
tiAvG
11.80 ​+ 1​ 8.32

1.18 ​+ 1​ .55 ​2 80,33,32,37

​ ​3 4 5 6
103.6
210.8 101.0 106.7 163.4 88.9
188.9 248.5
BM,
229.7 213.9
BMvv
67.5 ​-+ ​163.4
130.5 ​+ 2​ 48.5 - 10.92 - 8.32 - 9.83 ​- 1​ 0.52 - 10.85 -11.03 - 10.69 - 10.79 - 10.82

- ​ 10.70 ​10.39 -​ 09.55 ​ - ​10.19 11.26 - 10.35

-​ -​ ​

- 16.08 - 15.76 -12.11 - 20.58 - ​ 17.68 ​15.92 -​ ​11.80 -16.16 - 18.32

-​
1.32 ​+ 2​ .42 - 13.92 - 18.29
1.45 ​+ 1​ .90 -16.21
54 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS
I ​and
Node No.
Table ​7.8

1​ Range of

hAVG
BM, BM,
6AvG
BM, BM,
6AvG
BM, BM,
6AvG
BM, BMv,
&AvG
BM, BM,
. ​&AvG

BM ,.​ ​BM,,
Case Case Case Case Case Case 2 3 ​1 ​- - - ​4 ​- ​5 ​- ​6 ​- ​Vertical Vertical Vertical Vertical Vertical Vertical Loads Loads Loads
Loads Loads Loads Only
+ ​+ + ​+ ​+ ​Column Column Column Column Column Base Base Base Base BaSe Moments Moments Moments Moments
Moments ​+ + ​+ ​+ ​Seismic Seismic Seismic Seismic ​B M , i​ n
​ N&~
Ratio (​ W.R. ​T K
Case-l) ​Min. Mar
- ​296.7
0.81 ​-+ ​1.19 ​- ​299.7 ​- ​307.9
0.90 ​+ ​1 ​.I ​1 -287.7
0.75j1.27 ​- ​222.9 ​- ​376.3 ​- 3​ 02.9

0.83 ​+ ​1.16 ​- ​- 303.0

​ ​284.8
0.73 -​ + ​1.34 -316.8

-​ ​ ​-
0.72 ​+ ​1.28 ​- ​ 21 ​387.0 8.9 ​- 151.9
​ ​153.4

​ irection) Force(-x Direction) Force(+ ​y ​Direction) Force(- y Direction) ​Load

Force(+ ​x D
Average '6' Case
in mm

​ .76 ​- ​7.78 ​- ​6.44 ​- ​9.00 ​

- ​8.13 -8.13 ​- ​9.21 ​- ​6.88 ​- ​6.56 ​- ​9.71 ​- 7
- - 6​ .71 8.84 ​- ​9.26
1.80 ​+ ​1.18 ​- ​9.00 ​- ​- 10.89
​ ​7.40 -​ ​169.9
 - ​101.6 0.67 ​+ ​1.13 0.67 ​+ ​1.24 ​
- - ​9.09 9.1 ​1

- - ​188.7 119.2 -​ ​10.61 ​- ​301 ​.I ​0.95 ​+ ​1.05 ​- ​10.52 -​ ​10.34

- - ​298.3 292.5
0.88 ​+ ​1.16 ​- ​11.17
- ​259.6

0.86 ​+ ​1.03 ​- ​- 10.98

​ ​10.07

-311.4 ​- ​285.5 -​ ​- q28

​ ​6.40
-80.9 ​- ​79.5
0.85j1.23
- ​6.95
- ​72.8

0.84 ​+ ​1.43 ​- -​ ​7.70 ​7.50

-150.9 ​- ​94.3
0.79j1.87
- ​5.31

- ​63.7 -5.19 ​ - ​ 4.3 7.53 5.23

-​ -​ ​ ​1 -​ 1​ 04.7 ​- 1​ 04.1
- ​96.5 ​. ​- ​141.8
0.71 ​-+ 1​ .45

0.78 ​-+ ​1.23 0.52 ​+ ​1.47 ​- 6​ .76 -3.71

​

- ​ 153.6 ​54.8 BM, ​in KNm/m

-​ ​

- ​100.5 ​- ​101.3 ​- ​98.8 ​- ​107.4 ​ - ​112.0 90.5

-​ ​ ​- ​92.7 ​- ​93.5 ​- ​67.7 ​- ​123.8 ​- -​ ​11 ​75.9 ​1.1 -​ ​- - 60.2
​ ​58.7 77.5 -​ ​40.8 ​- ​68.7 ​- ​48.9 ​-

230.4 ​- ​- 234.2
​ ​230.0 -​ ​256.1 -201.9 ​- ​266.4 -91.3 ​- -91.9
​ ​114.2 -​ ​- - 130.8
​ ​107.3 76.3 -​ ​48.5 ​- ​- - 48.6
​ ​45.5 49.1 .​ ​- ​- 59.6
​ ​37.7

​ in. Max. -​ ​6.56 ​+ ​- 9​ .71

Variation M
- ​90.5 ​+ ​- 1​ 12.0 -222.9+-376.3
​ .00
- ​6.44 ​- 9

- -​ ​21 67.7
​ ​ ​+ -​ ​- ​387.0 123.8
​8.9 + ​ -7.40 ​+- ​10.89
- ​40.8 ​+ ​-68.7 ​- ​101.6 ​-3- ​188.7
- ​10.07 ​+ ​- ​11.77
- ​201.9 ​+ ​-266.4 ​- ​259.6 ​+ ​- ​31 ​I ​.4
-5.31+-7.70
- ​76.3 + ​- ​130.8 -63.7+-150.9
​ .53
- ​3.71 ​+ ​- 7
- - ​37.7 54.8 ​+ ​+ ​- - ​59.6
153.6
 55
STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS ​

7.2.2 Discussion of Results ​Table 7.7 shows that there is large variation in the bending moments in ​X
and Ydirections both, when effect of vertical horizontal load alone. loads. In is considered. case of plate

The element values No. vary 18 ​B.M ​over in a ​.- ​range direction which for lies vertical on both loads
sides alone of the is 79.0 value KNmIm. of the

​ i​ ​
f1
' ​1 ​t​
For different load case, this value increases in a range of 79.0 to 227.1 K N d m The minimum

increased values ​
have a ratio of 1.32 and maximum of 2.78 ​. ​Similar values for plate element No. 28 are 0.98

​
to 1.64 for ​BM, ​and 0.87 to ​1.32 ​for ​BM,,. Table 7.8 indicates variation but in this case the increase is

comparatively low as compared to Example ​1. ​This is on the expected lines as the raft for the central core is

symmetrical about one axis and lesser variation is expected.

