ADVANCED

Original Research Article

COMPOSITES LETTERS

Advanced Composites Letters

Volume 28: 1–15
ª The Author(s) 2019
Structural performance of frames Article reuse guidelines:
sagepub.com/journals-permissions
with concrete-filled steel tubular DOI: 10.1177/2633366X19894593
journals.sagepub.com/home/acm

columns and steel beams: Finite

element approach

Ahmed Dalaf Ahmed1,2 and Esra Mete Güneyisi2

Abstract
Composite columns such as concrete-filled steel tube (CFST) were adopted in many building constructions in recent years
because of carrying high loading with the ability to resist buckling and small cross-sectional area. The high behavior of the
CFST columns is due to the interaction between steel and concrete which called the composite action. This type of
composite column without main and tie reinforcements embedded in concrete gives high axial compression strength to
resist the external loadings with the economic sectional area. The work presented in this article includes simulation models
that tested by other researchers and a parametric study on the performance of frames that connected steel beam by
composed columns of circular CFST that subjected to lateral loading. A finite element (FE) approach is adopted to simulate
the models by ANSYS software. All models consider the linear and nonlinear material analysis of the concrete and steel. The
validity of the developed model was examined by comparing with the experimental data founded in the literature. Different
parameters such as the ratio of the axial load, the slenderness ratio of CFST column, the linear stiffness ratio of the beam–
column, the steel yield strength of the beam, the steel yield strength of the tube, and concrete strength on the performance
of the composite frames were also studied and the load-deformation performance was obtained over the different cases of
the study. Analysis results by FE modeling were in good agreement with the experimental results.

Keywords
composite frame, confined concrete, concrete-filled steel tube column, finite element modeling, lateral loading, stiffness,
ANSYS

Introduction to develop a nonlinear FE model to study the performance

The constructions of composite elements such as composite
columns become so fast that make and lead the composite
column is a key application that adopted by most of the
structural engineering designers during recent years.
Concrete-filled steel tube (CFST) columns have been con-
structed in different structures around the whole world due
to the combined properties of steel and concrete. Experi-
mental and theoretical approaches were adopted by many
researchers to try understanding the actual behavior of the
composite columns but little deals with the steel beam rest
on CFST columns. The most effective parameters that
affect the behavior and performance of such problems were
investigated.1–9 Schneider5 simulated a three-dimensional
(3-D) nonlinear finite element (FE) model for CFST mem-
bers by the ABAQUS software. Hu et al.6 used ABAQUS
of CFST columns with confinement effect. Han et al.7 also
used ABAQUS for modeling the torsional response of
CFST columns. They adopted a uniaxial compression
stress–strain relationship for the confined core concrete
of CFST columns, and the core concrete was simulated
1
Department of Civil Engineering, Anbar University, Al Anbar, Iraq
2
Department of Civil Engineering, Gaziantep University, Gaziantep,
Turkey
Date received: 24 July 2019; accepted: 19 November 2019
Corresponding author:
Ahmed Dalaf Ahmed, Department of Civil Engineering, Anbar University,
Al Anbar, Iraq; Department of Civil Engineering, Gaziantep University,
27310 Gaziantep, Turkey.
Email: ahmeddalaf44@gmail.com

