Structures 79 (2025) 109311

Contents lists available at ScienceDirect

Structures
journal homepage: www.elsevier.com/locate/structures

Axial performance of ultra-high strength concrete filled double-skin

circular steel tube columns after exposure to fire
Xiao Lyu a,b, Peng Zhang a , Yu-Zhuo Wang a,b,* , Xiao-Xi Hu a, Er-Feng Du c,d, Ze-Jian Liu a
a
School of Civil Engineering, Shandong Jianzhu University, Jinan 250101, China
b
Key Laboratory of Building Structural Retrofitting and Underground Space Engineering (Shandong Jianzhu University), Ministry of Education, Jinan 250101, China
c
School of Civil Engineering, Southeast University, Nanjing 211189, China
d
Key Laboratory of Concrete and Pre-stressed Concrete Structures (Southeast University), Ministry of Education, Nanjing 211189, China

A R T I C L E I N F O A B S T R A C T

Keywords: To investigate the mechanical properties of ultra-high strength concrete filled double-skin circular steel tube
Ultra-high strength concrete (UCFDST) short columns under axial compression after fire exposure, four axial compression tests were con­
Double-skin circular steel tube ducted on UCFDST short columns. The compressive strength of the sandwich ultra-high strength concrete is
Columns
160 MPa. The test results indicate that despite undergoing prior fire resistance tests, the composite members
After fire
Parametric analysis
retain good residual bearing capacity and ductility performance. The study analyzed the working mechanism and
failure modes of UCFDST short column specimens under axial compression post-fire, along with the load-
displacement curve of a typical specimen. Finite element models were established using ABAQUS and vali­
dated through experiments. Key parameters of the component were considered in parametric analysis including
the compressive strength of the concrete cube, yield strength of the outer steel tube, duration of fire exposure,
hollow ratio, slenderness ratio, and wall thickness of the outer steel tube. The results show that increasing the
concrete cube compressive strength has a minimal effect on the ultimate bearing capacity. However, the ultimate
bearing capacity increases by 32.31 % when comparing the yield strengths of Q355 and Q590 for the outer steel
tube, indicating a significant impact of the outer steel tube’s yield strength on the component’s ultimate bearing
capacity. The ultimate bearing capacity decreases with longer fire exposure times, and the load-displacement
curve becomes more gradual at maximum bearing capacity. A smaller hollow ratio results in greater ultimate
bearing capacity, while the slenderness ratio and wall thickness of the outer steel tube also significantly influence
the ultimate bearing capacity of the members. The proposed formula has good calculation accuracy and can
provide reference for the residual bearing capacity evaluation after exposure to fire of UCFDST columns.

1. Introduction members, is considered. Researchers worldwide have conducted

experimental studies and numerical simulations on CFDST columns at
A concrete filled double-skin steel tube (CFDST) is a composite normal temperatures. Le et al. [16] analyzed the square CFDST short
member consisting of two concentric steel tubes—an inner and an outer columns with outer stainless-steel tube under axial compression and a
tube—with concrete poured into the space between them [1–6]. It in­ design formula for the ultimate bearing capacity was proposed. Wang
herits the high bearing capacity, good ductility, convenient construc­ et al. [17] conducted an experimental study on the hysteretic perfor­
tion, and economic benefits of concrete filled circular steel tubular mance of high-strength CFDST columns with diameter-to-thickness ra­
members. Additionally, it offers lightweight, high bending stiffness, tios exceeding the limit and a theoretical model considering the tensile
sectional versatility, and excellent fire resistance [7–12]. These features strength effect of engineered cementitious composite was proposed to
make it highly regarded by researchers globally and promising for en­ calculate the flexural capacity. Manigandan et al. [18] performed
gineering applications [13–15]. In super high-rise buildings, to reduce eccentric compression tests and numerical simulations on circular
the load from the structure itself, replacing the sandwich ordinary CFDST short columns. The results show that the increase in the intensity
concrete with ultra-high strength concrete, thus creating UCFDST of eccentrically loaded circular CFDST short column decreases the

* Corresponding author at: School of Civil Engineering, Shandong Jianzhu University, Jinan 250101, China.
E-mail address: [email protected] (Y.-Z. Wang).

https://doi.org/10.1016/j.istruc.2025.109311
Received 14 December 2024; Received in revised form 29 April 2025; Accepted 25 May 2025
Available online 12 June 2025
2352-0124/© 2025 Institution of Structural Engineers. Published by Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and
similar technologies.
X. Lyu et al. Structures 79 (2025) 109311

ultimate compressive strength and initial stage of bending stiffness. 2. Experiments

