CHAPTER 2 BACKGROUND AND LITERATURE REVIEWshodhganga.inflibnet.ac.in/bitstream/10603/10080/7/07_chapter 2.pdf · CHAPTER 2 BACKGROUND AND LITERATURE REVIEW 2.1 GENERAL ... Dowel action

11

CHAPTER 2

BACKGROUND AND LITERATURE REVIEW

2.1 GENERAL

In this chapter a study of shear behaviour in RC beams and its

background is initially discussed. Subsequently a review of some of the steel

RC slender and deep concrete beams has been dealt with. Finally FRP

reinforced slender and deep concrete beams are reviewed.

In order to correlate FRP and steel reinforced concrete beams

effectively, the need for a study of steel reinforced concrete slender and deep

beams has been felt and this has also been carried out.

The research, performed throughout this project involves the use of

FRP in reinforced concrete deep beams which is a novel approach for the

analysis of the deep beam. There has been limited research work performed

using FRP as reinforcement in deep beams and hence, only a small number of

publications are available for reference work in this regard. The experimental

research work completed till date by other researchers with respect to FRP

reinforced deep beams is limited to FRP being used as main bars only,

without any web reinforcement. Till date, there is no data of any experimental

work carried out with respect to reinforced concrete deep beams with FRP

being used as web reinforcement. The few literatures pertaining to FRP

reinforced concrete deep beams which been published till date has been

reviewed in this chapter.

12

2.2 SHEAR BEHAVIOUR OF STEEL REINFORCED

CONCRETE BEAMS- BACKGROUND REVIEW

2.2.1 Shear and Its Importance

The study of shear behaviour in concrete structures has been going

on since a century. It was until the year 1955, when the shear failure of beams

that took place in the warehouse at Wilkins Air Force Depot in Shelby, Ohio,

researchers were of the view that shear was simple problem to deal with. Then

they realized that shear in concrete beams cannot be designed as traditionally

as it was done earlier. There has been a feeling among researchers to go back

and rethink about the fundamentals of shear design.

Going back, the work done by Talbot (1909) during the year 1909

was considered to give a clear and significant way to analyze the shear for

designing concrete structures. Talbot’s findings affirm that the shear stress is

a function of longitudinal reinforcement, length of the beam and the stiffness

of the beam.

Further, in the case of beams without web reinforcement, the

strength of the concrete is also to be considered for designing shear. The shear

failure at Ohio made many researchers to think about the seriousness of shear

behaviour in concrete beams.

It is only since the last four decades, researchers have been

focusing their work to evolve a common and a rationalised consensus on

design for shear which could be globally acceptable. As a result, many

theories have been developed to explain the shear behaviour in beams and

also to estimate its shear capacity.

13

2.2.2 Classification of RC Beams

The shear behaviour and capacity of concrete beams depends on

various factors and among them the length /span of the beam plays a crucial

role. The behaviour of beams varies depending on the span to over all depth

ratio. The beams are broadly classified as deep, short and slender depending

upon their behaviour and failure mode.

One of the important governing factor affecting the shear behaviour

is the shear span ‘a’ to effective depth ‘d’ ratio was strongly stated in ASCE-

AC1 Task Committee 426, (1973). According to MacGregor (1988) beams

can be classified as very short, short, slender and very slender according to

the a/d ratio. Very short spans (a/d <1.0) induce inclined cracks joining the

load and support. Due to internal redistribution of forces, the arch action

occurs in beams having short shear spans (1.0< a/d < 2.5) and due to which

the beam can take up additional load. This type of beam either fails by ‘shear

tension’ or by crushing of concrete near the loading points.

Shear tension failure occurs as a consequence of loss of bond

strength due to a horizontal crack at the level of the flexural reinforcement.

The other type of failure commonly known as ‘shear compression failure’

occurs by crushing of the concrete near the load point. Due to the presence of

inclined cracks their equilibrium gets disturbed which is common in beams

having slender spans (2.5<a/d < 6.0). Failure is by means of flexure in case of

very slender beams (a/d > 6.0) which takes place prior to the formation of any

inclined cracks due its large shear span.

14

2.2.3 Factors Influencing the Shear Behaviour and Capacity of RC

Beams

2.2.3.1 Factors influencing the shear behaviour

Shear failure in concrete beams are brittle in nature and are

catastrophic, which is quite contradictory to flexural failure where ductility is

dominant. In the concrete beams since the shear failure precedes the flexural

failure, the shear strength is designed to be greater than the flexural strength at

al1 points along the beam. Shear behaviour of a beam without shear

reinforcement is mainly determined by four factors: the ratio of shear span to

effective depth, the longitudinal reinforcement ratio, the tensile strength of the

concrete and the existence of axial forces Mac Gregor (1988). Shear forces in

a beam occur wherever the applied moment changes along its length.

The main assumption in the ACI 318 code specification is that the

shear capacity is proportional to the depth of the member. To discover the fact

about this assumption, many experimental investigations were conducted by

Shioya et al (1989) in which they tested reinforced concrete members that

ranged in depth from 100 to 3000 mm. All members were simply supported

without any shear reinforcement and were reinforced in flexure. The results

prove that the shear stress at failure decreases as the depth of the member

increases.

The effect of size has a significant role in shear carrying capacity in

RC beams was found by Kani (1967) who worked on “size effect” in

concrete beams. He demonstrated that the shear stress at failure decreases as

the depth of the member increases. Assuming the contribution of concrete

strength in design is a common practice, where the shear resistance is

assumed to be proportionate to the square root of the maximum cylinder

15

compressive strength. But, the latest research has proved that high cylinder

strengths does not result in high shear strength Angelakos et al (2001).

In concrete beams the Shear transfer mechanism may occur by any

of the following means:

(1) Shear stress in the uncracked concrete,

(2) Interface shear transfer,

(3) Dowel action and arch action.

