Journal of Constructional Steel Research 65 (2009) 1846–1854

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

journal homepage: www.elsevier.com/locate/jcsr

Confinement effect of stiffened and unstiffened concrete-filled stainless steel

tubular stub columns
Mohamed Dabaon, Saher El-Khoriby, Mahmoud El-Boghdadi, Mostafa Fahmi Hassanein ∗
Department of Structural Engineering, Tanta University, Tanta, Egypt

article info abstract

Article history: This paper presents a comparative study between stiffened and unstiffened concrete-filled stainless steel
Received 7 June 2008 hollow tubular stub columns using the austenitic stainless steel grade EN 1.4301 (304). Finite element
Accepted 11 April 2009 analysis of concrete-filled stainless steel unstiffened tubular stub columns is constructed herein based on
the confined concrete model recently available in the literature. It is then compared with the experimental
Keywords: results of concrete-filled stainless steel stiffened tubular stub columns. The stiffened stainless steel tubular
Cold-formed
sections were fabricated by welding four lipped angles or two lipped channels at the lips. The longitudinal
Experimental investigation
Finite element investigation
stiffener of the column plate was formed to avoid shrinkage of the concrete and to act as a continuous
Stainless steel tubes connector between the concrete core and the stainless steel tube. The behavior of the columns was
Slender plates investigated using two different nominal concrete cubic strengths of 30 and 60 MPa. The overall depth-to-
Stiffeners width ratios (aspect ratio) varied from 1.0 to 1.8. The depth-to-plate thickness ratio of the tube sections
Square and rectangular hollow-sections varied from 60 to 90. The stiffened and unstiffened concrete-filled stainless steel tube specimens were
Confinement subjected to uniform axial compression over the concrete and stainless steel tube to force the entire
Compressive strength section to undergo the same deformations by blocking action. The ABAQUS 6.6 program, as a finite element
Structural design package, is used in the current work. The results of the comparative study showed that the stainless steel
tubes in stiffened concrete-filled columns offered a high average of increase in the confinement of the
concrete core than that of the unstiffened concrete-filled columns.
© 2009 Elsevier Ltd. All rights reserved.

1. Introduction about physical, mechanical and resistance properties among

A composite column refers to any compression member in
which a steel element acts compositely with a concrete element,
so that both elements resist compressive force. There is a wide
variety of composite column types of varying cross-section, but
the most commonly used and studied are encased I-section and
concrete-filled steel tubes. In contrast to the encased composite
column, the concrete-filled column has the advantage that it
does not need any formwork or reinforcement. The concrete-
filled column offers several advantages related to the structural
behavior over pure steel, reinforced concrete or encased I-section
column. It can be said that a concrete-filled column delivers the
economies of a concrete column with the speed of construction and
the constructability of a steel column which results in significant
economies in the overall structure of a building project [1].
On the other hand, the use of stainless steel in structural
engineering has been limited in conventional constructions due
to the high initial material costs, limited structural design
guidance, restricted section availability and lack of knowledge
∗ Corresponding author. Tel.: +20 128898494.
E-mail address: mostafa.fahmi@yahoo.com (M.F. Hassanein).
0143-974X/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jcsr.2009.04.012
designers, practice engineers and architects. In recent years, the
introduction and the revision of the design codes have assisted
in the increased use of the material in conventional structures.
The better awareness of the additional benefits of the stainless
steel such as aesthetic appearance, durability, hygienic aspects
and cleanliness, ease of fabrication, high fire resistance, ductility
and impact resistance and reuse and recycling, made the material
suited to use in construction [2–4].
The behavior of stainless steel sections is different from that
of carbon steel sections. Stainless steel sections have a rounded
stress–strain curve with no yield plateau and low proportional
limit stress compared to carbon steel sections. Stainless steel
design rules have been based on assumed analogies with carbon
steel behavior. Consequently, stainless steel designs based on these
assumptions lead to overly conservative cross-sections. In addition
to the differences in the basic material properties, other differences
cannot be ignored such as the nature of the stress–strain curve,
and the material’s response for cold-working and high elevated
temperature. Because of the nonlinear stress–strain behavior, the
design of stainless steel compression members is based on the
tangent modulus theory, for column buckling. However, there
are limited test data on concrete-filled unstiffened stainless steel
tubular columns.
 M. Dabaon et al. / Journal of Constructional Steel Research 65 (2009) 1846–1854 1847

