The current issue and full text archive of this journal is available on Emerald Insight at:

https://www.emerald.com/insight/1726-0531.htm

Nonlinear modelling and finite Nonlinear

modelling and
element analysis on the load- finite element
analysis
bearing capacity of RC beams with
considering the bond–slip effect
Pandimani Received 14 June 2021
Revised 17 July 2021
Department of Civil Engineering, JNTUK, Kakinada, India Accepted 17 July 2021

Markandeya Raju Ponnada

Department of Civil Engineering, MVGR College of Engineering,
Vizianagaram, India, and
Yesuratnam Geddada
Department of Civil Engineering, JNTUK, Kakinada, India

Abstract
Purpose – This paper aims to present nonlinear numerical simulations using the versatile finite element
(FE) analysis tool ANSYS and theoretical analysis based on code provisions to assess the load-carrying
capacity of reinforced concrete (RC) beams under two-point monotonic static loadings.
Design/methodology/approach – Four quarter-size FE models with load and geometry symmetry
conditions were constructed, the load-bearing capacity and associated mid-span deflections at critical points
are verified against the full-scale experimental RC beams available in the literature. These developed FE
models incorporated the tension stiffening effects and bond–slip behaviour. Theoretical analyses based on
Indian standard code IS: 456–2000 and ACI 318–19 were also carried to verify the experimental and numerical
predicted moments at critical loading points.
Findings – The load-deflection curves predicted through FE models exhibit closer corroboration with the
experimental curves throughout the loading history. The contour plots for deflections, concrete principal
stresses, reinforcement yield stresses are satisfactorily predicted by the FE models, which reveal the complete
information of nonlinear behaviour of RC beams. The developed model well captured the initial and
progressive crack patterns at each load increments.
Practical implications – The FE modelling is an efficient, valid and economical tool that is an alternative
to the expensive experimental program and can be used to explore, analyse and fully understand the
nonlinear response of RC beams under static loadings.
Originality/value – The ultimate moment capacity evaluated based on ACI 318–19 code provision show a
better correlation with the experimental data as compared to the IS: 456–2000 code provision. The ultimate
loads and associated centre-span deflections predicted by RN-2, RN-3, RB-12 and RB-16 FE model show a
discrepancy of 1.66 and –0.49%, –4.68 and –0.60%, –9.38 and –14.53% and –4.37 and 4.21%, respectively,

The authors thank the management of MVGR engineering college (Autonomous) and JNTUK
university for their support and encouragement towards the collaborative research.
Conflict of interest: There is no conflict of interest with anybody concerning this work.
Ethical statement: The authors declare that this work is original work done by them and has not
been submitted to any journal.
Journal of Engineering, Design
Funding body: No external funding was received. and Technology
Replication of results section: All results have been presented in the manuscript in a format that © Emerald Publishing Limited
1726-0531
enables easy reproduction. DOI 10.1108/JEDT-06-2021-0310
JEDT against the experimental results, which reveals that the developed ANSYS FE models predict consistent
results and achieved a reasonable agreement with the experimental data.
Keywords Crack patterns, Bond–slip model, FE analysis of RC beam, Load-deflection behaviour,
Stress contour plots, Theoretical analysis, ANSYS
Paper type Research paper

1. Introduction
The various available constitutive numerical equations and the use of finite element (FE)
software like ANSYS to predict the nonlinear performance of RC beams that delivers
comprehensive results with time, material and cost-effectiveness in contrast to the
experimental and theoretical methods. ANSYS is a versatile software for modelling,
simulation and validation of various complex structures and contains huge collections of
FEs that are capable of simulate various material, geometrical and boundary types of
nonlinearity problems with highly consistent results. The ANSYS software has been used
effectively by several investigators to validate their experimental findings and further
explored parametric studies with their calibrated numerical models (Godínez-Domínguez
et al., 2015; Kadhim et al., 2019; Sayed, 2019; Hawileh, 2015; Hawileh et al., 2019; Al-Rousan
et al., 2020). Al-Rousan et al. developed FE models for fibre-reinforced plastic (FRP)
reinforced concrete deck-slab using ANSYS software and studied the influence of varying
compressive strength of concrete, bottom transverse steel ratio and type of reinforcements
(Al-Rousan et al., 2020). They concluded that the stiffness and ultimate loads of slab with
FRP bars had a superior performance related to the steel reinforcements. Hamrat et al.
explored the deflection behaviour and failure modes of pre-cracked and FRP-repaired RC
beams through experimental and numerical study (Hamrat et al., 2020). Choobbor et al.
performed an experimental and numerical investigation on the strengthened RC beams
(Choobbor et al., 2019). They suggested that the RC beams strengthened with hybrid FRP
composite sheets improve the strength and ductility behaviour. Sayed et al. numerically
explored the effect of circular opening at shear span on the flexural behaviour of RC beams
(Sayed, 2019). Hawileh et al. examined the influence of side-bonded FRP sheets on the
flexural behaviour of RC beams through FE analysis (Hawileh et al., 2019). El-Sayed et al.
investigated the influence of recycled lathe steel fibres on the load-carrying capacity of RC
beams through experimental and numerical analysis (El-Sayed, 2019). Kadhim et al.
performed a numerical analysis of near-surface mounted (NSM) shear strengthened T-
beams to examine the influence of concrete grade, percentage of NSM reinforcement and
percentage of steel stirrups (Kadhim et al., 2019). Prajapati et al. developed ANSYS FE
models to probe the failure modes and load-deflection response of U-wrapped and striped
FRP shear strengthen RC beams and validated the numerical results against the
experimental tested beams existing in literature (Prajapati et al., 2017). Vasudevan et al.
created a batch file in ANSYS to investigate the flexural performance of conventional and
retrofitted RC beams with external steel bars and verified the numerical results against the
experimental data and theoretical results (Vasudevan and Kothandaraman, 2014). Hawileh
generated quarter-size FE models for unstrengthened and NSM FRP strengthened RC
beams using ANSYS (Hawileh, 2012). Xiaoming and. Hongqiang probed the flexural
behaviour of corroded RC beams through FE analysis (Xiaoming and Hongqiang, 2012).
They adopted COMBIN39 elements to develop the bond–slip mechanism between the
concrete–steel interfaces. Jayajyothi et al. presented the nonlinear FEA of carbon-fibre-
reinforced polymer (CFRP)-strengthened RC beams in the shear and flexure region (Jayajothi
et al., 2013). Dahmani et al. conducted nonlinear FE analysis to investigate the cracking
patterns and load-deflection behaviour of RC beams through the ANSYS model (Dahmani Nonlinear
et al., 2010). Smeared reinforcement method was used to model the reinforcements. modelling and
This study demonstrates the complete nonlinear numerical simulations of RC beams on load-
deflection behaviour, crack patterns and other capabilities of ANSYS FE models. To predict the
finite element
reliability of ANSYS software, the four developed numerical models are verified through analysis
comparison against the experimental data available in the literature. In this study, SOLID65,
LINK180, COMBIN39 and SOLID185 elements are used to represent the concrete, steel rebars,
interface bond–slip and loading and support plates based on the intensive literature study (Adawi
et al., 2016; Jnaid and Aboutaha, 2014; Vasudevan and Kothandaraman, 2015; Shetty et al., 2015;
Nwankwo and Ede, 2020; Ramesh et al., 2021; Kadhim et al., 2019). A detailed mesh sensitivity
study is performed to assess the optimum number of elements required to achieve a good accuracy
of results. The cracking pattern and load-deflection plots of the mesh sensitivity models are
presented and discussed. Also, full-, half- and quarter-scaled FE models are generated to verify the
(loading and geometry) symmetry effect, and their load-deflection plots are drawn. The predicted
outcomes illustrate that the developed quarter-size numerical model shows a relatively similar
response with the half- and full-size numerical models. Therefore, to use the advantage of symmetry
condition and to reduce the simulation duration, only quarter-size FE models are used in this study,
and the results are validated with the full-size experimental RC beams. The deflection and stress
contours of RC beams are also discussed. Subsequently, two code provisions, IS: 456–2000 and ACI
318–19, are used to evaluate the theoretical ultimate moment capacity, and the calculated moments
are verified with the experimental and numerical values.