Failure due to cracking of beams occurs when the stresses in the

beams exceed the tensile shear stress and the failure due to crushing occurs

when the stresses exceed the compressive shear stress. As a result of

compressive force being applied in the concrete strut portion between the

loading point and support, the web portion of a beam carries biaxial stress.

The combined effect of both tensile and compressive loads reduces the stress

at failure ASCE-AC1 Task Committee 426 (1973).

Shear transfer across a crack in the concrete structure may be

possible by the aggregate interlocking at the interface of the crack. Some

earlier studies made by Nilson and Winter (1991) has proved that the

interlock forces developed at the interface can resist about one-third of the

total shear force. Once the failure takes place along this interface, then it leads

to slipping along that plane. In the presence of the longitudinal bars, the

cracks crossing them will be resisted by the dowel action of the bars against

the shear propagating. The dowel action widens the crack and also induces

tension in concrete surrounding the reinforcement. This leads to splitting

cracks around and along the longitudinal reinforcement. Compared to the total

shear resistance the contribution made by dowel action is not a dominant

ASCE-AC1 Task Committee 426 (1973).

16

The presence of the longitudinal reinforcement also influences the

crack by reducing its size and modifies resistance performed by the aggregate

interlock. The length and depth of the crack reduces considerably in the

presence of larger amount of longitudinal reinforcement. The reduction in the

size of cracks helps to prevent the beam from further crack propagation. Thus

by providing greater amount of longitudinal reinforcement, the shear capacity

can be increased to an extent.

When the shear resistance is more as in some cases, the flow of

shear gets affected and a resistance in the form of arch action is developed.

For this arch action to take place, the depth of the member should be

sufficiently large enough which should be comparable to its span. If not, the

structure will not be possible to develop the arch action. This type of arch

action is common in deep beams, where the inclined cracks are formed from

the point of loading to the support reaction.

2.2.3.2 Factors influencing the shear capacity

In case of concrete beams without shear reinforcement, the load at

cracking decides the capacity of the member. However, in the case of beams

with shear reinforcement, even after cracking there seems to be some

resistance to shear due to the presence of tensile stresses in concrete. This fact

about the increased capacity was found by Collins et al (1996). The design

capacity of these beams depends on the load at cracking.

Concrete beams without stirrups, having a longitudinal

reinforcement ratio between 0.75 to 2.5 percentage, fail only due to shear. In

this range, beams with Lower reinforcement ratios tend to fail at lower shear

stresses. Beams that fail in shear have greater amount of reinforcement than

what is minimum required for a flexure. Beams with very low ratios of

longitudinal reinforcement generally fail in flexure earlier before the shear

17

capacity is reached. The axial compression produced due to applied load,

serves to increase the shear capacity of a beam. Opposing this, the axial

tension greatly decreases the shear capacity this has been proved by

MacGregor and Wight (2005).

2.2.4 Shear Resisting Mechanism in RC Beams

Model studies in concrete beams with and without shear

reinforcement has been going on since 1973. Most of the model studies

related to the shear mechanism of concrete beams was first reviewed by

ASCE-ACI Committee in the year 1998, which was published in the ASCE-

ACI Committee report 445. An extensive review of most of the important

shear models of RC beams, evolved between 1973 and 1998, was

consolidated in this report. Some of the prominent shear models mentioned in

the ASCE-AC1 committee report (1998) are the Compression Field Theory

(CFT) proposed by Collins (1978), Modified Compression Field Theory

(MCFT) proposed by Vecchio and Collins (1986), the variable angle truss

model brought out by the Eurocode EC2 (1992), truss models using crack

friction principle developed by Dei Poli et al (1990), fixed and rotating angle

softened-truss models (Hsu 1993).

The foremost attempt made by a Swiss engineer Ritter (1899) and a

German engineer Morsch (1902) to explain the shear behaviour in reinforced

concrete beams was based on a 450 truss model in which the cracks were

assumed to be formed at 450 in the web region within the lever arm distance

(jd ). The concrete present in the region between adjacent inclined cracks is

assumed to take up the compressive stress due to the applied shear force.

However later studies done by Nielsen (1984) has proved that the 450

cracks

developed in beams are not always common and this model cannot be fully

accepted.

18

The “Compression Field Theory” (CFT) developed by Collins

(1972) was based on above truss model incorporating a compatibility equation

to resolve the angle of inclination of the compressive strut. Later, based on his

studies Collins was able to put forward a model that explains the shear

behaviour in concrete beams through which he postulated that the directions

of principal strain coincided with that of the corresponding principal stress

directions.

Vecchio and Collins (1981) and (1982) conducted some tests and

proposed a model through which they were able to explain that there was

some decline in the concrete compressive capacity due to the principal tensile

strain in concrete in the cracked region and hence the tensile stresses

contributed by concrete between cracks has to be taken into account to

estimate the shear capacity of concrete beams. Based on this thought, Vecchio

and Collins proposed the “Modified Compression Field Theory” (MCFT) in

the year 1986 and 1988.

Similar to the “Modified Compression Field Theory” Hsu

developed some models which were developed based on his experimental

results. Based on the assumption that the rotating angle of the concrete struts

varies with the shear load applied, the “rotating angle softened truss model”

was developed by Hsu (1992). However, this theory was not useful in

predicting the shear contributed by concrete. Hence Pang and Hsu (1996)

developed another model called the “Fixed Angle Softened Truss Model”

through which the concrete contribution term VC was predicted.

The shear friction models for concrete beams were developed by

Loov (1998) which was based on the test conducted by Clark (1951) and Kani

(1979). Loov predicted that the major shear crack occurs when there is a

possibility of slip occurring in the members. Based on this he formulated an

19

expression to derive the shear strength of the beams. However this expression

was found to be not very useful only for exceptional situations.

