CHAPTER 2

LITERATURE REVIEW
2.1 Preamble
The in-plane bending stiffness (EI) of a beam is the product of two variables:
(1) the in-plane second moment of area (the in-plane moment of inertia I), reflecting
the cross-sectional resistance to loading; and (2) the modulus of elasticity (E),
reflecting the material resistance to loading. In concrete beams, both variables are
subject to change during the course of loading. The variation in the modulus of
elasticity with the increasing load is caused by the inelastic stress-strain behavior of
concrete beyond the elastic limits, while the variation in the moment of inertia is
associated with the cracking of concrete due to the tensile strains greater than the
cracking strain of concrete. The cracked zones in a concrete beam are ineffective in
resisting stresses originating from applied loads and moments. Therefore, cracking of
concrete decreases the resistance of a concrete beam to loading, leading to greater
deformations in the beam. Hence, the study of reinforced concrete beams in pre-
cracked and post-cracked state becomes important.
A substantial research work has been done and is going on the flexural stiffness
of beams and columns. The research work done by various researchers in discussed
here in brief.

2.2 Experimental work on Beams

Hsiung and Frantz [4] examined shear tests of five approximately one-third
scale models of large, reinforced concrete beams with shear reinforcements. The tests
were carried out to determine how varying web widths and different transverse stirrup
distributions affect beam shear strength. In the present study, the authors discussed
load-deflection response, ultimate shear capacities, stirrup stresses, surface inclined
crack widths and internal crack widths. The authors observed that shear capacity of a
large reinforced concrete beam is not affected by the transverse spacing of stirrups
across the web width. The effect of confinement was considered in this study.
Similarly, Espion and Halleux [7] carried out a series of tests on rectangular reinforced
concrete beams submitted to bending and constant compressive normal force. The
moment curvature relationships were evaluated and compared with the prediction of
two theoretical models. The first model was CEB model proposed by Favre and

5
Koprna. It was a simplified model which refers to the un-cracked and fully cracked
stiffnesses in pure bending only. The second model was a proposition made by the
authors which takes into account the tensioning effects, the variation in the position of
neutral axis as a function of eccentricity of normal force and the non-linear behaviour
of concrete in compression.
Soroushian et al. [11] studied the effects of the design variables on
axial/Flexural behavior of reinforced concrete sections. These factors were concrete
compressive strength, steel yield strength, longitudinal steel ratio and lateral
confinement of concrete. The following conclusions were made:
 Variations in longitudinal steel yield strength have major effects on the
element flexural strength and some important effects on the element axial
strength for the typical beam and column cross sections considered in this
study. The axial and flexural stiffnesses are not much influenced by the
variations in yield strength.
 Variations in concrete compressive strength have relatively small effects on
the beam flexural behavior but major effects on the column compressive and
strength.
 The flexural behavior of reinforced concrete beam is strongly influenced by
the variations in the tension steel ratio but not much by the changes in the
compression steel ratio. The axial strength and stiffness of RC columns are
found to be varying to the changes in their total steel area. Increased
confinement does not have any major effects on the beam flexural behavior. It
slightly increases the compressive strength and ductility of the column.
Abdulrahman et al. [14] evaluated experimentally the effective moment of
inertia of rectangular reinforced concrete beams from the immediate deflection. The
beams were of different reinforcement ratios and loaded under a mid-span
concentrated load. The test results obtained were compared with the ACI Building
Code equation of effective moment of inertia. The authors observed a noticeable
difference especially for heavily reinforced beams. The authors discussed the effect of
using the moment of inertia of the un-cracked transformed section in the computation
of effective moment of inertia. The effect of reinforcement ratio was also considered
in a recently developed model. It was used to calculate the effective moment of inertia
of reinforced concrete beams under any type of symmetrical loading by considering

6
the variation in the beam’s cracked length. Using similar variable i.e. transverse
reinforcement, Plizzari et al. [20] experimentally investigated a beam under constant
bending moment. The objective of the study was to examine flexural and splitting
crack openings along the splice length, to determine the splice length and study the
effective stiffness of the beam length containing the splice especially during the
cracked stage. The authors underlined the importance of transverse reinforcement to
limit the splitting crack opening even outside the splice. The authors observed that
specimen failed because of the collapse of the overlapping splice. This failure took
place due to concrete splitting. It was also seen that at the ultimate moment a sudden
increase of the splitting crack opening occurred over the splice.
Khuntia and Ghosh [33] continued the work on flexural stiffness and verified
EI values of columns and beams experimentally. In the experimental verification, the
axial load – bending moment diagrams were obtained using the proposed stiffness
assumptions. The diagrams were of slender columns for a given initial eccentricity
(M/P) ratio. The results were compared with numerous published test data and found
to be in good agreement. According to the authors, the proposed stiffness expressions
were applicable for all levels of applied loading, both service and ultimate loads.
Beeby et al. [39] reviewed the current approaches to tension stiffening BS 8110
[74] and Euro Code 2 [73]. The authors summarized a major experimental
investigation which was undertaken to explore time dependent aspects. Revisions to
BS 8110 and Euro Code 2 were proposed as a result of this work. Further, the authors
described the refinements for modeling tension stiffening effects in reinforced
concrete members.
Kaewunruen and Remennikov [47] clarified the post-failure mechanism and
residual capacity of railway concrete sleeper. Australian manufacturer’s concrete
sleeper was used for the negative bending test in accordance with AS1085. LVDT was
used in deflection measurement at the mid-span while inclinometers were placed in
line of the rail support. The post-failure load-deflection curves were first presented in
the study. The remaining part of the concrete sleeper was used to core for some
samples. The concrete material of the sleeper was of 88.5 MPa strength and
prestressing wires were of a proof stress of 1860 MPa. It was found that each
prestressing wire restrains approximately 10 kN residual load-carrying capacity of the
concrete sleeper. The residual load-carrying capacity results in the safety allowance to
public domain, in order to escape from the accident.
7
 Earlier researchers focused on the effect of reinforcement ratio. Akmaluddin
and Thomas [48] carried out an experimental study on 36 reinforced concrete beams
of 150x240x3000 mm with a clear span of 2750 mm. Along with reinforcement ratio,
one more variable i.e. concrete grade was considered in this study. The beams were
subjected to two symmetrical point loads with various distances between the beam
support and the load position. The results showed that one of the models could be used
for calculating singly reinforced concrete beams deflection with reinforcement ratios
between 1% and 3.25%.
Carmona et al [51] presented the results of experimental research program
intended to investigate mixed-mode fracture reinforced concrete. The tests were so
designed that only one single mixed-mode crack generates and propagates through the
specimen. The specimens were of three different sizes and subjected to 3-point
loading. The specimens were provided with asymmetric notches and reinforced with
different ratios of longitudinal and inclined reinforcement. According to authors, these
experiments were expected to help understand the mechanisms of crack initiation and
propagation through reinforcing bars under mixed-mode load conditions.
Stramandinoli and Rovere [57] proposed a tension-stiffening model for
reinforced concrete elements. A parameter based on reinforcement ratio and steel-to-
concrete modular ratio was studied. The model was implemented into a computational
program that allows for nonlinear finite element analysis of reinforced concrete beams.
The reinforced concrete beams were tested subjected to 4-point loading. The
experimental results of these tests were in good agreement with the results of the
model proposed in this study.
After a long period, the equation for effective moment of inertia proposed by
Branson and Bischoff for deflection values was examined by Kalkan [59]. Deflection
values were compared with measured deflection values of reinforced concrete beams.
The beams were with reinforcement ratios ranging from 0.024 to 0.034. The author
found that both methods give deflection values in close agreement with the actual
values and the method proposed by Bischoff provides a slightly better correlation with
the actual in-plane bending deformations. These deformations were of heavily
reinforced concrete beams. It was also observed that the shrinkage cracking was
causing lesser deflection of a concrete beam compared to the response by effective
moment of inertia expression. The cracking moment calculated from experiments was
compared to the methods given in ACI 318-05 [72], Euro Code 2 and TS 500 [80]. It
8
was observed that the cracking moment values based on modulus of rupture expression
in Euro Code 2 were in close agreement with the experimental value.
Recently, Amir et al. [69] introduced an empirical procedure for determination
of crack opening in a reinforced concrete prism. The authors investigated effective
parameters and the results were compared with the available experimental data. The
parameters considered were bar size and the effective concrete area surrounding the
reinforcement.

