CHAPTER 7

FINITE ELEMENT TECHNIQUE

7.1 INTRODUCTION AND SCOPE

In this study, the finite element analysis is proposed as an alternative and

complementary method to the design of composite structures. Both strength and
ductility of CFST columns subjected to uniform axial compression loading are
considered. This design procedure only applies to the columns with simply supported
boundary conditions at the ends. A design procedure using the finite element method,
which was earlier given by chai *14', Zheng and few reports on ductility are
available so far from the review of literature.

To investigate the behaviour of the new CFST columns, expensive, time

consuming, and elaborate experimental analyses are required. If a suitable and reliable
numerical model is developed, however, a wider parametric investigation can be
performed and a reduction in time and cost can be achieved.

Typically, the behaviour of columns is studied by full-scale experimental

investigations. The results are compared to analytical calculations that estimate
ultimate strength and deflections in the columns. Finite element analysis can be used
to model the behaviour numerically to confirm these calculations, as well as to
provide a valuable supplement to the laboratory investigations, particularly in
parametric studies. Finite element analysis, as used in structural engineering,
determines the overall behaviour of a structure by dividing it into a number of simple
elements, each of which has well-defined mechanical and physical properties [l00]

This chapter contains a report on the analysis made using Finite

Element Method (FEM). Models were developed to simulate the behaviour of
concrete filled steel tubular composite columns. To create the models, commercial
finite element software ANSYS [l011 was used in this numerical study.

90
 Prior to analyzing the post-buckling behaviour of structure, a linear buckling
analysis is first performed on specimens to obtain its buckling mode shape. Following
this, the non-linear post buckling analysis is carried out to predict the load versus
lateral deformation, load versus end shortening characteristic curves and ultimate load
capacity.

Comparisons are made for load-deflection, stress-strain characteristic curve

and ultimate load. Modeling simplifications and assumptions developed during this
research are presented. The study compared ultimate load carrying capacity of the
columns from the FEM analysis with measured failure load from the experimental
investigation.

7.2 COMPUTER MODELING OF STRUCTURES

Modeling is one of the most important aspects for the finite element analysis.
Accuracy in the modeling of element type, size, geometry, material properties,
boundary conditions and loads are absolutely necessary for close numerical
idealization of the actual member. Modeling the complex behaviour of reinforced
concrete, which is both anisotropic and non-homogeneous, is a difficult challenge in
the finite element analysis of Civil Engineering structures.

7.2.1 Type of Elements

The following are the types of element used in the simulation.

(i) SOLID 65 for concrete

(ii) SOLID 45 for Steel plates
(iii) SHELL 63 for Steel tube
(iv) LINK 8 for Reinforcement

(i) SOLID65

An eight-node solid element, SOLID65, is used to model the concrete. The

solid element has eight nodes with three degrees of freedom at each node-translation
in the nodal x, y, and z directions. The element is capable of plastic deformation,

91
cracking in all directions, and crushing. The geometry and node locations for this
element type are shown in Fig. 7.1.

Fig.7.1 SOLID65 - 3D Reinforced Concrete Solid

(ii) S0L1D45

SOL1D45 is used for the 3-Dimen sional modeling of solid structures. This
element type is defined by eight nodes having three degrees of freedom (3 DOF) at
each node. The element has plasticity, creep, swelling, stress stiffening, large
detlections and large strain capability. The geometry and node locations for this
element type are shown in Fig. 7.2.

Fig.7.2 SOLID45 - 3D Plain Concrete Solid

(iii) SHELL63

SHELL63 is used to model the thin walled structures effectively. The element
type used is a 4-noded 3-dimensional quadrilateral shell element with six degrees of
freedom at each node. Elastic shell elements are used in the finite element model for

92
linear buckling analysis, whereas, plastic shell elements, which allow for plasticity
and large deflection behaviour, are used in the modeling for nonlinear buckling
analysis. The geometry and node locations for this element type are shown in Fig 7.3.

I
t

I tI
I
I

!
i

}i
!

Fig.7.3 SHELL63 - Elastic Shell

(iv) LINK 8

L1NK8 is a 3D spar element. It is a uni-axial tension-compression element

with three degrees of freedom at each node. Plasticity, creep, swelling and stress
stiffening capabilities are included. A L1NK8 element is used for the model. Two
nodes are required for this element. Each node had three degrees of freedom,
translations in the nodal x, y, and z directions. The element is also capable of plastic
deformation. The geometry and node locations for this element type are shown in
Fig.7.4.