There are several internal mechanisms within the concrete to

resistance shear failure. And it is difficult to predict its failure precisely. The

various factors involved are uncracked concrete in compression region,

interlock of aggregates, dowel action and the tensile stresses acting

perpendicular to the cracks. The uncracked compression zone is a portion of

concrete that can fully resist shear forces. Also, Collins et al (1996) confirmed

that cracked concrete has lot of tensile stresses that it can significantly

increase the capacity of concrete in resisting shear forces. Aggregate interlock

refers to the internal friction formed at a crack due to surface roughness and

can contribute to almost one third of the total shear force. Nilson et al (2004)

demonstrated the effect of dowel action which occurs as a result of the

vertical forces acting across the longitudinal steel reinforcement.

MacGregor and Wight (2005) developed a simple truss model

through which they illustrated that a beam which has inclined cracks, formed

due to the applied load, develops compressive and tensile forces in the top and

bottom flanges together with vertical forces in the stirrups and compressive

forces in the diagonals. For members which have very small amount of shear

reinforcement, the resistance offered to shear by this model predicts more

conservative results.

Later Nilson et al (2004) modified this original truss model into the

“variable angle truss model” in which it was assumed that the concrete strut

angle was not always inclined at 45°. Instead they vary within the range of

25° to 65°. This new proposal modelled by Nilson and named as “variable

angle truss model”. The variable angle model comprises of compression fans

and compression fields. The compression fans which takes place near the

20

supports or under the direct loads has numerous diagonal struts spread out

from this region. It is assumed that the total vertical load is fully resisted by

these radiating struts. The compression field consists of diagonal compression

struts that are formed parallel between the compression fans. All the stirrups

are assumed to have yielded at this point as assumed in the original truss

model.

Shioya (1989) studied the influence of beams depth and aggregate

size on the shear strength of concrete beams by conducting experiments on

beams of varying depth from 100 mm to 3000 mm. The investigation revealed

that as the beams depth increased the shear stress decreased. This decrease in

shear stress may be attributed to the reason that due to larger area of frictional

resistance across the failure crack due to greater depth, the shearing force

could have distributed to a relatively larger area. The study also revealed that

the size of the aggregate was inversely proportional to the shear stress at

failure.

2.2.5 Effect of Shear Reinforcements

The flexural failure in RC beams does not take place suddenly but

instead it shows some warning of distress going to take place in the near

future. Contradictory to this, the shear failure is sudden, catastrophic and

devastating. To avoid any such sudden shear failure, shear reinforcements are

provided. Also, the shear reinforcement has a control over the shear strength

of the beam. The shear stirrups and are used to increase the shear strength of

concrete beams, to avoid the shear failure and to cause a flexural failure.

Shear reinforcements, which are normally provided as vertical

stirrups, are placed at varying intervals depending upon the shear conditions

acting on a beam. Different configurations of stirrups are being used, such as

21

an open or closed stirrup, or stirrups with multiple legs which depends upon

the amount of applied shear. Shear reinforcement are also provided as

inclined bars in some cases. Shear reinforcement comes into effectiveness

only after the formation of diagonal cracks either crossing them or in its

vicinity.

At the instance of diagonal cracks formation, the stirrups come to

effect and the stirrups offer more resistance to the shear when the cracks cross

them. This controls the growth of the cracks and reduces the penetration of

the crack further. The stirrups oppose widening of the cracks, which

maintains the aggregate interlock within the concrete was explained by Nilson

et al (2004) in case of FRP RC beams. Moreover, stirrups are tied to the

longitudinal reinforcement and due to this confinement effect the splitting of

concrete along the longitudinal main bars is controlled more effectively. The

extent and amount of shear resisted by the shear stirrups and concrete depends

upon the design procedure which is being adopted

The role of shear reinforcement in concrete beams which are placed

in the form of stirrups contributes to the strength of the shear mechanisms and

enhances the shear capacity. A stirrup effectively confines the longitudinal

reinforcement and resists the crack. This in turn increases the contribution to

shear by effective dowel action. In the presence of stirrups the cracks are

minimized and due to which the shear transfer through aggregate interlocking

can be considerably reduced.

By providing stirrups with sufficient spacing, the concrete can be

confined with its region and this can enhance the compressive strength and

thereby increasing the shear capacity. The shear reinforcement used in

concrete beams conserves the concrete contribution to shear and allows the

development of additional shear force.

22

2.2.6 Summary

Considerable progress has been made in the past century in the

design of RC members subjected to shear. Some of the important

contributions to study the shear behaviour such as Kani’s model, modified

compression field theory, variable truss angle model, rotating angle softened

truss model have been discussed.

In spite of an enormous number of works done on shear behaviour

of beams, there is still no unified solution to predict the shear strength of a

beam irrespective of whether it is slender, short or deep beam. Still, research

work on shear behaviour of slender beams is currently carried out to find a

unified expression for shear strength that could be accepted and adapted

commonly.

2.3 SHEAR BEHAVIOUR OF STEEL REINFORCED

CONCRETE DEEP BEAMS – BACKGROUND REVIEW

2.3.1 General

In this chapter, focus is mainly towards the study of reinforced

concrete deep beams whose a/d ratio is less than 1.0 with or without web

reinforcement. Before proceeding to study the deep beams, the background of

shear behaviour is also reviewed initially.

The role of shear behaviour in steel reinforced concrete deep beam

members is discussed in this section. The recently introduced theoretical

concepts explaining the shear behaviour has also been considered for this

review.

23

2.3.2 Deep Beams and Its Concept

Concrete deep beams carry heavy load within a very short

supported span. In other words, a reinforced concrete deep beam can be

expressed as a beam having a depth comparable to the span length. They have

wide applications and are used in foundations works, tall buildings, offshore

structures, etc.

The conceptual changeover from ordinary-beam behaviour to deep-

beam behaviour is stated to be imprecise and has been well explained in a

book written by Kong (1990). He has mentioned that the transition from an

ordinary beam behaviour to a deep-beam behaviour is imprecise and is

difficult to exactly predict the point of change in the behaviour.

Reinforced concrete deep beams differ from other beams primarily

in their behaviour to take up the load. Due to the geometry of deep beams, the

failure in deep beams is totally governed by shear rather than flexural failure.

