Mahgoub, Mahgoub Osman (1976) Shear strength of prestressed concrete

beams without shear reinforcement. PhD thesis.

http://theses.gla.ac.uk/677/

Copyright and moral rights for this thesis are retained by the author

A copy can be downloaded for personal non-commercial research or

study, without prior permission or charge

This thesis cannot be reproduced or quoted extensively from without first

obtaining permission in writing from the Author

The content must not be changed in any way or sold commercially in any
format or medium without the formal permission of the Author

When referring to this work, full bibliographic details including the

author, title, awarding institution and date of the thesis must be given

Glasgow Theses Service

http://theses.gla.ac.uk/
[email protected]
 SHEAR STRENGTH OF PRESTRESSED CONCRETE BE91S

WITHOUT SHEAR REINFORCEMENT

by

MAHGOUB OSYAN MAHGOUB, B. Sc.,

being a thesis submitted for the degree of

Doctor of Philosophy in the University
of Glasgow .

Department of Civil Engineering 9th July, 19? 6-

0
 CONTEWTS
Page
Contents i
Acknowledgements iv
Summary v
Notation
Chapter 1
Introduction
Chapter 2
Historical Review
2.1 General 5
2.2 Shear in reinforced concrete beams 5
2.3 Shear-compression approach 12
2.4 Some code approaches to design of 15
beams for shear strength
2.5 Shear in prestressed concrete beams 19
2.6 Analytical approach using finite element 41
2.7 Concluding remarks 43
Chapter 3
Study of the Parameters Affecting the Shear
Strength of Prestressed Concrete Beams
and Criteria used in Predicting Failure of
Concrete.
3.1 Introduction 45
3.2 The geometric configuration of the cross- -46
section
3*3 Concrete strength 47
3.4 Prestressing force 48
3.5 The shear span 48
3.6 A failure criierion for concrete 49
3.7 Principal tensile stress and diagonal 52
tension cracking
3.8 Semi-empirical approach based on 53
dimensional analysis.
Chapter 4
Experimental work
4.1 Test specimens 57
4.2 Materials 57
4.3 Fabrication of specimens 59
4.4 Instrumentation, loading and ?0
apparatus
test procedure
Chapter 5 Page
Description of Tests
5.1 Introduction 76
5.2 Development of the shearpatterns
crack
and the observed modes of shear failure 76
5.3 Prediction of shear failure type 97
5.4 Comparison between the shear crack
patterns observed unler uniform
loading and point loading 99
Chapter 6
Analysis of Test Results
6.1 Prediction of the diagonal tension
cracking load 101
6.2 Prediction of shear-compression
failure lQad 121
6.3 Comparison between diagonal tension.
equation and shear-compression
equation 137
Chapter 7
Comparison with other Results and
Deqign Rules
701 BSCP 110: Part 1: 1972 and ACI
(318-71) Building Code design
equations 142
7.2 Comparison of equations 6.5 and 6.7
for*one-or two-point loading with
published equations and code rules 143
7.3 Comparison of equation 6.5 for one-
or two-point loading with published
test results 145
7.4 Comparison of the shear-compression
equation 6.28 with experimental
results and other published shear-
compression equations 153
7.3 Comparison of the expressions
developed for uniformly loaded
beams with test results and published
expressions 158
Chapter 8
Page
Conclusions and Recommendations for Further
Research
8.1 Conclusions 164
8.2 Recommendations for further research 167
Appendices
Appendix A: Figures showing the transmission 170
length with 7 m,.
n diameter indented
wires and 12.5 mm diameter strands
Appendix B: Estir. 'ation of prestress losses 172
in accordance with BSCP 110:
Part 1: 1972 and CEB-FIP
Reco=endations
Appendix C: Calculations steps in analysis 179
of one-or two-point load cases
Appendix D: Calculations steps in analysis 183
of uniformly dietributed load
Appendix E: Mohr's failure criterion 184
assuming a straight line envelope
References 187

iii
 I ACKNOINLEDGMENTS

The work described herein was carried out in the

Departmentýof Civil Engineering at the University of

Glasgow under the general guidance of Professor W.T..

Marshall, until his sudden death at the end of 1975.

The author would like to express his appreciation to

Professor Marshall for the facilities of the Department

and to Professor H. B. Sutherland for taking over the

final stages of the supervision of the research.

The author is indebted to Dr. P. D. Arthur for his

valuable supervision, encouragement and advice through-

out the course of the investigation'.

The author is grateful to Dr. P. Bhatt for useful

I
discussions and criticisms, and to Dr. I. A. Smith for

helping with computer work.

The author wishes to express his thanks to the con-

crete laboratory and workshop staffs for their interest

and assistance. In particular, the help and interest

of Messrs. J. Thomson, J. Coleman and A. Galt is grate-

fully acknowledged.
The author also wishes to thank Mrs. E. Carr, of
17 Woodbank Crescent, for typing the thesis.

Finally, thanks are extended to the Sudan Government

for financial support during the period of the research.

iv
 I SWIM

This thesis presents a detailed study, based on

dimensional analysis and confirmed by experiment, of

the factors which affect the shear strength of prestress-

ed concrete beams without shear reinforcement. Sixty -

eight pre-tensioned concrete beams of six different

I- sections and one rectangular section were tested

under one - or two-point loading and twenty-three pre-

tensioned concrete beams of five-different I- sections

were tested under uniform loading.

The final mode of failure as well as the ultimate

failure load were observed to be functions of many

variables, some of which cannot be evaluated. As a

result the shear force at diagonal tension cracking

rather than the ultimate failure load was taken as the

limit of the usefulness of the beam in shear. Accord-
ingly an expression was developed for predicting the

diagonal-tension cracking shear force under one - or

two - point loads, and this expression was modified to

piýedict the total uniform load at the diagonal tension

crack in the case of a uniformly loaded beam.

Based on Mohr's failure theory, an expression for
the shear - compression failure load was derived. It

was shown that the ultimate strength of beams without

shear reinfoýcement must be limited-to the diagonal
tension cracking load or the shear load,
- compression
whichever is the lesser.

The equations developed -were coipared, for specific

V
cases, with other published expressions and code design

rules. Other published test results were shown to be

in good agreement with the derived expressions.

vi
 I NOTATION

All used are the standard

symbols
(30)
symbols Of BSCP 110: Part lt 19? 2
except as indicated below: -

av shear span.
(28929).
C horizontal projection of a diagonal crack
d I( distance '33
4
between centroids of flanges
Eci
static secant modulus of elasticity of concrete
at transfer.

e eccentricity of the prestressing tendons from

centroidal axis of beam.
f stress; concrete compressive stress at
compression face at any stage of loading.
f
av average normal compressive stress in the
compressive zone of beam.
fc
concrete compressive stress at compression
face of seition at failure (which corresponds
to strain CO.
fI
C uniaxial. compressive strength of concrete
(taken as 0.8 f
cu
fI
ct tensile strength of concrete (cylinder-sPlitting
value).
f
cu characteristic concrete cube strength, taken as
1.25 x (compressive strength of 150 x 300 mm,
cylinders).
f0
maximum compressive strength of concrete in
flexure (= 0.67 fcu for CP 110).
f tensile in
pb stress prestressing tendons at beam
'
-failure.

pi stress in tendons before deduction of losses.

fptr
atresss in tendons after elastic shortening.
f
P. 2% 0.2% proof stress of prestressing tendons.

vii
fprism te)5 28,29)
I
compressive strength of concrete prisms" I
f fo
sv permissible tensile stress in web reinforcement"3
fý
uniaxial tensile strength of concrete. t

fxx in the compressive

normal flexure stress zone-
of beam (taken as f
av).
f the
yy stress normal to the longitudinal axis of
beam due to applied load and reaction.
flif in a two dimensional
2 principal stresses stress
system.
(34)
h clear distance between flanges 0
k1
ratio of average normal flexure compressive
stress to maximum normal flexure compressive
stress.
k2
ratio of the depth to the line of action of
the normal compressive force to the neutral
axis depth at failure.
k3
ratio of the maximum normal flexure compressive
stress, fo, to compressive strength of concrete,
f fo (27) 30))
cu(= = 0,68 for CEB and = 0.67 for BSCP 110
fcu

U ratio of neutral axis depth at failure to effect-

ive depth.
Lt
transmission length(83).

bending moment at any stage of loading.

MU
ultimate resistance moment at ultimate shear
failure.
M
moment - shear ratio at failure.
C
first moment of area of cross-section above and
about the neutral axis.
qC
uniformly distributed load per unit length of
the span at diagonal tension-crack.
qCL total uniformly distributed load at diagonal
tension crack (written as WC in the photographs).
qut total uniformly distributed load at failure
(written as Wu in the photographs).

viii
 v shear force at any stage of loading.

vC
ultimate shear resistance of concrete
shear force at diagonal cracking.

VP the prestress
vertical component of effective
at the section considered(32).

Vs force by web steel.

shear resisted

Vu force failure.
shear at

v shear stress.
''

shear stress in concrete (Table 5

c ultimate
of CP 110), nominal shear stress for concrete
"Ve
F-d
w

horizontal shear stress.

Ivh
v
max. maximum shear stress.

v XY in the compression zone

average shear stress
at failure.
v
xymax maximum shear stress in the compression zone
or in the web at failure.

distance from the centroid pf concrete to

IV
centroid of the tensile reinforcement.
y
t distance from centroid axis of cross-section,
neglecting the reinforcement, to extreme fibre
in tension(32).
ecleo

ratio of uniaxial compressive strength to,

uniaxial tensile strength of concrete =f
C/ft
strain; concrete compressive strain at
compression face of section at any stage of
loading.

EC concrete compressive strain at compression

face of section at failure.

eu ultimate concrete strain in compressi

*(= 0.0035 for CEB-FIP(27) and BSCP 1100ho))

ix
F'o corýcrete compressive strain at compression
face. of section when fo ig reached (= 0.002
for CEB-FIP and 0.244 xl57f--cu for BSCP 110).
Epa strain in tendons produced by the applied
loading.
tpb strain in tendons at beam failure.

Epe strain in tendons due to the effective pre-

stress.
Epi strain in tendons before deduction of losses.
Ept. strain in tendons due to concrete prestress
-' at level of tendon "= fpt
(52) Ec
a
slopes of cables
XCL distance of the critical section in shear
from a support in a uniformly loaded beam
failing by diagonal tension.

distance of the critical section in shear

from a support in a uniformly loaded beam
failing in shear compression.
ev Asv
shear steel ratio(= 'Ed-
reduction factor(32) (ACI(318 - 71)given as
0.85).

110te: - In Chapters 6 to 8, expressions containing the terfil 1000V

0
will be dimensionally correct only when Vc is expressed in
29
kNv ft in N mm, and other dimensions in mme
ct

x
 I APTER 1
_CF.

INTRODUCTION

In the design of concrete structures, it is generally

desirable to ensure that ultimate strengths are governed by

ý112)
flexure rather than by shear A prestressed concrete

beam under the combined action of a shear force and be ing

moment may fail in shear before its ultimatd flexural strength

is attained if it is not adequately designed for shear. The

problem of shear failure in prestressed concrete beams is

important mainly because, unlike flexural failure of correct-

ly designed beams, it is characterised by small deflections

and lack of ductility. Shear failure can occur very sudden-

ly and without warning and it is qometi-mes violent and cat-

astrophic as illustrated in Figures 5.1. d to 5. l. h.

The collapse of a large part of the roof of a U. S. Air

Force warehouse in August 1955, due to the failure of the

major structural frames by diagonal tension cracking, expos-

ed the inadequacy of the design methods suggested in the

then current codes and created fresh interest in the study

of shear in reinforced concreteý3) Nowadays the introduct-

ion of the concept of limit state design in the codes of

practice requires a thorough knowledge of shear failure

since design for the ultimate limit state results in size

reductionýin turn may increase the danger of shear failure.

In view of the large number of factors affecting the shear-

ing strength-, and the complexity of the conditions
- stress
in the web of a cracked prestressed concrete I- beam, a

fully mathematical solution is not a practical possibility.

1
 As I -' beams are in practice the most commonly used

prestressed concrete structural members, the majority of

the test specimens in this investigation were I- sections

with differing geometrical properties. The dimensions

were varied to permit a systematic study of the parameters

the shear strength of prestressed concrete beams

affecting

without shear reinforcement under point loads and uniform

loads, so as to establish an expression for predicting

the shear force below which shear reinforcement is un-

necessary. Beams without shear reinforcement are not

common in practice, but they were used in this investig-

ation because in them the diagonal cracking shear force

could be defined clearly and the variables affecting it

could be studied.
A shear failure in beams without shear reinforcement

may be defined as a failure for which the primary cause

is the formation of an inclined tension crack due to the

combined action of a bending moment and shear force. In

prestressed concrete beams without shear reinforcement

the following types of failure have b een observed: -

(a) Splitting of concrete due to diagonal tension crack*
(b) Web crushing under compress. ion-
(C) The compressive zone is subjected to compression and
shear and can fail either by splitting or crushing
of concrete in the compressive zone-
(d) Splitting along the longitudinal rein-
of concrete
forcement following the formation of the inclined
tension crack,
The final mode of shear failure depends on various
&4)
factor which in turn govern the reserve capacity of

a beam after the formation of the inclined tension crack-

2
ing. Sorýe beams in this investigation carried a con-

siderable load beyond t.he first inclined tension crack-

ing load. In some instances the failure load Vu Was

70% greater than the first inclined cracking C>load VC

but the amount of this excess could not be predicted

as it involves some unpredictable factors. Hence the

shear force at the formation of the first diagonal crack

rather than the actual maximum load has to be taken as

the ultimate load for a beam without shear reinforce-

ment. This load has been studied in this investigation

using dimensional analysis and has been expressed in

terms of the beam properties and either the av/d or

the Ud ratio depending on the type of loading.

In some instances, such as beams with rectargular

cross-sections tested at higher av/d or L/d ratiost

shear-compression failure initiated by a flexure -

shear crack has proved to be the dominant mode of fail-
(42943)
ure. -Kar's prestressed rectangular beams were

good examples of this type of shear failure. For such

cases an expression based on Mohr's failure hypothesis

with a straight line envelope to the failure stress

circles was developed in this investigation to predict

shear force at failure. Then the lesser of the first

diagonal tension cracking load and shear - compression'
failure load is taken as a limit of the useful capacity
I
of a beam without shear rpinforcement in shear.
The expressions developed as described above were

compared, for specific cases, with other published

expressions and code design rules. The equations

3
 developeb were also compared with the test results
published by other investigators and good agreement
was observed.

4
 I CHAPTER 2.

HISTORICAL REVIEW

2.1 General: -
The shear strength of concrete beams has been a subject

of considerable interest to various investigatorsý115-8)

In 1973, an excellent report was published by the joint

A. S. C. E. - A. C. I. Committeý2) which referTed to over 200

documents and reviewed recent research results and design

proposals .
con cerning the shear strength of reinforced con-

crete .
stru ctures. Despite the tremendous number of refer-

ences in this subject, the Committee pointed out that

the question of shear strength is far from eing

settled.. In some instances the explanations of behaviour

I
and-design concepts that are presented are somewhat speculat-
ive and may change as more information becomes available".

A comprehensive review of the published work on shear

in concrete beams seems impossible to accomplish in a thesis
this nature,. and accordingly reference will be made only
-of
to some major papers. a

As shear strength of prestressed concrete beams can

reliablybe related to that of "unprestressed" beams, a

review of some of these papers is necessary.
2.2 She ar in Reinforced Concrete Beams:
2-2.1 Concept of shear strength:
Controversy the early development of
characterised
(6)
shear design from 1900 to 1910 Some engineers believed
that horizontal shear, 'h' was the basic mechanism of shear

.5
strength in reinforced concrete beams. Accordingly,

shear stresses were computed by the equation:

VQ
h Ib (2.1)

A second group of engineers recognised that the

basic mechanism of shear strength is diagonal tension

computed by:
v

Zb (2.2)

It took a full decade of heated discussion to

arrive at diagonal tension and equation 2.2 as basic

design tools. This was accomplished largely through

the efforts of Yorsch in Germany and Talbot in United

States.
2.2.2. In 1951 Clark(9)carried out 62 tests on beams

with no web reinforcement and on beams with varying

ratios of-web reinforcement. From the test results,

he derived the following semi-empirical expression for

the maximum shear stress vc as:

f1v (0,8 fcu) (2.3)
v*c= 17*3 + ý0.12 48.4Q
av/d

Clark was the first to include the av/d ratio in shear

equations and he was the first to account quantitatively

for all the variables listed by Talbot(lO)in 1909 as
influencing the shear strength of reinforced concrete
beams.

2.2.3 In an investigation sponsored by the Reinforced

Concrete Research Council at the University of Illinois)

U. S. A., Moody, Viest, Elstner and Hognestad presented'a

6
 (lljl2)
series oý tests in three reports and these test

results were analytically studied by Moody and Viest

(13)
in the fourth report 0
The authors observed that the phenomenon of dia-

gonal cracking was one which involved the combination

of flexural and s.hear stresses. Various attempts were

made to express this phenomenon in terms of rational

theory based on the ordinary theory of flexure and they
have not yielded any solution. Hence an empirical

equation for cracking load was reported. This equation

includes both concrete strength and the av/d ratio, and

is as follows:

For 8.7 kf -4 43.3 N/mm2

cu
0.8 f (1,
vc av
(0 CU 091,
v 0.8? 5 bd 0.12x -
cd .8f1- cuý 69.2 d
2
and for 43.36; f &L51.9 I/=
cu
VC a.
0.12x (0- 8w 43.0.8
3)(1 x ý23.3 0.1
v d)
d
c 0.875 bd 69.
.
(2.4)

Equation 2.4 shows that the rate of increase of the

nominal shearing stress, vc, is decreasing v.ith increasing

concrete strength, and that for concrete strength greater

than-f 2,
='43.3 N/mm the nominal shearing stress is
cu
independent of f This
cu equation was derived from test
data with a limited range of variables. The range of
av/d in test beams was 0.57 to 3.03, and hence equation
2.4 is not necessarily applicable to longer shear spans.
For the ultimate shearing stress, they assumed that

7
 the ultimate moment could be expressed by the same type
( )o
of equation as for pure flexure see'Section 2.3
'and Viest (14)
2.2.4. V.orrow in their tests covered a

wide range of a ratios ranging from 0.96 to 7.79.

v/d
As a result of this wide variation in av/d ratios,
different modes of failure were observed, which could
be seen from their photographs. Analysing their test

results, they reported that the concentration of concrete

compressive stresses at a critical section was caused
by concentrated rotations at the bompressive end of the
diagonal crack, the tensile stresses on the 'compression

face' were caused by larch-action' present-after the

formation of the diagonal crack. The concentration of
compressive strains at the critical section led to a
premature crushing of the compressive zone of concrete

and thus to failure before the flexural capacity was

reached.

They gave two semi-empirical expressions similar

to those of Moody and Viest. Oneýfpr the diagonal
tension crack load in terms of the nominal shearing
stress which is given by:
3.19
v 0.12 +-0.8 f
C (14/Vd)c cu (2.5)

Where Ec is given from Kesler's data as

460 x(O. 8 fcu)+ 12456 NIM 2

is The othe for 'shear-compression' in terms

strength
of shear moment capacity.
They concluded that the presence of a diagonal
tension crack is dangerous even in beams. This is
short

8
because diagonal tension cracks in beams without shear

reinforcement are considerably wider than flexural cracks;

furthermore, a few repetitions of load may cause the dia-

gonal tension crack to spread and possibly result in

splitting along the reinforcement or in a premature shear

failure. Accordingly for beams without vieb, reinforcement,

the diagonal tension cracking load may have to be consid-

ered in design as the ultimate capacity in shear.

2.2.5. Whitney(16) reported that the value of the

unit shear at diagonal cracking, is-not a simple function

of concrete strength, since it depends largely on the

tension reinforcement. He proposed the fd1oviing equat-
ions, for one-or two-point loads.
Mu
v 0.346 -11/=2 + 0.26 2 II/=2 (2,. 6)
c bd

and for uniformly distributed load

2+0.54 MU 1 '?
N/mm 2 (2. )
v=0.484 II/mm
c bY2 ý 0.5 L/T
where
Mu0.8 feu for over-reinforcbd beams.
bd 3
1ý )reinforced
fL1-l.? fýy L
2y
bdý fo r under-
-X(0.8 f-
cu beams.
Balanced reinforcement is given by
(0.8 fcu)
Ro 0.456x
-'ý

Like Morrow and Viest, VilAtney considered the diagonal

cracking load as the ultimate strength in shear of the
beams without shear reinforcement.
2.2.6. In the United Kingdom, Taub and Neville(17)

9
 conducted a large number of tests and they emphasised

the importance of the moment-shear ratio. They showed

itlso that the lattice analogy is not satisfactory for

the design of bent-up bar reinforcement and recommended

the use of a combination of bent-up bars and vertical

stirrups to achieve the most effective resistance to

shear failure.
(18)
2.2.7. Smith developed an expression for the

cracking load taking 'account of the influences

of the ratio of the main reinforcement, ?, and the av/d

ratio. Smith gives for sections already cracked in

-
flexure:
v (9 2
c-0.? - 186 + 0.0015?
9FO--13Fcu 0.0) 1)
Y+7.5
Eh- - d
a > 2*4 (2.8)
v/d
where 0.79 0.8 f is the modulus of rupture.
CU,
(19)
On the basis of tests by Krefeld and Thurston
Smith concluded that the critical shear force at the

critical section of a. uniformly loaded simply supported

beam can be expressed by the equation:

(U. D. L. ) (1 fo-.
= 0.247 + 7.5 Q)xO. 79 Bfcu (2.9.1
bh
where the critical section is aý a distance X from
the support and -X is given by the following equation:
X (L (2.9.2
-X) T-1.2
d (L - 2X
Smith(l) later on revised equation 2.8 to give

more weight to the influence of the ratio of main

reinforcement and he gave the following: -
for eA 1.25%

VC ý0---B--ý-u (3 ay2 (1
= 1.57 0.79 0.13 + 0.0224 v + 14ý
bd d)

10
for Plý-1.25%
10.13
VC (3 (+
9 FO vI (2.10.2
= 0.? -8f cu + 0.0224 ?OP
bd d)2
when a "*Of the second term in the squ; re brackets
v/d,
of both equations is dropped.

2.2.8. Brock in a private communication to the Shear.

(1)
Study Group presented a more general approach to the
,
problems of shear in beams without shear reinforcement.
This was in the form of an interaction diagram between

the ultimate moment and the a ratio.

V/d
This can be su=arised by the following equations:
Mflex 0.268 (-
0.456
fcu bdiý F'o
.8
for under-reinforced sections, i. e. when P 4eo
Mflex
and
= 0.293 + 0.04 (P/Po)
0.8 f bd2
- cu
for over-reinforced sections i. e. when PA
Fo (2.11)
where = 0.456 x (0.8 f
L
/ fy
cu)
The section attains its full flexural moment of

resistance M, when
flex 0 I,

'IV 144 fy
>- L
d 8000

where av (or M at the critical section)is less than

Vd--
Id
144 f 8000 there is shear deterioration except for
YL
cases where M, fcu 2), e
bd 0.1()6
flex/(0.8 0
For very sinLall values of a there is a splitting
v/d
type of failure which is assumed to be analogous to
the Brazilian tensile test in which a cylinder of
concrete is loaded along its sides. Recently this

11
 (20)
approach veas used by Desayl too.

2.3. Shear-compression ApDroach.

The co=onest approach in determining the ultimate

load capacity of beams failing in shear is the so-called

shear-compression approach. This type of failure was

(21)
envisaged by Laupa et al as being essentially the

same as flexural failure, the only difference being

that the depth of the compressed zone was reduced by

diagonal cracka extending considerably'higher than

flexural cracks at failure. This results in crushing

of concrete at the head of critical shear crack.

In this approaC-h, the dowel force is ne--glected,
it being assumed that the stiffness of the dowel formed

by the main steel is greatly reduced when the horizon-

(22)
tal part of the crack is formed With this assum-

ption the internal structural behaviour is similar to

that of a simple tied arch. The external load is

supported by an inclined, arch-like, thrust in the,

concrete above the shear crack, and -ýhe horizontal

component of thrust at-the support is resisted by

the tension steel acting as a tie.

This concept of the structural behaviour of a
beam containing a shear crack is supported by measure-

ments of strains in the concrete which show that the

centre of compression in the concrete falls as the
is (23)
support approached 41
Provided that the anchorage0 of the main rein-
forcement is sufficient for the tie-force, and that
the geometry of the crack is such that the larch' does

1.2
not become-unstable, i. e. fails by instability of the

compression zone, collapse of the structure is caused

by the crushing failure of the concrete at the crown

of the arch. This ultimate compression load is given

generally, by:
Ilu = VU XaV=k1k3f bku d2 (1 -k2k (2.12)
cu U)
The coefficients (k and k2 are generally ascribed
1k3)
values equal to normal flexural ones, viz.
(1) (2.13.1)
Bjuggren klk3fcu = 0.5 fcu
(13 Bfcu i)
Moody and Viest kkf]. 121 -5
=1-4.58xl44K(o. X
1 cu 10
.3 (2.18.2)
(0.8 f
cu
(21)k (1 10.8xl44x(O. fcu)
Laupa et al 37 K
1ý3f cu= -
-LO5
(0.8 f (2.13.3)
cu
(24) (2.13.4)
Regan klk3fcu = 0.67 x(O. 8 f
CU)
(26)
Walther(25) and Ojha have assumed that the capacity

of the concrete to resist longitudinal stresses is

reduced by the existence of a shear crack. Walther

applied a biaxial failure criterion to the average

4
compressive and shear stresses above the head of a

shear crack and obtained:

fprism
k '2 (2.13.5)
lk3fcu 1+3.2(Vctp

where f is the compressive strength of concrete

prism
prisms.

The main difficulty in equation 2.12 lies in a-

realistic assessment of the*neutral axis depth kud.

The neutral axis depth, kud, in equation 2.12 differs

from the normal flexural one because the flexural

13
 assumption that all sections plane before loading
I
remain plane during loading is not valid in the presence

of shear cracks, At the same time kUd is not generally

equal to the ultimate flexural neutral axis depth

because in most shear failures the main steel has not

yet yielded at failure.

A detailed derivation of the expressions used in
determining the neutral axis depth is beyond the scope

of this review. The final expressions derived by

some investigators in calculating Iýu are shoym as

follows: -

--Bjuggren:
mu20.5
k1k3ku (1 - 0.4 ku)
Yc-Ub d
PEs Eu 2
where k -
f PES
cu
f
CU
and ýu =. 0.003

Moody and Viest:

m
11
Pf 0.42 ef6t
2 st (2.14.2)
TCU bd f kl k 3(5.8 fcu)
cu
r-
where 1450 k, kS(0.8
-4 fcu)
f ý 0.729s 6.9 10 Es + +
st PE
s
Laupa et al: '-

M
u kl k3ku
fcu bd2 .(1-0.45 ku) (2.14.3)
ý1.23
where k 1.11 lý
-!- --0.92
and ku (np) 2+
np + 2»np
ES 10000
where n= + 144xO. 8 f
Ec
cu

14
Laupals equation appears to predict a decrease of

beam strength with increasing concrete strength,

(see 2.13-3), but it is intended to ýe
equation not
.
applicable in such cases. '
Regan:
muI
kjkýku (1 0.375 k (2.14.4)
f bd U)
cu
n02
where ku
no +

n02 3eEs Eu
no 2xO. 8 f
cu
t=0.0035
u
The values of k1k3 are given by equations

2.13.2 - 2.13.4

Z. 4. Some Code AODroaches to Desipm of Bearns for

Shear Strength:

European Concrete Committee - International

Federation of Prestressing.
The 1970 CEB - PIP Reco=endat; that
-ons(27)state
the resistance mechanism of a beam or slab subject to

shear loading depends essentially on the mode of

cracking under design loads. The Committee recognises
three cases of behaviour, but from a practical point
Of view only two are discussed, viz. AB anI C.
Case AB is characterised by the development of
web-shear cracks without ftexural cracks at ultimate.
In the region in which web-qhear cracks occur, web
reinforcement must be provided for the difference between
the principal tensile stress at ultimate and a reduced

is
 I
value of the tensile strength of concrete (Clause

R 43 - 132). Failure due to crushing of the web

should be prevented (Clause R 43.131).

Case C corresponds to those regions having

flexure-shear crackinta,, (dlause R43.14). The max-

.,
imum shear is limited to prevent crushing of'the

web. The shear3 VC, carried by concrete is a

function of the amount of longitudinal reinforce-

ment and the square root of concrete strength.

For prestressed concrete beams the design proc-

edure involves checking Case AB and C. In case C

the shear, Vc, carried by the concrete is increased

as a function of'average prestress.

2.4.2. U. S. S. R. Building Codes.

The shear strength carried by concrete is

assumed(28129) to be a function of the tensile strength

of concrete and the horizontal projection of the diag-

onal crack, ICI. The valiie of 'C' is taken as that

length giving the minimum value of'(Vc + V. ), and
these are given by:
0.15 f bh2
VC prism
C
fDrism bh
C zs
v
fprism Asv fyv bh (2.15)
or Vc +V0.6
Sv
where fprism is the strength of concrete prisms.
The maximum shear stress on the web is limited to

prevent crushing of the inclined concrete struts in

the web.

3.6
2.4.3. British Code of Practice.

The provisions for shear in the new code BSCP 110:

Part 1: 1972(30)differ considerably from those in BSCP

(31)
115: 1959 BSCP 115; 1959 treats shear by limiting
.
the principal tensile str6ss in regions of a member
uncracked in flexure to values given in Table (6) of
the Code. For regions cracked in flexure this Code
is very vague and only gives the warning that "special

consideration should be given to the shear resistance

under ultimate load conditions where the section is

cracked in bending". However, the new Code, BSOP 110:

Part 1: 1972 gives separate expressions for the shear

carried by the concrete in regions of a member un-

cracked in flexure and regions'cracked in flexure.

For regions uncracked in flexure the BSCP 110

limits the tensile stress at the centroidal axis to

0.24ýý. The shear force, VC01 corresponding to
this-principal tensile stress, is the shear force

carried by the concrete at the limit state of collapse.

From elastic theory, and neglecting the stress normal
to the longitudinal axis, due to applied load
j
%7
and reactions, V is given as:
col
V bh Vft+0.8
2
= 0.6? ftf Code equation (45)
co cp
.where ft=0.24r-fcu

For regions cracked in flexure, the BSCP 110 gives -

the value of shear force ciarried by the concrete as
Vcr, given by:
fpe_ mo v
V 0., 5- vc bd + Code equation (46)
cr fpu
ý 0.1 bwd ýf-cu
where vc is the ultimate shear stress and is given

as a function of Pand fcu in Table (5) of BSCP 110,

and Y10 is the moment necessary to produce zero stress

in the concrete at the depth d, given by Mo = 0.6 fI
pt Y
The shear force at the inclined crack should be

taken as the lesser of the Code equations (45) and

(46).

2.4.4. American Building Code:

The American Building Code ACI (318 - 71)(32) 9

has two alternative methods for calculating the shear

force carried by the concrete in regions of a member

not cracked in flexure. One given in terms of aver-

age shear stress, v derived from the principal

ew,
tensile stress equation as:
(0 P-8 V
29 fcu + 0.3 f+ D Equation(11.12)
v0 . -2-
Ew'
d
cw cp d-)
of ACI(318-71)
a.nd the other by limiting the principal tensile stress

at the beam centroid to 0.33 VO---8fcu-

For regions in flexure, It it

cracked gives vci

as the average shear stress at diagonal cracking.