Deng et al. [19] investigated the torsional behavior of conical CFDST
columns through experimental studies. The results show that the cone 2.1. General
angle and hollow ratio have a great influence on the torsional bearing
capacity of the members. Chen J et al. [20] carried out eccentric The axial compression test of short columns after fire was conducted
compression tests on seven dodecagon CFDST members. The formula for following Xiong’s fire resistance limit experiment of long columns.[35].
calculating the bearing capacity of concrete filled double skin steel The specimens were truncated, retaining the fire-affected parts at the
tubular members with dodecagonal hollow section is proposed. Wang ends. The long column specimens were heated according to the ISO-834
et al. [21] replaced the outer steel tube of CFDST short column speci­ standard heating curve. Table 1 shows the yield strength of steel,
mens with stainless steel and explored their mechanical properties using compressive strength of concrete, duration of fire exposure, and thick­
finite element analysis. Ahmed et al. [22] tested 20 CFDST short column ness of the fire protection layer. The ultra-high strength concrete used in
specimens under eccentric compression, varying section size, the test was produced by DENSIT company. The mixed aggregate model
width-to-thickness ratios of inner and outer steel tubes, and eccentric is D4, which is composed of cementitious material superplasticizer and
distance. Li et al. [23]performed axial compression tests on four CFDST alumina aggregate with a maximum particle size of less than 4.75 mm.
short columns with high-strength steel outer tubes and found the outer Before the fire resistance limit test of the long column, the measured
steel tube was locally buckled outwards, the concrete at the buckling compressive strength of the sandwich concrete ranged from 163 to
was crushed, and the inner steel tube was buckled inwards. 165Mpa. Cylinder standard specimen material properties test ultra-high
In recent years, the frequent occurrence of high-rise building fires strength concrete mix as shown in Table 2. S355 steel was used for both
has prompted domestic and foreign scholars to study the mechanical the inner and outer steel tubes.
aspects of CFDST members during and after a fire. Yang and Han.[24] In the fire resistance limit experiment of long column, the thermo­
proposed a simple formula for calculating the thickness of the fire pro­ couple is embedded in the cross section of the longitudinal middle po­
tection layer and the fire resistance limit through finite element analysis sition of the specimens: the outer wall of the inner steel tube, the
of CFDST columns exposed to fire. Lu et al. [25–27] conducted experi­ midpoint of the sandwich concrete and the inner wall of the outer steel
mental research on the fire resistance of these members and performed tube are arranged with a thermocouple respectively. When the ther­
finite element simulations in fire conditions. Yao et al. [28] established a mocouple is embedded, it is protected by a plastic bag to prevent it from
calculation model for CFDST columns in fire and predicted the fire being destroyed during concrete pouring. During the first 10 minutes of
resistance limit by the traditional Rankine method. Romero et al. [29] the test, the system records the temperature data every 10 seconds, and
conducted fire resistance tests on high-strength concrete filled after 10 minutes, the temperature data is recorded every 20 seconds.
double-skin steel tubular columns, showing that the fire resistance of the To accurately study the axial compressive mechanical properties of
specimens increases with the thickness of the inner steel tube. Wan et al. ultra-high strength concrete filled steel tubes after fire, the length of the
[30] conducted axial compression tests on CFDST members after fire and intercepted specimen must be appropriate. Specimens that are too long
proposed a simplified formula for calculating the bearing capacity of may exhibit bending deformation, causing the resistance measured in
axial compression under a uniform temperature field. Lu et al. [31] the test to not accurately represent the true compressive strength.
analyzed the mechanical properties of six CFDST specimens after fire, Conversely, specimens that are too short cannot ignore the influence of
using axial compression ratio, hollow ratio, eccentricity, and fire pro­ the end effect. Generally, the longitudinal length of the short column
tection layer thickness as test parameters. Li et al. [32] conducted hys­ should be 3–3.5 times the diameter or side length. This test follows
teretic tests on 12 CFDST specimens after fire and found that the failure recommendations from relevant literature, with the longitudinal length
mode is similar to that at room temperature. Liu et al. [33,34] performed of the short column specimen being 3 times the diameter or side length.
13 sets of tests on the residual bearing capacity of CFDST columns under [36].
axial compression after varying fire durations and found that the CFDST
column has good fire resistance, while the concrete strength has little 2.2. Specimen making and test process
effect on the residual bearing capacity of CFDST column after fire.
In summary, the research on the mechanical properties of UCFDST In order to simulate the damage of the component in the case of pre-
columns after fire can provide theoretical basis and technical support for loading when encountering fire in the actual project., the naturally
its subsequent identification and reinforcement. At the same time, cooled UCFDST compression specimen was cut to form a short column
mastering the general law of axial compression performance is the basis after the fire resistance test. Compared with the traditional research on
for studying its working performance under complex stress conditions. It the residual bearing capacity of short columns after fire, it is more
also provides a theoretical basis for the analysis of the mechanical per­ realistic to cut short column specimens on long column specimens with
formance of such long columns and eccentric compression columns after pre-loading after fire. The cutting and post-processing of the short col­
fire. To address this, this paper utilizes ABAQUS finite element software umn are illustrated in Fig. 1(a). The geometric dimensions and related
to establish a finite element model of the axial compressive mechanical parameters of the short column are detailed in Table 3. The end plate
properties of CFDST columns after fire and validates the model through measures 400 mm × 400 mm × 20 mm. The end plates were groove
experiments. The results of the validation show good agreement. The welded to the outer steel tube of the short column specimen. To mitigate
study analyzes the effects of ultra-high strength concrete strength, yield the adverse effects of flame cutting and welding heat on the steel tube
strength of the outer steel tube, fire duration, hollow ratio, slenderness and internal ultra-high strength concrete, some researchers have
ratio, and wall thickness of the outer steel tube on the bearing capacity increased the thickness of the end plate.[37]. In this test, stiffeners were
of specimens after fire, and reveals the axial compression working welded at two ends of the specimen to reduce the negative impact of
mechanism of UCFDST columns after fire. According to the results of high heat input on the specimen’s bearing capacity.
parameter analysis, a simplified formula for calculating the residual To accurately measure the longitudinal and transverse deformation
bearing capacity of UCFDST columns after fire is given, which provides a of the specimen, longitudinal strain gauges (Z: longitudinal direction)
reference for calculating the residual bearing capacity of such compo­ were placed on one side in both the X and Y directions at 1/4 and 3/4
nents after fire in engineering practice. heights of a single specimen. Longitudinal strain gauges were set every
90 degrees at 1/2 height. Additionally, circumferential strain gauges (H:
circumferential direction) were positioned on one side in both the X and
Y directions at 1/2 height, with a total of 10 strain gauges used. Two
vertical displacement meters were symmetrically placed at the top and

2
X. Lyu et al. Structures 79 (2025) 109311

Table 1
Parameters of specimens in fire resistance experiment.[35].
Specimen fck (MPa) fy(MPa)/Es(GPa) time of fire(min) tp load ratio α λ χ ξ
(mm)

CZ− 2–1 165 432/203 60 0 0.33 0.72 10.9 0.61 1.88

CZ− 2–2 163 134 8.2 0.60 1.90
CZ− 2–3 165 468/183 25 7.9 0.94 ​ ​ ​ 1.88

Notes: CZ represents the long column. tp is the thickness of covering for fire protection. α is the nominal steel ratio of the section. α=As/Ac (As is the total area of inner
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
and outer steel tubes, and Ac is the area of sandwich concrete.) λ is the slenderness ratio, λ= 4 L/ (D2 + D21 ), L is the height of the specimen, D(D1) is the diameter of the
outer steel tube (inner steel tube).χ is the hollow rate, χ =D1/(D-2t), t is the wall thickness of outer steel tube. ξ is the hoop constraint coefficient, ξ= Asfy/Acfck= αfy/fck.

2.3. Loading equipment and loading system

Table 2
Mix proportion of ultra-high strength concrete.
The axial compression test was conducted using the 1500 t
mix proportion (water/D4) water (kg) D4 (kg) grouting volume (L) compression-shear test machine at the experimental center of the School
0.114 2.85 37.5 14.1 of Civil Engineering, Southeast University. To ensure the specimen was
under axial compression and to verify the functionality of the strain
gauge and displacement meter, a preload of 1000 kN was applied to the
bottom of the specimen, and one lateral displacement meter was posi­
specimen before formal test. This preloading step was performed to
tioned at 1/2 height. The arrangement of strain gauges and displace­
observe the reliability of the loading system and the performance of each
ment meters is shown in Fig. 1(b).
measuring point. The preloading process lasted 1.5 minutes, with the

Fig. 1. Fabrication of specimen and experiment arrangement.

3
X. Lyu et al. Structures 79 (2025) 109311

Table 3
Relevant parameters of specimens after fire.
Specimen L D×t D1×t1 The outer and inner walls experience the Concrete experiences the highest maximum temperature Pe Pu
(mm) (mm) × (mm) (mm) × (mm) maximum temperature (℃) (℃) (kN) (kN)

DZ− 2–1X 657.3 219.1 × 16 114.3 × 6.3 718 597 7900 8512
413
DZ− 2–2X 657.3 16 × 219.1 6.3 × 114.3 518 422 8200 8782
325
DZ− 2–3S 657.3 16 × 219.1 6.3 × 114.3 143 122 8700 9236
101
DZ− 2–3X 657.3 16 × 219.1 6.3 × 114.3 143 122 8700 9000
101

Notes: DZ represents the short column. X and S represent the bottom and top parts of the long column. t1 is the wall thickness of inner steel tube, Pe is the estimated
bearing capacity, Pu is the ultimate bearing capacity.

