Far East Research Centre www.fareastjournals.

com

Bearing capacity investigation of pin-ended composite columns with

eccentricity loading using finite element simulation

Ahmed Abdulhaq
Department of Civil Engineering, Universiti Kebangsaan Malaysia, 43600 UKM, Bangi,
Selangor, Malaysia
Email: ahmed.abdulhaq@yahoo.com

W.H. Wan Badaruzzaman2

Department of Civil Engineering, Universiti Kebangsaan Malaysia, 43600 UKM, Bangi,
Selangor, Malaysia
Email: whamidon@vlsi.eng.ukm.my

ABSTRACT

In this paper, a complete non-linear finite element model that could represent the behaviour of
the composite columns tested under axial compression in the laboratory was developed. A
numerical analyses study was presented based on experimental results for investigating the
effects of introducing column head on the structural behaviour of 10 finite element models of
fully encased composite columns with embedded I-beam. Effects of parameters such as
length of composite columns and loading condition on the ultimate load capacity have been
examined. It was observed that the whole section-loaded models shows greater load carrying
capacity compared with the core loaded models due to the effect of column head implies that
load condition has an influence on the behaviour and strength of the composite columns
moreover, it is suggested that the finite element models were able to simulate various loading
conditions, with very good accuracy.

Keywords: Composite column, Bearing capacity, Finite element simulation, Column head
Paper Type: Research Paper

INTRODUCTION
Composite systems refer to any members composed of more than one material. The parts of
these composite members are rigidly connected such that no relative movement can occur
and to make best use of their particular characteristics. Composite column is one of these
systems that comprise of steel sections with a concrete encasement or core, where is used as
load-bearing member in a composite framed structure. As for the components, reinforced
concrete is inexpensive, massive, stiff and advantageous in compressive strength to assist in
resisting the service loads while steel members are strong, lightweight and have the

34
 Far East Journal of Psychology and Business Vol. 5 No. 1 October 2011

advantages of high tensile strength and ductility in which used for the building erection and
resisting all construction loads. Additional reinforcement is placed in the concrete cover
around the steel section in order to prevent spalling under axial stress and in fire and to
improve the ductility of these columns somewhat under cyclic loading.

Fully encased columns are columns with standard steel I-beam or H-beam section embedded
within a concrete, see Figure 1. In the proposed system, the steel section is located at the
central section in the longer dimension and applying a head to the column, making it an
efficient load bearing system about the main axis. This has great advantages when the column
is subjected to combined axial loads and large major axis moments by frame action.

Longitudinal bars

I-beam Structural steel

Concrete

Stirrups

Figure 1: Typical cross-section of fully encased composite columns.

Many researchers have studied the bearing capacity behaviour of the concentrically loaded
encased composite columns and number of papers containing complete numerical and
experimental investigations has been found, but no research work has been done on encased
composite columns with column head by simulation. From current situations and problems
encountered, an idea about idealising a method to increase the load bearing capacity for the
composite column was formed regarding laboratory test results. Introducing head on top of
the composite column may have possibility of increase the load capacity of the column. This
will be the main focus of this study. Other parameter that will be varied is the column length.
Furthermore the author felt that he is in need of achieving the ability to conduct a non-linear
finite element analysis of such a structure as an important outcome of this study.

The model is to be capable of simulating numerically the full behavioural history including
peak capacities and the failure mode caused by local buckling of the steel plates and /or
crushing of the concrete. Determination of the composite column strength by varying the
column length and introducing column head on the load bearing capacity of the composite
column will be studied. The study involves a numerical modelling via non-linear finite
element approach employing commercially available software, LUSAS 14 of fully encased
composite column under axial static loading. In the present study, non-linear finite element
analysis is used to investigate the behaviour of fully encased composite columns. The

35
Far East Research Centre www.fareastjournals.com

columns that are subjected to axial loading, through the core and the whole sections require a
consideration of cross-section capacity as well as slenderness effects.

MATERIALS & METHODS

DESCRIPTION OF THE MODEL

Confirming the validity of a finite element model of fully encased composite column which
have been tested in the laboratory was the first attempt in this study. Initial verification of the
author’s LUSAS model is conducted to study the behaviour of fully encased composite
column when introducing head to the top of the column on the load carrying resistance.
Indeed, the experimental test results are published in the paper entitled “Influence of concrete
and fiber concrete on the load-carrying capacity and deformability of composite
steel-concrete columns (Szmigiera 2007)”. Columns were modelled and analyzed regarding
change in column length to study the columns behaviour under static incremental load.