Neglecting the dead load v is given by the follow-
Op I cil
ing expression:
VO-8
0.5 fcU + fpe
vei 0 0.05 0-81 fm +-
Yt b'd 2 (a
v/d)
Equation(11.1ý of
(,
ACI qlB-? l)
0 is a reduction factor of 0.85

18
.
2.5. Shiýar in Prestressed Concrete Beams:

2.5.1. Prestressing introduces extra compressive

stresses which are expected to reduce the final tensile

stress resulting from shear. Thus the prestressing

force creates a new variable in addition to those

already mentioned for ordinary reinforced corcrete

beams. The factors affecting the shear of prestressed

concrete beams and their behaviour unler various con-

ditions of loading and levels of prestressing have

(213-54)
been discussed in various papers
The shear strength of simply supported prestress-

ed concrete beams without under

shear reinforcement
(33)
one-or two-point loads have been studied by Hicks
(26)
Sethunarayanan Sozen, Zwoyer and Siess
(37) (38)
Evans and Hosny walther Warner and Hall
,
(41) (42,43)
'Evans and Schumacher(40) Swamy Kar
, ,
(45).
Arthur(44) and Arthur and Mahgoub
(46),
Kar(42,43) Wilby and Razir Hanson and
I
Hulsbos(47) and Arthur, Ehatt and Duncan(48) made
I fa
investigations on uniformly loaded beams. Bennett,

Abdul-Ahad and Neville(49) have studied the problem

of moving loads.

Among the investigators who studied the effect

of shear reinforcement on the shear strength of
prestressed concrete beams were 1.11'acGregor, ýozen

and Siess(50) and they also reported on the behaviour

of prestressed concrete be ams with draped reinforce-

(51)
ment A paper which dealt with the shear

strength of continuous beams of prestressed concrete

19
 (52)
was by Jena and Pannell
Now some of the papers which dealt with the

shear strength of prestressed concrete beams without

shear reinforcement will be reviewed in more detail.

2.5.2. Shear strenath under one- or two-Doint loads:

(33)
2.5.2.1. Hicks in his tests covered a wide range

of av/h ratiosin an attempt to relate a to the

v/h
various types of failure and to investigate the
C>
effect of shear span and concrete strength on shear

strength. The specimens were 16 pre-tensioned con-

crete beams. One cross-section of unsymmetrical

I- section was used. The bottom and top flange

breadths were 127 and 178 mm respectively, the web

breadth was 38 mm, the flange thickness was 70 mm

and overall depth was 254 mm.

The prestressing force was developed by using

ten -5 mm diameter indented wires having an ultimate
2.
strejigtn of 1590 N/mm The wires were distributed
in such a way to give zero and 133.8 N/mM2 compressive
stress at the top and bottom fibres respectively and
this was regarded as constant in all the beams. The

concrete strength, which was one of the main variables,

2.
was varied between 34.0 and 4?. 0 N1=
By varying the a ratio between zero and 8.15, -
v/h
four types of failure were observed which gave 4icks
his limits of the a ratio within which each type
v/h
of failure occurred. For a ": 1.5, 'shear distort-
v/h
ion' would prevail which describes the state of web

20
cracking"and crushing followed by failure of the
separate flanges in flexure, cracks appear at the
top flange above the reaction and in bottom flange
below the load. 'Diagonal compression' would be the
failure pattern for a between 1.5 and 4.5. The
v/h
main difference between this type of failure-and
the shear distortion was that, in the diagonal

compression failure the vertical cracks in the top

and bottom flanges occurred away from the reaction

and load points. At higher a ratiosýbetween

v/h
4.5 to 9.0, failures tended to resemble the

'diagonal tension' type. This occurs by a sudden

splitting of the web and causes the beam to fail

immediately without any traces of web crushing.
Above a m 9.0 flexure failure would be expected.
v/h
For those beams failing in diagonal tension,
Hicks suggested the following equation for the

principal tensile stress at failure at the neutral

axis: 4
I (21h
ft= fet - 0.187 -2 av/h A: 2 ý2.16)

He concluded his investigation by suggesting

a reduction factor of 0.7. to be applied to the

ultimate flexural moment for av/h 47.0 for design

purposes.

This was. an investigation of restricted scope

since Hicks used only one. type of cross-section

with a constant prestressing force.

21
2.5.2.2,, On the other hand tests reported by
(34)
Sethunarayanan on pre-tensioned I beams with

various cross-sections revealed that the transition

from one mode of failure to another depended not

only on a ratio, but also on the amount of pre-

v/h
stress and strength of concrete. This was later
(44)
confirmed by Arthur's tests.

Sethunarayanan's tests were conducted on 32

pretensioned I beams with top flange breadth ranging

from 127 to 178 mm and with a constant bottom flange

breadth of 1?8 mm. The thickness of the top flange

ranged from 50.8 to 81.2 mm and that of the bottom

flange was constant at 63.5 mm. The web breadth

ranged from 35.0 to 62.2 mm. The distance between

the centroids of the flanges was taken as the effect-

ive depth and this ranged, accordingly, between 171

and 184 mm.

The prestressing force was developed through

5 mm diameter indented wires varied in number between

5 and 10. The wires were stressed and distributed

over the cross-section to give a variation in f

cp
2
ranging from zero to 7.75 N/mm The concrete strength,
feul 20
varied from 84.4 to 41.6 N/mm

Prom his test results Sethunarayanan developed

the following expression for the diagonal cracking
load: -
(1 1+f
cp/fct h bv, (2.17)
vc = fct +
a
v/d'
Equatio4 2.17 shows that ,*
VC is inversely proportional

22
 to flange t1iickness which conflicts with ArthuýJ44)find-

ings. He also developed an analysis based on the truss

analogy for the cases in which web crushing followed the

initial inclined cracks.

2.5.2.3. As a part of an investigation of prestressed con-
(36)
crete for highway bridges., 8ozen, Zwoyer and Siess

carried out tests on 43 and 56 rectangular and I-section

concrete beams respectively. All the beams were 152 by

304 mm,overall in cross-section. For I-beams the b

w/b
ratio had values of 0.29 and 0.50, and the hf/d ratio

ranged from 0.28 to 0.50. The end-blocks were 456 mm,long.

5 mm,diameter hard drawn wire was used with f

pu
varying from 1660 to 1870 N/mm2. The stress in the pre-

stressing wires, at test, varied between zero and 970

N/mm2 and the prestressing steel ratio, P, varied from

0.10; 6 to 0.96% giving a variation in f between zero

cp
2
and 6.23 N/mm .
All the I-beams and eight of the rectangular beams

were pre-tensioned, the rest being post-tensioned and

grouted. All but three of the I-beaw had prestressed

external stirrups to prevent propagation of cracks into

the end-block. Those external stirrups were placed, one
at each junction of the web and the end-block and one
immediately on the outside of each reaction block.
The beams were tested simply supported with av/d

ranging from 2.7 to 5.4 and most of the tests were -

conducted between av/d = 3.2 and 4.2.
The strain measurements in the top surface of the
beam showed that after web cracking, the longitudinal

23

f
 strains adjacent to the load points were higher than in
I-
the zone'between-them, which led to the conclusion that

web cracking effect was to cause concentrations of strains

leading to crushing of concrete.
Ninety beams failed in shear and the remaining nine

failed in flexure either by crushing of the concrete or

fracture of the steel. The modes of shear failure

observed were given a detailed description in their

bulletin. They classified the shear failures into two

categories: lahear-compression' and 'web-distress'. The

shear-compression failure was described as similar to

flexural compression failure except that the concrete

crushed at the upper end of the inclined crack where

I
there was a very high strain concentration. This mode

of failure was observed in both rectangular and I beams.

On the other hand web-distress could take any one of the

following forms:

Secondary inclined tension cracks formed near the

supports and above mid-height of the beam, which
separate the compression flange from the web,
leading to violent failure.

Inclined cracks near the loading points extending

horizontally toward the supports tending to separate
the web from the bottom flange entirely, or
Crushing of the web under high compressive stresses
due to larch action', created by the loss of shear
flow between the stee 1 and compression flange.

Based on their definition of the inclined tension

cracking load as the load at which the inlined crack

started to affect the behaviour of the beam, they

24
 analysed thbir test results and derived the following

semi'empirical expression for the inclined cracking load:

(1 f
Vc =f +fP bwb d
t (2.18)
a
v/d
where fl,t was given as a function of (0.8 f as
cu
I--16.92 .
ft _ N/MM2
41.5
B+ (2.19)
0.8 f
cu
I- They gave B=1.0 for concrete with regular coarse

aggregate (maximum size 38 mm) and B=2.0 for small

size coarse aggregate (maximum size 10 mm).

They concluded that the inclined cracki-ng load

should be taken as the limit of the usefulness of a
beam in shear since the development of inclined cracks

was unstable..
The limited number of tests outside the range of

av/d ratio of 3.2 to 4.2, means that the confidence which

can be placed in equation 2.18 is limited, and may lead
to seriously unsafe overestimates of the strength
particularly for low values of av/d(45) 0
2.5.2.4. (37)
Evans and Hosny analysed the test results
carried out by Hosny on post-tensioned prestressed
concrete beams, together. with the results of similar
tests (33,6
) (37)
reported by Zwoyer and Thornton
Hosny's beams were 3 rectangular and 17 1 beams.
The'specimens were 101 by 304 mm in overall. The I
section had its top and bottom flange thicknesses as
62 and 70 mm respectively and the web breadth was 47.0

25
 to 50.8 mm. This gave a variation in,, hf/d of 0.240 to

0.245 and in bw/b from 0.46 to- 0.50.

The prestressing force was developed through I'lacalloy

I- - bars. All the beams were grouted qxcept two. The per-

centage of the prestressing steel, ?, ranged from 2.453

to 2.98?. The average prestress in the concrete at

test, fcpl ranged from l.? 2 to 5.25 N/=2. The concrete

strength measured as (0.8 f ) ranged from 34.? to

cu
2.
'58.1 jj/MM

Thornton's beams were 13 pretensioned I beams with

prestressed compression reinforcement placed at 11.0 mm

below the top fibres. Two cross-sections were basic-

ally used with b as 0.3? 5 and a variation in h

W/b f/d
of 0.188 to 0.210. The percentage of the prestressing

steel wires, e, was 0.437 and 0.485 giving a value Of

2.
fcp ranging from 0.61 to 3.16 NIMM The strength of
the concrete vvas intended to be constant and was 42.6
(=0.8 f, 2
N/mm The av/d ratio ranged from 2.32 to
u)
3.48.

Evans and Hosny gave a full description of the modes

of failure observed in Hosny's tests which were:

Shear-compression, diagonal crushing of the web and

shearing of the compression zone, the last being

observed in I sections with web reinforcement. In

their analysis of these test results, they took the

nominal shear stress to be'a function of F. '-feu,

the cross-section properties, the prestressing force

and a ratio and gave the following,, expression for

v/d
ultimate shear force: -

26
 ?,,ý0.350 5ý! Y-46-70
For rectangular section where and 2.57!
Id

(0.83
+ 0.15 p fpýj 0.8 fcu N, Z (2.20)
VU =
(av/d)
a,
For an I section where 2.32 d
3.60
L1.91
+ 0.15 fp" ?) bw Z
(2.21)
v/d)
From their analysis of the I section results, they

noticed that the ratio of maximum principal tensile

stress to the tensile strength of concrete was constant

for a 3.0. They gave the following relationships
v/d-ý"
for the value of the maximum principal tensile stress,

ft as: -
aV63.0
f=4.99 - 0.624) ; or (2.22)
t l.? 5 +a d
v/d
f=0.312 V 0.8 ff or -Lv ýkM (2.23)
t cu d -:

The applicability of equations 2.21 - 2.23 is rather

limited since they are based on a very narrow range of

av/d values.
(39)
2.5.2.5. Warner and Hall developeb expressions for
the principal tensile stress, ft) as a function of av/d

and fhe concrete strength.

For f 2
43.3 H/=
cu -ý,.
ft01 034 f (6.85 - 3.9 a for av
. v/d) 1-1.5
cu d
fC 0.034 f for av>1.5
cu d
For f 43.3 2
--' N/mm
cu
aa
f (6.85 - 3.9 v )(0.077 0.001 f )f for 1-1.5
td cu cu -Xd
f (0.077 - 0.001 f )f for (2.24)
t cu cu d

27
 40)
2.5.2.6. Evans and Schumacher( conducted tests on

5 rectangular and 49 1 section fully bonded post-

tensioned concrete beams. The percentage of the pre-

length of shear span, shape of beam

stressing steel,

cross-section and curing method were taken as major

variables.
Basically three shapes of cross-sections viere

employed, rectangular and two I-sections with nominal

b ratio of 0.84 and 0.50 and hf/d ratio was found

w/b
to range from 29 to 31. The I beams had solid end
. .
broader than flange breadth.
-blocks
Three types of steel were usea, viz, Lee-YcCall

bars or Gifford-Udall cables of 7 and 5 mm diameter

hard drawn prestressing wires. The percentage of

the prestressing. steel, p, was varied between 0.21

and 4.77. The beams were stressed, at minimum, 10

days after casting and grouted immediately afterv; ards.

fcp v alues ranged from 1.45 to 5.60 N and fcu

2*A
ranged from 40.5 to 69.5 NIMM

Most of the beams failed by shear-compression,

diagonal cracking alone, diagonal cracking followed

by web-crushing, or fle xur al failure. 13 of the

beams failed in flexure and two showed failure of

the web following the formation of cracks at the

upper web flange junction.
They anaiysed their test results and came out

with two expressions, one for shear-compression

failure and the other for diagonal crpcking failure.

28
For rectangAar sections failing in shear-compression,

they gave:
f
(1.400 fcu ku (1.00 42 ku)
Vu = - 1.388pjýR) - .
cu v/d
(2.25)
efpu/fcuL 0.72
where
For I-section failing in shear-compression,
ff
p Ru)(3.45FýIu- 0.19)kbd fc ku(l-. 42k
Vu = (1.400 - 1.,,, f f - u
cu cu av/d
(2.26)
f
13U 1 0.8
where 0.241 fcU

in the above two equations 2.25 and 2.25:

-2 (2.27)
k=1.2 - 0.66lx 10 fcu

RR (0.534 + 0.063 Pfpe) (2.28)

and kfU fcU
hf for I-sections.
d
For diagonal cracking load, they gave:
f 0.275ý
fl bwd (2.29)
V=3.33 ct 1+2,13) 0.10 +
c fl
ct
They concluded that the lesser of the shear-

compression load and the diagonal cracking load should

be taken as the limit of usefulness of a beam failing

in shear.
The authors realised that equation 2.29 might

not be applicable to other I- sections because

the ratio hf in
was not varied their tests, although
d
it would be expected to have an influence on the
0
diagonal cracking load.

29
 (41) investigation
2.5.2.7. -Swamy carried out a preliminary

on seven post-tensioned, unbonded, hollow rectangular

beams to study diagonal cracking failure.

All the beams were 229 = by 152 overall with

depth of 203 mm, having b 0.25 and

effective c:ý w/b
h = 0.25. The prestressing steel used was. Macalloy
f/d
high tensile steel bars and the percentage of the pre-

stressing steel, ?, ranged from 0.?? 0 to 1.540, giving

_
2.
a var iation in f from 4.43 to 8.23 IV= The con-
cp
. 2
crete strength, f ranged from 59.2 to ? 3.4 N/m.,,
a
cul
The beams v-#,ere loaded at third-points giving av/d as

3.? 5.

The tensile steel and the compressive concrete

strains and deflections all showed essentially elastic

behaviour until failure occurred. The principal tensile

from the gauge readings taken in

stress obtained strain
I three directions at the centre of the web at the mid-

point of the shear span were found to beer little

relation to either the theoretical stresses or the

tensile stress of concrete. Accordingly Swamy con-

eluded that the principal tensile stress was an

unsuitable criterion for predicting the cracking load.

Failure took place by diagonal cracking in all the

beams except two which failed by flexure initiated by

yielding of steel and they gave warning of distress

before failure. There was practically no difference

between the cracking load and the ultimate load, so

Swamy considered the diagonal, cracking load as a

measure. of shear capacity of the beam. Because of

30
the small h=ber of the test beams no attempt was made

to dev.elop an expression for the diagonal cracting load.

(42,43)
2.5.2.8. Kar conducted tests on 26 rectangular

and 9 I-beams which were all post-tensioned and grouted.

The rectangular beams were composed of ten 125

by 250 mm, nine 125 by 300 mm and seven 100 by- 200 mm.

The I-beams were all 150 by 300 in overall with b

w/b
as 0.254 and hl. ranging from 0.358 to 0.383. The

prestressing steel was 5 mm diameter and 7= diameter

high tensile steel wires. The concrete strength,

measured as (0.8 f ), was varied between 27.9 and

CU,
2
38.9 N/mm and the a ratio ranged from 2,0 to
v/d
6.0.

The majority of beams failed by shear-compression

and the rest by web-crushing. For the prediction of

the shear-compression failure load, Kar based his

analysis on a modified linear strain diagram,

equilibrium of internal forces and equilibrium of

internal and external moments. Then-he developed

expressions which require a trial and prror procedure

to determine VU for rectangular beams.
(42)
Kar examined the web of the I-beam and found
that the plane of rupture prior to cracking was subject
to non-uniform longitudinal compressive stress. Accord-
ingly Kar proposed Seth' S(42) failure under
criterion
0
combined stresses which is identical in nature with
Coulomb's internal friction theory for biaxial stress

31
conditions,, to be appTied to the rupture conditions on

the web of the I beam. Using the experimental results,

KIar(42)modified Seth's failure criterion to take into

the influence av/d and developed the

account of
following equation for predicting the web-shear

cracking shear force:

Cr I bvi(O. 8 fcu) (N'Prcu)

ve (i + C1)2- 2Cl(l + Cl)
2Q 112
L 2) )2
(i -Cl FT
- -.
C-_u

0.0565 av av
where Cr = 0.953 - ;ý2.0
-h 757.
or
Cr 1.88 - 0.52 av a. 2.0
.
h h
0.8 fcu + ftý 1
and G, - where ft=0 624 0.8 f
0.8 f . au
cu -ft
(2.30)
(44)
2.5.2.9. Arthur carried out tests on 55 pretensioned
concrete I beams, treating as major variables the solid

end-blocks, the strength of concrete, prestressing

force, beam cross-section av/d
and ratio.
Basically four cross-sections were used. Three of
the cross-sections were 152 by 304 = in overall with an

effective depth of 272 mm. The fourth one was 152 by

228 mm with an effective depth of 200 =. The ratio
b,
of w/b was varied between 0.33 and 0.50, three differ-
h
ent values viere chosen and f/d was 0.19 and 0.25.

The prestressing-steel was indented tensile steel

wires with diameters of 2.65 mm and 5 mm. The aver-
f
age effective prestress, cp, ranged from 4.17 to

32
 2. fcul between
6.20 ll/= The concrete strenja:, th, varied

06.5 and 68.9 N/mm2.

The av/d ratios covered ranged from 1.12 to 4.57.

The mode of failure observed was mainly diagonal crack-

ing either alone or followed by either web distortion,

web crushing or flexural compression. Shear-compress-

ion failure init. iated by a flexure-shear crack was

also observed in-some cases.

The solid end-blocks were found to have no effect

on the cracking load. In 86% of the tests conducted,

. the shear loads carried by the beams exceeded the

diagonal cracking load, but this excess could not be

predicted. Accordingly Arthur limited the useful

capacity of a beam failing in shear tO the diagonal

cracking load. The following expression was developed

for its prediction.
f 3.34
Ve =f (1 + T722) bw hf (0.73 + T--) (2.31)
ct et v/d

for hf/d
0.19 Is
where the,, fl values are those given by Dewarls(55)
ct
eXpr'essions:
fc't 2/3
= 0.25 f for crushed rock
cu
2/3 ý2.32)
or ft = 0.23 f for gravel
ct cu
Equation 2.31 above is'based on two values of h f/d

so its applicability is rather limited, and it tends

to over-estimate the value of VC at high values of

h. (see Section 7.2). It is not applicable in the
f/d

33
r
--

0
case of a 'rectangular beam. Arthur suggested a

reduction factor of 0.7 to equation 2.31 for design

purposes,
(52)
2.5.2.10. Jena and Pannellcarried. out tests on two

series of 20 two-span unbonded prestressed concrete

I-beams of bw = 76.0 mm. The first series was 152 mm

wide by 229 mm deep overall5with hf"= 57.5 and the

.
- second series 152 mm wide by 305 mm deep overall,

with hf = 70.5. Each of the twenty beams had a total

length of 4.9 m and was continuous over two sPans of

2.3 m each. The prestressing steel was two-strands
of 12.7 mm nominal diameter. The cable profile is

shown in Figure 2.1. The effective

prestress the in
2.
strands ranged from 765 to 1061 NIMM The concrete
2.
strength, fcul varied between 35.9 to 56.2 N/mm

Two-point loads were applied, at equal distances

on either side of the central support. Of the 20

beams tested, 17 failed in shear, remainder failing

in bending. a
Jena and. Pannell compared the limiting
principal tensile stresses at the centroids of the
17 unbonded prestressed concrete beams failing by
diagonal tension cracking with others obtained from

pretensioned and post-tensioned I beams with similar

modes of failure and found that there was no dist-
inguishable difference between the data derived from-

both cases. During their tests they observed that

the diagonal tension crack appeared suddenly in the

web at or near the point of contraflexure and very

34
 14
Is u tot st is Ikt

tn

-1
in
LIJ
it
w
Co
Co fi
(1) ca
r-i (1)
ri
w ýA
ca
'a
1

0 0
0

(1) 0)
r-i orl
12

Nu ! 94 A

0-
t:

I;t
1

35
close to fhe centroidal axis of the uncracked section.
Thus basing their analysis on the principal tension

theory, Jena and Pannell using their own results and

(33j34j42)
those of others derived expressions for the
,
maximum principal tensile stress at the neutral axis

as a function of a and concrete strength. They

v/h
gave the following expression for calculating the

cracking shear force: -

bw f 2. (
VI ft f+Af sin 9 S'3)
CQt+ cp s pe

is fcu
where ft equal to 0.55V
a
v/d
(8)
2.5.2.11. Borisanskij and Nikolaev at the NIiB
(Reinforced Concrete Institute) in Yoscow studied
the conditions for the formation of diagonal cracks
in beams. As well as I beams, prestressed T beams

were studied with flanges in the tension or the

compression zone. It was shown that diagonal cracks
in the web do not develop at the places where the
principal stresses, as calculated f; r elastic behav-
iour, are assumed to be greatest.
Gvozdev(8) in calculating the principal tensile

stress took into consideration the compressive stresses

acting on the horizontal planes near the points of
8PPlication of loads and reactions. By using the data
from Paduart's theoretical studies and the resultp
of the deformation tests on beams by Borisanskij and
Nikolaev, Gvozdev Proposed the following expressions
for determining local compressive stresses:

36
 ph-11-xf for h and x z--2y'
fyy y"Z- 0.4
:--7757 )

2h( Ye

fp--x 1 0.4
fo r0 y -21ý h and x'--'-0.8 h
yy 0.8 b-T -Y-
h'r 0.8h)
w (2.34)

P is the load, I is the distance

where value of point y

between the point on the beam from the compression

face and X'is the distance between the point on the

beam and the section 'where the load is applied.

Equation 2.84 predicts a zero value for fyy at

the mid of shear span at centroidal axis for a 2.0

v/dý
(d = 0.8 h)
9
2.5.3. Shear strength under uniformly distributed load.
(46)
2.5.3.1 Wilby and Nazir conducted tests on 5 post-

tensioned prestressed concrete I beams, 4 of them

with grouted ducts and all with unreinforced webs

to study their behaviour and strength under multiple
point loads. Thebeams were all simply supported

on sprins corresponding to L/d = 5.4. The beams were

152 by 304 mm overall. Only one crobs-section, with
'b
hf = 0.25 and effective depth = 254 =m,was
bw=0.33,
d
used, The prestressirig, force was developed through
4e-McCall bars, which were stressed initially to

413 N/mm2.
The only variable was the concrete

strength which varied between 28.0 and 58.0 N/=2

The modes of failure observed were the destruction

of the web either by splitting along the line of the
diagonal crack or by the crushing, of concrete due to

the arch action. It was found that the critical

37
diagonal crack crossed the neutral axis at an average
distance of 0.22Lfrom a support. Strain measurements

were taken with electrical strain gauges forming an

equilateral triangle at points on the centroidal axis

0.22 L from each support. Those measurements showed

that the discrepancy was quite appreciable when com-

pared with the theoretical values of the D.rincipal

tensile stress. This led 17ilby and Nazir to con-

elude that the theoretical principal tensile stress

as a means of predicting the initial diagonal cracking

load has no meaning. They also f6und that the princi-

pal tensile stress calculated from the measured strains

corresponding to the formation of the first diagonal

cracks did not compare with the tensile strength of

concrete. In this respect their results were in good

agreement with those obtained by Swamy(41) although

,
they pointed out that their experimental measurements

were susceptible to some slight amoqnt of error because

the strains at a point were measured along the three

sides of an equilateral triangle enclosing, instead of

precisely, at the point. Also they mentioned that

Owing to the short span, the portion undergoing

diagonal faiiure experiences some of the vertical

stresses due to the load points.

Although ! in number and
heir tests were restricted
the concrete strength %as the only variable, they der-
ived the following expression for web-cracking load
qcL = 4.94 feu + 61.3 MI (2.35)
1
38
 (43)
2.5.3.2. Kar carried out tests on 14 post-tensioned

prestressed rectangular concrete beams under uniformly

distributed load with simply supported ends. The spec-

imens were four beams 100 by 200 mm overall with L/d

= 15,3 and ten 125 by 300 mm overall with L/d = 13.1.

Ile found that the-critical section in shear-compression

type of failure lay at 0.315 L from the nearest support,

)
and this confirms Hanson and Hulsbos(4? observations

on two pre-tensioned I-beams under the same type of

loading who found the critical section in his I- beams

C>
at 0.33 L from the nearest support,
Kar gave the following relationship between the

point-load case and the uniformly distributed load

case for use in prediction of the shear-compression

failure load of the latter:
B-33 Vu (av/d)
qL (2.36)
u L/d

(48)
2.5.3 . 3. Arthur Bhatt and Duncan studied the
. ,
strength in shear of 19 pretensioneý concrete I-beams

with unreinforced webs under uniformly distributed

load with simply supported ends. The uniformly dis-

tributed load was applied using the fire-hose tech-

(47)
nique used by Hanson and Hulsbos and by Leonhardt
(61)
and Walther

Basically three different cross-sections were

employed. The specimens w6re 153 by 304 = overall

with b ranging from 0.416 to 0.500, hf/d = 0.187

w/b
and the effective deDth, d= 272 mm., All the bea=s

were pre-stressed by nine 5 mm diameter high tensile

39
steel wires. The value of prestressing force at test
f
ranged from 134 to 184 KN giving a variation in cp
2.
between 5.40 and 7.00 N/mm The concrete strength,

given as tensile strength derived from split cylinýer

2. L/d
value, ranged from 2.84 to 3.93 N/mm The ratio

covered ranged from 6.18 to 9.52.

It was observed that all the beams cracked on

a line through. a reaction. In 9 beams the ultimate

load was the same as or only slightly greater than

the web-craCking load. Thus Arthur et al regarded

the web-cracking load as the ultimate load for fail-

ure in shear with unreinforced webs. The following

semi-empirical expression was developed for predict-

ing the web-cracking load.
0
fcP) "bwa (0.1
qL=f" (1 + + 0.51f) (9.85 - 0.79
c ct f/ d
ct
(2.37)
The constant hf
and the coefficient of in the second
d
bracket were chosen arbitrarily as only one value of
h
f/d was used. Equation 2.37 predicts a zero value
for the cracking load at L/d = 12.47
which restricts
iis applicability at high values of Vd.

Arthur, Bhatt and Duncan in their analytical

approach, took the principal tensile stress theory

as a criterion for failure. A computer progra-mme

based on the. simple theory of bending of beams was
first prepared to. search the web for the of
position
maximum principal tensile stress. The results obtain-
ed showed the maximum occurring either over the suPoorts

40
 I
at 3-5= above the centroidal axis or at mid-span

depending on the value of the load and the length of

span. From this it was concluded that the simple

bending theory approach which neglected the effect of

local stresses due to the reactions gave misleading

information for maximum principal tensile stress in

value-and position. Then another approach was attempt-

-- ed based on elastic stress analysis which allowed

approximately for the effect of the reactions, and

vertical stresses due to the loading, in addition to

normal bending and shear stresses. The reactions

were treated as point loads on an infinite wedge.

By equating the resulting maximum principal tensile

stress in the web to the tensile strength of the

I
concrete, a reasonable agreement with their experi-

mental results was obtained.

2*6. Analytical Approach Using Finite Element:

.Any attempt at a detailed analytical stress

0
analysis over the whole loading range ofa prestr. essed
concrete beam has to recognise the non-linear behav-

lour of the beam due to cracking as well as non-

linear material properties.
The advent of computers
and modern methods of
(56)
analysis, such as the finite element method
(57-59)
led to attempts by some. investigators to

achieve the above objective.

An accurate analytical determination of the
deform-

displacements and the internal stresses and

41
ations in a reinforced or a prestressed concrete

structure throughout its load history is complicated

(58)
by a number of factors:

The structural system is composed of two materials,

concrete and steel.

2. The structural system has a continuously changing

topology due to the cracking, of concrete under
increasingqoad.
3.. Governing relationships and failure criteria under
combined stress states are difficult to obtain.
(See Section 3.6).

4. The stress-strain relationship for concrete is

non-linear and is a function of many variables.
(See Section 6.2.2. ).

5. Concrete deformations are influenced by shrinkage

and creep.

6. Deformations and stress are time-dependent on

load and environmental history.

7. The effect of dowel action in the steel reinforce-

ment, bond bet%een the reinforcement and concrete,
bond slip, and aggregate interlock at cracks are
difficult to incorporate into a general analytical
model.

Although no attempt has been made to include all

the above factors in an analytical'model, it seems
that there is a potential for this finite element
technique, mainly as a research tool, to explore the
behaviour of Yeinforced-and prestressed concrete beams

under the combined action-of shear and bending when

more information concerning the above mentioned

factors becomes available.

42
 2,7. Concluding Remarks:

The review of previous work on the shear strength

of prestressed concrete beams without shear reinforce-

ment shows that:

2.7.1. None of tne existing expressions is generally

applicable because:.

1. The geometric properties of the cross-section

(34)
treated differently. All but Sethunarayanan
-were
(44)
and Arthur ignore the flange thickness, hf, as a

variable. Sethunarayanan included the effect of

the flange thickness in his equation 2.17 in

terms of the clear distance between flanges. On

i the other hand Arthur used a constant value of the

flange thickness, hf, throughout his tests and he

included hf as a variable directly proportional to

V. in his equation 2.31 Sethunarayanan and

Arthur'S, - equations are o contradictory nature a

with respect to the variable hfo

2. All but Sozen et al(36)did not"include the effect

of b in their expressions.
/bw
3. The data recorded in the literature for concrete

strength is not uniform. A wide range of express-

ions is used for the tensile strength of concrete.
4. LTost of the tests were conducted at ratios
a
v/d
between (33,34,44)
2.5 and 4.2 except a few cases

which dealt with valu6s of a less than 2.5.

v/d
Sozen et al conducted tests on 99 beams, the

43
 majority of them with av/d between 3.2 and 4.2,, which
(45)
makes equation 2.18 of a restricted scope
2.7.2. Although only uniformly distributed load con-
stitutes a practical type of loading, we find that most
of the available test data relate to beams under one-
or two-point load.

2.7.3. The well-known failure criteria for concrete

subjected to normal and shear stresses cannot be

applied in the case of beams without being modified

to take into (25142)
account the affect of av/d*
2.7.4. A fully mathematical solution is not practic-
able in case of prestressed concrete beams subjected
to combined bending and shear.
2.7.5. There are gaps in the experimental results
available which indicate that a wider range of.
variables needs to be covered in a systematic way.