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons
Attribution-NonCommercial 4.0 License (http://www.creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial
use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open
Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage).
2 Advanced Composites Letters
using a plastic damage concrete model in which the anal-
ysis results compared with experimental ones. Liu and
Foster8 studied by FE approach the performance of con-
centrically loaded columns with concrete strength up to
100 MPa. Ellobody and Young9 studied the behavior of
normal and high strength CFST columns under the effects
of static loadings by ABAQUS software. They used the
cold-formed stainless steel tube. The concrete model and
the other parameters used were similar as in the study of
Hu et al.6 There are recent investigations on the perfor-
mance and behavior of the composite frames and connec-
tions such as presented by Han et al.10 They investigated
square CFST columns framed to steel beams by experi-
mentally and FE modeling under cyclic loading in which
the analysis was verified by six frame specimens. Also,
Han et al.11 investigated the behavior and strength of the
frames with circular CFST columns under a constant
axial load on the CFST columns and a lateral cyclic load.
Based on the analysis of the results, a simplified hystere-
tic lateral load versus lateral displacement model was
proposed. Mou et al.12 studied the earthquake response
of frame joints subjected to bidirectional cyclic loading
test and they concluded that in case of constant D/t ratio,
the bending moment of the specimens under the biaxial
symmetrical loading was about 20% less as compared to
specimens under uniaxial loading. Ding et al.13 studied
the frame of circular CFST column to steel-concrete
composite beam under lateral cyclic loading. The results
based on the FE modeling were closed with the experi-
mental results in terms of the failure modes, load–dis-
placement hysteretic curves, and skeleton curves.
Hajjar et al. 14 proposed a 3-D modeling of interior
beam-to-column composite connections. Wang et al.15
investigated several types of CFST column to steel
beam connections by using two types of CFST with
square and circular sections with outside stiffening ring
plate or with penetrating ring plate. It was pointed out
that these joints could improve the resistant of the pro-
gressive collapse when compared with the other types of
connections. Mou and Bai16 results indicated that rebar
stiffeners could be very effective to improve the seismic
behavior of joints for the CFST column and steel beam.
Mou et al.17 declared that the composite action of the
floor slab and the steel beam was particularly significant
under the sagging moment for the CFST column (square
section) and steel beam joint. Muhummud18 predicted
the seismic behavior of a multi-story CFST composite
frame using the nonlinear dynamic time history analysis
by DRAIN-2DX software. Herrera 19 investigated the
earthquake response of CFST column-wide flange beam
frames.
Aim and significance of research
Experimental investigations of the CFST elements are quite
costly and time spoliation. FE approach as a branch of
numerical analysis has become more desirable or suitable
due to the widespread. Many FE software can be utilized
to simulate the composite sections, one of them is ANSYS
software.20 The main focus on the analysis and design of
the CFST column is the strength capacity, ductility, and the
whole stiffness of the column in addition to the column–
beam connection relation. The important points that the
actual relationships between the concrete core and the sur-
rounding steel tube in the elastic, elasto-plastic, and in the
plastic zone but still the performance of the composite
frame in the elastic range is the base for structural design.
This study sets out a numerical realization of the perfor-
mance and strength of frames with CFSTs and steel beams
using the FE approach by ANSYS software. The CFST
column and their connection with the steel beam offer more
than an increase in strength capacity and stiffness but give
more ductile so that no sudden failure will occur. The
essential elements that adopted to simulate the composite
action of the CFST are concrete, steel, and the interface
between the outer concrete surface and inner steel tube
elements. The main objective of the present study is to
characterize and develop an accurate FE model to simu-
late the performance of the composite frames with circular
CFST columns to a steel beam by the FE software
ANSYS. Toward this aim, the effects of concrete strength
and concrete confinement were considered in the analysis.
The interface between concrete and the steel tube was also
modeled as a contact surface. To validate the FE model,
the experimental test data given in the literature were
used, then, a parametric study was conducted in which
31 composite frame cases were considered. The adopted
parameters included the axial load level, the slenderness
ratio of the columns, the linear stiffness ratio of the beam–
column, the steel yield strength of the beam, the yield
strength of the tube, and concrete strength. Each model
specimen consisted of two CFST columns and a steel
beam to represent a typical frame element in a building,
and each specimen was subjected to the lateral load. The
analyses of the results were given and discussed
comparatively.
Composite frames—FE approach
Thirty-one CFST composite frames were designed and ana-
lyzed to study the effect of the different parameters on the
performance of such frames. All specimen information
summary is listed in Table 1 while Figure 1 shows the
layout of the frame model and basic frame elements and
Figure 2 represents the composite frame with an illustration
of the beam–column and column–base connection. Differ-
ent parameters are considered as follows
(a) Axial load ratio (n) of CFST columns such as (0.0,
0.1, 0.3, 0.5, 0.7, 0.9, and 1.1): The axial load ratio is
described as n ¼ (No/Nu), where No is the applied
Ahmed and Güneyisi 3

Table 1. Details of the composite frames and basic parameters studied.