Young and Ellobody [5] investigated concrete-filled cold-

formed high-strength stainless steel tubular columns experimen-
tally. In this study, concentrically loaded rectangular hollow-
section columns were tested. The depth-to-plate thickness ratio
of the tube sections varied from 25.7 for compact sections to 55.8
for relatively slender sections. Generally the local buckling fail-
ure mode of the high-strength stainless steel tubes was observed
for specimens with relatively slender sections. A concrete crush-
ing failure mode together with local buckling of the high-strength
stainless steel tubes for specimens with compact sections was also
observed.
On the other hand, several methods were used for column
stiffening. The first method is to use longitudinal stiffeners. The Fig. 1. Equivalent uniaxial stress–strain curve for confined concrete.
second method is to fill the cross-section of the column with
concrete. The third is to use the previous two methods together.
However, one theoretical study was found in the literature
on concrete-filled high-strength stainless steel stiffened tubular
columns. In this investigation, the nonlinear behavior of concrete-
filled high-strength stainless steel stiffened slender square and
rectangular hollow-section columns was investigated [6]. The
stiffened slender tubes had overall depth-to-plate thickness ratios
in the range 60–160. The concrete strengths covered normal and
high-strength concrete. The investigation focused on short axially
loaded columns. In this theoretical investigation, intermediate
stiffeners of U-shape were proposed. A nonlinear finite element
model was developed to study the behavior of the concrete-filled
stiffened tubular columns. A parametric study was conducted to Fig. 2. Definition of symbols for concrete-filled stainless steel stiffened tubular stub
columns.
investigate the effects of cross-section geometry and concrete
strength on the behavior and strength of the columns. The results
remarkable confinement for the concrete core. Therefore, the
of the concrete-filled stiffened tubular columns were compared
concrete can be modeled as a confined concrete model [9], as
with the results of the companion concrete-filled unstiffened
shown in Fig. 1. The strength enhancement of confined concrete
tubular columns. It was shown that the concrete-filled stiffened
is empirically related to the maximum average lateral confining
slender tubular columns offer a considerable increase in the
pressure (σlat ). The lateral confining pressure (σlat ) imposed by
column strength and ductility over the concrete-filled unstiffened
the stainless steel depends on the depth-to-thickness ratio. The
slender tubular columns.
value of the lateral confining pressure can be obtained by using
In this paper, a theoretical series of models to investigate
the empirical equations given in [9]. However, the value of the
the behavior and strength of concrete-filled normal-strength
lateral confining pressure of the stainless steel tube is expected to
unstiffened stainless steel columns as well as experimental series
increase for stainless steel tubes stiffened longitudinally, as shown
of tests to investigate the behavior and strength of concrete-filled in Fig. 2. This means that the confinement of concrete will increase
hollow-section stiffened normal-strength stainless steel columns and affect the strength of concrete-filled hollow-section stainless
is reported and compared. Each stiffener was made of two single steel columns. This idea encouraged the authors to experimentally
folded stiffeners welded together by k-welds. The principal aim investigate this type of stainless steel column [7,8].
was to improve the current knowledge of the mechanical behavior
of concrete-filled stiffened stainless steel columns, leading to a
2.2. Experimental testing
more efficient use of concrete with higher compressive strength.
A series of tests was conducted on square and rectangular hollow-
The test program of concrete-filled stainless steel stiffened
sections by using two different in-filled concrete strengths with no
tubular stub columns was presented in [7,8] using stainless steel
use of discrete mechanical shear connectors to improve the bond
grade 1.4301 (304). The stiffened stainless steel tubular stub
in the stainless steel interface or additional reinforcement besides
sections were fabricated by welding four lipped angles or two
the stainless steel tubes. The dimensions of the stainless steel tubes
lipped channels at the lips. Therefore, the stiffeners were formed
were chosen so that they include only slender sections [7,8].
at the mid-depth of the sections, as shown in Fig. 2. Table 1
summarizes the dimensions of the specimens, while the test
2. Problem statement and background strengths (PTest ) are given in Table 5. The cross-section dimensions
D and B are outside measurements. The depth of the stiffeners was
2.1. Problem statement fixed to 30 mm for all test specimens. Results from five slender
stainless steel stiffened columns and ten concrete-filled stainless
Strength enhancement in excess of uniaxial strength and the steel stiffened columns have been presented. The test specimens
deformation improvement of concrete can be observed when had proof stress and tensile strength of 285 and 742 MPa,
concrete is subjected to a triaxial compressive stress state. A respectively. The complete stress–strain curve obtained from the
common example of triaxially loaded concrete could be seen tensile coupon tests for the stainless steel tubes is shown in Fig. 3.
in unstiffened concrete-filled stainless steel tubular columns. The square and rectangular hollow-section columns of constant
Concrete-filled hollow-section stainless steel columns with a high thickness were concentrically loaded. The dimensions of the steel
value of depth-to-thickness ratio provide inadequate confinement tubes were chosen so that they include only slender sections.
for the concrete core due to the local buckling of the stainless steel The average overall depth-to-width (D/B) ratios varied from 1.0
tubes. In contrast, concrete-filled hollow-section stainless steel to 1.8 while the average overall depth-to-thickness (D/t) ratios
columns with a small value of depth-to-thickness ratio provide varied from 60 to 90. The behavior of the columns was investigated
1848 M. Dabaon et al. / Journal of Constructional Steel Research 65 (2009) 1846–1854

Table 1
Dimensions of stiffened concrete-filled stainless steel columns.
Group Specimen Depth D (mm) Width B (mm) Thickness t (mm) D/t Length L (mm) L/D Area of stainless steel (mm2 ) Area of concrete (mm2 )

SHS1 SHS1C30 120 120 2 60 360 3 1351 12 985

SHS1C60 120 120 2 60 360 1351 12 985
SHS2 SHS2C30 160 160 2 80 480 3 1671 23 865
SHS2C60 160 160 2 80 480 1671 23 865
RHS1 RHS1C30 140 80 2 70 420 3 1061 10 096
RHS1C60 140 80 2 70 420 1061 10 096
RHS2 RHS2C30 170 120 2 85 510 3 1341 19 016
RHS2C60 170 120 2 85 510 1341 19 016
RHS3 RHS3C30 180 100 2 90 540 3 1301 16 656
RHS3C60 180 100 2 90 540 1301 16 656

concrete strength could be identified from the label. For example,

the label ‘‘SHS1C30N’’ defines the specimen with a square hollow-
section that belonged to test series SHS1, and the letter ‘‘C’’
indicates the concrete strength followed by its value in MPa
(30 MPa). The suffix letter ‘‘N’’ denotes unstiffened specimens.