44
 I CHAPTER 3

STUDY OF THE PARAMETERS AFFECTING

THE SIMAR STREINIGTHOF PRESTRESSED
C01*-CRETEBEAMS AND CRITEMIA USED
IN PREDICTING FAILURE OF CONCRETE

3.1 Introduction:
The review in the foregoing Chapter showed that the

parameters which. seem to affect the shear strength of

prestressed concrete beams without shear reinforcement

are : the geometric configuration of the cross-section,

the concrete strength, the intensity of the prestressing

force, the position of the prestressing tendons and the

shear span.
Professor R. Walther, in a lecture on 'shear proble=s

in reinforced concrete beams' given on 31st October 102 U,

in the Department of Civil Engineering at the University

of Glasgow, disagreed with Table 5 of B. S. C. P. 110 *.part 1:

(30)
19? 2, which takes the shear stress as a function of the

longitudinal steel ratio. He claimed that the shear stress

was a function of the amount of longitudinal steel in excess

of that needed for bending, i. e. shear stress is a function

of (As prov. ). In dealing with prestressed concrete beams

(As req. )
the above ratio will have little influence on diagoral

tension cracking since the area of excess longitudinal

steel likely to be used is small compared with that likely

to be used in reinforced concrete beams. Hence that factor

will be neglected and factors mentioned above will be

discussed in more detail.

45
3.2. The Geometric Configuration of the Cross-section:

The effect of the cross-section configuration

on shear has been shown by test results

strength
(17 60-62)
obtained by several investigators I. Ferguson
(60)
and Thompson and Leonhardt and Walther(613 showed

that failure loads increase with increasing web breadth.

For instance, Ferguson and Thompson's beams Al and D2

differed only in web breadths, which were in the ratio

of 1.75 -. 1. The ultimate shears of these two beams

.
were found to be in the ratio of 1.79 1 1, i. e. the

ultimate shear strength was-directly proportional

to the web'breadth.
The effect of flange dimensions on the shear
(17)
strength was demonstrated by Taub-and Neville

on rectangular and T beams with comparable web

breadths. Similar tests on flange width were report-
(62)
ed by Placas and Regan on T beams with 150 mm

web breadth. It was found that the beams with

300 mm or wider flanges had about 20% greater ulti-

mate shear strength than rectangular beams. It was

-concluded that only the portion of the flange

immediately adjacent to the web could transmit a

component of the shear in the compression zone.

The effect of beam size was shown by Leonhardt

(61) (63) (64)
and Walther's ,
Kani's and'Taylor's tests.

Leonhardt and Walther concluded that in beams with

external dimilarity but constant bond quality, shear

strength was fairly independent of the beam size.

46
Kani, using beams differing only in depth, found that the

shear stress at failure decreased with increasing boom

depth. Koni concluded that as the depth increas6d, the

splitting forces due to the 'wedging bond action' of the

reinforcement increased, and as a result, the failure of

the two deeper series of Kani's beams involved splitting

along the reinforcement while the smaller beams did not.

Taylor showed that the size effect could be reduced if

the size of the coarse aggregates was changed in proport-

ion to the beam size. Taylor 'suggested a reduction factor

of 0.6 to the she.ar stress carried by a concrelte beam with

d>4.0.
b
3.3. Concrete Strength.

Shear does not directly cause the failure of reinforced

or prestressed concrete beams, but its effects appear as
tensile stress leading to diagonal tension cracks. Although
the tensile strength of concrete is not convenient for use
in design, as the compressive is the
strength usual prop-
erty specified, it sounds reasonable to relate the diagonal

cracking load to the tensile strength of concrete. The

values for the tensile strength of concrete reported in Ike

literature were obtained in a variety of ways.

-Sethunarayanan used the cylinder splitting value. Sozen et al
related the modulus of rupture of concrete to cylinder
compressive strength and used two-thirds of this derived

modulus of rupture as a measure of the tensile strength of

concrete. - Evans and Hosny took the xensile strength of

concrete as directly proportional to end Evans

47
 S

and. Schumacher used 0.72 of the tensile'strength obtained

from briquette tests. Arthur used values derived from

Dewar's expressions.
As the indirect tensile strength of concrete derived

from the split cylinder value tends to give more uniform

results than other types of tensile tests, it is used

in this investigation as a measure of the tensile strength

of concrete.
3.4 Prestressing Force:

As already mentioned the prestressing force induces

---, ---extra compressive stresses which reduce the-final tensile

stress resulting from shear. The position of the prestress-

ing force creates a variation across the depth of the beam

in the value of the prestress in the concrete. Since the

position at which the crack will open is unknown, the value

of the prestressing stress at that point will be taken to

be proportional to*the average prestress in the concrete,

f Also, as the value of the inclined tension cracking
CT) 0
load depends to some degree on the presence or absence of
flexural cracks, it is expected that the eccentricity of
the prestressing force should have some effect. So the

effective depth, d, rather'than the overall depth, h,

will be considered as a variable. Some investigator&42,52)

took h as a variable where no flexural cracks were expect-

ed .
3.5 The Shear Span:

The importance of this variable on the shear strength

has been emphasised by the majority of investigators. Thus

the effect of the bending moment and the vertical stresses

48
due to the relative positions of the failure plane, the

load and the reaction will be introduced by considering

the shear span, av

3.6 A Failure Criterion for Concrete.

The various modes of shear failure all involve

cracking or crushing of concrete under complex state of

stress as, e. g., when diagonal cracking, shear-compression

failure, splitting or web crushing occurs. Several studies

have attempted to determine which of the classical failure

theories is most applicable to concrete, but Goode and

(67)
Helmy state that none of these theories successfully

pred. icts its failure under all complex states of stress.

The strength of concrete under combined shear and direct

stresses seems to be well predicted by -the octahedral

(68)
stress theory The simplest cracking criterion is
0
based on the principal tensile stress or principal tensile

strain theories of failure. These approaches have been

shown to be useful in predicting tensile failure when

applied to certain simple state of stress bnd have

been shown to give reasonable results when applied to
the tensile cracking of
a reinforced concrete beamý under
(70)
Combined shear and bending moment Frequently the

stresses in a structure can be idealized to a biaxial

state of stress with the stress in the third direction

equal to zero as shown in Figure 3.1. Kupfer, Hilsdorf

(71) ('79)
and Rýsch ondat er, in 19 72, Liu, Nilson and Slate' *-

have shown that as f1 and f2 are varied, the element

shown in Figure 3.1. a will have the strengths shown by

the solid lines in Figure 3.2. Figure 3.2 is sy=etrical

49
 f2

11 I1

(ü) (b)
-

v 0- v
12t

Figure 3. fe, Element subjected to (a) normal stresses

(b) normal and shear stresses.
C

f 21fc at
f (i
E
0
U D-64
96 fzt

r . 60 ---

Rel. ( 71

---32

80 .96 h/fcu
-.16ý-32 -48
Tension Compression

Ir

Figure 3.2: Biaxial strength of concrete

f
7 on

Figure 3.3: Possible Mohr's pture

r, %A, envelopes for
concrete
V
-*V
fVf

V
A6 to-20 fcu
ýef-(251422 74-76)

4 -1 iII IIN0,1
Tension ft ---
-161Cu -32fCU *&SICU *64fcu OtOlcu Co-mpression
Figure 3.4: 'Failure of concrete under direct
combined
and shear stresses.
50
 0 it into three
about a 45 axis and may be divided regions:
M: biaxial compression, (II) biaxial tension, and (III)

combined compression and tension. The failure criterion

based on uniaxial strengths is given by the rectangle
indicated by dashed lines.

Several observations can be made: (a) In Region I

biaxial compression increases the compressive stress

at failure above the uniaxial compressive stress, (b)

In Region II biaxial tension has little effect on the

tensile stress at failure, and (c) Region III shows
that the combined compression and tension may appreciab-
ly reduce both the tensile and the compressive stresses

at failure. ý
(73)
Alternatively,, Mohr's theory of failure yields

acceptable strength predictions for either of the cases

(52996) (74)
shown in Figure 3.1. Parabolic and straight line

envelopes have been proposed for the family of Mohr's

circles representing failure conditions. Any stress

condition that corresponds to a Mohr: s circle that is

tangent to or intersects this envelope is assu=ed to

represent a failure condition.

Based on either the Mohr theory of failure or the
Kupfer, Hilsdorf and Rusch diagramS3 relationships have
(25,42974-76)
been derived, Figure 3.4yfor the strength

of elements str6ssed as shown in Figure 3.1. b. This

envelope, Figure 3.4, represents all the combinations

of shearing and axial stresses on the vertical plane of

the element shown in Figure 8.1. b which results in
failure of the element. The implication, of such diagram

0 51
.

is that the. -presence of shearing stres3es will reduce the

compressive strength of concrete(25). Combined shear and

compressive stresses occur in the compression zone of

beams except in regions of constant moment.

3.7. Principal Tensile Stress and Diagonal Tension Cracking.
(77)
Guyon found that the diagonal tension c.,racks

always coincided with the paths of stress trajectories,

but he did not show at which positions the principal

tensile was critical. MacGregor et al(51)regarded the

principal tensile stress at the elastic centroid as

(33,37,39,41952)
critical, whilst others believed that

this principal tensile stress should be related to con-

(42)
crete strength and the a ratio. However, when Kar
v/d
to FO. 8 feu
tried to relate the principal tensile stress
&33,34,86)
and to a using his own results and those of other
v/d
the plotting showed a considerable scatter even for the

same value of av/d* He obtained slightly better predict-

ions of the cracking load by the use of Mohr's failure

theory for concrete subjected to combined shear and

normal stresses after he introduced a correction factor

evaluated from the experimental data. This was a function

of a equation 2.30.
v/d,
Bhatt(53)by using a method which allows approxim-

ately for the effect of local stress. concentrations on

the normal bending and shear stresses in a pre-tensioned

I-beam, took the maximum principal tension in the web

as a criterion and this was found to occur at the lower

web - flange junction for the majority of cases analysed.

He concluded that the maximum principal tension cannot

52
be the onlS, criterion in the case of the diagonal cracking

of the web. Then he used the maximum principal tension

(68) fail-
and the Bresler Pister criterion'to determine

loads in
tests carried out by several other investigat-
ure
(34936,44) ýhat
ors. He found both criteria predicted

approximately equal failure loads and the Bresler -

Pister criterion always predicted failure at mid-shear-

span along the lower web-flange junctions. He also found

that-both approaches showed that the Ov/d ratio had an

important effect on the predicted failure load.

From the above discussion it emerges that the stress

conditions in the web of a prestressed concrete I-beam are

extremely complex. It is, reasonable to assume

-however,
that failure will occur when a failure criterion for con-

crete. is reached at some point in the shear span.

In the circumstances, a rational semi-empirical approach

based on dimensional analysis has been adopted.

3. b. Semi-empirical Approach Based on Dimensional Analysis.

-
By applying the dimensional analysis technique to

the variables that seem to affect the value of the

diagonal cracking load, the basic format of the express-
ion can be obtained; then regression analysis can be

applied to determine the empirical constants and their

relationship in order to obtain agreement with test data.

Referring to Section 3.1, the relationship connect-

ing the variables that seem. to affect the value of
the diagonal cracking load for simply supported pre-

stressed concrete beams under one- or two-point loading

is indicated by the equation,

53
 f (Vcl'fcpl b, bw, d, fLt, av, )=0 hf (3.1.1)
(78)
By applying Buckingham's Pi theorem equation
,
3.1.1. can beexpressed in terms of a complete set

of dimensionless products. Any product TI of these

variables has the following form: -

k3 k* k5 k7
Vckl fk2 bf bw4 dt fCjk6 av hýB (3.1.2)
cp

where kl, k
29 ------9k. are the exponent of
-s
a dimensionless product.
If F (= Force) and L (= Length) are the basic
dimensions, then the dimensions of the variables
in equation 3.1.1 are:
-2 -2
V=F, f, FLfFL
cp and bw b= hf
c ct
= av L

Applying Buckingham's theorem,

No. of variables =8

No. of basic dimensions =

No. of dimensionless ratios =

Then a complete of 6 dimensionless
set products
av/d
of the variables can be derived. -Since and
h the
f/d are seen to be dimensionless products,
ah f may be tentatively disregarded.
variables v and
Then thb dimensional matrix is

vfb bw df

001
11 -2

where each column consists of the exponents in the

dimensional expression for the corresponding variable.

In the above dimensional matrix, the rank is Uo and

54
the nuýber of variables is six. Consequently, there

are four dimensionless products in a complete set.

The equations cor: ýesPonding to the dimensional

matrix are:
k1 4- k2+ k60

2k 0 (3.2)
2+k3+k4+k5 -2k6

The matrix solution is:

VC fcp bv,
b d ct

31000 -2 -1

40100 0. -1
11
50010 -1 0

171
60001 -1 0

Acc'ordingly, a complete set of dimensionless

products is:

'IV hV
171 --Tjf II cf
1234
fet
'44
b bw
and 7F
5d6

By referring to Section 3.2, the following transform-

ation is made to achieve greater experimental control

of the variables: -
av 41 hf Vc
11 = 113/1-1 Jj
d2-63 d bwd r
ct
fcP b0= bw
, 14 371 /Il 615 and Ilra
rc-t 5 -zr

The justification of replacing f(IT 111121

55
by f (11ý, IIý I: [ý) is that the relationship
1 21
among 1-I 11 1121 -_-and 1)IC is unknown, all that is

known is that a relationship exists. A relationship

among 11 11 and 1161 implies a relationship

11 ý1121 1
1
among 1711 1'121 _--and 11 since the variables
6
Ill IT and 11 are determined by 1719
21 _, 6
J* If
IT and 176 and vice versa.
21 -, --.
Consequently there is a function Fo, such that

fcP h av, LW
P0 0 (3.3)
Ibf,
f .0 bw dd
ct
This may be written in the explicit form:

ve hf BVI ýw (3.4)
J?l bI
I -d
bwd f bw 7F d)
ct
where I'll is a functional notation.
Th e term bw -
be dropped because of its negligible
will
a-
effect on the value of the cracking load within the
(63979). 'Kani (63)
practical range of normal proportions
did not find any significant change in shear stress

when the beam breadth was changed 150 to 600 mm

from
0 (79)
and statistical studies by Iyengar et al on a
large amount of data showed no significant effect of
bw in the range from 0.25!; bw 41.0.
d --ff- -
Consequently equation 3.4 will take the more

specific form:
vc P2 fb9 hf av (3.5)
bd f-1-- EW ,
f dd
w ct ct

where F2, is a functional notation.

I
56
 I 4
_CIIAPTER
EXPERIMENTAL WOMý

4.1. Test Soecimens:

The experimental progra=e was planned to demonstrate

the effect of each of the dimensionless parameters shown

in equation 3.5 on the value of the cracking load, VC.

The grouping of the variables in the manner shown, in

equation 3.5 requires a wide variation in the cross-

sectional properties of the beam. As design formulae

should be the logical outcome of any research of this

naturej the test specimens were designe'd to resemble

as far as possible, those which could be used in pract-

ice. Accordingly h and b /bw were allowed to vary

f/d
systematically within the bounds of realistic values.
hfid was varied from zero to 0.33 and b from 0.25
w/b
to 1.00. The av/d ratio ranged from 1.25 to 6.00 and

the L/d ratio from 6.0 to 17.78. The overall cross-

section dimensions were 200 by 300 mm, and the spec-

imens were 3.25 m and 4.75 m in length; the latter

being'used with L/d ratio of 17.78, Seven different

cross-sections were tested under one-or two-point

loading, and five cross-sections, under uniformly

distributed load. Full details of the test SDecimen

.
cross-sections and geometric properties are given in

Figures 4.1,4.2 and Table 4.1

1
4.2. Materials:

4.2.1. Cement:

"Ferrocrete" rapid-hardening Portland cement and

"Blue Circle" ordinary Portland ce=ent were used; the

57
30 )O

0 7mmwire i2-5mm strand

Figure 4.1. Figure 4.2

Table 4.1
I 4. 1 1
Becm ti t2 bw Axio-3 lxlo-7
c c
ma rk mm mm mm MM2 MM4

A 37-5 1 2-S 75 3 3.44 34-97

B # 20.0 so 29-25 34.1-3
c S7-5 12 -5 75 38.44 39-63
D It 20-0 so 35-25 39-30
E 77-5 1 2.5 75 43-44 42.5 6
F & 20.0 so 41-25 42-50
G 00.0 00.0 200 60-00 45-00

Geometric properties of cross-section

of the specimens

58
 latter was used owing to difficulty in obtaining the

rapid-hardening Portland cement at one period.

4.2.2, Aggregates:

Mid-Ross sand and gravel were used for all beams

but six. When Mid-Ross stopped production, IVndford

sand was"used for beams D11 to D14. The maximum size

of the aggregate
0 used was 10 m. Both of these mater-

lals are of morainic origin, and the gravels are

irregular in shape.
4.2.3. Concrete !..ix, *

The mix was designed(80)to attain about 35 II/=2

cube compressive strength in 5 days. Slight differences

I in the mix proportions were required from time to time

to achieve this strength. Full details of mix proport-

ions and their properties are given in Table 4.2.
4.2.4., Prestressing Steel:

In ?2 of the beams, the prestressing steel was

seven 7 mm diameter indented (Belgian pattern) high
0
tensile steel wires. Plain wires of the same diameter
had to be used-in six beams owing to difficulty in

obtaining indented wires at'one period. At a later

stage 12.5 mm diameter strands were used with 13 beams

so
, as to cover high ratios of av/d and '/d and to see
the effect of different arrangements of prestressing
tendons. Figures 4.3 shows the stress-strain behaviour

and the properties of a typical batch of each type.

4.3. Fabrication of Soecimens:
(81)suff
4.3.1. The formwork Nvas designed iciently rigid
 TABLE 4.2

Concrete Properties and

Details of Prestress.

Beam ' I
S, _ C. F. I
Mix proportions f f f f Ty- Apsfpe
mark at um- cu ci cu ct
test p 29 at at pe
Max
in tý t 97ý2 of 4l
2 te- 121
d ays W C S G mm N/ M l N&

A1 26 40 1.00 80 2.50 20 88 54.5 431.4 51.3 3.50 7P 247.5

. . .
2 137 35 94 58.8 44.0 54.3 3.49 71 225.7
.
20 25 91 54.5 42.8 55.3 3.41 " 249.5
.3 .
4 29 53.8 45.5 52.8 3.12 " 247.5
- -
5 29 57.8 43.0 55.6 2.94 " 243.4
- -
6 42 54.0 40.8 59.0 3.43 258.3
- -
7 27 20 85 60.3 40.5 51.0 3.68 251.5
.
8 13 44 1.00 95 3.00 20 92 50.1 34.2 43.7 2.71 258.0
. . .
9 28 is of 25 91 53.0 34.5 42.8 3.15 246.5
.
10 27 it it 53.8 37.3 45.2 3.45 219.4
- -
11 25 40 1.00 1.24 1.86 35 92 52.6 38.3 40.9 3.08 1? 9.9
. .
12 14 44 1.00 95 3.00 49.1 31.5 36.5 2.59 197.0
. . - -

B1 35 40 1.00 1.24 1.86 45 90 50.2 36.0 46.3 3.36 71 206.8

. .
2 86 1111 of fr 50 96 50.3 37.8 50.5 2.86 " 215.9
.
3 35 35 92 59.5 41.3 58.0 3.15 " 209.1
.
4 27 35 85 51.3 33.0 48.8 3.30 " 214.7
.
5 64 45 95 38.0 46.4 3.26 " 164.1
. -
.6 49 40 1.00 80 2.50 30 90 37.0 52.1 3.30 194.6
. . . -
7 23 of It it 35 88 39.0 50.6 3.52 209.7
. -
8 28 It to 25 89 8.0 41.8 3.04 204.8
. --; -3
9 33 it to 30 87 54.8 47.0 52.8 3.01 165.0
.
10 19 It it 45 91 - 40.0 53.5 3.72 219.1
.
11 184 It it 45 89 - 41.0 53.5 3.51 "
. 229.9-
12 26 41 1.00 84 2.61 34.5 46.0 3.00 "
. - - - 141 7
.
13 18 of it of 29.5 41.5 2.90
- - - " 144.1
14 24 it It It 33.5 47.0
- - 2.76 1??. 5
15 13 31.0 41.4 2.91 184.0

60
 TABLE 4.2 (Cont'd)

Beam Age Sl-',. '*. F. f Apsfpe

Mix proportionp ff f Ty-
at um- cu ci cu ct
mark 28, pe
test p max.
l at at of
in t est test te- KN
I
dayE w c 2 2 nd-
S G mm N- N/=2 NIMM NIMM
on,c

C1 21 . 40 1.00 80 2.50 20 88 54.5 43.4 50.0 3.52 ?P 25?. 9

. ..
2 120 If 20 88 58.8 44. C 46.5 3.33 ?1 240.4
.
3 26 25 ' . 91 54.5 42.8 55.9 3.68 ?1 242.6
4 25 If - - 53.6 45.5 53.1 3.01 246.0
5 2? - - 5?. 8 43.0 51.0 3.29 241.?
6 42 30 89 54.0 40.8 59.0 3.43 258.4
.
-7 28 20 85 60.3 40.5 50.8 3.40 250.7
.
8 15 44 1.00 95 3.00 20 82 50.1 34.2 45.0 2.74 252.5
. . .
9 29 44 1.00 95 3.00 25 91 53.0 34.5 43.8 3.17 244. %,
. . .
10 23 11 If it - - 53.0 37.3 42.3 3.22 221.8
11 21 . 40 1.00 1.24 1.86 35 92 38.3 45.0 3.14 188.4
. -52.6
12 13 44 1.00 95 3.00 - 49.1 31.5 35.0 2.49 200.3
. . -
13 20 41 1.00 84 2.61 25 95 - 34.5 46.9 3.48 153.4
. . .
14 26 If it it 40 92 - 29.5 46.5 3.14 154.6
.
15 23 it to it - 33.5 47.0 3.45 194.9
. - -
16 14 it to fl - - 31.0 41.4 2.91 207.2.
-
1? 28 40 1.00 84 2.61 - - 30.9 37.2 2.70 2S 98.0
. .
18 31 34.0 45.8 2.89 95.?
-
19 36 It - 37.0 51.0 3.01 94.2
20 379 41 1.00 0.81 2.61 34.0 34.2 2.97 108.0
.

D1 26 40 1.00 80 2.5 so 90 - 37.0 48.6 2.94 ?1 224.9

. . .
2 13 IF I# If 30 90 - 39.0 49.4 3.05 " 22?. 3
.
3 30 to 35 88- 38.0 53.3 3.. lo " 188.9
.
4 34 is 30 88 54.8 47.0 51.5 3.30 176.9
.
5 726 45 91 - 40.0 45.0 3.55 187.2
.
6 611 It 45 89 - 41.0 54.1 3.52 188.6
.
? 28 40 1.00 84 2.61 30.9 37.2 2.? 0 2S i98.7
. . - - -
8 31 is It 34.0 45.8 2.89 95.9
- - -
9 36 37.0 51.0 3.01 95.2

61
 TABLE 4.2 (Cont'd)

Beam Age, Apsfpe

sl- C. F f, f Ty- fI
at Mix proportions um- cu ci cu ct
28 Pe
mark test max*
p at at Of
in test test ,
te- , ICII
days W1 CI j l I ndj
Sl G mm '/mml N/mný IT/mm2 N/=2
on

D10 378 .4 1.00 . 81 2.6 451 . 89154.81 34.01 34.21 2.9712SI104.8

Ila 14 " it 46.5 56.0 3.45 " 109.3
lit 30 if it it 43.5 54.8 3.29 " 108.2
12 22 1.00 81 2.5 46.0 61.3 3.34 " 119.9
.4 .
is 15 11 it, 34.2 1 46.5 2.52 1 if 116.0
14 14 1.00 81 2.5 36.9 43.9 2.88 " 121.4
.4 .

E 11 42 44011.001 1 91157.11 41.01 64.11 3.731711190.3

.02.50120 .
-2- 11 it ff It 15 92 53.2 41.8 52.? 3.1? " 160.8
.
3 11 ft it 20 85 56.5 42.3 55.? 3.3? " 224.2
.
4 11 it it 20 96 54.0 45.5 50.0 3.29 " 225.0
.
5, 8 it 30 90 55.3 40.4 53.1 3.1? " 245.9
.
6 -14 20 8? 53.2 40.8 50.0 3.22 " 2? 4.5
.
? 15 20 91 53.1 38.8 46.0 3.10 7P 242.4
.
8 138 25 90 54.0 4?. 0 57.2 3.55 233.0
.
9 20 40 1.00 84 2.61 30.0 40.8 2.93 ?1 189.3
. . - - -
10 20 11 if ft 37.1 42.1. 3.31 195.5
- - -
11 27 40 1.00 1
. 84 2.61 - - - 32.5 39.5 2.87 216.8
.
12 47 " of ff it ,31.7 43.5 2.70 194.2
- - -

F 11 33 ý4011.00 11.2411.86145 1 90150.2 1 36.0 1 47.51 3.141711187.7

.
2 34 it 50 96 50.3 37.8 49.3 3.30 193.5
.
3 41 35 92 59.5 41.3 54.9 3.30 220.7
.
4 21 35 86 51.3 83.0 49.4 3.08 242.1
.
5 62 45 95 - 38.0 52.6 3.03 137.4
.
6 21 41 1.00 81 2.61 30.0 42.3 3.10 " 178.9
. - - -
7 20 of it 37.1 40.9 3.03 " 188.0
- -
8 28 32.5 42.1 2.99 " 207.5
9 47 31.7 43.7 2.80 " 186.0

62
 TABLE 4,2 (Cont'd)

If lApsfpe
Beam Age Sl- P. f i ff
cu c . cu ct Ty -w
at . Mix -prop orti ons um- 2S
mark test pe pax,
p at at
in test test of KN
te
days
w c S 2 nd
G mm /4 N/mE N/mg r /mm
.1 on

G1 40 40 1.00 80 2.5 C 20 91 57.1 41.0 65.8 3.69 71 192.0

. . .
-2 13 15 . 85 56.5 42.3 54.5 3.45 224.4
3 14 15 . 95 53.2 41.8 53.0 3.64 139.8
4 12 20 . 96 54.0 45.5 53.2 3.30 213.0
5 7 30 . 90 55.3 40.4 43.4 2.71 223.0
6 12 90 87 53.2 40.8 49.? 3.48 252.0
.
7 20 20 . 97 53.1 38.8 46.0 3.11 7P 239.3
8 903 25 . 90 54.0 47.0 '54.4 3.94 230.0

Notes.

Mix proportions were by weight. W= water, C= cement, S= sand

G= gravel(10= maximum). Where the cement proportion is under-
lined the cement used was ordinary Portland Cement, otherwise
R. H. P. C. was used. C. F. = compacting factor.
2. f = compressive strength on 100 mm cubes at 28 days.
cu 28
(Average of 3).
.
f= tensile strength of concrete. (Cylinder split test).
ct
at' test
3.71 = Seven 7 mm indented prestressing wires.
7P = it plain . of
2S = Two 12.5 mm prestressing strands.

63
 2000

1750 Stra nds

Plain wires
Indented wires

o-. 1500
C*4
E
E
z

1250
z

V)

WOO

Tendon f fp2,10 Es
pu
type 2 KN
750 - N/mm N/Mm *n
. ..

7mm indent ed 159 0- 15 60 195

wires 1700
500 7mm ptain 15 90 14 so 20S
wires
12, -5 mm stra ndsi 1740 15 70 210

250

0 0-2 0-4 0-6 008 1-0 1 -2 1-4

2
STRAI Nx10

Figure 4.3: Typical stress-strain curves for tendons

used.

64
 I
and tight to prevent loss of mortar from concrete at

all stages and to maintain the forms in their correct

position, shape and profile.

Specimens were cast in pairs in a semi-long-line

process in rectangular steel moulds with timber web-

formers fixed inside to give the required I-sections.

Seven wires were used in the wire-stressed beams, two

being placed in the top flpnge to give the correct

distribution of Prestress, as shown in Figure 4.1.

Later on, two 12.5 mm diameter strands were used, as

shown in Figure 4.2, together with one untensioned

7 mm diameter wire placed in the top flange to provide

an anchorage for the end zone stirrups. The prestress-

I
ing tendons for the manufacture of a pair of beams
.
were tensioned and the two beams were cast at the same

time'.. using the same. mix. The prestressing tendons

were passed through the bearing platesIshown in Figure

4.41-of the prestressing frame and the end plates of the

forms; the former provided the reýction for the tension-

ing force. The anchorage was provided by the use of

split-wedge and barrel-type anchor grips. Before tension-

ing, the tendons were free of rust and were cleaned of

surface oil. No other reinforcement. was used except in

the end zones where 3 mild steel stirrups of 3.2 =

diameter were provided. These stirrups were designed

according to Marshall and-blattock's(82) formula given by:

At=0.02 A f (
PS Pi 4.1)
f Lt
sv
where At = total cross-sectional area of web reinforce=ent.

65
 I
.4

(a) JncR Anchor Pracket m] Jack 'F-c e,plate

(b)ea-, ' li, -.cIior-! )ra c I-,

etv. n] Fa cop la te

Figure 4: The M-n-ring Plotes of the Pro-strccsýýing '12,

cl
-.

66
 One stirrup at each end was extended to form a hook for

lifting the beam.

4.3.2. Tensioning aDnaratus:

Two alternative tensioning devices were employed,

one for the wires and the other for the strands. For

the wires a hand-controlled P. S. C. monowire jack oper-

ated by a motor-driven hydraulic pump with a delivery

2
pressure of 70 N/mm was used. A C. C. L. 160 KN -

Stress-0-Yatic Jack was used to stress the 12.5 mm

strands. This Jack operates with a calibrated load-

cell. Figure 4.5 shows the two tensioning devices.

4.3.3. Tensioninp, process:

The distance between the outer faces of the bearing

plates was approximately 7.3 m. All the tendons were slrd: %3ýLL
/stressed individually. in
The stress each tendon was
increased at a gradual and steady rate. The tendons

were overstressed by about 5% for two minutes to reduce

(83)
stress loss due to relaxation of the prestressing steel 0
Then the stress was reduced to the required level and

the tendon was anchored. After the anchorage of the

tendon, the force exerted by the tensioning apparatu6

I
was decreased gradually to avoid any shock to the tendon

or anchorage.

The prestressing force applied was checked by strain

measurements on the wires by means of 203 mm Demee gauge

readings on collars attache%,.'A to the wires. The strain

developed in the wires was taken as the difference in

strain readings, after the slack was removed and the

tendons were. locked. This was checked against extension

67
 Figurd 4.5: Tensioning Devices

(A hand-controll-e-I P., '-j. C. monowire jack ,Land n C. C. L.

160 KNTStress-o-1, Jack)
latic

Di mensi 0 ns dePe ndon Zhe

g co me try Of cr oss-sectic n\

01

C, (3
2345678 etc.

,Sh

Iii@
lom
iý 101 mm C/C
-I-
Q3- C

Appr ox imu t ely 1000 mm -F

_]_
d'epends or. becim I

Figure 4.6. Arr3nr7, e-',l,ent of for

gauge rarks, aý,n-surinp,
the concrete strain. (The gauge marl-s were to both
ý-Iued
siýes o-P the bonn).

68
 measurements taken for the wires by a. scale attached

to the wire and the jack. The calibrated pressure gauge

built in the hydraulic system and Vogtmeter were used

also to check the wire force. For the strands, the

tendon force was checked against extension measurements,

the calibrated pressure gauge reading and the load-cell

attached to C. C. L. device reading.

4.3.4. Mixing and casting:

The concrete was cast after the stressing of the

tendons. It was mixed in a pan-type mixer. The aggreg-

ates and cement were mixed dry, then water_was added.

Slump, and compacting factor were determined immediately

I
after mixing. The number of batches needed for a pair

of beams varied between 8 and 12. The batches were

plaýed in layers of uniform height through the beam.

.
Meanwhile control specimens were made with concrete

taken from all the batches in standard steel moulds

placed on top of vibrating table. Four 150 x 300 mm

cylinders and six 100 mm cubes weremade with each beam.

Compaction was achieved by vibrating the mould

by means of a Tremix vibrator bolted to the centre of

the mould base. The mould was supported on hard rubber

cushions. A poker was also used in the early stages

when one of the Tremix vibrators was out of order.

0
A small vibrating table was placed on the top of the,

specimen near the end zones to improve compaction in

those zones.
After casting was complete, the tops of the cast

specimelis were trowelled smooth. After 6 hours, the

69
 bolts on the sides of the shutters were loosened and the

specimens were covered with damped sacks.

After 24 hours, the side shuttersvere removel,.