load maintained for 2 minutes before being unloaded once the data 2–3X collected data normally. However, for specimen DZ-2–2X, only the
stabilized. All data from the DH3816 data acquisition instrument were Z-5 and H-8 strain gauges collected data normally because the other
rebalanced before the formal loading. gauges were damaged during the test. The load-strain (N-ε) curves
Initially, the load for each stage was approximately 1/12 of the during the loading process of the specimens are shown in Figs. 2 to 5, it is
specimen’s expected bearing capacity. The loading time for each stage evident that the specimens exhibit a high capacity to continue bearing
was 1.5 minutes, with a holding time of 2 minutes. When the load the load and demonstrate good ductility after reaching the ultimate
reached 90 % of the expected bearing capacity, the loading speed was bearing capacity. The bearing capacity does not decrease rapidly after
reduced. The load for each stage was then approximately 1/20 of the reaching the ultimate bearing capacity, maintaining a long ductility
expected bearing capacity, while the loading and holding times platform. This is because the internal and external steel tubes provide
remained unchanged. Strain data were recorded by the data acquisition significant support and confinement to the concrete after reaching the
system DH3816, load values were collected by the MTS data acquisition ultimate bearing capacity, allowing the UCFDST column specimens to
system, and the longitudinal displacement of the specimen was recorded maintain overall integrity and function together effectively.
synchronously. As the specimen approached failure, loading continued The N-Δ curves of the four specimens after fire are shown in Fig. 6. It
at a slow speed until the ratio of axial compressive deformation excee­ is concluded that as the maximum temperature experienced by the
ded 5 % and the steel tube showed signs of wrinkling, at which point the specimen increases, the ultimate bearing capacity of the specimen de­
test was terminated. creases, the elastic segment shortens, and the elastic modulus decreases.
DZ-2–1X was used as a typical specimen for mechanical performance
2.4. Experimental results and analysis analysis, and the N-Δ curve was shown in Fig. 7. It can be seen that when
the axial load is small, the axial deformation is small, and the N-Δ
All strain gauges arranged in specimens DZ-2–1X, DZ-2–3S, and DZ- relationship is basically linear. When the applied load reaches 75~85 %

Fig. 2. N-ε curve of DZ-2–1X.

4
X. Lyu et al. Structures 79 (2025) 109311

Fig. 3. N-ε curve of DZ-2–2X.

Fig. 4. N-ε curve of DZ-2–3S.

of the ultimate load of the specimen, the displacement change of the end gradually increases, resulting in a curved load-displacement
gradually increases. The load corresponding to the end point of the relationship up to point B.
elastic stage of the N-Δ curve is defined as the yield load of the specimen. (c) BC phase: In the plastic stage, once the load exceeds point B, a
After the applied load reaches the yield load, the specimen still has a significant increase in displacement requires only a small in­
high bearing capacity. With the increase of axial deformation, its crease in load, continuing until the highest point, C, resembling a
bearing capacity is gradually increasing and tends to be gentle, showing yield platform.
good ductility. Four typical points of O, A, B and C are selected on the (d) Post-C phase: In the decline stage, after the load exceeds point C,
curve, and the stress process of the specimen can be divided into four the load gradually decreases as the displacement continues to
stages: increase.

(a) OA phase: In the elastic stage, the steel tube and the sandwich As seen in Fig. 8, when the specimen reaches the failure state, it
concrete are jointly stressed. The entire section of the specimen is exhibits buckling of the steel tube, typically with 2–3 buckling points.
compressed and remains in the elastic stage, with displacement When the load reaches approximately 80 % of the ultimate bearing ca­
increasing linearly as the load increases. pacity, local crushing sounds from the internal concrete and shedding of
(b) AB phase: In the elastic-plastic stage, as the load increases past the external oxide layer are observed. At this stage, shear slip lines
point A, displacement is no longer proportional to the load but appear on the steel tube wall, increasing in number as the load increases.

5
X. Lyu et al. Structures 79 (2025) 109311

Fig. 5. N-ε curve of DZ-2–3 X.

specimen. This buckling development progresses slowly until the spec­

imen ultimately fails.
After stripping part of the outer steel tube of the specimen, it was
observed that the ultra-high strength concrete at the bulging areas of the
steel tube, located at the upper and lower ends of the specimen, was
severely fragmented. There was also local fragmentation of concrete in
the middle bulging position, though the concrete still maintained good
overall integrity. Additionally, from the buckling shape of the inner
circular steel tube, there was an evident shear failure phenomenon, as
shown in Fig. 9. The inner circular steel tube exhibited two local con­
caves, which corresponded to the fragmentation positions of the con­
crete. The test results indicate that each bulge of the outer steel tube was
Fig. 6. N-Δ curve of specimens. caused by the local buckling of the tube due to the expansion of the
internal concrete, resulting in significant fragmentation and extrusion
flow deformation of the concrete.

3. Finite element model analysis

The numerical simulation adopts the sequential thermal-mechanical

coupling method. The model sets two analysis steps, step-1 and step-2.
The temperature field calculation results are introduced in step-1, and
the displacement load is applied in step-2. Considering that the defor­
mation of steel tube in the process of processing, transportation and
cutting in practical application will inevitably affect the axial
compression performance of UCFDST members, the maximum deflec­
tion is 1/1000 of the height of steel tube considering the influence of
geometric initial defects of members.[38].
Fig. 7. N-Δ curve of typical specimen.

3.1. Material constitutive relationship

Local buckling first occurs at unfavorable positions, such as near the
ends due to edge effects or internal concrete defects caused by the fire. The thermal parameters of steel and ultra-high strength concrete
Most initial local buckling is observed near the upper or lower end mainly include thermal conductivity, specific heat, density and thermal
plates. The development of local buckling is generally slow, but as the expansion coefficient, which are set by.[39,40]. The density of steel is
load continues to increase, buckling also appears in the middle of the little affected by temperature, so the density of steel in this modeling is

6
X. Lyu et al. Structures 79 (2025) 109311

Fig. 8. Failure mode of specimens.

Fig. 9. Internal and external failure modes of typical specimens.

ρs = 7850 kg/m3. stage is Es(T)=Es= 2.06 × 105 MPa, strengthening stage E´ s(T)=
The temperature of a fire significantly affects the properties of steel 0.01Es(T), εy(T)= fy(T)/ Es(T).
materials. However, after cooling, the physical properties of steel, such After a fire, interlayer ultra-high strength concrete undergoes irre­
as strength and elastic modulus, largely recover. Based on test results, versible chemical changes, resulting in significant degradation of its
the constitutive relationship of steel materials after natural cooling post- mechanical properties, which cannot be restored. In this paper, the
fire is determined using the double broken line model.[36], as follows. formula recommended.[41] is used to determine the constitutive rela­
{ tionship of ultra-high strength concrete after fire. The influence of steel
Es (T)ε ε ≤ εy (T) tube on the constraint effect of internal concrete is considered in the
σ= [ ] (1)
fy (T) + Eʹs (T) ε − εy (T) ε > εy (T) recommended formula, but the applicable range of concrete strength is
from 30 MPa to 90 MPa, and the concrete strength in UCFDST short
f y (T) = f y (1.01 − 4 × 10− 4 T + 10− 6 T2 − 10− 9 T) column specimens studied in this paper exceeds 160 MPa. Therefore, it
(2)
20∘ C ≤ T ≤ 1000∘ C is necessary to modify the recommended formula. Through the concrete
material test after fire, the relationship between the mechanical prop­
In the formula (1), T represents the maximum temperature (℃)
erties of materials such as compressive strength and elastic modulus and
experienced by the steel; fy(T) is the yield strength of the steel after being
the highest temperature after fire is obtained.[42]. Considering the in­
subjected to the highest temperature T and subsequently naturally
fluence of temperature, the basic form of the concrete’s stress-strain
cooled; Es(T) is the elastic modulus of steel after experiencing the highest
(σ-ε) relationship is selected based on its normal temperature form.
temperature T and natural cooling; E´ s(T) is the elastic modulus of the
Using the method recommended in reference.[36], the parameters σ0(T)
steel in the strengthening phase after being exposed to the highest
and ε0(T) in the expression are modified. The applicable range of the
temperature T and natural cooling; εy(T) is the yield strain of the steel
modified concrete is from 100 MPa to 180 MPa, as follows.
following the highest temperature T and natural cooling.
Formula 2 is derived from post-fire material property tests on steel [ ( )2 ]
2ε ε
and subsequent formula fitting. The elastic modulus of steel in the elastic σ = σ0 (T) × − , ε ≤ ε0 (T) (3)
ε0 (T) ε0 (T)