Determination of a suitable finite element model for the analysis was carried out by
performing a mesh convergence study to select the mesh density. Adequate number of
elements is used in a model and results were obtained. Steel composite columns were
modelled as simply supported similar to the one adopted in the experiment by Szmigiera
(2007). Yield point of the structural steel HEA160 was 282 MPa and yield point of
longitudinal bar steel 317 MPa. Length of models which are simulated are 2500 mm. In
LUSAS, columns were modelled using ungraded Mild Steel with Young’s Modulus equal to
210 kN/mm2 and Poisson’s ratio of 0.3. Young’s Modulus and Poisson’s ratio of concrete
components are 28 kN/mm2 and 0.2 respectively.

In non linear analysis of LUSAS software, incremental-iterative solution procedure was used.
In this procedure, the total required load was applied in a number of increments. With each
increment, a linear prediction of the non-linear response is made, and subsequent iterative
corrections were performed in order to restore equilibrium by the elimination of residual
forces. The iterative corrections were referred to some form of the convergence criteria that
indicate to what extent an equilibrium state has been achieved.

Choice of elements was based on accumulated experience. The I-beam structural steel section
was modelled by using TLS6 3D thin shell element in the LUSAS element library. It has six
nodes numbered anticlockwise. Generally, this element can accommodate curved geometry
with varying thickness and anisotropic and composite material properties. The element
formulation takes account of both membrane and flexural deformations. The concrete
element is simulated by 3-D Structural Solid Continuum element (TH10), with the addition of
special cracking and crushing capabilities. It is isoperimetric solid continuum elements with
higher order models capable of modelling curved boundaries. The most important aspect of

36
 Far East Journal of Psychology and Business Vol. 5 No. 1 October 2011

this element is the treatment of non-linear material properties. The concrete is capable of
cracking (in three orthogonal directions), crushing, plastic deformation, and creep.

RESULTS

MODEL VERIFICATION
The goal of the comparison of the finite element models and the columns from Szmigiera
(2007) was to ensure that the elements, material properties and convergence criteria are
adequate to model the response of the member and make sure that the simulation process is
correct. Therefore, experimental columns tested by Szmigiera (2007) were verified. The
composite column consists of rolled steel I-beam HEA 160 entirely covered by casing of
concrete, four 12 mm diameter longitudinal ribbed bars as well as 6 mm stirrups from plain
steel were placed into the section. Length of elements was 2500 mm, while dimensions of
cross-section: 260 x 260 mm, see figure 2.

9 Ø 6 @ 250mm stirrups

HEA 160 I-beam

2500 mm

4 Ø 12 longitudinal bars

260 mm

260 mm

Figure 2: Details and elevation of the composite columns.

37
Far East Research Centre www.fareastjournals.com

Table 1: Ultimate failure load for verifying the specimens

Specimen Failure load Failure load Failure load ratio

P exp (kN) P FEM (kN) P exp / P FEM

C 1000 1039.885 0.96

W 2190 2164.674 1.01

The comparison between the experimental ultimate load for the column was presented along
with the corresponding finite element values in each case, see Table 1. The results in Figures
3 and 4 shows close agreement between the experimental and finite element results. The
elastic behaviour and ultimate bearing capacity were predicted by the finite element
modelling with sufficient accuracy. Therefore, the finite element modelling is used in future
analysis of the composite columns in the following sections.

2500 EXPERIMENT
LUSAS
2000
Force (kN)

1500

1000

500

0
0 0.25 0.5 0.75 1 1.25

Strain (%)

Figure 3: Loading through the whole composite section (W) verification result

38
 Far East Journal of Psychology and Business Vol. 5 No. 1 October 2011

EXPERIMENT
1600
LUSAS
1400
1200
Force (kN)
1000
800
600
400
200
0
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
Strain (%)

Figure 4: Loading through the core (C) verification result

EFFECT OF LENGTH & COLUMN HEAD

The first parameter is the effect of various lengths (2.5 m, 4 m, 6 m, 8 m and 10 m) for both
group of columns, Five different C columns (core loaded) and five coresponding W columns
(whole section loaded). These columns were assumed as pin-ended. The results from the
analyses are shown in Figures 5 and 6 in the form of load-displacement curves.
1200

1000

800
Force (kN)

2.5m
600 4m
6m
400
8m
200 10m

0
0 1 2 3 4 5 6 7 8 9

Displacement (mm)

Figure 5: Loading the composite columns of various lengths through the core
As it can be seen from Figures, the ultimate load capacity of columns decreased by the
increase of the columns length. Infact, the axial buckling resistance of a real column will be
less than the squash resistance because of the effects of slenderness and initial imperfections.

