Engineering Structures 135 (2017) 68–80

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Concrete filled double circular steel tube (CFDCST) stub columns

Talha Ekmekyapar a,⇑, Baraa J.M. AL-Eliwi a,b
a
Civil Engineering Department, University of Gaziantep, 27310 Gaziantep, Turkey
b
Civil Engineering Department, University of Mosul, 41001 Mosul, Iraq

a r t i c l e i n f o a b s t r a c t

Article history: Concrete filled steel tube (CFST) columns are deployed as the main load carrying members in high per-
Received 12 August 2016 formance structures as they integrate the beneficial properties of constituent materials. In recent years,
Revised 25 November 2016 there has been an increased interest in the literature to further improve the performance of the CFST col-
Accepted 30 December 2016
umns. The work presented in this paper, for the first time, investigates through experimental and statis-
Available online 9 January 2017
tical analyses the potential of concrete filled double circular steel tube (CFDCST) columns to improve the
efficiency of the composite members. Experimental results validated the capability of CFDCST columns in
Keywords:
performing high strength, stiffness and ductility properties which can be utilized in challenging design
Circular CFST columns
CFDCST columns
scenarios. Employed statistical analysis characterized the impact of parameters on the collapse perfor-
Experiments mance of the CFDCST column. To establish whether the configuration of CFDCST members could be used
CFST column repair in repairing process of highly stressed and deformed CFST columns, a further set of experimental work
was undertaken. Deformed CFST stub columns were centred in a larger diameter steel tube and the region
between the existing CFST column and the larger diameter tube was filled with concrete. Repaired forms
of the columns were tested and the performances were compared against the performances of the
CFDCST counterparts. Results show that it is possible to employ such a repairing, since the repaired
CFST columns exhibit performances very close to that of CFDCST counterparts. The findings are quite
revealing for design purposes and have important implications for advancing the knowledge on compos-
ite compression members.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction costs and provides efficient structural solutions. The steel tube
eliminates the need for external formwork in the concreting pro-
Composite construction is fast becoming a key application in cess [11,17,19] and compared to reinforced concrete and bare steel
high performance engineering structures. Concrete filled steel tube columns, the robust mechanical characteristics of CFST compres-
(CFST) columns which combine the favourable properties of steel sion members allow reducing the cross-sectional size of columns
and concrete are very effective compression members in providing for a given load and leads to increased net floor space [5,8,20].
the composite action. The composite action provokes superior Aforementioned merits of CFST columns encourage the employ-
mechanical properties of constituent materials. High ductility ment of such compression members in many different types of
and large energy absorption capacity of CFST columns [1–3] lead challenging engineering fields [18,21].
to superior seismic resistance [4–7]. Core concrete delays the local There is a growing body of literature that recognises the promi-
buckling of the steel tube by providing an internal support [4,8– nent and distinctive characteristics of CFST columns. Nevertheless,
10], resulting in a higher compression resistance [11,12]. In addi- in recent years, there has been an increasing interest to further
tion, circular steel tubes offer substantial confinement and increase improve the performance of such composite members and over-
the core concrete strength [5,9,13–15]. The steel tube forms the come the drawbacks. The demand for larger and more efficient
outermost layer of the column, and consequently the bending structures to satisfy the burden of increasing load actions imposes
moment performance of the steel can be fully exploited [16]. Fur- bigger column cross sections to be designed. Consequently, CFST
thermore, the steel tube motivates the use of high strength core columns with diameter up to 1600 mm were constructed in the
concrete by precluding its explosive failure [17,18]. The CFST col- field [22,23]. This diameter significantly reduces the useful area
umn concept has also the potential to reduce the construction of the structure. A number of research has been directed towards
the improvement of strength, ductility and confinement effect of
⇑ Corresponding author. CFST columns to minimize the cross-sectional dimensions under
E-mail address: ekmekyapar@gantep.edu.tr (T. Ekmekyapar). severe compression actions. External rings [24,25], internal rein-

http://dx.doi.org/10.1016/j.engstruct.2016.12.061
0141-0296/Ó 2017 Elsevier Ltd. All rights reserved.
 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80 69

forcement steel bars [26,27], internal steel sections steel tube. Therefore, this type of composite column configuration
[20,22,23,28,29] and stiffeners [16,30] have been proposed for this leads to better fire and post-fire performances.
purpose. In addition to sharing the promising characteristics of steel sec-
Another major issue in lately CFST research concerned about tion reinforced CFST columns, CFDCST columns offer further
the performance of columns in elastic stage [17,24]. Since the Pois- advantages. Since the shell and core concretes are separated by
son’s ratio of steel is larger than that of concrete, lateral expansion the inner steel tube, different concrete classes and properties can
of steel is larger than the expansion of concrete in the elastic stage, be employed for core and shell regions. Also, unlike steel section
resulting in delamination at the interface between the steel tube reinforced CFST, inner steel tube cross section is some distance
and core concrete. This leads to an unconfined core concrete beha- away from the centroid of the column in CFDCST. This could signif-
viour until the development of initial cracks in concrete and lateral icantly improve the bending resistance of the column (bending
expansion of concrete becomes larger than that of steel tube. This stresses are larger away from the centroid). Furthermore, both
behaviour delays the confinement effect and reduces the effective- inner and outer tube sections are circular, so the core and shell
ness of the steel tube [17]. This phenomenon could be more com- concretes could be confined better. Although extensive research
promising for CFST columns with high strength core concrete and has been carried out on CFST columns with various properties, to
thin steel tube [24]. In order to improve the elastic performance of the knowledge of authors no single study exists which character-
CFST columns, external ring application has been proposed in the izes and quantifies the performance of CFDCST compression mem-
literature [17,24]. bers. Thus, the behaviour of CFDCST columns is particularly
Although CFST columns offer better fire resistance over rein- appropriate for investigation.
forced concrete and bare steel counterparts [31,32], high stiffness The first objective of this paper is to characterize the response
of the exposed steel tube can be adversely affected under certain CFDCST stub columns. In order to assess the performance of
fire conditions. This also might jeopardize the post-fire perfor- CFDCST columns, an in-depth experimental program was under-
mance of the structure. Considering this behaviour, internal steel taken. Towards this aim, CFDCST stub column specimens were pre-
sections [20,23,28], double skin CFST columns [33,34], steel bar pared and concentrically tested to delve into pre-collapse, collapse
reinforcement [35] and fibre reinforcement [36] have been pro- and post-collapse behaviours. Following a systematic way of
posed to enhance the fire performance of CFST columns. examination, two different concrete classes were employed for
both core and shell regions (normal and high strength concrete
classes). Also, in order to examine the confinement effect and its
contribution to compression capacity, two different outer steel
2. Aim and significance of research
tube properties were utilized. That is to say, effects of three param-
eters on composite column performance were investigated (core
In view of the fact that the major function of the CFST structure
concrete strength, shell concrete strength and outer steel tube
is to sustain the loads, strength, stiffness and ductility are the main
properties). The second objective of this paper is devoted to explor-
focus in the design. However, further simultaneous measures must
ing the potential of CFDCST column configuration in repair applica-
be undertaken to improve the interaction between the core con-
tions. CFST stub columns were loaded up to a certain fraction of the
crete and the steel tube in the elastic stage which is the target
capacity to develop plastic deformations in the specimens. After
range in structural design and provide adequate safety against fire.
reaching the fraction of the compressive capacity, the loading
This study, therefore sets out to experimentally investigate the
was released. Following, the deformed column specimens were
performance of concrete filled double circular steel tube (CFDCST)
centred into larger diameter steel tubes and the regions between
column concept which has the potential to ease the burden of
the outer and the inner steel tubes were filled with concrete. Since
above design concerns. CFDCST column consists of a circular steel
the fresh concrete can take any shape, this repair concept compen-
tube which is centrally located in a larger diameter circular steel
sates geometrical deformations of the inner CFST column. After
tube, the core concrete in the smaller diameter steel tube and
completion of curing stage of the shell concrete, the repaired col-
the shell concrete between the steel tubes, Fig. 1. This column con-
umns were tested to failure. The use of the same steel and concrete
cept offers very high strength, superior ductility and stiffness prop-
properties in the second part of the study enabled the authors to
erties due to an embedded circular steel tube which gives rise to
assess the performance of repaired CFST columns against that of
effective use of floor space. The inner tube is supported both inter-
the intact CFDCST columns which were tested in the first part of
nally and externally by concrete which allows fully utilization of
the paper. Filling the gap in the literature, this study provides novel
the tube strength. Better composite action in both elastic and plas-
insights into the behaviour of composite construction and
tic stages can be attained, since steel tubes and the concrete inter-
advances the potential and knowledge of CFST applications.
act efficiently in CFDCST cross section. The inner steel tube and the
core concrete are protected by the shell concrete and the outer
3. Experimental procedure