7
X. Lyu et al. Structures 79 (2025) 109311

( ) [ ] equation and yield surface equation are as follows: an expansion angle of

⎧ ε 0.1ξT ⎫
⎪
⎪ σ0 (T) × 1 − q + σ 0 (T) × q × , ξT ≥ 1.12 ⎪
⎪ 30, a flow eccentricity of 0.1, a biaxial isobaric yield strength to uniaxial
⎪ ε0 (T) ⎪
⎨ ⎬ compressive strength ratio of 1.6, and a ratio of the second stress
σ= ε 1 ε invariant on the tension and compression meridian of 2/3.
⎪
⎪
⎪ σ0 (T) × × [ ]2 , ξT < 1.12 ⎪
⎪
⎪
⎩ ε0 (T) ε ⎭
β × ε0 ε(T) − 1 +
ε0 (T) 3.4. Model validation
> ε0 (T)
To verify the accuracy and reasonableness of the constitutive model
(4)
and modeling method used in this paper, the simulation results are
In the formula: compared with the test results. Fig. 11 is the comparison between the
[ ( )9.55 ( )0.45 ] finite element calculation results of the temperature field of the
T 24 ( )
σ 0 (T) = fcr (T) × 1 + 1 − × × 0.5ξT − 0.044ξ2T component and the previous test results. Points 1,2 and 3 represent the
1000 fc
contact points between the sandwich concrete and the inner steel tube,
(5) the midpoint of the concrete thickness direction, and the contact points
( ) between the concrete and the outer steel tube, respectively. In the
f cr (T) = 0.7637 − 2 × 10− 5 T − 10− 6 T2 × f c , 20∘ C ≤ T ≤ 800∘ C (6)
simulation of component temperature field, CZ-2–1, the highest tem­
[ ( )] peratures of points 1,2 and 3 are 382◦ C, 530◦ C, 702℃ respectively. CZ-
fc − 24
ε0 (T) = εcr (T) 1400 + 800 × × ξ0.2 (7) 2–2, the highest temperatures of points 1,2 and 3 are 284◦ C, 480◦ C,
24 T
504℃ respectively. CZ-2–3, the highest temperatures of points 1,2 and 3
( ) are 78◦ C, 130◦ C, 165℃ respectively. Fig. 12 presents the comparison of
εcr (T) = 0.886 + 3.2 × 10− 3 T − 3 × 10− 6 T 2 (1300 + 12.5f c ) 20∘ C ≤ T
the load-displacement relationship curves between the finite element
≤ 800∘ C results and the test results. Fig. 13 illustrates a schematic comparison of
(8) the calculated failure mode of the finite element model with the test
results.
In the formula: σ 0(T) represents the peak stress of ultra-high strength
Fig. 12 (a) to (d) show the comparison between the test values and
concrete, considering the confinement effect after experiencing the
the calculated values of the N-Δ curve of the specimen. It can be
highest temperature T and natural cooling. ε0(T) denotes the peak strain
observed that after the fire, the bearing capacity of the test specimen
of ultra-high strength concrete under the same conditions. ξT is the
exceeds the yield load, displaying a long "yield platform" where the
confinement coefficient of the hoop after exposure to the highest tem­
displacement increases significantly while the load remains relatively
perature T and subsequent cooling. fcr(T) is the standard value of re­
constant. Using DZ-2–3X as an example, after yielding, the bearing ca­
sidual strength of ultra-high strength concrete following the highest
pacity of the specimen increases slightly with the increase in axial
temperature T and natural cooling. εcr(T) represents the peak strain of
deformation, indicating significant plastic hardening. This demonstrates
ultra-high strength concrete after experiencing the highest temperature
that although the specimen has experienced a fire and the mechanical
T and natural cooling.
properties of the sandwich concrete cannot be fully restored, the
confinement effect of the inner and outer steel tubes on the sandwich
3.2. Establishment of the temperature field concrete is pronounced, allowing the three components to work together
in coordination to bear the external load. The overall ductility perfor­
The temperature field model of the component is established. The mance of the specimen is fully utilized, and the calculated curve accu­
steel tube and stiffening rib use the four-node reduced integration shell rately simulates this behavior, aligning well with the test curve. This
element (S4R). The sandwich ultra-high-strength concrete employs the indicates that the constitutive model and modeling method used in this
eight-node reduced integration three-dimensional solid element paper are reasonable.
(C3D8R). The absolute zero is − 273℃, and the Stefan-Boltzmann con­ Therefore, the finite element model can be reliably used to conduct
stant is 5.67 × 10− 8 W/(m2⋅K4). The type of mutual contact between the parametric analysis of UCFDST short columns under axial compression
inner and outer steel tubes and the core concrete is set to Tie connection. after a fire.
The convective heat transfer coefficient is 25 W/(m2⋅℃), and the
comprehensive radiation coefficient is 0.5 [43,44]. The ISO-834 heating 4. Parametric analysis
curve is input into the amplitude table, and the amplitude is set in the
heat transfer and surface heat exchange of the component. The initial Based on the finite element model established above, the following
ambient temperature was set to 20℃ variable parameters were selected: concrete cube compressive strength
fcu (100, 120, 140, 160 MPa), hollow ratio χ (0.43, 0.61, 0.72), yield
3.3. Establishment of the force field strength of outer steel tube fy (235, 355, 460, 590 MPa), fire exposure
time tfire (60, 90, 120, 180 minutes), slenderness ratio λ (11 18 24 30),
The general finite element software ABAQUS is utilized for modeling and wall thickness of outer steel tube t (6 8 12 16). Eighteen groups of
and calculations. During model establishment, different attributes are axial compression model members of UCFDST columns were created
assigned to each component of the specimen. The end plate is modeled with these parameters. The specific parameters of the members are
as a rigid body. The contact interface between the ultra-high-strength detailed in Table 4.
concrete and the inner and outer steel tubes uses "hard" contact in the Data analysis was conducted using the control variable method,
normal direction and "penalty" contact in the tangential direction, with a where only one parameter is varied at a time while keeping the others
friction coefficient of 0.3.[45]. constant. This approach allows for the analysis of the influence of each
Through mesh size sensitivity analysis, the final mesh generation parameter on the ultimate bearing capacity of UCFDST short columns
method is determined to ensure both accuracy and computational effi­ after fire exposure.
ciency.[46]. The specimen’s mesh division is shown in Fig. 10. During
calculations, steel is modeled using an isotropic elastic-plastic model 4.1. Cube compressive strength of concrete
that satisfies the Von Mises yield criterion and incorporates the relevant
flow rule and isotropic hardening rule under monotonic load.[47]. The In Table 3, the fire exposure time, yield strength of the outer steel
basic parameters for the concrete model’s plastic potential energy tube, hollow ratio, slenderness ratio, and wall thickness of the outer steel

8
X. Lyu et al. Structures 79 (2025) 109311

Fig. 10. Specimen meshing.

