Thin-Walled Structures 138 (2019) 485–495

Contents lists available at ScienceDirect

Thin-Walled Structures
journal homepage: www.elsevier.com/locate/tws

Full length article

Seismic behavior of circular tubed steel-reinforced concrete column to steel T

beam connections
⁎
Guozhong Chenga,b, Xuhong Zhoua,b, Jiepeng Liua,b, , Y. Frank Chenb
a
Key Laboratory of New Technology for Construction of Cities in Mountain Area (Chongqing University), Ministry of Education, Chongqing 400045, China
b
School of Civil Engineering, Chongqing University, Chongqing 400045, China

A R T I C LE I N FO A B S T R A C T

Keywords: The tubed steel-reinforced concrete (TSRC) column is a special type of SRC columns where the longitudinal
TSRC column reinforcement and reinforcement cage in the SRC column are replaced by a thin-walled steel tube. Without any
Connection traditional reinforcement, concrete pouring becomes easier in a TSRC member. Although the TSRC columns
Seismic behavior possess high load-carrying capacity and good ductility performance in seismic zones, the seismic behavior of
Design model
TSRC column to beam connections has not received much attention, which limits the application of TSRC col-
Joint strength
umns. This paper aims to investigate the seismic behavior of circular TSRC column to steel beam connections
with cross diaphragms. Four such connection specimens were tested under cyclic loading, considering the
thickness of joint tube, the height of extended tube, and the axial compression ratio of columns. Based on
experimental results, the failure progression, load-displacement curves, and stresses for the circular TSRC
column to steel beam connections are discussed. The favorable seismic performance for such connections is
demonstrated and a design model for determining the joint strength is proposed in this paper.

1. Introduction utilized. Without any traditional reinforcement, concrete pouring be-

comes easier in a TSRC member. Gan et al. [5] tested 6 TSRC column
Composite columns have been increasingly used in many modern specimens under the combined constant axial compression and lateral
structures. As a main type of composite columns, steel-reinforced con- cyclic loads and investigated their seismic behavior. The test results
crete (SRC) columns can provide considerable advantages compared to show that the flexural strength, ductility, plastic deformation capacity,
steel columns. Since the steel section is encased with concrete, it be- and energy dissipation capacity of circular TSRC columns are sig-
comes a solid section requiring no local buckling check on the steel nificantly higher than those of common SRC columns with the same
section and the fire resistance of a SRC column is also enhanced [1]. steel ratio and axial compressive load. Liu et al. [6] developed a non-
However, longitudinal and transverse reinforcement is needed to pre- linear three-dimensional finite element model to simulate the hysteretic
vent the concrete from spalling off [1,2]. The transverse ties cannot behavior of TSRC columns and proposed a design formula to predict the
effectively prevent the longitudinal rebars and the flanges of the en- shear strength of short square TSRC columns. Qi et al. [7] carried an
cased steel shape from buckling in a situation where the SRC column is experimental study on the behavior of stub TSRC columns subjected to
subjected to combined high axial and cyclic lateral loads. Thus, the axial compressive loads, including 14 circular and 15 square specimens.
permitted axial compression ratio for SRC columns is the same as that The test results indicate that height to diameter/width ratio of the
for RC columns [3], which limits the use of high strength concrete. discontinuous tube has little effect on the failure mode and strength of
Moreover, the reinforcement in a SRC column also may hinder concrete TSRC columns. Yan et al. [8] investigated the axial behavior and sta-
pouring, especially at beam-column connections (Fig. 1(a)). bility strength of circular TSRC columns by testing 8 specimens. A series
The tubed steel-reinforced concrete (TSRC) column proposed by of experiments have been conducted to study the behavior of TSRC
Zhou et al. [4] is a special type of SRC columns where the longitudinal columns under eccentric compression, followed by the development of
and transverse reinforcement in the SRC column is replaced by a thin- axial load versus moment interaction diagrams [9–12].
walled steel tube (Fig. 1(b)). No direct axial load is applied on the steel Based on the literature review, previous studies mainly focused on
tube as the steel tube is discontinued at the connection ends with the the static and dynamic behaviors of TSRC columns. However, the
column. As such, the effectiveness of the steel tube in confining the core seismic behavior of TSRC column to beam connections has not received
concrete is maximized, enabling the high strength concrete to be fully much attention. To the knowledge of the authors, no work has been