h  bf  tw  tf

Model mark Member D  ts (mm) L (mm) k n l f 0 c (MPa) fy Column (MPa) fy Beam (MPa)
F1-1 Beam 327.66  309.88  15.5  25.4 4775 3.0 0 48 28 290 290
Column F254  14.76 3048
F1-2 Beam 327.66  309.88 15.525.4 4775 3.0 0.1 48 28 290 290
Column F254  14.76 3048
F1-3 Beam 327.66 309.88  15.5  25.4 4775 3.0 0.3 48 28 290 290
Column F254  14.76 3048
F1-4 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 290 290
Column F254  14.76 3048
F1-5 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.7 48 28 290 290
Column F254  14.76 3048
F1-6 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.9 48 28 290 290
Column F254  14.76 3048
F1-7 Beam 327.66  309.88  15.5  25.4 4775 3.0 1.1 48 28 290 290
Column F254  14.76 3048
F2-1 Beam 333.24  319.92  18  28 4775 3.0 0.5 24 28 290 290
Column F254  5.92 1524
F2-2 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 290 290
Column F254  14.76 3048
F2-3 Beam 311.15  305.8  10.9  17.01 4572 3.0 0.5 72 28 290 290
Column F254  14.76 3048
F2-4 Beam 309.62  254.25  9.14  16.25 4775 3.0 0.5 96 28 290 290
Column F254  14.76 6096
F3-1 Beam 327.66  309.88  15.5 25.4 7690 1.8 0.5 48 28 290 290
Column F254  14.76 3048
F3-2 Beam 327.66  309.88  15.5 25.4 4775 3.0 0.5 48 28 290 290
Column F254  14.76 3048
F3-3 Beam 327.66  309.88  15.5 25.4 3584 4.2 0.5 48 28 290 290
Column F254  14.76 3048
F3-4 Beam 327.66  309.88  15.525.4 2900 5.4 0.5 48 28 290 290
Column F254  14.76 3048
F3-5 Beam 327.66  309.88 15.525.4 2466 6.6 0.5 48 28 290 290
Column F254  14.76 3048
F4-1 Beam 327.66  309.8815.525.4 4775 3.0 0.5 48 28 290 235
Column F254  14.76 3048
F4-2 Beam 327.66 309.8815.525.4 4775 3.0 0.5 48 28 290 290
Column F254  14.76 3048
F4-3 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 290 345
Column F254  14.76 3048
F4-4 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 290 390
Column F254  14.76 3048
F4-5 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 290 420
Column F254  14.76 3048
F5-1 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 235 290
Column F254  14.76 3048
F5-2 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 290 290
Column F254  14.76 3048
F5-3 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 345 290
Column F254  14.76 3048
F5-4 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 390 290
Column F254  14.76 3048
F5-5 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 420 290
Column F254  14.76 3048
F6-1 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 28 290 290
Column F254  14.76 3048
F6-2 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 48 290 290
Column F254  14.76 3048

(continued)
4 Advanced Composites Letters

Table 1. (continued)
h  bf  tw  tf

Model mark Member D  ts (mm) L (mm) k n l f 0 c (MPa) fy Column (MPa) fy Beam (MPa)
F6-3 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 68 290 290
Column F254  14.76 3048
F6-4 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 88 290 290
Column F254  14.76 3048
F6-5 Beam 327.66  309.88  15.5  25.4 4775 3.0 0.5 48 108 290 290
Column F254  14.76 3048

Figure 1. Typical frame element (unit: mm).

axial load to the column and Nu is the axial com-
pressive column capacity.
Nu ¼ f y As þ f c Ac ð1Þ
(b) Slenderness ratio (l) of CFST columns (24, 48, 72,
and 96): The slenderness ratio (l) is
4H
l ¼ ð2Þ
D
(c) Beam-to-column linear stiffness ratio (k) (1.8, 3.0,
5.4, and 6.6): k is computed as iB/iC, iB is determined
as EsIeq/lb, where Ieq is the moment of inertia of the
transformed section in accordance with GB 50017-
2003,21 and lb is the length of the steel beam. iC is
calculated as EhIh/H. The stiffness of the circular
CFST column (EhIh) is calculated as
E h I h ¼ E s I s þ 0:8 E c I c ð3Þ
Based on AISC-LRFD22
(d) Steel beam yield strength ( fy Beam) (235, 290, 345,
390, and 420 MPa).
(e) Steel column tube yield strength ( fy Column) (235,
290, 345, 390, and 420 MPa).
(f) Concrete compressive strength ( f c0 ) (28, 48, 68, 88,
and 108 MPa).
FE modeling of the frames
Material modeling
The composite action between the steel and the concrete
that caused by the contact between the steel tube and the
concrete in a CFST column leading to exerted radial lateral
confining pressure by the steel tube on the concrete induces
the confinement in concrete. A biaxial state of stress is also
Ahmed and Güneyisi 5

Figure 2. Typical composite frame and connections in FE model. FE: finite element.

induced in steel, which may cause some reduction in its axial

load capacity. Thus, any numerical model that intends to
capture the behavior of CFST in compression must use a
suitable constitutive model for the steel tube and concrete.
Further, to effectively replicate the inherent advantages of
CFST, it is necessary that the composite action between steel
and concrete could be carefully modeled.