Fig.7.4 LINK 8 - 3D Spar (or Truss)

93
7.2.2 Material properties

7.2.2.1 Concrete

Development of a model for the Peak compressive stress

behavior of concrete is a challenging fc’

task. Concrete is a quasi-brittle

material and has different behaviors in
Compres ; softe
sion -n'mg
compression and tension. Fig.7.5
shows a typical stress-strain curve for ■"V Max. comp, strain
normal weight concrete. Material
Tension
nonlinearity was used in the analysis.
For concrete the following nonlinear
Fig.7.5 Typical stress-strain curve
material properties are considered. for normal weight concrete

Following are the input data required to create the material model for concrete
in ANSYS.
• Elastic Modulus, (Ec)
• Poisson's Ratio, (v)
• Ultimate uniaxial compressive strength, (/ck)
• Ultimate uniaxial tensile strength, (/,)
• Coefficient for opened crack. (po)
• Coefficient for closed crack, (pc)

The elastic modulus and ultimate uniaxial tensile strength of concrete are
found using the following equations given in Table 7.1

Table 7.1 Properties of different grades of concrete

Grade Designation
M20 M25 M30 M35 M40
(N/mm2)
/ck 24.8 29.75 34.83 39.86 44.95
/t=0.6V/ck
2.99 3.27 3.54 3.79 4.02
Ec= 5000 V/Ck 24900 27272 29508 31567 33522

94
Where,
fck ~ Characteristic compressive (cube) strength of concrete
ft = Mean tensile strength of concrete
Ec = Elastic modulus of concrete

Poisson's ratio for concrete was assumed to be 0.2 for all the columns Damien
Kachlakev et.al [I02J conducted numerous investigations and they found out that the
coefficient for opened crack is 0.2 and for closed crack is 1. As per the ANSYS
concrete model, two coefficients, one for open cracks and other for closed ones, are
used to consider the retention of stiffness in cracked concrete.

Even though the above parameters are enough for the ANSYS non-linear
concrete model, it is better to keep a stress-strain curve of concrete as a backbone for
achieving accuracy in results. Hence it was attempted to input the stress - strain curve.

The stress-strain curve for concrete can be constructed by using the Desayi
and Krishnan equation tl041 Multi-linear kinematic behaviour is assumed for the
stress-strain relationship of concrete, which is shown in Fig. 7.6.

Fig.7.6 Simplified Compressive Uniaxial Stress-Strain Curve for Concrete

95
 It is assumed that the curve is linear up to 0.3 /ci<. Therefore, the elastic stress-
strain relation is enough for finding out the strain value.

£|=/cVEc = (0.3/c')/Ec. (7.1)

where,
/c’ i = Compressive strength of confined concrete cube.
Ec = Elastic modulus of concrete

The Ultimate strain can be found out from the following formula.

£o = 2 f'J Ec (7.2)

The total strain in the non-linear region is calculated and corresponding

stresses for the strains are found out by using the following formula.

f c (2, 3 & 4 ~
) (Ece)/(l+(£/8o)2) (7.3)

The above input values are given as material properties for concrete to define
the non-linearity.

In compression, the stress-strain curve of concrete is linearly elastic up to

about 30% of the maximum compressive strength. Above this point, the stress
increases gradually up to the maximum compressive strength, and then descends into
a softening region, and eventually crushing failure occurs at an ultimate strain £cu. In
tension, the stress-strain curve for concrete is approximately linearly elastic up to the
maximum tensile strength. After this point, the concrete cracks and the strength
decreases to zero.

ANSYS has its own non-linear material model for concrete[l051 Its reinforced
concrete model consists of a material model to predict the failure of brittle materials,
applied to a three-dimensional solid element in which reinforcing bars may be
included. The material is capable of cracking in tension and crushing in compression.
It can also undergo plastic deformation and creep. Three different uniaxial materials,
capable of tension and compression only, may be used as a smeared reinforcement,
each in any one direction. Plastic behaviour and creep can be considered in the
reinforcing bars as well. For plain cement concrete model, the reinforcing bars can be
removed.