2.3 Analytical work on Beams

Various researchers have focussed on different aspects of variables such as
reinforcement ratio, effect of confinement, concrete grade, and crack pattern. The
analytical work carried out on reinforced concrete beams is discussed in brief.
Gupta and Akbar [3] proposed a simple model of forming cracks in reinforced
concrete. The authors found out that the crack direction may change in the course of
loading. It was assumed in this study that the cracks are formed in the direction of
major principal tensile strain and also that the crack directions may change with the
change in strains. The proposed model resulted in crack directions which were
consistent with the limit state. Further, Chan et al. [12] proposed an analytical model
for crack analysis of reinforced concrete tension members in which the tension-
stiffening effect has been taken into account. The ultimate bond stress and bond stress
transfer length were introduced in the bond stress distribution function. The
predictions of crack loading and elongation of a reinforced concrete member from the
proposed model were satisfactory when compared with some available experimental
data and analytical results. Polak and Blackwell [27] presented a constitutive model
for tension in cracked reinforced concrete members. Special importance was given to
the bond between concrete and reinforcing steel because this is the main factor
influencing tension stiffening in cracked reinforced concrete. The model was used to
analyze 14 specimens which were subjected to bending and in-plane loading. The
authors presented the comparison between experimental and analytical results.
Creazza and Marco [16] presented a mathematical model which was used to determine
the bending moment-mean curvature relationship for beams subjected to combine
bending and axial forces taking tension stiffening effects into account. It was observed
that the structural behaviour was discontinuous due to crack opening. The analytical
formulation and solving techniques were presented after this observation. To show the
9
application of the procedure and useful comparison with the results in literature, some
numerical examples were solved. Amara [22] developed a new method based on
Griffith energy balance postulate for brittle fracture. It was developed to find a beam
model for crack growth in plain and reinforced concrete sections. The model was used
on a local scale to simulate the mechanisms of crack formation during crack-growth
process. The focus was on aspects of crack-growth stabilization in reinforced sections.
It was observed that comparison of this model with fracture mechanics based model
resulted as a more accurate tool for fracture analysis of reinforced concrete sections.
Pulmano and Shin [6] presented a simpler finite element method for predicting
the instantaneous and long-term deflections of statically determinate or indeterminate
ordinary reinforced and partially / fully prestressed beams. This method presented had
a parabolically varying bending stiffness along its length. It also considered the non-
linear effects due to cracking and tension stiffening and the time dependent effects due
to shrinkage and creep. In the solution procedure, the predicted values by proposed
method were compared with measured values to assess its performance. It was
concluded that it could be implemented on a microcomputer or even on a
programmable calculator. The predicted instantaneous and long-term deflections were
in good agreement with tests of simply supported, continuous ordinary RC beams as
well as partially prestressed concrete beams.
Raed [8] proposed a rational method of analysis for the calculation of creep
deflections of reinforced concrete beams under sustained service loads. After a period
os sustained load, the initial stress in concrete was related to the final strain by the
introduction of fictitious modulus. The flexural rigidity of the member was determined
from new neutral axis depth and new modular ratio. It was used further to calculate
final deflections. The results obtained were compared with several experimental
measurements of beam deflections. It was observed that the results were in good
correlation with each other. Further, the author [24] proposed a method to describe the
stress, strain and deflection behaviour of reinforced concrete beams. The role of
compression reinforcement was also taken into account through an enhanced rational
approach. Creep deflections were computed by applying a direct procedure which was
similar to that then used by ACI318-95 to calculate immediate deflections. It was
observed that the beams with high compression steel ratios were particularly shown to
have calculated results in good correlation with experimental results. It was also

10
observed that the deflections of the examined cases were lesser by 9% less to 12%
more than the measured values.
Ghali [15] critically reviewed the equations of ACI 318-89 and Canadian
standard CAN3-A23.3 for the prediction of the immediate and long-term deflection in
one-way non-prestressed construction. A more accurate approach was proposed which
was comparatively easy to apply in practice.
The equation for effective moment inertia was