Concrete. Confinement of concrete column by means of steel

tube section enhances the strength capacity of composite
column and concrete ductility. The steel tube provides pas-
sive confinement to the concrete and over the entire compo-
Figure 3. Stress–strain relation of steel tube confined concrete.23
site member. The adopted compressive strength f c0 in case of
unconfined and the f cc0 for confined. The adopted stress–
strain for confined concrete is shown in Figure 3,23 and the ecc is the strain corresponding to maximum stress fcc and eo
confined concrete computed as follows23,24: is the strain corresponding to maximum compressive stress.
f cc0 ¼ f c0 þ k 1 f 1 ð4Þ Supposing that the confinement of concrete in the initial stage
of loading (elastic) does not exist because the Poisson’s ratio
The value of k1 conducted by a test that ranged between of concrete in cases of normal and high strength concrete is
4 and 6 in which the present study 4.5 is adopted.7 The lower than that of the steel tube (0.3) so that at the initial stage,
average radial pressure fl relies on the ratio of diameter to the stress–strain performance is same as for unconfined con-
thickness of the tube (D/t) that was calculated by adopting crete. After the initial stage, the confinement will work and
the following empirical formula23: the stress is calculated by Saenz equation as follows23:
 1:027
fl D Eo e
¼ 0:3111  2 ð5Þ s ¼      ð7Þ
fy t 1þ
Eo
2
e
þ
e
E cs e cc e cc
0
!
f cc fcc
e cc ¼ eo ð6Þ E cs ¼ ð8Þ
f c0 e cc
6 Advanced Composites Letters