3.2. Finite element type and mesh

Fig. 3. Stress–strain curve of stainless steel.
using two different nominal concrete cubic strengths of 30 and
60 MPa. The column strengths, load–axial strain relationships and
failure modes of the columns have been reported. The tests showed
that to increase the capacity of slender stainless steel stiffened
tubular stub columns, in-filled concrete may be used. However, the
test strengths (PTest ) increased with the increase of the nominal
strength of the in-filled concrete. Thus, increasing the nominal
compressive strength of the in-filled concrete leads to a smaller
column size which accordingly increases the amount of usable
floor space in the structure. Once local buckling had occurred by
the time of reaching the test strength, the steel tube was not able
to provide confinement to the concrete. The column capacity, at
this stage, was governed by the local buckling failure mode. The
failure was achieved by the local buckling of the stainless tube and
the crushing of concrete. The stiffeners contributed largely to the
test strength (PTest ) of columns even when the stiffener rigidities
were small because the local buckling of longitudinal stiffeners was
prevented by the concrete.
The test strength (PTest ) for each test specimen was compared
to the design rules specified by ENV 1994-1-1 [10] and ASCE [11].
The results of the experimental study showed that the design rules
specified in European specifications as well as the American one
are highly conservative for cold-formed stainless steel square and
rectangular concrete-filled slender stainless steel hollow-section
stub columns.
3. Current finite element model
3.1. General
The finite element program ABAQUS [12] was used to
simulate the unstiffened concrete-filled slender stainless steel
stub columns. The model used the measured geometry and
the nonlinear material properties of the stiffened concrete-filled
slender stainless steel stub columns which were experimentally
tested [7,8]. The analysis is called a load–displacement nonlinear
analysis. From this analysis, the ultimate loads and failure modes
are determined.
The unstiffened concrete-filled stainless steel tube columns
were labeled such that the shape of stainless steel tube and
The S4R shell element was used to model the slender stainless
steel stiffened tubular columns. The S4R element has six degrees
of freedom per node and provides an accurate solution to most
applications. The mesh providing adequate accuracy and minimum
computational time in modeling the stainless steel stiffened
columns was chosen by selecting approximate global size equal to
25 mm. For the concrete core, a fine mesh of three-dimensional
eight-node solid elements, so-called C3D8R, was used.
To simulate the bond between the stainless steel tube and the
concrete core, a surface-based interaction with a contact pressure-
overclosure model in the normal direction, and a Coulomb friction
model in the directions tangential to the surface, are used. In order
to construct contact between two surfaces, the slave and master
surfaces must be chosen successfully. Generally, if a smaller surface
contacts a larger surface, the best option is to choose the smaller
surface as the slave surface. If the distinction cannot be made, the
master surface should be chosen as the surface of the stiffer body
or as the surface with the coarser mesh if the two surfaces are on
structures with comparable stiffness. The stiffness of the structure
and not just the material should be considered when choosing the
master and slave surfaces. Herein, a thin sheet of stainless steel
may be less stiff than a larger block of concrete core even though
the stainless steel material has a higher stiffness than the concrete
material. Therefore, the stainless steel surface is chosen as the slave
surface whereas the concrete core surface is chosen as the master
surface.
3.3. Boundary conditions and load application
The concrete-filled stiffened stainless steel columns considered
herein had fixed ends. Only a displacement at the loaded end in
the direction of the applied load was allowed. The other nodes
were free to translate and rotate in any directions. A distributed
load was applied at the top of the upper cover plate. The load was
applied in increments using the modified RIKS method available in
the ABAQUS library. The load was applied as static uniform loads
at the loaded end using the displacement control. The nonlinear
geometry parameter (*NLGEOM) was included to deal with the
large displacement analysis. The load application for RSH2C30N is
presented in Fig. 4.
3.4. Material modeling
3.4.1. Stainless steel material
In a study by Rasmussen et al. [13] describing the development
of numerical models for analysing stainless steel plates in
 M. Dabaon et al. / Journal of Constructional Steel Research 65 (2009) 1846–1854
Fig. 4. Boundary conditions and load application on RHS2C30N.
compression, it was indicated that anisotropy may not be
important for numerical analyses involving monotonic loading.
Consequently, the anisotropic nature of the stainless steel material
was not included in the current modeling.
The material behavior provided by ABAQUS allows for a multi-
linear stress–strain curve to be used. The first part of the multi-
linear curve represents the elastic part up to the proportional
limit stress with measured Young’s modulus E◦ , and Poisson’s
ratio was taken as 0.3. Since the analysis of post-buckling involves
large inelastic strains, the nominal static stress–strain curve was
converted to a true stress and logarithmic plastic true strain curve.
pl
The true stress σtrue and plastic true strain εtrue , as required by
ABAQUS, were calculated using Eqs. (1) and (2):
σtrue = σ (1 + ε) (1)
pl σtrue
ε true = ln(1 + ε) − . (2)
E◦
3.4.2. Concrete material
The concrete-filled stainless steel tubular columns of small
depth-to-thickness ratios provide considerably high confinement
for the concrete. In this case, an equivalent uniaxial stress–strain
relationship for the confined concrete should be used. On the
other hand, high depth-to-thickness ratios of concrete-filled
stainless steel tubular columns provide inadequate confinement
for the concrete, therefore a uniaxial stress–strain relationship for
unconfined concrete should be used. Mander et al. [14] defined
the limiting depth-to-thickness (D/t) ratio between confined and
unconfined concrete to be equal to 29.2. However, this limiting
depth-to-thickness (D/t) ratio is recommended for unstiffened
concrete-filled tubular columns.
Poisson’s ratio (υc ) in the elastic part of concrete under uniaxial
compression stress ranges from 0.15 to 0.22, with a representative
value of 0.19 to 0.2 according to ASCE [15]. In this numerical
modeling, Poisson’s ratio (υc ) of concrete is taken as 0.2.
Fig. 1 presents the equivalent uniaxial stress–strain curve for
confined concrete as well as the unconfined stress–strain concrete
curve. Three parts of the curve have to be identified in the case
of confined concrete. The value of εc is usually around the range
0.002–0.003. A representative value suggested by ACI Committee
318 [16] and used in the analysis is 0.003. The confined concrete
strength (fcc ) and the corresponding strain (εcc ) may be calculated
from Eqs. (3) and (4):
fcc = fco + kσlat (3)