 M cr 
3
  M 3 
Ie    I g  1     I cr  I g ……………………………..…… (2.1)
cr

 a
M   a  
M

It was replaced by the following equation for mean curvature for use in
deflection calculation of a cracked section.
 m  (1   ) 1   2
where,   1   (M cr / M )2 forM  M cr
  0.4
M= Bending moment at the section considered
β= β1 β2, β1=1 for high bond bars
β2= 1 & 0.5 respectively for first loading and for loads applied in a sustained
manner or for a large number of cycles.
The equation for the factor λ was omitted. Instead, three equations were
proposed applicable to un-cracked and cracked sections by using the appropriate
cross-sectional area properties.
 (t0 )  M / Ec (t0 ) I
Khuntia and Ghosh [32] suggested simplified and reasonably accurate
expressions for the computation of effective moments of inertia of beams and columns.
The factors considered were the reinforcement ratio ρc, the concrete compressive
strength fc’, the magnitude of axial load P and the eccentricity ratio e/h. this paper was
divided into three parts:
1. An expression for moments of inertia of RC columns derived using a
parametric study-
 e P 
EI e  Ec I g (0.80  25 g )  1   0.5 u   Ec I g  Ec I beam ………...…. (2.2)
 h P0 

11
 2. An expression for moments of inertia of RC beams derived using a
parametric study-
 b
EI e  Ec I g (0.10  25 g )  1.2  0.2   0.6 Ec I g ……………..….……. (2.3)
 d

where 1.2  0.2   1.0

b
 d
For high strength concrete, equation (20) becomes
 b
EI e  Ec I g (0.10  25 g )  1.2  0.2   1.5  4 105 f c'   0.6 Ec I g …. (2.4)
 d

where 1.2  0.2   1.0

b
 d
3. A brief review of current ACI code provisions w.r.t. EI and EI for slender
columns in particular was provided. The authors also explained the need for
modifications to the current code provisions. Liew and Choo [38] presented the
procedures for determination of bending moment based on best-fit curve method on
measured deflection profile of a R.C. flexural element. The overall procedure consists
of three major steps – (a) determination of curvature k, (b) determination for effective
moment of inertia Ie and (c) determination for bending moment M. The equation
1 v ''
proposed for the curvature k was k   . The authors have focused on

1  v  
3
, 2 2

d2y dy
effect of v’’ and v’ i.e. 2
and on deflection k. A third to fifth degree polynomial
dx dx
may tend to flatten out some of the local peaks in the curve and hence may
underestimate the corresponding curvature. Therefore, a minimum sixth degree
polynomial function was suggested in order to capture the localized larger curvature.
Moment of inertia was determined using an expression proposed by Branson. It was
used in the analysis of a cracked section. The authors also derived an equation for
bending moment. This value was substituted in the Branson’s equation for Ie. These
iterations are to be repeated for other points along the measured deflection profile until
the whole bending moment profile is determined.
Ugur and Guney [26] developed a computer program to determine the moment-
curvature relationship of reinforced concrete sections. The program was developed in
a spreadsheet environment. The material model used for steel reinforcement was strain

12
hardening of steel. The program was used to study the effect of parameters like
confinement and strain hardening of steel, level of axial load, amount of tension and
compression reinforcement. An important conclusion of this study was that
analytically generated moment-curvature curves were not correct because it is not
possible to define exact stress-strain relationship of concrete since it is influenced by
various parameters. Olivia and Mandal [42] examined the influence of three variables
on curvature ductility of reinforced concrete beams. A computer program was
developed to predict moment-curvature and available curvature ductility of reinforced
concrete beams with or without axial loads. Ten beams with different variables were
analyzed using this program. The variables measured were concrete strength, amount
of longitudinal reinforcement and spacing of transverse method. The input given was
beam geometry, material properties and loading. A confined stress-strain curve for
concrete proposed by Saatcioglu and Razvi was applied in the program. On the other
hand, steel stress-strain model was adopted from BS 8110 (British Standard Institution
1985). It was observed from the study that the curvature ductility increases with the
increase in longitudinal strength and concrete strength. It was also concluded that the
spacing of transverse reinforcement does not have any significant influence on the
curvature ductility. Srikanth et al. [52] presented a procedure for finding the analytical
Moment Curvature behaviour of statically determinate reinforced concrete beams. The
confinement offered by shear reinforcement to concrete in compression zone was also
considered. The experimental program consisted of casting six beams of three
different concrete strengths. For each concrete strength, one under-reinforced and one
over-reinforced beam were cast. The authors noticed that strain in steel was the
governing criteria in under-reinforced beam while it was the concrete strain in over-
reinforced beams. The experimental and analytical values obtained were used for the
numerical comparison. The ratio of analytical/experimental values was calculated at
all the significant points. These values were compared with six confinement models
reported in the literature in the last decade. These models were used as a stress block
for confined concrete for generating the complete analytical Moment Curvature
behaviour. The prediction of the curvature corresponding to the 85% of ultimate
moment descending portion made by Mendis model was found to be better when
compared to the other models.
Many of the authors presented the work based on flexural failure. However,
Scollard and Barlette [35] investigated the reliabilities of 33 reinforced concrete
13
cantilever beams using Monte Carlo simulation. For this, the authors used ASCE 7-98
load factors and strength reduction factors for flexure from ACI 318-99 and ACI 318-
02. In this study, three limit states were considered corresponding to flexural failure
neglecting strain hardening of the reinforcement, flexural failure neglecting strain
hardening and flexural failure as limited by bond. The authors concluded that bond
failures were more numerous than flexural failures for all cases investigated in the
simulation. This is not desirable because bond failures are more brittle in nature.
In the analysis of beams, the parameters like non-linearity in case of material
and geometry are very important. Siddique and Rouf [44] developed a non-linear
numerical model considering material and geometrical properties. Material non-
linearity was simulated by considering parabolic stress-strain relationship of concrete
and bi-linear relationship of reinforcing steel. The modified Newton-Raphson
technique was used for the solution of non-linear equations. The load-deflection
behaviour of over-reinforced high strength concrete beams was carried out with the
model. It was observed that for the high strength concrete, the increase in steel content
increases the strength and stiffness but decreases the ductility. Barbosa and Ribeiro
[60] considered the practical application of nonlinear models in the analysis of
reinforced concrete structures. It was observed from the load deflection curves that the
model for concrete in compression plays a major role in the achievement of numerical
solution. The authors also suggested that satisfactory prediction of response of
reinforced concrete structures could be obtained owing to good analytical results.
The reinforcement bars used in any R.C.C. work are high yield strength
deformed bars. The effect of the strength of these bars is an important parameter in the
design of beams. Mast et al. [55] presented a methodology for the flexural strength
design of concrete beams reinforced with high-strength reinforcing steel that confirms
to the requirements of ASTM A1035-07. The design method was based on simple
analysis techniques that satisfy fundamental principles of equilibrium and
compatibility. Strain limits for tension-controlled sections and compression-controlled
sections were proposed and these were according to current and past ACI 318 codes.
It was expected that the proposed method will enable the designers to take full
advantage of high-strength steel bars as reinforcement for concrete structures.
The structural actions of beams and slabs are similar. Both these structural
members are subjected to flexural action. Hence, the study of effective stiffness of
RCC slabs is also important.
14
 Maria [23] developed an effective stiffness model for reinforced concrete
slabs. The objective of this paper was to present a finite-element program for practical
calculations of deflections of RC slabs. When the concrete cracks, the stiffness of the
concrete changes. In case of bending, it changes through the thickness of the member
which is accompanied by the reinforcement. This behavior was approximated by the
concept of effective moment of inertia suggested by Branson. In the finite element
formulation, the quadratic Serendipity-plate-bending element was an eight nodded
element with 24 DOF. Each node having 3 DOF: one transverse translation (w) and
two rotations (өx and өy). The finite element formulation was followed by material
model and analysis procedure. In the experimental program, one of the specimens
subjected to uniaxial constant bending was used for the analysis. The response of this
specimen to the reinforcement ratios and boundary conditions, effect of modulus
rupture, effect of reinforcement orientation was studied. This response was compared
with the previous literature. It was concluded that analyses done by the proposed
model show adequate accuracy when compared to the results from tests on slabs with
various reinforcement, boundary and loading conditions. The practical advantage of
this type of model was the reduced computer time required to execute the program and
the reduced time spent on creating input for the analysis. Katalin and Gyorgy [25]
carried out a research to find the correct moment diagrams of the column and moddle
strips of a flat slab. The authors also worked out a method for the computation of
deflections of the slab. The method was compared with the provisions in ACI and
Eurocode-2. The authors arrived at an important conclusion that the ACI method for
determining the deflection of a cracked member was not correct. The authors proposed
a new relationship for deflection of cracked member considering cracked moment of
inertia and gross moment of inertia.