96
7.2.2.2 Failure Criteria for Concrete

ANSYS non-linear concrete model is based on William-Wamke failure

criteria[105] As per this failure criteria, at least two strength parameters are needed to
define the failure surface of concrete. Once the failure is surpassed, concrete cracks if
any principal stresses are tensile while crushing occurs if all the principal stresses are
compressive. Tensile failure consists of a maximum tensile stress criterion. Unless
plastic deformation is taken into account, the material behaviour is linearly elastic
until failure. When the failure surface is reached, stresses in that direction have a
sudden drop to zero, provided there is no strain softening neither in compression nor
in tension. This indicates that the descending portion in stress-strain curve of concrete
is not considered in ANSYS non-linear concrete model.

'yp

Fig.7.7 3D Failure Surface for Concrete

A three-dimensional failure surface for concrete is shown in Fig 7.7. The most
significant non-zero principal stresses are in the x and y directions respectively. Three
failure surfaces are shown as the projections on the axp . ayp plane. The modes of
failure are the function of the sign of azp (principal stress in Z direction). For example,
if both axp and oyp are negative (compressive) and ozp is slightly positive (tensile),
cracking would be predicted in a direction perpendicular to ozp. However, if ozp is
zero or slightly negative, the material is assumed to crush. In a concrete element,

97
cracking occurs when the principal tensile stress in any direction lies outside the
failure surface. After cracking, the elastic modulus of concrete element is set to zero
in the direction parallel to the principal tensile stress direction. Crushing occurs when
all principal stresses are compressive and lie outside the failure surface. Subsequently,
the elastic modulus is set to zero in all directions and the element effectively
disappears.

7.2.2.3 Steel tubes and steel rods

The material model of steel tube and reinforcement rods for the finite element
models was assumed to be of linearly elastic-perfectly plastic material[29] Properties
like Young's modulus and yield stress, for steel tube and reinforcement used in this
FEM study were found out by conducting the required tests on the sample specimens.
Poisson's ratio of 0.3 was used for the steel reinforcement. Bilinear kinematic material
model is adopted and Fig. 7.8 shows the stress-strain relationship used in this study.

Fig.7.8 Stress-Strain Curve for Steel Tube and Reinforcement

A summary of material properties used for modeling all the columns are
shown in Table 7.2. These values are used for calculating the important properties
required for specifying material non-linearity.

98
 Table 7.2 Material properties

Outer Properties
Thickness diameter
Material Modulus of
(t) in 'mm' (D) in Stress
(N/mm2) elasticity
'mm' (N/mm2)
Steel tube 2 114 288.7 2.06x105

8 446.8 2.10xl05
-
Reinforcement
6 371.5 2.10xl05
-

7.2.3 Modeling the Geometric Shape

The geometrical details of the column modeled are given in Table 6.1, 6.2, 6.3
and 6.4. By taking advantage of the shape of the columns, a full scale steel tube
column in-fdled with concrete was modeled. Nodes of the steel tube shell elements
were connected to those of adjacent concrete solid elements in order to satisfy perfect
bond assumption. Fig. 7.9 illustrates the element connectivity.

Ideally, the bond strength between the concrete and steel reinforcement should
be considered. However, in this study, perfect bond between materials was assumed.
To provide the perfect bond, the link element for the steel reinforcing was connected
between nodes of each adjacent concrete solid element, so that the two materials
shared the same nodes.

Concrete solid elements

Fig.7.9 Element Connectivity

(a) Concrete Solid and Link Element
(b) Concrete Solid and Steel tube Shell Element

99
 Fig.7.10 Full scale model of CFST composite column

Partially steel
encased portion

(with out steel encasement)

Fig.7.11 Full scale model of partially steel encased RCFST column

Discretely steel
encased portion

Concrete exposed portion

(with out steel encasement)

Fig.7.12 Full scale model of discretely steel encased RCFST column

100
7.2.4 Finite Element Discretization

As an initial step, a finite element analysis requires meshing of the model. In

other words, the model is divided into a number of small elements, and after loading,
stress and strain are calculated at integration points of these small elements. An
important step in finite element modeling is the selection of the mesh density. A
convergence of results is obtained when an adequate number of elements is used in a
model. This is practically achieved when an increase in the mesh density has a
negligible effect on the results.

Therefore, in this finite element modeling a convergence study was carried out
to determine an appropriate mesh density. The finite element models dimensionally
replicated the full-scale columns. That is, CFST composite columns with the same
material properties were modeled in ANSYS with an increasing number of elements.
A number of response parameters is compared, including deflection and stress at the
mid height extreme fiber of the column.