σlat
 posite columns. Therefore, the thickness was varied between the
εcc = εc 1 + 5k . (4)
fco
Because the concrete in the concrete-filled tube columns is usually
subjected to triaxial compressive stresses, the failure of concrete
1849
is dominated by the compressive failure surface expanding with
increasing hydrostatic pressure. Hence, a linear Drucker–Prager
yield criterion is used to model the yield surface of concrete.
The first part of the curve is assumed to be an elastic part up to
the proportional limit, which is taken as 0.5fcc . The initial modulus
of elasticity (Ecc ) is highly correlated to its compressive strength
and can be calculated with reasonable accuracy from the empirical
equation of ACI Committee 318 [16] as follows:
Ecc = 4700 fcc .
p
(5)
The second part of the curve is the nonlinear portion starting
from the proportional limit stress (0.5fcc ) to the confined
concrete strength (fcc ). The stress–strain relationship proposed by
Saenz [17] has been widely adopted as the uniaxial stress–strain
curve for concrete and it has the following form:
Ecc ε
fc =    2  3 (6)
1 + (R + RE − 2) εε − (2R − 1) ε
εcc
+R ε
εcc
cc
where
RE (Rσ − 1) 1
R= − (7)
(Rε − 1) 2 Rε
Ecc εcc
RE = . (8)
fcc
Rσ = Rε = 4 may be used, as recommended by Hu and
Schnobrich [18].
In the analysis, Eq. (6) is taken as the equivalent uniaxial
stress–strain curve for concrete when the concrete strain (ε ) is less
than εcc , as can be seen in Fig. 1. When ε > εcc , a linear descending
line (the third part of the curve) is used to model the softening
behavior of concrete. If k3 is defined as the material degradation
parameter, the descending line is assumed to be terminated at
the point where fc = rk3 fcc and ε = 11εcc . To account for the
effect of different concrete strengths, the degradation parameter
(k3 ) should be multiplied by an additional reduction factor (r) as
introduced in [6]. The value of r is taken as 1.0 for concrete with
cubic strength (fcu ) equal to 30 MPa. The value of r is taken as
0.5 for concrete with fcu greater than or equal to 100 MPa. Linear
interpolation is used to determine the value of r for concrete cubic
strength between 30 and 100 MPa.
Generally, the parameters σlat and k3 should be provided in
order to completely define the equivalent uniaxial stress–strain
relation. These two parameters apparently depend on the width-
to-thickness ratio (D/t or B/t), cross-sectional shape, and stiffening
means. Consequently, their appropriate values were determined
by matching the numerical results with the experimental results
via a parametric study, as given in [9].
3.5. Verification of previous concrete-filled steel columns
Before a comparative study could be carried out, it was nec-
essary to prove that the established finite element model is
capable of simulating the structural behavior of unstiffened
concrete-filled stainless steel tubular stub columns. Accordingly,
four short concrete-filled steel tubular columns tested by Schnei-
der [19] were modeled herein using the ABAQUS computer
program. The columns were concentrically loaded in compression
to failure. The tests were conducted to investigate the effect of the
thickness of the steel tube on the ultimate strength of the com-
four columns while the length was fixed to 609.6 mm. The columns
were of square cross-section, 127 × 127 mm. The compressive
strength of the in-filled concrete was different in the four columns.
The columns are labeled, herein, by the same labels as used in the
1850 M. Dabaon et al. / Journal of Constructional Steel Research 65 (2009) 1846–1854

Fig. 5. Load–axial strain for columns tested by Schneider [19].

Table 2
Details of columns tested by Schneider [19].
Specimen B (mm) t (mm) L (mm) Fy (MPa) Fc (MPa) σlat (MPa) k3

S5 127 7.47 609.6 347 23.8 8.0 0.9

S4 127 5.67 609.6 312 23.8 0.9 0.7
S3 127 4.55 609.6 322 23.8 1.0 0.8
S2 127 4.34 609.6 357 26.0 0.0 0.6

Table 3
Finite element results of columns tested by Schneider [19].
Specimen PTest (kN) PFE (kN) PFE /PTest

S5 2069 2029 0.98

S4 1202 1214 1.01
S3 1113 1074 0.97
S2 1095 1093 1.00

original paper [19]. The full details of the four columns are pre-
sented in Table 2 whereas the comparisons of the finite element
results with the experimental results are shown in Table 3. The Fig. 6. Definition of symbols for concrete-filled stainless steel unstiffened tubular
load–axial strain relationships for the four short columns are plot- stub columns.
ted in Fig. 5. It can be seen that a good agreement was achieved
between the experimental results and the numerical modeling 4. Results and evaluation
results.
The main objective of this research was to investigate the ef-
fect of concrete infilling in stainless steel stiffened tubular stub
3.6. Current concrete-filled unstiffened stainless steel columns
column. However, the columns in this investigation were stiffened
by two methods of stiffening. These two methods were using lon-
In order to make the comparison between stiffened and
gitudinal stiffeners in addition to in-filled concrete. The main vari-
unstiffened concrete-filled tubular stub columns, finite element
ables were the cross-section shape of the stainless steel tube, the
models were constructed to simulate the unstiffened concrete-
type of the stainless steel tube (stiffened or unstiffened) and the
filled columns, as shown in Fig. 6. These finite element models
concrete strength. The test strengths (PTest ) of the concrete-filled
were based on the confined concrete model recently available in
stainless steel stiffened tubular columns of the SHS and RHS are
the literature [6,9]. The lateral confining pressure (σlat ) imposed
shown in Table 5. Three comparisons are made in this section. The
by the stainless steel was found to be zero for all the models, as the
first comparison is made with the design specifications [10,11].
depth-to-thickness ratios for the whole columns are bigger than The second is made between the stiffened concrete-filled and the
29.2 [9]. The model neglected the effect of residual stresses because unstiffened concrete-filled test specimens. Finally, the confine-
its effect on the column capacity was found by Ellobody [20] to be ment of the stainless steel tubes on the concrete core is discussed.
small. The full details of the unstiffened concrete-filled stainless
steel columns are presented in Table 4. The measured stress–strain 4.1. Comparison with design specifications
curve obtained from the tensile coupon tests for the stainless steel
tubes was used in this numerical modeling as well as the same The strengths (PTest and PFE ) for each stiffened and unstiffened
concrete cubic strengths; see [7,8]. specimen, respectively, were compared to the design rules
 M. Dabaon et al. / Journal of Constructional Steel Research 65 (2009) 1846–1854 1851

Table 4
Details of unstiffened concrete-filled stainless steel columns.
Group Specimen Depth D Width Thickness T D/t Length L L/D Area of stainless steel Area of concrete r [6] k3 [9]
(mm) B (mm) (mm) (mm) (mm2 ) (mm2 )