3.1. Material properties

To be able to address the behaviour of CFDCST stub columns in

detail, experiments were conducted in a systematic manner. Two
different concrete mixes were produced. Cylinder moulds of
100  200 mm were filled with concrete in three layers. After each
layer, the concrete was compacted carefully to eliminate entrapped
air. Once the curing stage was completed, concrete specimens were
crushed in accordance with ASTM C39/C39M [37]. Measured com-
pressive strengths of two different concrete batches are 30.55 MPa
and 68.09 MPa. The motivation behind the selection of these con-
crete classes is to allow the examination of CFDCST columns with
Fig. 1. CFDCST column cross section. normal strength and high strength concrete. Moreover, the CFDCST
70 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80

column configuration offered the advantage of employing hybrid in five layers and the compaction process after each layer was
concrete properties (normal strength core concrete and high again applied. It is worth to note that, in order to ensure the com-
strength shell concrete or vice versa). All possible concrete combi- paction in narrow shell regions, some precautions were taken in
nations were tested to thoroughly characterize this type of com- the design of normal strength and high strength concrete mixes.
posite columns. The coarse and the fine aggregates used in the concrete mixes (nor-
A single type of inner steel tube was utilized throughout the mal strength and high strength) were river aggregates which have
investigation with an outer diameter of 88.9 mm and thickness rounded shapes. Also, the maximum size of coarse aggregate was
of 4.25 mm. On the other hand, two different outer steel tubes limited to 10 mm. Furthermore, a hyper-plasticizer was used to
were employed. Both outer steel tubes have outer diameters of provide highly workable concrete mixes. These precautions
139.7 mm. However, the first one has a thickness of 3.30 mm and allowed the authors to obtain good compaction in both core and
the second one has a thickness of 5.87 mm. Along with the prepa- shell regions. Obtaining uniform and flat surfaces at the upper ends
ration of CFDCST columns a number of supplementary tensile cou- of the columns is also necessary. To achieve this aim, the top last
pon specimens which comply with ASTM E8/E8M [38] were 3–4 mm of the tubes were left empty. After completing the curing
sectioned from the steel tubes to measure mechanical properties, process, the top of the columns was levelled using high strength
Fig. 2. The yield strengths were measured to be 375 MPa, levelling epoxy. Besides, to be able to assess the contribution of
290 MPa and 355 MPa for 88.9  4.25, 139.7  3.30 and the inner steel tube to the compression capacity, CFST columns
139.7  5.87 steel tubes, respectively. (without inner steel tube) with larger diameter steel tubes were
manufactured.
Prior to testing, the glass plates at the lower ends of the larger
3.2. Column specimens and testing
diameter steel tubes were removed and silicon residues were
cleaned. In order to eliminate eccentric loading and associated
All of the CFDCST column specimens have a length of 270 mm
bending actions, the CFDCST columns were centred carefully in
which corresponds to a stub column configuration. Since the global
the testing machine. A displacement controlled compressive load-
failure mode is prevented in stub columns, mechanical properties
ing with a slow rate was applied to all specimens up to the end of
of materials could be effectively exploited. Prior to casting the con-
the tests, Fig. 3. To be able to address pre-collapse, collapse and
crete, lower and upper ends of the steel tubes were machined to
post-collapse behaviours, compression loading versus end shorten-
insure maximum uniformity of contact with the loading heads of
ing data of the columns were recorded.
the test machine [18,21]. Providing uniform and flat ends is of
Also, to establish whether this type of geometrical configuration
key importance in column testing to obtain reliable test results
is appropriate to be employed in repair of the CFST columns, a fur-
[39].
ther series of experiments were undertaken. The smaller diameter
Surfaces of all steel tubes were cleaned from dust to guarantee
steel tubes were filled with the same concrete classes. After the
an appropriate interaction between the steel tubes and the con-
curing stage, CFST stub columns with smaller diameter steel tubes
crete. Thick glass plates were used at the bottom ends of the tubes
were concentrically loaded up to a fraction of compressive capac-
to restrict the fresh concrete during the casting and obtain a lev-
ity. At this level of loading, columns exhibit plastic longitudinal
elled flat surface of concrete at this end. Glass plates were con-
and lateral deformations. Following, the loading was released and
nected to the bottom ends of the smaller and larger diameter
after a period of time (this time is used to allow any possible spring
tubes using silicon. In order to manufacture CFDCST stub columns,
back behaviour of the steel tubes) deformed CFST columns were
first smaller diameter steel tubes were filled with concrete in five
centred in larger diameter steel tubes. Shell regions were filled
layers. After each layer, the concrete compacted carefully. A mini-
with concrete and repaired columns were subjected to curing con-
mum, but enough period of time was permitted to elapse between
ditions. Finally, these columns were fully tested to failure and pre-
concreting the core and shell regions to insure uniform curing of
collapse, collapse and post-collapse behaviours were recorded.
both concretes. Towards this aim, on completion of the core con-
crete setting, silicon at the bottom ends of the smaller diameter
steel tubes were removed and cleaned carefully. Following, the
smaller diameter steel tubes with hardened core concrete were
centred in larger diameter steel tubes which were previously con-
nected to thick glass plates. Shell regions were filled with concrete

Fig. 2. Tensile coupon specimens. Fig. 3. Column testing.