Before a deep beam could take up its full flexural strength, diagonal cracks

are formed which tend to cause shear failure. Hence, shear strength is

considered as an important factor in the design of concrete deep beams.

The very basic ideology of classifying a concrete deep beam has

not become universally common. The design of reinforced concrete deep

beam for shear which is adopted by various design codes differs mainly in

classifying RC deep beams. Different countries follow dissimilar ideology to

define a deep beam in their relevant code books. However, in this research the

classification of deep beams is based on ACI-ASCE Committee 445(1998)

which states that a beam with shear span to-depth ratio (a/d) less than 1.0 as a

“deep beam” and a beam with a/d exceeding 2.5 as an ordinary shallow beam.

Any beam between these limits (1< a/d < 2.5) is classified as a short beam.

24

2.3.3 Factors Influencing Shear Behaviour in RC Deep Beams

Based on the review made on earlier research work done on deep

beams, one can infer that the vital parameters that control the shear strength in

deep beams are:

1. Effective depth (d)

2. Width of the beam (b)

3. Effective span (le )

4. Shear span (a)

5. Cylindrical compressive strength of concrete ( fc’ )

6. Yield strength of horizontal web reinforcement

7. Yield strength of vertical web reinforcement

8. Reinforcement ratio of main tension bars ( )

The structural behaviour of deep beams has been proved to be

different when compared with slender or short beams. One of the important

parameters controlling this change is its ‘shear span to effective depth’ (a/d)

ratio which depends on the depth of the beam. Since this ratio is small in deep

beams, there is a significant change in the strain distribution across the deep

beam’s depth. This variation of strain is non-linear and is not seen in ordinary

slender beams.

Shear deformation which is insignificant in ordinary beams is

considered to be substantial in deep beams and hence it cannot be ignored as

this factor is also associated with the depth and effective span of the beam. It

has been proved by many researchers that the width of the deep beam

increases its stiffness and shear strength and reduces the lateral buckling.

25

Considerable amount of work has been done under the title “size

effect” of reinforced concrete deep beams. Studies on the effect of web

reinforcement strength on the shear carrying capacity of beams have been

carried out by many researchers. Web reinforcement of different type of

materials, shapes and orientation has been experimentally tried in deep beams

in many earlier works.

2.3.4 Earlier Studies on Shear Behaviour of RC Deep Beams

De Paiva and Siess (1965) conducted experiments on small beams

having an effective depth range of 150 mm to 300 mm and with a small

“shear span-to-depth” ratio in the range of 0.7 < a d < 1.3. This study was

done by varying the transverse reinforcement ratio from 0 to 1.4%.

Based on this experiment the following points were concluded:

1) By adding the vertical and/or inclined stirrups the inclined

cracks formation cannot be altered. Such an addition also has

very little effect on the ultimate strength.

2) By providing more vertical stirrups deflections at the ultimate

load was found to be reduced.

3) A higher loading capacity was observed beyond inclined

cracking for beams with small shear span-to-depth ratios

without shear reinforcement.

4) They concluded by saying that for beams with small ‘shear

span-to-depth’ ratio, transverse reinforcement has no influence

on beam’s strength. This was formed due to the ideology that a

single direct strut formed between the loading point and the

support reaction in which the transverse reinforcement was

26

assumed to have no role in altering the beam’s strength.

Nevertheless, they have stated that an increase in the transverse

reinforcement can reduce crack widths.

Leonhardt and Walther (1966) conducted tests on nine simply

supported deep beams and two continuous deep beams. The beams tested

were with a ‘span to depth’ ratio ranging between of 0.9 to 1.0. The tests

conducted by these researchers were on one of the largest deep beams ever

tested having a depth of 1.6 meters. Their tests proved that deep beams which

experience a uniform tension force in the bottom main reinforcement develop

a “tie-arch” action. Based on their test results, they suggested that it is not

possible to increase the shear capacity of deep beams by providing additional

web reinforcement. This was however disproved by later researchers.

They also observed that for a ‘overall length to overall depth’ (L/D)

ratio equal to1.0, the horizontal web reinforcement distributed over

approximately 1/5 to 1/10 of beams overall depth (D) was found to be more

effective in countering the shear cracks. They also indicated that in the case of

beams loaded at the bottom, the vertical or inclined shear stirrups are

considered to be significantly important in taking up the shear load.

The researchers have also expressed the importance of detailing the

anchorage zone of the main longitudinal reinforcement as playing an

extremely important role in the design of deep beams. They also observed that

at ultimate loads, the deflection measured along the bottom chord of the

beams was small and thus compared to ordinary beams, deep beams had more

stiffness. European concrete code committee (1970) and the CIRIA Guide 2

(1977) formulated the basic design of deep beams based on their work.

Ramakrishnan and Ananthanarayana (1968) conducted experiments

on tested reinforced concrete deep beams having a ‘span to depth’ ratio

27

ranging from 0.9 to 1.8. Testing was done on twenty six deep beam specimens

out of which twenty two beams were without any web reinforcement. The

amount of steel web reinforcement provided in the remaining beams was very

less. They observed that most of the beams failed due to shear.

Based on their observations an expression for the shear strength of

reinforced concrete deep beams was proposed which is based on the splitting

strength of the concrete.

They concluded that shear failure in deep beams was almost the

same as that of shallow beams with in a ‘shear span to depth’ ratio of less than

2. The main cause of shear failure in deep beams is due to splitting of

concrete as observed in a cylindrical split tensile test.

Kong et al (1970) conducted experiments on simply supported deep

beams to study the effect of web reinforcement. The research objectives were

the span-depth ratio and seven types of web reinforcement on deflections,

crack widths, crack patterns, failure modes and ultimate loads in shear. A total

of 35 reinforced concrete deep beams were tested. Keeping the span constant

and by varying only the depth of the beam, all beams were tested within a

range of 1 to 3 of ‘over all span to depth’ (L/D) ratio. The beams were tested

under two point loading until failure.