SHS1 SHS1C30N 120 120 2 60 360 3 930 13 448 0.97 0.43

SHS1C60N 120 120 2 60 360 930 13 448 0.77 0.43
SHS2 SHS2C30N 160 160 2 80 480 3 1250 24 328 0.97 0.4
SHS2C60N 160 160 2 80 480 1250 24 328 0.77 0.4
RHS1 RHS1C30N 140 80 2 70 420 3 850 10 328 0.97 0.4
RHS1C60N 140 80 2 70 420 850 10 328 0.77 0.4
RHS2 RHS2C30N 170 120 2 85 510 3 1130 19 248 0.97 0.4
RHS2C60N 170 120 2 85 510 1130 19 248 0.77 0.4
RHS3 RHS3C30N 180 100 2 90 540 3 1090 16 888 0.97 0.4
RHS3C60N 180 100 2 90 540 1090 16 888 0.77 0.4

Table 5
Comparison of test strengths with unfactored design strengths according to ENV 1994-1-1 [10] and the ASCE Standard [11] for stiffened concrete-filled columns.
Group Stiffened specimen PTest (kN) Aeff ,EC 3 (mm2 ) PEC 4 (kN) PTest /PEC 4 Aeff ,ASCE (mm2 ) PASCE (kN) PTest /PASCE

SHS1 SHS1C30 870 1275 671 1.30 1346 691 1.26

SHS1C60 1090 1275 910 1.20 1346 930 1.17
SHS2 SHS2C30 1422 1429 972 1.46 1647 1034 1.37
SHS2C60 1706 1429 1412 1.21 1647 1474 1.16
RHS1 RHS1C30 875 953 511 1.71 1053 539 1.62
RHS1C60 928 953 697 1.33 1053 725 1.28
RHS2 RHS2C30 1225 1014 739 1.66 1187 788 1.55
RHS2C60 1444 1014 1089 1.33 1187 1139 1.27
RHS3 RHS3C30 1006 1013 683 1.47 1184 732 1.38
RHS3C60 1488 1013 990 1.50 1184 1039 1.43

Table 6
Comparison of finite element strengths with unfactored design strengths according to ENV 1994-1-1 [10] and the ASCE Standard [11] for unstiffened concrete-filled columns.
Group Unstiffened specimen PFE (kN) Aeff ,EC 3 (mm2 ) PEC 4 (kN) PFE /PEC 4 Aeff ,ASCE (mm2 ) PASCE (kN) PFE /PASCE

SHS1 SHS1C30N 516 580 484 1.07 694 516 1.00

SHS1C60N 864 580 731 1.18 694 764 1.13
SHS2 SHS2C30N 839 604 748 1.12 737 786 1.07
SHS2C60N 1531 604 1196 1.28 737 1234 1.24
RHS1 RHS1C30N 470 561 404 1.16 660 433 1.09
RHS1C60N 761 561 595 1.28 660 623 1.22
RHS2 RHS2C30N 739 594 625 1.18 719 660 1.12
RHS2C60N 1070 594 979 1.09 719 1015 1.05
RHS3 RHS3C30N 680 586 567 1.20 704 600 1.13
RHS3C60N 948 586 878 1.08 704 911 1.04

specified by ENV 1994-1-1 [10] using the effective area approach. forming this research. A reduction factor ρ should be taken as
The comparisons of strengths with design strengths (PEC 4 ) for follows
the test and finite element series for stiffened and unstiffened For cold-formed or welded internal elements, which are used
concrete-filled stainless steel stub columns, respectively, are given with flat portions, as shown in Figs. 2 and 6:
in Tables 5 and 6. Generally, the predicted strength was calculated
0.772 0.125
using Eq. (6.30) of ENV 1994-1-1 [10]. In this equation, the plastic ρ= − but ≤ 1. (10)
resistance of the concrete-filled stainless steel stiffened tubular λp λ2p
columns should be calculated by adding the plastic resistance For welded outstand elements, which are used with longitudinal
of the tube material and the concrete core, in the case of no stiffeners, as shown in Fig. 2:
reinforcement being used, as in the current test and finite element
specimens, as follows: 1 0.242
ρ= − but ≤ 1 (11)
PEC 4 = Aa fyd + 0.85Ac fcd (9)
λp λ2p

where PEC 4 is the design value of the plastic resistance of the
composite section to compressive normal force; Aa is the cross-
sectional area of the structural steel section; Ac is the cross-
sectional area of concrete; fyd is the design value of the yield
strength of structural steel; and fcd is the design value of the
cylinder compressive strength of concrete.
However, Table 5.2 in EN 1993-1-4 [22] was used to classify
the cross-section type of the test specimens. By applying the
limitations in Table 5.2, the test specimens were all of Class 4.
As is well known, the effective widths may be used in order to
reduce the resistance of the cross-section due to the effect of
local buckling. For this reason, clause 5.2.3 was in use during
the calculations of the effective widths in Class 4 cross-sections
where λp is the element slenderness, defined as,
b/t
λp = √ (12)
28.4ε kσ
where t is the relevant thickness.
kσ is the buckling factor corresponding to the stress ratio
ψ and boundary conditions from Table 4.1 or Table 4.2 in EN
1993-1-5 [9] as appropriate: it is 4.0 for flat portions and 0.43
for longitudinal stiffeners. These values mean that the case of
simple support is considered. The assumptions about simple
support when determining kσ are normally unfavorable. The
assumption is therefore normally a simplification intended for
manual calculation.
1852 M. Dabaon et al. / Journal of Constructional Steel Research 65 (2009) 1846–1854
b is the relevant width as follows:
b= flat element width while b = c for outstand flanges; see
Fig. 2.
ε is the material factor defined in Table 5.2 in EN 1993-1-4 [4].
The comparison of the test and finite element strengths (PTest
and PFE ) was also made by using the predicted value according to
the ASCE Standard [11]. The comparisons of strengths with design
strengths (PASCE ) according to the ASCE Standard [11] for the test
and finite element series for stiffened and unstiffened concrete-
filled stainless steel stub columns, respectively, are also given in
Tables 5 and 6. In general, clause 3.4 in the ASCE Standard [11] was
used to classify the cross-section type of the test specimens. The
resistance of a stainless steel cross-section subject to compression
with a resultant acting through the centroid of the effective section
shall be calculated as follows:
PASCE = Ae Fn + 0.85Ac fc
where Ae is the effective area calculated at stress Fn .
A reduction factor ρ should be taken as follows:
For uniformly compressed stiffened elements, this is used with
flat portions:
1 − 0.22/λ
ρ=
λ 61.9 kN. From Tables 5 and 6, it can be seen that the strengths (PTest