A convergence of results is obtained when an adequate number of elements is

used in a model. If the mesh density is increased higher, then, convergence problems
arise. Based on trial solutions only, the required mesh density is selected. All the
nodes were merged with one another to provide a stiff model. Fig.7.13 to 7.17
represents FEM meshed model of columns.
Bharathlar University
Library

T — 1022

Fig.7.13 Meshed model for plain concrete column

101
 Fig.7.14 Meshed model for hollow steel tubular column

Fig.7.15 Meshed model for CFST column

Fig.7.16 Meshed model for partially steel encased RCFST column

102
 Fig.7.17 Meshed model for discretely steel encased RCFST column

7.2.5 Boundary conditions and loading

The boundary conditions for all finite element models are chosen to simulate
actual conditions in the experimental setup. Mild steel end plates were glued to both
ends of each specimen during the experiment in order to prevent the loaded ends of
the specimens from moving laterally, while offering little resistance to out-of-plane
rotation of the plate elements. This type of end support so achieved is regarded as a
close approximation to simple support condition. The finite element models are
loaded at the same locations as in the full size columns.

Loading location" Movable enclN

Ux=Uy=0

Column specimen-

Immovable end
Ux=Uy=Uz=0

Fig.7.18 Boundary conditions

103
(A) Concrete core only loaded (B) Fully loaded (C) Steel shell only loaded
Fig.7.19 Loading conditions

A thick steel plate, modeled using Solid45 elements, is added at the end
support location in order to avoid stress concentration problems. This provides a more
even stress distribution over the support area. The section is then loaded at the
movable end by prescribing suitable increments of axial load. For each incremental
step of loading, the shortening at the end can be obtained.

7.3 ANALYSIS

Initially linear analysis is carried out. After getting confirmation of the results
in the linear range, nonlinear analysis is performed.

7.3.1 Linear Analysis

Results of the proposed finite element model are verified against results
obtained experimentally from tests. The behaviour of the model is investigated
throughout the loading history from the first application of the load to ultimate load.
The ultimate loads are obtained from this study. The load, lateral deflection and axial
shortening results are also obtained from the FEM. During the initial stages of
loading, all the curves show a linear relationship between the load and the end
shortening. As the end-shortening increases and exceeds the critical buckling loading,

104
 AN SYS Graphics

Fig.7.21 Stress contours of CFST for fully loaded column

Fig.7.22 Element solution for hollow steel tubular column

Tables 7.3, to 7.6 show the comparison of the results obtained using the
proposed finite element model with those obtained from the experimental tests. It is
evident from the results that the numerical analysis can predict both the failure load
and the displacement up to service load of the new system with acceptable accuracy.

107
 Table 7.3 Experimental and numerical results for confinement effects

Ultimate Ultimate Axial Ultimate Ultimate Axial

Column Lateral Shortening Axial Strain Stress
SI. No. Description of Specimens Deflection (mm) (mm) lO'3 (N/rnrn2)
ID

<

<1
b

IO
to
Q.

Ui
Urn
Ui
Um

S
u
S
X
X

Ui
Q.
Lii

ft.X
X
u

<
4.21 5.33 8.79

-
o
00
c
Z IP
;
CFST fully loaded A2 3.75 5.12 5.69 53.885
A3 4.31 5.71 8.55

<N
4.10 2.74 4.30
CFST concrete core only
B2 3.80 4.20 2.94 2.54 4.10 63.130
loaded
B3

108
3.50 2.99 4.52

U
1.93 2.87 3.45
ZD

m
CFST steel shell only loaded 2.10 1.60 2.97 2.45 4.20 252.940
O
OO

C3 1.70 2.71

Oh
<N
o
OO
2.30 4.10

o
OO

OO
in
Plain concrete loaded P2 2.70 2.61 3.90 6.949
P3 2.35 2.95 4.90
<N
r"

GO
OO

3.80 12.10

*n
Hollow steel tube loaded S2 3.60 3.90 8.00 8.02 11.50 188.550
o
o

OO
m
4.10 8.35
 Table 7.4 Experimental and numerical results for slenderness effects

Ultimate Axial Ultimate Ultimate Axial

Ultimate Lateral
Description of Column Shortening Axial Strain Stress
SI. No. Deflection (mm) 10'3 (N/mm 2)
Specimens ID (mm)

w
b

to

<1
<
K>

UJ
UJ
0.

%
X
X

a.