 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80 71

Table 1
Properties of stub column specimens.

Column Length Core fc Inner tube D Inner tube t Inner tube fy Shell fc Outer tube D Outer tube t Outer tube fy Test
(mm) (MPa) (mm) (mm) (MPa) (MPa) (mm) (mm) (MPa) case
CC1-SC1-OT1 270 30.55 88.9 4.25 375 30.55 139.7 3.30 290 Full
CC1-SC1-OT2 270 30.55 88.9 4.25 375 30.55 139.7 5.87 355 Full
CC1-SC2-OT1 270 30.55 88.9 4.25 375 68.09 139.7 3.30 290 Full
CC1-SC2-OT2 270 30.55 88.9 4.25 375 68.09 139.7 5.87 355 Full
CC2-SC1-OT1 270 68.09 88.9 4.25 375 30.55 139.7 3.30 290 Full
CC2-SC1-OT2 270 68.09 88.9 4.25 375 30.55 139.7 5.87 355 Full
CC2-SC2-OT1 270 68.09 88.9 4.25 375 68.09 139.7 3.30 290 Full
CC2-SC2-OT2 270 68.09 88.9 4.25 375 68.09 139.7 5.87 355 Full
CC1-OT1 270 30.55 – – – – 139.7 3.30 290 Full
CC1-OT2 270 30.55 – – – – 139.7 5.87 355 Full
CC2-OT1 270 68.09 – – – – 139.7 3.30 290 Full
CC2-OT2 270 68.09 – – – – 139.7 5.87 355 Full
CC1 270 30.55 88.9 4.25 375 – – – – Full
CC2 270 68.09 88.9 4.25 375 – – – – Full
CC1-R 270 30.55 88.9 4.25 375 – – – – Partial
CC1-R 270 30.55 88.9 4.25 375 – – – – Partial
CC1-R 270 30.55 88.9 4.25 375 – – – – Partial
CC1-R 270 30.55 88.9 4.25 375 – – – – Partial
CC2-R 270 68.09 88.9 4.25 375 – – – – Partial
CC2-R 270 68.09 88.9 4.25 375 – – – – Partial
CC2-R 270 68.09 88.9 4.25 375 – – – – Partial
CC2-R 270 68.09 88.9 4.25 375 – – – – Partial
CC1-SC1-OT1-R 270 30.55 88.9 4.25 375 30.55 139.7 3.30 290 Full
CC1-SC1-OT2-R 270 30.55 88.9 4.25 375 30.55 139.7 5.87 355 Full
CC1-SC2-OT1-R 270 30.55 88.9 4.25 375 68.09 139.7 3.30 290 Full
CC1-SC2-OT2-R 270 30.55 88.9 4.25 375 68.09 139.7 5.87 355 Full
CC2-SC1-OT1-R 270 68.09 88.9 4.25 375 30.55 139.7 3.30 290 Full
CC2-SC1-OT2-R 270 68.09 88.9 4.25 375 30.55 139.7 5.87 355 Full
CC2-SC2-OT1-R 270 68.09 88.9 4.25 375 68.09 139.7 3.30 290 Full
CC2-SC2-OT2-R 270 68.09 88.9 4.25 375 68.09 139.7 5.87 355 Full

Employment of the same material properties (both concrete and 4.1. CFDCST columns with normal strength core concrete
steel) in repair part of this paper enabled the authors to assess
the performance of repaired columns against CFDCST columns, Compression loading versus end shortening curves of the com-
resulting in obtaining key findings. posite columns with normal strength core concrete are presented
Table 1 presents all properties of the column specimens. Speci- in Fig. 4. Specimen CC1 achieved a compression load capacity of
men labelling system provided in Table 1 includes core concrete, 734.59 kN with a very good ductility behaviour. It is worth to note
shell concrete and outer steel tube types, respectively. For exam- that all specimens in this set of stub columns exhibited high duc-
ple, the specimen CC1-SC2-OT1 represents the CFDCST column tility as shown in Fig. 4. Compression capacities of the CFDCST
with core concrete strength of 30.55 MPa, shell concrete strength specimens CC1-SC1-OT1, CC1-SC1-OT2, CC1-SC2-OT1 and CC1-
of 68.09 MPa and outer steel tube with smaller thickness. There- SC2-OT2 were recorded to be 1435.05 kN, 1977.94 kN,
fore, in the labelling system, the numbers 1 and 2 refer weaker 1606.89 kN and 2044.44 kN, respectively. Considering the capacity
and stronger properties, respectively. Since there exist a single type of the specimen CC1, examined CFDCST column configurations
of inner steel tube section, in the specimen labelling system this provide performance increments in the range of 95.35% and
tube is not referred. Also, for the CFST specimens with larger diam- 178.31%. These levels of performance increments are significantly
eter steel tube (without inner steel tube) two level labelling system high and show the effectiveness of CFDCST concept. Also, it is
was used. Furthermore, a suffix of ‘‘R” is used to represent CFST col-
umns with smaller diameter steel tubes which were loaded up to a
fraction of the capacity (partially tested). The same suffix ‘‘R” is
also used in Table 1 to distinguish between CFDCST columns and
final form of the repaired CFST columns (deformed CFST columns
plus shell concrete and outer steel tube).