2.4 Experimental work on Columns

Macgregor et al. [1] presented and documented a proposal for revising the
slender column design procedure of ACI 318-63. The authors proposed the use of a
second order structural analysis wherever possible and practical. In place of such an
analysis, an approximate design method was proposed based on a moment magnifier
principle. Through a series of comparisons with analytical and test results, the
accuracy of the design procedure was established. It was shown that the proposed
procedure was more rational, more accurate and more consistent than the procedure
15
used that time. The proposed method called the attention of the designer to the basic
phenomenon in slender columns. It was also possible to evaluate the additional
moment requirements in restraining members. Hence, according to the authors, this
method resulted in a superior and safer design.
Donald [2] presented a new approximate method for estimating the strength of
reinforced concrete columns in rigid framed structures. The method used a
deteriorating stiffness stability concept combined with a modified use of the
conventional short-column moment-thrust interaction diagram. When these results
were compared with exact computer based solutions, it was observed that the results
obtained by the proposed method predicted the failure loads accurately. However, the
method was limited to frames with symmetrical single-curvature, no-sway only.
According to the author, when the method is properly modified, it can provide a good
platform for treating more general problems also.
Mirza [10] carried out experiments and proposed two sets of equations, set 1
for initial sizing and preliminary design of structures and set 2 for use in final or more
accurate structural designs.
Set 1:
For columns in bottom stories supporting more than three floors plus the roof
EI  [0.27 Ec I g  Es I se ] /1.7 …………………………………………..…………. (2.5)

For columns in intermediate stories supporting one to three floors plus roof
EI  (0.21Ec I g  Es I se ) /1.6 ……..………………………………..……...…..…. (2.6)

For columns in top stories supporting just the roof

EI  (0.1Ec I g  Es I se ) /1.5 .……………………………………………………... (2.7)

Set 2:
 Ec I g  Es I se
EI  ……………..…………………………………….…….. (2.8)
(1   d )
  (0.27  0.003l / h  0.3e / h)  0 ………………….………………..…...… (2.9)
  (0.3  0.3e / h)  0 …………………………………….……….……..... (2.10)
Based on these proposed equations, the author concluded that the results
presented indicate that prediction variations of the ACI EI expressions were about
twice as high as those of the proposed equations. Also, the ACI 318-89 EI equation
(2.8) was in most cases less conservative than the ACI 318-89 equation. (2.28). This
was the contrary to the statement in the ACI Building Code.
16
 Resheidat et al. [17] carried out the research work for circular columns also on
same line of action as that for the rectangular columns. The proposed equations were
as follows:
EI   Ec I g  Es I se ………………………..…………….………….…..…… (2.11)

 2   2 
   max     ;    max ………………………..……….……….. (2.12)
  max   max  

max  0.6  0.004 f y / fc,   . ………………………………..……..………. (2.13)

 max  0.65  0.02 f y / fc '   . ………………………………………………. (2.14)

Resheidat et al. [18] proposed a new equation for the EI and the variables were
material properties, steel ratio and axial load. The proposed effective flexural rigidity
for columns under short-term loads was expressed as follows-
EI   Ec I g  Es I se ……………………………….……...………………… (2.15)

where α = stiffness reduction factor, Ec= modulus of elasticity of concrete, Ig =

moment of inertia of gross concrete section about centroidal axis, Es = modulus of
elasticity of steel and Ise = moment of inertia of steel about centroidal axis of cross-
section.
The value of α was expressed by the following equation-
 2 N  N 2 
   max     ; N  N max ……….…………………..…………. (2.16)
 N max  N max  

where αmax= maximum value of α, N=P/P0, P= the axial load, P0= the axial load under
pure compression, Nmax= value of N corresponding to αmax.
For N>Nmax; N shall be replaced by (2 Nmax -N-). The factor α is a function of the
material properties, the steel ratio and the axial load. The authors obtained 768 α-
curves. Using multiple regressions for the data, the following expression was obtained:
Nmax  0.58  0.003 f y / fc,  N . ……………………………………….….. (2.17)

 max  0.64  0.02 f y / fc '  N . ……………...…………………………….... (2.18)