From their test results it can be concluded that deflection in deep

beams can be substantially reduced by a reasonable amount of horizontal web

reinforcement placed close to the bottom of the beam. Also it is distinct from

their study that with closely spaced horizontal web reinforcement, the

deflection of the beam can be minimized. They concluded by stating that the

effectiveness of horizontal web reinforcement decreased with increase in

‘overall span to overall depth’ (L/D) ratio and ‘clear shear span to overall

depth’ (x/D) ratio.

28

From this study they concluded that,

1) The closely spaced horizontal web reinforcement near the

bottom of the beam was very effective in controlling the crack

width.

2) The shear reinforcement to control cracks and deflections was

very much dependant on the clear shear span-to-depth ratio.

When the clear shear span was larger than the effective depth

(x/D > 0.35) vertical stirrups were more effective in controlling

the crack widths. When the clear shear span was greater than

the effective depth (x/D > 0.7) vertical stirrups were more

effective than the horizontal bars or orthogonal reinforcement

placed in two directions.

3) The primary cause of failure observed was diagonal cracking

and crushing of the compression strut between the bearing

support and the applied load.

Smith and Vantsiotis (1982) carried out a wide range of

experiments on concrete deep beams to investigate the influence of web

reinforcement and ‘shear span to effective depth’ ratio in contributing to the

shear strength of deep beams. Testing was conducted on fifty-two simply

supported deep beam specimens under two point top loads. All the beam

specimens which were tested had a rectangular cross section. Five of the

fifty-two beams were provided without web reinforcement.

All beams which were tested were grouped into four series based

on their ‘shear span to effective depth’ (a/d) ratio. The (a/d) ratios so chosen

were 0.77, 1.01, 1.34 and 2.01. Both vertical and horizontal web

reinforcements were provided as shear reinforcement, the spacing of which

29

was also considered as a variable apart from the ‘shear span to effective

depth’ ratio.

The conclusions arrived at were:

1) All beams generally failed in shear.

2) No significant change in the failure mode observed between

different series.

3) The use of minimum amount of vertical and horizontal web

reinforcement reduced the crack width and deflection.

4) In general, the web reinforcement increased the ultimate shear

strength for all beams that were tested.

5) The effect of vertical web reinforcement was greater above a/d

< 1.0.

6) The horizontal web reinforcement had more influence in

beams with a/d <1.0

7) Concrete strength has a greater influence over the ultimate load

capacity, especially for beams with a/d <1.0.

Lehwalter (1988) conducted experiments on sixty simply supported

deep beams specimens under three point bending to investigate the capacity

of the compression struts, which were designed to fail in shear. In the first

phase of the experiment the characteristics of the aggregate, the shear span-

depth ratio a/h (0.5 < a/h < 1.5), and the beams over all depth were made as

variable parameters. The effect of web reinforcement on the bearing capacity

was investigated in the second phase of the experiment.

The influence of web reinforcement placed under different

conditions was investigated. Beams with horizontal web reinforcement

30

distributed over the web were initially investigated, and then beams with

horizontal web reinforcement concentrated at the top of the beam were

examined. Finally beams with vertical web reinforcement distributed over the

web portion of the beam were examined.

To investigate the influence of the shear span-depth ratio some of

the beams were tested with varying ‘a/h’ ratio ranging between 0.5 and 1.5. In

addition to the above, test were conducted to study the effect of variation of

depth on the ultimate strength of the beam by varying the height of the beam

between 200 to1000 mm.

The observation was that the inclined cracks started forming when

the applied load reached close to 45-50% of the ultimate load. They found

that the shear span depth ratio ‘a/h’ had a large influence on the ultimate

strength. The ultimate strength rapidly decreased with increasing in ‘a/h; ratio.

The type of aggregate or the maximum particle diameter of aggregate in

concrete was found to have no influence according to their observations.

The experiments results from their second phase of experiment

showed that the shear strength was influenced by web reinforcement. A slight

increase in strength was found with increasing horizontal web reinforcement.

They observe that when the a/h ratio was reduced, the effectiveness of vertical

web reinforcement decreases. While considering the depth as a variable

parameter, they conclude that in beams height do not influence the shear

strength of beams with varying web reinforcement.

Tan et al (1995) studied the effect of variation of a/d ratio on the

shear strength of deep beams. Nineteen simply supported deep beams were

tested in this work. The shear capacity of beams were examined under eleven

different ‘shear span to effective depth’ ratios ranging from 0.27 to 5.38 for

varying concrete strength ranging from 50 to 68 MPa. All specimens had

31

uniform cross-section and were provided with same longitudinal and shear

reinforcement. The spans were varied to obtain the desired span-depth ratio.

Beams at the lower range of the a/d ratio were observed to fail due to pure

shear condition. It was observed by researchers that when the a/d ratio was

increased the failure mode changed from shear to ‘flexure-shear’ mode of

failure.

2.3.5 Effect of Web Reinforcement in Deep Beams

Kong et al (1994) conducted experiments to mainly study the

influence of high strength concrete (HSC) in the shear behaviour of deep

beams. The variable parameters considered for this study were the concrete

strength, the ‘shear span-to-depth’ (a/d) ratio and the slenderness ratio of

beam specimens. A total of 30 beams tested which were designed to fail in

shear. A variety of beams were tested which can be grouped as simply

supported, continuous, slender or stocky beams. The test results were

compared to predictions given in many design codes to study the shear load at

failure. The ‘shear span-to-depth’ ratios adopted ranged between 0.22 and

1.50. The strength of the concrete cube ranged from 43 to 96 MPa.

They observed that the first crack developed was a flexural crack,

and later the diagonal inclined cracks developed and finally the beam failed

by crushing which was similar to the behaviour of deep beams. The types of

failures witnessed were diagonal splitting, diagonal compression strut

crushing and at times, bearing failure. They also noticed that there was a

sudden increase in the beams’ deflection subsequent to the formation of initial

inclined crack particularly in beams with web reinforcement.