4. CFDCST stub columns

The compressive loading versus end shortening curves of

CFDCST columns are presented in this section to discuss the com-
pression performance, stiffness and ductility behaviours of the col-
umns. The influence of each parameter; core concrete strength,
shell concrete strength and outer tube properties, on the behaviour
is addressed. Also, comparison of the performance of CFDCST col-
umns against CFST columns with larger diameter steel tubes
reveals the contribution of the inner steel tubes. Furthermore,
depicted performance of CFST columns with smaller diameter steel Fig. 4. Compressive loading versus end shortening curves of columns with normal
tubes shows the effectiveness of CFDCST concept. strength core concrete.
72 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80

apparent from Fig. 4 that CFDCST columns offer much greater ini- than the squash load of the inner tube. Consequently, significant
tial stiffness. conclusions can be drawn. First, with CFDCST column configura-
Examining the influence of CFDCST column parameters, it can tion, mechanical properties of the inner circular steel tube are fully
be certainly deduced that the most effective component is the utilized. Second, the inner circular steel tube triggers a further con-
outer tube. The increment in thickness and yield strength of the finement effect on core and shell concretes. Furthermore, the
outer tube in CFDCST column with normal strength shell concrete toughness of the CFDCST columns is strongly influenced by the
led to 37.83% capacity improvement (comparison of CC1-SC1-OT1 employment of the inner steel tube, Fig. 4.
and CC1-SC1-OT2), whereas this improvement is 27.23% for Failure modes of the stub columns in this set are shown in
CFDCST column with high strength concrete shell (comparison of Fig. 5. All columns experienced compressive yielding of steel and
CC1-SC2-OT1 and CC1-SC2-OT2). Further examination shows that local failure of the tubes. Local failure of the steel tube is more
for the composite columns with thinner outer tubes, high strength obvious and takes a crippling mode in columns with the thinner
shell concrete increases the compression capacity by 11.97% (com- outer tube. A crushing mode which was followed by associated
parison of CC1-SC1-OT1 and CC1-SC2-OT1). On the other hand, shear failure was observed in concrete. Interestingly, it was
high strength shell concrete just improves the compression perfor- observed that the existence of the inner steel tube could not pre-
mance by 3.36% for columns with thicker outer tubes (comparison clude shear failure of concrete. The sectional views of some of
of CC1-SC1-OT2 and CC1-SC2-OT2). Local buckling of the thinner the tested specimens are presented in Fig. 6. Smooth inner surfaces
outer tube plays an important role in this behaviour. Since the fail- of the shell concretes verify the appropriate compaction of the con-
ure of the CFDCST columns with thinner outer tubes is mainly gov- crete in this region. The interactions between the inner tube and
erned by local buckling of the outer tube, shell concrete strength is the shell concrete and between the shell concrete and the outer
more visible in these specimens. tube can be observed clearly in this figure. The residue and dust
Assessment of compression capacities of CFST stub columns of the inner steel tube which come out after the yielding of the
(without inner tubes) against the compression capacity of associ- steel, left black marks on the inner surfaces of the shell concrete
ated CFDCST columns reveals important knowledge about the con- as seen in Fig. 6. By observing these marks and the shapes of the
tribution of the inner circular steel tubes. The compression shell concretes, it can be concluded that the shell concretes appro-
capacities of the specimens CC1-OT1 and CC1-OT2 were priately cover the inner steel tubes. On the other hand, considering
930.47 kN and 1480.55 kN, respectively. Considering the perfor- the distortions of the inner tube, shell concrete and the outer tube,
mances of specimens CC1-SC1-OT1 and CC1-SC1-OT2, these results it is apparent that the distortions in the specimen with the thinner
suggest 504.58 kN and 497.39 kN compression capacity incre- outer tube spread over larger regions, whereas the distortions in
ments due to the inner steel tube. These contributions are greater the specimen with the thicker outer tube intensify over a specific

Fig. 5. Failure modes of columns with normal strength core concrete.

 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80 73

Fig. 6. Sectional views of the specimens CC1-SC2-OT1 and CC1-SC1-OT2.

region of failure. This behaviour implies the superior confinement than performances of corresponding CFDCST columns with normal
effect and the governing role of the thicker outer tube which has strength core concrete. However, the percent increments provided
the main control over the failure of the shell concrete and the inner for columns with normal strength core concrete are greater as
tube. detailed in Section 4.1. This result confirms that the normal
strength core concrete was confined better compared to the high
4.2. CFDCST columns with high strength core concrete strength core concrete. It should be also noted that CFDCST col-
umns have greater initial stiffness compared to the stiffness of
Fig. 7 depicts compression load versus end shortening curves of specimen CC2.
columns with high strength core concrete. The compression capac- Comparison of the performances of CFDCST columns with nor-
ity of the specimen CC2 was 844.35 kN. Although specimen CC2 mal strength core concrete and corresponding CFDCST columns
has 14.91% greater compression capacity compared to the speci- with high strength core concrete reveals crucial understandings.
men CC1 with normal strength concrete, specimen CC1 exhibited The specimen CC2 has 109.76 kN greater load capacity compared
a superior ductility performance, Figs. 4 and 7. The hydraulic con- to the capacity of the specimen CC1 which corresponds to a
figuration and software of the testing machine do not allow the 14.91% performance increment. However, the specimens CC2-
specified displacement rate to be followed after 2200 kN due to SC1-OT1, CC2-SC1-OT2, CC2-SC2-OT1 and CC2-SC2-OT2 achieved
safety reasons. Beyond 2200 kN, a very slow displacement rate is 135.10 kN, 175.17 kN, 177.50 kN and 230.33 kN greater compres-
dictated by the testing machine. Because of this reason, after sion capacities compared to the capacities of the specimens CC1-
obtaining the compression capacity of the specimen CC2-SC2- SC1-OT1, CC1-SC1-OT2, CC1-SC2-OT1 and CC1-SC2-OT2, respec-
OT2 the test was terminated. CFDCST columns CC2-SC1-OT1, tively. These results indicate that the shell concrete and the outer
CC2-SC1-OT2, CC2-SC2-OT1 and CC2-SC2-OT2 developed compres- steel tube function very well in confining the core concrete and
sion load capacities of 1570.15 kN, 2153.11 kN, 1784.39 kN and greater strength can be obtained from the core concrete by
2274.77 kN, respectively. Therefore, the CDFCST column concept employing the CFDCST column concept.
with high strength concrete core offer performance increments in In an effort to compare the effectiveness of the CFDCST column
the range of 85.96% and 169.41% compared to performance of parameters, it was again found that the most contributing compo-
the specimen CC2. As expected, the compression loading capacities nent is the outer steel tube. Considering the CFDCST columns with
of CFDCST columns with high strength core concrete are greater normal strength shell concrete, thicker steel tube with higher yield
strength led to 37.13% greater compression capacity (comparison
of CC2-SC1-OT1 and CC2-SC1-OT2). However, the contribution of
the thicker tube with higher yield strength is 27.48% for CFDCST
column with high strength shell concrete (comparison of CC2-
SC2-OT1 and CC2-SC2-OT2). Interestingly, these percent incre-
ments provided by the outer steel tubes are in the same level of
percent increments provided by the outer steel tubes in the
previous set (Section 4.1). Thus, it can be concluded that in CFDCST
column concept contribution of the outer steel tube to
compression capacity is same for normal and high strength con-
crete cores. Examining the performance of the specimens with high
strength shell concrete compared to the columns with normal
strength shell concrete, it was observed that high strength shell
concrete increases the capacity by 13.64% in the columns with
thinner outer tube (comparison of CC2-SC1-OT1 and CC2-SC2-OT1),
whereas it increases the performance just by 5.65% in the
columns with thicker steel tubes (comparison of CC2-SC1-OT2 and
Fig. 7. Compressive loading versus end shortening curves of columns with high CC2-SC2-OT2).
strength core concrete.
74 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80

Fig. 8. Failure modes of columns with high strength core concrete.