39
Far East Research Centre www.fareastjournals.com

2500

2000
Force (kN)

1500 2.5 m
4m
1000 6m
8m
500 10m

0
0 2 4 6 8 10
Displacement (mm)

Figure 6: Loading the composite columns of various lengths through the whole section

The second parameter was considering the effect of column head, a way of which the load is
applied through the core or the whole section for the columns having the same length.
Column head was made from fibre concrete in the laboratory, while it was simulated by steel
plate (thin shell element) 30 mm thickness for the place, where the load is applied. Such
application changes a scheme of applied compressive force, aiming at distribution of load in
the whole composite section and prevent the local crushing of concrete just under the loading
location.The verification of results show that this assumption is accurate. Figure 7 shows the
way of distrbuting the axial load through the column.

(a) Loading through the whole section W, (b) Loading through the steel core C,
by using column head protruded from column section
Figure 7: Two ways of applying load to the column

40
 Far East Journal of Psychology and Business Vol. 5 No. 1 October 2011

The second parameter is the effect of using column head (the way of applying the load
through the core or whole section). The evaluation of ultimate load capacity of columns C
and W, which have the same length, were considered by (2.5m, 4m, 6m, 8m, and 10m)
lengths and the results will compare the ultimate load for each certain length .For all the
columns with same heights and different loading conditions the ultimate load for columns W
obtained from the software exhibits higher stiffness and ultimate strength compared to the C
columns that proves the advantage of the column head and the concrete contribution, which
were able to sustain more bearing capacity. These results are shown in following figures as a
form of load-displacement curves and summarised in Table 2.

2500

2000
Force (kN)

1500

1000

500
whole section loaded
Core loaded
0
0 0.5 1 1.5 2
Strain (%)

Figure 8: Strains of the composite column 2.5 m length in various loading conditions

2000
1800
1600
1400
Force (kN)

1200
1000
800
600
400 whole section loaded
200 core loaded
0
0 2 4 6 8 10
Displacement (mm)

Figure 9: Displacement of the composite column 4 m length in various loading

conditions

41
Far East Research Centre www.fareastjournals.com

1400

1200

1000
Force (kN)

800

600

400
whole section loaded
200
core loaded
0
0 2 4 6 8 10

Displacement (mm)

Figure 10: Displacement of the composite column 6 m length in various loading

conditions

700

600

500
Force(kN)

400

300

200

100 whole section loaded

core loaded
0
0 2 4 6 8 10
Displacement (mm)

Figure 11: Displacement of the composite column 8 m length in various loading

conditions

42
 Far East Journal of Psychology and Business Vol. 5 No. 1 October 2011

300

250

200
Force (kN)

150

100

50 whole section loaded

core loaded
0
0 1 2 3 4 5 6
Displacement (mm)

Figure 12: Displacement of the composite column 10 m length in various loading

conditions

Table 2: Maximum failure load for each group of loaded columns

length Maximum failure load (kN)

%
(m) (W) Columns (C) Columns

2.5 2164.647 1039.885 51.96052751

4 1834.874 839.955 54.22274227

6 1246.209 571.2 54.16499159

8 593.4393 363.627 38.72549391

10 277.92 222.336 20

DISCUSSION
Comparison between the ultimate load bearing capacities of W and C columns shows that the
effect of column head decrease by increasing the length as it shown in the first parameter
study (length). The analysis showed also that the behaviour and characteristics of failure of
such columns are of a very complex kind and they depend on many factors. Not all of them
are considered in recommendations of actually valid standards. The increment of the ultimate
load bearing capacity for each column length due to column head is as shown in Figure 12 It
decreases by increasing the length due to the increase in column slenderness. By observing

43
Far East Research Centre www.fareastjournals.com

the columns with length 2.5 m up to 6 m, it is noticed that the effect of the column head on
the ultimate load of column is very significant. This is due to the ultimate load bearing
capacity governs by strength of material and area of the section. However, by increasing the
length of column from 6 m up to 10 m, it is observed that the effect of the column head
dropped because the ultimate load bearing capacity governed by length.