They concluded by expressing that there was no substantial change

in the shear behaviour of the deep beam specimens cast with normal strength

compared to those deep beams with high strength except that the failure in

32

deep beams with high strength concrete was further brittle in nature. They

also concluded by expressing that the web reinforcements were most efficient

in deep beams when placed perpendicular to the diagonal crack immaterial

whether they are simply supported or continuous, or stocky or slender.

In addition, they stated that both simply supported and continuous

specimens demonstrated similar shear behaviour. Also, they expressed that

the horizontal web reinforcement was more effective in countering shear for

simple and continuous deep beams.

Some of the other conclusions arrived at were:

1) For beams with greater depth, the horizontal web reinforcement

was found to be very effective in preventing the steeper cracks.

2) The vertical web reinforcement are more effective when the

beam develops a diagonal cracks which are relatively flat

closer to the horizontal axis.

3) This tied-arch action compensates for the reduced aggregate

interlock.

4) Finally, comparing their test results with various code

provisions they concluded that the CIRIA (1977) method was

the most accurate in predicting the ultimate shear capacity

among the methods considered. For deep beams with higher

concrete strengths, the ACI Code predictions were conservative,

but in case of the Canadian Code the predictions were found to

give unconservative results.

Braestrup (1990) studied the shear strength of deep beams by using

the theory of plasticity. In his study the lower and upper bound approaches are

33

analysed. He concluded that the theory of plasticity for structural concrete

gives insight into the behaviour of deep beams at failure, in addition to

providing reasonable predictions of the ultimate loads. He conclude by

expressing that for steeply inclined shear failure plane in beams with low a/d

ratio, the horizontal web reinforcement is more effective and this was just the

reverse in beams with high a/d ratio.

Ashour et al (2002) proposed an empirical modeling obtained by

using the genetic programming to predict the shear strength of reinforced

concrete deep beam. The various parameters influencing the shear strength of

RC deep beam was analyzed using genetic programming.

The following points were concluded in their model study:

1) The shear strength was found to increase as the shear-span to

depth ratio was decreased, which indicates that the shear

strength is inversely proportional to the ‘shear span to depth’

ratio.

2) The ‘shear span to depth’ ratio and amount of the main

longitudinal reinforcement has a significant influence on the

shear strength of RC deep beams.

3) The main longitudinal bottom reinforcement which has an

influence over the shear strength was found to be only till a

certain limit, beyond which there seem to be no effect on the

shear strength.

Zararis (2003) proposed a theory on shear compression failure in

deep beams which is based on the analysis in cracked reinforced concrete

members along with the internal forces that are at the diagonal shear cracks.

34

Based on this study a formal theory to describe the shear strength in deep

beams was proposed. The theory was used to predict the depth of the

compression zone above the critical diagonal crack and also the ultimate shear

capacity of deep beams with or without web reinforcement.

The study concluded that the contribution of the horizontal web

reinforcement to the shear strength of a deep beam is insignificant. The

proposed theory a predicts with accuracy the experimental observations for

the ultimate shear of deep beams with various strengths of concrete, main

steel ratios, shear reinforcement ratios, and shear span to depth (a/d) ratios

between 1.0 and 2.5.

2.4 LITERATURE REVIEW OF SHEAR BEHAVIOUR OF FRP

REINFORCED SLENDER CONCRETE BEAMS

2.4.1 FRP as Reinforcement in RC Beams

FRP materials, in spite of having the advantage to take up more

tensile load than the conventional steel, are brittle in nature. This brittleness of

FRP is due to its low percentage of strain coupled with increased load bearing

capacity. The FRP materials show a linear stress-strain relationship up to their

failure without yielding.

On the other hand, concrete is also a brittle material. A combination

of FRP and concrete in RC structures both of which are brittle leads the

structure to fail without sufficient warning. Hence more care has to be taken

in designing FRP reinforced RC structures.

At the very beginning, most of the research work related to FRP

reinforced concrete structures was carried out in Japan and many research

articles were published.Machida et al (1995) and Sonobe et al (1997)

suggested the use of higher material safety factor for FRP reinforcements

35

compared to steel, to avoid brittle failure possibly arising out of rupture of

FRP reinforcements. Later their findings were incorporated in the Japanese

design guidelines JSCE (1997)

2.4.2 The Shear Transfer Mechanism

2.4.2.1 Aggregate interlocking

Tottori and Wakui (1993) experimentally investigated the shear

capacity of concrete beams using FRP as flexural and shear reinforcement. In

their experiments CFRP composite cables were used as longitudinal

reinforcement. Shear reinforcement in the form of spirals made of GFRP,

AFRP, CFRP and Vinylon FRP bars were used in their test specimens.

To evaluate the dowel capacity of CFRP flexural reinforcement,

specimens were specially designed for tested. With different types of shear

reinforcement and with different combinations, many beams were tested for

shear. Strain gauges were installed on the FRP spirals to measure the strains

and calculate the shear force contributed by the shear reinforcement.

They were able to conclude that the shear force carried by concrete

in the compression zone and aggregate interlock were related to the tensile

stiffness of the longitudinal reinforcement. Further, the contribution of

concrete to the shear resisting force was observed to be equal to the shear

cracking load of the beams which is calculated based on the measured shear

force contributed by the FRP spirals.

Raffaello Fico et al (2007) in their work have discussed about the

importance of the aggregate’s mechanical interlock which takes part during

shear transfer across a crack in the tensile zone. This aggregate interlock

accounts for about 33% and 50% of the amount of shear capacity of

uncracked concrete as estimated by Taylor et al (1970). The amount of shear

36

further gets reduced when the crack width increases due to increased loading

as stated by Walraven et al (1981). They have concluded that the aggregate

interlock is a function of the crack roughness and the crack width which

depends on the maximum aggregate size and on the reinforcement stiffness

respectively. Also, in their study they have mentioned another significant

function influencing the aggregate interlock due to the concrete strength.