In order to explore the role of the inner steel tube in compres- this figure. The distortions of the inner tube, shell concrete and the
sion performance of the CFDCST columns with high strength core outer tube again confirm the superior confinement effect and the
concrete, it is necessary to focus on the performance of the speci- governing characteristics of the thicker outer tube. The local distor-
mens without inner steel tube (CC2-OT1 and CC2-OT2). The com- tions of the parts and the black marks of the inner steel tube spread
pression capacities of the specimens CC2-OT1 and CC2-OT2 were over larger regions in the specimen with the thinner outer tube. On
recorded to be 1372.87 kN and 1884.10 kN, respectively. The the other hand, distortions of the inner steel tube, shell concrete
CFDCST columns CC2-SC2-OT1 and CC2-SC2-OT2 have 411.52 kN and the outer tube again intensify over a specific region of failure
and 390.67 kN greater compression load capacities compared to in the specimen with the thicker outer tube.
the columns CC2-OT1 and CC2-OT2, respectively. These increments Since the specimens tested in this study are stub columns, local
in loading capacities are very close to squash load of the inner steel material failures in both steel and concrete are dominant in the
tube which means an effective use of the inner tube properties. collapse modes. Stub column configuration minimizes the length
However, it can also be deduced that unlike CFDCST columns with effects and precludes the global buckling modes and coupled insta-
normal strength core concrete, in CFDCST columns with high bility modes (global buckling plus local material failure) to take
strength core concrete the inner steel tube could not provide addi- place. Herein, it is worth to note that the discussions made on
tional confinement to the core and shell concretes. Nevertheless, the performances of the specimens represent the behaviour of
the existence of the inner steel tube significantly improves the duc- the stub columns. CFDCST columns with higher length to diameter
tility and toughness behaviours of the column. The characteristics ratios would experience potentially different failure modes and
of the compression load versus end shortening curves in Fig. 7 collapse load performances. Coupled instability modes and global
show the significant contribution of the inner steel tube to the duc- buckling modes could be expected in medium length columns
tility and toughness behaviours. and long columns, respectively. The increment of the length to
Failure modes of the columns in this set are similar to that of diameter ratio and the associated failure mode would also lead
the specimens in the previous set with normal strength core con- to a reduction in collapse load performance.
crete, Fig. 8. Yielding of steel plus local failure of the tube and
crushing of concrete with associated shear failure was observed. 4.3. Ranking the contributions of the parameters
Local distortions were more apparent in CFDCST columns with
thinner outer tubes. The inner steel tubes could not prevent shear In the experimental set of CFDCST stub columns, there exist
type of failure in concrete. The sectional views of the specimens three distinct parameters; core concrete, shell concrete and outer
CC2-SC2-OT1 and CC2-SC1-OT2 are shown in Fig. 9. The appropri- steel tube properties. By observing the collapse load capacities of
ate compaction of the concrete in the shell regions can be seen in the CFDCST stub columns, the discussions have been done on the
 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80 75

Fig. 9. Sectional views of the specimens CC2-SC2-OT1 and CC2-SC1-OT2.

Table 2 Table 4
Lower and upper boundaries of the CFDCST column parameters. Percent contribution of the parameters to collapse load of CFDCST column.

CFDCST parameter Lower boundary Upper boundary CFDCST parameter Percent contribution to collapse load (%)
Core concrete strength 30.55 MPa 68.09 MPa Core f c 10.07
(MPa) Shell f c 6.44
Shell concrete strength 30.55 MPa 68.09 MPa Outer tube squash load 82.37
(MPa)
Interactions
Outer tube squash load (N) 1414 mm2  290 MPa 2468 mm2  355 MPa
Core f c  shell f c 0.18
Core f c  outer tube squash load 0.16
Shell f c  outer tube squash load 0.76
influence of each parameter on the compression performance. Error 0.02
However, to be able to clearly understand the variation in the Total 100
experimental performances and remove any uncertainty from the
discussions, the characterization of the parameters must be under-
taken. This characterization process requires a statistical analysis mention that different outer steel tube thicknesses lead to a small
to be followed. Based on the experimental results, an ANalysis Of difference in shell concrete areas. However, this difference could be
VAriance (ANOVA) was undertaken to numerically specify relative ignored in the analysis to simplify the procedure, since its
contribution of the individual parameter to collapse load of CFDCST influence on the behaviour would be very small compared to the
stub column [40]. influence of the outer tube properties. Table 2 provides lower
In the applied statistical method, all parameters were examined and upper boundaries of the parameters.
at only two levels. The first level represents a lower boundary of Collapse capacities of the CFDCST specimens have already been
the parameter and the second level represents an upper boundary provided in previous sections. However, to clarify the statistical
of the particular parameter. Two concrete grades (normal and high input employed in ANOVA analysis, Table 3 presents the results
strength) were employed for core and shell regions and two differ- from the experimental study in conjunction with parameter
ent outer steel tubes were employed. However, the geometrical boundaries in a summarized format.
configuration and mechanical properties of the outer steel tube As a result of ANOVA analysis, Table 4 presents percent contri-
do not allow to specify lower and upper boundaries of this tube butions of CFDCST column parameters to the collapse capacity.
using the yield strength of the material. Consequently, for the con- Examining the Table 4, over the range of parameter magnitudes
venience it is required to represent the boundaries of the outer studied, the outer tube squash load is seen to have the most signif-
steel tube using its squash load (cross-sectional area of the outer icant influence on the collapse capacity. The second important
tube multiplied by the yield strength). It is also necessary to parameter for the collapse capacity is the core concrete strength.

Table 3
Experimental results used in ANOVA analysis.