1.052  w  f
  
if λ = √ ≥ 0.673.
k t E◦
The uniform compressed unstiffened elements, which are used
with longitudinal stiffeners, use the same equations as for stiffened
elements with the exception that k = 0.5.
Fn is the flexural buckling stress for doubly symmetric sections,
closed cross-sections, and any other sections which are not
subjected to torsional or torsional–flexural buckling. The flexural
buckling stress, Fn , is determined as follows:
π 2 Et
Fn = ≤ Fy
(KL/r )2
E◦ F y
Et =
n−1
Fy + 0.002nE◦ Fn /Fy
E◦ is the initial elastic modulus. n is the Ramberg–Osgood
parameter, which is the strain-hardening exponent that defines
the degree of roundness of the curve (n = 4.67). Ac is the cross-
sectional area of concrete. fc = 0.8fcu is the design value of the
cylinder compressive strength of concrete.
The determination of flexural buckling stress and then the
design of axial strength of concentrically loaded cold-formed
stainless steel compression members, according to the ASCE
Standard [11], requires an iterative process.
From Table 5, it can be seen that the predicted values for the
stiffened concrete-filled stainless steel stub columns according
to ENV 1994-1-1 [10] and the ASCE Standard [11] are highly
conservative. However, using the ASCE Standard [11] provides
relatively closer design strengths than using ENV 1994-1-1 [10]
for stiffened and unstiffened concrete-filled stub columns. On the
other hand, using ENV 1994-1-1 [10] and the ASCE Standard [11]
provides less conservative strength values in the case of the
unstiffened concrete-filled stainless steel stub columns, as can
be seen in Table 6. This is a result of using stiffeners in the
stainless steel cross-section. These stiffeners had improved the
local buckling resistance of the stainless steel tube. At the same
time, the stiffeners contributed largely to the test strength (PTest )
of columns even when stiffeners rigidities were small because
the local buckling of longitudinal stiffeners is prevented by the
concrete core.
4.2. Comparison between stiffened and unstiffened concrete-filled
stub columns
From Tables 5 and 6, which tabulate the results of both the
stiffened and unstiffened concrete-filled stub columns, it can be
seen that the column strength of the stiffened concrete-filled
stainless steel tubular stub columns are considerably higher than
those of the unstiffened concrete-filled stainless steel tubular stub
columns. In general, the average value of the column strength
of the stiffened concrete-filled columns divided by that of the
unstiffened concrete-filled columns is equal to 1.43. This means
that the stiffened concrete-filled columns offered an average of
43% increase in the column strength over that of the unstiffened
concrete-filled columns. However, this increase in the column
capacity was accompanied by a relatively small increase in the
cross-sectional area of the stainless steel tube. Herein, the whole
(13) cross-sectional area of the stiffened stainless steel tubes divided
by that of the unstiffened stainless steel tubes is equal to 1.30.
4.3. Confinement effect
Two different nominal concrete strengths were used in tested
stub columns. The mean cubic concrete strengths were 34.8 and
(14)
and PFE ) increased with the increase of the nominal strength of the
(15) in-filled concrete for both stiffened and unstiffened concrete-filled
stub columns. The contributions of both the stainless steel tube (Ps )
and the concrete core (Pc ,s and Ps,u for stiffened and unstiffened
columns, respectively) to the overall strength of the concrete-filled
columns were calculated as follows:
Ps = Aeff ,EC 3 × σ0.2 (18)
Pc ,s = PTest − Ps (19)
Pc ,u = PFE − Ps . (20)
From Table 7, it can be seen that the contribution of the
(16) stainless steel tube to the overall strength of the stiffened
concrete-filled columns is generally higher than its contribution
for the unstiffened concrete-filled columns. This is because the
(17)
local buckling resistance of the slender stainless steel tube was
improved remarkably as a result of using longitudinal stiffeners.
In contrast, the contribution of the concrete core to the overall
strength of the stiffened concrete-filled columns was found to be
generally less than its contribution for the unstiffened concrete-
filled columns. However, the average values of the concrete core
strength of the stiffened concrete-filled columns divided by that
of the unstiffened concrete-filled columns is equal to 1.31. This
means that the stainless steel tubes in stiffened concrete-filled
columns offered an average of 31% increase in the confinement
of the concrete core over that of the unstiffened concrete-filled
columns. Subsequently, the strength enhancement of the confined
concrete was increased by using stiffened stainless steel tubes.
The resistances of the stiffened stainless steel tubes were
calculated without taking the effect of initial imperfection and
welding residual stress into account. These residual stresses are
induced by the longitudinal welding of the stainless steel tubes
on two or four sides. These residual stresses were investigated
by Dabaon et al. [21]. In this investigation, a comparative
experimental study between stiffened and unstiffened stainless
steel tubular stub columns was conducted. It was concluded in [21]
that the residual stresses should be taken into consideration
due to their pronounced effect. Accordingly, the strengths of the
experimentally tested stiffened stainless steel tubes were used
herein instead of the calculated values from Eq. (18), as tabulated
in Table 8. In general, the contribution of the stainless steel tubes in
the overall strength of the concrete-filled columns was decreased
 M. Dabaon et al. / Journal of Constructional Steel Research 65 (2009) 1846–1854 1853

Table 7
Strength of confined concrete in stiffened and unstiffened concrete-filled columns.
Group Stiffened PTest Ps [22] Ps [22]/PTest Pc ,s [22] Pc ,s [22]/PTest Unstiffened PFE Ps [22] Ps [22]/PTest Pc ,u [22] Pc ,u [22]/PTest Pc ,s /Pc ,u
specimen (kN) (kN) (kN) Specimen (kN) (kN) (kN)