Comparison between the failure loads of composite columns and that of their respective
change in length as it is observed from the load-displacement curves in Figure 13 clearly
shows that stiffness, stability and load capacity increases due to the effect of column head in
comparison with corresponding column without column head. Observations made after the
failure load had been reached confirmed that the steel component and the concrete encased
within the steel act together even after failure and are capable of sustaining considerable
loading while maintaining a ductile behaviour.

60
Percentage of ultimate failure

54.2 % 54.1 %
51.9 %
50
38.7 %
40
loads

30
20 %
20

10

0
2.5 4 6 8 10
Column lengths

Figure 13: Percentage of ultimate failure loads due to affect of column head
in each certain length
CONCLUSION

The study was confined on the composite columns that have been simulated by using finite
element theory. The structure stability observations on the composite columns showed that
buckling phenomena is associated with long columns, while short columns will squashed
under axial compression. Szmigiera (2007) developed a technique to increase the bearing
capacity by applying a column head to resist the squash load for short length composite
columns. Based on the results obtained, column head theory was more than adequate for
medium length columns in which the bearing resistance reached to its ultimate capacity and

44
 Far East Journal of Psychology and Business Vol. 5 No. 1 October 2011

proposed 3D finite element models were capable of predicting the ultimate bearing resistance
for the composite columns with good accuracy.

REFERENCES
Ahmed Abdulhaq. 2010. The non-linear behaviour of pin-ended hybrid headed columns
under uniaxial compression: Master Thesis. Universiti Kebangsaan Malaysia.
British Standards Institution. 1994. Design of composite steel and concrete structures. Part
1.1. General Rules and Rules for Buildings. Eurocode 4. London.
Cook, R.D., Malkus, D.S. & Plesha, M.E. 1989. Concepts and Applications of Finite Element
Analysis. 3rd edition: John Wiley and Sons.
Elghazouli, A.Y. & Treadway, J. 2008. Inelastic behaviour of composite members under
combined bending and axial loading. Journal of Constructional Steel Research
64: 1008–1019.
EL-Tawil, S. & Deieleinr, G.G. 1999. Strength and ductility of concrete encased composite
columns. Journal of Structural Engineering 125(9): 1009–1019.
European Committee for Standardisation 2004. Design of composite steel-concrete
structures. Part 1.1: General rules and rules for buildings. Eurocode 4.
Brussels. Belgium.
Johnson, R.P. & Smith, D.G.E. 1980. A simple design method for composite columns. The
Structural Engineer 58A: 85-93.
Lachance, L. 1982. Ultimate strength of biaxially loaded composite sections. Journal of
Structural Division. ASCE 108: 2313-2329.
Memon, B.A. & Xiao-zu, S. 2004. Arc-length technique for nonlinear finite element analysis.
Journal of Zhejiang University SCIENCE 5(5): 618-628.
Mirza, S.A. 1989. Parametric study of composite colum strength variability. Journal of
Constructional Steel Research 14: 12137.
Munoz, P.R. 1994. Behavior of biaxially loaded concrete-encased composite columns. PhD
thesis, New Jersey Institute of Technology.
Shanmugam, N.E. & Lakshmi, B. 2001. State of the Art Report on Steel concrete Composite
Columns. Journal of Constructional Steel Research 57: 1041-1080.
Szmigiera, E. 2007. Influence of concrete and fibre concrete on the load-carrying capacity
and deformability of composite steel-concrete columns. Journal of Civil
Engineering and Management 8(1): 55-61.
Venugopal, R., Shanmugam, N.E. & Richard Liew, J.Y. 2003. Built-up columns encased in
concrete. Advances in structures proceeding international conference
ASSCCA, (vol 2).
Zienkiewicz, O.C. 1971. The finite element method in engineering science: McGraw-Hill
Ltd.

45
Far East Research Centre www.fareastjournals.com

ACKNOWLEDGEMENT

I would like to express my sincere gratitude to Prof. Ir. Dr. Hj. Wan Hamidon bin Hj. Wan
Badaruzzaman for the guidance, knowledge, and numerous hours spent helping me achieve
this work. I truly appreciate everything he has done for me as a lecturer, supervisor and as a
friend. His generosity and support will not be forgotten.

46