CFDCST column Core f c (MPa) Shell f c (MPa) Outer tube squash load (N) Collapse load (kN)

CC1-SC1-OT1 30.55 30.55 1414  290 1435.05

CC1-SC1-OT2 30.55 30.55 2468  355 1977.94
CC1-SC2-OT1 30.55 68.09 1414  290 1606.89
CC1-SC2-OT2 30.55 68.09 2468  355 2044.44
CC2-SC1-OT1 68.09 30.55 1414  290 1570.15
CC2-SC1-OT2 68.09 30.55 2468  355 2153.11
CC2-SC2-OT1 68.09 68.09 1414  290 1784.39
CC2-SC2-OT2 68.09 68.09 2468  355 2274.77
Maximum variation (kN) [as percentage] 839.72 [58.52%]
76 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80

Finally, shell concrete strength seems to have a third degree influ- reserve strength would be useful to provide safety in repairing pro-
ence on the collapse capacity. With regards the interactions cess. Moreover, the deformation responses of the CFST stub col-
between the parameters, the ANOVA analysis determined that umns filled with normal and high strength concretes are
the interactions have no statistical significance on the collapse significantly different. CFST stub columns filled with high strength
load. All of the interactions have percent contribution values smal- concrete reach the collapse load at a very small level of axial and
ler than 1%. lateral deformations, Fig. 10. On the other hand, the ductile beha-
It is worth noting that these percent contribution results which viour of CFST columns filled with normal strength concrete leads
allow ranking the influence of the parameters on collapse load are to significant magnitude of axial and lateral deformations at the
not fully generic, and are depend on the range of the parameter level of column capacity, Fig. 10. Therefore, in case of stressing
magnitudes. However, these findings have important implications the CFST columns filled with normal strength concrete to further
for developing the correct sense in CFDCST stub column design. levels of the capacity, due to excessive lateral deformations it
would be very difficult to surround the deformed column with
an outer tube which has a reasonable diameter. Thus, 85% of the
5. Repair of CFST columns collapse load is a rational stage of column performance to facilitate
the repairing process.
The geometrical configuration of CFDCST columns can also be Compression loading versus end shortening curves of the CFST
utilized to repair or strengthen stressed and deformed CFST col- columns which were loaded up to 85% of the compression capaci-
umns. Structures, especially the public ones are stressed to further ties were recorded for comparison purposes. Before repairing the
levels under the increasing load actions of daily life. Consequently, CFST columns, a certain time was permitted to elapse to eliminate
unexpected deformations which could compromise the perfor- any adverse effect of spring back behaviour of the deformed steel
mance and safety of structure might arise. Instead of rebuilding tubes. In case of an immediate repair process, this spring back
the structure, the load carrying members of the structure could behaviour of the deformed steel tube could cause cracks in the
be repaired or strengthened. If the components of the structure shell concrete and disable the bond between the shell concrete
did not lose the full strength, repairing would be more economical and the steel tubes. Following, deformed CFST columns were cen-
compared to the rebuilding process. In this manner, deformed CFST tred in the larger diameter steel tubes and the shell regions were
columns could be surrounded by a larger diameter steel tube and filled with concrete. After completion of curing stage, eight stub
the space between the existing column and the larger diameter columns were tested to failure. Load displacement behaviour of
steel tube could be filled with concrete. Since the fresh concrete columns were recorded to be able to assess the performances of
can take any shape, it has the advantage of compensating the the repaired columns against the performances of CFDCST columns
deformations of the existing CFST column, resulting in an out- which are presented in Section 4. In order to make clear discus-
standing repair method. sions each repaired column is handled separately in this section.
To investigate the repairing process, eight CFST stub columns
(four with normal strength core concrete and four with high
strength core concrete) were loaded up to 85% of the compression 5.1. Repair of CFST columns with normal strength core concrete
load capacities. The motivation behind the selection of such a frac-
tion of compression loading is based on two reasons. First, the Figs. 11–14 present the compression performances of compos-
structures, including CFST columns are designed to serve in the ite columns in a comparative form. Performances of the CFST col-
elastic range. For the studied configurations, 85% of the collapse umn (loaded up to 85% of the capacity), its repaired form
load capacity corresponds to a level of plastic behaviour at which (covered by shell concrete and larger diameter outer steel tube)
the deformations in the columns could be clearly observed. Also, and corresponding CFDCST column (discussed in Section 4) are
at this stage, the column does not lose the full strength and the shown in one figure to be able to make clear judgements. The

Fig. 10. Normal and high strength CFST column behaviours.

 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80 77

Fig. 11. Compression performance of CC1-SC1-OT1-R. Fig. 14. Compression performance of CC1-SC2-OT2-R.

The compression capacities of the repaired stub columns CC1-

SC1-OT1-R, CC1-SC1-OT2-R, CC1-SC2-OT1-R and CC1-SC2-OT2-R
were 1483.20 kN, 1969.90 kN, 1576.73 kN and 2115.89 kN, respec-
tively. These capacity loads correspond to +3.36%, 0.41%, 1.88%
and +3.50% performance differences between the CFDCST stub col-
umns and repaired counterparts, respectively. Therefore, it can be
concluded that the repaired CFST stub columns could exhibit com-
pression performances very close to CFDCST counterparts.
Although the repaired CFST columns were highly deformed in plas-
tic region, there is no an adverse effect of such deformations in the
columns covered with shell concrete and outer steel tube. This
behaviour could be attributed to several reasons, especially the
mechanical characteristics of the steel. First, the CFST columns
were not fully tested to failure which means that the constituent
materials steel and concrete have still reserve strength in compres-
sion. According to the classification of AISC 360-10 [41], the smal-
Fig. 12. Compression performance of CC1-SC1-OT2-R.
ler diameter tube forms a compact section due to geometrical and
mechanical characteristics. A compact circular section has a suffi-
cient thickness to develop yielding of the steel in longitudinal com-
pression and to provide confinement to the concrete infill to
develop 0.95f c . In support of this, the compact section can develop
the full plastic strength in compression [42]. This capacity was not
fully consumed in CFST columns which were loaded up to 85% of
the collapse load. Furthermore, in case of reloading, plastically
deformed steel can exhibit greater yield strength compared to
yield strength of parent material [43].
Also, in addition to collapse load capacities, the stiffness and
ductility responses of the repaired columns are of importance in
comparisons. It can be clearly observed from Figs. 11–14 that
repaired columns have stiffness and ductility behaviours very close
to CFDCST counterparts. Failure modes of the repaired columns are
depicted in Fig. 15. Yielding of the steel and local failure of the tube
were observed in failure modes. However, it is important to note
that unlike CFDCST columns in which shear type of failure of con-
crete is more dominant, in repaired columns compressive crushing
Fig. 13. Compression performance of CC1-SC2-OT1-R.
is more dominant in concrete, Fig. 15.