The descending part of the stress–strain curve is repre- Table 2. Material properties of the steel—CFST frame.11
sented as follows23:
  Yield Ultimate Modulus of
  a e  e cc strength strength elasticity Poisson’s
s ¼ f p þ f cc  f p e e cc
ð9Þ Member fy (MPa) fu (MPa) Es (MPa) ratio v
Tube of CFST 327.7 397.9 2.063  105 0.266
In which fp is the residual compressive stress and k is the Steel beam 303.0 440.9 2.061  105 0.262
parameter representing the descending rate equal to 0.6.11
The magnitudes of fp rely on the D/t ratio and are calculated CFST: concrete-filled steel tube.
as follows:
!0:1
0:1 f y Table 3. Verified theoretical and experimental results.
f p ¼ f cc in case ðD=tÞ  40 ð10Þ
f }c Theoretical
  Experimental FE model
  ðD=tÞ 40 Item results11 results Theoretical
% Experimental
0:1 f y 0:1 
f p ¼ f cc g þ ð 1  g Þe 40
ð11Þ
f cc Ultimate lateral 55.25 55.60 100.6
load (kN)
in case ðD=tÞ  40 Lateral displacement of 12.50 12.46 99.7
the frame at yield
(mm)
Steel tube column and base plate. The steel tube that sur-
rounding the concrete core subjected to hoop tension stress FE: finite element.
during the applied loading on the composite columns. The
yield criteria based on Von Misses are adopted as following
with axial and hoop stress sa and sh, respectively:
Concrete–steel tube interface
The coefficient of friction between the concrete and steel
s 2a þ s a s h þ s 2h ¼ f y2 ð12Þ surfaces used was 0.6 as suggested by Han et al. 11 ANSYS
The stress–strain curve as the numerical analysis is interface elements were used in order to model the contact
assumed to be linear with a slope of Es up to the yield point area between the steel tube and the concrete. The two sur-
and then the slope changes to 0.1Es to improve the numer- faces of the element in the normal direction is a hard con-
ical stability analysis.23 tact, while the tangent contact surface is modeled by the
Coulomb friction model. As far as the faces remain in
contact, the friction between them is maintained.
Elements description
Solid 186 element was used to model the steel tube, its FE model verification
stiffening diaphragms, and the steel beam and adopted to
To verify the validity of the FE model, the analysis results
model the base plate of 50 mm in thickness to prevent any
obtained from the FE model were compared with the
stress concentration. This element is a higher order 3-D
experimental results. Composite frames with circular CFST
20-node solid element that exhibits quadratic displace-
columns to steel beam tested by Han et al.11 were utilized in
ment behavior. Each node had three degrees of freedom
this study to verify the proposed FE model using ANSYS
translations x, y, and z directions. While for the concrete,
software.
the solid element Solid 65 which has eight nodes with
For the frame CF-13,11 with circular CFST columns, the
three degrees of freedom at each node—translations in
beam span was 1450 mm and column height 2500 mm. The
the nodal x, y, and z directions. The element is capable
column size is F140  2 mm thick tube, while the beam
of plastic deformation, cracking in three orthogonal direc-
size (depth  flange width  flange thickness  web thick-
tions, and crushing.
ness) is 140  65  3.44  3.44 mm4. Table 2 lists the
mechanical properties for the beam and CFST column.
The verification analysis results are given in Table 3. The
Loading and the boundary conditions comparison shows closed for predicting the load–displace-
The load condition which is used in these models had con- ment behavior of the composite frames. Figure 4 shows the
stant axial load through nodes on columns according to deflected shape of the frame under the FE model analysis.
special values, and lateral load subjected on the end of the Figure 5(a) to (c) shows the maximum Von Misses stresses
steel beam through nodes increases step-by-step until it at the regions of the frame connections. The application of
reaches failure. The support conditions at the left and right Von Misses yield criteria as the square of sum difference
of columns bases are restraints in all directions(x, y, and z) between the principle stresses must be less than the square of
so that working as fixed. the magnitude of the yield strength. The Von Misses stress at
Ahmed and Güneyisi
the frame connections was greater than the yield strength so
that the frame fail at the final load stage. The largest recorded
stress values were in the CFST column–steel beam connec-
tion and the lower part of CFST column nearest to the loading
region was observed. Figure 6 also represents the comparison
between the experimental and numerical analysis result based
on the load-displacement behavior.
Nonlinear analysis
In the nonlinear static analysis, the ultimate applied load
was divided into a number of load increments as steps. At
the end of each load step, the model stiffness matrix was
recalculated and updated the nonlinear changes in the struc-
tural stiffness and then it proceeds to the next load step by