I

;​7.2.3 Conclusion

The study shows that values of bending moments in a raft of particular thckness and for the same value
of so11 modulus will vary considerably when the effect of horizontal loads on the super-structure as
transmitted to the foundation is considered. The increase could be several times. The usual practice of
assuming that ​25 ​to 50% increase can be permitted when seismic force are considered is not sufficient to
cover these excessive values and the structural designs, if carried out neglecting the effect of horizontal
loads may not be safe. Similar results are expected if raft of the other thickness and soil modulus are
considered

z3; ​Study ​3: ​Comparison ​with ​Conventional ​Rigid

Methods

7.3.1 Details of Conventional Method: Combined Footing Approach I​ n the conventional rigid combined footing
approach (explained in detail by TENG'), the raft is analysed using simple statics without any
consideration of the elastic properties of the raft and the soil. Here the raft is analy sed as a large beam
member independently in both the directions. The row of column loads perpendicular to the length of the
beam are coupled together in single column load. Then for these column loads acting on the beam, the
upward soil pressure is calculated and the moments and the shears at any section is determined by simple
statics. Hence, the moment per unit width of the raft is determined by dividing the moment values by
corresponding width of the section. In the situation when the width of the raft is changing both the
values, ​i.e., c​ orresponding to lesser width and the bigger width are determined and the section designed
 to allow for this sudden change which in fact means concentration of stresses at this junction. Then the
same analysis is repeated for the other direction. ​('Y d​ irection) considering raft as a whole. These in
general are the lower bound values.

The above analysis gives the total bending moments and shears across the whole raft. But the distribution of this
moment and shear along this section, ​i.e., ​the width, is a problem of highly indeterminate nature and the
average moment obtained above may not exactly indicate the sign and the magnitude of the bending
moment at a particular location. In order to obtain some idea ​as ​to the upper bound values the raft is
divided into strips bounded on the centre line of the column bays in each direction. Each of these strips is
then analysed -as independent combined footing by simple statics. Using the column loads on each strip
the soil pressure under each strip is determined without reference to the planar distribution determined for
the raft as a whole. The eccentricity of the load and the pressure distribution below the raft which is
considered to be linearly varying are taken into account in this analysis. This method undoubtedly gives
very high stresses because it ignores the two way action of the raft and transfer of stresses from one strip
to another strip. Therefore, certain arbitrary reduction in values (30% in this case) is made.
STRIP ​1 ​,2690 STRIP
​ ​ ​! I​ ​STRIP ​3 3100 ​, ​I ​STRIP ​4 3660
​4130 2
1749 1674 ​! ​935
​ TRlP ​7 I​ ​
STRlP ​5 ​STRlP ​6 S

300-0 600tl
​ ​
​ ​350x350 ​RAFT SLAB THICKNESS STRIPS
- 1- ​ ​@+ ​@ ​COMBINED

EXAMPLE- ​1 ​1000 ​MM ​o v​ , ​m @+

​ ​(@ ​COMBINED ​@ ​
0 C​ OMBINED ​ Fig. ​7.9 (ALL
​

DIMENSIONS ​ARE ​IN MM) ​5 ​ O​

E ​ D ​5 ​v, ​V)

STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS ​r ​

EXAMPLE ​2
 ' ​STIP.5 ​' ​STRIP G'STRP.7 'STRP.~ STRP.9 ​I (​ ​AL ​.L ​DIMENSIONS ​ARE

Fig. 7.10

7.3.2 Examples ​Selected ​The raft in Examples ​1 ​and ​2 ​has been divided into strips, and this division ​is ​indicated
in Figs. 7.9 and 7.10. While examining these strips which are made on the centre line of column bays, it would be
seen that there ​are ​some strips which have columns only at one end of the strip or there is ​a ​big cantilever beyond
a particular column. Here a judicious examination is required to be carried out by the designer to see the
likelihood of particular strip acting individually or in combination with the adjoining strip. Keeping in view the
location of columns in Example 1​ , ​Strip No. 1 and 2​ .5 ​and ​6,8 ​and 9 were considered to be combined into one
each and
58 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS

also analyzed. Similarly, in Example 2, strip Nos. ​6, ​7 and ​8 ​were combined and analysed as a combined
footing. -The values of the bending moments were calculated in all these strips at the column point and
mid point of the span and also at the points where the raft changes width.
 Both these rafts have already been considered in Study Nos. 1 and 2. Study No. 1 was carried
out in 2 parts. In the first part thickness was kept constant and values of modulus of sub-grade reaction
were changed. Ratio of maximum and minimum value of moments in ​X ​and ​Y ​direction were worked out
​ as kept constant and values of thicknesses
for each element and tabulated. In the second part value of ​Ks w
were varied. Similar values as above were calculated and tabulated.

In Study 2 (refer ljara 7.2 above), the effect of horizontal loads was considered for single thickness of ​, 1​
metre of raft slab and one value of modulus six loads cases were considered for the purpose of comparison and the
variation in bending moments in the raft due to earthquake loads were worked out. If I​ ​acomplete picture of the
variation in the values of moments in the raft was to be obtained, all permutations and combinations of the raft
thickness, values of soil modulus and seismic forces is required to be considered. This, however, would have
involved much more computational efforts. Since the objective have been only to get an idea of variation, the cases
considered have been restricted in the manner explained above.

In fact, it is difficult to say as to what thickness and what value of ​Ks ​out of the four values
would be adopted by any designer, he being free to adopt any of them but the effect of seismic forces
would be applicable to all of them. Therefore, percentage increase or decrease in the values of moments
due to earthquake effect has been worked out and then applied to maximum values of moments for any
combination of thickness and ​K,. v​ alues of ​M, ​and ​My t​ hus obtained are expected to be values likely to
be met with if analysis is carried out for all combinations.