⁎
Corresponding author.

https://doi.org/10.1016/j.tws.2018.10.041
Received 15 May 2018; Received in revised form 22 October 2018; Accepted 30 October 2018
Available online 27 December 2018
0263-8231/ © 2018 Elsevier Ltd. All rights reserved.
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Nomenclature tj thickness of joint tube

tw width of steel web
Ac cross-sectional area of core concrete Vc horizontal force resisted by the concrete compression strut
As cross-sectional area of joint tube Vcal calculated joint strength
Av effective cross-sectional area of joint tube Vj shear force
C coefficient set as 82.5fs Vt horizontal force resisted by the joint tube
dp increment of accumulated equivalent plastic strain Vu maximum shear force
dαij increment of backstress tensor Vw horizontal force resisted by the steel web panel
dσij increment of stress tensor αij tensor of backstress
dεij increment of strain tensor γ coefficient set as 150
dεijp increment of plastic strain tensor γ12 engineering shear strain
D diameter of column γj shear deformation
Es elastic modulus δ story drift
fc axial compressive concrete strength εij tensor of strain
fs yield stress of joint tube ξeq equivalent damping coefficient
fyvs shear yield stress of joint tube μ ductility coefficient
fyvw shear yield stress of steel web σ11 transverse stress
he height of extended tube σ22 longitudinal stress
hw sectional height of steel web σij tensor of stress
H story height σMises equivalent von Mises stress
k0 coefficient set as 0.85 τ12 shear stress
L story span υ Poisson's ratio
N axial compression load applied on column △ lateral displacement applied at the top of column
P lateral load applied at the top of column △y lateral yield displacement
Pu lateral peak load △u lateral displacement corresponding to 0.85Pu
tf thickness of beam flange

Fig. 1. SRC and tubed SRC column-steel beam connections in a building frame.

published on the seismic behavior of TSRC column to steel beam con- joints. The dimensions and details of ⅓ scaled specimens are shown in
nections. This paper thus attempts to study the behavior of such con- Fig. 3. As shown, the tested circular TSRC column consists of a steel
nections under cyclic loading. A new connection system for circular tube of 3 mm (thickness) × 300 mm (diameter) × 2090 mm (length)
TSRC column and steel beam is proposed (Fig. 2), where cross dia- and an encased steel shape of 200 mm (depth) × 100 mm (flange
phragms are added to ensure the transfer of moments at the beam ends. width) × 4 mm (web thickness) × 14 mm (flange thickness). The steel
Meanwhile, the embedded/inside portion of the cross diaphragm en- beam has the cross-section of 250 mm x 150 mm × 8 mm × 10 mm
sures the shear transfer at beam ends. The joint tube (Fig. 2(b)) is used (Fig. 3) and is 3000 mm long. The beam flanges were made continuous
to improve the joint strength. The exterior cross diaphragms prevail to with the corresponding cross diaphragms (Fig. 4), thus eliminating the
improve the sectional height of steel shape. To verify the feasibility of uncertain effect caused by flange welds. The beam webs were welded to
such connection and investigate its seismic performance, four tests were the joint tube. The circular steel tube was cold-formed by rolling the
carried out. steel plate and the butt weld joint was enhanced by a 40 mm wide steel
plate to prevent the possible premature weld failure. The investigated
parameters include the thickness of joint tube (tj), the height of ex-
2. Experimental program tended tube (he), and the axial compression ratio of TSRC columns. The
specimen details are given in Table 1 and the mechanical properties of
2.1. General information of specimens steel are listed in Table 2. The axial compressive strength of concrete fc
is 35 MPa.
All specimens tested in this study were designed to have a weak link
in the connections, allowing the evaluation of the inelastic response of

486
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Fig. 2. Circular TSRC column to steel beam connection.