Experience showed that the removal of the form sides

could be achieved without damaging the specimens by

attaching two steel plates of appropriate dimensions

to the surface of the timber in contact with concrete

near the ends of the form sides and then by jacking

against those plates through a bolt acting-in a nut

welded to the outside of the form.

The control specimens were then demoulded and

placed on the top of the specimens, and all were then

cured under damp sacking covered with polythene for 3

I
days.
. After 5 days, that
and provided a cube test
(30)
indicated that the required strength had been reached
I
the wires were released all together and uniformly
by an inviard. movement of one of the bearing plates.
0
Before and after transfer, readings7Yere taken on
Demec points which had been fixed to the specimens
for estimating the prestressing losses and to investi-

gate the transmission length as shown in Figure 4.6.

The results of these are shown in Figures Aeý and A. 2

in Appendix A.

4.4. Instrumentation. Loadinc, ADDaratus and Test

Procedure:

4.4.1. Instrumentation is required to give quantitative

results to test the theoretical involved.
work In

concrete structures due to local cracking and uncertaint.,;

70
 I
about the stress-strain relationship of concrete, the

instrumentation can hardly be as efficient as desired.

Despite this some instrumentation intended to reveal

the behaviour at different stages of loading was

undertaken. Load-deflection curves for the mid-span

were recorded for all the beams. Those load-deflection

(84)
curves were used in the way described by I forice

to. obtain an estimate of the prestressing force actually

present at the time of test whenever this was possible.

The shapes of these plots were not always sufficiently

clearly defined, so in the final calculations, the

effective prestressing force estimated from the surface

strain measurement was used. A typical example of

calculating the prestressing losses using C. E. B. -

(27) (30)
F. I. P. and B. S. C. P. 110 Recommendations are

shown in Appendix B.
4.4.2. All the beams were tested simply supported.
The majority under either-central point loading or
It
symmetrical two-point loading'in a 900 KN capacity
Olsen crew-type universal testing machine. Later
on, 23 tests were conducted on beams under uniformly
distributed load. The load was uniformly distributed

over the entire beam-span surface by means-of the

(47948161)
water-filled fire-hose technique These

specimens were tested at six span to effective depth

ratios ranging from 6.00 io 1?.? B. Typical testing

arrangements are shown in Pigure 4.? .

4.4.3. The specimen was set centrally in the testinky

71
(a) Unýer two-point loading.

(b) Under unifori. rily distrilýuted load.

(using water-Ifillel I -fire-hose technique)

Figure 4.?: Typical Tcstinf, - Arrangements

 P
machine on pads of wet plaster which accommodated

any irregularities of the beam shape. After the

plaster had set, the load was applied in suitable

increments. After each increment of loading, the

load was held constant while strains were measured

and any cracks marked. The magnitude of the increments

in loading depended on the development of the crack

pattern. The loading measuring apparatus was flex-

ible enough to follow up any change in the beam behav-

iour; the inclined cracking loads after vilhich the be-

haviour of the beam would be affected were easily

obtained from an autographic plot of the central

deflection against load for each specimen. Some

typical plots are shown in Figures 5.2 and 5.3.

In some cases where diagonal cracking was observ-

ed in one of the shear spans of the test specimen, the

shear span showing the cracking was clamped externally

by channels and threaded steel rods. By so restraining
0
the development of failure in the span already cracked,
the other shear span frequently showed diagonal tension

cracking. 'Shen the beam failed suddenly by a diagonal

crack in one of the shear spans, it was sometimes poss-

ible to make more than one test in the intact portion,
thus making extra useful results available.
During testing, the development of crack patterns

was carefully studied, the distance of the intersection

of the critical crack with the centroidal axis in the

case of a uniformly distributed load being measured

(see Figure 6.8 The load at the first inclined

73
 0
tension crack and ultimate load were noted. All the

results are shown in Tables 5.1 and 5.2. Photographs

were taken of each beam after the completion of the test.

Some typical examples of different types of shear fail-

ure are shown in Figures 5.1. a to 5.1. L and Figure 6.7.

For each beam, four 150 x 300 mm cylinders were

split and four 100 = cubes were crushed in accordance

with B. S. 1881; 'Part 4: 1970. A relationship between

I
the tensile strength of concrete, f derived from
ct,
cylinder split tests, and the cube strength of concrete,
f was developed and the following expressions were
cul
obtained:
f/=f2 (4.2)
et cu + 1.45
28
-cu (4.3)
or f/
ct
= 0.456 1f
and equation 4.2 was used in the final calculations.
Figure 4.8 shows equations 4.2 and 4.3 compared with

other equations for predicting the tensile strength

(27,30,36, 55)
of concrete against the"exPerimental
.
results obtained in this investigation.

74
 el to
Z)
In
0
0
.

-je
.
Z 4M)
r,
.

in
10 40 rý

CL 2 ei
U -, m to 2c Z C ll-
u
* *-
u »
U JU -
3: ýu Y
% w u-
-, + 0 c 44
km 41 -- 41 2 *111 2 Co
to 99 N
M m ft
0N Co 44 «c
. ' t4
ýi ty in 0
. .- *u
in
ei ei to 81 1
u ei

u ei

tn

a0

CQ
0

co
0 ob
to
in %
Icý4
0

0
0

0
U

. .0
ft

ý,' 0
0"

Wb ei

E-4

C) 00
to
co

a
V 0

j0
19 0
cz
 0
CHAPTER 5

DESCRIPTION of TESTS

5,1 Introduction:
Distress in shear begins with shear cracking,,

Two types of shear cracks are distinguished*. web-

(S2,85)
shear cracks and flexure-shear cracks The

cracks originate independently in the web

web-shear

and they may lead to the type of shear failure shown

in Figure 5.1. a. Flexure-shear cracks occur in

regions already_cracked in flexure and these shear

cracks are extensions of the flexure cracks and they

may bring about the type of shear failure illustrated

in Figure 5.1. b.

5.2. Development of the Shear Crack Patterns and the

Observed Ijlodes of Shear Failure:. 11
,
5.2.1. Shear-failures develoned from web-shear_cracks,:
.
In the majority of the tests carried out in this

investigation the shear crack started as a web-shear

crack. These web-shear cracks can be divided into

.
two types depending on the speed of their formation.

In one type the formation of the web-shear crack is

a gradual process and in the other it is sudden,

explosive and destructive.

The gradual formation of the web-shear crack

starts with very minute inclined web cracks in the

middle of the web without any appreciable charge in

the load - deflection curve being observed. As the

load inqreeses these minute cracks start to link up

76
Figure 5.1. a. Shear failure as a result of web-shear
crack followed by bond failure (mode of failure

classifiea as diagontil crack m-id bond failure).

vu=1.00 Vc Ex *
a
v/d = 2.0ý Pt* = 1.01
VC I Vc Calc.

Figure 5.1. b. Sheor failure initiated by a flexure-

shear crack (mode of failure cicnassified as shear-
comoression ).

a vu Vc ExPt-
v/d 2.0) 7c- 1.617 0.77
Vc Ceic.
P%ote
In the photographs the standard international form of
the date is used, i. e. Year/1"onth/Day.
See Pa,7e 82.
77
to form a continuous web-shear crack as illustrated
in Figure 5.1. c and the load sustained by the beam

suddenly drops as shown by the load-deflection curve

of Beam B7 in Figure 5.2. At this stage the principal

tensile. stress might be equated to the tensil. e*strength

of concrete. This was observed with I-beams at a

v/d
1.25. Depending on the quality of bond, further

increase in loading may lead to the formation of more

parallel inclined cracks and eventually web-crushing

which was observed to be the predominant mode of

failure in such cases as shown in Figure 5.1. c for

Beam B6 and Beam B? and other examples as given in

Table 5.1. The web-crushing failure. in itself is a

gradual process by which the'load decreases gradually

as the web starts. to crush as shown by the inclined

curve of upper part of the load-deflection curve of

Beam-B ? shown in Figure 5.2. In these cases the

ultimate load could be as high as 3,..7 of the first

inclined cracking load.
-6
The second type of web-shear crack appears
.
auddenly and without warning in the uncracked web-

zone traversing a considerable height of the web

and extending rapidly both ways followed immediately

by an explosive collapse of the beam along the inclined

crack. With sections and depending on the value of

fcp and the a ratio, the sudden formation of this
v/d
web-shear crack (diagonal tension crack) will take

place be.fore any vertical flexure crack appears in

78
Figure 5.1. c. Showing (a) Non-explosive
web-shear
crack formation (b) web crushing mode of failure
as shown by circles in 'B6' and IB? '
-
a vu Vc "Not
v/d = 1.25, 1.1? - 1.691 - 0.3? - 1.12,
vc Vc Calc.

Figure 5.1. d. Sudden explosive formation of a diagonal

tension cracý in an uncracked web zone. No slipping
was observed.
a f 21f Lu
v/d = 3.0, cu = 51.0 N/mm cp = ?. 52 N/mmý = 1.09,
VC
Vc Expt.
.Vc C-)Ico -=1.03

Note; The figures indicate the load (=2V) in kips at

which a crack had nenetrated as far as the horizontal
line indicated.

79
 p

the beam. This is shown in Figure 5.1. a for Beam C1

at av/d 2.0. However, with greater values of av/dI

i. e. a 3.0 flexure cracks first start to form in

v/dý!:
the flexure span and as the load is increased, flexure

cracks appear in the s4ear span in the regions of

maximum bending moment or a distinct flexure-shear

crack might take place at a distance from the load

point greater than or equal to the effective depth,

depending on the a ratio. While the flexure -

v/d
shear crack is gradually widening and proceeding
towards the point load, a sudden opening of an

explosive destructive diagonal tension crack may

I take place in the uncracked web. This extends from

1.

the support to the load point and sometimes results

in the destruction of bond between the concrete and

steel leading to immediate collapse of the beam.

Examples of these are shown in Figures 5.1. d to

5.1. f. In some instances it was observed that,

while the load was held constant for a few minutes

while readings were taken, this sudden explosive

diagonal crack formed, as shown by Beam E4 at

a 4.0 in Figure 5.1. g. This was also observed

v/d
by Sozen et al(36)in their tests.

5.2.2. Shear failures developed from flexure-shear

cracks.
The development of shear failures from flexure-

shear cracks was observed in this investigation in all

ranges of av/d :ýt2.0 depending on the cross-sectional

properties and fcp values. With the rectangular

80
 Figure 5.1. e. Sudden formation
explosive of a
diagonel tension crack in an uncracked web-zone.
Slipping of' wires was observed.

v f
3.0) 2 f 2.
d, cu = 43.7 I.,/nlr, on,J cp = 7.4_3 N/mm
Vu Vc Expt.
Vc 1.007 1.08
VC Caic. - 1.13

"19

5.1. f.
-.
oJe of failure in A10.
,
Sudden formation
of a diagonal tension crack
followed by bond destruction.
avf2f2
3*01 cu = 43.2 N/mm
dm and cp = 6.56 N/mm
VU Vc Expt.
= 1.0of 1.29.
vc Vc Calc.

81
Figure 5.1. g. Sudden formation of diLgonal tension
crack in E4 while the load was held constant for few
minutes.
a f2f
v/d cu = 50.0 N/mm and cp = 5.18 N/nh
v Vc Eynt.
0.96
-,
vc = i. cop VC Calc.

Figure 5.1. h.
L-ode of failul-e in A6. Explosive opening
oa-0 *JiaE.on,-d tension craciý originating from a flexure
crack, accompaniel by bond destruction and separation
oj.' the top flange.
avf2f2
d=3.0, cu 59.0 N/MM and cp N/mm

VV
'L
F'xOt.
1.02
Vu=Vc Calc.
cC
VVC
I. ote ct Expei =ental
VC Vc Cplculate] 6.5
C31C. using CDequation
-

82
 S

cross-section at av/d = 2,0, vertical'flexure cracks

appeared first in the flexure span. With increase in

load, a flexure crack might appear in the shear span

in the regions of maximum bending moment or a distinct
flexure-shear crack might take place at a distance

equal to or less than the effective depth (a = 2.0)9

v/d
followed by an appreciable drop in the load. Further

increase in loading would lead to a gradual widening

of the crack which in turn would start a process of

internal redistribution of stress between concrete

and steel resulting in strain concentration at the

(11,86,42)
top of the inclined crack At this stage
9
the crack would become sufficiently inclined and

start to extend downwards. The beam could either

fail very gently with crushing of concrete near the

point load or by the sudden opening and extension

of the flexure-shear crack both ways resulting in the

explosive destruction of the beam.. This is exemplified

I
by Beam G3 tested at av/d = 2.0, which also showed
some traces of crushing near the point load as shovm
in Figure 5.1. b. Similar behaviour and crack patterns
were observed with an I section A6 at av/d '= 3*01
2 2.
f = 59.0 N/mm and f 7.34 N/mm The diagonal
cu cp 1
crack opened suddenly with propagation along the level

of tension reinforcement. The increased shear force

pressed down the longitudinal steel and caused the

destruction of the bond between the concrete and steel
which led to splitting of the concrete as illustrated

83
in Figure 5.1. h. This case could be compared with

Beam A 10 with similar properties to Beam A6

except for f and f Beam A6 had a higher value

CU. cp-
of (f + f, ) compared with either A 71 A8 or A 10
cp ct
and this might explain why the diagonaý tension cracks

in these three beams formed independently of the exist-

ing flexure-shear crack. This behaviour is reflected

also in their load-deflection curves as shown in

Figure 5.3 for Beam A6 and Beam A 10.

With
a increasing to 4.0, f ranging between
v/d - cp - 2
5.21 and 5.77 N/m2 and f from 35.0 to 42.3 N/mm
Cu
diagonal tension cracks originating from flexure.
.
cracks were found to initiate shear failure in many

cases. The inclined crack showed a distinct tendency

to become horizontal toward the nearest support at the

level of the bottom web-flange junction. However in

some-instances, a series of hor#ontal cracks, which

seemed to extend from the inclinedcrack, developed

along the bottom web-flange junction. With, increase

in load, these cracks linked up and widened so leading
to loss of bond. This resulted in the separation of
the bottom flange from the web and eventually failure

of the beam as shown for type C beam in Figures 5.1A

and 5.1. j. The corresponding load - deflection curves

for Beam C 10 and Beam C 12 are shown in Figure 5.3.
It is possible that this horizontal cracking was
initiated by microscopic 'bond' cracking caused by the
drying shrinkage of cement paste, which induces high

84
Fi. -ure 5.1. i. Nlode of shear failure in C10. Diagonal
U
tension crack originating from a flexure crack led to
.,
the separation of bottom flange from the web and was
followed by bond failure.

a f 2 f 2
v/d = 4.01 cu = 42.3 N/mm and cp = 5.77 N/mm

vuvc
Expt.
1.001 - 0.90
v v

Figure 5.1. j. ýtlode of shear failure in C12.

Diagonal tension crack developing from a flexure
crack and extending along the bottom web-flange
junction leadine to the separation of the bottom
flange from the web. The 'arching' eff"ect and the
crushing of the web near tne reaction was observeJ.
a f 2 f
v/d = 4.01 cu = 35.0 N/mm and cp = 5.2'1 N/mm2.
VU Vc Expt.
= 1.001 = 1.00
Vc Vc Ca1c.

85
internal'tensile stresses at the mortar-aggregate

interface Beam C 11 in Figure 5.1. k. shows

an example of the
effect of reducing f from 5.77
cp
2,
N/mm Beam C 10 case, to 4.90 N/mm2 on the initiation

of shear failure compared with Beam C 10. In the case

of Beam C 11 the diagonal tension crack formed inde-

pendently of any flexure crack and its behaviour

afterwards with increase in load was similar to that

described aboVe for Beam C 10 and Beam C 12.

5.2.3. Shear failures developed from secondary

inclined tension cracking.
What is described as a secondary inclined tension
(36)
cracking was observed in Beams A 12, B1 and F 1.

With Beam B, 1 tested at av/d = 4.0, an inclined

tension crack opened suddenly in the vicinity of the

support at a shear force of 89.9 KN, and a series of

short inclined cracks followed i=ediately thereafter

along the upper web-flange junction. With increase in

load these short inclined linked

cracks up and widened.
0
Then this damaged shear span was strengthened by clamp-

ing it by channels and threaded rods to force shear

failure on the other shear span and the test continued
from this stage. a shear force of 44.5 KN a non-

explosive inclined tension crack opened in the vicinity

of the reaction and propagated just beneath the upper

web-flange junction towards the load point and the -

shear force dropped to 35.6 XN. Further attempts to
increase the load led to the formation of an inclined

crack extending from the support to the load point

86
 followed by crushing at its upper end*near the upper

web-flange junction and the formation of a horizontal

tension crack in the top flange near the middle of
the shear span. The load then dropped to zero. A

load-deflection of behaviour of Beam B1 is shown in

Figure 5.2.

With Beam A 12 tested at av/ .d ý-' 3.0, an inclined

tension crack. opened suddenly in the web at a shear

force of 60.1 n just over the reaction and propagated

aloq,, y the upper web-flange junction. This was followed

immediately by the formation of another inclined tension

crack in the uncracked part of the web in the same

shear span which extended along the upper web-flange

junction into the flexural span and backwards along
\-I the lower bottom junction to the
web-flange. support

as shom on the L. H. S. of Figure 5.1. L. The load-

deflection curve is shown in Figure 5.3. This damaged

shear span was clamped and the*tesý was repeated to

force the shear failure to occur in the intact shear

span. A similar behaviour and crack patterns were

observed in the unclamped shear span, but at a shear
force of 55.6 Ell. Some crushing was observed in the

upper web-flange junction as shown on the R. H. S. of

Figure 5.1. L.

87
Figure yode of shenr ftailure ill C11. Diagonal

doveloped indepenýeiat of anýr fl e:K-ure

-ick
crack and extending along-, týie bottom web-flanEe
Junction separating the bottom flange from the web.
,
No slipping was observed.
>
C:

f2f
4.0, cu = 45.0 Nlrml aný cp 4.90 14/i-am
vu

Vc r.
1.07 '>T)t '- = 0.97
VC Vc Cale.

Figure 5.1. L. Shear failure in A12 as a result of

secon, ary inclined tension cracking.
VU. Vc Expt,
vc - = 1.001 Vc Calc. 0.97 - 1.05

88
 250

291
0- Appearance flexure crack.
3 of
i
LOAD Diagonal tension crock formation.
ZY) f7% Ultimate faitu I oad
KN shear re
222 G Web. crushing
(A ta different Web -distortion
scale Of load. HorizOntcl-tension crack
Ing us shown In the upper flange C arcoing
2 beside It ef f ect
U7 W1 res- st I ppi nq

Z 150

04

100

n.,
VA
so
Beam I cl
fcu fcp v cý) I cý/d
N0 WnM2 WMM2 mm mm
1
BI 46-3 S.'tS9
S-S9 106C 630 4-0
B4 48-8 7.05
7-05 530 1690 2-0
B7 50-6 6- 45 330 597 1-25
11
89 52-6 5-64 1190 570 4-5
Blo 53*5 7,49 930 $90 3-5

0-0 2.5 S-0 7*5 10-0 12-S is-0

CENTRAL DEFLECTION IN mm.

Figure 5.2 Load-deflaction curves for type B beaLa,

89
 *

200

150

-0

loc

060 2.5 Soo 7.5 10.0 12-5 15-0

,CENTRAL DEFLECTION IN rnrn

Figure 5.3: Typical load deflection curves

(Types A and C beams)

90
 TABLE 5.1

Experimental results under

one - or two-point loads.

Beam av (1-2a d a f f Vc vu Mode

V, v C"t cp vu
mark KN KN of
MM Mm mm d I vies 1 .
vieo EaST vves t' fail-
2 V
N/mm2 N1= c , ure

Al ? 95 1360 265 3.00 3.29 7.40 87.1 87.3 1.00 DCI*VM

- -
0 of 86.7 86.7 1.00 DO
-
2 930 890 It 3.50 3.39 6.75 64.5 77@,C 1.20 DCAM
- - -
3 530 1885 it 2.00 3.42 6.33 77.8 ? 7. E 1.00 DO
- -
810 It it 76.4 76.4 1.00 DC/BF
-
it it, 0 It 7.46 101.2 101.2 1.00 DO
- -
4 it 1890 to 3.34 6.81 1012 104., r 1.03 DC/V,D
- -
IT if 820 It to It It 81.4 81.4 1*00 DC
- -
it it 0 it It to 94.3 90.1 100.1 1 . 11 DC/"",
',])
-
5- to 18vo to it 3.42 6.67 87.8 87. E 1.00 DO
- -
it to 1355 it of It 94.5 94.5 1.00 DO
- -
it It 830 is it If 7.28 102.3 95.6 109.0 1.14 DO
-
It it 0 it It If 98.7 98. -1 - 1.00 DO
-
6 795 1155 3.00 3.56 7.34 80.1 93.4 1.17 DVTRýS'
- -
7 1160 it it 3.27 7.52 79.5 87. C 1.09 DO
- -
it 0 to it It It 74.5 87.9 87.9 1.00 DO
-
8 1355 to 2.47 7.43 80.3 80.3 1.00 DO
- -
it it 0 it It it It 75.6 75. E 1.00 DO
. - -
9 It 1160 If 2.98 7.37 79.0 80.1 1.01 DC/BF
- -
it ft 0 it It of to 79.4 79.8 1.01 DO
- -
10 it 1160 11 11 3.05 6.56 82.3 70.1 82. 1.00 DCAM
-
11 930 890 It 3.50 2.91 5.38 53.4 58.5 71.42 1.33 DO
if 0 to of of 55.6 64. r%. 1.16 DO
-
12 It 890 it to 2.75 5.86 60.1 55.6 60.1 1.00 DC/, WD

Bl 1060 630 265 4.00 3.10 5.89 38.9 44.5 44.5

- 1.00 DCAIW
TFS
it 3.26 7.07 47.4 48.9 48.9 1.00 DCAWD
-
3 530 1690 2.00 3.52 6.79 66.7 71.2 - 76.1 1.07 DC/SCC
11, 630 if it of 74.5
- - 74.5 1.00 DC/SC#C
4 1690 it it 3.19 7.05 66.2 67.7 80.1 - 1.21 DCPM
It It 630 to it It 7.34 83.4 67.7 63.4 - 1.00 DC/VtT
5 11193 1 560 If 4.50 3.28 5.61 40.0 40. C 1.00 DC/BF
_- ,- . ,
91
 TABLE 5.1 (Cont'd)

ý-2, f1t fcp VC Vý Y ode

Beam av aý d av Vu
c of
mark N/mm2 NIMM 2 KN KN f ail-
mm mm mm
Vý u re.
East West East West
1
B6 330 2090 265 1.25 3.31 5.99 80.1 77.8 - 117. C 1.50 DCAM
it It 1430 It It It 93.4 89.0 - 149. c- 1.68 DCA'.'C
7 2085 It 3.26 6.45 90.1 90.1 - 122.1 1.36; DC/'.i-D
- /BF
1200 of 102.3 85.1 - 111.2 1.31 DC/ViD
TF
J
to It 597 It of It - 86.2 - 145.4 1.69 DCAVC
8 2085 265 1.23 3.23 6.30 78.1 78.1 - 99.5 1.27 DCAM
/BF
It It 1445 It it is 83.7 100.4 117.0 1.17 DC,,'iVL
-
9 1190 570 4.50 3.30 5.64 37.2 37.2 1.00 DC/BF
- -
10 930 890 3.50 3.23 7.49 55.6 55.6 1.00 DC/B:-
- -
it 400 680 1.50 It it 71.2 84.5 84.5 1.00 DC/B1
-
11 ?30 1190 2.75 3.36 7.86 57.8 57.8 1.00 DC/Bz-
- -
It it 0 it 11 of 58.9 62.3 62.3 1.00 DC
-

C1 530 1890 265 2.00 3.24 5.77 87.3 87.8 clam 87.8 1.00 DC/BF
It it 530 it it 6.71 102.3 87.8 pýýd 123.2 1.40 DCAY1
2 330 1890 1.25 3.11 5.94 93.7 122.3 151.2 1.62 DC1,111
-
to It 715 It It 129.0 129.0 189.0 1.47 DC/V,'C
-
3 1060 810 4.00 3.45 6.31 60.0 68.9 1.15 IF
-
4 It $1 3.35 6.40 57.8 66.7 1,15 F
-
5 530 1890 2.00 3.27 5.66 77.8 77.8 1.00 DC/M
- -
1358 It It it it 87.0 85.4 87.01 - 1.00 DC/B1
It, it 830 it of 81.0 96.7 1.19 DCAI
- -
It It 0 it It it It 93.4 124.5 1.33 DC
- -
6 320 2090 1.25 3.33 6.46 120.1 129.0 137.9 1.07 DC/'.'#"i
-
It 1430 It It it 111.2 111.2 1.00 DC/81
- -
of 1100 It It to 109.0 133.5 1.22 D'C/=' 1
- -
It It 640 It It 142.4 133.5 169.1 1.27 DC
-
7 if 2090 of 3.26 5.87 101.2 115.7 133.5 1.32 DC/B]
-
It 1430 It 100.1 113.4 1.13 DCI*;rj
- -
/BF
it 770 120.1 127.7 209.1 1.64 DC/`.V(
-

92
 TABLE 5.1 (Cont'd).

Beam a -2 V d a f fcý VC v vu Yo de
v n u
KN of
nark fail-
N/me NIMM2 East West ure
MM mm d East -West
VC
-08 530 1890 26.1- 2.00 3.061 6,24 74.9 74.9 1.00 DC/BF
1355 82.3 - 82.3 - 1.00 DC/BF
640 - 94.5 - 98.4 1.04 DCAO
9 330 2110 1.25 3.01 5.72 105.2 105.2 ll?. 5 1.12 DC/VM/
-
BP
It it 960 11 124.6 124.6 - 195.8 1.5? DC/Vv'C
10 1060 630 11 4.00 2.96 5.77 - 52.5 - 52.5 1.00 DC/BFS
11 111 IT It 3.06 4.90 51.7 48.9 - 52.1 1.0? DC/BFS
lo it to 2.? 0 5.21 51.2 52.8 52.8 1.00 DC/VTD
-
17 1350 0 225 6.00 2.73 2.55 27.8 40.9 1.47 DC/VD
- -
18 0 3.04 2.49 25.6 40.0 1.56 DC/BFS
- -
19 1180 490 5.25 3.27 2.45 31.1 54.5 1.75 SC
-
D1 530 1690 265 2.00 3.19 6.38 66.7 66.7 - DC
--
#I If, 630 " it It It 71.2 66.7 86.5 - 1.21 VC
DC11.
2 1690 265 2.00 3.21 6.45 68.9 68.8 1.00 DC
- -
3 795 1230 3.00 3.35 5.36 38.9 45.6 45.6 1.00 DC/BF
-
4 570 1190 it 4.50 3.29 5.02 36.1 36.1 36.1 clam 1.00 DCAM
iý ped
5 1060 530 4.00 3.06 5.31 43.9 clamp 43.9 1.00 DCAM
6 795 1230 3.00 3.39 5.35 51.2 td 51.2 1.00 DC/WD
7 1350 0 225 6.00 2.73 2.80 35.0 23*4 35.0 1.00 DC/TFS
8 It to It If 3.04 2.72 23.4 26.7 1.14 DCAM
-
9 1180 490 5.25 3.27 2.70- 24.5 36.1 1.47 DC/'11D
-

El 795 1360 265 3.00 3.74 4.38 70.6 71.2 75.9 1.07 DG
it 415 it If* If 72.3 - 72.3 1.00 DC
-
it 530 0 2.00 85.6 81.0 85.6 1.00 DC
-
2 IT 1890 3.33 3.33 69.1 73.6 1.07 DC
- -
If It 1240 It III It 75.8 75.8 1.00 DC
- -
If it, 0 it it If - 69.9 - 75.8 1.08 DC
3 it 1880 it 3.37 4.88 86.2 86.2 1.00 DC
it It 1240 Ir it it 86.7 86.7 1.00 DC
- -
680 it it it 5.16 88.4 98.4 153.5 1.74 DC/'I-'C'
-
1060 830 4.00 3.29 5.18 56.2 57.8 57.8 1.00 DCAVD
-
.45 I 'VID
it It 3.17 5.6 68.9 68.9 1.00 DC/
-
5,90 530 2.00 84.5 92.3 92.3 1.00 DC

93
 TABLE 5.1 (Cont'd)

ý-2a, )' ,t VC i VU
Beam a d af
v c fcp Mode
v Tu

mark KN KN of
Fail-
MM MM mm d N/mM2 2
East West East West ure
VC

E6 79Z 1360 265 3. OC 3.22 6.32 92.3 80.1 92.3 - 1,00 DC

7 IT 11 of TV 3.10 5.58 - 66.7 - 66.7 1.00 DC
tI it 448 it TV IT 11 68.9 - 68.9 1.00 DC
It it 0 IT IT TV 72.1 - 81.0 1.12 DC/IND
a 1060 630 IT 4.00 3.52 5.32 - - 68.9 - F

Fl 106C 680 265 4.00 3.14 4.55 41.4 29.7 41.4 1.00 DVTFS/ý
BF
2 635 3.21 4.69 41.1 39.5 41.1 1.00 DC/".VD
3 530 1690 2,00 3.41 5.12 64.5 66.7 98.5 1.48 DC/',7C
-
It 400 IT it, 5.35 ? 1.2 75.9 105.3 1.39 DCANC
-
4 IT* 1690 11 TV 3.28 5.58 66.7 66.7 93.0 1.39 DC/WC
-
IT, it 630 if IT IT 5.87 71.2 79.0 87.2 1.10 DC/1"01C
-
5 1200 570 11 4.5 3.33 3.33 31.1 32.8 1.05 DC/TFS
- -

Gl 530 1890 1" 2.00 3.80 3.04 126.8 126.8 1.00 DC

- -
It it 595 IT IT, IT It 136.6 136.6 - 1.00 DC
-,
2 1880 IT it 3.40 3.74 144.6 151.6 1.05 Sc
- -
3 1890 265 2.00 3.34 2.00 84,5 136.0 1.61 SC
- -
TV of 880 2.00 11 97.9 129.3 1.82 SC
-
it TO 0 it TV It 2.83 122.3 122.3 1.00 DC
- -
4 1060 830 11 4.00 3,35 3.55 75,3 F
- - -
5 530 1890 2.00 2.94 3.73 108.4 108.4 1.00 DC
- -
IT IT 0 IT to IT 3.73 120.1 120.1 1.00 DC
-
6 795 1355 tv 3.00 3.23 3.88 94.5 98,0 1.04 SC
- -
IT 530 0 11 2.00 11 w 111.2
- --
IT It 0 fI IT IT 4.20 160.1 165-0 1.48 SC
- -
7 400 2150 11; 1.50 3.26 3.39 174.1 BF
- - -
8 560 1790 It 2.00 3.39 3.83 147.7 p
- -

94
 TABLE 5.1 (Cont'd)

Notes.

f
f cu Equation (4.2).
ct -+1.45 -
28
2. Mode of failure:

BF bond failure (observed by the slipping of the

tendons after the formation of the diagonal
tension crack).
BFS bottom flange separation.
DC diagonal tension cracking.
F flexure failure.
SC shear compression.
SCO shrinkage crack opened.
TFS top flange separation.
WC web crushing.
WD -web distortion.
followed by.
Vu
Ultimate failure load
V
c Di a ona l crac ki ng l oa d
,

95
 TABLE 5.2

Experimental results under

uniformly distributed load.