It would be noted that the values of bending moments obtained in finite element analysis are
for each element and represent the values of the bending moments at the centre of that element in both the
directions. The values obtained in the conventional rigid method are at the sections along the length or
width of the raft. For comparing both, some method is to be evolved which could bring these to a common
base. After careful consideration, it was decided that comparison may be made of the maximum positive,
maximum negative, minimum positive, minimum negative and the average of all positive and negative
values determined by each method. This comparison would reasonably indicate extent of variation in the
values of moments determined by the two methods. These values have been tabulated in Tables 7.9 and
7.10 for Examples 1 and 2, respectively.

7.3.3 ​Discussion of Results

Table 7.9 for Example 1 shows that the values of bending moments in the rigid raft analysis are always higher than
those which could be possible for any combination of raft thickness and modulus of sub-grade reaction and the
earthquake effect. The extent of difference between the 2 values varies but in general these values are L​ ​=I ​higher than
those likely to be obtained by each analysis.

In Table 7.10 relating to Example 2 also the values of moments in the rigid method analysis are
higher than the values expected in any combination of raft thickness, soil modulus and considering
earthquake effects except in case of ​M y ​maximum and ​My ​average. This indicates that there may be
sections wherein values of moments determined by rigid raft, approach may be lower than those expected
in the actual structure under seismic consideration.
Table 7.9
Example 1 :​

(- ​1 2 3 4 5 6 7 8
1 ​Mxx (+ve)
Max.
162.7
409.1
151.4%
187.6
471.63
914.01
605.56
423.89
M, (-ve)
My, (+ve) KNm
Min.
Avg.
0.142
- ​18.27
0.98
- ​19.80
590%
8.37%
0.106
-36.24
0.73 -39.27 126.51 ​-
69.74
-413.6
48.82
-289.5
M,, ​(- ​ve) KNm
Min.
- ​1.28
- ​0.235

-81.64%
-​

-0.13
-0.0238

- ​ 6.0
-​
-4.2
KNm
Min.
3.34
19.42
481.4%
1.544
8.977
82.21
52.046
36.43
KNm
ve Implies Tension at Bottom)
Min.
Finite Element Analysis
- ​14.39 Considering Vertical Loadr on1 ​y ​for ​Ks ​= 10996 KN/m" andd= 1.0m Finite Element Analysis
- ​0.50 considering Seismic Loads also for ​Ks =​ 10996 K N / ~ ~ and ​d = 1​ .0m Percentage Increase of
-96.5% 2 over 1 Finite Element Analysis
- ​0.92 considering Vertical Loads only for any Combination of ​Ks a​ nd ​d ​Expected Values for Finite
-0.032 Element Analysis for any Combination of ​Ks ​and ​d ​if Seismic Loads also were Considered i.e. 4 ​x ​3 Rigid Raft
Analysis
- ​19.91 considering the Raft as a whole Rigid Raft Analysis
- ​1.0 considering Individual Strips Values for Rigid Raft Analysis for Strips Decreased by 30 ​% ​i.e 7 ​x ​0.7
Avg.
Max.
Max.
Avg.
72.0
-54.67
205.4
89.0
174.38
-40.78
487.9
167.47
142.2%
-25.4%
137.54%
88.17%
76.58
- ​59.53
233.9
112.78
185.47
-44.4 555.6
212.2
317.68

-
536.06
266.5
532.52
- ​1011.3
574.6
346.0
372.76
-707.9
402.2
242.2
Max.
Avg.
- ​16.13
- ​15.26
- ​37.08
- ​15.39
-​129.8%

0.851 % ​
- ​24.7
- ​15.26
- ​56.8
- ​15.39
- ​19.91
- ​19.91
-293.7
- ​173.30
-205.6

Table Example 1 ​: ​
1​ Mxx (+ve) KNm

I​
7.10 ​M, ​(-ve) ​KNm ​ My, (+ve) KNm ​ --

M, ​(- ​ve) ​KNm

(-

1 ​2 3 4 5 6 7 8 -​ ​ve I​ mplies Tension at Bottom)

Min.
Finite Element Analysis
-48.46 considering Vertical Loads also for ​Ks ​= 10996KN/m3 andd= 1.0m FiniteElementAnalysis
-33.89 considering Seismic Loads only for K ​ s ​= 10996 KN/rn3 andd= ].OM Percentage Increase of
- ​30.07% 2 over 1 Finite Element Analysis
- ​1.629 considering Vertical Loads only for any Combination ​KF ​and d Expected Values for Finite
- ​1.139 Element Analysis for any combination of ​Ks ​and ​d ​if Seismic Loads also were considered i.e 4 ​x ​3 Rigid Raft
Analysis
-221.99 considering the Raft as a whole Rigid Raft Analysis
- ​88.0 considering Individual Strips Values for R~gid Raft
-61.6 Analysis for Strips decreased by 30 ​% ​i.e 7 ​x ​0.7
Max. ​
- - ​- ​- -​ ​154.9
400.02
280.014
Min. ​Avg. ​Max.
Avg.
Max. Min. Avg. Max.
Min.

Avg. ​
- - -​ ​239.9
- ​125.54

- ​- ​- -​ 5​ 42.9
- ​80.95
- ​194.14

- - -280.2
​
- ​133.9

- - ​- -​ ​683.1
-45.88
- ​193.17

- - 16.79
​ ​%
6.65 ​%

- ​- - 25.82 ​%
​
- ​43.32%
0.50 ​%

- - -​ ​356.6
- ​192.4

- ​- ​- -​ ​644.10
- ​1.67
- ​341.06

- -​ ​-416.5
154.9

154.9
-631.1
300.24
400.02
- ​191 ​.O
210.170 280.014
- ​1337.7

-205.2 ​- - -416.4
​ ​78.20

14.4
- ​1030.86
21 1.47
7.6
- ​721.6

_​ 148.03

5.321 ​
L -​ -​ ​ 810.4

- ​0.95
- ​342.76
51.7
-87.75
- ​12.41
- ​37.29
89.31 1
- ​314.04
- ​6.0
- ​283.43
62.517

J​ - ​254.83
1​ -4.20

-198.401

Y' ​I ​9 ​% ​ V, ​D​ 2 ​ 3 ​5 ​z ​n ​v, ​

0​ V,

STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS ​61

T​ Dir. of Moments ​Table

7.1 ​1
Strip NO.
(KN- d m )

​ ​Example I​ 2​ 3 4 7 1​ ​+2 5+6

M, (​ Max.) 1
Raft as ​a ​whole
Myy (​ Max.)