event. The specimens were tested under the combined axial and cyclic
lateral loads. The cyclic lateral load was applied at the top of column by
a manual jack which was held horizontally, while a constant axial load
was applied through a self-balanced system consisting of a vertical
actuator, a reaction rack, and four steel rods (Fig. 5(a)). According to
Chinese code JGJ/T 101–2015 [13], the applied cyclic loading history
(lateral load/displacement-cycle curve) as shown in Fig. 6 was adopted.
All key displacements were measured by the linear variable differ-
ential transformers (LVDT), as shown in Fig. 7, including the lateral
displacements at column top (LVDT-1), the lateral displacements at
beam ends (LVDT-2, LVDT-3), and the joint shear deformations (LVDT-
4, LVDT-5). Besides, the slip between test-up and floor was monitored
by LVDT-6, which was observed to be about 2 mm during the whole
experiment. The strains of steel beam, column tube, and joint tube were
Fig. 3. Connection details (unit: mm). measured using strain gauges. A load cell was used to monitor the ap-
plied lateral loads at column top. The details of instrumentation layout
are shown in Fig. 7.

3. Failure progression

Figs. 8–11 show the lateral load (P) versus lateral displacement (△)
curves and the corresponding failure modes of the tested connections.
The details of failure progression are described as follows.

Fig. 4. Geometry of cross diaphragm (unit: mm). 3.1. Specimen 1

Fig. 8(a) shows the failure modes of Specimen 1. The lateral peak
2.2. Test set-up and instrumentation layout
load was attained at 3.1% story drift (δ), with no obvious failure phe-
nomenon prior to that. The corners of the joint tube started to bulge at
The bottom of column was pinned to the strong floor of a laboratory
δ = 3.7% and the middle portion started to bulge at δ = 5.6%. The joint
and the beam ends were connected to this strong floor by steel links
tube near the enhanced plate started to tear at δ = 6.8%. After the test,
which permit the rotations and horizontal translations of beams. The
the joint tube was removed to observe the failure mode of core con-
pins shown in Fig. 5 represent the inflection points that are likely to
crete. As a result, concrete crushing failure was noticed. Additionally,
occur at the column mid-height and beam mid-span during a seismic
the flange of steel shape bulged slightly over a portion of the joint

Table 1
Details of specimens.
Specimen label tj (mm) he (mm) N (kN) △y (mm) △u (mm) Pu (kN) μ ξeq Vu (kN) Vcal (kN) Vcal/Vu

Specimen 1 2 120 1584 50.7/− 47.3 126.7/− 132.1 168/− 170 2.6/2.8 0.11 1308 1044 0.80
Specimen 2 2 120 2375 48.9/− 44.3 119.8/− 124.2 169/− 175 2.4/2.8 0.12 1331 1044 0.78
Specimen 3 2 0 1584 44.5/− 45.9 117.2/− 112.9 157/− 150 2.6/2.5 0.11 1188 1044 0.87
Specimen 4 4 120 1584 62.7/− 62.7 136.8/− 136.7 221/− 213 2.2/2.2 0.15 1680 1224 0.73

487
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Table 2
Measured mechanical properties of steel.
Steel type Thickness (mm) Yield strength (MPa) Ultimate strength (MPa)

Steel plate with a thickness of 2 mm 1.76 273 392

Steel plate with a thickness of 3 mm 2.70 438 557
Steel plate with a thickness of 4 mm (Web of steel shape) 3.67 367 493
Steel plate with a thickness of 4 mm (Joint tube) 3.66 447 563
Steel plate with a thickness of 8 mm 8.40 435 545
Steel plate with a thickness of 10 mm 9.85 425 533
Steel plate with a thickness of 14 mm 14.40 385 492

Fig. 5. Test set-up.

Fig. 7. Details of instrumentation layout.

Fig. 6. Cyclic loading history.

beam tore completely. After removing the joint tube, severe damage of
core concrete was observed and the flange of steel shape bulged ob-
region.
viously over a portion of the joint region. Compared to Specimen 1,
Specimen 2 showed more severe damages and worse deformability,
3.2. Specimen 2 indicating that the higher axial compression ratio resulted in more se-
vere damages.
Fig. 9 shows the failure modes and P-△ curves for Specimen 2, which
indicates higher axial load than Specimen 1. The lateral peak load was
attained at δ = 2.7%，which is lower than Specimen 1. The corners of 3.3. Specimen 3
the joint tube started to bulge at δ = 3.4% and the joint tube near the
enhanced plate started to tear at δ = 4.8%. At δ = 6.9%, the middle Fig. 10(a) shows the failure modes of Specimen 3, where the height
portion of the joint tube bulged and the joint tube near the web of steel of extended tube is zero. At δ = 3.0%, the lateral peak load was

488
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Fig. 8. Failure progression of Specimen 1.