Fig. 11. Comparison of experimental and calculated values of component temperature field [35].

9
X. Lyu et al. Structures 79 (2025) 109311

Fig. 12. Comparison of experimental and calculated values of N-Δ curve.

Fig. 13. Comparison between finite element analysis and experiment.

tube for UCFDST-1, UCFDST-2, UCFDST-3, and UCFDST-4 members are From Fig. 14, it can be observed that the ultimate bearing capacity of
consistent. The corresponding ultra-high strength concrete strengths are the component shows a slight improvement with an increase in concrete
100 MPa, 120 MPa, 140 MPa, and 160 MPa, respectively. The load and cube compressive strength. Specifically:
displacement values for different ultra-high strength concrete strengths
under axial load were extracted through finite element simulation,
shown in Fig. 14.

10
X. Lyu et al. Structures 79 (2025) 109311

Table 4
Parameters of finite element model.
Specimen L t t1 D D1 tfire fy fcu χ λ
(mm) (mm) (mm) (mm) (mm) (min) (Mpa) (Mpa)

UCFDST− 1 657.3 16 6.3 219.1 114.3 60 355 100 0.61 11

UCFDST− 2 657.3 16 6.3 219.1 114.3 60 355 120 0.61 11
UCFDST− 3 657.3 16 6.3 219.1 114.3 60 355 140 0.61 11
UCFDST− 4 657.3 16 6.3 219.1 114.3 60 355 160 0.61 11
UCFDST− 5 657.3 16 6.3 219.1 80.4 60 355 160 0.43 11
UCFDST− 6 657.3 16 6.3 219.1 134.6 60 355 160 0.72 11
UCFDST− 7 657.3 16 6.3 219.1 114.3 60 235 160 0.61 11
UCFDST− 8 657.3 16 6.3 219.1 114.3 60 460 160 0.61 11
UCFDST− 9 657.3 16 6.3 219.1 114.3 60 590 160 0.61 11
UCFDST− 10 657.3 16 6.3 219.1 114.3 90 355 160 0.61 11
UCFDST− 11 657.3 16 6.3 219.1 114.3 120 355 160 0.61 11
UCFDST− 12 657.3 16 6.3 219.1 114.3 180 355 160 0.61 11
UCFDST− 13 1112.1 16 6.3 219.1 114.3 60 355 160 0.61 18
UCFDST− 14 1483.6 16 6.3 219.1 114.3 60 355 160 0.61 24
UCFDST− 15 1853.4 16 6.3 219.1 114.3 60 355 160 0.61 30
UCFDST− 16 657.3 12 6.3 219.1 114.3 60 355 160 0.61 11
UCFDST− 17 657.3 8 6.3 219.1 114.3 60 355 160 0.61 11
UCFDST− 18 657.3 6 6.3 219.1 114.3 60 355 160 0.61 11

Fig. 14. N-Δ curves of different concrete cube compressive strength. Fig. 15. N-Δ curves of different yield strength of outer steel tubes.

(1) When the compressive strength increased from 100 MPa to Q355, and the capacity with Q355 is 13.90 % higher than with Q235.
120 MPa, the ultimate bearing capacity increased from Overall, the ultimate bearing capacity increases by 32.31 % from
7213.42 kN to 7610.8 kN, a rise of 5.51 %. Q235 to Q590, indicating that the yield strength of the outer steel tube
(2) When the compressive strength increased from 120 MPa to significantly affects the component’s ultimate bearing capacity. All four
140 MPa, the ultimate bearing capacity increased from groups of members exhibited a decrease in load-bearing capacity after
7610.8 kN to 7868.36 kN, a rise of 3.38 %. reaching the ultimate value. The decline is more gradual for members
(3) When the compressive strength increased from 140 MPa to with higher yield strengths (Q590 and Q460) compared to those with
160 MPa, the ultimate bearing capacity increased from lower strengths (Q355 and Q235).
7868.36 kN to 8204.53 kN, a rise of 4.27 %. Taking the load-displacement curve of the UCFDST-8 component as
an example (Fig. 16), five points A, B, C, D, and E are highlighted. The
It can be concluded that while increasing the concrete cube bearing capacity at point B is 1122.04 kN higher than at point A, a
compressive strength slightly improves the ultimate bearing capacity of 20.99 % increase. From point B to C, the capacity increases by
the component, the effect is not significant. 449.88 kN (3.88 %). From point C to D, the capacity decreases by
42.21 kN (0.54 %), and from point D to E, it decreases by 57.83 kN
(0.62 %). The overall decrease from point C is 2.5 %, indicating a minor
4.2. Yield strength of outer steel tube reduction.
Thus, increasing the yield strength of the outer steel tube enhances
In Table 3, the fire exposure time, concrete cube compressive
strength, hollow ratio, slenderness ratio, and wall thickness of the outer
steel tube for UCFDST-7, UCFDST-4, UCFDST-8, and UCFDST-9 mem­
bers are consistent. The yield strength of the outer steel tube varies as
Q235, Q355, Q460, and Q590, respectively. The load and displacement
values for different ultra-high strength concrete strengths under axial
load were extracted through finite element simulation, shown in Fig. 15.
According to Fig. 15, the ultimate bearing capacity of the component
increases with higher yield strength of the steel. Specifically, the ulti­
mate bearing capacities corresponding to steel yield strengths Q235,
Q355, Q460, and Q590 are 7203.49 kN, 8204.53 kN, 9464.68 kN, and
10855.41 kN, respectively. The ultimate bearing capacity of the
component with steel yield strength Q590 is 14.70 % higher than that
with Q460, while the capacity with Q460 is 15.36 % higher than with Fig. 16. N-Δ curve (UCFDST-8).

11
X. Lyu et al. Structures 79 (2025) 109311

both the yield strength and the ultimate bearing capacity of the
component. After reaching peak capacity, the curve’s gentle decline
suggests improved ductility.