10 I 11 ​12 13
8+9
Raft as a whole
Table ​7.12
Dir. of Moments
M, (​ Max.) Example 2
My (​ Max.)
I
Rigid Raft Method
(Continuous Beam) ​wL2/10 (​ KN- d m )
1011.3
223.25 503.37
260.19 505.07
253.03 373.0
281.09 605.56
207.96 529.86
260.55 410.58
281.09
*914.01
137.0
*Smaller Width of Section Considered
386.08
181.18 566.0
176.25 574.6
 248.65 118.0
44.678
293.7
156.35
536.06
175.7
Strip No.
Rigid Raft Method
(Continuous Beam) ​wL2/10 (​ KN -​ ​dm)
(KN -​ ​dm)
5 578.4
286.54 ​6 839.0
​
392.2 ​7 650.0
​
217.86 ​8 1911.0
​
392.2 ​9
403.9
370.34 6+7+8
868.0
3922
Raft as a ​whole
631.10
392.2
1 ​21 1.50

156.7 ​2 139.0
​
186.0 ​3 364.0
​
116.6 ​4
143 .O
11 6.6
Raft ​as ​a whole
87.75
65.10

RAFT FOUNDATIONS-DESIGN AND ANALYSIS ​

I​ ​ the other method, ​i.e.,

7.3.4 Znverted Floor Method In

Inverted floor also called beam analogy method by some designers, the raft is treated as an inverted floor.
Analysis of each strip bounded by the column bay centre line is carried out as a continuous beam in both the
directions as compared to analysis by simple statics in the combined footings approach. Though not mentioned
specifically in any text book there, has also been a practice in design offices not to adopt the values for design
purposes lower than WL2/10 at any of the sections of this case. For Examples 1 and 2 the maximum value of
bending moments determined by combined footing approach and inverted floor approach in these directions for
each strip and raft as a whole have been tabulated in Tables 7.11 and 7.12. A study of these tables would indicate
that bending moments determined by ~ ~ ~ 1 1 0 are always lower than those determined by combined footings
approach in general.
This study also shows that WL2/10 values shown in Tables 7.11 and 7.12 ​are ​not always higher
than the expected values of moments in row No. ​5 ​of each Tables 7.9 and 7.10. Situations may, therefore,
arise wherein continuous beam analogy values may not lead to safe designs. This is in line with the
recommendations of Mr. M ~.~omlinson'O.

7.3.5 Conclusions ​This study has shown that values of moments determined by conventional
combined footings approach are higher than the values likely to be obtained after considering the variation
of changes in rigidity of the raft, soil pressure distribution, values of modulus sub-grade reaction and
earthquake forces. This may, however, be not true in case of inverted floor system even where using
wL2/10 as lower bound values.

7.4 ​Study ​4. ​Another ​Office ​Building

This is an office building which has six blocks. Two central blocks are eight-storeyed with basement while
the side blocks are two storeyed with basements. Each central block has a lift machine room, stair case,
mumty and water tanks above the general terrace level. The two central blocks are separated from the two
storeyed blocks. In this study, these two central blocks have been considered. These blocks are separated
by a separation joint from each other but the raft foundation is continuous. Through the building looks
symmetrical in plan but due to loadings and position of few columns this is not so. A plan of the raft is
shown in Fig. ​ZS?

7.4.1 Example Details T ​ he design of the raft has been done utilising beam and slab constructl'on with
beams projecting upward. For this purpose raft was considered to consist of beams only in both
longitudinal and transverse directions. Each of these beams was analysed as a beam on an elastic
foundation utilising value of modulus of sub grade reaction, calculated from soil properties. The loads
considered were only vertical loads. However, taking into account the assumptions made, design bending
moment values obtained from the analysis were subjected to a minimum value of w12 /​ ​10 where W is the

load carried by the beam per unit length considering the raft as an inverted floor. ​The superstructure has

been analysed on a ​three ​dimensional analysis programme on computer considering the​ earthquake effects
and the values oi column loads, base moment, increase and decrease in column loads due to the
earth-quake obtained. These values have been utilised to study the variation in the values of shear forces,
bending moments, contact pressures and deflections of the raft when earth-quake effects are taken into
account against only vertical loads considered in the first analysis.
The analysis for raft has been done utilising SAP ​5 ​Computer Programme based on finite
elements method. The ​finite ​elements chosen are the beam elements of the beam slab foundation system.
The rigidity of the raft
SLAB THICKNESS =400mm ​BEAMS ALONG ​X-X ​=I000 XI200 mm BEAMS ALONG ​Y-Y ​=I000 X800 mm

(ALL ​DIMENSIONS ​ARE ​IN MM)

Fig. ​7.1 ​1
64 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS
slab is not considered. Each beam element is defined by two nodes, each node is also the point of application of
concentrated loads (vertical load forces and moments). The soil has been idealised as a spring under each node.
The spring constant is a function of soil sub grade modulus and the area under the node.
7.4.2 Comparison of Results
Six different cases of loading i.e., vertical loads only, vertical loads plus base moments due to vertical loads
only, vertical loads and their values for seismic loads acting in positive ​X ​direction, negative ​X ​direction,
positive ​Y ​direction and negative ​Y ​direction have been considered. For comparison of results, some ​beam
elements shown in Fig. 7​ .11 ​have been selected. The maximum (larger of the values on the two nodes) shear
force, bending moment, contact pressure and deflection in ​Z ​direction have been tabulated along with the range
of these values. All these values ​are ​shown in Table 7.13. Also shown are the values obtained in simplified
flexible design explained in 7.4.1 above. Also shown in this table are ​the ​ratios these values have with the value
considering vertical load alone.
7.4.3 Discussions of Results
A s​ tudy of the table would show that:
(a) Shear varies force from 0.07 for beam ​KN ​to element 467.5 ​KN ​No. giving 40, for a ratio vertical of 0.003
load alone ​to ​19.26. is 24.27 Maximum ​KN. ​For increase other in load shear cases, force it
is seen in case of beam element No. ​44 ​where ratio goes as high as 92.90.
Table ​7.13
Node Nos. a​ nd E
​ lement No.