X
X

UJ
Sim
u.

b.
Ui

uj
u
S
on

<
10.47 7.21
CFST for

-
<
<N
9.20 10.15 6.35 6.15 1.30 39.880
L/D=12 on

<
m
10.68 7.35

fS
s
oc
in
4.95 6.40

CN

CQ
CFST for L/D=9 5.56 4.70 6.45 5.13 5.80 44.500

00

pq
m

in

to
6.79 6.80

109
00 o

Q
4.21 5.12

to
O
CFST for L/D=6 4.31 3.75 5.34 5.85 7.90 53.880

U
fO
4.12 5.71 9.15

5
3.90 3.51 14.30
CFST for L/D=3 D2 3.21 3.20 3.75
ore 3.15 19.80 64.830
o
0\
O)

Q
to
3.20
 Table 7.5 Experimental and numerical results for concretes of different grades

Ultimate
Ultimate Lateral Ultimate Axial Ultimate Axial
Axial Strain Stress (N/mm 2)
Description of Column Deflection (mm) Shortening (mm) 10'3
SI. No.
Specimens ID

14

K>

O
&

K>
0L

LU
X
LU

u
*
u
u

Cl.
S

X
X
X

cu
u.
Q.

<
9.20 7.21 13.00
0Z9

-
<
(N
CFST for M20 9.13 9.00 6.55 12.68 48.990
A3 9.00 7.23 13.78

(N
*
o
o

s
6.50 5.34

o
in

q
o
OO
CFST for M25 B2 6.80 5.12 10.50 53.880

o
o
o

CQ
CO
5.71

110
6.40

U
6.00 5.10 9.70
00
o
in
OO

zo

o
00

CO
o
00
in
CFST for M30 5.20 4.90 9.00
©
cS
r~~

CO
O
5.90 4.90
o
o

5
4.90 4.70
CFST for M35 D2 4.80 4.60 4.40 4.10 6.70 63.680
D3 4.85 4.75 7.30
1

s
j 3.68 4.40 6.00

m
<N
w
CFST for M40 3.55 3.50 4.60 4.32 5.90 63.680

m
CO
3.80 4.20 5.70
 Table 7.6 Experimental and numerical results for fully, partially, discretely RCFST, CFST and RC columns

Ultimate Lateral Ultimate Axial Ultimate Axial Ultimate Axial

Description of Deflection (mm) Shortening (mm) Strain 10~3 Stress (N/mm 2)
SI. No. Column I D
Specimens

<3
b

K5
<4

VO

u
Cu
X
u>
X
X

ui
s
Ui
ui
0.

a.

w
X
H
s

fiu
lim
1
1
5.00

K,
6.55 5.75
Fully steel encased of Zd

-
o
o
oo
o

•/•i
4.99 6.35 6.40 5.60
RCFST
P3 5.20 6.65 5.95

(N
on

O'
0.99 1.45
Partially steel encased of on

o
o

<N
00

a
RCFST (middle two 1.07 1.05 1.60 25.250
third)
on

CO

O'
1.15
on 1.60

o
o
uo
O

Discretely steel encased 660 on

o
o

CO
00
uo

©
R2 0.95 24.940
of RCFST on
1

q
o
R3 1.20
l

yn
oo
CO
6.50 6.40
Fully steel encased of
S2 4.93 4.99 6.66 6.60 6.80 48.860
CFST
S3 5.56 6.79 5.80
q
o

2.7 5.30
q
o

DO
Conventional RC T2 2.65 2.7 0.90 5.00 12.060
T3 2.65 0.95 5.50
7.3.2 Nonlinear Analysis

In nonlinear analysis, the total load applied to a finite element model is

divided into a series of load increments called load steps. At the completion of each
incremental solution, the stiffness matrix of the model is adjusted to reflect nonlinear
changes in structural stiffness before proceeding to the next load increment. The
ANSYS programme (ANSYS 1998) uses Newton-Raphson equilibrium iterations for
updating the model stiffness. Newton-Raphson equilibrium iterations provide
convergence at the end of each load increment within tolerance limits.

A force convergence criterion with a tolerance limit of 5% is adopted for

avoiding the divergence problem. Equilibrium iterations to be performed are relaxed
up to 100. Failure loads of each column are obtained and are presented in Table 7.7,
7.8, 7.9 and 7.10.

7.4 RESULTS AND DISCUSSION

This section compares the results from the ANSYS finite element analysis
with the experimental data for the full-size columns. The following comparisons are
made: load-lateral deflection plots at mid height, load-axial shortening at the ends and
loads at failure. The data from the finite element analysis are collected at the same
locations as the load tests for the full-size columns. The following results are obtained
from ANSYS for all the tested specimens.