4.3. Fire exposure time

In Table 3, the concrete cube compressive strength, yield strength of

the outer steel tube, hollow ratio, slenderness ratio, and wall thickness of
the outer steel tube for UCFDST-4, UCFDST-10, UCFDST-11, and
UCFDST-12 members are consistent. The corresponding fire exposure
times are 60 minutes, 90 minutes, 120 minutes, and 180 minutes,
respectively. The load and displacement values for different fire expo­
Fig. 18. N-Δ curves of different hollow ratios.
sure times under axial load were extracted through finite element
simulation, shown in Fig. 17.
From Fig. 17, it is evident that the ultimate bearing capacity of the decreased from 8204.53 kN to 7592.66 kN, representing a
members gradually decreases with increased fire exposure time. decrease of 7.4 %.
Specifically: (3) The ultimate bearing capacity of the members with a hollow ratio
of 0.72 decreased from 8829.13 kN to 7592.66 kN compared to
(1) When the fire exposure time increases from 60 minutes to those with a hollow ratio of 0.43, showing a reduction of 14 %.
90 minutes, the ultimate bearing capacity decreases from
8204.53 kN to 7476.89 kN, a reduction of 7.65 %. Therefore, a smaller hollow ratio corresponds to a greater ultimate
(2) When the fire exposure time increases from 90 minutes to bearing capacity. This is because, the decrease in the actual cross-
120 minutes, the ultimate bearing capacity decreases from sectional area subjected to force of concrete with a higher hollow ratio
7476.89 kN to 7039.52 kN, a reduction of 7.12 %. (0.72), leading to the reduction in the ultimate bearing capacity.
(3) When the fire exposure time increases from 120 minutes to
180 minutes, the ultimate bearing capacity decreases from 4.5. Slenderness ratio
7039.52 kN to 6507.51 kN, a reduction of 7.56 %.
In Table 3, UCFDST-4, UCFDST-13, UCFDST-14, and UCFDST-15
Therefore, the ultimate bearing capacity of the components de­ members have consistent concrete cube compressive strength, yield
creases as the fire exposure time increases. When the maximum bearing strength of the outer steel tube, hollow ratio, fire exposure time, and
capacity is reached, the curve tends to level off. This behavior is wall thickness of the outer steel tube. The corresponding slenderness
attributed to the degradation of concrete and steel performance due to ratios are 11, 18, 24, and 30, respectively. The load and displacement
prolonged high-temperature exposure. values for different slenderness ratios under axial load were extracted
through finite element simulation, shown in Fig. 19.
4.4. Hollow ratio Fig. 19 illustrates that the ultimate bearing capacity of the compo­
nent decreases as the slenderness ratio increases:
In Table 3, UCFDST-5, UCFDST-4, and UCFDST-6 members have
consistent concrete cube compressive strength, yield strength of the (1) When the slenderness ratio increases from 11 to 18, the ultimate
outer steel tube, fire exposure time, slenderness ratio, and wall thickness bearing capacity decreases from 8204.53 kN to 7552.55 kN, a
of the outer steel tube. The corresponding hollow ratios are 0.43, 0.61, reduction of 7.95 %.
and 0.72, respectively. The load and displacement values for different (2) When the slenderness ratio increases from 18 to 24, the ultimate
hollow ratios under axial load were extracted through finite element bearing capacity decreases from 7552.55 kN to 6841.44 kN, a
simulation, shown in Fig. 18. reduction of 9.42 %.
From Fig. 18, it can be observed that the ultimate bearing capacity of (3) When the slenderness ratio increases from 60 24–30, the ultimate
components decreases with increasing hollow ratio: bearing capacity decreases from 6841.44 kN to 5643.79 kN, a
reduction of 17.51 %.
(1) The ultimate bearing capacity of the component with a hollow
ratio of 0.61 decreased from 8829.13 kN to 8204.53 kN The data show that members with a higher slenderness ratio exhibit a
compared to the component with a hollow ratio of 0.43, reflect­ more pronounced decrease in ultimate bearing capacity. In contrast,
ing a decrease of 7.07 %. members with a slenderness ratio of 11 reach their ultimate bearing
(2) Compared to the members with a hollow ratio of 0.61, the ulti­ capacity and show a more gradual decline in the curve.
mate bearing capacity of those with a hollow ratio of 0.72 Fig. 20 shows the failure modes of two UCFDST columns with

Fig. 17. N-Δ curves of different time of fire. Fig. 19. N-Δ curves of different slenderness rations.

12
X. Lyu et al. Structures 79 (2025) 109311

different slenderness ratios under axial compression. When the slen­

derness ratio is small, the buckling failure occurs in the middle of the
component (Fig. 20(a)). When the slenderness ratio is large, the overall
instability failure of the component occurs (Fig. 20(b)). Thus, the slen­
derness ratio significantly affects the ultimate bearing capacity of
UCFDST columns after fire, because the larger the slenderness ratio of
the component, the worse the stability.

4.6. Wall thickness of outer steel tube

In Table 3, UCFDST-4, UCFDST-16, UCFDST-17, and UCFDST-18

members have consistent concrete cube compressive strength, yield
Fig. 21. N-Δ curves of different wall thickness of outer steel tube.
strength of the outer steel tube, hollow ratio, fire exposure time, and
slenderness ratio. The corresponding wall thicknesses of the outer steel
tube are 11 mm, 18 mm, 24 mm, and 30 mm, respectively. The load and 6830.23 kN, a reduction of 9.10 %.
displacement values for different wall thicknesses under axial load were When the wall thickness of outer steel tube decreases from 8 mm to
obtained through finite element simulation, shown in Fig. 21. 6 mm, the ultimate bearing capacity decreases from 6830.23 kN to
Fig. 21 shows that the ultimate bearing capacity of the component 6149.22 kN, a reduction of 9.97 %.
decreases as the wall thickness of the outer steel tube decreases, indi­ These results indicate that the wall thickness of the outer steel tube
cating a positive correlation: significantly impacts the ultimate bearing capacity of UCFDST columns
When the wall thickness of outer steel tube decreases from 16 mm to after fire exposure.
12 mm, the ultimate bearing capacity decreases from 8204.53 kN to
7514.12 kN, a reduction of 8.41 %.
When the wall thickness of outer steel tube decreases from 12 mm to
8 mm, the ultimate bearing capacity decreases from 7514.14 kN to

Fig. 20. Failure modes of members with different slenderness ratio.

13
X. Lyu et al. Structures 79 (2025) 109311

5. Simplified calculation formula of residual bearing capacity Table 5

after fire Values of coefficients d1~d10.
Coefficient d1 d2 d3 d4 d5
Through the analysis of the above influencing factors, the influence
Values 19.30 10.44 0.13 − 45.67 − 14848.37
law of each influencing factor on the residual bearing capacity of UCFDST Coefficient d6 d7 d8 d9 d10
columns after fire is obtained. The simplified calculation formula of re­ Values 12811.94 − 3.10 − 14.10 8.92 − 197.77
sidual bearing capacity (Nc) of UCFDST columns after fire is obtained. The
scope of application of the proposed formula (9): 100 MPa≤fcu≤ 160 MPa,
235 MPa≤fy≤ 590 MPa, 60 minutes≤tfire≤ 120 minutes, 0.43 ≤χ ≤ 0.72,
11 ≤λ≤ 30, 6 mm≤t ≤ 16 mm.

Nc = d1 f cu + d2 f y + d3 t 2fire + d4 t fire + d5 χ 2 + d6 χ + d7 λ2 + d8 λ + d9 t2 + d10