2. Experimental beams review

To corroborate the developed FE models, four experimental tested RC beams are chosen from
two authors. Two beams RN-2 and RN-3 tested by Ashour and two beams RB-12 and RB-16
tested by Vasudevan et al. were adopted to model and validate the finite element analysis
(FEA) results (Ashour, 2000; Vasudevan et al., 2013). The geometry and material properties
of the experimental RC beam available in the literature are illustrated in Figures 1 and 2 and
Table 1, respectively.

3. Nonlinear finite element analysis

This section demonstrates the complete numerical modelling techniques of RC beams using
the ANSYS19 package. A detailed mesh sensitivity analysis, beam geometric symmetry
study and efficient modelling techniques of RC beam are illustrated. The numerical results

Figure 1.
Details of Ashour
tested RC beam
JEDT of the four FE models are corroborated against the experimental data available in the
literature. The load-deflection plots, crack patterns and stress contour plots are presented
and discussed broadly. The various discrete elements typically used to model the RC beams
are shown in Figure 3.

Figure 2.
Details of Vasudevan
et al. tested RC beam

fc’ ft f st Ast pt fy
Beams type (MPa) (MPa) (mm) (mm2) (%) (Mpa)

Material properties of Ashour beam

RN-2 48.61 5.64 2#18 508.93 1.18 530
RN-3 48.61 5.64 3#18 763.4 1.78 530
Material properties of Vasudevan et al. beam
Table 1. RB-12 35.8 4.19 2#12 226.2 0.52 525
Material property of RB-16 33.4 4.05 2#16 402.12 0.92 522
experimental RC
beams Source: Ashour, (2000); Vasudevan et al., (2013)

Figure 3.
Elements type used
for RC beam
modelling
3.1 Development of finite element models Nonlinear
3.1.1 Concrete material. Concrete is a composite and brittle material that possesses distinct modelling and
nonlinear material properties in tensions and compression (Al-Rousan et al., 2020). The
eight-node SOLID65 brick element is used to model the concrete material property with each
finite element
node that can free to move in translational x-, y- and z-directions (El-Sayed, 2019; Vasudevan analysis
and Kothandaraman, 2014; Ramesh et al., 2021; El-Shaer and Samaan, 2018; Salama et al.,
2019). This element is capable of modelling the cracking and crushing behaviour of concrete
in tension and compression in three orthogonal directions (ANSYS, 2007). In the ANSYS
model, the SOLID65 element experiences cracking or crushing behaviour when the tensile or
compressive principal stresses of the developed model reach the maximum tensile or
compressive strength of concrete respectively (Abu-Obeidah et al., 2015; Hawileh et al.,
2010). Consequently, the elastic modulus of those elements is set to zero in that direction by
the ANSYS program to converge the solution (Jawdhari and Harik, 2018; Banjara and
Ramanjaneyulu, 2019). Young’s modulus and Poisson’s ratio of concrete material are input
to model the linear behaviour (Al-Rousan et al., 2020; Choobbor et al., 2019). A multilinear
stress–strain data based on Thorenfeld et al. model is used to define the nonlinear behaviour
of concrete as presented through equations (1)–(5) and shown in Figure 4 (Adawi et al., 2016;
Wight and MacGregor, 2011). The ANSYS program uses a default bilinear stress–strain
relationship to define the tension behaviour of concrete, which assumes the model follows
linear behaviour until the maximum tensile stress of concrete ft (modulus of rupture stress)
as given in equation (6), and then a 40% vertical drop in stress occurs (0.6 ft), and thereafter,
it linearly descends to zero stress at a strain value equal to 6« t, as shown in Figure 5(b)
(Abu-Obeidah et al., 2015; Kadhim et al., 2019; ANSYS, 2007). It should be noted the
relaxation coefficient of 0.6 (ANSYS default) is adopted only to speed up the convergence of
solution when cracking is imminent (ANSYS, 2007). In ANSYS, William and Warnke’s
model governs the failure criteria of concrete, which require a minimum of four parameters
to define the failure that is ultimate tensile stress of concrete (ft), the maximum compressive
stress of concrete (fc’) and the open and closed shear transfer coefficients b o and b c,
respectively, at cracking face (Dahmani et al., 2010; Jawdhari and Harik, 2018; Kachlakev
et al., 2001; Wolanski, 2004). The b o and b c values range from 0–1 (Godínez-Domínguez
et al., 2015; Barour and Zergua, 2020; Barour et al., 2019). At the cracking face, 0 denotes a
complete loss of shear transfer and 1 denotes no loss of shear transfer (ANSYS, 2007; Barour
et al., 2019). In this study, b o = 0.3 and b c = 0.5 are adopted based on the literatures
(Choobbor et al., 2019; Hawileh, 2012). The actual and idealized stress–strain relationship in
compression is based on the Thorenfeldt et al. model, as shown in Figure 4(a) and (b),

Figure 4.
Compressive stress–
strain relationship of
concrete
JEDT respectively. The ANSYS software does not support the stress softening effect through the
load controlling method due to high convergence problems, and hence, the softening effect
(descending portion) of the curve is not considered in this study. The idealized curve follows
a parabolic shape up to the maximum compressive stress followed by a horizontal plateau
up to the strain limit of 0.0035, as shown in Figure 4(b).
 
n ««0c
fc ¼ fc0
c
 nk (1)
ðn  1Þ þ «c
«0
c

p 0
Ec ¼ 3320 fc þ 6900 (2)

0
n ¼ 0:8 þ fc =17 (3)

0 0


k¼ « c = « c 0 < 1:0 or 0:67 þ fc =62 « c = « c > 1:0 (4)

« c 0 ¼ fc 0 n=Ec ðn  1Þ (5)

p
ft ¼ 0:62 fc 0 (6)

where fc = compressive stress of concrete (MPa), e c = strain at fc, fc’ = maximum

compressive stress of cylinder (MPa), e c 0 = corresponding strain at fc’, k = a factor that
controls curve profile, n = curve fitting factor, ft = ultimate tensile stress of concrete (MPa)
and Ec = initial tangent modulus (MPa).
3.1.2 Steel reinforcements. Two-node LINK 180 spar element having three degrees of
freedom (DOF) at each node is used to model the longitudinal and transverse steel
reinforcements (Sayed, 2019; Hamrat et al., 2020; Nwankwo and Ede, 2020; Yousaf et al.,
2017). The steel reinforcements are assumed to be bilinear isotropic with identical elastic–