SHS1 SHS1C30 870 363 0.42 507 0.58 SHS1C30N 516 165 0.32 351 0.68 1.27
SHS1C60 1090 363 0.33 727 0.67 SHS1C60N 864 165 0.19 699 0.81 0.95
SHS2 SHS2C30 1422 407 0.29 1015 0.71 SHS2C30N 839 172 0.21 667 0.79 1.61
SHS2C60 1706 407 0.24 1299 0.76 SHS2C60N 1531 172 0.11 1359 0.89 1.00
RHS1 RHS1C30 875 272 0.31 603 0.69 RHS1C30N 470 160 0.34 310 0.66 1.99
RHS1C60 928 272 0.29 656 0.71 RHS1C60N 761 160 0.21 601 0.79 1.11
RHS2 RHS2C30 1225 289 0.24 936 0.76 RHS2C30N 739 169 0.23 570 0.77 1.69
RHS2C60 1444 289 0.20 1155 0.80 RHS2C60N 1070 169 0.16 901 0.84 1.31
RHS3 RHS3C30 1006 289 0.29 717 0.71 RHS3C30N 680 167 0.25 513 0.75 1.47
RHS3C60 1488 289 0.19 1199 0.81 RHS3C60N 948 167 0.18 781 0.82 1.58

Table 8
Strength of confined concrete in stiffened concrete-filled columns.
Group Stiffened PTest [8] (kN) Ps [21] Ps [21]/PTest Pc ,s [21] Pc ,s [21]/PTest Ps [22] Ps [22]/PTest Pc ,s [22] Pc ,s [22]/PTest
specimen (kN) (kN) (kN) (kN)

SHS1 SHS1C30 870 425 0.49 445 0.51 363 0.42 507 0.58
SHS1C60 1090 425 0.39 665 0.61 363 0.33 727 0.67
SHS2 SHS2C30 1422 350 0.25 1072 0.75 407 0.29 1015 0.71
SHS2C60 1706 350 0.21 1356 0.79 407 0.24 1299 0.76
RHS1 RHS1C30 875 258 0.29 617 0.71 272 0.31 603 0.69
RHS1C60 928 258 0.28 670 0.72 272 0.29 656 0.71
RHS2 RHS2C30 1225 263 0.21 962 0.79 289 0.24 936 0.76
RHS2C60 1444 263 0.18 1181 0.82 289 0.20 1155 0.80
RHS3 RHS3C30 1006 254 0.25 753 0.75 289 0.29 717 0.71
RHS3C60 1488 254 0.17 1234 0.83 289 0.19 1199 0.81

Table 9
Stresses in the concrete-filled columns’ components.
Group Stiffened Ps [21] Aeff [22] Ps /Aeff Ac ,s fcc a Unstiffened Ps [22] Aeff [22] Ps /Aeff Ac ,u fcc a
specimen (kN) (mm2 ) (MPa) (mm2 ) (MPa) specimen (kN) (mm2 ) (MPa) (mm2 ) (MPa)

SHS1 SHS1C30 425 1275 333 12 985 42.9 SHS1C30N 165 580 285 13 448 32.6
SHS1C60 425 1275 333 12 985 64.0 SHS1C60N 165 580 285 13 448 65.0
SHS2 SHS2C30 350 1429 245 23 865 56.1 SHS2C30N 172 604 285 24 328 34.3
SHS2C60 350 1429 245 23 865 71.0 SHS2C60N 172 604 285 24 328 69.9
RHS1 RHS1C30 258 953 271 10 096 76.4 RHS1C30N 160 561 285 10 328 37.5
RHS1C60 258 953 271 10 096 83.0 RHS1C60N 160 561 285 10 328 72.8
RHS2 RHS2C30 263 1014 259 19 016 63.3 RHS2C30N 169 594 285 19 248 37.0
RHS2C60 263 1014 259 19 016 77.6 RHS2C60N 169 594 285 19 248 58.5
RHS3 RHS3C30 254 1013 251 16 656 56.5 RHS3C30N 167 586 285 16 888 38.0
RHS3C60 254 1013 251 16 656 92.6 RHS3C60N 167 586 285 16 888 57.8
a
fcc : Confined cubic concrete strength.