Beam b fI f qcL- qut qut Mode

w -d Ct cp -
KN KN of
mark d 2 'r fail-
mm MM mm I/mm N/mmC qct
est East West ure

B12 50.0 265C 265 10.0 3.09 4.48 153.5 133.4 - 181.3 1.36 DC
13 212C 8.0 2.93 4.? 3 166.8 180.1 - 235.? 1.31 DC,/'.'IC
14 193C ?. 28 3.13 6.0? 200.2 220.2 - 314.9 1.43 DCAID
15 159C 6.0 2.93 6.29 231.0 209.0 - 275.8 1.32 DC/T:-ý

C13 75.0 -265C 265 10.0 3.13 3.79 199.0 267.0 - 1.34 DC/SF
-
14 212C 8.0 3. *ll 3.86 222.4 235.7 - 315.8 1.34 DC/TK
15 193C 7.28 3.13 5.07 266.9 289.1 346.9 1.20 DC/dD
-
16 159C 6.0 2.93 5.39 275.0 275.0 360.3 1.31 DC/
-
TFS
20 265C 225 11.78 2.6? 2.81 152.3 187.9 1.25 DCAIM
- -

D10 50.0 2650 225 11.78 2.67 2.97 126.3 154.6 1.22 DCAM
- -
lla 58.0 4000 17.78 3.45 3.10 114.0 114.0 1.00 DC/V;D
- -
llb 3.41 3.07 100.1 115.6 146. 8 1.27 DC/TT
12 2650 11.78 3.64 3.40 160.0 129.0 222.4 - 1.39 DCAVD
13 3930 17.47 3.11 3.29 91.2 103.4 - 1.13 DC/!WD
-
14 3.02 3.45 104.5 1.06 DCAM
- - -111.2
E9 ? 5.0 2650 265 10.0 2.91 4.14 213.5 - 271.1 1.27 DC
10 2120 8.0 2.95 4.32 240.2 246.7 346. 9 1.41 DC/TT
11 1590 6.0 3.10 4.99 311.4 311.4 378. 1 1.21 DC/TT
12 1930 7.28 3.00 4.47 298.0 275.8 299.8 1.01 DC/BF

F6 50.0 2650 265 10.0 2.98 4.12 177.9 169.0 249.1 1.40 DCA-VD
? 2120 1" 8.0 2.91 4.42 213.5 213.5 258.0 1.21 DC/iM
-
8 1590 6.0 2.? 5 5.03 235.7 231.3 342.5 1.48 D%lf/
-
TFS
9 7.28 3.01 4.57 200.1 204.6 35.7 1.15 DC/
11930 -
TFS

Note.
Symbols for mode of failure are as shown on page 95
TT = test terminated.

96
 0
5.3. Prediction of Shear Failure Type:

The above analysis of the occurrence of shear

failures and the photographs provided show that, within

the range of variables covered in this investigation,

the final mode of shear failure and hence the ultimate

(4)
shear failure load could be affected by (a) the

locaiion of the diagonal tension crack in the shear

span, (b) the. position of the upper end of the inclined

crack with respect to the compression face, and (c) the

quality of bond. The first two factors can be consider-

ed to be functions of the cross-sectional-properties,

the av/d ratio, the prestressing force and the concrete

strength. Bond, the failure of which is characterised

by the longitudinal splitting of concrete along the

tendons, is an indeterminate quantity because of the

many factors affecting its quality.

Although all shear failures in I-sections origin-
ating from flexure cracks took place in the higher
4
range of values of av/d and all the shear failures

initiated by the simple diagonal crack (web-shear

crack) occurred in the lower range of values of a

v/dI
the two categories overlapped in the middle of the

range. Web-crushing following the formation of a

diagonal crack was observed to be the predominant mode

of failure for I-sections at lower ranges of a

v/dI
i. e. a ý 1.25, provided that there was a good
v/d
bond between concrete and steel; but it was also

97
observed'that the web-crushing also-took place at the

upper range of av/d accompanied by 'arching' effect.

From the above discussion and the photographs

shown in Figures 5.1. a to 5.1A it is clear that. the

,
value of a at which the change from one type of shear
v/d
failure to the other depends on several indeterminate

factors in addition to the properties of the beam.

Thus it is not. possible to predict with any certainty

what will follow the formation of a diagonal crack

from a knowledge of only one variable, namely a
v/d*
I

98
5.4. ComDarlson between the Shear Crack Patterns

observed under Uniform Loadinz and Point Loading:

The photographs (Figures 5.1-a. to 5-1-L ) show

that under point loading the diagonal tension crack

generally forms in such a position that the upper

end of the crack points towards the point of applicat-

ion of the load. The louer end of the diagonal crack

extends in the direction of th-- support, and this

extension may be in the form of a nearly straight

line continuation of the diagonal tension crack as

shovm in Figures 5.1. al 5.1. c, 5.1. d. 5.1. e and

5.1. g. In contrast, in some cases, the inclined

crack continues down to the lower web-flange junction

or to the level of the tendons and thence along the

lower web-flange junction to the support, as shown

in Figures 5.1. b, 5.1. fl 5.1. h2 5.1A, 5.1. j. and
5.1. k.

All the beams tested under uniform loading in

this investigation showed similar shear crack patterns

4
to those observed under point loading and the crack

pattern seems therefore to be unaffected by the

arrangement of loading. This similarity in behaviour

and mode of failure imp'lies that there should also

exist some relationship between their shear cracking

loads, and tnis will be discussed in Section 6.1.3.

Under ýniform loading it was observed that more

than one diagonal tension crack may form in either
half of the beam as shown for Beam B 14 in Figure

99
 I
6.7. b; some of these cracks may develop from flexure

cracks as shown for Beam D 14 in Figure 6.7. e. These

two cases are similar to the cases of Beams A12 and A10

as shovm in Figures 5.1. L and 5.1. f respectively. This

crack pattern can also be seen in the photographs of-

(48)
Arthur et al, i

-The critical diagonal tension crack for all the

I-boams tested under uniform loading in this investigat-

ion formed on a line through a reaction. In this respect

they showed behaviour similar to that of the reinforced
(61)
rectangular beams reported by Leonhardt and Walther

and to that of the prestressed I-beams tested by Arthur

(48)
et al None of them showed the shear-compression
(43)
mode of failure observed in Kar's prestressed
rectangular beams. Thus it may be concluded that as
.
well as the span to effective depth ratio, the magnitude

of prestress and-the web breadth play an important role

in d6termining the mode of failure under uniform loading.

100
'I

Islotet Readem are rcminded of V.10 footnoto to page x vhcra It is stated

that when expresalons containing the to= 1OOOV are used in

Chapters 6 to 8, these expreusions wIII only be dimensionally

2
correct V
vhen e in expre3sed In kll ff in V/rmi , zml othox! 1.11
ct
dimerislons in =.
 CHAPTER 6
,
ANALYSIS OF TEST RESULTS

6.1. Prediction of the Diagonal Tension Cracking Load.

Diagonal tension cracking loadand ultimate

shear failure load.

Unaer some ranges of variables, the ultimate shear

failure load, for a beam without shear reinforcement,

was equal to the first diagonal tension cracking load,

and the beam could either fail to sustain further

increase in loading or collapse simultaneously with

the fomation of the diagonal tension'crack. For a
different range of variables it might exceed the crack-
ing load by a substantial amount as shown in the second

last column of Table 5.1. Since the magnitude of the

ultimate shear failure load depends on the final mode

of shear failure and the latter was shown in Section

5.1 to be a function of many indeterminate variables,

it seems necessary to limit the useful capacity in shear

of a beam without shear reinforcement to the first

Qs
diagonal tension cracking load, and will be

defined as the shear force at which the diagonal tension

(36)
crack will start to affect the behaviour of the beam

Prediction of the first diagonal tension

cracking load under one- or two-noint loading.

As mentioned in Section 4.1, the experimental

progra=e was plpnned to bring out the eff6ct of each

of the dimensionless variables shown in equation 3.5

re-written below:
v (fýt
C--L F2 fcP hf av (3.5)
bwd fclt = Ib '6;
. dId)

101
 I the format-
where V corresponds to the shear force at
c
ion of a diagonal tension crack which develops when
the diagonal-tension stress (principal tensile stress)

exceeds the tensile strength of the concrete. The-

effect of the stresses normal to the longitudinal

axis, which are usually ignored by present codes of

(30,32)
practice is taken care of by the variable
,
(av/d). Thus equation 3.5 is basically a principal
tension criterion.
Figure 6.1 shows the relationship between

1000 Vc/(N df1 and f It for four cross-sections

IcC t) cplf
and one value of av/d for Despite the scatter
each.
in the test results which may be due to the random
occurrence of voids and flaws resulting from imper-
feet compaction, a linear relationship could be

assumed to exist in which the ordinate increases as

f /fc't increases. This could be formulated as:
CP
1000 VC fc1o
A+B
ft bd
et w
where the values of A and B for a given cross-section
depend on the value av/d.
of Values of A and B depend
also upon the way in which the tensile strength of
concrete is evaluated, as the latter can be determin-
ed in a variety
of ways. The ratio A/B was taken as
(36) (40)
unity by Sozen et al Evans and Schumacher
(44),
and Arthur when the tensile strength of concrete
was taken to be approximately equal to f, So, by
Ct.
taking A/B as unity, 6.1
equation could be modified

102
 1.5

0
0

0-5 1-5 2-0 A-S

1-0
fCP/ fc,t

If I
Figure 6.1: Effect of f upon Vc
cp ct

1.0

V..

-u

lou.

0
0

II

-250 -26 6 -364 -667 -800

i/! Ov/d
Pigure 6.2: Effect, of a upon VC
v/d
(Results for type B beams)

103
 I
to have the following form:

1000 VC F3 hf (6.2)
, _"ý_ = .0
ff cp db W-)
Ct(l +f Trc-t

where P3 is a functional notation.

The test data obtained from type B beam over
av/d in Figure
a wide range of ratios were plotted
6.2. These test data demonstrate that cracking shear
force, VC, is inversely to a
proportional v/d ratio.
Figure 6.3 shows a linear relationship between
0 and h av/d.
f/d for a given value of For'a'given
h
value of f/d, the value of 0 increases as b/bw
increases. This leads to the assumption that the
flange projections contribute to thq value of
They may be related in the following form:
0C+D hf (6-3)

in which the second term on the R. H. S. is proportional

to the area of flange projections. Both C and D

depend on av/d. Figure 6.4 shows this linear relat-

10
hf
ionship between 0 and (b
d

for a given value of v/d.

Therefore the final semi-empirical expression
for the diagonal cracking load can be written as:
1000 vc (1 ; gct L
f
bwd bw d) (6.4)
ct 0
the ratio C/D being taken. constant for simplicity.
Using a least-squares method based on the data
plotted in Figure 6.4 the value of D$was found to be
1.50. Similarly Bland Cowere found to be 0.10 and

104
 1.0 For alt b=2 00 mm tr Clv/d = 2- 0
bw
.
6 56 mm
75
0
x 200

For bw = 50 mm 0

,+,
-.
or 02
Z-ý Cm --0

, 0-5
6u
For bw= 75 mm
U-
0

For bw =50175&200 mm

0-0 0-1 0-2 0-3

h f/d

Figure 6.3: Effeet of h f/d on Vc

1-0 For alt bl= 200 mm & av/d m 2.0 -

bw
01, so mm
75
x 200 Pf 0.356 + 0-240(-h- -1 )2L
b« d
f or bw = 50 9 75 & 200 mm
Z.

Ci. - ýOO
u- For bw=SOMM --. 00
N- I .-
+ 000

For bW= 75 mm

0.0 0.2 0.4 0.6 0.0 1-0

b hf
bw d
Figure 6.4: Effect of hf/d and b /bv; on V.

105
0.31 respectively from the data plotted in Figure

6.5. Accordingly equation 6.4 may be written as:

1000 vc
0.10 + 0.31
(i if! ýF]
fb + pt d[l. 5 +f v/d
ct w
(6.5)

The ratios of the experimental val. ues of VC to

the values predicted by equation 6.5 to all the test

results obtained in this investigation under one -

or two-point loading gave a mean value of 0.98 with

standard deviation of 0.09.

Figure 6.5 shows that the curve tends to be

horizontal av/d, i. e. the rate

at higher values of
of decrease of W tends to be zero at higher values
av/d. for
of This shows that, the specimens whose

failure in shear was not initiated by flexural

cracks, the effect of stress normal to the longitud-

inal axis due to the load and the reactions is either
becoming constant or diminishing at higher values of
a This *is illustrated by Figu3ýe 6.6 which shows
v/d.
the variation of the stress normal to the longitudinal
axis, fyy, just prior to the formation. of the diagonal

tension av/d
crack with ratio at different levels of
f fj
cp/ ct for type B beam assumed to be in the centroidal

axis in the vicinitY of the mid shear span. These relat-

ionships were derived from-the tension
principal
equation. If the principal tensile stress, ft, is
taken as positive, elastic theory gives

106
 4A C- 00
vi ID
c 4ý
t4

OA cr
01 10 w0%.
C%l
C
0
0
El
co
'A
+w
%
Ul rc)
ci
Cl)
ro
ci
%0
-jol
0

ca

1
1-1

+1& & \to 10

4'1ý - %r
-Q A, >
-->
co

7/ or

-1 +a
H o
toco
.0

1 'tl% ro
1,
-6
or
I, N" r-4
- C-4
4J ca

Ln
. 01,
00, El o
00, V)
(D
010
00

*0 0

U)

rq
Ol 4-)

0
0
mq
--)+Got) pmq WDA )131
q 0
3A 0001,

107
 %0

I, W

r:
0

U ro
4-)

+1

C14
to
a
00
;4 El
Q

om
Zo
0
4-) 4J

U')

CL

%oo
0
F-I-r4

rX4
j
9; tz -0
e31 xx
aAissaidwo: )

108
 fcp + fvv fCp
ft24 ++v -2 xy max (6.6

where v I bw
xymax

Putting f= fl and rearranging, equation 6.6f

t ct
can be written as:

f±-3Z V2
xy max
=
f ct
.4 f ý2 fcp/fct)
C41 +

V
where xy max is the maximum shear stress just prior
to the diagonal cracking and is determined from the

value of Vc given by equation 6.5 Thus Figure 6.6

.
shows that equation 6.5 pr&dicts, for a given value
fcp/fcit, fyy
of nearly constant values for at values
a
of v/d>, 3.0. Figure 6.6 also shows that the use of
the principal tension equation taking fyy as zero may
lead to an overestimate of the value of Vc at higher
af. f.
values of v/d especially with low levels of cp/ cý.
This is indicated by the fact that Figure 6.6 gives

tensile f av/d
valu'es for yy at higher v4lues of and
f fl
low values of cp/ ct.

Under certain sets of material and cross-section

a
variables, some specimens at high v/d ratios exhibit

shear failures initiated by flexural cracks, as

illustrated in Figures 5.1. b and. 5.1. h to 5.1. j.

In such cases, any reduction in the diagonal cracking

load increase in av/d
with may be attributed to the
influence of the flexural cracks, as they reduce the

stiffness of the specimen. The load-deflection

curves for such specimens confirm this, as they show

109
 I
a definite reduction in slope, indicating a reduction
in stiffness beyond a point corresponding to the

occurrence'of flexural cracks. This is shown for

Beam B9 in Figure 5.2 and for Beam C10 in Figure 5,3.
(30)
This was taken into account by BSCP 110: 1972
032)
and by ACI (318-71) .
Both of these codes give
two expressions for calculating Vc, one for sections

uncracked in flexure and the other for those cracked

in"flexure. However, despite the general indication-

that is likely to occur with a

shear-compression v/d>
3.0 the f
strand-stressed I- beams-with cp ranging
between 2.45 and 2.81 N/mm2 tested av/d
at of 5.25
and 6.0 all failed by diagonal tension cracking, the

experimental value of the diagonal cracking shear

force being accurately predicted by equation 6.5

as shown in Figure 6.5 by circles. But as a design

expression should be the logical outcome of researcli

of this nature, a general solution for diagonal crack-

ing taking tte 0 flexure

into account effect of cracks
is felt to be desirable. For design purposes this

could be done empirically by assuming the rate of

decrease ý increase a to be constant
of with in v/d
to the a
and equal rate of decrease-at v/d = 3.0
for all values av/d> 3.0. This is justified
of
because for beams tested a
all at v/d = 3.0, the

ultimate shear failure load resulting from a sudden

major diagonal-tension crack was sensibly the same

as that at the formation of the inclined'tension

110
.
crack (uS'ually flexure-shear crack), as shown in the

second last column in Table 5.1 tangent to the

.A
in Figure 6.5. av/d follow-
curve at = 3.0 gives the

ing expression
1000 vc av
= 0.31-0.034
ýP- bwI dL bfd f
-rfct1 .5 4(
ct t)

av- > 3.0 (6.7)

for
d

Equation 6.7 is shown in Figure 6.5 and its limit-

ations will be discussed later on when it will be

compared with other published experimental results.

Prediction of firstdiagonal tension cracking
load in a uniformly loaded beam.

6.1.3.1. General:

This project was extended to cover the uniformly

distributed load case, as this constitutes a practical
type of loading, in order to discover whether the

results already obtained in the case of one - or two-'

point-loads could be applied in the case of the uni-

formly distributed load,
In this investigation all the uniformly loaded
beams cracked on a line through a re'action as shown
in Figures 6.7. a to 6.?. e and the distance from the
support to the point of intersection of this inclined
tension crack with the centroidal axis was found to-
bear some relation to ýdyatio,
as shown in Figure
G. B. Figure 6.9 shows the relationship bIe.tween the
distance from the support to the failure in
section
the face as reported (42,43),
compression by Kar Hanson

ill
 TTI
Figure 6.7: She,,ir C'rack Patterns unier Load, -*

_ni-form, _nr-
n
(form, ation of a Jiagonal tension crack on a line thxou. ý7h

a re2etion in a unifor: mly loadeý prestressed I-beam)

Ir
omppFlAwli4w
!ic
'iou
I

I I

Figure 6.?. a: Beam Ell, L/d = G. C

Ii 7-,

112
Figure (3.7 -c '. ý/j
E 10 7 = 8. C)

Figure 6.7. d: Bo,-, n C20, t/d = 11.78

I.,igure G. 7. e: Beam D14, L,/3

= 17.47

11-3
 B-type

X C-type
6
D-type
a E-t y pe
0 F-t ype
A
MIS
,

040
15 17 19
579 11 13
Figure (6.8): Relation between L/d and the distance
from the support to the intersection of the diagonal
crack with the centroidal axis.

;K Leo nha rdt

Walthe r

0a0Kar
0-3 0000 0 Hanson
0 Hutsbos

0 -2

41414,4 4 444 4-
-4
0- 1

0.-0
13 Is 17 19
S79 11
Figure (6,9): Relation between L/d distance
and the
from the suppok to the fbilure section in the
compression face.
114
 V

(47) (61)
and Hulsbos and Leonhardt and Walther . Leonhardt

and Walther's results were obtained from ordinary

reinforced concrete beams which cracked on a line

through the reaction, but Hanson and Hulsbos' results

and Kar's were for failbres caused by flexure-shear

crack. The latter had an average value of o. 33Lfrom

the nearest support.

From Figures 6.8 and 6.9, it may be concluded

that, for beams that crack through a reactiong the

position of the critical section is a function of

Vd,
while for thoseln which failure is initiated by
,
a flexure-shear crack the failure sections in the

compression face will occur at a distance of about

0.33L from a support..
6-i. 3.2. Prediction of the diagonal cracking load..
qcL, distributed load from
under uniformly
the one - or two-point loading results.,
(MV)cinstead
By using the moment-shear ratio term

of av in 6.5, then the laAer become:

equation will

(0- 10 0031 (6,8)

VK + FV-dt
a
(+ r bw df
where K= f' 1 C-P +(bw 1L
at fl at) 1000 -9 -) df]

VC is assumed to be the shear force at the critical

section, i. e. at a distance; h from the nearest

ct
support of a simply supported beam with a uniformly
distributed load. If qc is the load length
per unit
I
at cracking, then
qct (1
VC 2X (6.9)
2 - C)

115
 t2
Me c
2c
XCL
Im N=
andFVd)c d1 2Ac

By subs tituting equations 6.9 and 6.11 into equation

6.8, we will have:

qcL 0.31 (1 2ýc)

2ý K 0.10 + -
2 L/d Ac (1 -A c)-)

or
qcL 0.10 0.31 (6.12)
2K 2Ac + Vd ýC (1

Then the section of' least shear resistanc-6-according

to equation 6.5 will be obtained by differentiating

equation 6.12 with respect to)\c and equating to

zero, so we get:
dqj (1
2-2 0.1 0.31 2ýc)-)
2k . I/dA2 0
dýc (1 2Xc) (1 >ýc r-)
c)
or
)? (1 ý)2 (l )3
0.2 L/d c =A0.31 c
c -2
After rearranging we get:
3+ 2
ýc 112.40 (i
+ 2)Xc +

9.80ýc 1.55 0 (6.13)

L/d 77-d
-
The solution of the above equation will give the
Xc in
value of which will result a minimum value

for qJ in equation 6*12. The solution of equation

6.13 is obtained using the Newton-Raphson iterative

formulaPg) Thevalues ofý c for different values for

116
 PA
-r IA
0 r: ro
9.1
E
E I cu

I gib

2 Zý 0 %- 0
I
0
0 jc
c
X

0 (3 0

0
%-,

C) ri
co 00

1ý: u E-4

M
0
.H

\ý
9) (D
4-J

:J
0
EIM G)

'm
ýý u
4-4 ý4

ri 0

4-3 Co 13
Cb

XX x c)
P

X
G) (0
r4
1

-0 c)
gn cý 1,5
ew 0 Co 44
i;
c1) -ý
to co p

orl V CO

44 C)
12
booo&

117
 ý/d in Table 6 .1 below:
ratios are shown

Table 6.1

U/d Xc M) qc
Vd
c
K
6.00 244 360 L/d 9508
. . .
6.18 243 358 Yd 9345
. . .
7.28 233 334 Vd 8511
. . .
8.00 228 323 Vd 8080
. . .
8.42' 224 316 Vd 7859
. . .
9100 221 308 ýd 7590
. . .
9.52 218 301 Vd 7366
. . .
10.00 216 298 L/d 7182
. . .
13.00 200 267 L/d 6310
. . .
ýd 5911
15,00 192 252
. . .
L/d
20.00 175 222 5224
. . .

The last 6,1, qJ is in,

column of Table plotted
K
Figure 6.10 together with the present experimental
(42943,47,48)
results and those reported by other investigators

6.1.3.3. Development-of anexpression forpredicting

q
ct using the exogrimental results of the
uniformly loaded beams.

An alternative approach to the attempt above, to

adjust the results for point loaded beams to suit the

uniformly loaded case, is. to attempt to develop an

expression for the uniformly loaded case entirely

independently*

Starting with an equation similar to equation

6.2 one can write:
0

118
 q hf
1OC30 22- (6.14)
V
)bw bIV d
(1
+4 ;PP_
-fet
ct
where F4 Is a functional notation. Equation 6.14

can be rewritten as follows:

qcL hf
1000 1) (6,15)
,=A bw d +B
(1 &P
-fc't +f fct) bw d
h
(bw the
where the term A 1) reflects effect of
-d
the flanges of an I section on the value of-q CL&
W can either be a constant or a function ofý/d
Yd.
and B*is a function of
Figure 6.11 shows the relationship between

1000 qcý hf
and for different
f (1 bw d
ct + bw d

L/d. The straight lines indicate that

values of
different, values for A and B apply for the different
ý/d
ratios.
6.12 shows the plot of A and B derived
Figure
ý/d.
from Figure 6.11 in terms of Because of the
L/d,
scatter of A- values with no general trend for
L/d
A with respect-to can be formulated. Thus an

average value of 0.? 5 for A- values was chosen. The

value of B can be represented by the equation:

6.74
B + 0.36 (6.16)
L/d

Thus equation 6,15 will become:

1000 qj b1 Lf 6.74
- - 0.36 + 0.75 + L/d
fl (1 Lfc6 d
at + bwd
(6.17)

119
 2.5
t1d

0 10.0
-------8.0
7.25
2-0 El 6.0

a 1-5 00
u

vu

1-0

0.5

00 92 .4 s6 06 1-0
( b/bw- i)h f/d

I! igure 6.11: Effect of beam properties on diagonal

cracking load.

2-0
0-8
9 -. AI OC)()q ct )h qd +A 5)
- =A( b/bw-i
f -1* fr bwd
CFj( p/icl t)

8=6.74 + 0.36
I/d

i4o

A 0.75

-0.0 111111-
56 10 12 14
I/d
Figure 6.12: Diagonal cracking load and beam properties
in terms of Vd.
120 *
 v

6.1.3.4. CO"MDarison between equations 6.12 and 6.17.

Equations 6.12 and 6.17 were used independently

to the the load, qcl-, for.

predict value of cracking

all the tests in this series and the results published

(42,43,47,48)
by others The values predicted were

compared witn the corresponding experimental values

and the results are shown in Table 6.2. The mean

value of the ratio of the experimental load to

calculated load and corresponding standard deviation

are also shown for each set of data in Table 6.2.

)
There is little difference between the values

predicted by the two equations, and this confirms

that all the. assumptions Made in deriving equation

6.12 from equation 6.. 5 were reasonable. Accordingly

equation 6.12 can be used for predicting the diagonal

load, qCL, in the

cracking case of a uniformly
loaded beam.

6.2. Prediction of Shear-compressi-on Failure Load:

6.2.1. General.

Equations 6.5 and 6.12' predicted the first

diagonal tension cracking load reasonably well for

all the beams in which shear cracking was not

influenced by the presence of flexure cracks. At

higher a the the

v/d ratios resistance of web section
to the formation of major. diagonal tension cracks is

high because the breadth, bwj is very thick as in

rectangular sections or because the prestress level,

121
 TABLE 6.2_
v

Comparison of results of tests on

uniformly loaded beams with values
predicted by diagonal tension equations
(6.12) (6.1? -)
and

Author Beam L qC L' ExT)t.

Ratio
mark d
q0L Cale.
Eqn. 6',,.12 Eqn. 6.17

Present B12 10.00 1.05 1.07

Investigat-
ion 0.91 0.93
13 8.00 1.00 1.03
1.08 1.11
14 7.28 0.95 0.97
1.04 1.07,
15 6.00 0.98 1.00
- 0.88 0.91-
C13 10.00 1.06 1.09
14 8100 1.05 1.07
1.06 1.08
15 7.28 1.01 1.03
1.10 1.12
10- 6.00 0.92 0193
0.92 0.93
20 11.78 1.27 1.29
D10 it 1.08 1.07
I
11a 17.78 1.10 1.04
llb 0.98 0.92
1.13 1.06
0
12 11.78 1.10 1.10
0.90 1189
13 17.47 0.90 0.84
14 1003 0.96
E9 ,
10.00 1.05 1.07
10 8.00 1.01 1.04
1.04 1.07
122
 TABLE 6.2 (Cont'd).

Author Beam L qcL

Ratio Expt.
mark d
q LC
c alc.
Eqn. 6.12 Eqn. 6.17

Present Ell 6.00 1.00 1.03

Investigat-
ion 1.00 1.03
.
12 7.28 1.16 1.20
.
1.08 1.11
F6 10.00 1.06 1.07
1.00 1.01
7 8000 1.09 1.13
1.09 1.13
8 6.00 0.95 1.00
0185 0.91
9 7.28 0.95 0.99
0.97 1.01

Mean value of ratio 1.021 1.034

Standard deviation 085 089
. .

Hanson and
Hulsbos(47) F17 10.58 '1.07 1.08
1.11 1.12

Arthur C9 9.52 0.84 0.86

(48)
et al 10 7.28 0.95 0.96
1.05 1.07
12 9.52 0.80 0.82
13 8.42 0.88 0.90
0.93 0.95
14 9.52 0.90 0.93
1.06 1.09
15 8.42 0.77 0.79
0.90 0.92
16 7.28 0.92 0.94
.
0.97 0.99
.
 TABLE 6.2 (Cont'd)

Author I Beam L
Ratio qcL, Ex-pt.
mark d
qL- Calc.
c
Eqn. 6.12 lEqun. '6.17

Arthur C17 6.18 0.85 0.86

(48)
et al 0.89 0.91
D6 8.42 0.97 0.99
1.03 1.06
7 9.52 0.93 0.95
8 7.28 1.12 1.14
9 6.18 0.87 0.87
0.87 0.87
10 6.18 0.88 0.88
11 8.42 1.02 1.05
.
1.07 1.09
E5 9.52 1.17 1.21
6 8.42 1.09 1.11
7 7.28 1.28 1.29
8 6.18 1.18 1.18
9 8.42 0.99 1.01
1.18 1.20

Mean value of ratio 0.98 1.00

Standard deviation 126 128
. .

Kar(43) A-U-2 13.10 0.68 0.? 3

-3 0. ?7 0.82
-4 0.71 0.76
-5 0.66 0. ?l
-6 0.? 4 0.? 9
-8 tt 0.? 4 0.? 9
-9 0.73 0.79
-11 0.64 0.69
15.30 0.71 0.82
-2 it 0.59 0.69
-4 0.74 0.86
-5 0.58 0.67
Mean value of* ratio 0.69 0.76
Standard deviation 0.061 0.061
124
fcp, is very high. In such cases, for example that

beams high or av/d

of rectangular at relatively
L/d (42,43) in is
ratios the ultimate failure shear
,
frequently shear-compression initiated by a flexure-

shear crack. In this type of failure it is difficult

to define precisely the first diagonal tension crack

and hence the load at which it occurs. In the follow-

ing sections an attempt will be made to calculate the

shear-60=pression failure load which in turn will be

compared either with equation 6.5 or with equation

6.12 as appropriate. It is recommended that the

lesser be taken as the limit of usefulness in shear

of a beam without shear reinforcement.

To achieve this. object, it is necessary to have

full-knowledge of the stress-strain behaviour of

concrete in compression and to find a simple but

appropriate failure criterion for concrete in the

compression zone. 0
6.2.2.. Equation for the stress-strain curve of
concrete in COMDression..
,
The stress-strain curve for concrete in compression
is adequately represented by a parabola3 either termin-
0
ating at a maximum stress f0 corresponding to a strain
to, beyond this
or continuing point at a constant
to the maximum, up to a limiting 6u.
stress equal strain
The exact geometry cfthe stress-strain curve is

dependent'on a large number of factors, the most

125
 0

significant of which are the concrete mix, the strain

(90)
rate'and duration of loading and the strain gradient
(91,92). im-
across the concrete In design methods it is

practicable to take proper account of the effects of

the rate of loading and the st -rain gradient upon the

stress-strain characteristics, and therefore most

of the expressions given for the stress-strain curves

are in terms of concrete crushing strength alone. The

stress-strain curves for differing concrete strengths

are generally, for convenience, taken to be geometric-

ally similar and are expressed in the form:

f1f (E /E (6.18)
0)
0=F5
where F is a'functional An example notation.
of
5
(93)
thia type is given by Desayi and Krishnan as:
[1
f1f (el E 2]
2t/Eo + 0) (6.19) *
0=
As it is normal in design to consider only the

rising branch and the part over which the stress is

approximately constant, the idealized stress-strain

(30)
curve given by BSCP 110: Part 1: 1972 will be

used here. This curve is shown in Figure 6.13

and can be represented by the following equations:

-ýcu 2t
f 5.5 x 103 11.29 106 for =to
f 0.67 f Fo for 0.0035
cu
Eo If -C-U-,
where = 0.244 x 153 (6.20)
0
6.2.3. Failure criterion for the-shear-comDression zone:
The strength of an element in the compression zone
depends on the normal flexure stress fxx, the shearing
stress vxy, and the stress normal to the longitudinal

126
 Co

OH

0,0--
C)
9.40
0 cri
2 CH %--*
u C\i
(1) r-
,V)zr
, >m

x 0H
%r H
13 -p
u) P. 4
le
0

0 C(I 0

-P
to

4 r-i
cö
>

*0
cri
3
vi

(D
b
tto
OH
p4

127
 it

axis f The first two of these stresses are related

yy .
to one another by the configuration of loading so

that their effect can be taken into account by the

effective moment-shear ratio(X-ýV, )C *

Full consideration o-f the strength of concrete

under a complex state of stress i's outside the scope

of this investigation, and further discussion will be

based on the Mohr failure theory. Mohr's theory

(25,69)
has been found to be valid for biaxial stresses

and this is generally the state of stress in the

compression zone-of concrete beam. -So in-order to

_-
establish the relation between normal and shearing

stresses which, when acting together, cause failure

to occur in the compression zone, the shape of Mohr's
. failure envelope should be known.

Various suggestions about the shape. of the envel-

ope have been made in the past, and this is discussed

in Section 3.6. In this investigation a straight line

envelope to M-ohr's circles of failure, based on pure

tension and pure compressive strength circles was

adopted. By using the geometric relationships derived

from that envelope together with the principal stress

equation, the following expression was derived for the

failure criterion in the compression zone.