Figure 5.
Constitutive models
perfectly plastic behaviour in compression and tension, as shown in Figure 6(b). Young’s Nonlinear
modulus of 200 GPa and Poisson’s ratio of 0.3 are adopted for all the link 180 elements (steel modelling and
reinforcements) in this study (Hawileh et al., 2019; Jnaid and Aboutaha, 2014). The yield
stress values, as shown in Table 1, are adopted for the steel reinforcements for different
finite element
models. Steel reinforcements can be modelled based on the discrete, embedded or smeared analysis
method, as illustrated in Figure 6(a) (Bitencourt et al., 2018). The embedded and smeared
methods do not enable the bond–slip mechanism at the steel–concrete interfaces (Kachlakev
et al., 2001; Wolanski, 2004). The discrete method that facilitates the relative slip behaviour
of steel reinforcements is adopted in this study (Bitencourt et al., 2018; Adawi et al., 2016).
3.1.3 Bond–slip mechanism. The bond stress at the concrete–steel interfaces and the
relative slip of tensile rebars are modelled using dimensionless two-node spring element
COMBIN39 with an assigned longitudinal force–slip relationship as a real constant
(Xiaoming and Hongqiang, 2012; Adawi et al., 2016; Sakar et al., 2014). The spring elements
with three DOF at each node in translational x-, y- and z-directions are introduced between
the identical (coincident) nodes of tensile rebars and surrounding concrete along the
reinforcement direction to enable the relative slip of tensile rebars, whereas the other two
coordinates of the coincident nodes are constrained in the y- and z-directions using node
coupling (Adawi et al., 2016; Shetty et al., 2015). The other coincident nodes of compressions
and transverse steel reinforcements in the developed FE models are restricted in all
directions by merging them. The bond–slip relationship based on the CEB-FIP code 2010 for
monotonic loadings is used in this study, as depicted in Figure 5(a) (CEB-fip, 2010). The
given expression in equation (7) is derived for the ascending segment of the adopted curve,
which specifies the stiffness of the spring element followed by a horizontal plateau that is
equal to the maximum bond stress (t max) (Sakar et al., 2014; Hawileh, 2015). The complete
curve need not be assigned in this study because the discrete tensile reinforcements do not
come out completely from the concrete element (CEB-fip, 2010). According to the CEB-FIP
model, t max and the associated slip (s1) depends on the types of rebar and the compressive
strength of concrete. The value of t max = 2.5Hfc’ (MPa), curve fitting parameter a = 0.4, s1 =
1 mm and the slip value at the end of the horizontal plateau (s2) is taken equal to 2 mm based
on CEB-FIP code (Sakar et al., 2014; CEB-fip, 2010). The bond force F(s) can be achieved by
multiplying the bond stress t (s) with the circumferential area of steel reinforcements (p dl),
as given in equation (8). Where, s = slip (mm), l = length of single link180 element (mm) and
d = diameter of the rebar (mm). Figure 7(c) shows the developed bond–slip model assigned
between the coincident nodes of tensile rebars and concrete elements.

Figure 6.
Reinforcement model:
(a) types of methods
for reinforcement
modelling, (b) stress–
strain curve for steel
reinforcements
(Godínez-Domínguez
et al., 2015; Kachlakev
et al., 2001)
JEDT t ðsÞ ¼ t max ðs=s1 Þ a (7)

FðsÞ ¼ t ðsÞ p dl (8)

3.1.4 Loading and support elements. An eight-node SOLID185 brick element with three
translational DOF at each node is used to model the steel plates at the support and loading locations
(Kadhim et al., 2019; Barour and Zergua, 2020). The SOLID185 element acts as cushion plates,
which reduce the stress concentrations at these locations (Vasudevan et al., 2013; Wolanski, 2004).
Only the linear material property (Es and m ) identical to the tensile steel reinforcements is assigned
to this element. The bottom surface of this element is fixed to the SOLID65 elements using the
volume sweep option. The loading and support plate configuration and their position are shown in
Figures 7 and 10.
3.1.5 Boundary conditions. In this study, to minimise the computational effort and time,
only quarter-scale FE models are developed with a proper boundary condition at symmetric
planes and supports, as illustrated in Figure 10(a). The nodes of the mid-span (yz) and mid-
width (xy) symmetric planes are given a zero displacement in the x- and z-directions,
respectively, as depicted in Figure 10(a) (Abu-Obeidah et al., 2015; Kachlakev et al., 2001).
The hinged support boundary condition is created by assigning a zero displacement value in
the y- and z-directions to the centreline nodes (along the z-direction) of the support plate

Figure 7.
Development of FE
model
 Nonlinear
modelling and
finite element
analysis

Figure 8.
Mesh density details
of developed FE
models

Figure 9.
Mesh sensitivity
analysis

(Elyasian et al., 2006; Dahmani et al., 2010). The boundary conditions for half- and full-scale
FE models and reinforcement details are shown in Figure 10(c)–(f), respectively.
3.1.6 Nonlinear solution and failure condition. The ANSYS default Newton–Raphson
iteration method is used to achieve the nonlinear solutions of FE models in this study
(ANSYS, 2007; Banjara and Ramanjaneyulu, 2017). Using the automatic time-stepping
option, the total applied load is divided into several load increments by user-defined
maximum and minimum load steps. The convergence of the solution can be achieved in two
ways, either change the default tolerance limits or increasing the total number of load
JEDT

Figure 10.
Geometry and
reinforcement details
of FE model

increments and substeps. The default ANSYS program checks for displacement and force
convergence criteria (Godínez-Domínguez et al., 2015; Kachlakev et al., 2001). With the
default tolerance limits, the solution is unable to converge due to high nonlinear response,
particularly at the formation of the first crack and beyond the yielding stage. Therefore, to
overcome this problem, the force criteria are increased to 0.1 in this study to obtain the
complete solution up to the failure stage (Sakar et al., 2014; Hawileh, 2012). The typical range
of convergence criteria lies between 0.05 and 0.25 and is available in the literature study
(Elyasian et al., 2006; Prajapati et al., 2017; Hawileh et al., 2010). In this study, the FE model
is assumed to fail when a 10 N load increment causes the solution to diverge associated with
a large deflection, and accordingly, ANSYS displays a message stating that the deflection
limit of the FE model has reached the maximum limit of the ANSYS program.
3.1.7 Creation of quarter-size finite element model. In this study, only one-fourth size of
the RC beam is considered with geometry and load symmetry in the x- and z-directions,
which gives the same results as that of a full-size beam, as shown in Figure 7(a). To model Nonlinear
the quarter-size beam, the lines of steel reinforcements are created first using key points. modelling and
The lines have meshed with 25, 31.25 and 25 mm element length in the x-, y- and z-directions, finite element
respectively, to convert into steel reinforcements (LINK180), as shown in Figure 10(b). The
concrete block is incorporated over the steel reinforcement cage using solid volume with 170,
analysis
100 and 250 mm dimensions along the length, breadth and depth direction, respectively, as
depicted in Figure 7(a). The concrete volume is then discretized using the meshing attribute
command with the identical mesh density as that of steel reinforcements, as shown in Figure
7(b). Finally, steel plates at loading and support locations are created and fixed to the
concrete element using the volume sweep option, as shown in Figure 7(a)–(d). The coincident
nodes of concrete and tensile reinforcements are picked alone and the spring element is
assigned between each coincident node (shared nodes) along the length direction, while the
other two orthogonal directions are coupled together as shown in Figure 7(c). The boundary
conditions at two symmetric planes and support are generated using displacement
constraints. All the nodes in both the symmetric planes are constrained in a perpendicular
direction, as shown in Figure 10(a). As it is a quarter-size model, so one-fourth (P/4) load is
applied through the centreline of loading plates, which is equally distributed to all the nodes
such that every single node carries a magnitude of P/4n (n = number of nodes), as depicted
in Figures 7(c) and 10(a). The complete geometry configuration of all the elements used to
develop the quarter symmetry FE model is shown in Figure 7(d).
3.1.8 Mesh sensitivity analysis. The significant aspect of FEA is to achieve the optimum
element density in the developed model, which reflects on the accuracy of numerical solution
in terms of load-deflection response and stress parameters of RC beams (Al-Rousan et al.,
2020; Banjara and Ramanjaneyulu, 2017). This can be achieved practically when the
increasing element density exhibits an insignificant effect on the FEA results (Al-Rousan
et al., 2020). Figure 8 represents the four constructed FE models with varying element
densities. The element size along the width and depth direction is kept constant as 25 and
31.25 mm, whereas the longitudinal mesh size is changing from 150, 75, 50 and 25 mm,
respectively, as depicted in Figure 8. The FE model with 2,581 elements shows the ultimate
deflection (d u, FEA = 61.23 mm) value closely matches with the ultimate deflection of the
experimental beam (d u, Exp. = 61.10 mm), as shown in Figure 9 (a). The load-deflection
plots of the four mesh sensitivity FE models reveal that all the model exhibits similar
behaviour up to the yield stage, except the model with the lowest mesh density of 561
elements, as depicted in Figure 9(b). The result analysis of the four sensitivity models is
presented in Table 2, which compares the predicted and experimental values. The load-
deflection plot of the developed model with 2,581 elements exhibits a closer agreement with
the experimental curve, as shown in Table 2 and Figure 9(b), and therefore, this model is

FE model mesh densities

Parameters Exp. 561 934 1,351 2,581

Cracking load, Pcr (kN) 12.43 21.09 18.05 16.61 16.6

Crack deflection, d cr (mm) 1.15 1.01 1.08 1.29 1.23
Yield load, Py (kN) 77.9 97.98 81.98 80.00 80.00
Yield deflection, d y (mm) 17.99 15.64 15.27 17.88 17.76 Table 2.
Ultimate load, Pu (kN) 90.18 106.09 92.00 88.86 88.74 FEA results from
Ultimate deflection, d u (mm) 60.93 33.27 66.91 57.49 61.23 mesh density study
JEDT confirmed as the required optimum mesh density. Hence, the same mesh density is used for
all the RC beams analysed in this study.