due to the effect of initial imperfection and residual stresses. Then,
the contribution of the concrete core in the overall strength of
the stiffened concrete-filled columns was found to be increased.
However, the contribution of the concrete core to the overall
strength of the stiffened concrete-filled columns was found also to
be generally less than its contribution for the unstiffened concrete-
filled columns. However, the average values of the concrete core
strength of the stiffened concrete-filled columns divided by that
of the unstiffened concrete-filled columns is equal to 1.33. This
means that the stainless steel tubes in stiffened concrete-filled
columns offered an average of 33% increase in the confinement
of the concrete core over that of the unstiffened concrete-filled
columns. Subsequently, the strength enhancement of the confined
concrete was increased by using stiffened stainless steel tubes.
Table 9 provides the stresses in the stainless steel tubes and the
confined concrete cores for the stiffened and unstiffened concrete-
filled columns. The stress in the confined concrete core (fcc ) is
calculated with respect to the cubic strength.
5. Conclusions
This paper is mainly concerned with the confinement eval-
uation of concrete in stiffened concrete-filled stainless steel
hollow tubular stub columns. Therefore, a comparative study
between stiffened and unstiffened concrete-filled stainless steel
hollow tubular stub columns was conducted. Austenitic stainless
steel grade EN 1.4301 (304) was used. A finite element nonlinear
displacement analysis of unstiffened concrete-filled stainless steel
tubular stub columns was constructed herein based on the con-
fined concrete model recently available in literature and then com-
pared with the experimental results of stiffened concrete-filled
stainless steel tubular stub columns. The stiffened stainless steel
tubular sections were fabricated by welding four lipped angles
or two lipped channels at the lips. Therefore, the stiffeners were
formed at the mid-depth and mid-width of square sections and at
the mid-depth of rectangular sections. The longitudinal stiffener
of the column plate was formed to avoid shrinkage of the concrete
and to act as a continuous connector between the concrete core and
the stainless steel section. The behavior of the columns was inves-
tigated using two different nominal concrete cubic strengths of 30
and 60 MPa. A series of tests and models was performed to investi-
gate the effects of cross-section shape and concrete strength on the
behavior and strength of stiffened and unstiffened concrete-filled
stainless steel tubular stub columns. The measured average over-
all depth-to-width ratios (aspect ratio) varied from 1.0 to 1.8. The
depth-to-plate thickness ratio of the tube sections varied from 60
to 90. Different lengths of columns were selected to fix the length-
to-depth ratio to a constant value of 3. The stiffened and unstiff-
ened concrete-filled stainless steel tube specimens were subjected
1854 M. Dabaon et al. / Journal of Constructional Steel Research 65 (2009) 1846–1854
Fig. 7. Modes of failure of concrete-filled stainless steel square stub columns.
Fig. 8. Modes of failure of concrete-filled stainless steel rectangular stub columns.
to uniform axial compression over the concrete and stainless steel
tube to force the entire section to undergo the same deformations
by blocking action. The ABAQUS 6.6 program, as a finite element
package, was used in the current work.
The results of the comparative study showed that the stiffened
concrete-filled columns offered an average of 43% increase in
the column strength over that of the unstiffened concrete-filled
columns. However, this increase in the overall strength of the
stiffened concrete-filled columns is accompanied by a relatively
small increase in the cross-sectional area of the stainless steel tube.
Also, the comparative study showed that the stainless steel tubes
in stiffened concrete-filled columns offered an average of 33%
increase in the confinement of the concrete core over that of the
unstiffened concrete-filled columns (taking into account the effect
of initial imperfection and welding residual stresses in the stiffened
stainless steel tubes). It was found that the design rules specified
in the ASCE Standard are generally conservative for unstiffened
concrete-filled stainless steel square and rectangular stub columns,
while they are highly conservative for stiffened concrete-filled
stainless steel square and rectangular stub columns. However, the
mode of failure can be drawn as in Figs. 7 and 8.
Acknowledgements
The authors would like to acknowledge the Faculty of
Engineering, Tanta University, for its support. They are grateful
to the Concrete and Heavy Structures Laboratory staff, Faculty of
Engineering, Tanta University, for their great help and technical
support. Further thanks are due to the Materials Properties
Laboratory staff, The Housing and Building Research Center, Cairo,
for their help. Special thanks to Prof. Milan Veljkovic at LTU, Lulea,
Sweden, for his advice and expertise concerning the thesis. The
authors would like to thank Wylliam Husson at LTU, Lulea, Sweden,
for advice and valuable discussions with the fourth author.
References
[1] Johansson M. Composite action and confinement effects in tubular
steel–concrete columns. Ph.D. thesis. Goteborg (Sweden): Chalmers Uni-
versity of Technology; 2002.
[2] Gardner L. A new approach to structural stainless steel design. Ph.D. thesis.
London: Imperial College of Science, Technology and Medicine; 2002.
[3] Gardner L. The use of stainless steel in structures. Progress in Structural
Engineering and Materials 2005;7(2):45–55.
[4] Mirambell E, Real E. On the calculation of deflections in structural stainless
steel beams: An experimental and numerical investigation. Journal of
Constructional Steel Research 2000;54:109–33.
[5] Young B, Ellobody E. Experimental investigation of concrete-filled cold-
formed high strength stainless steel tube columns. Journal of Constructional
Steel Research 2006;62:484–92.
[6] Ellobody E. Nonlinear behavior of concrete-filled stainless steel stiffened
slender tube columns. Thin-Walled Structures 2007;45:259–73.
[7] Hassanein M. Behavior of composite columns: Concrete-filled hollow-section
stiffened steel columns. Ph.D. thesis. Tanta (Egypt): Tanta University; 2009.
[8] Dabaon M, El-Boghdadi M, Hassanein M. Experimental investigation on
concrete-filled stainless steel stiffened tubular stub columns. Engineering
Structures 2009;31:300–7.
[9] Hu HT, Huang CS, Wu MH, Wu YM. Nonlinear analysis of axially loaded
concrete-filled tube columns with confinement effect. Journal of Structural
Engineering, ASCE 2003;129(10):1322–9.
[10] ENV 1994-1-1. Eurocode 4: Design of composite steel and concrete structures
— Part 1-1: General rules and rules of building. CEN; 2004.
[11] ASCE, Specification for the design of cold-formed stainless steel structural
members (SEI/ASCE 8-02), American Society of Civil Engineers, 2002.
[12] ABAQUS Standard User’s Manual. Hibbitt, Karlsson and Sorensen, Inc. Vol. 1, 2
and 3, Version 6.6, USA, 2007.
[13] Rasmussen KJR, Burns T, Bezkorovainy P, Bambach MR. Numerical modelling
of stainless steel plates in compression. Journal of Constructional Steel
Research 2003;59:1345–62.
[14] Mander JB, Priestley MJN, Park R. Theoretical stress–strain model for confined
concrete. Journal of structural Engineering, ASCE 1988;114(8):1804–26.
[15] ASCE, ASCE Task Committee on Concrete and Masonry Structure, State of the
Art Report on Finite Element Analysis of Reinforced Concrete, ASCE, New York,
1982.
[16] ACI, Building code requirements for structural concrete and commentary, ACI
318-95, American Concrete Institute, Detroit, USA, 1999.
[17] Saenz LP. Discussion of ‘equation for the stress–strain curve of concrete’ by
P. Desayi and S. Krishnan. Journal of American Concrete Institute 1964;61:
1229–35.
[18] Hu HT, Schnobrich WC. Constitutive modeling of concrete by using
nonassociated plasticity. Journal of Materials in Civil Engineering 1989;1(4):
199–216.
[19] Schneider SP. Axially loaded concrete-filled steel tubes. Journal of the
Structural Engineering, ASCE 1998;124(10):1125–38.
[20] Ellobody E. Buckling analysis of high strength stainless steel stiffened and
unstiffened slender hollow section columns. Journal of Constructional Steel
Research 2007;63:145–55.
[21] Dabaon M, El-Boghdadi M, Hassanein M. A comparative experimental study
between stiffened and unstiffened stainless steel hollow tubular stub columns.
Thin-Walled Structures 2009;47:73–81.
[22] EN 1993-1-4. Eurocode 3: Design of steel structures - Part 1-4: General rules-
Supplementary rules for stainless steel. CEN; 2006.