• Lateral deflection contours at failure load

• Axial shortening

• Failure load

Deflections are found out for various load values. The development of cracks
is captured at various load intervals. Since the columns were covered with the steel
tubes, the crack patterns were not noticed during tests. The load-lateral deflection and
load-axial shortening are shown for selected specimens and are given in Figs.7.23 to
Figs.7.60 for various parameters.

112
 The failure loads for CFST columns are obtained from ANSYS and compared
with the experimental data. The ratios between the failure load from experiment
(Pexp) and ANSYS (Pfem) are found out for all the columns. All the results are
presented in the Tables 7.7, 7.8, 7.9 and 7.10. The graphical comparison of failure
load for different parameters (confinement, effects of slenderness and concrete of
different grades for CFST, and also fully steel encased, partially steel encased and
discrete steel encased for RCFST) is presented in Figs. 7.61 to 7.68 respectively.
From the results, it is found that the variation between the experimental and ANSYS
failure loads was around 15%. This variation is due to the stiffer models created in
ANSYS. Perfect bond is assumed at the steel concrete interface of CFST columns,
which is not the case in practice. Moreover, the moduli of elasticity of the loading
plates at top and at bottom support were not determined experimentally.
Honeycombing during concreting might also be attributed to these effects.

Table 7.7 Experimental and numerical failure load of confinement effects

Average Failure
Description of Load (kN) Load Ratio Percentage
SI. No.
Specimens (Pexp/Pfem) Difference
Pexp Pfem

1 CFST fully loaded 602.00 623.00 0.97 3.37

CFST concrete core
2 613.33 625.00 0.98 1.87
only loaded
CFST steel shell
3 191.33 197.00 0.97 2.88
only loaded
Hollow steel tube
4 154.00 148.00 1.04 -4.05
loaded
Plain cement
5 113.46 118.00 0.96 3.85
concrete loaded

113
 Table 7.8 Experimental and numerical failure load of slenderness effects

Average Failure
Description of Load (kN) Load Ratio Percentage
SI. No. (Pexp/Pfem)
Specimens Pexp Pfem
Difference

1 CFST for L/D=3 663.33 670.00 0.99 1.00

2 CFST for L/D=6 596.67 623.00 0.96 4.23

3 CFST for L/D=9 562.67 580.00 0.97 2.99

4 CFST for L/D=12 456.33 452.50 1.01 -0.85

5 RC for L/D=3 143.00 158.00 0.91 9.49

6 RC for L/D=6 123.00 133.00 0.92 7.52

7 RC for L/D=9 107.00 126.00 0.85 15.08

8 RC for L/D= 12 96.00 102.00 0.94 5.88

Table7.9 Experimental and numerical failure load of different concrete grades

Average Failure
Description of Load (kN) Load Ratio Percentage
SI. No. (Pexp/Pfem)
Specimens Pexp Pfem
Difference

1 CFST for M20 485.00 505.00 0.96 3.96

2 CFST for M25 602.00 620.00 0.97 2.90

3 CFST for M30 647.33 658.00 0.98 1.62

4 CFST for M35 689.00 712.00 0.97 3.23

5 CFST for M40 723.86 738.00 0.98 1.92

6 RC for M20 103.63 102.00 1.02 -1.60

7 RC for M25 123.00 125.00 0.98 1.60

8 RC for M30 154.87 148.00 1.05 -4.64

9 RC for M35 169.90 175.00 0.97 2.91

10 RC for M40 178.50 185.00 0.96 3.51

114
Table 7.10 Experimental and numerical failure load of fully, partially and
discretely RCFST, CFST and RC columns
Average Failure
Description of Load (kN) Load Ratio Percentage
SI No. (Pexp/Pfem)
Specimens Pexp Pfem
Difference

Fully steel encased of

1 612.00 622.00 0.98 1.61
RCFST
Partially steel encased
2 of RCFST (middle 565.33 582.00 0.97 2.86
two third)
Discretely steel
3 255.00 235.00 1.09 -8.51
encased of RCFST
Fully steel encased of
4 267.33 256.00 1.04 -4.43
CFST

5 Conventional RC 107.00 118.00 0.91 9.32

115
 88
LOAD IN kN.

8
£ 03
8

100 100

0 0
0 1 2 3 4 5 0 1 2 3 4 5
LATERAL DEFORMATION IN mm. LATERAL DEFORMATION IN mm.