applying the Newton–Raphson method that built in
ANSYS with a displacement control tolerance of 0.001.
Figures 7 and 8 illustrate the failure modes, the maximum
stresses, and the deformed shape of the composite frame,
all the composite frames exhibited similar failure mode.
Plastic hinges were formed at the top and the bottom of the
CFST column when the loads reach the ultimate. The
Figure 4. Deflected shape of Han et al.’s11 frame under FE model
analysis for the verification purpose. FE: finite element.
Figure 5. Maximum principle stresses at the region of (a) the right column–base connection, (b) the left column–base connection, and
(c) the beam–column connection of Han et al.’s frame.11
7
maximum stress propagations are shown in Figures 7 and
8 at the ultimate load stage. The stress concentration at the
beam–column connection and at the top and bottom regions
of the CFST due to this area is the weak zone. The maximum
stress is 312 MPa which is greater than the yield strength 290
MPa that indicates a plastic hinge formation at the connec-
tion of the steel beam and at the top of the CFST.
Figure 9 illustrates the stages of loading and the resulted
stresses. At stage 1, the lateral load is low and elastic condi-
tion prevails, when the load increased and the elastic range is
over, higher stresses are shown at the tension sides of the
column and beam. At stage 3, the final stage, the maximum
stresses are reached and the failure has occurred as shown in
the beam just near the connection, which confirms that the
failure was based on the weak beam–strong column principle.
Parametric study
Different parameters are considered in the numerical anal-
ysis such as axial load ratio (n), slenderness of the column,
linear stiffness ratio of the beam–column (k), steel yield
strength of the beam ( fyBeam), steel yield strength of the
tube of the column ( fyColumn), and concrete strength ( f 0 c ).
It is worthy to note that the computation only alters one
parameter while preserving other parameters unaltered. In
all cases, the toughness was calculated.
Axial load ratio
The axial load ratio (n) varied as 0.0, 0.1, 0.3, 0.5, 0.7, 0.9,
and 1.1. The other parameters were kept constant as shown
in Table 1 (F1-1 to F1-7). Figure 10 shows the fitted rela-
tions of lateral load with lateral displacement of the com-
posite CFST frames at various axial load ratios. Increase in
the lateral load capacity for the frames with increasing
n values from 0.1 to 0.7 and closed increasing for n values
from 0.7 to 1.1. Figure 11 reveals the maximum lateral load
versus n values. The lateral stiffness of the frame improves
with the increase of the axial load ratio, at all stages, with
no descending whatsoever. As the axial load ratio increases
8 Advanced Composites Letters
from 0 to 0.7, see Figure 12, the frame stiffness at all stages
also increases clearly. From 0.7 to 1.1, the frame stiffness at
the elastic and yield stages looks very close, while the
lateral post-yield stiffness of the frame increases slightly.
Slenderness of column
The column slenderness (l) varies as 24, 48, 72, and 96. The
other parameters were kept constant as shown in Table 1
(F2-1 to F2-4). Figure 13 shows the relationship of lateral
load–lateral displacement of the composite CFST frames at
Figure 6. Comparison between the experimental and numerical
analysis result.
Figure 7. Failure mode and maximum stresses at beam–column and column–base connections.
different slenderness values. As shown from the figure, the
structural responses of the case study frames are depended
on the level of the column slenderness. The performance of
all models is straight up to the inflection point and then the
curve behaves as nonlinear in the direction of horizontal due
to a decrease in stiffness. Moreover, as observed in Figure 14,
there is a clear trend that the rapid decrease of the maximum
load as the slenderness increased. Similarly, the lateral stiff-
ness of the frame decreases with the increase of the slender-
ness value (see Figure 15). However, the less-stiff frames
show greater lateral displacements and therefore greater duc-
tility. None of the curves is shown descending, however,
lateral displacement looks larger. Hereby the yield and ulti-
mate load are seen constant, which indicates imminent struc-
ture failure. With the increase of the column’s slenderness
due to increase in column height, the yield load and ultimate
load decrease very clearly, in case of l ¼ 24 compared to
l ¼ 48, l ¼ 72, and l ¼ 96, respectively, the ultimate load
decreases 1.77, 2.82, and 4.63 times. It is shown that the
slenderness of the column is more than 48; the overall frame
instability is due to the formation of plastification.
Linear stiffness ratio of the beam–column
To study the influence of the beam–column linear stiffness
ratio (k) on the frame structures response, the steel beam
length was varied while the other input parameters kept
Ahmed and Güneyisi 9

Figure 8. Deformed shape of a composite frame (specimen F2-3).