2,3 ​2
10, 11
16
24,25
40

Load ​Case ​6,, ​

-​ 1 ​-45.73 2 -47.65
​ ​3 -47.67
​ ​4 -47.526,
​ ​5 -​ ​51.21 ​6 -42.98
​

1 -​ ​38.76 ​2 3 4 5 -​ ​- - - 40.90
​ ​40.45 39.90 39.50 6 -40.12
​ ​1 ​2 3​ -​ ​- - 53.93
​ ​58.22 58.07 4 -59.56
​ ​5 ​- ​57.68 6
- ​57.12
​ in. ​Ma*.
Ratio (W.R.T Case ​I) M
0.16 ​+ 0​ 0.81 3.12+ 11.52 0.94+ 01.12 0.94+01.12
1.24 0.10 1.01 1.02 ​+ + + + ​2.88 1.00 1.05 1.06
0.003 0.54 1.02 1.06 ​+ + ​+ + ​19.26 22.02 01.10 01.10 ​V,, ​KN Conventional
​
Design Values.

30.20
1319.00 07.91
1359.00 24.36
130.01 195.0 15.07
-125.00 04.70 04.81
04.79
1797.00 09.74
1851.30 13.80 05.96

​ ​195.0 10.46
130.0 ​- ​125.00 1
24.27
1797.00 05.58
1851.30 450.20 460.50

​ ​195.0 00.07 19.00

130.0 ​- 1​ 25.00 1

BM,, Range of ​KNm

Variations
Min. M
​ ax
- ​21.22
4.70 ​+ 3​ 0.20 66.11 ​- ​172.25 118.58
116.11 j ​+ 2​ 4 138.32 4 . 5 0
-42.98+-51.21 ​- ​172.25 244.50

-550.60 ​- ​- 529.40
​ ​549.08 -​ ​513.50

-513.50 ​- 104.65
​ ​38.76 ​4.79 ​+ + + ​+ -550.60
​ ​-40.90 ​110.37 13.80
- ​527.25 09.40-535.17108.33
102.20 159.90 2250.00 2080.83
2080.83 -53.95 145.63 0.07 ​+ + ​+ + ​2250.00 -59.56
467.50
160.54
240.58

​
BP,, KNm/ ​mZ
123.70 128.71 125.51 89.00128.34 138.32 116.11
104.65 110.37 109.23 '​ ​105.75 106.66
145.63 157.21 148.89 160.54 54.99.555.73
154.22

V, ​BM, BP, V,, ​BM, ​BP, ​6,

V,, ​BM, ​BP, ​6,

6 ​5
STUDIES CARRIED OUT ON EFFECT OF VARIOUS PARAMETERS ​
Table 7.13 (Continued)
​ ase Moments, '​ ​Case 3 ​: ​Vertical Loads ​+ S​ eismic Loads (Seismic
Case 1 ​: ​Vertical Loads only, Case 2 ​: ​Vertical Load ​+ B
​ irection), Case 4 ​: ​Seismic Force in ​-X ​Direction, Case 5 ​: ​Vertical Loads ​+ S​ eismic Loads(seismic Force in
Force in ​+X D
+Y ​Direction), Case 6 ​: ​Seismic Force in ​-Y ​Direction. All Loads Correspond to un Factored Condition.
The variation is more in beams which are heavily loaded in the nodes and the beams which is discontinued at
both nodes. (b) Bending moment for Beam element case, it varies from ​- ​2080.83 KNM No. to ​+ 4​ 0 2250.0 for
vertical KNM. load Minimum alone is positive 102.2 ​KNM. ​Bending For moments other load in this.element is
54.99 giving a ratio of 0.54 to 22.02. Maximum increase is also occurring in this beam. These are the beam which
are heavily loaded at the nodes. (c) Bearing pressure due to vertical loads alone in case of beam No. 2 is 123.7 K
N / with ~ ~ maximum of 138.22 giving a ratio 0.94 to 1.12. Maximum increase in bearing pressure is also in this
beam. The safe bearing pressure recommended by soil consultant is 130 K N / ~ ' . (d) Deflection due to vertical
load alone in case of beam No. 2 is 45.73 mm. For other load cases, it varies from -​ ​42.98 to 5 1.2 1 mm giving a
ratio of 0.94 to 1.12. Maximum increase is also in this beam.
535.73
0.67 ​+ ​03.91
59.42 ​+ ​346.58
805.00 740.57
1.36 ​+ ​05.20
-1284.17j2788.33
723.30 -1284.17
1.02 +​ ​01.10
145.63 ​+ ​160.54
130.0 1 195.0 2788.83
​ ​01 .I0
1.06 +
- ​53.93 ​+ ​59.56
- ​125.00 727.75 763.33 235.60 272.53
0.29 ​+ ​2.43
24.03 ​+ ​198.92
1026.00 0.36 j 2 . 5 9
-84.67 j611.08
707.30 292.67 233.67
0.92 ​+ ​1.09
129.43 ​+ ​152.66
130.0 1 195.0 0.92 ​+ ​1.09
- ​47.93 ​+ ​- ​56.53
- ​125.00 ​- ​84.67 611.08
-536.47
- ​92.00
0.01 j 0 . 3 5 0.17 ​+ ​0.46
1.45 ​+ ​145.80 ​- ​92.00 ​+ ​- 5​ 36.47
1105.30 762.00 -214.67
1.03 ​+ ​1.10
105.62 ​+ ​116.20

130.01 195.0 28.60-112.50116.21 ​- ​245.92

1.03+1.14-42.25+-48.15
-125.00
- ​200.25

01.63
V, 4.18 + ​ ​92.90 06.82
BM,, 2.23 ​+ ​14.83 29.69 ​+ ​438.92 603.30 09.21 20.23 25,26
1 88.67
145.63

- ​53.93 ​2 137.27
​
157.21

- ​58.22 4 1 ​3 59.42
​
148.89

- ​59.20 ​4 346.58
​
160.54

- ​59.56 ​5 153.00
​
155.73

-57.68 ​6 130.70
​
154.22

- ​57.11 ​4.8 1 81.87

​
140.37
- ​51.97 ​2 89.27
​
143.71

-53.22 ​7 3 98.42
​
152.66

- ​56.53 ​4 76.98
​
129.63

-47.93 ​5 24.03
​
148.68

- ​55.07 ​6 198.92
​
138.95
-51.46

14, 18

1 145.80
​
105.62

-42.25 ​2 01.45
​
115.01
-46.01 26

3 28.06
​
108.46

-43.38 ​4 -48.156,,,
​ ​5 50.62
​
113.97 ​- ​45.58 ​6
50.98
110.75

​44.30
- ​-

23.29
29.60 65.86 44
79.26 90.58

BP,, ​
​ ​+ ​01.08 139.57 ​+ ​153.69 130.01 195.0 ​0.98 ​+ ​01.08 ​- ​51.66 ​+ ​56.88 ​- ​125.00 137.75
amax 0.98 ​
420.67 151.42
438.92