Fi2. 7.23 P-A of CFST for fullv loaded Fig. 7.24 P-A of CFST for core only loaded

250 ------------------------------------------------------------------------------------------------------ 140 -------------------------------------------------------------------------------------------------------

0.5 1.0 1.5 1.0 1.5 2.0

LATERAL DEFORMATION IN mm. LATERAL DEFORMATION IN mm.

Fig.7.25 P-a of CFST for steel only loaded Fig. 7.26 P-A of PCC loaded
.
LOAD IN kN

0 ---------- T- - - - - - - - - - T- - - - - - - - - - - - - 1- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 1- - - - - - - - - T- - - - - - - - - - - - - - T- - - - - - - - - - - - - - - - - T- - - - - - - - - - 0 --------------------t------------------------------------------ r-------------------- --------------------- .---------------------

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 0 1 2 3 4 5 6
LATERAL DEFORMATION IN mm. AXIAL SHORTENING IN mm.

Fig. 7.27 P-A of hollow steel loaded Fig. 7.28 Load-Axial shortening of CFST
for fully loaded

700
200 ------------------------------------------------

-♦— C2

—*—C3

•- FEM

1.0 1.5 2.0

AXIAL SHORTENING IN mm. 1.0 1.5 2.0
AXIAL SHORTENING IN mm.

Fig. 7.29 Load-Axial shortening of CFST tor Fig. 7.30 Load-Axial shortening of CFST
core only loaded for steel only loaded

116
 8

8
8

.
.

LOAD IN kN
LOAD IN kN

8
§

* S1

8
8

—•—S2

S
—a— S3

.. . FEM

3 4 5 6
1.0 1.5 2.0
AXIAL SHORTENING IN mm.
AXAL SHORTENING IN mm.

Fig.7.31 Load-Axial shortening of PCC loaded Fig. 7.32 Load-Axial shortening of hollow
steel tubular column loaded
.
LOAD IN kN

23456789
LATERAL DEFORMATION IN mm.

Fig. 7.33 P-A of CFST for L7D = 12 Fig.7.34 P-A of CFST for L/D = 9

600

500
,

400
LOAD IN kN

• /
300
o

200 • # y l

•- FEM ...... FEM

100

0 it------- T- - - - - - - - - - - - - - - - - - T- - - - - - - - - - - T- - - - - - - - - - - - - - - - - T- - - - - - - - - - T-------------- »---------.--------------T--------

2 3 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5

LATERAL DEFORMATION IN mm. LATERAL DEFORMATION IN mm.

Fig. 7.35 P-A of CFST for L/D = 6 Fig.7.36 P-A of CFST for L/D = 3
LOAD IN kN

0 1 2 3 4 5 6 7
2 3 4 5 AXIAL SHORTENING IN mm.
A VIAI QMORTFMIMn IN mm

Fig. 7.37 Load-Axial Shortening of CFST for L/D = 12 Fig.7.38 Load-Axial shortening of CFST for L7D = 9
-4
 0 300
< —D1
—•—C2 O
-1 200 pi —•—D2
C3
•- FEM 100 • FEM

12 3 4 1 2 3
AXIAL SHORTENING IN mm. AXIAL SHORTENING IN mm.

Fig. 7.39 Load-Axial shortening of CFST for L/D = 6 Fig. 7.40 Load-Axial Shortening of CFST for LVD = 3

• B1
—*— B2
—*— B3
•- FEM

1 2 3 4 5 6 7 8
LATERAL DEFORMATION IN mm. LATERAL DEFORMATION IN mm.

Fig.7.41 P-dof CFST for M20 Fig. 7.42 P-A of CFST for M25

700 700

600 600

500 500 D1

2 400 400 —*— D2

o 300 D3
< 300

200 FEM

100

1.0 2.0 3.0 4.0 5.0 6.0

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
LATERAL DEFORMATION IN mm. LATERAL DEFORMATION IN mm.

Fig. 7.43 P-4 of CFST for M30 Fig. 7.44 P-4 of CFST for M35

800 600
700
600
.

.
LOAD IN kN

LOAD IN kN

500 • E1
400 —♦— E2
300 —*— E3
200 • FEM

100

0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 1 2 3 4 5 6
LATERAL DEFORMATION IN mm. AXIAL SHORTENING IN mm.

Fig.7.45 P-A.of CFST for M40 Fig. 7.46 Load-Axial Shortening of CFST for M20

11 8
 .
LOAD IN kN

o *---------------------------------------------------------------------
0 1 2 3 4 5
AXIAL SHORTENING IN mm. AXIAL SHORTENING IN mm.