stiffness and ductility performances of the repaired and corre- 5.2. Repair of CFST columns with high strength core concrete
sponding CFDCST stub columns can also be compared.
The presented results of the repair study are quite revealing in Compression performances of repaired columns with high
several ways. It is apparent from Figs. 11–14 that CFST stub col- strength core concrete are illustrated in Figs. 16–19. Behaviours
umns filled with normal strength concrete experience significant of CFST column (CC2-R), its repaired form and CFDCST column with
amount of plastic deformations at 85% of the compression capacity. same material configuration are compared.
Four CC1-R columns reached the 85% of the compression capacities As expected, CFST stub columns with high strength core con-
at the level of end shortenings varying from 7.18 mm to 9.10 mm. crete reached 85% of the compression capacities at very small
These levels of axial deformations also led to visible permanent lat- end shortening levels. The CC2-R columns experienced 85% of the
eral deformations in CC1-R columns. capacities at the end shortening levels varying from 2.03 mm to
78 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80

Fig. 15. Failure modes of repaired columns filled with normal strength core concrete.

Fig. 16. Compression performance of CC2-SC1-OT1-R. Fig. 18. Compression performance of CC2-SC2-OT1-R.

Fig. 19. Compression performance of CC2-SC2-OT2-R.

Fig. 17. Compression performance of CC2-SC1-OT2-R.

performances of corresponding CFDCST columns, there exist

2.24 mm. At this level, the CFST columns experience plastic defor- 1.06%, 2.20%, +2.98% and 0.66% differences in collapse capac-
mations as can be observed in Figs. 16–19. However, unlike CFST ities, respectively.
stub columns with normal strength core concrete, permanent Stiffness and ductility behaviours of repaired forms of the stub
deformations were not highly visible. columns are also very close to that of CFDCST columns as shown in
Like repaired CFST stub columns with normal strength core con- Figs. 16–19. Therefore, it can be deduced that the CFST stub col-
crete, repaired CFST stub columns with high strength core concrete umns with high strength core concrete can be repaired by employ-
exhibited remarkable compression performances. The specimens ing larger diameter circular steel tube and infill concrete between
CC2-SC1-OT1-R, CC2-SC1-OT2-R, CC2-SC2-OT1-R and CC2-SC2- existing deformed CFST column and the larger diameter steel tube.
OT2-R developed collapse loads of 1553.42 kN, 2105.73 kN, The failure modes of the repaired forms of the columns are shown
1837.57 kN and 2259.66 kN, respectively. Compared to the in Fig. 20. As with the repaired CFST columns with normal strength
 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80 79

Fig. 20. Failure modes of repaired columns filled with high strength core concrete.

core concrete, compression crushing of concrete is more dominant strength and shell concrete strength have second and third
in failure modes of these specimens. degree importance on the collapse capacity, respectively.
 Experimental results showed that the geometrical configuration
6. Conclusions of double circular steel tubes is highly convenient to be utilized
in repairing the CFST columns. The outer steel tube has signifi-
This article examines through experimental and statistical anal- cant potential to increase the compression performance and the
ysis the potential of CFDCST stub columns for improving the struc- shell concrete could compensate the deformations of the exist-
tural performance of composite applications. Two different ing CFST columns.
concrete classes (normal and high strength) were employed for  The repaired columns exhibit compression, ductility and stiff-
core and shell regions, and two types of outer steel tube (same ness performances very close to performances of the CFDCST
outer diameter, but one with thinner section and lower yield counterparts. The CFST columns with normal strength core con-
strength and another one with thicker section and higher yield crete were highly deformed at the level of 85 % of the compres-
strength) were used. The ranges of the parameters were chosen sion capacity. Even with this deformations, the repaired
with the intent to explore the impact of each parameter on the col- columns developed remarkable performances very close to that
lapse performance. The experimental results validated the effi- of CFDCST counterparts.
ciency of CFDCST column configuration. Statistical analysis
demonstrated the contribution of each parameter to collapse load
of the CFDCST stub column. Additionally, further experimental Acknowledgements
tests were undertaken to investigate the potential of CFDCST col-
umn configuration in repairing deformed CFST columns. Based on The authors acknowledge financial supports of the Scientific
the results, following conclusions can be drawn: Research Unit of the University of Gaziantep on the Project MF
14.06. The authors especially would like to thank the Borusan Man-
 CFDCST column configuration is very effective in improving nesmann/Turkey for providing the steel tubes for the experiments.
compression, ductility and stiffness performances and could The valuable support of the BASF/Turkey (AKAS) in supplying the
be deployed in structures to lead to reductions in cross- concrete admixtures is also appreciated.
sectional dimensions of the compression members.
 The boundaries of the inner steel tube provide separated con-
References
crete regions in CFDCST columns (core and shell regions).
Therefore, this composite column configuration offers the [1] Han LH, Li W, Bjorhovde R. Developments and advanced applications of
advantage of using two different concrete classes in the cross concrete-filled steel tubular (CFST) structures: members. J Constr Steel Res
section of the composite member. 2014;100:211–28.
[2] Liu H, Wang YX, He MH, Shi YJ, Waisman H. Strength and ductility
 Experimental results validated that the mechanical perfor- performance of concrete-filled steel tubular columns after long-term service
mance of the inner steel tube could be fully exploited owing loading. Eng Struct 2015;100:308–25.
to effective supports of core and shell concretes. Additionally, [3] Chang X, Luo XL, Zhu CX, Tang CN. Analysis of circular concrete-filled steel tube
(CFT) support in high ground stress conditions. Tunn Undergr Sp Tech
the inner steel tube provides a further confinement effect in 2014;43:41–8.
case of the use of normal strength concrete in core and shell [4] Zeghiche J, Chaoui K. An experimental behaviour of concrete-filled steel
regions. tubular columns. J Constr Steel Res 2005;61:53–66.
[5] Huang C, Han XL, Ji J, Tang JM. Behavior of concrete-filled steel tubular planar
 The inner steel tube significantly improves the ductility and intersecting connections under axial compression, Part 1: Experimental study.
toughness of the columns. The improvements are much more Eng Struct 2010;32:60–8.
perceptible in columns with high strength core and shell [6] Dong JF, Wang QY, Guan ZW. Structural behaviour of recycled aggregate
concrete filled steel tube columns strengthened by CFRP. Eng Struct
concretes.
2013;48:532–42.
 The longitudinal sectional views of the tested columns verified [7] Guo LH, Zhang SM, Kim WJ, Ranzi G. Behavior of square hollow steel tubes and
the superior confinement effect of the thicker outer tube. steel tubes filled with concrete. Thin Wall Struct 2007;45:961–73.
 The results of the statistical analysis demonstrated that the [8] Giakoumelis G, Lam D. Axial capacity of circular concrete-filled tube columns. J
Constr Steel Res 2004;60:1049–68.
outer steel tube properties have the greatest impact on the col- [9] Fam A, Qie FS, Rizkalla S. Concrete-filled steel tubes subjected to axial
lapse capacity of the CFDCST stub columns. The core concrete compression and lateral cyclic loads. J Struct Eng – Asce 2004;130:631–40.
80 T. Ekmekyapar, B.J.M. AL-Eliwi / Engineering Structures 135 (2017) 68–80