Further in their work they have stated that the total stiffness of FRP

RC members reduces due to lower strain values and higher reinforcement

ratio of FRP bars as compared to steel RC members. This lower stiffness

makes the members to deflect larger and make wider cracks in the concrete.

Ultimately, this leads to the carrying of smaller amount of shear force by

aggregate interlock in FRP-reinforced members.

Razaqpur et al (2004) in their study have tested seven RC beams in

bending to determine the concrete contribution to their shear resistance. All

the RC beams had only flexural reinforcements and there was no shear

reinforcement used in their experiments. The variables considered in their

tests were the shear span to depth ratio varying from 1.82 to 4.5 and the

flexural reinforcement ratio, varying from 1.1 to 3.88 times the balanced

strain ratio. The test results were compared to the values obtained by using the

various design codes. Finally, they concluded by indicating that their

experimental results much closer to the Canadian standards values compared

to that of the values obtained by JSCE code.

They have also discussed about various factors influencing the

shear in FRP RC members as mentioned in the ACI-ASCE Committee 445

(1998) which is used for conventional steel reinforced concrete members.

According to ACI-ASCE Committee 445, in a steel reinforced concrete

member subsequent to the formation of diagonal tension cracks, a member

resists the shear by means of a number of mechanisms:

37

(1) The shear resistance of uncracked concrete compression zone

(2) Aggregate interlock

(3) Arching action

(4) The dowel action of the longitudinal reinforcement

(5) Residual tensile stresses across cracks

(6) The shear carried by the shear reinforcement.

Further, they have stated that the shear contribution of the

uncracked concrete depends on the depth of the uncracked zone and the

concrete strength. They have further mentioned the importance of the

roughness of the crack’s inner surface, which depends on the maximum

aggregate size and on the crack opening size both of which have an influence

over the aggregate interlock resistance.

2.4.2.2 Dowel action in FRP RC beams

Grace et al (1998) made an extensive study on simple and

continuous FRP reinforced concrete beams. Seven simply supported beams

and seven continuous T-section beams with different combinations of

reinforcements and stirrups were tested. The materials used as reinforcements

were steel, CFRP and GFRP. Under this study with different combinations of

main reinforcement and stirrups, the load shared by the individual reinforcing

materials was evaluated.

They identified the importance of the dowel effect in preventing the

beam from a sudden collapse due to shear failure. They also indicated that the

dowel effect was very low in case of FRP bars when compared to steel bars.

They further observed that among the FRP bars, GFRP bars had the lowest

dowel effect and CFRP bars were slightly better in taking up the dowel effect.

38

They concluded the work by indicating that the use of GFRP stirrups

increased the shear deformation and deflection and reduced the ductility of

the beam. Also, they pointed out that the type of shear failure mode occurred

when GFRP stirrups were used with FRP main reinforcement in case of

ordinary simply supported beams. The dowel effect was very much critical in

case of continuous beams with FRP reinforcement.

Tottori and Wakui (1993) carried out some experimental tests on

FRP reinforced members and concluded that the dowel capacity of FRP

reinforcement members was less than that of steel reinforced members by

about 70%. They also inferred that when FRP was used as flexural

reinforcement, the dowel contribution was very small and could be neglected.

The low dowel strength of FRP bars was due to FRP’s low transverse

stiffness and strength.

Tim Stratford and Chris Burgoyne (2003) have done extensive

analytical model studies on shear analysis of concrete reinforced with FRP

reinforcement to understand how the compatibility, equilibrium and the

material constitutive laws can be combined to establish the actual conditions

within a FRP-reinforced beam subjected to shear. This study investigates the

compatibility of FRP reinforced concrete members being modelled using with

the lower bound theory. The importance of this study is to emphasis the fact

that the lower bound theory depends on stress redistribution which is not

possible in case of FRP reinforced concrete members.

With respect to the dowel action they have stated that the load

carried by the dowel action of the reinforcement across a crack is negligible in

steel reinforced beams which have also been proved by Kotsovos and

Pavlovic (1999). But with FRP reinforcement which has low transverse

stiffness, a much smaller load will be carried by the dowel action as stated by

Kanematsu et al (1993) in their work. Consequently, this dowel action causes

39

longitudinal cracking of the concrete along the flexural reinforcement as

identified by Kotsovos and Pavlovic (1999) and also Sakai et al (1999).

In this study, models were developed to explain the dowel splitting

which causes a sudden increase in the un-bonded length of reinforcement

leading to uneven crack propagation into the compression-zone, causing the

beam to fail. The effect of dowel rupture due to the dowel action is considered

as an important mode of failure in beams has been discussed in their study

which depicts the rupture of reinforcement under combined shear and tensile

actions. Their prediction on FRP being used as beams main tensile

reinforcement which has low transverse strength, illustrates the vulnerability

of FRP to dowel rupture which does not occur with steel reinforcement as

mentioned in Naaman and Park (1997), Bank and Ozel (1999).

Maruyama et al (1989) made some research studies to understand

the effect of inclined shear cracks in FRP reinforced concrete beams by

conducting experiments with FRP stirrups being placed inclined to the

artificially made vertical crack in concrete specimen. The experiments were

performed using concrete block specimens in which the FRP stirrups were

embedded and tensile forces being applied across an artificial vertical crack

which was intentionally formed at various angles to cut across the FRP

stirrups.

CFRP, AFRP and GFRP bars made of Epoxy resin matrix were

used in these experiments having diameters between 5 to 6 mm. An important

finding made in this work was that when FRP bars were subjected to tension

by placing them at an angle to the applied load the tensile capacity of FRP

bars reduced drastically. In case of GFRP bars, when the fibers were inclined

at 30o to the horizontal, the diagonal tensile strength was observed to be only

65% of the strength of the bar when the fibers were oriented horizontally.

40

From the study it was concluded that the maximum tensile capacity

of FRP bars can be achieved only when the fibers in the FRP bars are placed

perpendicular to the predicted crack in concrete. In other words, the dowel

forces are maximum when the fibers in the bars are perpendicular to the

cracks.