Fig.7.47 Load-Axial Shortening of CFST for M25 Fig. 7.48 Load-Axial Shortening of CFST for M30
LOAD IN kN

AXIAL SHORTENING IN mm. AXIAL SHORTENING IN mm.

Fig. 7.49 Load-Axial Shortening of CFST for M35 Fig. 7.50 Load-Axial Shortening of CFST for M40
■'4

8
cn
8
cn
8
LOAD IN kN
-c*.

8
co
8
r\o

11 2 3 4 5
LATERAL DEFORMATION IN mm LATERAL DEFORMATDN IN mm.

Fig. 7.51 P-A of fully steel encased RCFST Fig. 7.52 P-Aof fully steel encased CFST
O
S
8
CM
8 CVJ

LOAD IN kN
LOAD IN kN

8
8

o -------------------------------------------------------
0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5
LATERAL DEFORMATION IN mm. LATERAL DEFORMATION IN mm.
Fig. 7.53 P-Aof partially steel encased RCFST Fig. 7.54 P-Aof discretely steel encased RCFST

l 19
LOAD IN kN.

LOAD IN kN.
0 ----------------------------------------T---------------------------------------- T----------------------------------------

0.0 0.5 1.0 1.5 2.0 2.5 3.0

LATERAL DEFORMATION IN mm.
AXIAL SHORTENING IN mm.

Fig. 7.55 P-Aof conventional RC Fig. 7.56 Load-Axial Shortening of

fully steel encased RCFST
LOAD IN kN.
LOAD IN kN.

R1

-R2

-R3

FEM

1
AXIAL SHORTENING IN mm. AXIAL SHORTENING IN mm.

Fig. 7.57 Load-Axial Shortening of Fig. 7.58 Load-Axial Shortening of

fully steel encased CFST Partially steel encased RCFST
LOAD IN kN.

LOAD IN kN.

AXIAL SHORTENING IN mm. AXIAL SHORTENING IN mm.

Fig. 7.59 Load-Axial Shortening of Fig 7.60 Load-Axial Shortening of

discretely steel encased RCFST Conventional RC

120
 Steel only
197
loaded □ FEM

Core only
625
loaded

Fully loaded 623

0 200 400 600 800

Load in kN

Fig.7.61 Failure load of CFST column for different loading

conditions using ANSYS

Load in kN

Fig.7.62 Failure load for various materials using ANSYS

121
 FEM

□ Plain concrete

□ Hollow steel

□ Increased failure load of CFST

Fig. 7.63 Increased failure load (kN) of CFST column using ANSYS
Load in kN.

Slenderness Ratio

Fig.7.64 Failure load of CFST column for different slenderness

ratio using ANSYS

122
 200
180
160
140
Load in kN.

120
100
80
60
40
20
0
L/D=3 L7D=6 L7D=9 L7D=12
Slenderness Ratio

Fig.7.65 Failure load of RC column for different slenderness

ratio using ANSYS

Concrete Grades

Fig.7.66 Failure load of CFST column for concretes of different

grade using ANSYS

123
 300

□ FEM

z 200
_c
-a
c3
c
100

0
■ ft II
M20 M25 M30 M35 M40
Concrete Grades

Fig.7.67 Failure load of RC column for concretes of different

grade using ANSYS

0 100 200 300 400 500 600 700 800

Load in kN
Fig.7.68 Failure load of fully, partially and discrete RCFST column,
fully CFST column and RC column using ANSYS

124
7.5 COMPUTATION RESOURCES

In this thesis work, Pentium IV processors with 2 GB RAM and a speed of

3 GHz was used to perform the analysis. The time taken for modeling and analyzing
each CFST column was around two hours. Each problem required around 21 MB for
database memory.

7.6 KEY FINDINGS

From the finite element analysis the following conclusions are drawn.

• Numerical simulation using ANSYS on a series of CFST columns was

performed. For concrete Multi linear kinematic material model was used
whereas Bilinear kinematic gave excellent predictions for steel tubes.

• Results of the numerical simulations were compared with the experimental

findings. Apparently, good agreement has been obtained from the comparison
showing that the proposed numerical simulation method is applicable for
analyzing the similar structures.

• Verification and comparison of material models for steel tube and in-filled
concrete column test make it possible to predict the failure load, load- lateral
deflection, load-axial shortening and stress-strain response up to service load
with higher confidence.

125