3.2 Comparisons of geometry symmetry models

The developed FE models for half-, full- and quarter-type RC beams with the symmetric
plane and support boundary condition s along with the reinforcement details are illustrated
through Figure 10(a)–(f). In the literature on FE simulation of RC beams using ANSYS, most
of the researchers have only modelled the quarter-size RC beam for validation purposes
without checking the full-size model effect. In this study, we developed full-, half- and
quarter-size FE models and simulated them up to their ultimate failure load to assess their
effect and the ultimate load-deflection curves are compared with each other. It is evident
from Table 3 and Figure 11(a) that all the three types of developed FE models have exhibited
a similar behaviour up to the ultimate stage and the load-deflection plots, which are in
excellent agreement with the experimental plot at all loading stages. Therefore, to save the
computational effort and duration, the quarter symmetry model is the better choice, which
gives a similar response as compared to the half- and full-scaled models. The bending
moment comparisons at critical loading points of the three developed models are verified
against the experimental beam data, as shown in Figure 11(b), which further suggests that
all the models exhibit closer agreement with experimental results.

3.3 Validations of finite element models

To validate and ensure the reliability of the ANSYS FE models, the four experimental-tested
RC beams discussed in the preceding sections were modelled and analysed up to their

ANSYS FE analysis
Sr. No. Parameters Experimental Quarter size Half size Full size

1 First crack load, Pcr (kN) 12.43 16.60 17.24 17.18

2 first crack deflection, d cr (mm) 1.15 1.23 1.18 1.16
Table 3.
3 Yield load, Py (kN) 77.90 80.00 82.3 81.99
Results comparison 4 Yield deflection, d y (mm) 17.99 17.76 16.74 16.70
of experimental, full, 5 Ultimate load, Pu (kN) 90.18 88.74 91.15 91.01
half and quarter 6 Ultimate deflection, d u (mm) 61.10 61.23 61.06 61.072
beam models 7 Disk memory consumed (GB) NA 4.40 15.25 31.10

Figure 11.
FEA results of full-,
half- and quarter-size
models
ultimate failure using ANSYS software. The predicted numerical results regarding loads Nonlinear
and deflections at critical loading points and the load-deflection plots are presented and modelling and
compared with the experimental data available in the literature. finite element
3.3.1 Load-deflection response of RN-2 and RN-3 beams. Table 4 compares the load,
deflection and displacement ductility values between the experimental beams and numerical FE
analysis
models. The predicted ultimate loads of RN-2 and RN-3 FE models show a deviation of 1.66 and –
4.68 percentages related to the experimental values. The maximum deflection of RN-2 and RN-3
numerical models exhibits a divergence of –0.21 and –0.60 percentages concerning the experimental
beams. The yield load and deflection of RN-2 and RN-3 numerical model exhibits –10.65 and 1.27%,
–4.03 and 8.24% differences, respectively, with the experimental beams. Table 4 shows evidence
that the yield deflection increases with increasing tensile steel percentage and in contrast, the
ultimate deflection shows the adverse effect with increasing steel content. Table 4 also reveals the
experimental load at first crack and steel, yielding show a larger discrepancy with the numerical
results. It should be noted that the exact prediction of load and deflections at crack and yield points
are highly difficult to achieve through experimental investigation, and most often, can be judged
through the load-deflection plot, and hence, errors are probable. Whereas, the FE models can able to
predict the first crack and yield loads accurately. Hence, the results displayed by the FE models
look more sensible at those points as compared to the experimental data in this study. The yield and
ultimate load-bearing capacity of the RN-3 FE model reinforced with three bars exhibit 50 and
47.49% higher capacity compared to the RN-2 FE model, which is reinforced with two bars. The
RN-3 model shows a 5% increment in yield deflection and a reduction in ultimate deflection of
16.21% as compared to the RN-2 model. The load-deflection curve plotted in Figure 12(a) and (b)
reveals that the RN-2 and RN-3 FE models exhibit a closer response with the experimental
behaviour, respectively, at all loading stages. The RN-2 model exhibits a better correlation with the
experimental load-deflection curve as compared to the RN-3 model, as shown in Figure 12.

Beams type Pcr (kN) d cr (mm) Py (kN) d y (mm) Pu (kN) d u (mm) md Table 4.
RN-2 Exp. 12.43 1.15 72.30 17.99 90.19 61.10 3.39 Verification of
RN-2 FEA 16.60 1.22 80.00 17.76 88.69 61.23 3.44 experimental and
RN-3 Exp. 13.4 1.28 115.35 20.38 124.96 50.99 2.50 FEA results of RN-2
RN-3 FEA 18.46 1.38 120.00 18.70 130.81 51.30 2.74 and RN-3 beams

Figure 12.
Validation of load-
deflection plot
JEDT 3.3.2 Load-deflection response of RB-12 and RB-16 beams. The load-deflection and ductility
behaviour of RB-12 and RB-16 FE models are compared with the experimental data, as
illustrated in Table 5. The cracking loads predicted by the RB-12 and RB-16 FE models
exhibit –2.34 and 4.06% deviation, respectively, against the experimental results. However,
the FEA deflections show large differences with the experimental values, as the numerical
models can effectively capture the first crack pattern, whereas it is highly difficult to
identify the first cracking through experimental testing. Hence, the numerical predicted
cracking deflections are more relevant for these beams. The predicted yield load and
deflections show a disparity of –20.14 and –8.05, –6.80 and 8.02% with the experimental
data for RB-12 and RB-16 FE models, respectively. The numerical ultimate loads of RB-12
and RB-16 models show a divergence of –9.38 and –4.38 percentage as compared to the
experimental data. The ultimate deflection s of RB-12 and RB-16 numerical models show a
difference of –14.53 and –4.21 percentage with the experimental data. Table 5 reveals that
the yield load, ultimate load and yield deflections are increased with increasing tension
reinforcement percentage, whereas the ultimate deflection is drastically reduced as the steel
content increased, which ultimately reduces the displacement ductility of RC beams. The
load-deflection plots depicted in Figure 13(a) and (b) show that the FE models predict a
closer agreement with the experimental curves with reasonable accuracy. The RB-16 FE
model captured the load-deflection response closer to the experimental behaviour compared
to the RB-12 model, as depicted in Figure 13. It is concluded through FEA that the developed
ANSYS models are reliable and can be used to predict the flexural behaviour of RC beams
effectively.
3.3.3 Finite element analysis results discussions of RN-2 and RN-3 beams. The output
results predicted from the validated numerical FE model RN-2 and RN-3 FE are displayed in
Figures 14 and 15, respectively. The output results of the FE models are advantageous over the

Table 5. Beams type Pcr (kN) d cr (mm) Py (kN) d y (mm) Pu (kN) d u (mm) md
Verification of FEA RB-12 experimental 32.00 1.425 74.91 5.46 96.00 19.20 3.52
and experimental RB-12 FEA 32.75 0.42 90.00 5.90 105.00 21.99 3.73
results of RB-12 and RB-16 experimental 34.91 0.98 145.09 7.98 160.00 17.56 2.20
RB-16 beams RB-16 FEA 33.49 0.43 154.96 7.34 167.00 16.82 2.29

Figure 13.
Validation of load-
deflection plot
 Nonlinear
modelling and
finite element
analysis