2 fxx ývy
+ ft
t ft
21
(6.21)

128
whereYis the ratio between uniaxial compressive and

uniaxial tensile stress, i. e. ý=fq C/

The derivation of equation 6.21 is. given in Appendix E.

Shear-compression failure is observed either at

high av/d the failure loads
values of where are not
f
high enough to produce any appreciable value of yy

or in rectangular sections with moderate shear spans

f
where the value of yyj if it is assumed to be inverse-

ly f
proportional to b, will be small. Neglecting yy
in equation 6.21 may thus lead to-a small error on
the safe side, and equation 6.21 will be reduced to

the following form:

6- fxx (fxý) 2
+U (6.22)
(1+02 t

The value of ý can be obtained-from equation 4.2 as:

1 fcu
:rc 22.4
= (6.23)
:rt fcu -1- 40.6 N/m=m
Ia
04

and f' c is taken here as 0.8 f cu and*ft t f ct

so fI=0.8
t
f W (6.24)
CU/

For simplicity, from Figures 6.13 and 6.14. .,

a 7f
xx
is taken as f where f is the
av av average normal
flexure stress in the compression by:
zone given
f = k, x 0.67 fcu (6.25)
av

where k, is the ratio between average. normal flexure

stress and the maximum normal flexure stress.

129
 fc
F- SC
'C2kud

\,. t
kilo- kud b

ue tO prestress
3tfa itur e
,-ý
av 10

ApSfpb

EC E
Ptp
('a) Forces and normal stress I
E-Pb
(bending), fxx,
diptribi#ion. -
(b) Concrete and tendons
strains at failure

Figure 6.14

I
vxymax
k2kUd

vu

d( I-k2 ku)
VXY

shear stress, v,
(a) distribution
qu
I
k2 kud
0 . 67 ki fcu kudb

d( i- k2 ku

-L

Figure 6.15

130
 I
k1 is a function of the value of Ec/8o at the extreme

compressed fibre (see Section 6.2.6. ).

By substituting equations 6.24 and 6.25 into

equation 6.22 vie will have:

1 (6 - 1) k 2
0.64 + 0.536 0,. 449 Iýk
VCU)

(6.26)

Equation 6.26 i's a failure criterion for an element

in the compression zone under normal and shear stresses.
6.2.4. Equilibrium condition:
Referring to Figure 6.15. a, ana neglecting the

dowel action of the longitudinal reinforcement and

the vertical component of the force due to aggregate

interlock on the inclined crack, one can write from
the equilibrium of external and internal moments that:
aV0.67 kfkdbd (l. - kk (6.27)
vuI cu U2u

and from equilibrium of vertical forces:

vu =v kUdb (6.28)
xy 1$
where kU is the ratio of the neutral axis depth to

effective depth and k2 is the ratio of the depth to

the line of ac#on, of the normal compressive force
to-the neutral axis depth.
,E%I-
equations 6*2? and 6.28 we get:
-From
v 0.6?. kl (1_ k2 ku)
XY (6.29)
f a
cu v/d
To determine the value of the average shear stress,
but not the position of the neutral axis itself, it

131
 v
(25y26)
is reasonable to adopt the simplified-assumption
that (1 -k2ku)=0.9 (6.30)

By combining equations 6.26,6.29 and 6.30 we get:

((,
k2 +. 6)2 1.2411)- 1.48 (16 - 1) k, 1.77
1 - =0
(a,, (6.31)
r/d)2+

v-
So u can be calculated using equations 6.28,6.29

and 6.31 if the neutral axis depth factor, kul is

known.
6.2.5. Neutral axis depth factor ku

The tendon-strain e at failure may-be consider-

pb
ed to bý made up of two parts: (a) the strain ýpe due

to f and (b) the additional strain e produced by

pe pa
the applied loading as shown in Figure 6.14. b . Thus

Epb = Epe + Epa (6.32)

Assuming an idealized case of perfect bonding between

steel-and concrete, E pa can be written as:

Epa 1- ku. ) fpt
Ec + (6.33)
= ku EC
Substituting equation 6.33 into equation 6.32 then
Epb Epe (1- ku fpt
= + q8o + (6.34)
ku Ec
Ec
whe're c<
E0

From equation 6.04

0C 0F,
k (6.35)
U Epb + Eo Epe L=
- -
Ec

Referring to Figure 6.14. a and aDplying the equilibrium

132
 0
con itions:

0.67 k1f k db Aps fpb (6.36)

cu U,
From equations 6. C5 and 6.36
klfcu bEd-
f 0.670(Eo (6.37)
pb fpt
Aps (Epb + O(Eo
pe Ec
ý< (= e- F-
The value of c/ o can be determined from

Section 6.2.6, and equation 6.87 can be-solved

graphically using the stress-strain curve of the

tendon on a strain compatibility basis. Having found

f Epb,
and hence the tendon strain at -Pailure
, then
pb
ku can be calculated from equation 6.35.

6.2.6. Evaluation of k and k in terms of

12
c/, Lr'o ý---Snd their limitations:

Referring to Figures 6.13 and 6.14 a, and by

definition:

fo Ec
fd.
k fo
ec

for'l. 0 0.0035
e_
-ö 7fcu
0.244 x 10 x
and k2 by definition is equal to:
k2 f
=1 k1f0 E2
c0
-1.2.
4 o( (3 cK (6 'e _4 lk + 1) (6139)
for 1.0 0.0035
Vf
10.-3K
0.244 x

133
where to ýc K- 0.0035'
and
to

6.2.7. Evaluation of Vu:

klj k2 and ku can be calculated as shown in the

previous Sections, and therefore (1 -k2kU) can be

determined. If the latter differs considerably from

0.9 (see equation 6.80), then the new value of

(1 - k2ku) should be used in equation. 6.29 to con-
0 6.26. This lead to
struct equation will a new value

of k1. This procedure should be repeated till there

is a reasonable agreement between the calculated and

(1 -k2ku Vu be
the assumed value of can then

calculated using the final values of kl, k2 and kU

in equations 6.29 and 6.28.

6.2.8. Prediction of shear-compression failure load
in a uniformly-loaded simDl)r suT)r)orted beam.

6.2.8.1. Failure section and equilibrium condition:

The position at which. failure will occur in the

shear-compression case in a simply. supported uniformly

loaded beam is not immediately apparent. The process

of flexure-shear cracking is progressive and does not

stop suddenly when a crack reaches a particular section,
but the crack tends to progress towards the centre of
the span,

Neglecting the effect of aggregate interlock the

shear crack should theoretically progress with increas-

ing load until the shear resistance of the concrete at
its head is equal to the external shear force. The
length of the crack should thus increase increasing
with

134
 0
10
load.

Assume that the failure plane in shear-compression

lies at a distance Nut from the nearest support. Then

by referring to Figure 6.15. b, and neglecting the
dowel f oi 7ce and aggregate interlock, one can write
from-the equilibrium or external and internal moments
that:

qu >ýu q22 Au L-0.67 k1 fCUku dbd (1-k2 ku)

u
2 2
(6,40)

and from equilibrium of vertical forces:

qu qu XuL v XY ku bd
2
Combining equation 6.40 and equation 6.41, gives: -

vxv 0.67 kl (1 - k2 ku) (1

= u). (6.42)
f L/d Xu _2X
cu (1 - Xu)

As mentioned before, experiments showed that uniformly

loaded simply supported beams whicý failed in shear-

compression, had critical sections at a distance 0.3,3L

from Hence putting )ýu in

a support. = 0.33 equation
6.42 gives: -
,

kl (1 - k2 ku)
(6.43)
f L/d
CU.
Assuming (1 - k2ku) = 0.9 and combining equations
6.26 and 6.43, gives: -

k2 )2 + 0.554 - 0.662 1) kl - 0.79 =0

1 (L/d)2 (6.44)

135'
The solution 6.44 yields values for k,
of equation
24 L/d>,, 7.01
for f >,,ZO. 0 N/mm and for all values of
cu
which are all greater than

0.33 x 0.0035
0.244%1703ýýCu

the for kI (see

which is maximum value equation
ýd >,,7.0
6.38). Hence for all values of and fc
ýU>z
30.0 N/mm the value of k'l should be'limited to

the maximum value of k-, given above. The value of

1ý from 6.35
.U can be caiculated equations and 6.37
i
and by substituting the values of kj, k2 and kU

into Vu,
equations 6.42 and 6.28, the shear force

at the critical section,. can be calculated.

To determine the value of quL, from Vu, it will
be assumed that this value of Vu is the shear force

at a distance 0.5 d from the critical section measur-

ed in the direction of decreasing bending moment.

Hence
qu L. d
vu. = 2 qu

Vu L/d
L= 6 (6.44)
or qu L/d +3

A more conservative assumption is to take Vu at d

from the critical section This gives
V, qu L
u= qu
2 --
6 Vu L/d
or qu (6.45)
L/d
+6

Equations 6,44 and 6.43 will be discussed later in

Section.?. 5.2 when they be compared
will with experi-

136
 0

mental results.
6.3. Comparison between Diagonal Tension Equation
and Shear-6ompression Ecuation.

One - or two-point loading case:

A typical specimen of the general dimensions of
the type C beams-used in the tests and stressed by

seven ? mm diameter wires was studied in order to

study the results predicted by equations 6.5 and

6.28 for various values of concrete strength, pre-

stress and shear span to effective depth ratio.

Pigures 6.1.7.1 and 6.17.2 show the value of

shear force at failure as predicted by the diagonal

tension equation 6.5, the shear-compression equation

6.28 and ultimate flexure(94) as a function a
, of v/d.
f 2 fcP
In Figure 6. l?. 1 cu, was taken as 50,0 N/mm and
2.
was chosen as 3,5 and 7 NIMM Figure 6.17.1

shows that, within these ranges of variables, beams

.
of type C will in shear-compression.
never fall
Figure 6.17.2 shows the effect 4f
of varying cu between
f
35.0 and 45.0 N'1=2 with cp constant at 5.0 N/=2
on the mode of failure of specimen C. This case, too,
shows that diagonal tension is the dominant mode of
failure of a type C spedimen. For a 5.0,
v/d> ultimate
flexural failure will be reached before diagonal tension
failure can occur.
6.3.2. Uniformly-distributed load case:
For this case two-typical specimens were studied,
one with the general dimensions of type D in the present

137
 Btam ty PtC undtr one-or two-point tooding
00 -
fC N/mt4 and t*p 3)S &7 N/ M4
U50
( 94)
)k - Ultsmatit 'Ftexure
hear co rnpressio nEqn. C6.2 8
120 ----S
Diagonat Tension Eqn-C 6.5)
VC
KN

1%.
so - -
ft-% _.
-

40 r
.0
0L
367
0 CtVjZd
f
Figure 6.17.1: Type C beam with varying values of cý
and a (Concrete strength constant
v/d

Beam type 6 C1 under one-ort'wo-point loading

i0o
45 fc u354/. I
5 NIMJ and !cp5M t4/m
4 (94)
35 - .. 'Flexure
Ultimate

80 4S Shear-compression Eqn. (6-26)

, -ft
% DICIgonal Tension Eqn. (6-5)
VC
35
%
KN 45
3s
60

40

20 r

011
-1-I--9 8-
23567
"V f
Figure 6.17.2: Type, C beam aiý/hdvarying values of cu and
a (Prestress constant)
v/d
f
Figure 6.17. Point-loaded beams. Effect of ep, cu and
sv/d on mode of failure and shear force at failure or type
C beam.

138
 I
series of tests stressed by two 12.5 = diameter

strands, and the other similar to type A-U-5

(43)
reported by Kar For the D type beam with
fcu 2 fcp
bw = 50 mm, was taken as 50.0 NIMM and

was varied'from. 2.0 to 5.0 NI=2. For the A-U-5

type b= 125 mm, feu was assumed to be 45,0

with
H/MM2. Figure 6.18.1, for the D type beam, shows

that diagonal tension is the dominant mode of fail-

fcp. '-4.0 N/mm2, but that flexure failure
ure for
fcp >z5. O N/mm2
will become the mode of failure at
G/d >,,12.0. the breadth of the web-increases
and -As
or in a rectangular case as shown in Figure 6.18.2

for the A-U-5 type beams, shear-compression is dom-

b/d 1-10 and fcP 2. fcp >
inant for = 2.0 NIM For

2.0 N/mm2 and L/d> 8.0, flexure failure dominate.

will
fcp ý/d: >10.0 the fail-
For 42,0 and shear-compression

ure and flexure failure overlap. For fcp =0 and for

all L/d4 20 the diagonal tension failure is

values of
dominant. This latter case can also be observed in

Leonhardt and Waltherýs photographs for rectangular

beams of b= 190 mm.

From the above it can be concluded that the web

bre'adth plays an important role in determining the

mode of failure in a uniformly loaded case. Prom the

case studied which is shown in Figure 6.18.1 and con-

firmed by experiments in this investigation in the

higher U/d on type D-beam, it

range of can be concluded
that the shear-compression type of failure observed in

139
 240

( 94)
6eam type D Ultimate Flexure
Locldirg U. D. L -Shear-Compression
200 fcu 50 Nlffvn2 Eq. (6 . 44 )
fcp 2014 L5 N/ n-.
-n2 ____O_ Diagona I Te ns ion
Eq. ( 6-12 )

%%
qc t 160 0% %:

KN

120
f 5 N/mm2
cp=
5.,
. 4

3'
80
I

40 f
6 10 12 14 16 18 20

Figure 6,18.1: Type D beam wi th varying fcp and 6/d

(Concrete strength constant)
200

CP Oy2 IL4 N MT12 ( 94)

Beam type A-U-543) X Flexure
-Ultimate
Loading U-D-L ------ Shear-Compression
160 fCU 45 N/MM2 Eq. (6- 44 )
0 Diagonal Tension
Eq. (6 12
%b

qC 1 120 0 fc ,nm2
11%
% p
% -'Z -
KN

so 2

0
40

0
68 10 U 14 16 It 20
I team L/d
Figure 6.18.2 Kar's type (&-U-Sj with varying fcP and
(concrete strena, th constani)
Figure 6.18 Uniformly-loaded beams. Effect of beam properties and
Yd
on modo of failure and shear force at failure,

140
Kar's beams will L/d
rectangular never occur at all
b Z 0.25.
ranges for sections with w/d

141
 CHAPTER 7,

COITARISON WITH OTHER RESULTS AIM DESIGN RULES

(30) (318
7.1. BSCP 110: Part 1: 1972 and ACI - 71)
BuildiTL7 Code(32) Design Equations.

Equations (45) and (46) of CP 110: Part 1: 1972

contain partial. safety factors of 1.50 relative to

concrete strength and 1.25 relative to the prestressing

force. For the purpose of this comparison, these safety

factors incorporated in the Code have been removed,

giving: -
f
VcO= 0.67 bw ýft 1+ CP Code equation (45)
ft

where ft0.361fcu
Le mo
and Vcr 0.55 vc bwd +a
fpý 0.5) d
v/d -
Code equation (46)
I
where Mo = fpt 5r

The shear force at diagonal cracking should be taken

as the lesser of the Code equations (45) and (46).

If the dead load is neglected ancl straight tendons

are assumed, the ACI (318-71). Building Code gives under

a web,-shear crack, the ultimate shear resisiance of a

section as Vcw where:
(0
VO.
Vcw 0 bwd 0 29 8 fcu + 0.3 fc P) ACI equations
(11-3 and 11-12)

and Vci as an upper limit to the ultimate shear strength

of a section und; r flexure-shear where:

[FO.6 fcU (0.5 V0.8 feu

Vei 0 bvid 0.05 + + fpe)
bw d 2(4)
av
ACI equations
(11-3 and 11-11)
142
To permit comparison, the capacity reduction factor

was removed and the lesser of the above two express-

ions was taken as the shear force at diagonal cracking.

7.2. Comnarison of Equations 6.5 and 6.7. for
one- or two-point Loading with Published
Equations and Code Rules.
To compare equations 6.5 and 6.7 with other pub-
liqhed design equations and code rules, two typical

specimens were studied, one with the general dimens-

ions of type B in the tests and the other with those

of type F, both being stressed by means of seven 7 mm

diameter indented wires. The concrete strength and

the prestressing force were chosen to be within the

practical range. The concrete cube strength was taken

as 50.0 NIMM2 in both cases.
For B type beam, fcp was taken. as 8.0 N/=2

and fpe/fpu as 0.55. For the F type, fcp was taken

as 6.0 11/mm2and fpe/fpu as 0.58. The value of the

cracking load*predicted by each of the published
a
av/d
expressions as a function of was evaluated and
the results were plotted as shown in Figures 7.1 and
7.2, together with the shear force corresponding to
(94)
ultimate flexural moment
For the B type beam (hf = 47.5. mm), Figure 7.1,
the expression .s (34)
of Sethunarayanan q Sozen et al(86)
(40)
and Evans and Schumacher overestimateVC, but
(36)
Sozen et. al come out witý close agreement at
av/d >5.0, (44)
and the expressions of Arthur and
Jena and Pannell (52)-
underestimate Vc over the whole

143
 fcu=so. o N1W StthuncrOyonan
200 2
fc 8-0 H/MM2 2-0-2 Sozen et at
p=
303 E vonsASchumccher
tpe
44 Arthur
1pu
ax-I 5-----5 Jena& Pannell
6------6 ACI( 316 -7 1

7-o ..ý- 7 BSCP 110% 1972

VC
KN u Ito rn ate. Ft ex ure
9-9 Author-Eqn-( 6-5
9-4 Author. Eq n-C 6-7
100
-
s S4
-

" - -01
3

0 23
/Vd
CL

Figure 7.1: Typical type B-beam. " Diagonal cracking load

as predicted by published expressions and
design rules in terms of a
v/d
fcu=50.0 Nlmrý I- Sethunurayanan
200 -o- -i
f= 6-0 N/md 202 Sozen et'al
9 CP
fpe 303 Evans A Sctý'umacher
=O. Sb
fPU 4 )k 4 Arthur
S-----S Jenci & Pa nnell
6 6 A CI (31g-71 )
------
VC I s, 7 BSCP 110; 1972
..
&---6 Ultimate Flexure
9-9 Author. Eqn. (6-5)

ý 9-9 Author - Eqn, t6.7)

100
1

7' Z0

9'

0123.6

a V/d
Figure ?. 2: Typical type F-beam. Diagonal cracking load
as predicted by published expressions and design rules
in terms of av/d

144
 0
av/d. Both BSCP 110: 19? 2 and ACI (318-? l)
r6nge of

for a /-2.0, but results

are conservative v/d give

close to the modified curve (equation 6.7) at higher

a
values of v/d.

Figure 7.2 shows the comparison based on the

F type beam with a thicker flange (hf = 87.5 mm).

Here 8ethunarayanan and Jena and Pannell give

conservative predictions whilst Evanb and Schumacher

and Arthur overestimate the value of Vc. Sozen et al

for av/d> 4.0. BSCP 110: 1972

show some agreement

and ACI (318-71) Building Code unaerestimate the

av/de-2.0, but they to the

value of Vc at are close

(equation 6.7) for av/a >3.0. Both

modified curve

the British and the American Codes of necessity, being

design rules, stipulate lower, bounds to the expected

range of test results, so they would be expected to

appear conservative.
7.3. Comoarison of Equation 6.5 for one- or two-point
Loading with Published Test Re4ults:

Figures 7.3.1 and 7.3.2 and Table 7.1.1 show ,

a comparison between equation 6.5 and test results

(34-362 41-45,47,49)
reported by different authors for

various rectangular, I and hollow sections with differ-

levels av/d
ent concrete strength, prestressing and

ratios. Some of these beams were of zero* prestressing

(47l49).
force(36149)and I others contained web reinforcement

From Figure 7.3.1 and Table 7.1.1 it can be seen

A (47)
that, of the beams reported by Hanson and Hulsbos
(49)
and Bennett et al with web reinforcement, the test

145
 0

Ei
00A
40
+ V -P

1
c c cu C)
40 40 0 iLf
cii 0

x0
9
ý
006 .H

1
(4)
Co ;-,

0 -p
oft M to.,. f
.
11
-

-p
119 r-1 %-.,
x x -M :i
0ý
OCDS 0)
4j Cl
XX. x
4(X In ;4
V
NI
>
y16 ci
xx
m ; -4
.000
*o0 -ri 0
000 H
(0 r2

cli

no
m
p mg Hp

ooot

146
 C-4
t

2aI1 4-ý
Al .0 0
ro (1)
1V t-
.C. : 1%
Ix

ix tic)
os 0 Ix 0
C3
0
.cNI E H

40 0

04x E) co
cc
C) 0)
144 - to
0% V
U? r-4
C3 :J
rl 0

tio 4j,

00
x
+ý co
0)
x r

(1) 0
IV
q-

14 Ul
t-
-M P0
04 *rq
*90 a 0 -P ýý
0) cu
/
**1 1
1;
141
a) P
w '.*- & +3
'ca .. / LA
V
C4
ro
-ý

4 co
0)
ý01
#0
00
->- or4 4-4
ol
rq 0
.0
00

4J

00 CO

p :1) ýhq 14
74 + S. 1) p C+1
Ell Q.
q 1.
Ao C)oL

147
(j ý-

TABLE 7.1.1.

Comparison of test
results for beams
under one-or two-point loads by other
investigators(34,36,41,43 - 45,47949)

with values calculated with the diagonal

tension cracking equation(6.5) -

Author Beam av vc Author Beam avI Vc

Extit. Exot.
mark -d Vc Calc. mark -d -Vc
Calc.
(36)
Sethu-na 1 2.84 0.94 Sozen ýC1233 3.57 1.17
rayanan 2 4.26 0.90 et al 1257 3.63 1.14
9 1.20 fc
p Fý 0 2231 3.31 1.25
10 2.12 0.89 i 2240 3.65 1.11
(36
11 2.13 1.25 Sozen B3211 3.46 0.64
13A 2.12 0.92 et al
19 3.53 0.71
13B 4.24 1.07 fcp 0 31 3.53 0.99
14 2.83 0.92 34 3.56 0.86
15A 2.13 0.86 41 3.40 0.92
15B 4.26 0.82 3.47 0.85
ý54
17A 2.13 1.07 C3211 3.25 89
.
17B 4.26 1.01 22 3.60 1.05
21A 2.13 0.95 37 3.60 1.07
21B 4.26 1.00 42 3.56 0.77
(36) 50 3.37 0.87
Sozen M43 6.55 0.73 80 3.60 1.06
et al 1151 6.40 0.86 1
(41
/0 1153 6.73 0.77 Swamy Sj 3.75 1.04
cp
1196 6.42 0.73 sii to 1.19
31120 5.29 0.79 siii it 1.08
1129 5.40 0.78 siv it 0.94
1140 0.91 v to 0.86
1210 3.24 0.81 svi 1.02
1261 3.64 1.10
2126 5.29 0.73 (43)
- A-1 5.00 0.74
-Kar
2265 3.62 0.96,
-2 it 0.73
3115 5.29 1.04
-4 3.50 1.20
3231 3.53 1.01
-5 3.00 1.22
3254 3.47 0.85
-6 4.00 0.98

148
 .11
TABLE 7 1.1. (Cont'd)
.

Author Beam av Vc Author Beam av vC EýC!

ExT)t. Dtp
mark -d -
Vc mark d Vc Calc.
Calc.

(43) (44)
Kar A-7 3.86 1.07 Arthur A24 3.43 1.19
-8 4.14 0.98 25 111 1.25
-9 3.86 -1.03 26 is 1.21
-10 5.00 1.07 Bl 3.36 0.82
-12 4.00 1.00 2 of 0.88
B-3 3.50 1.11 3 2.52 1.02
-4 4.00 0.92 of
-5 4.50 0.97 4 1.68 1.12
-6 4.67 1.00 5 2.52 1.18
-7 3.50 1.00 6 1.68 0.86
-9 5.00 0.84 7.. 11 0.87
-10 5.00 1.09 It it 1.22
(44) 8 2.52 1.02
Arthur Al 2.28 1.09 of if 1.07
2 4.57 1.06 9 3.36 0.75
4 it 0.97 of it 0.77
5 2.28 1.26 10 1.68 1.03
6 it 1.27
8 it 11009 1.12 0.98
9 11. 1.13 ll 3.36 1.16
10 it 1.19 1.20
of 'It 1.12 12 2.24 0.92
11 It 1.08 of If 0.96
12 If 1.06 Cl 1.68 1.00
13 it 1.19 to it 1.05
17 4.57 1.12 2 1.12 0.94
18 it 1.06 3 2.24 1.09
19 It 0.87 1.12
IT it 1.12 6 1.12 1.18
20 q5 7 2.24 0.89
ý1. 03
T It 11
41 It

0.93
21 1.03 0.90
22 It 0.98 It It 1.02
23 3.43* 1*23 8 1.68 0.86

149
 TABLE 7.1.1 (Cont'd)

Author Beam av VC Expt. Author Beam av v

mark c Expt.
-d Vc mark - d Vc Cale.
Cale.
(44)
Arthur Ce 1.68 0.95 Hanson & 7.05 0.99
-F6
to ft 0.14 (47)
Hulsbos ff
Dl 3. S6 1.21 (with 7 4.24 1.33
2 3.92 0.82 web to ff 1.28
3 2.52 0.78 reinfor- 8 it 1.24
cement)
0.91 9 5.64 1.15
4 1.68 0.69 it 6.34 1.04
0.70 10 4.94 1.31
0.82 ff of 1.20
0.91 13 5.64 1.83
0.70 1.21
0.84 14 6.35 1.12
ti it 0.90 ff It 1.07
ft ff 0.88 15 7.05 1.02
El 2.80 '0.79 16 7.76 1.02
2 It 0.93 It it 0.99
3 3.92 0.93 (49) A03 0.75 1.00
4 1.68 0.66 Bennett
13 1.50 1 12
0.80 et al .
(with 23 2.25 1.41
0.81
web 33 3.00 1.25
0.82 reinfor- B02 0 . 75 1 23
cement) .
12 1.50 1.07
Arthur & 1 1.68 0.70 f
(45) cp= 0 22 2.25 1.17
Mahgoub 2 1.76 0.80 '
C03 0.75 1.04
3 0.91 0.73
13 1.50 1.31
23 2.25 1.20
Hanson & FXI 3.39 1.26
(47) 33 3.00 1.12
Hulsbos it it 1.19
D13 1.50 1.21
F1 2.12 1.08
(with 23 2.25 1.37
web VI 11 1.05
33 3.00 1.03
reinfor- 2 2.82 1.35
cement) E13 1.50 1.21
of 1.19'
33 3.00 1.17
3 1.20
F13 1.50 1.36
1.08
23 2.25 1.30
5 3.52 1.24
33 3.00 1.12
iso
 F TABLE 7.1.1. (Cont'd).
V

Author Beam av vc
Expt.
0 mark d Vc Calc.

(49)
Bennett G23 2.00 1.10
et al 33 3.00. 1.14
(with B13 1.40 1.06
web 23 2.10 1.11
0 reinfor-
cement) 23 2.80 1.30
33 1.00 1.07
Cc
.,: J12 2.00 1.14
22 2.10 1.14
E13 1.40 1.14
23 2.10 1.21
33 2.80 1.14
L13 1.40 J. lg
.,
2.10 1.23
ý23
33 2.80 1.26
Ta3 1.40 0.99
13 is 1.04
23 2.10 1.15
.
33 2.80 1.11
IT13 1.40 0.99
23 2.10 1.07

09

151
 TABLE 7.1.1. (Con Vd)

Mean value of ratio and standard deviation.

Author No-0 of Mean Value Standard

results. of ratio deviation

Sethunarayanan, 14 0.99 0.125

Sozen et aLl 18 0.93 0.171
fcp X0
I
Sozen et al 12 0.89 0.138
f
cp 0
Swamy 1.02 0.114
-6
Kar 17 1.00 0.136
Arthur 76 1.00 0.155
Hanson and 25 1.16 0.114
Hulsbos
(with web >
reinforce-
ment)
Bennett et al 19 1.19 0.119
(with vieb
reinforce-
ment)
fcp
=0
Bennett et al 20 1.13 0.082
(with web ý10
reinforce-
ýnent)

cp X0

il I

152
results were more or less within or slightly above
the + 20% bound. This may mean that the presence

of web reinforcement has a negligible effect upon the

value of cracking load calculated according to equation

6.5. Figure 7.3.2, for beams without web reinforce-

ment, shows that at higher values of*Pv/d, some tests

(36) *
reported by Sozqn et al seem to lie below the
I
Ilower bound (-r20%) of equation 6.5, but not outside
,,,
the modified curve given by equation 6.7. On the other
I (42,43)
hand about 90% of the tests-'repo rted by Kar on
i,,
rectangular beams failing in shear-ccmpression at
3.0, /,, av/d, c'-ýz5.0, gave cracking load values above the
lower bound (-20%). of equation 6.5 as shown on Table
7.2 column 7. So it can be seen from Figure 7.3.2

thal 6.7, ýfter being multiplied by a reduct-

equation
ion factor of 0.8, can be taken as a lower bound for

predicting tne cracking load value for design purposes

for-av/d 45.0 av/d>
and that for 5.0 the effect of
a av/dý,,', 5.0
v/d can be ignored. Thus, for

1000 vc av
0.25 - . 027
[1.5 hil d
1f +f CD) bw d +k
ct J# i1 3j'
ct (7.1)

whilst for av/d> 5.0

1000 vc
* 0.1 (7.2)
,
ff + bw d
et .5+

7.4. Comnarison of the Shear-c'wul

o-nression--Equation 6.28
with Experimental Results and other Published
Shear-compression Eauations.
133
7.4.1. Some of test results in this investigation and
(43)
some others reported by Kar and Arthur(44) on rec-

tangular and I- section beams show that the final

mode of shear failure for those specimens was shear-

compression failure init. iated by a flexure-shear crack.

The ratio of the experimental to the predicted shear

force at failure is shown in Table 7.2 column 8. The

experimental values of 81% of test results Were within

±20% of the calculated values given by equation 6.28.

7.4.2. Comoarison of equation-6.28 with other

published shear-compression eguations:

To compare equation 6.28 with other published

(40143) (43)
expressions ,
the test results of Kar's

prestressed concrete beams were studied. The compa3ý-

ison was made between the experimental and the calcul-

ated value predicted by each expression as shown in

Table 7. a which also shows the mean value of that

ratio and the. standard deviation in each case. The

scatter of the experimental to calculated values in

termý of av/d is shown in Figure 7.4. From Table 7.3

and Figure 7.4, equation 6.28'seems to predict the

shear force at shear-compression failure reasonably

well when compared with Kar's., expression which was

derived from the same experimental data from which this

comparison was made. Kar's, expression is tedious, and

(40)
Evans and Schumacher,'s i. e. equations
., -expressions,
2.25 ., 2.28 have their limitations as they are governed
by the percentage of the,, main steel, e as discussed

154
 TABLE 7.2.
0

Comparison of results of tests on beams

failing in
shear-compression under one-
(44)
or two-point loads by Kar(43), Arthur
and the present investigator, with equation
(6.5) and shear-compression equation (6.28).

(3) (4) (5) (7) (8)

-(2)
Beam f f av vc Vu Vý Vu
cu 2 cp 2 Expt. Expt. Expto Expt.
mark N/mm NXm d KN KN v v
c Cale. u Cale.
(43)
(Kar)
-A-1 44.9 2.57 5100 22.0 26.4, 0.74 0.95
--
-2 43.5 2.07 5.00 19.6 24.0 0.73 0.96
-4 36.0 2.63 3.50 40.1 55.0 1.20 1.31
-,g 43.0 3.35 3.00 51.4 68.5 1.22 1.27
-6 35.0 1.57 4.00 24.5 38.2 0.98 1.12
-7 37.8 2.64 3.86 34.3 45.0 1.07 1.02
-8 42.7 4.03 4.14 39.2 43.1 0.98 0.94
-9 42.3 4.66 3.86 46.5 53.9 1.03 1.04
-10 39.8 4.66* 5.00 34.3 40.1 1.07 0.98
-12 '43.6 3.29 4.00 36.7 43.6 1.00 1.02
B-3 36.4 2.45 3.50 24.5 28.4 1.11 0.93
-4 40.0 2.42 4.00 19.6 28.9 0.92 1.01
-5 35.0 2.06 4.50 17.6 25.5 0997 0.98
-6 37.7 2.98 4.67 23.0 26.9 1.06 0.93
-7 41.4 3.28 3.50 26.4 39.7 1.00 1.17
-9 41.6 3.28 5.00 19.1 26.0 0.84 0.94
-10 44.3 4.14 5.00 28.4 33.3 1.09 1.12

(44)
Arthux
D2 56.7 5.09 3.92 37.8 50.3 0.82 0.83
E3 53.3 5.05 3.92 46.5 52.7 0.93 0.85
G6'* 49.7 3.88 3.00 94.5 98.0 0.84 0.97
C19 51.0 2.45 5.25 21.1 54.5 1.03 0.91
1 1
* Author's beams.
155
 TABLE 7.3
.