Figure 9. Stages of loading and corresponding stresses (specimen F2-3).

10
Figure 10. Lateral load–displacement relationships for various
axial load ratios.
Figure 11. Maximum lateral load versus n value.
Figure 12. Initial stiffness versus n value.
unchanged, to achieve the case of changing the linear stiffness
ratio of the beam–column as shown in Table 1 (F3-1 to F3-5).
The lateral load versus lateral displacement relationship is
shown in Figure 16. Figures 17 and 18 show the relationships
of the lateral load versus k values and the initial stiffness
versus k values, for the composite CFST frames, respectively.
With the increase of the linear stiffness ratio of the beam–
column, the lateral stiffness of the frame at the elastic stage
Advanced Composites Letters
Figure 13. Lateral load versus displacement relationships for
various column slenderness ratios.
Figure 14. Maximum lateral load versus l value.
Figure 15. Initial stiffness versus l value.
increases with significant differences. When the linear stiff-
ness ratio of the beam–column increased from k ¼ 1.8–6.6,
both the lateral stiffness and the maximum lateral load exhibit
approximately a constant increase. The initial stiffness
increases from 11.8 to 33.2 (180% increment) and the lateral
load increases from 585 kN to 864 kN (48% increment). The
linear stiffness ratio of the beam–column represents the con-
straint level, and furthermore, it puts an effect on the capacity
for the lateral deformation of the framed column. Thus, the
studied ratio is directly influencing the bearing capacity of the
frame.
Ahmed and Güneyisi
Figure 16. Lateral load versus displacement relationships for
various linear stiffness ratios.
Figure 17. Maximum lateral load versus k value.
Figure 18. Initial stiffness versus k value.
Steel yield strength of the beam
To study the influence of the strength of the steel beam on
the behavior of the frame, steel yield strength of 235, 290,
345, 390, and 420 MPa was used by keeping the other para-
meters constant as shown in Table 1 (F4-1 to F4-5). Figure 19
shows the relationship of lateral load–lateral displacement of
the composite CFST frames at different steel yield strengths
of the beam. With the increase of the strength of steel, the
lateral stiffness of all frames at the elastic stage remains so
close. It can be seen that the frames also can continue to
11
Figure 19. Lateral load versus displacement relationships for
different beam yield strengths.
Figure 20. Maximum lateral load versus beam fy.
Figure 21. Initial stiffness versus beam fy.
carry load and deformation. It is shown that the frames of
lower beam steel strength have greater deformation capacity
and thereby greater ductility, however, the shape of the
curves and trends of the lateral load–lateral displacement
of the frame almost remain the same. The maximum lateral
load increased from 635.8 kN to 741 kN (16.5% increment)
when the beam steel yield strength increases from 235 MPa
to 420 MPa, as shown in Figure 20. The initial stiffness also
12
Figure 22. Lateral load–displacement relationships for different
column tube yield strengths.
Figure 23. Maximum lateral load versus column fy.
Figure 24. Initial stiffness versus column fy.
rises from 14 kN/mm to 22 kN/mm (57% increment) as
observed in Figure 21.
Steel yield strength of the column
To examine the effect of the steel yield strength of tube of
the CFST columns on the lateral performance of the com-
posite frames, the following steel strengths of the tubes
were used: 235, 290, 345, 390, and 420 MPa by keeping
Advanced Composites Letters
Figure 25. Lateral load versus displacement for various concrete
strengths.
Figure 26. Maximum lateral load versus fc.
Figure 27. Initial stiffness versus fc.
the other parameters constant as shown in Table 1 (F5-1 to
F5-5). Figure 22 shows the fitted relationship of lateral
load–lateral displacement of CFST frames for different
column tube steel strengths. With the increase of the
strength of steel, the lateral stiffness of all frames at the
elastic stage remains close and increased at later stages as
the tube strength is increased; it is shown that the frames
with higher column steel strength have improved deforma-
tion capacity.
The shape of the curves and trends of lateral load–lateral
displacement of the frame almost remain the same.
The maximum lateral load increased from 624 kN to
Ahmed and Güneyisi
Figure 28. Toughness versus various studied parameters.
710.5 kN (14% increment) when tube steel strength
increases from 235 MPa to 420 MPa, as shown in Figure 23.
The initial stiffness increases when tube steel strength
increases from 235 MPa to 420 MPa. As shown in
Figure 24, it is increased from 14.3 kN/mm to 17.6 kN/
mm (23% increment). The effect is less pronounced than
what resulted when studying the beam steel strength. This
may be due to the composite action of the column.
Concrete strength
Due to the influence of the compressive strength of con-
crete core of CFST columns on the behavior of the compo-
site frame, different concrete cylinder strengths of 28, 56,
68, 88, and 108 MPa are adopted, as listed in Table 1 (F6-1
to F6-5). Figure 25 shows the relationship of the lateral
load–lateral displacement of the composite frames at dif-
ferent strengths of concrete up to ultimate load (note that
the final displacement not drawn). With the increase of the
compressive strength of concrete, the maximum lateral
load increases due to the increase in modulus of elasticity
of the concrete that increases in strength capacity of the
13
member due to the increase in stiffness. As concrete
strength increases from 28 MPa to 108 MPa, the maximum
lateral load increases from 662.9 kN to 758.9 kN (only
14.5% increment). It means a little gain of the CFST frame
strength is obtained when a huge increase in concrete
strength is made (see Figure 26). Moreover, the initial stiff-
ness increases from 14.8 kN/mm to 22.1 kN/mm (49%
increment) as given in Figure 27.
Toughness
The area under the load–displacement curve is measured as
an indicator of the toughness of the composite frame.
Figure 28 shows the toughness versus various studied para-
meters. As the axial load ratio (n) is increased from n ¼ 0 to
n ¼ 0.7, a clear increase in toughness is shown (84% incre-
ment), then a less increase from n ¼ 0.7 to n ¼ 1.1 (5%
increment). It can be indicated that as n is increased (as the
axial compression becomes larger), the composite frame
displays both strength and ductility.
As the column slenderness ratio (l) is increased, an
increase in the toughness is shown. For instance, as l is
14
increased from 24 to 96, the toughness increases by more
than 70%. Meanwhile, as the stiffness ratio of the beam–
column (k) is increased, a decrease in toughness is
observed. For instance, as k is increased from 1.8 to 6.6,
the toughness increases by more than 50%. It is indicating
that a higher beam to column linear stiffness ratio may
result in lower energy absorption; however, this may be
compromised by the increase in the axial load ratio. On the
other hand, when investigating the material strengths,
lower effects on the toughness might be obtained. As the
beam’s steel strength is increased, a small decrease in the