(9)
In the formula: d1 ~ d10 are coefficients, and the values are shown in
Table 5.(Fig. 22)
Fig. 21 shows that the calculation results of the simplified formula
for the residual bearing capacity of UCFDST columns after fire are
compared with the numerical simulation results. The correlation coef­
ficient is 0.99981, the average value of the ratio of the two is 1.0002, and
the mean square error is 0.038. It can be seen from Fig. 21 that the ac­
curacy of the simplified formula of residual bearing capacity is high, Fig. 22. Comparison of calculation results and numerical simulation results of
which can provide reference for the calculation of residual bearing ca­ residual bearing capacity.
pacity of this type of component after fire.
steel tube has the most significant increase in the bearing ca­
6. Conclusions pacity. When the slenderness ratio of the member increases from
11 to 30, the ultimate bearing capacity decreases from
(1) Analysis of the N-Δ curve and overall failure mode reveals that 8204.53 kN to 5643.79 kN, a decrease of 31.21 %, indicating that
the UCFDST axial compression specimens maintain high bearing the slenderness ratio has a great impact on the ultimate bearing
capacity and ductility even after exposure to fire. The N-Δ curve capacity of the member. This is because the larger the slenderness
for the ultra-high strength concrete filled steel tube columns ratio of the member, the worse its stability. The simplified
shows a decrease in ultimate bearing capacity along with an in­ calculation formula of residual bearing capacity of UCFDST col­
crease in deformation. Additionally, the curve’s elastic section umns after fire are obtained by regression. The formula has high
shortens progressively, the elastic-plastic section expands, and accuracy and can provide reference for the evaluation of residual
the descending section becomes more gradual. bearing capacity of UCFDST columns.
(2) The working mechanism and failure mode of UCFDST short col­
umn specimens under axial compression after fire were exam­ CRediT authorship contribution statement
ined. Post-fire, the failure mode primarily involves symmetrical
buckling, with some instances of shear failure. This behavior re­ Xiao Lyu: Software, Methodology, Conceptualization. Er-Feng Du:
sults from the irreparable damage sustained by the ultra-high Writing – review & editing. Xiao-Xi Hu: Visualization, Investigation.
strength concrete during the fire resistance test prior to the Wang Yuzhuo: Writing – review & editing. Peng Zhang: Writing –
axial compression test. original draft, Data curation. Ze-Jian Liu: Validation, Software.
(3) From the failure mode and crack distribution of the internal
concrete of the ultra-high strength concrete filled steel tube after Declaration of Competing Interest
the fire, it can be seen that the internal concrete of the ultra-high
strength concrete filled steel tube after the fire is mainly locally No conflict of interest exits in the submission of this manuscript, and
crushed, and the crushing position is consistent with the bulging manuscript is approved by all authors for publication. I would like to
position of the external steel tube. declare on behalf of my co-authors that the work described was original
(4) The constitutive relationships for steel and ultra-high strength research that has not been published previously, and not under
concrete post-fire have been revised. The updated finite element consideration for publication elsewhere, in whole or in part. All authors
model aligns well with the N-Δ curve obtained from experimental listed have approved the manuscript that is enclosed.
results. The finite element model can be applied to the analysis of
the stress mechanism of such specimens after fire. Acknowledgments
(5) The effects of cube compressive strength of concrete, yield
strength of outer steel tube, fire exposure time, hollow ratio, The research work described in this paper is part of Project
slenderness ratio and wall thickness of outer steel tube on the 52478532 supported by the National Natural Science Foundation of
bearing capacity of specimens after fire were analyzed. The re­ China, Project ZR2021ME190 supported by Natural Science Foundation
sults show that the cube compressive strength of concrete is of Shandong Province, Project 2024TSGC0537 supported by Key
increased from 100 Mpa to 160 Mpa, and the ultimate bearing Research and Development Plan of Shandong Province. The financial
capacity is increased from 7213.42 kN to 8204.53 kN, with an support is highly appreciated. The authors are grateful to Professor J. Y.
increase of 13.74 %, indicating that increasing the cube Richard Liew, Department of Civil and Environmental Engineering,
compressive strength of concrete, the ultimate bearing capacity National University of Singapore, for his useful comments.
of the component increases less. When the yield strength of the
outer steel tube increases from 235 MPa to 590 MPa, the ultimate
bearing capacity increases from 7203.49 kN to 10855.41 kN, an
increase of 50.70 %, indicating that the yield strength of the outer