Fig. 9. Failure progression of Specimen 2.

attained and the corner portions of joint tube started to bulge. The concrete compression strut was formed by the joint core concrete. The
middle portions of joint tube started to bulge at δ = 3.8% and the joint flange of steel shape bulged over a portion of the joint region, mainly
tube near the enhanced plate started to tear at δ = 4.7%. After re- caused by the shear deformation of the steel shape web. Clearly, the
moving the joint tube, the concrete failure by crushing was observed joint strength was mainly contributed by the joint tube, joint core
and the flange of steel shape bulged slightly over a portion of the joint concrete and steel shape web. During to the use of the enhanced plate,
region. Specimen 3 showed similar failure modes to Specimen 1, the failure of Specimens 1–3 was caused by the tearing of joint tube
showing that the extended tube had little effect on the failure mode. near the enhanced plate. The failure of Specimen 4 was due to the
fracturing of welds between the joint tube and the enhanced plate.
3.4. Specimen 4 Therefore, it is essential that the seamless steel tube be adopted for the
joint tube.
Fig. 11 shows the failure mode of Specimen 4 in which the thickness
of joint tube is 4 mm. Due to poor weld quality, the weld between the 4. Analysis and discussion
joint tube and the enhanced plate fractured completely at δ = 5.5%.
After removing the joint tube, slight cracks on the core concrete were 4.1. Load-displacement curves
noticed.
In summary, the damages of the specimens mainly concentrate on 4.1.1. P-△envelope curves
the joint tube, joint core concrete, and flange of steel shape. The middle Fig. 12(a) shows the effect of axial load ratio on P-△ envelope
portions of joint tube bulged, indicating a shear buckling of joint tube. curves, indicating that the axial compression ratio has little effect on
The concrete failure by crushing was noted, suggesting that the the shear strength of joints within the range of parameters considered in

489
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Fig. 10. Failure progression of Specimen 3.

Fig. 11. Failure progression of Specimen 4.

this study. Fig. 12(b) indicates that the extended tube increases the joint apparatus and δ1 + δ2 and δ3 + δ4 are respectively the displacements
strength by about 6.5%. The effect of thickness of joint tube on P-△ measured by the LVDTs positioned diagonally over the joint panel zone
envelope curves is shown in the Fig. 12(c), demonstrating that the joint (Fig. 13(b)).
tube thickness has a significant effect on the joint strength. Fig. 14 shows the shear force versus shear deformation curves es-
tablished based on the above method. These curves show significant
4.1.2. Shear force-shear deformation (Vj-γj) curves nonlinearity and descending branch, further verifying the joint failure.
The shear force Vj (Fig. 13(a)) is given by
4.2. Elastic-plastic analysis
PH (L − D)
Vj = T1 + T2 − P = −P
L (hb − t f ) (1) Since the strain development in the joint tube could not reveal the
where P is the lateral load applied at the top of column, H and L are the components behavior directly, a developed computation procedure is
story height and story span respectively (Fig. 7), D is the diameter of proposed to analyze the stresses based on the measured strains. The
column, hb is the height of beam, and tf is the thickness of beam flange. developed computation procedure is described as follows.
In this study, the shear deformation γj was calculated by In the elastic range, the stresses of joint tube under the state of
plane-stress can be determined by
1⎡ a2 + b2
γj = |δ1 + δ2 |⎟⎞ + ⎜⎛|δ3 + δ4 |⎟⎞ ⎤ σ 1 ν 0 ⎤ ε11
2⎢ ⎥ ab (2) ⎡ σ11 ⎤ Es ⎡ ν 1 0 ⎥ ⎡ ε22 ⎤
⎣ ⎠ ⎝ ⎠⎦ 22 = ⎢
⎢ ⎥ 1 − υ2 ⎢ 0 0 1 − ν ⎥⎢ ⎥
⎣ σ12 ⎦ γ
where a and b are respectively the width and height of the measuring ⎣ 2 ⎦ ⎣ 12 ⎦ (3)

490
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Fig. 12. Effects of experimental parameters on the P-△ envelope curves.