Kim and Park [28] investigated experimentally the strength and deformability
pf specimens with interlocking spirals subjected to concentric axial loads. The
columns were 36 in number with single spirals and 108 in number with interlocking
spirals. The variables in this study considered were strength, spacing of spirals, yield
strength of spirals, and interlocking length of spirals. Based on test results, the authors

17
proposed model equations for prediction of strength and axial strain at peak stress of
specimens with interlocking spirals.
Chung et al. [31] carried out a study to present a stress-strain relation of
confined concrete from an empirical study of 65 columns. The experimental
parameters included were strength of concrete, volumetric ratio, strength and
confinement type of rectilinear ties, and distribution of longitudinal bars. The authors
proposed an equation to determine the tie stress caused by lateral concrete pressure.
The experimental data was analyzed by nonlinear multiple regression method. This
study provided the empirical equations to determine the peak stress and its
corresponding strain of confined concrete. This was expressed by tie stress, the
effectively confined distance ratio, and the strength of concrete and the configuration
of ties. Studying all the experimental models, an empirical model for stress-strain
curve of laterally confined concrete was developed by three co-ordinates. Comparison
with the other stress-strain curves illustrated the validity of the proposed relation.
Bonet et al. [36] proposed a simplified method to design slender rectangular
reinforced concrete column. The method was based on moment-magnifier method
(ACI 318-02) and EuroCode-2 (draft for stage 49) and it was suitable only for columns
with equal effective lengths in two principal bending planes. The proposed method
was compared with 371 experimental tests from the literature and a good degree of
accuracy was obtained. The proposed method was supposed to be suitable for practice
engineers to check as well as design slender reinforced concrete columns.
Tan and Nguyen [40] conducted an experimental investigation of 30 plain and
reinforced concrete column specimens with 200 mm square sections. The specimens
were tested under different loading conditions i.e. concentric compression, eccentric
compression with a fixed neutral axis position and eccentric compression with a
constant eccentricity. The main variables were concrete strength ranging from 46 to
101 MPa and confinement reinforcement with a volumetric ratio ranging from 0.7 to
2.4%. A new equivalent rectangular stress block needing only one parameter was
proposed. A large volume of published test data with a wide range of concrete
strengths was used to examine this rectangular stress block and that of ACI code. It
was concluded that the new equivalent stress block was suitable for both normal and
high strength concrete.
Woods et al. [50] conducted a study on eight high strength concrete columns
of 69 MPa with dimensions 203x203x2030 mm. The objective of the study was to
18
differentiate the effects of spacing and volumetric content of the transverse
reinforcement on the bending ductility of high strength concrete columns. For this,
four columns were reinforced with ties with different spacing but same volumetric
content i.e. 1.1%. The other four columns were reinforced with ties of constant spacing
of 76 mm and different diameter ranging from 3.2, 4.8, 6.4 and 8.0 mm. It was
observed that the effects of parameters considered in the study yielded the results
within a range of values outside of which the effects were minimal.

2.5 Analytical work on Columns

Ehsani and Alameddine [5] presented a simple and accurate method for
calculating the stiffness of a circular column. The authors developed a computer
program to study the effect of variables for a circular reinforced concrete section. The
variables included in the study were material properties for reinforcing steel and
concrete, diameter of the column, number and area of each longitudinal bar, the cover
distance measured from the outside of the cross-section to the centroid of the
longitudinal steel. A parametric study was carried out to develop the EI expressions.
A quadratic expression was developed for high axial load region [(P/P0)>0.5] and low
axial load [(P/P0) <0.5]
I / I t  Cc Cd C y [a0  a1 (1  P / P0 )  a2 (1  P / P0 ) 2 ]  1…………..…… (2.19)

where a0  3.4  1.25, a1  3  3.09, a2  6.7   3.21

Cc, Cd & Cy are termed as coefficient of variation for compressive strength of
concrete, diameter of column and reinforcing steel yield stress.
 
Cd  1  0.05[1  e 0.134( d 12) ]

C y  [0.0075  0.015(1  P / Po )] f y  0.9(1  P / P0 )  0.55

I / I t  Cc Cd C y [b0  b1 (1  P / P0 )  b2 (1  P / P0 ) 2 ] ….…………..…… (2.20)

b0  1  0.01(67 / J  25 / J 2  1 / 06)
J  1.0
3
b1  3.84
(0.67 J  0.34)
3
J  1.0
b2  7.42
( J  0.14)
3
J  1 .6
b3  3.46
(J )

19
 It was observed that the proposed equations very closely fit the computer
program with a maximum error of less than 5%. These expressions vary with the axial
load levels. Hence, minimum stiffness expression was proposed as follows:
I / I t  0.118  17.8  135 2 ………………………………………… (2.21)

ρ is the reinforcement ratio.

It was concluded that equations (2.20) and (2.21) give an excellent
approximation of the variation of moment of inertia at different axial loads. It was also
concluded that the proposed method would result in considerable savings of material.
Sheikh et al. [9] suggested a simple relationship where the concrete strength
was reduced from fc’ at the balanced load to 0.85 fc’ for concentric compressive load.
A second-degree parabolic stress-strain curve with strain at peak stress equal to 0.002
was used. It was used for reasonably accurate and conservative prediction of section
capacity. The proposed relation for strength of concrete as a function of axial load on
the section was as follows:
For P<Pb, fcp= fc’……………………………………………………….. (2.22)
From the analytical investigations and experimental results, the authors
concluded that the strength of concrete depends on strain gradient and the state of
stress in concrete at ultimate. Below the balanced axial load level, the concrete strength
does not have very significant influence on the flexural capacity of the section.
Sun and Lu [13] presented a general computer model to investigate the ultimate
strength behaviour of regularly and irregularly shaped reinforced concrete columns.
The columns were subjected to axial load and biaxial bending. The computer model
was based on Chinese Code GBJ 10-89 and ACI 318-83. The model presented was
also applicable to reinforced concrete columns loaded in one direction. The position
of neutral axis and the area of reinforcement were obtained using classical Newton-
Raphson method. The accuracy of the method was verified with experimental results
for reinforced concrete columns subjected to axial load and biaxial bending.
Kim and Yang [19] proposed a numerical method for predicting the behaviour
of a reinforced concrete column using the layered finite element method. Experimental
results revealed that the ultimate load for a short high-strength concrete column was
significantly increased but not for a slender column. The possibility of stability failure
for a slender column was increased with the increase of concrete strength. The
predicted behaviours of the concrete columns by numerical method proposed by the