V, ​BM,

I​
​ ​r ​I ​1 V,
amax BP,, ​ ​BM,,, ​amax BP,,,
​ V,, ​BM,,, ​ BP,,
I​

, ​1 ​141.97

-52.55 ​2 148.54
​
- ​54.98 ​3 139.57
​
- ​51.66 ​4 153.20
​
 - ​56.71 ​5 148.96
​
- ​55.14 ​6
153.69
- ​56.88
RAFT ​FOUNDATIONS-DESIGN AND
beanng pressure ​and ​settlement. This is what was expected. This is a be
ANALYSIS
1.2 ​m ​x ​1 m, ​0.8 ​m ​x ​1 ​m with slab thickness of ​0.4 ​m) which is compar
a raft of uniform thickness. (b) The design done by the method actually
(e) When shear force values of any beam for any of the load case is compared
point, confirmingwith .F. ​value in
the​Sstatement
method -

adopted for actual design these are always lower. ​(f) ​When bending moment values for any of
the load are compared with the values on the method adopted for design (keeping in view the higher
permissible stresses in working stress method and lower load factorthatin limit
evenstate
inverted
design
floor
permitted
approach could lead to unsafe design. Sim
while considering seismic effect) it is seen that these values are lower. Except in case of element
No. ​41. that while detailing the reinforcement in members like raft, freq

​ onclusions
7.4.4 C view of various restrictions like minimum development length, ea
These results show
that: when the resisting moment of the section as actually provided w
found to ​be ​safe.
(a) While variation in bending moment and shear force values are very large, this is not so for

STUDIES CARRIED OUT ON

ANALYSIS AND DESIGN OF PILED
RAFTS

8.1 ​Design Procedures being Used

Rafts supported on plles are being increasingly used for multi-storeyed buildings with basements in poor
soils with high water table conditions. The piles are necessary to transmit the super-structure loads to a
deeper competent soil strata and the raft is required to transmit the column/wall loads evenly to the piles
and also to resist the buoyancy forces of the ground water. Piles are sometimes used to decrease the
settlement of the raft. The raft as a solid medium integrated with the retaining walls with necessary water
proofing layer also serves as a water proofing medium.
 The analysis of piled raft is a complex problem even more than that of a soil supported raft as too many parameters
influence the behaviour of the system. Very little is known about the exact behaviour of piled raft
foundations in service. The problem is to be understood by considering the composite behaviour of the
entire system, ​viz., ​super-structure, sub-structure, raft, piles and the soil medium. These factors influence
sharing of load berween piles and raft, between piles themselves and consequently the settlements, shears
and moments in the raft.
For design of piled raft, different practices are followed by various designers. Most simple method followed is the
conventional rigid approach, wherein the raft is assumed to be rigid. Piles are uniformly distributed
throughout the raft and a planar distribution of pressure is considered on the raft due to the piles. As a
variation of this method, some designers ​try ​to concentrate more pile under the heavily loaded columns as
compared to lightly loaded columns assuming that it would give a better uniform distribution on the piles.
In another approach individual pile caps below each column are provided and are connected either by a
slab of the thickness equal to that of the pile cap or of a lesser thickness, neglecting the effect of one pile
cap on the other. Where computer facilities are available, some designers use the concept of ​beam ​on
elastic foundation. Here again various methods are available. Some designers assume a uniform
distribution of load on piles replacing the piles by a soil medium having a hypothetical bearing capacity. ​K
value corresponding to this bearing capacity is selected and used for analysis. ​As ​further improvement to
this method, raft is taken as a plate supported on springs. The properties of the spring are determined
depending upon the type and elastic properties of the piles neglecting the effect of one pile on the other
and different soil layer on each other. Effect
​ AFT FOUNDATIONS-DESIGN AND ANALYSIS
68 R
of one pile on another and different soil layers can also be considered. This is done in computer programmes
wherein the spring replacing piles are coupled both horizontally and vertically meaning that the deflection of any
spring is affected by adjacent springs. The mat super-structure inter-action is usually neglected. None of these
methods take into account the effect of the soil foundation interaction except to the limited extend mentioned in
each of the method above. Sharing of loads between piles and raft soil system has also been suggested n d followed
by some. However, there are no practical methods available for working out this extent of sharing. ~oulos" has
suggested a method, but it is applicable for piles less than 4​ 0 ​in number and makes assumptions which are
controversial.
There has also been a practice of designing the raft foundation for vertical loads alone excluding the effect of
column base moments or the effect of horizontal load, ​ie., ​earthquake and wind. In all these analysis, thickness of
the raft and the safe load carrying capacity of the pile is decided before hand. In beam on elastic foundation
concept the rigidity of the raft plays a very important role and, therefore, the presumed thickness of the raft affects
final shears and bending moments in the raft and the loads on the piles. No effort is, however, generally made to
quantify this effect and optimise the thickness of the raft.
In this study, effect of rigidity of the raft, i​ .e., t​ hickness selected, the effect of superstructure rigidity. variation in
column loads and base moments due to earthquake and the type of piles, on pile loads and raft moments, has been
calculated and studied.
8.2 ​Example Selected
The problem is studied by considering a real building with piled raft.
The side blocks are two storeyed with an eight storeyed central block. The entire building has a basement and has a
T-shaped plan. The basement is to accommodate substation, air conditioning plant, stores etc. The building is
proposed to be constructed in a pond by retaining the water body.
The foundation is proposed to be provided with piled raft with overall dimensions ​as ​shown in Fig. 8​ .1. ​There
would be 7​ 08 ​bore cast-in-situ piles. The piles ​are ​predominantly friction piles of 5​ 0 ​cm diameter and 21m length.
The thickness of raft considered is ​1.20 ​m.
8.3 Soil ​Data
The subsoil below the bed levcl of the pond consists of successive layers of silty claylclayey silt down to ​30 ​m
depth and beyond. The sequence of soil stratification is summarised in Table ​8.1.
Table 8.1 ​Sequence of Soil Stratification
Stratwn Description