Comparison of the shear-compression

equation(6.28)and other published
expressions(40,43) ! dthexperimental
results obtained by Kar(43) on rectangular
prestressed concrete beams.

Beam a Experimental shear-compression failure load

v
mark d Calculated shear-compression failure load
Equation 2. 25(40) (43)
Equation 6.28 Kar

A-1 5.00 1.25 0.95 0.96

-2 5.00 1.24 0.96 1.03
-4 3.50 1.11 1.31 1.19
-5 3.00 1.05 1.27 1.16
-6 4.00 1.09 1.12 1.18
-7 3.86 1*00 1.02 1.07
-8 4.14 O. M. '0 . 94 0.92
-9 3.86 0.94 1.04 1.01
-10 5.00 0.92 0.98 0.98
-12 4.00 0.84 1.02 1.05
B-3 3.50 0.85 0.93 0.88
-4 4.00 0.93 1.01 0.99
-5 4.50 0.95 0.98 1.08
-6 4.67 1.05 0.93 0.96
_? 3.50 1.03 1.17 1.10
-9 5.00 0.96 0.94 -1.01
-10 5.00 1.12 1.12 1.19

mean 1.011 1.041 1.045

standard 0.127 0.118 0.095

deviation

156
 I.

0
40
40
00 9 &0 0.

+i

-0 xo
rf
x 00

0
-P
Co
mi
0 Co
+i
0
60
*K 0 FA
x 40
0 0 cu 0
ex c4 c) r4
14 ri -p
(D 0) G3
zt

00 -X0
0 . *x
0x0 0 ei >

0
ý 0
-4

es
0

10 40

157
in Sý-ction 2.5.2.5.

As mentioned in Section 6.1, because of the many

variables involved in determining the final mode of

failure in shear for beams without shear reinforcement,

and because of indeterminate nature of some of these

variables, for design purposes the useful capacity of

such beams- in shear under one - or Awo-point loads

should be limited to the values given by equations

7.1 and 7.2.

7.5. ComT)arison of the-Expressions Developed for
Uniformly Loaded Beams with Test Results and
Published Expressions.
,
Comnarison of -the diagonal' tension equation
6.12 and test results..
Figure 6.10 shows the test results in the present
investigation (43j47,48)
and those reported by others on

rectangular and I- section beams plotted with equation

6.12. Table 6.2 shows this scatter in the form of a

comparison*between the experimental and calculated

diagonal tension cracking load predi; ted by equation

6.12. From Table 6.2 and Figure 6.10, it can be seen

that all the
experimental results obtained by Hanson
(4? ) (48)
and Hulsbos ,
Arthur et al and this investigation
for the diagonal tension cracking-l oad were within 4-

20% of the value given by-equation 6.12, even for those

specimens tested L/d as high (D-type

atl, as 17.78 beam'

stressed with two 12.5 mm strands). In this investigat-

ion a common feature in all those beams was that the

iss

I... I
diagonal crack opened on a line either through or in the

vicinity of a reaction, as mentioned in Section 6.1.3.1,

On the hand, Kar '9ý'43)specimens

other which were of rect-

angular cross-section, all failed in shear-compression

initiated by a flexure-shear crack. His experimental

results are also shown in Figure 6.10 for the sake of

comparison with equation-6.12. The overestimation of

cracking load given by equation 6.12 in such instances

is shown in Table 7.4 column 7. Equation 6.12 even

predicts values greater than the shear-compression

failure load as can be seen by comparing columns 6

and ? of Table ?. 4, but this is to; be expected since

the final mode of failure. was not diagonal tension

cracking. * The only exceptional case, was (A-U-9),

with zero prestressing forcel'in which case the

diagonal tension cracking load predicted by' equation

6.12 was 0.78 of the experimental ultimate shear-
compression ýailure load. This case was discussed
in section 6.3.21 and was illustratqd by the example .
given in Pigure 6.18.2,

7.5.2. Comparison of the shear-compr ssion equati,

(43)
6.44 ana 6.45 with Kar's test results.

Kar's rectangular beams were found to fail in

shear-compression before the formation of the diagonal

tension cracks. In the previous section it was pointed
out that values predicted by equation 6.12 were greater
than the experimental shear-compression failure loads.
The expressions developed for the shear-compression

159
.
 TABLE 7.4

Comparison of results of-tests on uniformly

loaded rectangular beamsby Kar(43) failing
in shear-compression with values calculated
by shear-compression equations(6.44)andý6.45)
1
(1) (2) (3) (4) (5) l (6) (8) (9)
Beam f- f L q qLu qL
cu
ý%. 2 cp d c c qL Exot.
mark NýL2 Expt Expt. Calc, U.
KN I KN IKN qu. L calc.
Eqn. Eqn. 1 Eqn.
6-12 6.45 6.44

A-U-2 ] 44.3 11.96 '112.871 69.9 1'91.0 1102.8 1 0.98 10.82

-3 48.3 3.35 11 103.2 128.8 1`34.0

%.0
1.23 1.03
-4 45.7 4.02 103.2 119.9 145.4. 1.15 0.96
-5 35.9 3.23 81.0 97.7 122.7 1.17 0.99
-6 35.9 3.34 92.1 li6.6 124.5 1.34 1.13
-8 34.8 2.64 81.0 107.8 109.5. 1.30 1.10
-9 38.0 0.00 42.1 74.3 57.7 1.03 0.86
-11 42.3 4.02 108.8 125.5 170.0 1.19 1.06
.
B U-1 48.3 3.28 15.00 63.9 80.6 90.0 1.21 1.07
-2 42.3 4.14. 57.9 65.0 98.1 1.05 0.90
-4 45.7 4.15 75.0 79.4 101.4 1.18 1.01
-5 39.6 3.33 47.2 62.8 81.4 1.09 0.94.
-

Mean value of ratio''' 1.16 0.99

Standard deviation 0.107 0.096

,L

160
failure load in Section 6.2.81 equations 6.44 and 6.45

were used to predict the shear-compression failure load

for Karls'beams, and the results of comparison with the

experimental values are shown in Table 7.4 columns

and 9. Column 9 shows that the values predicted by

equation 6.44 are in reasonable agreement with the

experimental values, while equation 6.45 gave conserv-

ative predictions, as shown in column 8. Thus it can

be concluded that the value of the total uniformly
distributed load at shear-compression failure can

well be predicted by equation 6.44.

Hence, for design purpqses, the total uniformly

distributed load at failure should be limited to 0.8

times the lesser of that given by equations 6.12 and

6.44.

7.5.3. ComiDarison of the diagonal tension eauntions

with other published exDressions.

A comparison between equations 6.12 and 6.17

developed for predicting the total uniformly distribut-

ed load at diagonal tension, cracking and that. given

(48)
by Arthur et al. equation 2.37,, is shown in
,
Figures 7.5 and 7.6, together with the experimental

results obtained. from type B and F beams in this

investigatio n. Arthur et al's equation 2.37 tends
to underestimate qc W-ý9.0
the value of for as it
ý/d'=
predicts a zero value'at 12-47 (see Section
2-5.3.3. ). For the type. beams, which were of small
_B
flange depth compared-with",. those reported by Arthur

161
p
 I

2 Eq u cit on C 2.3 7

-OF C 6-12

2-0 - Experiment aIresuIts

-0
u

s6789 10 12
I/d

Figure 7.5. a. Type B beam.

41 -

3-0 48)
---Equation 2.37

6-17
"a
2. S Experimental r esults

U
a. '11ý
0
0
0
.,
0.2-0 - -

1.5 -

1-0
567 .9 10 1 12
0 t/d
Figure 7.5. b. Type F beam.

Figure 7.5. Comparison of results of tests on type B

and type F beams(ýgjer uniform load with the proposed
and. Arthur et al expressions for diagonal iension
cracking load, q
c

162
 L/d
et al, eqfiation 2.37 underestimates at all values.
This is because, as. mentioned in Section 2.5.3.3p of
the arbitrary choice of the'coefficient and the constant

which describes the geometry of the flange.

163
 CHAPTER 8
0

CONCLUSICNS AND RECOISU21DATIONS FOR

FURTHER RESEARCH

8.1. Conclusions:

For the number of specimens tested and the

range of variables tackled in this investigation, the

I
following conclusions may be drawn: -

8.1.1. Distinctly different characteristics of behav-

iour were shown by the specimens after ýhe formation

of the diagonal tension crack. This behaviour could

from a L/d
not be predicted a knowledge of v/d or alone.
8.1.2. Because there arel, at,, presentlindeterminate
variables controlling any margin of shear strength
beyond inclined cracking, the useful ultimate shear
strength of prestressed concretebeams-wi .thout shear
reinforcement should be limited to the inclined crack-
ing shear force.

The flange projectionsof jan I- section can

contribute up to 40% of the., value of the inclined
cracking shear force depending on their configuration.
8.1.4. The diagonal tension cracking load can be
predicted to an accuracy of 20% by the following

expressions.
8.1.4.1. Under one - or two-point loading.
1000 VC
f (0.10 0.31 )
+bh ,115 ý+ 7 . _. 1) +
f# ft w
ctbwdctda v/d
h Equation (6,5)
b "f ý 1.0
where I-1d
Tw

164'
8.1.4.2. Under uniformly distributed load:

IOCO qj fcP- )h f
--., 5 4b-1 -x
fýt bw d -21+ fo b,,; d
ctt -
0.10 + 0.31
1-2AC -7d Nc (1.

Equation (6.12)

where ýC: is given by: -

2)A3 + 18.6ý\2 9.30ýc 1.55

+- + T"d _ =0
c Urd-j/lc T-7d-
Equation (6.13)
and 10
d

L/d is in
The values ofýc as a function of shown
Ath L/d (equation
Table 6.1 and the relation of qCL
6.12) is shown graphically in Figure 6.10

The shear-comp: pession, failure -load, based on

ohr's criterion of failure using, a straight line

-11-!
envelope, can be calculated to within :L20% using
the following equations: -
8.1.5.1. Under one or two-point loading:

VU =V kUbd Equation (6.28)

XY
where
2 10
64 + 0.536 V- 1) k 0.449 9 2
kl]
ý ýc +-6)4 .
c
and k is given by: Equation (6.26)

2(l'(1 +j)2
+ b2
+ 1.24ý)- 1.48 1) k, - 1.77 =0
/d)
"Equation (6.31)

165
 F-0 Equation (6.. 35)
Epe - OF
-Lpb o-

Under uniformly distributed load:

Vu L/d Equation (6.44)

q6u L/d +3

.e ku bd Equation (6.28)
wher Vu =V XY
as

and v is given by-equation 6.26- shown in

XY
(8.1.5.11 where

XO. 0035
k=10.33
1 0.244 x 163xrf-cu
2
for f >_.,30.0 N/mm and '1/,d >/7.0
cu
k is given by equatio_n_6.3ý shown in (8.1.5.1).
U
8.1.6. The'Code rules of BSCP 110: Part 1: 1972

ACI (318-71) lower a v/d

and are ratherconservative at

ratios. The other published, bxpressions tend to over-

load lower av/d

estimate the value. of the, cracking at

ratios.

8.1.7. When published test resulýs are compared with

equations 6.51 6.71 6.122- 6.28 and 6.44, close agree-

ment is obtained over a wide range of variables.

Compared with beams having shear reinforcement, the

effect of the latter-upon the'value of-the cracking

shear force predicted by, e"quation 6.5 is negligible.

8.1.8. For design purposes,. the, following express-

ions may be used: -

8.1.8.1. Under one or two-point loading

3.66
0
-Its 0 1000 vc av
0.25 - 0.027
(1 h
fl +bwd 5', +d
ct

Equation (7.1)
k

for av 4 5.0
d
0.10 for av > 5.0 Equation(?. 2)
d
(bw hf
wheýre -d l'O

8-1.8.2. Under uniformly distributed load:

The lesser of the values given by equations 6.12

and 6.44 multiplied by a reduction factor of 0.80.

8.1.9. From the experimental obserVations, it seems

necessary to provide-a thin-webbed section with '

bw/d 140.33 av/d
zz: 44.0,
at 2.0 -. with a minimum amount
of web reinforcement which will"increase ductility

and reduce considerably the likelihood of a sudden

and catastrophic shear failure.

8.1.10. The experiments conducted on type D- beam

under uniformly distributed load and the case studied

in Figure 6.18.1 showed, that sheO-compression failure
in loaded beams with b C
will never occur, unif ormly w/d
0.25 for all L/d
values.
8.2. Recommendations for Further Research:
8.2.1. Equation 6.5 was.ýdeveloped from tests on
hf
beams with 1.0 w-hich are the types of
wd
beams most commonly used in general building construct-
ion, but it is questionable,, the present stage, if
_at
we could allow the, flange-projections, to contribute

167
more than 40% to the value of V. for higher values
R;b_ hfd (97)
of 1) -Tang quoted some long span box
b >12.0.
girder bridges of Hence further invest-

igation is needed on box beams with b /bw >4,0.

8.2.2. Some shear tests on full-scale prestressed

concrete box beams are. necessary to study the scale
effect on equation 6.5. A-few of these tests have
(45)
been reported

8.2.3. The shear failure behaviour of box beams

and T-beams under uniformly distributed'load should

be investigated.
8.2.4. A slower rate of loading might be expqcted

to lead to a lower, shear cracking load. Beam E4

in Figure 5.1. g is". an'example"of'-this possibili ty.

Hence the effects of different rates of loading

should be studied*
8.2.5. Equation 6.5 could be modified to predict
the cracking load. for the type of loading shown in
Figure 8.1. The end reaction q, nL,,,', for such
.
loading will be. assumed to be given by the follow-
ing equation:

q. (I
0.10 g. 62
Ob 11.5 n + __.(ýn-Ac)Vd
fc$t 1+ b
(ý; hf] (n-; kc) Ac
+ -uj
.wd
(8.1)
Lis the distance
where>ý. of tlie critical section from
a reaction and is given by the following equation:

168:
 ýC (4n2 ),
4n)ý3 + 3?. 2 ý2
+ 7dTc + 3?. 2n2Ac
c
3
12.4n 0
Wd '(8.2)

Further research is needed covering this type of

loading to verify equation 8.1,

qc qc

hd

Figure 8.1
 0

0 Lts to
9ý gn -
13cX In Z; Z; z
0
o 13 x

x cl

90
X Oe
0

0X0

013 X W
ro
ýf Grq

0
4->
ci

0> 0 A
tE
r xIa
CL
*13X 0

?A
orl a
13OX0 40

.0
4J
ho
0
CEI X 0)
H

.0
0
Sri
Osk*
E3
Sri
fl 40
00 x

E-A
40
OX 013
"H I LO
0

X 13

-0

1ý 1? 4? "
00 ýv I 0
'. 0

170
 0
N

Im 0
X 13 C4

0
0
so
0

0
0 to
ro

+A

0
0)
0
IA
x E
CP E

to a
x Xe ul
sr-I 0
ro
0 LO
XIC

Cý
H
0

C,
I-
4-3
frl
U
C,

co t vi
c
U

0
0
Ef)
co

vi to
si
03
$4
E-4

00

0
(C
49
C*4

14
:3
txo

W0 40 le e4
-Ii-

.0.
30C

171
 APPMTDIX B

B. 1 Estigration of -1)ýestress Losses in a Typical Beam

B. 1.2. Geometric andImaterial nroperties of the snecizen

eam Age Cement Cement "later Humidi- ff ýrestress-ý

type ty ci cu
ark at content cernent NýL2 ing
test kg/m3 ratio at
test 2 steel
CN/C
N1,rpm
days

Al 26 R. H. P. 500 0.40 Normal 43.4 51.5 7=

C. air indented
wires

_%3 -f
Aps Aver- Es Aver- Acx 10' 1x 10 Peri- Eccen-
age 2 age L2 meter tricify
6. pix1O 5 KNIMM mM2
=2 f exposed (e)
pi
RýMM21
Mm/Mm K mm

69.5 1 546 1200 1109.2 133.44 '134.97* 11205 46.4 *

- ,

R el. 3. Estimation of the losses in accordance w

BSCP 110'. Part-l*. 1972(30)

B. 1.3.1. Elastic losses:

AfAf
ps ptr. ps Di
1+ Es Aps + e2
Eci Ac -I

Aps fpi '294.3 KN

Eci(f = 43.4) =-ý32, Mj/=2' (Table-1, BSCP 110)
cu
E -7 2 'x -7
s 6.25,299
A x 10, .'e 61'. 6 10
E c
ci
2 -7 -2
and 1+e 360.6 x 10 mm
Ac I

172
 294.3 -7 294.3
Af
Ps ptr 6.25 x 269.5 x 360.6 x 10 1.061

277.5 KN i. e. Prestressing force in the_

wires after elastic shortening.
15.1.3.2. Relaxation, shrinikage and creeD losses:

Shrinkage strain (Table 41) 300 x 10

-6 -6 2
Specific creep (Clause ý4.8.2.5. ) x 48 x 10 = 24 x 10 mm/N

Prestressing force at transfer 277.5 KN

Relaxation loss (from BS 2691: 1969) 5%

13.9 KN

Average force during loss period ý77.5 13.9 = 270.6 KN

2
B. 1.3.3. 'First approximation- 111* -

Average stress in concrete at centroid of steel

32
270.6 x 10 x(yl- e )N/mM2
c
9,76 N/m ý2
-6
creep strain = 9.76 x 24 x 10 mm/=
-5
= 23.4 x 10 mm/mm.

ciýeep + shrinkage strain = (30.0 + 23.4) x 10

= 53.4 x 10 mm/mm.
2
Steel stress loss = 107 NIMM

Steel force loss = 107 x 269.5 x 10 KN

='28ý8 KN

Af 277.5 -, (13.9 + 28.8) KN

Ps pe
(approximately)
234,8 KN

B. 1.3.4. Estimation of the effective prestressing force:

Average force during loss period' 277.5 (13.9 + 28 . 8)

2

Average stress in concrete at

-3 22
centroid of steel 256.2 +e
x 10 X(WC -I )N/mm
9.24 N/Mm2
173
 -6
Creep stýain 9.24 x 24 x 10 mm/mm
-5
22.2 x 10 mm/mm
-5
Creep + shrinkage = (80.0 + 22.2) x 10 mm/mm
strains -5
= 52.2 x 10 mm/=

Steel stress loss = 104 N/=2

Steel force loss 28.1 KN

,=
*. Af 277.5 - (13.9 + 28.1) = 235.5 KN
. ps pe
Total loss at test = 100 - 235.5, x 100
294.5

= 20.0%

F. 1.4. EstimatiOn of losses-using CEB Recommend-

ations(27).
F. 1.4.1. Elostic, losses:

Aps f 294.3
ptr Es Aps( 1+e) 2
1+
Eci Ac
2 (Clause
Eci 6.6 ýfcu
FO--8
2
KN1m
.R
12 - 22)

38.9
Es
= 5.14
Eci
Aps f 294.3
ptr -7
5.14 x 269.5 x 360.6 x 10

294.3 280,3 IKN

1+ 05
.
i. e. Prestressing force in the wires after
elastic shortening, "

B. 1.4.2. Creep strain: f (Clause R 12.31)

f- Stress in concrete at centroid of prestressiLig

ýt /Ec
eelýJ,
steel X-
X. t
.) 28
Ot is the creep coefficient
where
Ec 28=secant
modulus of concrete at
28 days

174
 is the product of five partial coefficients, each
given in a form of a chart (see Figure B-1)
Ot
= ke kdkb ke kt

where
k the conditions and it
c depends on environmental
the effect of drying under load, which is
gives
large if the relative humidity is low.
very
kd represents the combined effects of ageing and
hydration as a function of, the type of cementýused.

kb depends on the mix. It includes the effects of

water cement ratio and cement content.

k covers the influence ofýthe size and shape of

e
the member in delaying drying.;

kt shows the development of creep, with time, inciuding

the delaying effect of the-_larger,
-sections.
In this case:

ke = 2.30 (Wormal air)

kd = 0.70 (R. H. P. Cement and 21 day loading,

detensioned after 5 days).
kb=1.00 (Water/cement. ratio =. O,.40. and cement
content = 500, kg/m,
k is a function of the theoretical thickness e
em
where

e Area of section
M
perimeter'in contact, ivith atmosphere
3
33.44 x 10 cm = 5.6 cm
60.25 x 10

k 1.20
e
kt 0.40 (em = 5.6 and t,ime under load 21 days)
-0.7 x 1.0 x 1.2'-x 0.4,,
t=2.3 x

= 0.? 73

175-
 ( 27)
Coefficientke (environmentalconditions). E iclu re B-1 *- CEB- FI P Recommendatio ns
Charts for the -nOT-
Est imatio
Creep and ShrinkQ e

Coefficient kb (composition of the concrete).

ei

-ooo

2-30

.Z
10
_y

4
0 0-2 0-4 6

WATER, CEMENT RATIO

100 90 80 70 60 50 40 30

RELATIVE HUMIDITY OF AIR Coefficicntkt (yariation as a function of time)

Coefficient kd (hardening at the age or loading).
21 111111

t22/
+0

,a F'*-ý
[1-0 1 1
_
1.1

_
iý
ng
ý 0-75-
0.7
- E',

"I 0-5ý
T- , 104 lu, 104

0
137 ISO JW days
1 10 t-
14 I's 56 90
AGE OF CONCRETEAT TIME OF LOADING (T-^. O'C-C0n5L)
(t representsthe number of days after the applicationof
11 loads).
L-1 II 111 1111 11111 1 11 1111
30 100 IUuO 10WO
DEGREEOF IIARDENING D Coefficientke(thtorctical thickness).

60

5' oo

4 Cr eep
'045
Am

0-60 0-71
gin
01 7
-70
«z
0
ShrInI age
27-5
0.65

_ v;v - i 0. 0.5 5 )-so

E

tA 0.

. 0-;

-11 7v ov lu toi NJ 40
-
10 - 20 3u 40 W
RELATIN E HUMIDITY OF AIR CM
THEORETICALTHICKNESS e M-

175
 Shrinkage strainEr (Clause R 12.32)

kk k' k
c b. qpt

where

depends on the environment

c
kb as for creep

ke depends on the theoretical thickness of the member em

(Approximately equals. to the creep values)

k depends on-the-percentageýof the steel and is given

p
as 100 to allow for restraint due to steel.
100 + 20P

kt defines the development of shrinkage'as a function

of time (as for creep).

In this example
-5
27.5 x 10
c=
kp 100
0.86
100 + 20 x 269.5 x lCO
33.44 x 1000

kk
bet and k as f. or creep.
-5
r=
27.5 x 10 x 1.0 x'1*2 X-0.86 x 0.4
-5
= 11.4 x 10 MM/MM.

B. 1.4.4. First approximation of losses:

Thý relaxation loss is .

by CEB (Clause
given R 11.22)

as a function of time by a straight-, line law of

logarithmic type by:

log ap, t k, +k2 log t

ap, o
A
where ap, t represent, s, --the loss. through relaxation
at time t (hours) and C5- ap',, o, the- initial tension.
The values of the coefficient's k anTk 2 'depend
on the type of prestressing steel.

176
For simplicity of calculations, the r'elaxation loss

will be assumed as 5% (as given by BS 2691: 1969)

Relaxation loss = 14.0 IM

Average force during loss period 280.3 - 14.0

2
= 273.3 EN

Average stress in concrete at centroid of steel

:, 3 2
273.3 10 1 e) N/=2
x X(-

9.87 N/mmý

creep strain 9.87 x 0.773

E
c28
fo0.8 2-
E 6.6 1 x 52.0 42.6 KNI=
c28
fcu2g estimated as 52.0 NImm2
-5
creep strain = 9.87 x . 773 17.9 x 10 mm/mm
U
42.6 x 10
-5
creep + shrinkage strain (17.9 + 11.4) x. lo mm/mm
-5
29.3 x 10
2-
Steel stress loss 58*6 N/mm

Steel force loss 15.8'KN

Aps f (Approximately) = 280.3 (14.0 + 15.8)

pe
= 250.5 KN

E. 1.4.5. Estimation''Of the' e'ffective prestressing, force:

Average force during loss period
280.3'', (14.0 + 15.8)
2
265.4 KN.

Average stress in
concrete,
I -: -Iý ý 1. ý 12
at centroid of the steel =*9.57 N/mm
Creep strain 17.4 x 10 mm/mm
 -5
Creep + shrinkage strain (17.4 + 11.4) x 10
-5
28.8 x 10

Steel stress loss 57.6 N/1=2

Steel force loss 15.5- KN

Aps fpe ý' 280.3 - (14.0 + 15-5)

= 250.8 KN

Total loss at test = 100 250.8 x 100

294.5
I
15.0%

B. 1.5. Total losses estimated-from s-urface strain

measurement on speckaný'Al = 16.01ZI.

The CEB and BSCP 110 Recommenda!t ions were applied

to 12 specimens theý! four seasons .

of the
covering

year, the cross sections used,, the prestressing

tendons employed and different types of, concrete

mix. CEB Recommendations, were, al%,jays very close

to the values estimated from, surface strain

measurement. BSCP 110 in. cases'studied

'allýthese
overestimated the values-of theýprestressing losses.

In all the "of the .p,re .stressing losses

calculations
in this investigation,, th '6"losses'" e stimaý-ted' from the

surface strain meas'urements were, ýused.

Use of the CEB PIP Recommendations requires,

some knowledge about the: environme'ntal conditions

and this may not be, known-toth6 designer'.,

The overestimation of the losses'_ given Vy, the

crude figures of CP, 110 is on the-safe side. Hence

in the author's opinion"-CP ilo'i s most, suitable for
design, at least for the time being.
178
 Annendix C

TABLE C. 1
Calculations steps in analysiB,
-of
one-or two-point load cases

hf b f f v
Ct -cp v cExOt
M, a37-k d t-, 2 - '
-W- N/mm f d
ct East West East West East Sest

165 2.66? 2? 5 3.29 2.25 3.00 2309 1.14

. . 04098 -, . -
of If of It to - 4080 - 2299 - 1.13
. .
3.39 1.99 3.5 0 3200 - 180C - 95 -
of ý . . .
11 It If 3.42 1.85 2.00 4089 - 2304 - - 90
. . .
OPT of of If is it to 4 015 2263 89
. .
19 If it If it 2.18' 11, 46? 9 2636 1.03
. . . -
If it If 34-)34 2.04 5069 2856 1.12
. . -
It -
it, is to 01 - 4033 22? 2 - 89
. . .
oror it it it to of it 2632 2515 1.02 99
. . .
of If It 3.42 1.95 it 43ý? ' '2466
. . - 9? -
.
it it It it of to 4? 11 - 2654 - 1.04
. .
to of to it 2.13 'Iý 4809 2? 09 2582 1.06 99
. 04464 . . .
of it It it to III 464.0 2614 - 1.02
. . -
-6 11 11 of 3.56 2.06 3.00 - 3696 - 2082 1.02
, -- ý, . . ý- -
is it It i , II% III. to
-

7 3.27 2.00 3706 2088

. -- .
it If #I If 3438 4056 1937 2285 95 1.12
. . . . .
2.47 3.01, - . 4080 - 2299 - 1.13
.
If' It to It it of it 3891 7 2192 1.08
- . . - -
It it If 2.98 2.46 3839 1.06
. e2163
It of If is it It- 3859 2174 1.07
ý. .
JLO it of If 3.05 2.15 4308 3669 2427 1.19 1.02
. . . -2067
is 2.91 1.85 3.50 3240 3550 1825 2000 97 L 06
. . . . .
it it of It it it it 8374 - 1901 1.01
- . . - -
3-2 is It it 2.75 2.13 of 3511 1978 1830 1.05 0.97
. 03248 . .
131 . 179 4.000 537 3.10 1.90 4.00 3267 3738 1835 90 1.03
, . 41'1604 . .
2 of If it 3.26 2.17 '1701 1755 *.96 . 99
*3464 *'3575 .
:3 it it of 3.52 1.93 4
2.00 -4884 . 5215 . 2398 2560 94 00
. . .
it of of to to to 5439 2670 05
. . - .
4 #1 If to 3.19 2.21 '4881 4992 2451 94 90-
. . *230 . . .
of If it If If 2*30 L 5980E
ý
19 2936 2383 1.15, 93
. *'4854 . t.
1- .
5 It of to 3.28 l.? l 4.5 339? 99
. - . . -
-I-- ý1668

179
 TABLE C-1 (Cont'd)

I hf f1m fc! I Ve
e ýMnl b
ct2l 3 av p Ex nt.
d bw fýt d vc
aýk I /M Cý9,1c.
ast lWest lEast West West

17914.0001.53713.311 1-811,1.25 1.65031 1.31121 92 89

. -63391-3192 . .
1 11 7610 7252 3736 3560 1.07 02
. . . . .
3.25 1.95 i, 7006 7006 7006 3439 99 99
. . . . . .
It . 7979 6638 3917 3259 1.13 94
. . . .
96724 - . 3301 - 95
.
3.23 1.9-5 6187 . 6189 . 3037 3037 87 87
. . . .
7443 7977 3654 3916 1105 13
. . . . .
3.30 1.71 4.50 3142 - 1542 - 91
. . .
: Lo 3.23 2.32 3.50 3916 - 19ý2 1.02
. .
1.50 5951 - 2921 - 95
. . .
3.36 2.34 2.75 3889 - 1909 - 90
. . .
*3960 . 4188 2056 91 97
91944 . . .

cl 241 2.667 401 3.24 1.78 2.00 4902 2580 01

. . . . .
IT to If It 4874 -- 2565 - 1.01
. .
If 2; 07 t' 5172 '. 4439 2722 23,90- 1.01 92
. . . .
It If 3.11 1.91 1.25 5208 6797 '. 2741 3597 79 03
. . . . .
It of it It it 71? 0 '. ? 170 3774 3774 1.08 08
. . . .
It it to 3.45 1.83 4.00 - 1644 - 93
93124 . .
4 It 3.35 1.91 2982 - 1563 - 88
. .
5 3.27 1.73 2.00 4357 - 2293 - 90
. . .
If It 4872 '. 4782 2564 2517 1.01 99
. . . .
ft It If 4536 - 2387 - 94
. . .
R 9) fl A
41j %.j w I- 0 rý I UU J. Ovo
0 1610
6 3.33 1.94 1.25 6125 -.6579 3224 3463 93 1.00
. . . .
96571 -- . 2985 - 86
.
o5559 . 2926 -* 84
.
. 7273 6809 3828 3584 1.10 1.03
. . .
7 3.26 1.80 5566 6364 2929 3349 84 96
it . . . . 9 . . 89
6606 2Z? 7 1 30 1:
. . .
3.06 2.04 2.00 - 3970 2089 82
. . .
. 4362 - 2296 - 90
. . L.-
5009 2636' 03
. .
 TABLE C. 1 (Cont'd)
I I
v
iýýt
Be am h b f fcT) 'v -
o" ct ý ý VcCalc.
mark -jjf -7,7 2 f d j
W I/Mm ct
East' West East West East Iýest
C9 241 2.667 401 3.01 1.90 1.25 6102 . 6102 . 3212 . 3212 92 . 92
. . . .
of it it it It tl it- 7227 22? 3804 3804 1.09 1.09
. .? . .
10 it to 2.96 1.95 4.00 - 3025 ,
-, . 1592 90
. .
11 Ir it it 3.06 1.60 3267 . 3090 . 1719 . 1626 97 92
. . .
12 it It ft 2.? 0 1.93 3256 . 8358 . 1? 14 . 176? 96 1.00
. .
Dl 255ý 4.00 764 3.19 2.00 2.00 . 5269 . 5269- . 2326 . 2326 9.L . 91
. . .
it If it it ff it ff 5625 5269 . 2483 . 2326 97 91
. . . .
2 11 it of 3.21 2.01 11 - 5385 - 23? 7 - 93
. . .
3 it it it 3.35 1.60 3.00 3372 . 3953 . 1489 . 1? 45 '. 73 '. 86
.
4 3.29 1.53 4.50 3280 - 1448 - 86 -
. . .
5 it- it it 3.06 l.? 4 4.00 3951 - 1745 98 -
. . .
6 It It 3.39 1.58 3.00 4418 - 1951 - 96 -
. . .
7 it it 2.73 1.03 6.00 5614 3744 . 2339 . 1560 1.54 1.0Z
. .
8ý it 3.04 89 11 - 3620 - 1308 - 9c.
. . . .
9 3.27 83 5.25 3639 - 1516 - 95 -
. . . .