Fig. 13. Shear force and shear deformation of joint.

where σ11 and σ22 are the transverse and longitudinal stresses of tube model the behavior of metals subjected to cyclic loading, in which the
respectively, σ12 is the shear stress, ε11 and ε22 are the transverse and yield surface is defined by
longitudinal strains of tube respectively, γ12 is the shear strain, Es is the
elastic modulus, and υ is the Poisson's ratio (= 0.18). f (σij, α ij) = 3J2 (σij − α ij) − k 0 fs = 0 (4)
In the plastic range, the kinematic hardening model was used to
where fs is the yield stress and σij and αij are the tensors of stress and

491
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Fig. 14. Shear force-shear deformation curves.

where dεijp is the increment of plastic strain tensor, dp is the increment

of accumulated equivalent plastic strain, and k0, C, and γ are the ma-
terial constants (= 0.85, 82.5fs, and 150 respectively) [14]. According
to the Von-Misses yielding criterion, the increment of accumulated
equivalent plastic strain dp is given by

2 p p
dp = dεij dεij
3 (6)

with the associated plastic flow assumed, the increment of plastic

strain dεijp is given by
∂f 3sij (σij − α ij)
dεijp = dp = dp
∂σij 2k 0 fs (7)

The consistency condition gives the following function

∂f ∂f
dσij − dα ij = 0
∂σij ∂α ij (8)

where dσij is the increment of stress tensor determined by

p
⎡ dσ11 ⎤ 1 ν 0 ⎤ ⎡ dε11 − dε11 ⎤
E ⎡ν 1 0 ⎥ ⎢ dε − dε ⎥p
⎢ dσ22 ⎥ = ⎢ 22 22
⎢ dσ ⎥ 1 − υ2 ⎢ 0 0 1 − ν ⎥⎢
p⎥
⎣ 12 ⎦ ⎣ 2 ⎦ ⎣ dε12 − dε12 (9)
Fig. 15. The developed computation procedure. ⎦
where dεij is the increment of strain tensor being known for a particular
backstress respectively. The increment of backstress tensor is calculated strain path. Solving Eqs. (3)–(9), the increment of stress tensor can be
by determined and the stresses are obtained through a numerical in-
tegration as shown in Fig. 15. The general finite element program
2 ABAQUS [15] was utilized to validate the developed computation
dα ij = C dεijp − γα ij dp
3 (5) procedure, where the same material model (kinematic hardening

492
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Fig. 16. The process of verification.

Fig. 17. Verification of the developed computation procedure.

model) and material properties (material constants) were assumed. this study. However, the deformation ability is increased by the ex-
Since it is difficult to directly calculate the corresponding stresses based tended tube and decreased by the axial load. The μ factor for Specimen
on a particular strain path in ABAQUS, a verification process is war- 4 is about 2.2, which is decreased due to the premature weld fracturing.
ranted, as illustrated in Fig. 16. Fig. 17 shows the comparisons of results μ factors for TSRC joints without the premature weld fracturing range
predicted by ABAQUS [15] and the developed computation procedure. from 2.4 to 2.8, which is higher than the average μ of 2 for a conven-
As seen, the results predicted by the developed procedure are consistent tional reinforced concrete (RC) joint.
with those predicted by the finite element analysis, thus validating the An equivalent damping coefficient ξeq [17] was adopted to re-
developed computation procedure. present the energy dissipation ability of a TSRC joint, which is defined
Adopting the developed calculation procedure mentioned above, by
the transverse, vertical, and shear stresses were obtained to evaluate the
ξeq = (SABC + SADC)/[2π (SOBE + SODA )] (10)
mechanical mechanism of joint tube. As a typical location, the center of
joint tube was selected. Fig. 18 shows the stresses of the joint tube for where SABC is the area enclosed by the curve ABC shown in Fig. 20.
all specimens tested in this study, which indicates that joint tube Similar definitions were used for SADC, SOBE, and SODA. ξeq for all spe-
yielded before the peak load and the vertical stress changed from cimens at the peak loads are given in Table 1, which indicates that ξeq of
compressive to tensile gradually. The transverse stress increased first about 0.11 for Specimens 1–3 (joint tube thickness = 2 mm) and 0.15
and then decreased with increasing shear forces. However, the shear for Specimen 4 (joint tube thickness = 4 mm). This implies that the
stress increased with the shear force continually and became dominant joint tube thickness can significantly increase ξeq. The average ξeq for
when the peak load was attained. Based on the elastic-plastic analysis conventional RC joints is about 0.1, while it is about 0.12 for TSRC
on the joint tube, it can be stated that the joint tube contributes to the joints, indicating that the TSRC joints generally have better energy
joint strength mainly by direct shearing. dissipation ability than the conventional RC joints.