I ​I1 ​111 ​IV V VI

Thickness (​ m)
Average SPT Range Average
Values (N)
Fill Material
2.4 ​- 4​ .5
​ rey silty claylclayey sandy silt
3.5 ​- G
4.0 ​- 7​ .0

5.5 ​1 Greyldark
​ grey silty clay ​with ​organic matterldecayed wood
5.0 ​- 6​ .0

5.5 ​3 Mottled
​ greyhluish grey silty clay ​with ​kankar
3.0 ​- 4​ .0

3.5 ​11 ​Yellowish greylmottled grey silty claylclayey silt

2.5 ​- 4​ .5

3.5 ​17 Mottled

​ grey clayey siltlclayey sandy silt with laminations
7.5 ​- 9​ .5
8.5

25
Fig. ​8.1 ​Plan ​of ​raft ​showing strips considered
8.4 ​Methods of Analysis Studied
In this study the following methods have been adopted ​:
(a) Conventional rigid method with simplified models as
(ii) continuous beam analogy or inverted floor
(i) Combined footing approach ​! ​
(b) Using finite element approach
' ​8.4.1 ​Conventional ​Rigid M
​ ethod with Simplified Models

8.4.1.1 ​Combined footing approach ​i ​The drastic simplification adopted in this model is that the closely spaced

piles, (spacing nearly equal to or less

​ than the ​d l ​thickness) can be approximated as a bed of equivalent soil strata.

The ​raft ​is analysed ​by ​: c​ onventional rigid approach using simple statics without any consideration for the elastic

properties of the raft and

​ the soil. Here the raft is idealised as a large beam member independently in both the
direction. The row of column loads perpendicular to the length of the beam are clubbed together as single column
load. Then for these known column loads acting on the beam the upward soil pressure is determined and the
moments and the shears at any section is determined by simple statics. Then the moment per unit width of the raft
 is

70 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS ​$ determined

​ by dividing the moment values by the
corresponding width of the section. Then analysis is repeated for the other direction ​(Y ​direction). The analysis is

also repeated for individual strips with relevant column ​load. This approach adopted by the early designers tends to

give high values of moments with long rafts. Assumption that such a long raft acts as a rigid member is the main

cause of such high values of moments obtained by this method. The plan configuration of the raft with the location

of the strips are shown in Fig. ​8.1. ​f ​8 f​ ​I-. g​ ​-; ​

'​ ' The
​ values of moments obtained by this method are given

in Table ​8.2.
8.4.1.2 Continuous beam analogy: invertedjoor T ​ he other approach of the rigid method is to use the continuous
beam analogy by treating the raft as an inverted floor. First, the average soil pressure of the equivalent soil strata
.​
under the raft is determined by dividing the ​ total column loads by the plan area of the raft neglecting the effect of
eccentricity of loads and moments. Raft is considered as a whole in both the directions and idealised as a long
continuous beam with known intensity of upward pressure based on average soil pressure mentioned above. With
no consideration of actual column loads, the column points are only taken as rigid supports of the inverted
continuous beam. The moments and shears of the continuous beam is determined by method of moment
distribution or by any other standard method. As an approximation, the maximum positive and negative moments
(excepting the cantilevering portions at the ends) is taken as ​~ ~ ~ ​. ​The ​1 1 ​procedure ​0 ​is repeated also for
individual strips with their corresponding average upward, equivalent soil pressure. The values of moments
obtained by this approach are also indicated in Table ​8.2.
8.4.1.3 Cornpanson of results F ​ rom Table 8​ .2 ​it could be seen that. in case of rigid method with simplified models
of combined footing approach momentdshears determined by simple statics, the value of the bending moments
​ ​directions are quite large. If one attempts to proportion the raft with these values, the raft
obtained in ​X ​and Y
thickness shall be very large with heavy requirement of steel. However in case of continuous beam analogy
method, the value of bending moments obtained are comparatively lower. Many designers consider the former
value to be unrealistically high and these values reasonable and adopt them for the purposes of approximate
designs. The basis of this differentiation is, however, not known.

Table ​8.2 ​Raft Moments obtained by conventional Rigid Methods ​Sf ​No. ​ 1 ​Description
Raft ​as ​whole in ​X ​direction Raft as whole in ​Y ​direction ​X ​direction Strips ​A B ​ ​direction Strips ​1 ​2
​ CY
Combined footing approach (KNdm width)
As ​continuous beam analogy (KNm per metre width)
STUDIES CARRIED OUT ON ANALYSIS AND DESIGN ​OF ​PILED RAFTS ​71

8.4.2 Piled Raft Analysis Based on Finite Element Approach ​A more sophisticated method of analysis
models the complete system, ​vjz., t​ he super-structure, raft, piles and soil medium with appropriate finite
 element types and carry out the analysis by considering the interaction between thesecomponents. In such
analysis, the superstructure is modelled as a three dimensional space frame, the raft discretised as plate
bending elements, piles as compressible elastic axial elements. The supporting soil is treated as consisting
of different layers of homogeneous linear elastic material with corresponding elastic modulus determined
with reference to soil properties. Normally the soil medium is discretised into a number of rectangular
prism elements. This generalised approach requires enormous computational efforts, time consuming and
quite expensive and hence cannot be used in normal design practice.
However, simplified versions of finite element approach are commonly adopted with the use of computers. In this
study a general purpose three dimensional finite element package (SAP-IV Structural Analysis Package)
has been used. In the present case of piled raft, the raft has been modelled as plate bending elements and
the piles are modelled as axial elements. The piles being predominantly frictional piles, as recommended

by ~owles" and ​ en^' ​the axial stiffness of the pile element has been taken as ​EAIL, ​where ​E i​ s the
modulus of elasticity, ​A ​is area of cross section of the pile and ​L, i​ s the effective length of the pile. No
exclusive modelling of the soil medium has been done, although the confining effect of soil on the
frictional piles is considered by considering the effective length of the pile as half the length of the pile.
Similarly, no separate modelling of the superstructure has been done. However, its stiffness contribution on the
overall behaviour of the system has been approximately considered as discussed subsequently. The raft is
considered to be entirely supported on the piles and do not have any soil support underneath. This is
particularly true in the present case where the building is located in a pond.