2.4.3 Shear Cracking in FRP RC Beams

Tim Stratford et al (2003) in their model study have indicated that

the importance of shear reinforcement in resisting the shear to ensure the

failure of a beam is only by flexure. It has been stated by them that by

providing shear reinforcement, the shear carrying mechanisms in a beam are

affected in many ways as mentioned below:

1) Shear reinforcement carries tensile actions across cracks,

which resist the further widening of cracks.

2) Shear reinforcement confines the compression-zone concrete

and thus increases its shear-capacity.

3) Dowel-splitting of the concrete is prevented by providing shear

reinforcements which are placed enclosing the flexural

reinforcements. However, the dowel-rupture of FRP

reinforcement cannot be avoided.

4) The equilibrium of a cracked section in a concrete member

with stirrups requires a shorter crack length for a given applied

load.

5) The shape of the crack will also differ when the stirrups are not

provided in the concrete beams.

41

6) Concrete softening mechanisms are less effective across a

wider crack if the surfaces of the crack are completely

separated as aggregate interlock cannot occur and this reduces

the shear capacity of the beam as mentioned by Kotsovos and

Pavlovic (1999).

7) Due to the yielding of stirrups in steel reinforced concrete

members, stress-redistribution occurs and the application of

lower-bound plasticity theory to analyze the various models is

possible. However, this is not applicable in the case of FRP

reinforced concrete members due to the lack of stress-

redistribution.

Raffaello Fico et al (2007) investigated the shear strength

contribution by uncracked concrete before the assessment they made on of

Eurocode-like design equations for the evaluation of the shear strength of

FRP RC members. The main objective of their study was to make an

assessment of Eurocode-like design equations for the evaluation of the shear

strength of FRP RC members, as proposed by the guidelines of the Italian

Research Council CNR-DT 203 [CNR-DT 203/2006].

They reviewed of current design provisions for shear strength

contributed by concrete in FRP RC members. ACI 440.1R-06, CAN/CSA-

S806_02, JSCE and the CNR-DT 203 design guidelines were considered for

this assessment work.

The following conclusions were arrived at:

1. The equation proposed by the CNR DT 203 which takes into

account the concrete contribution to the shear strength can be

conservatively used by practitioners.

42

2. The strength of stirrups bent portion is not seem to be a

significant factor affecting the FRP stirrups contribution to

shear in FRP shear reinforced members.

3. The stirrups contribution to the shear strength obtained by

using the proposed equation by the CNR DT 203 was found to

give relatively good results compared to other design

guidelines.

4. Increasing the amount of stirrup reinforcement ratio beyond

1% seems to have no effect on the shear capacity in FRP

reinforced concrete beams.

5. The outcomes of this research plan will be used to optimize the

proposed equation which takes into the influence of FRP bar

properties on different mechanisms contributing to the

concrete strength.

2.4.4 Size Effect in FRP RC Beams

Niwa et al (1997) investigated the size effect on the shear strength

of FRP reinforced concrete beams analytically by using a FEM technique and

based on the predicted crack models. The researchers evaluated the size effect

on the shear behaviour of concrete beams without shear reinforcement. To

study the shear behaviour, modelled bar elements where placed and sliding

adjacent to one another were assumed to study the shear crack.

Perpendicular to the crack, the nonlinear elements representing the

concrete's fracture properties has been investigated. They concluded this study

by expressing that the shear strength of concrete beams reinforced with FRP

decreases significantly with the decrease in the elastic modulus. Also, they

43

observed that there was no change in the ‘size effect’ between a FRP

reinforced and a steel reinforced concrete beam.

Matta et al (2007) studied the ‘size effect’ on shear strength of

beams reinforced with FRP bars. The results of the experimental tests on four

large-scale concrete beams reinforced with GFRP bars in flexure and shear

were observed and studied. the results indicate that a decrease in concrete

shear strength has a influences in the size effect of the GFRP reinforced

beams.

The following conclusions can be drawn:

1) The concrete shear strength appears to be strongly affected by

size effect.

2) Negligible difference on concrete shear strength has been noted

in sections with an increase in the amount of shear

reinforcement.

3) The proposed design equation for concrete shear strength takes

into account the factor that counters the size effect.

2.4.5 Developments in FRP Stirrup Fabrication

Cheolwoo Park and Jongsung Sim (2006) fabricated GFRP stirrups

and used them as shear reinforcement in concrete beams. They developed a

novel way to fabricate the GFRP stirrups. Twelve beams were tested for shear

strength in this work out, of which nine beams were provided with the newly

developed stirrups fabricated by the authors. The beams were divided into

three groups and were tested under three different shear spans. The efficiency

of the GFRP stirrups was evaluated. The results show that the newly

44

developed stirrups have performed well and are equally comparable with steel

stirrups in taking up the shear load.

2.5 SHEAR BEHAVIOUR OF FRP REINFORCED SHORT AND

DEEP CONCRETE BEAMS – LITERATURE REVIEW

2.5.1 Behaviour of FRP Reinforced RC Short Beams

Nehdi et al (2008) made the first attempt to study CFRP reinforced

RC short beams having a ‘shear span to effective depth’ ratio less than 2.5.

Twelve beams were cast without web reinforcement and tested for under four

point bending condition. Out of the twelve beams, four beams were reinforced

only with steel as reinforcement. The “shear span to depth” ratio was varied

between 1.36 to 2.33 to investigate the crack pattern, propagation of crack,

mode of failure and shear strength.

The researchers concluded by saying that the CFRP bars has a

greater influence on the shear strength and deflection of short beams Further

they observed that the increase in deflection is due to the low modulus of

elasticity of CFRP bars. The shear strength capacity of the CFRP short beams

was found to be greater than that of a similar steel reinforced short beam

indicating an increase in arch action.

Strut and Tie Models (STM) were created to predict the shear

strength with reasonable accuracy. A modification in the efficiency factor

used in the STM model was suggested at the end of this work.