toughness is shown. For instance, as the steel yield strength
is increased from 235 MPa to 420 MPa, the toughness
increases by only 15%. Moreover, as the column’s tube
steel strength is increased, little change in toughness is
observed. The same is true when discussing the effect of
the strength of concrete infill. As the composite column’s
concrete strength fc is increased, no change in toughness or
a very small reduction is seen as fc is increased from 28
MPa to 108 MPa. It may be indicated that the loading
condition, stiffness, and slenderness ratios for the compo-
site frames have greater effects on the toughness, which is
an indicator of energy absorption.
Conclusions
The structural performance of the composite frames includ-
ing CFST columns connected to the steel beam under the
lateral load was studied. The FE modeling was adopted to
analyze the frames by ANSYS software. The lateral load
was applied up to the failure under the effects of various
parameters. From the study results, at an early stage, no
confinement of steel tube around the perimeter of the con-
crete core due to the Poisson’s ratio of concrete lower than
that of steel tube so that there is no full capacity for the
composite column. When the applied load increased, the
longitudinal strains increased and then transverse strains
also increased so that the lateral expansion of concrete
gradually becomes greater than that of the steel tube. The
presence of concrete inside the steel tube delays the local
buckling of the tube wall and the concrete itself and reduces
the buckling at the final stage loading. The contact element
makes the force transfer between the steel tube and the core
concrete due to the friction between the contact surface so
that the composite action looks like full interactions. It was
observed that the composite frames exhibited a similar fail-
ure mode. Plastic hinges were formed at the top and at the
bottom of the CFST columns. No concrete core crush was
observed. The lateral stiffness of the frame improved with
the increase of the axial load ratio at all stages. The lateral
stiffness and lateral load capacity of the composite frame
increased as the slenderness of the CFST column
decreased. When the linear stiffness ratio of the beam–col-
umn increased from 1.8 to 6.6, both the lateral stiffness and
the maximum lateral load exhibited approximately a con-
stant increase. Composite frames with lower steel beam
Advanced Composites Letters
strength had greater deformation capacity and thereby
greater ductility than those with higher steel strengths.
Frames with higher steel yield strength of the tube of CFST
column had limited, improved deformation capacity. Little
gain of the frame lateral strength was obtained when a huge
increase of concrete core strength of CFST column was
made. The composite frame displayed higher ductility, that
is, higher toughness when the axial load ratio and column’s
slenderness ratio were increased. While the stiffness ratio
of the beam–column was increased, the frame exhibited
lower toughness. Also, little effect on the toughness of the
frame was found as the composite column’s steel strength
and its concrete core strength were changed.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with
respect to the research, authorship, and/or publication of this
article.
Funding
The author(s) received no financial support for the research,
authorship, and/or publication of this article.
ORCID iD
Ahmed Dalaf Ahmed https://orcid.org/0000-0002-2694-9180
References
1. Shams M and Saadeghvaziri MA. State of the art of concrete-
filled steel tubular columns. Struct J 1997; 94(5): 558–571.
2. Gourley BC, Tort C, Hajjar JF, et al. A synopsis of studies of
the monotonic and cyclic behaviour of concrete-filled steel
tube beam-columns. Minneapolis: University of Minnesota,
2001.
3. Shanmugam NE and Lakshmi B. State of the art report on
steel–concrete composite columns. J Constr S teel Res 2001;
57(10): 1041–1080.
4. Nishiyama I. Summary of research on concrete-filled
structural steel tube column system carried out under the
US-Japan cooperative research program on composite
and hybrid structures. Tsukuba: Building Research Inst,
2002.
5. Schneider SP. Axially loaded concrete-filled steel tubes.
J Struct Eng 1998; 124(10): 1125–1138.
6. Hu HT, Huang CS, Wu MH, et al. Nonlinear analysis of
axially loaded concrete-filled tube columns with confinement
effect. J Struct Eng 2003; 129(10): 1322–1329.
7. Han LH, Yao GH and Tao Z. Performance of concrete-filled
thin-walled steel tubes under pure torsion. Thin-Walled Struct
2007; 45(1): 24–36.
8. Liu J and Foster SJ. Finite-element model for confined con-
crete columns. J Struct Eng 1998; 124(9): 1011–1017.
9. Ellobody E and Young B. Design and behavior of concrete-
filled cold-formed stainless steel tube columns. Eng Struct
2006; 28(5): 716–728.
Ahmed and Güneyisi
10. Han LH, Wang WD and Zhao XL. Behavior of steel beam to
concrete-filled SHS column frames: finite element model and
verifications. Eng Struct 2008; 30(6): 1647–1658.
11. Han LH, Wang WD and Zhao XL. Performance of circular
CFST column to steel beam frames under lateral cyclic load-
ing. J Constr Steel Res 2011; 67(5): 876–890.
12. Mou B, Li X, Qiao QY, et al. Seismic behavior of frame
corner joints under bidirectional cyclic loading test. Eng
Struct 2019; 196: 109–316.
13. Ding FX, Yin GA, Jiang LZ, et al. Composite frame of cir-
cular CFST column to steel-concrete composite beam under
lateral cyclic loading. Thin-Walled Struct 2018; 122:
137–146.
14. Hajjar JF, Leon RT, Gustafson MA, et al. Seismic response of
composite moment-resisting connections. II: Behavior.
J Struct Eng 1998; 124(8): 877–885.
15. Wang W, Li H and Wang J. Progressive collapse analysis
of concrete-filled steel tubular column to steel beam con-
nections using multi-scale model. Structures 2017; 9:
123–133.
16. Mou B and Bai Y. Experimental investigation on shear beha-
vior of steel beam-to-CFST column connections with irregu-
lar panel zone. Eng Struct 2018; 168: 487–504.
15
17. Mou B, Zhao F, Qiao QY, et al. Flexure behavior of beam-to-
column joints with concrete slab. Eng Struct 2019; 199(11):
109–616.
18. Muhummud T. Seismic design and behavior of composite
moment resisting frames constructed of CFT columns and
WF beams. PhD Thesis, Lehigh University, Bethlehem,
2004.
19. Herrera R. Seismic behavior of concrete filled tube column-
wide flange beam frames. Bethlehem: Department of Civil
and Environmental Engineering, Lehigh University, 2006.
20. ANSYS. ANSYS user’s manual revision 5.5. Canonsburg:
ANSYS Inc, 2016.
21. GB 50017-2003. Code for design of steel structures. Beijing:
China Planning Press.
22. AISC. Load and resistance factor design Specification for
structural steel buildings. Chicago, IL: American Institute
of Steel Construction, 1999.
23. Xu T, Xiang T, Zhao R, et al. Nonlinear finite element anal-
ysis of circular concrete filled steel tube structures. Structural
Engineering and Mechanics, 2010; 35(3): 315–333.
24. Hu HT, Huang CS, Wu MH, et al. Nonlinear analysis of
axially loaded concrete-filled tube columns with confinement
effect. J Struct Eng 2003; 129(10): 1322–1329.