Figure 14.
FEA results of RN-2
beam

Figure 15.
FEA results of RN-3
beam

experimental testing in several aspects. The linear variable differential transformer (LVDT) and
finite strain gauges are used in the experimental study to assess the output parameters at a certain
location within the beams, which could only predict the results at those points, whereas the
numerical FE models provide comprehensive results throughout the beam section at each load
increments up to failure stage. For instance, Figures 14 and 15 show the ultimate deflection, third
JEDT (compressive) principal stresses in the concrete at failure, steel reinforcement stresses at yielding
and the ultimate crack patterns for RN-2 and RN-3 beams, respectively. It is evident from Figures 14
(a) and 15(a) that the nodal deflection in the y-direction is maximum at the mid-span region and
lowest at the support region and the ultimate deflection reduces with increasing tensile steel
percentages. As depicted in Figures 14(b) and 15(b), the principal compressive stresses in concrete at
failure are maximum and concentrated over the compression fibre between the regions of the
applied loads. Figures 14(c) and 15(c) reveal the distributions of stress contours for tensile,
compression and shear reinforcements at yielding of tension steel, which reveal that the FE model
accurately predicted the yield stress, and the stresses in compression steel increases with increasing
tensile steel percentage. Finally, the crack patterns of concrete at ultimate failure are illustrated in
Figures 14(d) and 15(d), respectively, which evidence that the developed numerical models
effectively captured the flexural, diagonal and crushing type of cracks. The first (flexural) cracks
form at the bottom fibre within the maximum moment zone for all the models. The first crack
appears for RN-2, RN-3 models at a load equal to 16.60 and 18.46, respectively. Beyond the cracking
load, numerous flexural cracks are formed at the bottom fibre within the maximum moment region
and also at the shear span. With higher load increments, the diagonal tensile cracks are developed
and propagated towards the loading points. At the yielding stage, the RN-2 and RN-3 FE models
show several diagonal cracks that are originated from the flexural cracks at the shear span at a load
magnitude of 80 and 120 kN, respectively. The ultimate cracking of RN-2 and RN-3 FE models
occurs at 88.69 and 130.81 kN, and the ultimate crack patterns are shown in Figures 14(d) and 15(d),
respectively. The FE models can crack in three orthogonal directions at the integration points. The
ultimate crack patterns of all these FE models are shown with three cracking signs with red, green
and blue colours, which represents the first, second and third cracks at each integrated points are
shown in Figures 14(d) and 15(d) for RN-2 and RN-3 models, respectively.
3.3.4 Finite element analysis results discussions of RB-12 and RB-16 beams. Figures 16 and 17
represent the output results in terms of ultimate mid-span deflection, principal compressive
stresses of concrete at yielding, tensile reinforcement stresses at yielding and the ultimate

Figure 16.
FEA results of RB-12
beam
 Nonlinear
modelling and
finite element
analysis

Figure 17.
FEA results of RB-16
beam

crack patterns in concrete for the validated FE models RB-12 and RB-16, respectively.
Figures 16(a) and 17(a) reveal that the ultimate mid-span deflection of the developed
numerical models is reduced from 21.99 to 16.82 mm when the area of steel
reinforcements is increased from 226 to 402 mm2 respectively. For the FE models with
similar concrete compressive strength as shown in Figures 16(b) and 17(b), it is evident
that the principal compressive stresses at failure are nearly equal for both beams, even
though there is a big variation in the steel reinforcement area. The yield load and
deflection of RB-12 and RB-16 models have been increased from 90 to 155 kN and 5.90 to
7.34 mm, respectively, with increasing area of tension reinforcements, as depicted in
Figures 16(c) and 17(c). The ultimate crack patterns and types of cracks developed for
RB-12 and RB-16 models are shown in Figures 16(d) and 17(d), respectively. It should be
noted that the compressive stresses in the top reinforcement increase from 70 to
132 MPa with increasing tensile steel reinforcements, as shown in Figures 16(c) and 17
(c), respectively. The first tensile crack appears between the maximum moment zone at
bottom fibre for RB-12 and RB-16 FE models at a load magnitude of 32.75 and 33.49 kN,
respectively. Beyond the cracking load, several flexural cracks are formed within the
maximum moment zone and also at the shear span region. With further increments in
load, the diagonal tensile cracks are appeared and propagated towards the loading
points. At yielding of tension reinforcements, RB-12 and RB-16 FE models exhibit
several diagonal cracks at the shear span with a load magnitude of 90 and 154.96 kN,
respectively. The ultimate cracking of RB-12 and RB-16 FE models occurs at 105 and
167 kN loadings, and the respective crack patterns are shown in Figures 16(d) and 17(d),
respectively. The developed FE models can capture and store the cracking patterns at
each load increment efficiently. The first, second and third cracks (red, green and blue
colour) appear in three mutually perpendicular directions at each integrated points of
these FE models with flexural, diagonal tensile and crushing types of cracks, which are
depicted in Figures 16(d) and 17(d), respectively.
JEDT 4. Theoretical analysis
This section presents the stress block diagrams, compatibility expressions used for the
theoretical analysis of RC beams based on the two code provisions. This section also
provides the moment (cracking, yield and ultimate) comparisons between the experimental,
FEA and theoretical results. The classifications of the beam section based on the strain
limits in the steel reinforcements are presented in Figure 18(a) and (b) for the singly
reinforced RC beams based on the IS: 456–2000 and ACI 318–19 codes, respectively.

4.1 Based on Indian standard (IS) code provision

According to the IS:456–2000 code provision, the RC beam sections are classified based on
the neutral axis depth (xu). When xu < xu,lim or xu = xu,lim or xu > xu,lim are the conditions for
the section being under, balanced or over reinforced section, respectively, where, xu,lim is the
limiting neutral axis depth, which is derived from the strain compatibility diagram, as
shown in Figure 18(a). In this study, all the beams exhibited ductile behaviour where the
tensile steel reinforcement yielded first before the crushing of concrete at ultimate failure.
The stress–strain distribution of under reinforced concrete beam section is shown in
Figure 19, which depicts the equilibrium of forces at the ultimate stage by considering the

Figure 18.
Strain limits in
tension-steel
reinforcements

Figure 19.
Stress–strain
distribution diagram
for under reinforced
section based on IS
456-2000 code
provision
(considering FOS = 1
for both concrete and
steel material)
factor of safety of concrete and steel material equal to 1 instead of 0.67fck and 0.87fy, Nonlinear
respectively [11,29]. The theoretical calculations based on IS: 456–2000 and ACI: 318–19 modelling and
code provisions for RN-2 beam are illustrated in Appendices A and B, respectively. The finite element
theoretical cracked depth of neutral axis (xcr) and moment of inertia (Icr) are evaluated based
on the transformed area method by the expressions given in equations (9) and (10),
analysis
respectively (Ashour, 2000):

bxcr 2 =2 þ mAst –mAst d ¼ 0 (9)

2
Icr ¼ bxcr 3 =3 þ mAst ðd  xcr Þ (10)

where m = modular ratio = Es/Ec

Depth of neutral axis ðxu Þ ¼ fy Ast =0:542fck b (11)

p
Modulus of rupture stress ft ¼ 0:7 fck (12)

Cracking moment Mcr ¼ ft I=yb (13)

Yield moment Mu ¼ fy Ast ðd  xcr =3Þ (14)

Ultimate moment of resistance Mu ¼ fy Ast ðd  0:42xu Þ (15)

where fy = yield stress of steel (MPa), Ast = total area of tensile reinforcements, fck =
characteristics compressive strength of cube (MPa) b = breadth of beam, I = gross moment
of inertia, yb = depth of neutral axis from bottom, d = effect depth.

4.2 Based on American Concrete Institute code provision

According to ACI 318–19 code provision, the RC beam sections are classified based on the
strain limit (« t) in the extreme tensile reinforcements at failure. When « t # 0.002 or 0.002 <
« t < 0.005 or « t  0.005, the beam sections are classified as compression controlled,

Figure 20.
Stress–strain
distribution diagram
for tension-controlled
section based on ACI
318–19 code
provision
JEDT transition controlled and tension controlled, respectively (ACI Committee 318-19, 2019). The
limiting strain values in the extreme tensile steel for the tensioned and compression
controlled section are shown in Figure 18(b). In this study, all the beams fall under tensioned
controlled one with ductile-type failure at ultimate loads. The stress–strain distribution for
the tensioned controlled section is depicted in Figure 20.