20
authors showed good agreement with the test results. The authors also showed that the
ACI’s moment magnifier method may be un-conservative for very slender high-
strength concrete column.
Mirza [21] presented an overview of a research and it summarized the
assumptions and analyses used related to probability based design for RC columns.
The statistical study was conducted to calculate the cross section and stability
resistance factors (ϕc, ϕs) for reinforced concrete tied columns subjected to single
curvature bending and short time loads. The important conclusions of the study were
as follows:
1. When the ACI load factors are used, the cross section resistance factor ϕc
should be taken as 0.7 for compression failures, increasing linearly from
0.7 to 0.9 as the axial load decreases from balanced load to zero.
2. When the ASCE/ANSI load factors are used, the ϕc value should be 0.65
for compression failures, increasing linearly from 0.65 to 0.85 as the axial
load decreases from the balanced condition to the pure bending condition.
A constant value of stability resistance factor ϕs = 0.75 should be used for
computing the moment magnification of slender columns. This value of ϕs is
compatible with both the ACI and the ASCE/ANSI load factors if the appropriate ϕc
factors have been included in the cross section strength.
Kim and Lee [29] proposed a numerical method for predicting the behaviour
of reinforced concrete columns subjected to axial force and biaxial bending
considering curvature localization. A series of 16 tied reinforced concrete columns
was carried out. The columns were of square size 100x100 mm and rectangular size
200x100 mm. the boundary conditions at both ends of the column were hinged and 40
mm eccentricities equal in same directions. The angles between the direction of
eccentricity and the major principal axis of gross section were 0, 30, 45 for square
section and 0, 30, 45, 60, 90 for the rectangular section. The comparison between
numerical prediction and the test results were in good agreement with each other in
ultimate loads, axial force-lateral deflection relations and lateral deflection
trajectories. The authors also concluded that ACI’s moment magnifier method was
conservative in both uniaxial and biaxial eccentric loading conditions.
Hong [30] proposed a simple approach for estimating the strength of a slender
reinforced concrete column under biaxial bending and axial load. The model used in
the work considered the non-linearity of stress-strain relationship of concrete and steel.
21
It could be used for columns with arbitrary cross-sections. The author explained the
use of the method for design purposes. It was observed that the predicted results
obtained from the proposed method compare well with experimental results found in
literature.
The current ACI equations were critically examined by Tikka and Mirza [34].
These equations were compared with theoretical EI equation. By this study, the
authors developed and proposed the equation for short term EI in three steps:
1. A format of proposed EI equation was selected that included variables
affecting EI significantly.
2. A multiple linear regression analysis of the generated theoretical stiffness
(EI) data was conducted to evaluate coefficients related to some of the
variables included in the proposed EI equation.
3. The proposed EI equation was then finalized by curve-fitting to one-
percentile values of the generated theoretical stiffness data.
The proposed EI equation was
   
 e 1  l
EI   0.47  3.5    0.003  Ec I g  0.8Es I rs …...…………. (2.23)
 h  1  e  h

  h 
where β=0.7 for rs  2% and β= 8 for rs  2% .
This equation is subjected to the following limitations:
e / h  0.1, l / h  30, rs  1% & fc '  55MPa.Ife / h  0.1, usee / f  0.1.

The authors presented a statistical evaluation of the parameters like end

eccentricity ratio, axial load ratio, slenderness ratio and longitudinal reinforcement
ratio. These parameters affect the flexural stiffness EI of slender tied reinforced
concrete columns. The columns were subjected to short-term loads in symmetrical
single curvature bending. A stiffness reduction factor ϕk= 0.8 was proposed for the
new equation. The authors also concluded that variations of the proposed design
equation are less than half as compared to the current ACI EI expression. Also, there
is no significant variation in the proposed EI equation due to effects of variables
investigated.
Bonet et al. [37] proposed an analytical approach for calculating failure
surfaces in rectangular reinforced concrete cross sections. The sections were provided
with symmetrical reinforcement and subjected to axial load and biaxial bending. This

22
method was valid for concrete with strength between 25 and 80 MPa. The proposed
expression was assessed with results obtained from experimental tests of the literature
and from a numerical model. The section was verified using this approach. According
to authors, the reinforcement could be designed to a sufficient degree of accuracy for
everyday professional practice. The method was easy to use and extensively applied
to reinforced concrete columns in buildings.
Tikka and Mirza [41] conducted a study to investigate the influence of different
variables on Cm of slender, tied, reinforced concrete columns. The existing expressions
for Cm were also investigated. The simulation of approximately 17,000 square
reinforced concrete columns each with a different combination of specified variables
was done. The columns studied were subjected to short-time ultimate loads and
unequal moments causing moment gradient in single curvature and double curvature
bending.
The theoretical equation for a pin-ended slender column subjected to unequal
end moments for Cm was proposed which was of the form Cm= Mcol/M2. The authors
selected two equations of Cm from the previous literature. These equations were
developed for steel columns. It was decided to investigate the applicability of these
equations to RC columns. The results of these equations were simulated for further
analysis and discussion. It was concluded from the study that there is no real advantage
in using these equations. Hence it was suggested that for design of RC columns, the
current ACI equation for Cm be replaced by the following two proposed equations:
 M 
Cm   0.6  0.4 1   0.3 …………………………………….……….. (2.24)
 M2 

  M1  
1.1

Cm  0.2  0.8  0.5  0.5    0.3 ……………….………………. (2.25)

  M2  

Kwak and Kim [43] introduced an analytical model which can simulate
nonlinear behaviour of slender RC columns. The authors adopted layer approach to
determine the equilibrium conditions in a section and to consider the different material
properties across the section depth effectively. The material nonlinearity including the
cracking of concrete was taken into account. Geometric nonlinearity due to load-
deflection effect was taken into consideration by using the initial stress matrix. The
literature related to present study and experimental results were correlated with each
other to verify the analytical results. Considering various parameters, the response of