8.5 ​Study ​of ​Parameters Influencing the Raft Behaviour

8.5.1 Effext of Raft Stiffness on the Pile Loads and Raft Moments ​In this study the effect of raft stiffness
is considered by increasing the ​E v​ alue of the raft instead of increasing the thickness of raft. This has been
adopted for computational convenience. This analysls has been camed out starting from single E ​ ​value
upto~a value of 50 times ​E, a​ t increments of 10E. This is done essentially to ​as- c​ ertain the effect of raft
stiffness on pile loads and raft moments.
The increase in moment values in the raft with hypothetical Increase in stiffness values along a Xdirection ​; ​strip
and ​Y ​strlp are graphically shown in F~gs. 8.2 and 8.3. A lo-, 20- and 50-fold increase in stiffness value of the
raft increases the raft moments upto 325. 450 and 6 ​ 00 ​per cent, respectively, at certain locations. Due to the
variation in raft rigidity the pile load variation is upto 75 per cent over the safe load. Considering the stiffening
effect of superstructure (​ i-e., v​ alue corresponding to ​3 ​E) t​ his pile load variation decreases to 50 per cent. With
the 10-fold increase in raft stiffness (which amounts to more than doubling the raft thickness), thls variation
reduces nearly to 10 per cent.
The effect of increase in the raft sbffness on the pile load variation along a strip is shown in Fig. 8.4. It is seen ​that

beyond a ten fold increase in raft stiffness the pile load variat~on is very small. This clearly shows ​' ​that raft

with assumed thickness is not rigid enough so as to distribute the loads on to the piles uniformly and
number of piles wlll be carrying loads upto 1.75 times the average load.

8.5.2 Effect of Superstructure and Retaining WaUs on Foundation Stiffness T ​ o obtain a realistic
assessment of foundation performance, it is essential to include the additional stiffening effect of the
structure above the raft level. In this case the stiffness effect of superstructure is included in the
RAFT FOUNDATIONS-DESIGN AND ANALYSIS
Element numbers

Fig. ​8.2 ​Effect of Raft Stiffness on ​Raft ​Moments Along X-direction Strip.

Element numbers

Fig. ​8.3 ​Effect of ​Raft ​Stiffness on ​Raft ​Moments Along Y-direction Strip.
STUDIES CARRIED OUT ON ANALYSIS AND DESIGN OF PILED
RAFTS
 Ekment numbers

Fig. ​8.4 ​Effect of Raft Stiffness on Pile Loads Along X-direction

Strip.

-60 ​-

Only rat t stiffness

Combined st ​i t ​f ness

1 ​a 5​

Element numbers

Fig. ​8.5 ​Effect of Superstructure and Retaining Wall on Raft Moments Along
X-direction Strip.
74 ​RAFT FOUNDATIONS-DESIGN AND ANALYSIS
raft analysis by adding the cumulative bending stiffness of the floor slabs to that of the raft stiffness and equivalent
raft thickness is obtained as recommended by ~ o o ~ e r ~ ~ . As per this procedure for this eight-storeyed
building, the bending stiffness contribution of super-structure is of the order of 20 per cent of the raft stiffness and
the original raft thickness of 1.2 m get enhanced to equivalent thickness of 1.28 m only.
Similarly, the stiffness effect of retaining walls is also included in the raft analysis by adding the bending stiffness
of retaining walls and the superstructure to that of the raft. In this case theequivalent bending stiffness of the raft
works out to 3 times and the raft thickness gets enhanced to an equivalent thickness of 1.73 m.
Due to the stiffening effect of these components, the raft moments tend to increase and the increase being in the
order of 20-70 per cent over the values obtained considering the stiffness of the raft alone. This effect has been
shown for chosen X direction and ​Y ​direction strips in Figs. 8.5 and 8.6. The variation between the loads carried
by the individual piles also tend to decrease due to this effect.
Beside the raft, the super-structure frame and the infill walls will also come in play in resisting the foundation
moments due to the composite action. However, this has not been considered in the present study as the effect is
very small.
8.5.3 Effect of Earthquake Loads and Moments W ​ hen subjected to earthquake forces (lateral loads), column
loads and base moments get affected causing increase in column load, on the one side, and corresponding decrease,
on the other. This causes a large eccentricity of loads with respect to the C.G of the raft. This in turn also results in
uneven loading on the piles. This behaviour will induce differential settlement causing additional moments in the
raft. Hence, it is necessary to investigate the effect of earthquake loading for the design of the raft.
In the present study, the vertical load and seismic analysis of the building was done using a three dimensional
analysis package, TABS. The loads and moments of the columns at the foundation level are determined for ​I
seismic loads applied in two orthogonal directions independently. The raft is analysed by finite element modelling
explained earlier for each of the above cases, and the critical values of loads and moments are obtained. The

I
stiffness of the raft is kept at its actual stiffness. ​

The effect of earthquake has been studied with respect to the variation of the moments in the raft and the pile loads.

1​
Analysis is made for the non-seismic condition (load factor ​= ​1.5) and the seismic condition (load ​ factor ​=

1.2) as per Limit State Method. ​ 1​ These analysis indicated that for Xdirection the raft moments in non-seismic

conditions are generally more '​ ​compared to the seismic condition except at few locations where the seismic
moments are more (Fig. 8.7). This behaviour is not surprising while the raft is continuous, super-structure is in
three blocks separated by the expansion joints. These three blocks behave independently in seismic condition and
their effect on foundations ​. with
​ earth quake in ​X ​direction is compensating rather than additive. However, for the
​ d​ irection the seismic moments are found to be more. This shows the raft moments
earthquake forces along the ​+ Y
in seismic conditions ​1​ cannot be ignored as they ​are ​likely to govern in certain situations (Fig. 8.8). The
​ pile

'​
loads tend to increase in the regions in the far end of the raft in the direction of the earthquake due ​ to the

overturning effect of lateral loads on the foundation. ​8.5.4 Effect of End Bearing and Friction Piles ​ 1; ​In
practice both types of piles, frictional or end bearing or a combination of both, may be encountered at site. The
behaviour of the pile raft will change due to change in the axial stiffness of the pile which depends upon its
effective length, the other factors remaining same. This will, in turn, alter the pile loads and raft moments.