Depth of rectangular block a ¼ fy Ast =0:85fc 0 b (16)

Depth of neutral axis c ¼ b 1 a (17)

p
Modulus of rupture stress ft ¼ 0:62 l fc 0 (18)

Cracking moment ¼ ft I=yb (19)

Nominal moment of resistance Mn ¼ fy Ast ðd  0:5aÞ (20)

4.3 Comparison of numerical, experimental and theoretical analysis

The moment comparison of experimental, FEA and theoretical calculations at first cracking,
steel yield and the ultimate stage is illustrated in Tables 6, 7, and 8, respectively. The
cracking moment predicted from FEA and theoretical calculation and the percentage
deviation in comparison with the experimental moments are shown in Table 6. The cracking
moments of the FE model show a difference of –33.54 and –37.84% for RN-2 and RN-3
beams and –2.38 and 4.06% for RB-12 and RB-16 beams, respectively. as presented in
Table 6. The theoretical cracking moments based on IS: 456 and ACI 318 codes show a
maximum percentage deviation of –26.80 and 12.18%, –12.34 and 20.41% for RN-2 and RB-
16 FE models, respectively. Table 6 also reveals that the cracking moments of FEA show
large percentage differences with the experimental values but are in reasonable agreement

Table 6. Mcr,EXP Mcr,FEA Mcr,IS:456 Mcr,ACI:318

Cracked moment Beams type (kN-m) (kN-m) % difference (1) (kN-m) % difference (2) (kN-m) % difference (3)
comparison between RN-2 8.02 10.71 –33.50 10.17 –26.78 9.01 –12.29
experimental, FEA RN-3 8.64 11.91 –37.81 10.17 –17.68 9.01 –4.23
and theoretical RB-12 8.80 9.01 –2.34 8.73 0.84 7.73 12.18
values RB-16 9.60 9.21 4.07 8.43 12.21 7.46 22.24

Table 7. My,EXP My,FEA My,IS:456 My,ACI:318

Yield moment Beams type (kN-m) (kN-m) % difference (1) (kN-m) % difference (2) (kN-m) % difference (3)
comparison between RN-2 50.25 51.60 –2.69 52.06 –3.61 51.91 –3.30
experimental, FEA RN-3 74.40 77.40 –4.03 76.53 –2.86 76.27 –2.52
and theoretical RB-12 20.60 24.75 –20.15 23.55 –14.34 23.50 –14.08
values RB-16 39.90 42.61 –6.80 40.60 –1.77 40.49 –1.47
with the theoretical values. There is a judgement error to identify the first crack through Nonlinear
experimental study, whereas the FE models can accurately capture the first flexural crack. modelling and
Therefore, the FEA results are more realistic at this loading stage.
The theoretical yield moments are calculated based on the expression given in equation (14)
finite element
by assuming linear stress distribution. The yield moment comparisons between the analysis
experimental, FEA and theoretical calculations are illustrated in Table 7. The percentage
differences between the FEA and experimental yield moments show a percentage divergence
of –2.69, –4.03, –20.15 and –6.80 for RN-2, RN-3, RB-12 and RB-16 beams, respectively. The
theoretical yield moments based on both IS:456–2000 and ACI code provision exhibits less than
4% discrepancy with the experimental data, except for RB-12 beams, which show a difference
of nearly 15%. It should be noted that the theoretical yield moment evaluated using equation
(14) gives overestimate values that are higher than the ultimate moments, which does not give
a true sense, and therefore, the expression should be modified to justify the yield moment
calculation. The advantage of the numerical FE model is that it can able to predict the stresses
contours of tensile steel reinforcements at yielding load with superior accuracy as compared to
the other methods, and therefore, the yield moments predicted through ANSYS FEA gives
more reasonable results over the theoretical as well as experimental results in this study.
The FEA, experimental and theoretical ultimate moment capacity and their percentage
variation in contrast to the experimental results of all the RC beams are presented in Table 8.
The ultimate moments predicted by RN-2, RN-3, RB-12 and RB-16 FE models show a
percentage difference of 1.66, –4.68, –9.39 and –4.38, respectively, as compared to the
experimental beams. The theoretical ultimate moments evaluated through the ACI 318–19
code predicts closer results with the experimental and FE values in comparison to the

Mu,EXP Mu,FEA Mu,IS:456 Mu,ACI:318 Table 8.

Beams type (kN-m) (kN-m) % difference (1) (kN-m) % difference (2) (kN-m) % difference (3) Ultimate moment
RN-2 58.17 57.21 1.66 52.19 10.27 53.59 7.87 comparison between
RN-3 80.60 84.37 –4.68 73.94 8.26 77.09 4.36 experimental, FEA
RB-12 26.40 28.88 –9.38 24.01 9.07 24.37 7.68 and theoretical
RB-16 44.00 45.93 –4.38 40.02 9.05 41.25 6.25 values

Figure 21.
Comparison of
ultimate moments
between
experimental, FEA
and theoretical values
JEDT results based on the IS:456–2000 provisions, as illustrated in Table 8. Overall, the numerical
and theoretical predicted ultimate moment capacity of RC beams shows consistent results
with the experimental data with a maximum discrepancy of less than 11%, as evident in
Table 8. Figure 21 depicts the comparison of ultimate moment capacity between the
experimental, numerical and theoretical values, which reveal that the FE model exhibits a
good accuracy of results with the experimental and theoretical data. Figure 21 also reveals
that the ultimate moment capacity evaluated based on the ACI code provision exhibits
closer results with the FEA as compared to the IS code provisions.