23
slender RC columns was studied. The authors carried out a rigorous nonlinear analysis
for more accurate prediction of ultimate capacity of slender RC columns. The authors
[44] further introduced a simple design formula to predict the resisting capacity of
slender RC columns. On the basis of numerical analysis results obtained from the
companion paper, the resisting capacity reduction factors were determined. These
factors represent the reduction rates of resisting capacity in a long column
corresponding to a short column on a P-M interaction diagram. The authors also
established the relationships between the resisting capacity reduction factors and
design variables from regression analysis. This relationship was used to determine
resisting capacity of slender RC columns without any rigorous numerical analysis. In
order to establish the relative efficiency of the proposed formula, the results were
compared with ACI formula.
Tikka and Mirza [46] carried out a study to determine the influence of a full
range of variables on EI used for the design of slender, tied, composite columns in
which steel shapes are encased in concrete. The authors also examined the ACI EI
equations then available in the code. The columns considered in the study were 12000
in number. All the columns were having a different combination of specified
properties of variables. A new nonlinear equation for EI was developed for use in
design of slender composite columns subjected to major axis bending. This proposed
equation was an alternative to the existing ACI EI equations.
Elwood and Eberhard [49] estimated the effective stiffness of reinforced
concrete columns from the PEER structural performance database. It was compared
with stiffness calculated by following the guidelines recommended in Federal
Emergency Management Agency (FEM A) 356 seismic rehabilitation guidelines. The
authors noticed that FEMA 356 overestimates the stiffness of columns with low axial
loads in which there can be significant bar slip in the eam-column joint or footings.
The digest authors practical recommendations for improving estimates of effective
stiffness.
Lejano [53] addressed the problem of accurately predicting the behaviour of a
reinforced concrete column with biaxial bending through fiber modeling. It was used
to establish its capacity at the ultimate stage. In implementing this method, Bazant’s
Endochronic theory was used as a constitutive model for concrete and Ciampi model
was used for steel. The effects of different structural parameters were considered in
establishing the interaction surfaces. The author conducted numerical analyses of
24
square reinforced concrete columns with symmetrical reinforcement. The strength of
concrete considered in this study varied from 21 MPa to 62 MPa. The result of fiber
method modeling agreed well with some available experimental data. According to
author, the proposed method provides structural designers with an alternative way to
analyze and design such column sections.
Lam et al. [56] evaluated the ductility of columns that were high strength in
nature. It required nonlinear moment-curvature analysis taking into account the stress-
path dependence of the steel reinforcement. Based on this analysis, a parametric study
was conducted to investigate the effects of various factors on the ductility of columns.
It was concluded that the effect of concrete strength is dependent on the axial stress
level and axial load level. At the same axial stress level, the use of high strength
columns has little or no adverse effect on the ductility. On the other hand, if the same
axial load level is maintained to reduce the column size, the use of high strength
concrete would significantly reduce the ductility. The authors developed two formulae
for direct evaluation of the ductility of columns.
Pallarés et al. [58] proposed an expression to determine the equivalent factor
Cm that is valid for concrete pin ended columns subjected to a non-uniform moment
diagram. The methodology followed to obtain the proposed expressions for the
equivalent factor Cm was as follows:
(a) First, a theoretical approach to the problem was carried out assuming linear
elastic behavior of the material. In this case, the differential equation that
governs the behavior of the column subjected to axial load and to a non-
uniform moment diagram was solved.
(b) Second, the factor Cmi was deduced as the ratio between the first-order moment
of model case and the first-order moment of each load case-i. Then, an
adjustment by least-squares approach on the derived theoretical formulae of
Cmi was carried out for all load cases.

(c) A numerical simulation taking into account the non-linear behavior of

reinforced concrete using finite element method was carried out.
(d) The main variables on which the factor Cm depends were studied.
(e) The accuracy of applying the proposed theoretical expressions compared to
results from the numerical simulation was obtained.

25
 It was concluded that in elastic and linear behaviour, the equivalent factor Cm
depends on the type of load and the axial load (α). It is obtained as the ratio of the
applied axial and buckling load. The axial load, the slenderness of the column, the
creep coefficient, the location and the reinforcement area and the type of load are the
variables which affect the equivalent factor Cm.