14
X. Lyu et al. Structures 79 (2025) 109311

References [24] Yang YF, Han LH. Concrete filled double-skin tubular columns under fire. Mag
Concr Res 2008;60(3):211–22. https://doi.org/10.1680/macr.2007.00074.
[25] Lu H, Han LH, Zhao XL. Fire performance of self-consolidating concrete filled
[1] Han LH, Lam D, Nethercot D. Design guide for concrete filled double skin steel
double skin steel tubular columns: experiments. Fire Saf J 2010;45(2):106–15.
tubular. Structures. CRC Press; 2018. https://doi.org/10.1201/978042944040.
https://doi.org/10.1016/j.firesaf.2009.12.001.
[2] Haas TN, Amika AU, Koen J, Patel VI. Parameter priority of circular concrete-filled
[26] Lu H, Zhao XL, Han LH. FE modelling and fire resistance design of concrete filled
double-skin tubular slender beam-columns under eccentric loading. Struct Concr
double skin tubular columns. J Constr Steel Res 2011;67(11):1733–48. https://doi.
2024. https://doi.org/10.1002/suco.202300666.
org/10.1016/j.jcsr.2011.04.014.
[3] Zhang JH, Hassanein MF, Cashell KA, Hadzima-Nyarko M, Xu Y, Shao YB.
[27] Lu H, Zhao XL, Han LH. Testing of self-consolidating concrete filled double skin
Experimental and numerical investigation on the behaviour of square concrete
tubular stub columns exposed to fire. J Constr Steel Res 2010;66(8-9):1069–80.
filled cold-formed double-skin steel stiffened tubular short columns. Eng Struct
https://doi.org/10.1016/j.jcsr.2010.03.004.
2024;303:117560. https://doi.org/10.1016/j.engstruct.2024.117560.
[28] Yao Y, Li H, Tan K. Theoretical and numerical analysis to concrete filled double
[4] Zarringol M, Patel VI, Liang QQ, Hassanein MF, Ahmed M. Machine-learning-based
skin steel tubular columns under fire conditions. Thin-Walled Struct 2016;98:
predictive models for concrete filled double skin tubular columns. Eng Struct 2024;
547–57. https://doi.org/10.1016/j.tws.2015.10.024.
304:117593. https://doi.org/10.1016/j.engstruct.2024.117593.
[29] Romero ML, Espinós A, Portolés JM, Hospitaler A, Ibañez C. Slender double-tube
[5] Ding R, Bai HH, Sun YL, Chen Y, Fan JS. Experimental and numerical study on
ultra-high strength concrete filled tubular columns under ambient temperature and
shear behaviour of concrete filled double-skin steel tubular structures. Eng Struct
fire. Eng Struct 2015;99:536–45. https://doi.org/10.1016/j.
2024;303:117587. https://doi.org/10.1016/j.engstruct.2024.117587.
engstruct.2015.05.026.
[6] Bhat P, Jamatia R. Compressive behavior of concrete filled double skin steel
[30] Wan CY, Zha XX, Dassekpo JBM. Analysis of axially loaded concrete filled circular
tubular columns: an analytical approach. J Constr Steel Res 2024;214:108483.
hollow double steel tubular columns exposed to fire. Fire Saf J 2017;88:1–12.
https://doi.org/10.1016/j.jcsr.2024.108483.
https://doi.org/10.1016/j.firesaf.2016.12.007.
[7] Tao Z, Han LH. Behaviour of concrete filled double skin rectangular steel tubular
[31] Lu H, Han LH, Zhao XL. Fire performance of self-consolidating concrete filled
beam–columns. J Constr Steel Res 2006;62(7):631–46. https://doi.org/10.1016/j.
double skin steel tubular columns: experiments. Fire Saf J 2010;45(2):106–15.
jcsr.2005.11.008.
https://doi.org/10.1016/j.firesaf.2009.12.001.
[8] Tao Z, Han LH, Zhao XL. Behaviour of concrete filled double skin (CHS inner and
[32] Li W, Wang T, Han LH. Seismic performance of concrete filled double-skin steel
CHS outer) steel tubular stub columns and beam-columns. J Constr Steel Res 2004;
tubes after exposure to fire: experiments. J Constr Steel Res 2019;154:209–23.
60(8):1129–58. https://doi.org/10.1016/j.jcsr.2003.11.008.
https://doi.org/10.1016/j.jcsr.2018.12.003.
[9] Ding F, Lu D, Lai Z, Liu X. Study on restraint coefficient of the stirrups-stiffened
[33] Liu X, Liu Z, Hui Y, Wang J. Research on the Concrete filled Double Skin Steel
square concrete filled double-skin steel tube axial compression stub columns.
Tubular (CFDST) columns subjected to axial force after fire. Structures 2023;57:
Structures 2024, February;60:105847. https://doi.org/10.1016/j.
105213. https://doi.org/10.1016/j.istruc.2023.105213.
istruc.2023.105847.
[34] Liu X, Liu S, Hui Y, Wang B. Mechanical properties of CFDST column under
[10] Wang X, Li T, Wang Y, Xu Y. Postblast damage assessment of concrete filled double-
eccentric compression after exposed to fire. Structures 2022;43:1200–15. https://
skin tube columns by intelligence-based techniques. J Constr Steel Res 2024;215:
doi.org/10.1016/j.istruc.2022.07.036.
108486. https://doi.org/10.1016/j.jcsr.2024.108486.
[35] Xiong MX, Liew JR. Fire resistance of high-strength steel tubes infilled with ultra-
[11] Manigandan R. Investigation on behavior of concrete filled double steel tubular
high-strength concrete under compression. J Constr Steel Res 2021;176:106410.
columns infilled with nanomaterial-based concrete subjected to axial compression.
https://doi.org/10.1016/j.jcsr.2020.106410.
Structures 2024, March;61:106122. https://doi.org/10.1016/j.
[36] Han LH. Concrete Filled Steel Tubular Structures-Theory and Practice. 3rd ed.
istruc.2024.106122.
Beijing: Science press; 2016.
[12] Yan XF, Ahmed M, Hassanein MF, He MN. Performance analysis and design of
[37] Li W, Chen B, Wang T. Seismic performance of concrete filled double-skin steel
circular high-strength concrete-filled double-skin aluminum tubular short columns
tubes after exposure to fire: analysis. J Constr Steel Res 2019;162:105753. https://
under axial loading. Struct Concr 2023;24(5):5677–96. https://doi.org/10.1002/
doi.org/10.1016/j.jcsr.2019.105753.
suco.202200768.
[38] Zhu T, Liang H, Lu Y, Li W, Zhang H. Axial behaviour of slender concrete filled
[13] Yu M, Zha X, Ye J, She C. A unified formulation for hollow and solid concrete filled
steel tube square columns strengthened with square concrete filled steel tube
steel tube columns under axial compression. Eng Struct 2010;32(4):1046–53.
jackets. Adv Struct Eng 2020;23(6):1074–86. https://doi.org/10.1177/
https://doi.org/10.1016/j.engstruct.2009.12.031.
1369433219888726.
[14] Han TH, Yoon KY, Kang YJ. Compressive strength of circular hollow reinforced
[39] Lie TT, Irwin RJ. Fire resistance of rectangular steel columns filled with bar-
concrete confined by an internal steel tube. Constr Build Mater 2010;24(9):1690–9.
reinforced concrete. J Struct Eng 1995;121(5):797–805. https://doi.org/10.1061/
https://doi.org/10.1016/j.conbuildmat.2010.02.022.
(ASCE)0733-9445(1995)121:5(797).
[15] An G, Wang R, Zhao H, Wang Y, Lam D, Li T. Tapered RAC-filled high-strength
[40] Lie, T.T., & Denham, E.M.A. Factors affecting the fire resistance of circular hollow
double-skin steel tubular stub columns: tests and numerical analysis. J Constr Steel
steel columns filled with bar-reinforced concrete. NRC-CNRC Internal Report,
Res 2024;215:108524. https://doi.org/10.1016/j.jcsr.2024.108524.
No.651, 1993. https://doi.org/10.4224/20358846.
[16] Le TT, Patel VI, Liang QQ, Huynh P, Ha NS. Numerical analysis of square concrete-
[41] Han LH, Yang H, Cheng SL. Residual strength of concrete filled RHS stub columns
filled double-skin tubular columns with outer stainless-steel tube. Struct Concr
after exposure to high temperatures. Adv Struct Eng 2002;5(2):123–34. https://
2022;23(5):2968–85. https://doi.org/10.1002/suco.202100649.
doi.org/1369-4332(2002)l:2< 123:RSOCFR> 2.0.TX;2-0.
[17] Wang JT, Sun Q, Li YW, Liu XH. Hysteretic performance of circular high-strength
[42] Lyu X, Jia GH, Shu GP, Zhang X, Du EF, Wang WM. Post-fire study on mechanical
concrete filled double skin steel tubular columns: experiment. Mar Struct 2024;93:
properties of damaged ultra-high strength concrete. Int J Steel Struct 2022;22(6):
103519. https://doi.org/10.1016/j.marstruc.2023.103519.
1918–29. https://doi.org/10.1007/s13296-022-00686-z.
[18] Manigandan R, Kumar M. Tests and numerical behavior of circular concrete filled
[43] Liu F, Yang H, Gardner L. Post-fire behaviour of eccentrically loaded reinforced
double skin steel tubular stub columns under eccentric loads. 000-000 Struct Eng
concrete columns confined by circular steel tubes. J Constr Steel Res 2016;122:
Mech 2023;88(3). https://doi.org/10.12989/sem.2023.88.3.287.
495–510. https://doi.org/10.1016/j.jcsr.2016.04.008.
[19] Deng R, Zhou XH, Wang YH, Bai JL, Deng XW. Experimental study on tapered
[44] Tao Z, Wang ZB, Han LH, Uy B. Fire performance of CFRP-strengthened concrete
concrete filled double skin steel tubular columns under torsion. Thin-Walled Struct
filled steel tubular columns. Methods Enzymol 2007;121(2007):228–34.
2022;177:109444. https://doi.org/10.1016/j.tws.2022.109444.
[45] Lyu X, Shu GP, Liew JR, Du EF. Fire resistance of steel tubular columns infilled
[20] Chen J, Wang J, Xie F, Jin WL. Behavior of thin-walled dodecagonal section double
with ultra-high strength concrete. Advanced Steel. Construction 2018;14(3):
skin concrete filled steel tubes under bending. Thin-Walled Struct 2016;98:
438–60. https://doi.org/10.18057/IJASC.2018.14.3.8.
293–300. https://doi.org/10.1016/j.tws.2015.10.002.
[46] Lyu X, Shu GP, Du EF. Parameters effect on predicting fire resistance of ultra-high
[21] Wang FC, Han LH, Li W. Analytical behavior of CFDST stub columns with external
strength concrete filled protected square steel tubular columns. Int J Steel Struct
stainless steel tubes under axial compression. Thin-Walled Struct 2018;127:
2020;20:1783–95. https://doi.org/10.1007/s13296-020-00410-9.
756–68. https://doi.org/10.1016/j.tws.2018.02.021.
[47] Mohammadbagheri S, Shekastehband B. Fire resistance of stiffened CFDST columns
[22] Ahmed M, Liang QQ, Patel VI, Hadi MN. Experimental and numerical studies of
after earthquake-induced damages. Thin-Walled Struct 2020;154:106865. https://
square concrete filled double steel tubular short columns under eccentric loading.
doi.org/10.1016/j.tws.2020.106865.
Eng Struct 2019;197:109419. https://doi.org/10.1016/j.engstruct.2019.109419.
[23] Li W, Cai YX. Performance of CFDST stub columns using high-strength steel
subjected to axial compression. Thin-Walled Struct 2019;141:411–22. https://doi.
org/10.1016/j.tws.2019.04.021.

15