5. Conclusions
This study demonstrates the nonlinear FEA on the flexural behaviour of RC beams using
the sophisticated numerical package ANSYS (APDL). FEA indicates that the developed
numerical models can efficiently predict the complete nonlinear behaviour of RC beams.
Also, additional information regarding the formation and distribution of cracks, stresses
distribution of steel and concrete elements at critical loading stages are well captured by the
ANSYS models. The following are the conclusions and observations drawn from this
investigation:
 The reliability of the developed FE model is demonstrated through comparison
against the experimental results, which show closer agreements in terms of load-
deflection plot, ultimate loads, ultimate deflections and the numerical models that
can efficiently capture the cracking patterns of RC beams at all loading stages.
 The first cracking in concrete, steel reinforcement yielding and the respective
principal stress contours of concrete and steel reinforcements are accurately
predicted by the developed numerical models, which reveals the practicability of
ANSYS FE models for performing nonlinear analysis of RC beams.
 The yield and ultimate load-carrying capacity and yield deflection of numerical
models increased drastically with increasing area of tension, steel but consequently,
the ultimate deflection and the displacement ductility reduced significantly.
 The maximum load and deflections predicted by RN-2 and RN-3 FE models show a
difference of þ1.67 and –0.49% and –4.68 and –0.60%, respectively, against the
experimental results.
 The deviation between the FEA and experimental results for maximum load and
mid-span deflection of RB-12 and RB-16 beams exhibits –9.37 and –14.53%, –4.37
and þ4.21%, respectively.
 The predicted ultimate moment-carrying capacity based on FEA, IS:456-2000 and
ACI 318-19 code provision shows a discrepancy of 1.66, –4.67, –9.37, –4.37 and
10.27, 8.26, 9.05, 9.05 and 7.87, 4.35, 7.68, 6.25% against the experimental ultimate
moment capacity for RN-2, RN-3, RB-12 and RN-16 beams, respectively. These
values reveal that the FE models exhibited a closer response with the experimental
data as compared to the theoretical analysis.
 The theoretical analysis based on ACI 318-19 code provision predicts a better
correlation with the experimental and FE results as compared to the IS:456-2000
code.
 The numerical FE model presented in this study offers a good understanding of the
nonlinear behaviour of RC beams at a lower cost with effective and rapid results.
References Nonlinear
Abu-Obeidah, A., Hawileh, R.A. and Abdalla, J.A. (2015), “Finite element analysis of modelling and
strengthened RC beams in shear with aluminum plates”, Computers and Structures,
Vol. 147, pp. 36-46.
finite element
ACI Committee 318-19 (2019), “Building code requirements for structural concrete (ACI 318-19)”.
analysis
Adawi, A., Youssef, M.A. and Meshaly, M.E. (2016), “Finite element modeling of the composite action
between hollow core slabs and the topping concrete”, Engineering Structures, Vol. 124 No. 30,
pp. 1-15.
Al-Rousan, R.Z., Alhassan, M. and Al-Wadi, R. (2020), “Nonlinear finite element analysis of full-scale
concrete bridge deck slabs reinforced with FRP bars”, Structures, Vol. 27, pp. 20-31.
ANSYS (2007), A Finite Element Computer Software and User Manual for Nonlinear Structural
Analysis, ANSYS, Inc, Canonsburg, PA.
Ashour, S.A. (2000), “Effect of compressive strength and tensile reinforcement ratio on flexural
behaviour of high-strength concrete beams”, Engineering Structures, Vol. 22 No. 4, pp. 13-23.
Banjara, N.K. and Ramanjaneyulu, K. (2017), “Experimental and numerical investigations on the
performance evaluation of shear deficient and GFRP strengthened reinforced concrete beams”,
Construction and Building Materials, Vol. 137, pp. 520-534.
Banjara, N.K. and Ramanjaneyulu, K. (2019), “Effective CFRP retrofit strategy for flexural deficient RC
beams”, Structural Engineering Mechanics, Vol. 69 No. 1, pp. 63-75.
Barour, S. and Zergua, A. (2020), “Numerical analysis of reinforced concrete beams strengthened in
shear using carbon fiber reinforced polymer materials”, Journal of Engineering, Design and
Technology, Vol. 19 No. 2, pp. 339-357.
Barour, S., Zergua, A., Bouziadi, F. and Abed Jasim, W. (2019), “Finite element analysis of CFRP-
externally strengthened reinforced concrete beams subjected to three-point bending”, World
Journal of Engineering, Vol. 17 No. 2, pp. 183-202.
Bitencourt, L.A.G., Manzoli, O.L., Trindade, Y.T., Rodrigues, E.A. and Dias-da-Costa, D. (2018),
“Modeling reinforced concrete structures using coupling finite elements for discrete
representation of reinforcements”, Finite Elements in Analysis and Design, Vol. 149, pp. 32-44.
CEB-fip (2010), “Model code for concrete structures”, International Federation for Concrete Structures,
Ernst and Sohn, Lausanne.
Choobbor, S.S., Hawileh, R.A., Abu-Obeidah, A. and Abdalla, J.A. (2019), “Performance of hybrid
carbon and basalt FRP sheets in strengthening concrete beams in flexure”, Composite
Structures, Vol. 227, pp. 1-11.
Dahmani, L., Khennane, A. and Kaci, S. (2010), “Crack identification in reinforced concrete beams using
ANSYS software”, Strength of Materials, Vol. 42 No. 2, pp. 232-240.
El-Sayed, T.A. (2019), “Flexural behavior of RC beams containing recycled industrial wastes as steel
fibers”, Construction and Building Materials, Vol. 212, pp. 27-38.
El-Shaer, M.A.A. and Samaan, M.F. (2018), “Experimental and numerical investigation of flexural
behaviour of new DSG reinforced concrete mixes reinforced with GFRP bars”, Ain Shams
Engineering Journal, Vol. 9 No. 4, pp. 37-49.
Elyasian, I., Abdoli, N. and Rounagh, H.R. (2006), “Evaluation of parameters effective in FRP shear
strengthening of RC beams using FE method”, Asian Journal of Civil Engineering, Vol. 7,
pp. 249-254.
Godínez-Domínguez, E.A., Tena-Colunga, A. and Juarez-Luna, G. (2015), “Nonlinear finite element
modeling of reinforced concrete haunched beams designed to develop a shear failure”,
Engineering Structures, Vol. 105, pp. 99-122.
Hamrat, M., Bouziadi, F., Boulekbache, B., Daouadji, T.H., Chergui, S. and Labed, A. (2020),
“Experimental and numerical investigation on the deflection behavior of pre-cracked and
JEDT repaired reinforced concrete beams with fiber-reinforced polymer”, Construction and Building
Materials, Vol. 249, pp. 1-13.
Hawileh, R.A. (2012), “Nonlinear finite element modeling of RC beams strengthened with NSM FRP
rods”, Construction and Building Materials, Vol. 27 No. 1, pp. 61-71.
Hawileh, R.A. (2015), “Finite element modeling of reinforced concrete beams with a hybrid combination
of steel and aramid reinforcement”, Material and Design, Vol. 65, pp. 1-9.
Hawileh, R.A., Abdalla, J.A. and Naser, M.Z. (2019), “Modeling the shear strength of concrete beams
reinforced with CFRP bars under unsymmetrical loading”, Mechanics of Advanced Materials
and Structures, Vol. 26 No. 15, pp. 1290-1297.
Hawileh, R.A., Musto, H.A., Abdalla, J.A. and Naser, M.Z. (2019), “Finite element modeling of reinforced
concrete beams externally strengthened in flexure with side-bonded FRP laminates”, Composites
Part B, Vol. 173, pp. 1-36.
Hawileh, R.A., Rahman, A. and Tabatabai, H. (2010), “Nonlinear finite element analysis and modeling
of a precast hybrid beam-column connection subjected to cyclic loads”, Applied Mathematical
Modeling, Vol. 34 No. 25, pp. 62-83.
Jawdhari, A. and Harik, I. (2018), “Finite element analysis of RC beams strengthened in flexure with
CFRP rod panels”, Construction and Building Materials, Vol. 163 No. 7, pp. 51-66.
Jayajothi, P., Kumutha, R. and Vijai, K. (2013), “Finite element analysis of FRP strengthened RC beams
using ANSYS”, Asian Journal of Civil Engineering, Vol. 14, pp. 631-642.
Jnaid, F. and Aboutaha, R.S. (2014), “Residual flexural strength of reinforced concrete beams with
unbonded reinforcement”, ACI Structural Journal, Vol. 111 No. 6, pp. 19-30.
Kachlakev, D.I., Miller, T.H., Potisuk, T., Yim, S.C. and Chansawat, K. (2001), Finite element modeling
of reinforced concrete structures strengthened with FRP laminates, Technical Reports For
Oregon Department of Transportation Research Group, Washington, DC.
Kadhim, M.M.A., Adheem, A.H. and Jawdhari, A.R. (2019), “Nonlinear finite element modelling and
parametric analysis of shear strengthening RC T-beams with NSM CFRP technique”, International
Journal of Civil Engineering, Vol. 17 No. 8, pp. 1295-1306, doi: 10.1007/s40999-018-0387-8.
Kadhim, A.M.H., Numan, H.A. and Özakça, M. (2019), “Flexural strengthening and rehabilitation of
reinforced concrete beam using BFRP composites: finite element approach”, Advances in Civil
Engineering, Vol. 2019, pp. 1-17.
Nwankwo, C.O. and Ede, A.N. (2020), “Flexural strengthening of reinforced concrete beam using a
natural fibre reinforced polymer laminate: an experimental and numerical study”, Materials and
Structures, Vol. 53 No. 6, pp. 1-13.
Prajapati, G.N., Pandey, A.D. and Aalok Reddy, T.B. (2017), “Finite element simulation of shear
strengthened RC beams externally reinforced with FRP”, Asian Journal of Civil Engineering,
Vol. 18, pp. 357-370.
Ramesh, B., Eswari, S. and Sundararajan, T. (2021), “Experimental and numerical studies on the
flexural behaviour of GFRP laminated hybrid-fibre-reinforced concrete beams”, Innovative
Infrastructure Solution, Vol. 6 No. 18, pp. 1-13.
Sakar, G., Hawileh, R.A., Naser, M.Z., Abdalla, J.A. and Tanarslan, M. (2014), “Nonlinear behavior of
shear deficient RC beams strengthened with near surface mounted glass fiber reinforcement
under cyclic loading”, Materials and Design, Vol. 61, pp. 16-25.
Salama, A.S.D., Hawileh, R.A. and Abdalla, J.A. (2019), “Performance of externally strengthened RC
beams with side-bonded CFRP sheets”, Composite Structure, Vol. 212 No. 2, pp. 81-90.
Sayed, A.M. (2019), “Numerical study using FE simulation on rectangular RC beams with vertical
circular web openings in the shear zones”, Engineering Structures, Vol. 198, pp. 1-10.
Shetty, A., Venkataramana, K. and Babu Narayan, K.S. (2015), “Experimental and numerical
investigation on flexural bond strength behavior of corroded NBS RC beam”, International
Journal of Advanced Structural Engineering, Vol. 7 No. 3, pp. 23-31.
Vasudevan, G. and Kothandaraman, S. (2014), “Finite element analysis of bearing capacity of RC beams Nonlinear
retrofitted with external bars”, Strength of Materials, Vol. 46 No. 6, pp. 31-42.
modelling and
Vasudevan, G. and Kothandaraman, S. (2015), “RC beams retrofitted using external bars with
additional anchorages-a finite element study”, Computers and Concrete, Vol. 16 No. 3, pp. 15-28. finite element
Vasudevan, G., Kothandaraman, S. and Azhagarsamy, S. (2013), “Study on non-linear flexural behavior analysis
of reinforced concrete beams using ANSYS by discrete reinforcement modeling”, Strength of
Materials, Vol. 45 No. 2, pp. 31-41.
Wight, J.K. and MacGregor, J.G. (2011), Reinforced Concrete Mechanics and Design, 6th ed., Prentice-
Hall.
Wolanski, A.J. (2004), “Flexural behavior of reinforced and prestressed concrete beams using finite
element analysis”, M.Sc thesis, Marquette University, Milwaukee, Wisconsin, USA.
Xiaoming, Y. and Hongqiang, Z. (2012), “Finite element investigation on load carrying capacity of
corroded RC beam based on bond-slip”, Jordan Journal of Civil Engineering, Vol. 6 No. 1,
pp. 34-46.
Yousaf, M., Siddiqi, Z.A., Sharif, M.B. and Qazi, A.U. (2017), “Force- and displacement-controlled non-
linear FE analyses of RC beam with partial steel bonded length”, International Journal of Civil
Engineering, Vol. 15 No. 4, pp. 499-513.