2.6 Software based work on RCC

Ahmed et al. [54] investigated the effect of cracking on the lateral response of
building structures. The controversies in the expressions of important parameters
related to the cracking of reinforced concrete were discussed. The authors concluded
that the significant increase in deflections and drifts is observed with concrete cracking
considerations. For high rise structural systems, the story drifts after incorporation of
concrete cracking effect were enhanced appreciably. It was also concluded that the
drift requirements may fail after incorporation of concrete cracking effect with the
present guidelines of the country.
Vellasco et al. [61] presented a semi-rigid low-rise portal frame parametric
analysis. The parameters considered were connection stiffness and strength, steel and
composite structural system and lateral frame stability. ANSYS finite element
program of a structural system was developed based on Euro-code 3. The model took
into consideration geometrical and material nonlinearities, semi-rigid behaviour of the
connections. The results of the same model were obtained using the traditional most
economical method. These results were compared with ANSYS model and it was
observed that the ANSYS results yielded an economical solution with as much as 15%
economy in terms of weight of steel.
Dahmani et al. [62] developed three-dimensional nonlinear finite element
model of reinforced concrete beam. The general purpose finite element package,
ANSYS 8.0, was employed for the numerical analyses. Using SOLID65 solid
elements, the compressive crushing of concrete was facilitated using plasticity
algorithm while the concrete cracking in tension zone was accommodated by the
nonlinear material model. Smeared reinforcement was used and introduced as a
percentage of steel embedded in concrete. Comparison with hand calculated results
was presented for the concrete beam. Convergence of analytical results was shown.
The capability of the model to capture the critical crack regions, loads and deflections
for various types of loadings in reinforced concrete beam was illustrated. The authors
26
observed that because of smear reinforcement approach, beams with moderate shear
span failed to reproduce satisfactory results when compared with hand-calculated
results. The use of discrete reinforcement approach was suggested by the authors.
Srinivasan and Sathiya [63] article presented a study on the flexural behavior
of reinforced concrete beams through finite element analysis. The authors noted the
significance of understanding the response of concrete structural components during
loading to the development of overall safe and efficient structure. The authors
compared the results from elastic analysis of reinforced beam under transverse loading
to that attained from normal theoretical analysis. It was observed that the minor
difference in results was due to the fault of modelling such as lack of deduction of
volume of the mild-steel reinforcement from the concrete volume.
Vasudevan and Kothandaraman [64] carried out finite element analysis of RC
beams under 4-point loading. The authors carried out a large number of trials by
changing various parameters. The variables considered were concrete strength, mesh
density, effect of shear reinforcement on flexural behaviour, convergence criteria,
percentage reinforcement. It was concluded that tension and shear reinforcements are
to be precisely incorporated using discrete modelling technique to get accurate
behaviour.
Saifullah et al. [65] carried out flexural test on reinforced concrete beams to
study the flexural behaviour. The beam specimens were modeled in ANSYS, SAS
2005 with different reinforcement ratio. It was observed that 1st cracking location was
0.43L to 0.45L from the support. Maximum load carrying capacity at 1st cracking was
observed for over-reinforced beams. This load carrying capacity was equal to the
ultimate load for beams with balanced section.
Al-Shimmari et al. [66] presented a theoretical investigation on composite
reinforced concrete. The authors studied nonlinear three dimensional analyses of the
tested beams. It was assumed that there exists a perfect bond between concrete and
reinforcing steel. The elements used for concrete and steel reinforcement were SOLID
65 and LINK8 respectively. The authors concluded from the analytical results that
there were no cracks at the bottom of beam because of channel section. The fine cracks
were observed in concrete web but were negligible.
Xiaoming and Hongqiang [67] conducted a finite element investigation on load
carrying capacity of corroded RC beam. The elements used for concrete and steel
reinforcement were SOLID 65 and LINK8 respectively. The bond between
27
reinforcement and concrete was simulated by using COMBIN39 element. It was
observed from the results that as the corrosion ratio increases, the stiffness of corroded
beam would decrease, slip between bars and concrete would be larger and ductile
failure of RC beam would turn to brittle failure.
Patil et al. [68] described the analysis of deep beams subjected to two point
loading with three different L/D ratios (1.5, 1.6 1.71) using non-linear finite element
method of ANSYS. The elements used for concrete and steel reinforcement were
SOLID 65 and LINK8 respectively. The authors obtained the results of flexural strains
and stresses at the mid-span of the beam and shear stresses near the support of the
beam. It was observed that for smaller span/depth ratio, the deviation of strain pattern
at mid-section of beam was more.

2.7 Current International Scenario

As per IS:456-2000 [70], this parameter is considered only of plain concrete
section. A rigorous study on this parameter has been carried out by various researchers.
The current ACI expression for effective flexural rigidity for columns under
short-term loads is expressed as follows-
EI  0.2Ec I g  Es I se / (1  d ) …..……………………………………. ………...(2.26)

EI  0.4Ec I g / (1  d ) …………………………………………………..……... (2.27)

where βd = ratio of maximum factored dead (or sustained) load to maximum

total factored load and is always positive. For short-time loads, its value is taken as
zero. Hence, the simplified form is as follows:
EI   Ec I g  Es I se ………………………………….………………………… (2.28)

where α = stiffness reduction factor, Ec= modulus of elasticity of concrete, Ig

= moment of inertia of gross concrete section about centroidal axis, Es = modulus of
elasticity of steel and Ise = moment of inertia of steel about centroidal axis of cross-
section.
From literature review, it is observed that several investigators have proposed
expressions for load carrying capacity of columns, flexural stiffness of columns as
well as beams and effective moment of inertia. Various parameters affecting the
stiffness were considered in the study. ACI code recommends the expression for
modulus of rupture as
𝑓𝑟 = 7.5√𝑓𝑐′ ………..…………………………………………………...………(2.29)

28
where fc’ is mean compressive strength of concrete in ksi.
Eurocode represents modulus of rupture as
ℎ
𝑓𝑟 = 𝑚𝑎 ⌈𝑓𝑐𝑡𝑚 , (1.6 − 1000) , 𝑓𝑐𝑡𝑚 ⌉…………………..……………….……….(2.30)

fctm= mean axial tensile strength of concrete.

As per IS:456-2000, the modulus of rupture of concrete is given by
𝑓𝑟 = 0.7√𝑓𝑐𝑘 ………………..…………………………………………...………(2.31)
where fck is characteristic compressive strength of concrete in N/mm2. These
equations can be used in calculating the cracking moments.
The equations suggested by various codes for modulus of elasticity of concrete
are shown below.
𝐸𝑐 = 4700√𝑓𝑐𝑘 (ACI Code)……………………...…….…………………..…..(2.32)
720−𝑇
𝐸𝑐 = ( ) × 100 (Australian Code) [79] …………………...……………..(2.33)
670

𝐸𝑐 = 22[(𝑓𝑐𝑚 )/10]0.3 (Euro Code)…………………………………………...(2.34)

𝐸𝑐 = 𝐾0 + 0.2𝑓𝑐𝑢 (BS Code)……………………..….………..………...........…(2.35)
Euro-Code specifies the equation for deflection of beam specimens which
involves cracking moment, maximum moment in the beam due to loads. The
deflection values calculated by Euro-Code equation can be verified experimentally as
well as analytically. IS:456-2000 is silent about the provisions of such equation for
deflection of beam in cracked as well as un-cracked condition.
In structural analysis, especially in indeterminate structures, it becomes
essential to know material and geometrical properties of members. The IS codal
provisions recommend elastic properties of concrete and steel which are fairly accurate
enough. These elastic properties are known independently for concrete and steel. But
reinforced concrete is a composite material containing steel and concrete. While
performing analysis by any software for high rise building, cross area of plain concrete
section is taken into consideration whereas effects of reinforcement bars and concrete
confined by stirrups are neglected.
Present aim of study is to determine elastic properties of reinforced cement
concrete material. Two important stiffness properties such as AE and EI play
important role in analysis of high rise RCC building idealized as plane frame. It is also
hoped that knowing exact material and geometrical properties, analysis may become
more accurate and economical design can be achieved.

29
2.8 Objectives of the Study
The objectives of the present study are finalized based on the literature study.
1. RC section behaviour in pre and post cracking conditions
2. Role of reinforcement on strength and stiffness
3. Analytical and experimental investigation
4. To arrive at empirical relation for effective AE and EI values in axial and
flexural loading based on percentage of reinforcement
5. To propose a realistic design procedure for RCC members of structural frame
and suggestions for IS: 456 Codal provisions.

30