El 314 2.667 524 3.74 1,17 3.00 4377 . 4414 . 2159 . 21? 8 1.06 1.0r,
. . .
it It it it it it it 4483 - 2212 - 1.09 -
. .
it it it to it it 2.00 5307 5022 . 2618 . 2478 1.03 9,
. .
2 11 It it 3.33 1.00 11 - 5252 - 2591 - 1.0t, a
. .
it it it it 11 Is it 5761 - 2842 - 1.11
. .
It, to it It it It It 5312 - .
2621 - 3
.
3 of it if 3.37 1.45 1, 5202 - 2566 - 1.0, L
. .
it tt It of of it - 5202 2566 1.0, L
it -- . -
.
of of ff it it 1.53 11 5127 5707 . 2529 . 2815 99 1.11 )
. . .
4 to 3.29 1.57 4.00 3372 3468 1664 1711 94
. . . . . .9
.5 It, 3.17 1.79 11 - 3858 - 1903 - 1.0 7
. -.
it it to 4732 5169 23204 . 2550 92 1.0 0
it 2.00 .
. . .
6 it it it 3.22 1.96 3.00 4892 4245 2413 . 2094 1.19 1.0 3
. . .
7 3.10 1.80 3869 - 1909 - 4
. . .9
of it 3996 1971 97 -
. - . - .
of it. it it 4182 2063 - 1.01 -
. - .
8 if 3.52 1.52 4.00 -
- - - - -
CIT 283 3-OC7 472 2.73 93 6.00 3110 - 154 1.03
. . -. . . - -
ft it it 3.04 82 3.00 2730 138 91
-. - . - .
19 of tr 3.27 75 5.25" 3200 162 1. Q
. . - . -

Note: C17, C18, and C19 are shown in the bottom of the page.
 TABLE C. 1 (Contd)

f f 13
3eam h4 b av Vc Expt.
f ct
ct CP -d -
nark --d- - b%v Cý
tm2 f I.
ct
--
ve%
I r%
Eastl West EastlWest East In
West
I
:Fj 830 4.000 990 3.14 1.45 4.00 4057 2911 1629 1169 92 66
. . . . . . .
2 3.21 1.46 3946 3792 1585 1523 89 86
. . . . . .
3 3.41 1.50 2. CO 5741 5936 2306 2384 90 93
. . . . . .
1.57 6123 6527 2459 2621 96 1.03
. . . . .
4 3.28 1.70 5684 3684 2283 2283 90 90
. . . . . .
1.79 5910 6557 2373 26ý`2 93 1.03
. . . . .
3.33 1.00 4.5 - 3526 - 1416 - 84
.5 . . .

GI 000 1.000 000 3.80 80 2.00 S558 - 2367 - 93 -

. . . . .
it 3825 - 2550 - 1.00 -
. .
2 3.40 1.10 3760 T 2507 - 98 -
. . .
3 3.34 60 - 2958 - 1972 - 77
. . . .
- 3427 - 2285 - 90
. . .
70 4 03o6 - 2691 1.06
% .
4 3.35 1.06 4.00
5 2.94 1.27 2.00 3035 - 2023 - 79
. . .
1.27 2.00 - 3363 - 2242 - 88
. . .
3.23 1.20 3.00 2552 - 1701 - 84
. . . -
2.00 0.0 . 3002 - 2001 - 78
. .
1.30 It 4073 2715 1.06
. .
7 3.26 1.04 1.50 - -
8 3.39 1.13 ý. Oo -

'llotes.
hf 13 1000 Ve and

I+r cp bw d
ct f
Vc ct
1000
f
(1+ LM)b (hb hf
fc't d 1,5 + I
.f, w w
ct
Vc Expt.
-2* v Experimental
v c
c Calc. v Calculated,
c us i ng equation 6.5

182
 Appendix D

TABLE D. 1

Calculations steps in analysis of

uniformly distributed load case.

F
eam h b f f
f C< ct cp v
a 37k -5 - T c. Pxnt
b 2 d ,-
w N/mm ct '
c Calc.
East, West East West East lest

133_2 179 4.000 53? 3.09 1.45 10.0 1.530 1.330 7511 653C 1.05 91
. . . . .
13 It it of 2.93 1.61 8.0 1.643 1.643 8066 8? lC 1.00 1.08
. .
3-4 of it it 3.13 1.94 7.28 1.642 1.806 8060 886C 95 1.04
. . .
3-5 tr It if 2.93 2.15 6.0 1.891 1.891 9280 840 98 88
. . . .

C]L3 . 241 2.667 401 3.13 1.21 10.0 1.447 7620 1.06
. . - -
IL4 of If 3.11 1.24 8.0 1.605 l.? 02 8450 856 1.05 1. OC-
. .
:L5 it 3.13 1.62 7.28 1.638 1.774 8620 934 1.01 L 1C
. .
-3.6 it it 2.93 1.84 6.0 1.663 1.663 8750 875 92 92
. . . .
20 283 472 2.67 1.05 11.78 1.649 8362 1.27
. . - . - -

jc):Lo 300 4.000 900 2.67 1.11 ll.? 8 1.718 7158

. . - . - - LOS
,:Lla 3.450 735 3.45 0.90 17.78 1.333 5963
. - . - - 1.1c
:Llb 3.41 0.90 1.184 1.367 5298 6116 0.98 1.1-2-4
. .
:L2 3.64 1.07 11.78 1.627 1.312 7280 5870 1.10 0.9C
. .
:L3 3.11 1.06 17.47 1.091 4881 90
- . - . -
3.4 3.02 1.14 1.239 5544 1-OiS
- . -
9 314 2.667 524 2.91 1.42 10 0 1.524 7520
. . - . - 1.03-
:Lo of 2.95 1.46 8.0 1.662 1.707 8200 8420 1.01 1.04
. .
:Ll it 3.10 1.61 6.0 1.934 1.934 9540 0540 1.00 1. OC
. .
: L2 it 3.00 1.49 7.28 2.007 1.858 9900 9170 1.16 LOE
. .
6 330 000 990 2.98 1.38 10.0 1.891 1.796
. . .. 7590 7210 1.06 L OC
. .
7 2.91 1.52 8.0 2.198 2.198 8830 8830 1.09 1.0s
. .
8 2.95 1.71 6.0 2.229 2.079 8950 8110 95 8q
. . . . % w
-9 Il 3.01 1.50 7.28 2. CO8 2.053 8060 8240
1" . 95 911,
. .
0
_.
ljoteq:
c< b/bvl- 1 hf/d 1000 qcL)/(fct (1+ fcp/fCt)
IL bwd)
(j()oo q, )/(fct (1+ fcplft ct) bwd(l. 5 +'(b/bw 1) hf/d)
-
q qc L Ex --
ct Ernt nP PI rnan + in.
qcL Cale. qc calcul ated u8ir4l. equation 66.12
183
 A,ppendix E

E. 1 Mohr's ýailure Criterion Assuming a Straight

Line Envelope:

From the geometrical relations in Figure E. 1,

and talking compression as positive.

f1+f2 `ý 2 OCI OC CE 0."3 where CE = CD

ff t
0 FDE TE Q
cosec since and =c cot
22
fl f2 9
j cosec cot
2
2 OC ý f, +f2= (f' 1f2 )cosec- 2c cot 9 (1)
or

With the same sign convention, the following

expressions can be deduced from the well known

principal stresses:

fl - f2 fxx + fyy
(2)
f+f (fxx fyy)2
12 _ +4 V2
XY
I
(1) and (2)
Prom
V(PXX 22 (f
- fýy) +4 vxy , x. + fyy) cosec -2 c cot G* (3)

From the geometrical relations in Figure E. 2 with

compress ion taken as positive

/.
-- f *ff
GQ = HK M- PQ where ' PQ c-I sin o
2 2
f, 1
GQ cf (% _f (4)
2-2 sin

E f, t fc ff
and K - c-t sin 0 (5)
222

From (5)

fc
- I+ sin 'G (6)
ft 1- sin 9
Solving (6) fore gives E) sin-' fe + ft (7)

184

. eý
 From Figure E .2:
fIf 04

t OK t
OE = EK + OK where EK cosec 0 and
22

Lt"
OE +s -*L
n
2 sin 9

tan c2c sinG

M, + sine)

or I
ft (1 + sinG (8)
c
2 cos 9

From (7) and (8)

(9)
c=- 2
Substituting (7) and (9) into (3), we'get:

F(fxx f ýV. (f fe'fý

2c V) 2
fyy) +4 vxy
fc + ft)11 (fc +f

or
(f 2 2 (fxx
+f f= -\2 +4 Vx Y) = +212
cf lyý Cf+
#,2 ,2fI(
4ft fe + 4f'f'(fxx+fyy)(f'-f-,
tcc t)

After simplifying, rearranging, dividing by fý4

and substituting % as c we get

ft ,-

+0 X-X+
t+ t
fxx
. xx

186
 REPERTENCES

SHEAR STUDY GROUP. The shear strength of reinforced

concrete benms. London, Institution of Structural
Engineers, January 1969. pp. 1 -170.
.

2. A. S. C. E. - A. C. I. CMUTTEE 426. The shear strength

of reinforced concrete members. Proceedings of
the American Society of Civil Engineers. Vol. 99,
No. ST 6. June 1973. pp. 1091 - 118?.

3. ANDERSO, B. G. Rigid frame failures. Journal of

the American Concrete Institute. Proceedings
Vol. 53, No. 7. January. 1957. pp. 625 - 6:36.

4. IJAHGOUB, M. O. Shear prestressed

strength concrete
of
beams without web reinforcement. Magazine of
Concrete Research. Vol. 27, No. 93. December
1975. pp. 219 - 228.

5. HOGNESTAD,E. 'Uhat do we know about diagonal tension

and web reinforcer. ent in concrete? Urbana,
University of Illinois, Engineering Experiment
Station, 1951. Circular Series No. 64, pp. 47.

6. A. C. I. - A. S. C. E. COTMAITTEE326. Shear and diagonal

tension. Journal of the American Concrete
Institute. Proceedings Vol. 59, Nos. 1 3.
January to March 1962. pp. 1- 309 277 333
and 353 - 395.

7. BRESLER, B. and YACGREGOR,J. G. Review of concrete

beams failing in shear. Proceedin7s of the
American Society of, Civil Engineers. Vol. 93,
No. ST. 1. February 1967. pp. 343 - 372.

187. "''
 GVOZDEVj A. A. Research on prestressed concrete
in U. S. S. R. and neighbouring Europeaii

countries. Fifth Congress of the Federation

Tnternationale de la Precontrainte,
-Paris,
1966. London, 1967. pp. 12 - 29.

CLAM, A. P. Diagonal tension in reinforced

concrete beams. Journal of the American
Concrete Institute. Proceedinp! s Vol. 48,
No. 2. October 1951. pp 145 - 156.

10. TAMOT, A. N. Tests reinforced concrete beams:

-on
Resistance to web stresses. University Urbana,
, Station,
of Illinois, Engineering Experiment
January 1901. pp. 85. Bulletin No. 29.

MOODYj K. G., VIEST, I. M., ELSTNER, R. C., and

IIOGNESTAD, E. Shearýstrength of reinforced
concrete beams. Journal of the American
Concrete Institute. Proceedings Vol. 51,
Nos. 4 and 5. December 1954 and January
1955. pp. 317 332 and 417 - 4,C.
A.

12. ELSTNER, R. C., JrOODYj K. G., VIEST, I. M. and

HOGNESTAD, E. Shear strength of reinforced
beams: Tests of restrained beams with web
reinforcement. Journal of the American
Concrete Institute. Proceedipzs Vol. 51,
No. 6. February '19,55. pp. 525 - 539.

13. MOODY, R. G. and VIEST'I I. M. Shear strength of

reinforced concrete beams: Analytical
studies.
Journnl of the American Concrete Institute.
Proceedings Vol. 51, No. -7. Mlarch 1955.
pp. 697 - 730.

ise.
14. '/ I J. and VIEST ,*I.
LFORIIU, M. Shear strength of
reinforced concrete frame members without
web reinforcement. Journal of the American
Concrete Institute. Proceeding8 Vol. 53,
No. 9. March 1957. pp. 833 - 869.

KESLER, C. E. Statistical relation between

cylinder, modified cube and beam strength
of plain concrete. Proceedings of the
American Society for Testinp,, Materials.
Vol. 54.. 1954. pp. 1178 -, 1187. *

16. WHITNEY, C. S. Ultimate shear strength of re-

inforced concrete flat slabs,, footings,
beams and frame members without shear
reinforcement. Journal of the American
Concrete Institute. Proceedings Vol. 54,
No. 4. October 195?. pp. 265 - 298.

17. TAUB, J. and NEVILLE, A. M. Resistance to shear

of reinforced concrete beams. Journal of
the American Concrete Institute. Proceedings
Vol. 57,Nop. 2 - 5. August - November 1960.
pp. 193 - 220,315 - 336,443 - 463 and
517 - 532.

SMITH, R. B. L. The interaction of moment and shear

on the failure of reinforced concrete beams
without web reinforcement. Civil Engineering
and Public Works Review. Vol. 61, Nos. 719,
720 and ? 21. June - August 1966. pp. ? 23 -
7259 869 - 8? 2 and 967 - 970.

KREFELD, W.J. and THURSTON, C. We, Studies of the

shear and diagonal tension strength of simply
supported reinforced concrete beams. Journal
of the American Concrete Institute. Proceedings
Vol. 63, No. 3. ýJarch 1966. ppo 451 - 476o

189
20. DESAYI, P. A method for determining the shear
strength of reinforced concrete beams with
small av/d ratios. Magazine of Concrete
Research, Vol. 26, No. 86. March, 1974.
pp* 29 - 08. --

21. LAUPA, A., SIESS, C. P., and NZ7MARK, N. M. Strength

in shear of reinforced concrete beams. Urbana,
Universitý of Illinois, Engineering Experiment
Station, March 1955. pp. ? 3, Bulletin No. 428.

22. WATSTEIN, D. and YWPHEY, R. G. Strains in beams

having diagonal cracks. Journal of the
American Concrete. Ins-titute. Proceedings
Vol. 55, No. 6. December. -1958. 'pp.
717 - 728*

23. KANI, G. N. J. The riddle of shear failure and

its solution. Journal of the American Concrete
Institute. Proceedings Vol. 61, No. 4.
April 1964. pp. 441 - 467* - I;

24. REGAN, P. E. Combined and bending of

shear
reinforced concrete members. Thesis submitted
to the University of London for the degree
of Ph. D. April 1967. pp. 722,

25. VIALTMER,R. Calculation of the shear strength of

reinforced and prestressed concrete beams by
the shear failure theory. Beton-und Stahlbetonbou.
Vol. 57, No. 11. November, 1962. pp.. '261*- 271.
English translation by C. V. Amerongen. London,
Cement and Concrete Association. 1964.,, pp. 28.
C& CA Translation-No. -Cj-110., -

190
26. OJIIAp S. K. The shear strength of' rectangular
reinforced and prestressed concrete beams.
Magazine of Concrete Research. Vol. 19,
No. 60. September 1967. pp. 173 - 184.

27. COMITE EUROPEENdu 13ETCN- FEDERATION INTERNATICNALE

de la PRE
-CONTRAINTE. International recor=end-
ations for the design and construction of
concrete structures. London, Cement and
Concrete Association, 1970. Principles and
recommendations. pp. 80.

28. GVOZDEV, A. A., D111ITRIEV, S. A. and KALATOUROV,B. A.

Russian limit design method for prestressed
concrete. A contribution in French to a
meeting of the F. I. P. - C. E. B. Joint Committee
No. 1, Venice, 1963. English translation
by C. V. Amerongen. London, Cement and
Concrete Association. 1964. pp. 16. C& CA I.

Translation No. Cj- 117.

29. SIGALOV, E. and STRONGIN, S. Reinforced concrete.

Moscow, Foreign Languages Publishing House,
1962. pp. 393*

30. BRITISH STANDARDSIlfSTITUTION. CP 110*. Part 11.

1972. The Structural-use of
concrete. Part 1.
Design, materials and workmanshi2. London.
pp. 154*

'31. BRITISH STANDARDSINSTITUTION. CP 115: 19590

The structural use of prestressed concrete
in buildings. London. pp. 44.

32. AMERICAN CONCRETEINSTITUTE. ACI 318 - 71.

Building code'requ ,irements for reinforced
concrete., Detroit, American Concrete
Institute, 1971. pp, 78.

. 191
33. HICKS, A. B. The influence of Shear span and
concrete strength upon the shear resistance
of a pre-tensioned prestressed concrete
beam. Magazine of Concrete Resenrch.
Vol. 10, No. 30. November 1958. pp. 115 - 122.

34. SETHUNARAYANAN,R. Ultimate strength of pre-

tensioned I beams in combined bending and
shear. Magazine of Concrete Research.
Vol. 12, No. 35. July 1960. pp. 83 - 90.

35. SETHUNARAYANANIR. Ultimate strength of pre-

stressed beams in combined bending and
shear. A paper presented to the
Internationsl Conference on Shenr,
--Torsion
and Bond in Reinforced and Prestressed
Concrete, Coimbatore, India, 1969. pp. 20.

SOZEN, M. A., ZWOYERyE. M. and SIESSI, C. P.

Strength in shear of beams without web
reinforcement. Urbana, University of
Illinoist Engineer-ing Experiment Stations
April 1959. pp. 69. Bulletin No. 452,

37. EVANS, R. H. and HOSNY, A. H. H. The shear

strength of pobt-tensioned prestressed
concrete beams. Third Congress of the
Federation Internationale de la Precontrainte,
Berlin, 1958. London, 1958. pp. 112 - 132.

38. WALTHER, R. The shear strength of prestressed

concrete beams.. Third--Conpress of the
Federation Internationale de la Precontrainte'.
Berlin, 1958. London, 1958. pp. 80 - 100.

WARNER, R. F. and HALL, A. S. The shear strength

of concrete beams without web reinforcement.
Third Congress of the Federation Internationale
de la Precontrainte, Berlin, 1958. London, 1958.

pp. 101
192'
40. EVANS, R. H. and SCIRMIACHER,E. G. Shear strength
of prestressed beams without web reinforcement
Journal of the American Concrete Institute,
Proceedinq. q Vol. 60, No. 11. November 1963.
-
pp. 1621 - 1642.

41, SVIAMY, '11. Some aspects of the diagonal cracking

failure of prestressed hollow beams without

web reinforcement. The Indian Concrete

Journal. Vol. 38, No. 9. September 1964.

pp. 828 - 333.

42. KAR, J. N. Diagonal cracking in prestressed

concrete beams. Proceedings-of the American
Society of Civil Engineers. Vol. 94, No. ST1.
January. 1968. pp. 83 - 109.

43. KAR, J. N. Shear strength of prestressed concrete

beams without web reinforcement. Magazine
of Concrete Research. Vol. 219 No. 68.
September 1968. pp. 159 - 170.

44. ARTHUR, P. D. The shear strength of pre-tensioned

I- beams with unreinforced webs. Thesis
submitted to the University of St. Andrews
for the degree of Ph. D. June 1968. pp-64.

45. ARTHUR, P. D. and IMAHGOUB,M. O. Shear test on

full-scale prestressed concrete box beams.
Proceedings-of the Institution of Civil
Engineers. Part 2. Research and Theory,
Vol. 59. September 19? 5. pp. 535 - 540.

46. WILBY, C. B. and NAZIRI C. P. Shear strength of

uniformly loaded prestressed concrete beams.
Civil Engineeringand Public 'lorks Review.
Vol. 59, No. 693. April 1964. pp. 457 - 461.

193
4?. HANSON, J. M. and HULSBOý, C. L. Ultimate shear
tests of prestressed concrete I beams under
concentrated and uniform loadings. Journnl
of the Prestressed Concrete Institute. Vol. 9,
No. 3. June. 1964. pp. 15 - 28..

48. ARTHUR, P. D., BHATT, P. and DUNCAN, W. E-..perimental

and analytical studies on the shear failure
of pretensioned I beams under distributed
loading. Journal of the Prestressed Concrete
Institute. Vol. 18, No. 1. January -
February 19? 3. pp. 50 - 67.

49. BENNETT, E*W*, ABDUL-AHAD, H. Y., and NEVILLE, A. M.

Shear strength of reinforced and prestressed
concrete beams subject to moving loads.
Journal of the Prestressed Concrete Institute.
Vol. 17l No. 6. November - December 1972.
PP. 58 - 69.

500' MACGREGORI'J. G., SOZEN, M. A. and SIESS, CePe

Strength of prestressed concrete beams
witn
web reinforcement. Journal of the American
Concrete Institute. Proceedinps. Vol. 62.
No. 12 December 1965. pp. 1503 1519.
-

51. MACGREGOR, J. G., SOZEN, M. A. and SIESS, C. P.

i
Effect of draped reinforcement on behaviour
of prestressed concrete beams. Journnl of
the American Concrete Institute. Proceedings
Vol. 57, No. 6. December 1960. pp. 649 - 677.

52. JENA, B. and PANINIELL, F. N. The diagonal cracking

strength of continuous prestressed concrete
beams. Magazine of Concrete Researcn. Vol.
24, No. 78. Yarch 1972. pp. 3- 14.

194
53. BHATTp P. The diagonal cracking strength of
prestressed I beams with unreinforced webs.
Building Science. Vol. 9, No. 4. December
1974. pp. 315 - 323.

54. MAYER, M. and REGAN, P. E. Shear strength of

orestressed and reinforced concrete T
beams. Shear in reinforced concrete.
Detroit, American Concrete Institute,
19? 4. Publication SP - 42. Vol. 1.
pp. 183 - 213.

55. DUNAR, J. D. The indirect tensile strength of

concrete of hiRh compre-ssive strength,
London, Cement and Concrete Association.
March. 1964. pp. 12. Technical Report
TRA/3??
*

56. ZIENKIEVIICZ, O. C. The finite element method

in engineering science. London, McGraw Hill,
1971. 521.
-pp.

57. NIISONI A. H. Nonlinear analysis of reinforced

concr. ete by the finite element method.
Journal of the American Concrete Institute.
Proceedinr! s Vol. 65, No. 9. September 1968.
pp. 757-7669

58. SCORDELISj A. C., NGO, D. and FRANEUN, H. A.

Finite element study of reinforced concrete
beams with diagonal tension cracks. Shear
in reinforced concrete. Detroit, American
Concrete Institute, 1974. Publication
SP - 42. Vol. 1. pp. 79 - 102,

59. HOUDE,J. and MIRZA, M. S. A finite element analysis

of shear strength of reinforced concrete beams.
Shear in reinforced concrete. Detroit, American
Concrete Institute, 1974. Publication SP - 42
Vol. 1. pp. 103 - 128.
195
60. FERGUSON, PAI. and THOTOSON, J. N. Diagonal
tension in T- beams without stirrups.
Journal of the American Concrete Institute.
Proceeding Vol. 49, No. 7. March 1953.
pp. 665 - 676.

61. LEONHARDT, F. and WALTHER, R. The Stuttgart

shear tests. Beton-und Stahlbetonbau.

Vol. 56, No. 12.1961 and Vol. 57, Nos.
21 31 61 7 and 8.1962. English translation
by C. V. Amerongen. London, Cement and
Concrete Association, 1964. pp. 134.
C& CA Translation No. Cj 111.

62. PLACAS, A. and REGAN, P. E. Shear failures of

reinforced concrete beams. Journal of
the American Concrete Institute. Proceedings
Vol. 68, No. 10. October 19? 1. pp. ? 63 - ?? 3.

63. KANI, G. N. J. How safe are our large reinforced

concrete beams? Journal of the American
Concrete Institute. Proceedings Vol. 64y
No. 3. Mardh. 190. pp. 128 - 141.

64. TAYLOR, H. P. J. Shear strength of large beams.

Proceedings Of the American Society of Civil,
Engineers. Vol. 98, No. ST 11. November
1972. pp. 2473 - 2490*

65. WRIGHT, P. J. F. Comments on an indirect tensile

test on concrete cylinders. Ylagazine of
Concrete Research., Vol. 7, No. 20. July,
1955. pp. 87 - 96.

66. OLADAPO, 1.0. failure

Cracking
and in plain
concrete beams. Magazine of Concrete Resear
Vol. 16, No. 4?. June 1964. pp. 103 - 110.

196
67. GOODE, C. D. and HELMY, M. A. The strength of
concrete under combined shear and direct
stress. Magazine of Concrete Research.
Vol. 19, No. 59. June 1967. pp. 105 - 112.

68. BRESLER, B. and PISTER, K. S. Strength of concrete

under combined stress. Journal of the
American Concrete Institute. ProceediLirs
Vol. 55, No. 3. September 1958. pp. 321 -
345.

69. COIWAN,H. J. The strength of plain, reinforced

and prestressed concrete under the action
of combined stresses)with particular reference
to combined bending and torsion of rectang-
ular sections. Magazine of Concrete Research.
Vol. 5, No. 14. December 1953. pp. 75 - 86.

70. KRAHLI N. W., KHACIIATURIAN, N.,

and SIESS, C. P.
Stability of tensile cracks in concrete
beams. Proceedings of the American Society
of Civil-Engineers. Vol. 93, No. ST 1.
February, 1967. pp. 235 - 254.

00
71. KUPFER, H., HILSDORF, H.,
and RUSCH, H. Behaviour
of concrete under biaxial stress4S. slnou-.
rnsl
of the American Concrete Institute.
Proceedings Vo. 66, No. 8. August 1969.
pp. 656 - 666*

? 2. LIUj T. C. Y., NILSON, A. H. and SLATE, F. O.

Biaxial stress-strain relations for concrete.
Proceedings Of the American Society of
Civil Engineers. Vol. 98, No. ST. 50 May
19? 2. pp. 1025 - 1034*

73. NADAII A. Theory of flow and fracture of solids.

Vol. I. Second edition. New York, 111jcGrawHill
Inc. 1950. pp. 572.
,

197
74. GURALNICK, S. A. Strength ft
of reinforced concrete
beams. Transactions of the American
Society of Civil Engineers. Vol. 125,
Part I, Paper No. 3036.1960. pp. 603 - 645.

75. SHEIKII, M. A., de PAIVA, H. A. R. and NEVILLE, A. M.

Calculation of flexure - shear strength
of prestressed concrete beams. Journal of
the Prestressed Coficrete Institute. Vol. 13,
No. 1. February. 1968. pp. 68 - 85.

76. MUKHOPADHYAY,M. and SEN, B. R. Ultimate load

analysis of reinforced concrete members
with plain
web under combined flexure,
shear and axial force. Building Science.
Vol. 9, No. 3. September 1974. pp. 167 - 174.

77. GUYONj Y. Prestressed Vol. II.

concrete.
Statically indeterminate structures.
London, _
Contractors Record Ltd., 1960y
pp. 74le

78. LANGHAAR, H. L. Dimensional analysis and theory

of models. New York, John Wiley and Sons
Inc. London, Chapman and Hall Ltd., 1951.
pp-1660

IYENGAR, K. T. S. R., RANGAN,B. V- and PALANISWAI.

ff, R.
Some factors affecting shear strength of
reinforced concrete beams. Indian Concrete
Journal. Vol. 429 No. 12. December 1968.
pp. 499 - 505*
-

80. ROAD RESEARCHLABORATORY. Desi of concrete rixes.

Second edition. London, H. M. S. O., 1950.
pp. 16. Road Note No. 4.

198
81. HURD, M. K. Formwork for concrete. Second edition,
Detroit, American Concrete Institute, 1969.

pp-364. ACI Special Publication No. 4.

82. MARSHALL, W. T. and MATTOCK, A. H. Control of

horizontal cracking in the ends of pre-
tensioned, prestressed concrete girders.
Journal of the Prestressed Concrete Institute.
Vol. 7, No. 5. October 1962. pp. 56 - 74.

83. LEONIIARDTj F. Prestressed concrete. Design and

construction. Second edition. Berlin,
Wilhelm Ernst and Sohn, 1964. pp. 6??.

84. MORICE, P. B. Full

scale test on a prestressed
concrete beam. Engine_ering, Vol. 1? 79
Nu. 4613.25 June 1954. pp. 815 - 817.

85. BATE, S. C. C. 'et al. Handbook on the Unified

Code for structural concrete (CP 110:
1972). London, Cement and Concrete
Association, 1972. pp-153.

HSU, T. T. C., SLATE, F. O., STUPMAN, G. M. and

WINTER, G. Midro-cracking of plain concrete
and the shape of the strees-strain curve.
Journal of the American Concrete Institute.
Proceedings. Vol. 60, No. 2. February
1963. pp. 209 - 224*

87. HSU, T. T. C. Mathematical analysis of shrinkage

stresses in a model of hardened concrete.
Journal of the American Concrete Institute.
Proceedings Vol. 60, No. 3. March. 1963.
pp. 3? 1 - 390.

199
 88. HSTJsT. T. C. and SLATE, F. O. Tensile bond strength
between aggregate and cement paste or mortar.
Journal of the American Concrete Institute.
Proceedipgs Vol. 60, No. 4. April 1963.

pp. 465 - 486*

89. BALFOURý A. and McTEPUAN, A. J. The numerical

solution of equations, London. Heinemann
Educational. 1967. pp. 85.
00
90. RUSCH, H. Researches toward a general flexural
.
theory for structural concrete. Journal of
the American Concrete Institute. Proceedings
Vol. 57, No. l. July 1960. pp. 1- 28.

STUIUAN, G. M., SHAH, S. P. and WINTER, G. Effects

of flexural strain gradients on micro-crack-
ing and 'stress-strain behaviour of concrete.
Journa the American Concrete Institute.
-of
Proceedings Vol. 62, No. 7. July 1965.
pp. 805 - 822e
92. CLARK, L. E. GERSTLE-yK*Hej and TULIN, L. G. Effect
)
Of strain gradient on the stress-strain curve
Of mortar and concrete. Journal of the
Ainerican-Concrete Institute. Proceedings
Vol. 64, No. 9. September 1967.
pp. 580 - 586.
93. DESAYI, P. and KRISHNAN, S. Equation foi- the
stress-strain curve of concrete. Journal of
the American Concrete Institute. Proceedings
,
Vol. 61, No. 3. March 1964. pp. 345 - 350*

94. SMITH, I. A. Calculations of ultimate moment

capacities of-prestressed concrete beam.
Computer program No. 10? 0. Civil Engineering
Department, University of Glasgow. 1974,
f
CARINO, N. J. and SLATE, P. O. Limiting tensile
strain criterion for foilure of concrete.
Journal of the American Concrete Institute.
Proceedings Vol. 73, No. 3. March 1976.
pp. 160 165.

200
 JONES, L. L. A theoretical solution for the
ultimate strength of rectangtilar reinforced
concrete beams without stirrups. A paper
presented to the European Conunittee for
Concrete at Wiesbaden, 1963, pp. 12.

97. TANG2 MATI-CHUNG. Shear design of large

concrete box girders. Shenr in rcinforced
concrete. Detroit, American Concrete
Institute, 1974. Publication SP - 42.
Vol. 1. pp. 305 - 319.

ýGLg GOW
x
v17,
,V

201