Further reading
Arafa, M., Alqedra, M. and Hammad, M. (2018), “Effect of fiber orientation on shear behavior of RC
beams externally strengthened with CFRP”, International Journal of Composite Materials, Vol. 8,
pp. 64-71.
IS 456-2000 (2000), Plain and Reinforced Concrete Code of Practice, Bureau of Indian standards, New
Delhi, pp. 1-114.
JEDT Appendix 1. A1 Theoretical analysis of under reinforced concrete beams using Indian
standard IS: 456–2000 code provisions
A1.1 at first crack
The modulus of rupture:
p p
ft ¼ 0:7 fck ¼ 0:7  48:61 ¼ 4:88 MPa

The gross moment of inertia:

Ig ¼ BD3 =12 ¼ 200  2503 =12 ¼ 260:42  106 mm4

The elastic modulus of concrete:

p p
Ec ¼ 5000 fck ¼ 5000  48:61 ¼ 34:86  103 MPa

The cracking moment:

Mcr ¼ ft Ig =yt ¼ 4:88  260:42  106 =125 ¼ 10:17  106 N  mm

The cracking load:

Pcr ¼ 2Mcr =av ¼ 2x10:17  106 =1290 ¼ 15:76  103 N

A1.2 at steel yielding Nonlinear
The modular ratio: modelling and
m ¼ Es =Ec ¼ 200000=34:86  103 ¼ 5:74
finite element
analysis
The area of steel reinforcement:

Ast ¼ 2ð p =4Þd2 ¼ 2  0:79  182 ¼ 508:94 mm2

The cracked neutral axis depth:

bxcr =2 þ mAst xcr  mAst d ¼ 0

100xcr þ 5:74  508:94xcr –5:74  508:94  215 ¼ 0

xcr ¼ 65:97 mm

The cracked moment of inertia:

2
Icr ¼ bxcr 3 =3 þ mAst ðd  xcr Þ

2
200  65:973 =3 þ 5:74  508:94  ð215  65:97Þ

Icr ¼ 83:99  106 mm4

The yield moment:

My ¼ fy Ast ðd  xcr =3Þ

¼ 530  508:94  ð215  65:97=3Þ ¼ 52:06  106 N  mm

(My is over estimated here)

The yield load:

Py ¼ 2Mcr =1290
¼ 2  52:06  106 =1290 ¼ 80:72  103 N

A1.3 at ultimate failure

The actual depth of neutral axis:

xu ¼ fy Ast =0:542fck b
¼ 530  508:94=0:542  48:61  200 ¼ 51:19 mm

The limiting depth of neutral axis:

JEDT xu;lim ¼ 0:429d ¼ 0:429  215 ¼ 92:33 mm

As xu is less than xu,lim, the section is under reinforced.

The ultimate moment of resistance:

Mu ¼ fy Ast ðd  0:42xu Þ
¼ 530  508:94  ð215  0:42  51:19Þ ¼ 52:19  106 N  mm

The ultimate load:

Pu ¼ 2  52:19  106 =1290 ¼ 80:91  103 N

Appendix 2. B1 Theoretical analysis of tension-controlled RC beam using American

concrete institute ACI 318–19 code provisions
B1.1 at first crack
The modulus of elasticity of concrete:
p 0 p 3
Ec ¼ 4700 fc ¼ 4700  48:61 ¼ 32:77  10 MPa

The modulus of rupture:

p p
ft ¼ 0:62 fc 0 ¼ 0:62  48:61 ¼ 4:32 MPa

The cracking moment:

Mcr ¼ ft Ig =yt ¼ 4:32  260:42  106 =125 ¼ 9  106 N  mm

The cracking load:

Pcr ¼ 2Mcr =av ¼ 2  9  106 =1290 ¼ 13:95  103 N

B1.2 at yielding of reinforcement

The yielding moment:

My ¼ fy Ast ðd  xcr=3Þ
¼ 530  508:94  ð215  65:97=3Þ ¼ 52:06  106 N  mm

The yield load:

Pcr ¼ 2Mcr =av ¼ 2  52:06  106 =1290 ¼ 80:72  103 N

(Pcr value is overestimated here)

B1.3 at ultimate failure
The stress intensity factor:


b 1 ¼ 0:85  0:05ðfc 0–28Þ=7
Nonlinear
¼ 0:85  0:05  ð48:61  28Þ=7 ¼ 0:702 modelling and
finite element
The depth of stress block: analysis
a ¼ fy As =0:85fc 0b
¼ 530  508:94=0:85  48:61  200 ¼ 32:64 mm

The depth of neutral axis:

c ¼ a= b 1 ¼ 32:64=0:702 ¼ 46:50 mm

The tensile strain in reinforcement:

« t ¼ « cu ðd–cÞ=c ¼ 0:003  ð215  46:50Þ=46:50 ¼ 0:0108 > 0:005

Hence, the section is tensioned controlled.

The nominal moment:

Mn ¼ fy As ðd  0:5aÞ
¼ 530  508:94 ð215  0:5  32:64Þ ¼ 53:59  106 N  mm

The ultimate load at failure:

Pu ¼ 2  53:59  106 =1290 ¼ 83:09  103 N

The design moment:

Md ¼ f Mn ¼ 0:9  53:59  106 ¼ 48:23  106 N  mm

The design load at failure:

Pd ¼ 2  48:23  106 =1290 ¼ 74:78  103 N

Corresponding author
Markandeya Raju Ponnada can be contacted at: markandeyaraju@yahoo.com

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com