[10] Lu YY, Li N, Li S, Liang HJ. Behavior of steel fiber reinforced concrete-filled steel [27] Yang Y, Wang Y, Fu F. Effect of reinforcement stiffeners on square concrete-
tube columns under axial compression. Constr Build Mater 2015;95:74–85. filled steel tubular columns subjected to axial compressive load. Thin Wall
[11] Lai MH, Ho JCM. A theoretical axial stress-strain model for circular concrete- Struct 2014;82:132–44.
filled-steel-tube columns. Eng Struct 2016;125:124–43. [28] Zhu M, Liu J, Wang Q, Feng X. Experimental research on square steel tubular
[12] Uy B. Local and post-local buckling of concrete filled steel welded box columns filled with steel-reinforced self-consolidating high-strength concrete
columns. J Constr Steel Res 1998;47:47–72. under axial load. Eng Struct 2010;32:2278–86.
[13] Kitada T. Ultimate strength and ductility of state-of-the-art concrete-filled [29] Xu F, Chen J, Jin WL. Experimental investigation of thin-walled concrete-filled
steel bridge piers in Japan. Eng Struct 1998;20:347–54. steel tube columns with reinforced lattice angle. Thin Wall Struct
[14] Chitawadagi MV, Narasimhan MC, Kulkarni SM. Axial strength of circular 2014;84:59–67.
concrete-filled steel tube columns – DOE approach. J Constr Steel Res [30] Tao Z, Uy B, Han LH, Wang ZB. Analysis and design of concrete-filled stiffened
2010;66:1248–60. thin-walled steel tubular columns under axial compression. Thin Wall Struct
[15] Li YL, Zhao XL, Singh RKR, Al-Saadi S. Experimental study on seawater and sea 2009;47:1544–56.
sand concrete filled GFRP and stainless steel tubular stub columns. Thin Wall [31] Shakir AS, Guan ZW, Jones SW. Lateral impact response of the concrete filled
Struct 2016;106:390–406. steel tube columns with and without CFRP strengthening. Eng Struct
[16] Dabaon MA, El-Boghdadi MH, Hassanein MF. Experimental investigation on 2016;116:148–62.
concrete-filled stainless steel stiffened tubular stub columns. Eng Struct [32] Song TY, Han LH, Yu HX. Concrete filled steel tube stub columns under
2009;31:300–7. combined temperature and loading. J Constr Steel Res 2010;66:369–84.
[17] Kwan AKH, Dong CX, Ho JCM. Axial and lateral stress-strain model for circular [33] Li W, Han LH, Zhao XL. Axial strength of concrete-filled double skin steel
concrete-filled steel tubes with external steel confinement. Eng Struct tubular (CFDST) columns with preload on steel tubes. Thin Wall Struct
2016;117:528–41. 2012;56:9–20.
[18] Ekmekyapar T, Al-Eliwi BJM. Experimental behaviour of circular concrete filled [34] Lu H, Han LH, Zhao XL. Fire performance of self-consolidating concrete filled
steel tube columns and design specifications. Thin Wall Struct double skin steel tubular columns: experiments. Fire Safety J 2010;45:106–15.
2016;105:220–30. [35] Lie TT, Kodur VKR. Fire resistance of steel columns filled with bar-reinforced
[19] Yu ZW, Ding FX, Cai CS. Experimental behavior of circular concrete-filled steel concrete. J Struct Eng – Asce 1996;122:30–6.
tube stub columns. J Constr Steel Res 2007;63:165–74. [36] Kodur VKR, Lie TT. Fire resistance of circular steel columns filled with fiber-
[20] Wang QX, Zhao DZ, Guan P. Experimental study on the strength and ductility reinforced concrete. J Struct Eng – Asce 1996;122:776–82.
of steel tubular columns filled with steel-reinforced concrete. Eng Struct [37] ASTM C39/C39M-14a, Standard Test Method for Compressive Strength of
2004;26:907–15. Cylindrical Concrete Specimens, ASTM International. West Conshohocken,
[21] Ekmekyapar T. Experimental performance of concrete filled welded steel tube PA2014.
columns. J Constr Steel Res 2016;117:175–84. [38] ASTM E8/E8M-15a, Standard Test Method for Tension Testing of Metallic
[22] Chang X, Wei YY, Yun YC. Analysis of steel-reinforced concrete-filled-steel Materials, ASTM International. West Conshohocken, PA2015.
tubular (SRCFST) columns under cyclic loading. Constr Build Mater [39] Ziemian RD. Guide to stability design criteria for metal structures. 6th
2012;28:88–95. ed. Wiley; 2010.
[23] Cai J, Pan J, Wu Y. Mechanical behavior of steel-reinforced concrete-filled steel [40] Roy R. A primer on the Taguchi method. Society of Manufacturing Engineers;
tubular (SRCFST) columns under uniaxial compressive loading. Thin Wall 1990.
Struct 2015;97:1–10. [41] AISC 360-10. Specification for structural steel buildings. Chicago, Illinois: AISC;
[24] Lai MH, Ho JCM. Confinement effect of ring-confined concrete-filled-steel-tube 2010.
columns under uni-axial load. Eng Struct 2014;67:123–41. [42] Commentary on the Specification for Structural Steel Buildings AISC-360-10.
[25] Lai MH, Ho JCM. Optimal design of external rings for confined CFST columns. Chicago, Illinois: AISC; 2010.
Mag Concr Res 2015;67:1017–32. [43] Gere JM, Goodno BJ. Mechanics of materials. Brief SI ed. Cengage Learning;
[26] Xiamuxi A, Hasegawa A. A study on axial compressive behaviors of reinforced 2012.
concrete filled tubular steel columns. J Constr Steel Res 2012;76:144–54.