4.3. Ductility and dissipated energy 5. Design model for joint strength

The ductility factor μ used to characterize the ductility is defined as The total strength of the joint is the sum of the contribution from the
Δu/Δy where Δy is the lateral yield displacement (Fig. 19) and Δu is the joint tube (Vt), joint core concrete (Vc), and steel shape web (Vw) as
lateral ultimate displacement taken corresponding to 85% of lateral indicated in Eq. (11). The mechanism for each component is illustrated
peak load Pu [16]. The Pu, Δy, Δu, and μ values for the four test speci- in Fig. 21 and the mechanisms are discussed as follows.
mens are listed in Table 1. It is found that the axial load and extended Vcal = Vc + Vt + Vw (11)
tube have little effect on μ within the range of parameters considered in

493
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Fig. 18. Stresses of joint tube.

Fig. 19. Method for determining Δy and Δu.

Fig. 20. The calculation of equivalent damping coefficient.

5.1. Concrete compression strut

where Ac is the cross-sectional area of core concrete and taken as 0.8D2.
The concrete compression strut (Fig. 21(a)) is mobilized by the joint
tube, which is similar to that used to model the shear for RC joints 5.2. Joint tube
(Chinese code for GB 50010–2010) [18]. The horizontal shear force Vc
resisted by the strut is calculated by Base on the elastic-plastic analysis on the joint tube described
above, it can be concluded that the joint tube contributes to the joint
Vc = 0.3fc Ac (12) strength mainly by direct shearing (Fig. 21(b)). The horizontal shear

494
G. Cheng et al. Thin-Walled Structures 138 (2019) 485–495

Fig. 21. The mechanism for each component.

force Vt resisted by the joint tube is calculated by [19] concrete, and steel shape web. The proposed design model for the
Vt = fyvs Av joint strength is somewhat conservative but practically acceptable.
(13)
where fyvs is the shear yield stress of joint tube and Av is calculated by Acknowledgements
Av = 2As / π (14)
The authors greatly appreciate the financial supports provided by
where As is the cross-sectional area of the joint tube. the National Natural Science Foundation of China (No. 51438001, No.
51622802). The opinions expressed in this paper are solely of the au-
5.3. Steel web panel thors, however.

The steel web was idealized as carrying the pure shear over a por- References
tion of the joint region. Since the steel section is encased with concrete,
local buckling of steel web is prevented. The distribution of shear [1] Eurocode 4: Design of composite steel and concrete structures, Part 1.1, general
stresses (Fig. 21(c)) in a web panel proposed by Parra-Montesinos et al. rules and rules for Building. London: British Standards Institution; 2004 (BS EN
1994-1-1: 2004).
[20] was adopted in this study. For interior joints, the horizontal shear [2] ANSI/AISC 341-10. Seismic provisions for structural steel buildings. Chicago,
force Vw resisted by the steel web panel can be expressed as Illinois, USA: American Institute of Steel Construction, 2010.
[3] JGJ 138-2016, Code for design of composite structures [in Chinese], 2016.
Vw = 0.9fyvw tw h w (15) [4] X. Zhou, J. Liu, Seismic behavior and strength of tubed steel reinforced concrete
(SRC) short columns, J. Constr. Steel Res. 66 (7) (2010) 885–896.
where fyvw is the shear yield stress of steel web and hw and tw are the [5] D. Gan, L. Guo, J. Liu, X. Zhou, Seismic behavior and moment strength of tubed
height and width of steel web respectively. steel reinforced-concrete (SRC) beam-columns, J. Constr. Steel Res. 67 (10) (2011)
1516–1524.
The ratio of calculated strength to experimental strength ranges
[6] J. Liu, J.A. Abdullah, S. Zhang, Hysteretic behavior and design of square tubed
from 0.73 to 0.87 as indicated in Table 1, demonstrating that the design reinforced and steel reinforced concrete (STRC and/or STSRC) short columns, Thin-
model for joint strength is somewhat conservative but acceptable for Walled Struct. 49 (7) (2011) 874–888.
[7] H. Qi, L. Guo, J. Liu, D. Gan, S. Zhang, Axial load behavior and strength of tubed
practical designs.
steel reinforced-concrete (SRC) stub columns, Thin-walled Struct. 49 (9) (2011)
1141–1150.
6. Conclusions [8] B. Yan, J. Liu, X. Zhou, Axial load behavior and stability strength of circular tubed
steel reinforced concrete (SRC) columns, Steel Compos. Struct. 25 (5) (2017)
545–556.
This article describes and discusses the seismic behavior of circular [9] X. Zhou, B. Yan, J. Liu, Behavior of square tubed steel reinforced-concrete (SRC)
TSRC column to steel beam connections based on an experimental study columns under eccentric compression, Thin-Walled Struct. 91 (2015) 129–138.
involving four specimens. The following conclusions can be drawn from [10] J. Liu, X. Wang, H. Qi, S. Zhang, Behavior and strength of circular tubed steel-
reinforced-concrete short columns under eccentric loading, Adv. Struct. Eng. 18
the study: (10) (2015) 1587–1595.
[11] X. Wang, J. Liu, X. Zhou, Behavior and design method of short square tubed-steel-
1) The failure of the test specimens is typically caused by the joint tube, reinforced-concrete columns under eccentric loading, J. Constr. Steel Res. 116
(2016) 193–203.
signifying a joint failure. The joint tube thickness has a significant [12] X. Zhou, X. Zang, X. Wang, J. Liu, Y.F. Chen, Seismic behavior of circular TSRC
effect on the joint strength, while the height of extended tube and columns with studs on the steel section, J. Constr. Steel Res. 137 (2017) 31–36.
the axial compression ratio of columns have little effect on the joint [13] JGJ/T 101-2015, Specification for seismic test of buildings [in Chinese], 2015.
[14] H. Hu, Concrete-Filled Steel Plate Composite Coupling Beam and its Application to
strength within the range of parameters considered in this study.
Shear Wall Structures, Tsinghua University, 2014 (in Chinese).
2) The developed computation procedure for elastic-plastic analysis on [15] Karlsson Hibbitt, Sorensen. ABAQUS/Standard User’s Manual. Version 6.7.1,
steel tube is shown to be valid. Based on the elastic-plastic analysis Hibbitt, Karlsson & Sorensen, Inc., 2007.
[16] Z. Guo, X. Shi, Reinforced Concrete Theory and Analyse, Tsinghua University Press,
on the joint tube, the joint tube contributes to the joint strength
Beijing, China, 2003 (in Chinese).
primarily by direct shearing. [17] J. Tang, Earthquake resistant design of reinforced concrete frame connection,
3) The ductility factor μ ranges from 2.2 to 2.8 and the equivalent Southeast University Press, Nanjin, China, 1989 (in Chinese).
damping coefficient ξeq is about 0.12 for the TSRC joints, as evi- [18] GB 50010-2010, Code for design of concrete structures [in Chinese], 2010.
[19] Eurocode 3: Design of steel structures, Part 1.8, design of joints. London: British
denced by the tests. Both indicate that the TSRC joint generally has a Standards Institution; 2004 (EN 1993-1-8: 2005).
better seismic performance than the conventional RC joint. [20] G. Parra-Montesinos, J.K. Wight, Modeling shear behavior of hybrid RCS beam-
4) The joint strength is mainly contributed by the joint tube, joint core column connections, J. Struct. Eng. 127 (1) (2001) 3–11.

495