Engineering Structures 171 (2018) 445–452

Contents lists available at ScienceDirect

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

CHS X-joints strengthened by external stiffeners under brace axial tension T

a a,⁎ a b c
Yunan Ding , Lei Zhu , Kuang Zhang , Yu Bai , Hailin Sun
a
Beijing Advanced Innovation Center for Future Urban Design, School of Civil and Transportation Engineering, Beijing Higher Institution Engineering Research Center of
Structural Engineering and New Materials, Beijing Key Laboratory of Functional Materials for Building Structure and Environment Remediation, Beijing University of Civil
Engineering and Architecture, Beijing, China
b
Department of Civil Engineering, Monash University, Clayton, VIC 3800, Australia
c
China Architecture Design and Research Group, China

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

Keywords: This paper presents an experimental and numerical investigation of the use of external stiffeners to reinforce
Circular hollow section circular hollow section (CHS) X-joints under brace axial tension. Six full-scale specimens with different brace to
X-joint chord diameter ratios (β) of 0.25, 0.50, and 0.73 are tested, including three unreinforced X-joints and three
Axial tension reinforced X-joints. The experimental setup and detailed parameters are presented and results including failure
External stiffener
modes, load–displacement curves, and ultimate capacity are compared. The experiment results show that the
Full-scale testing
Finite element modeling
reinforced external stiffeners clearly increase the mechanical performance of joints in tension, supported by the
improved ultimate strength and initial stiffness in comparison to the unreinforced specimens. Furthermore, the
enhancement of ultimate strength increases along with the increase of brace to chord diameter ratio (β). Finite
element (FE) modeling using SHELL181 element is also established and precisely describes the static behavior of
the X-joints under brace tension with and without external stiffeners.

1. Introduction subjected to brace tension has been investigated through a series of

The use of circular hollow sections (CHS) is becoming increasingly
popular because of their technological advantages over open sections,
such as resistance in compression, torsion, and lateral bending as well
as their aesthetically pleasing shape. Nowadays, CHS are applied in
structures such as stadiums, towers, bridges, long-span roofs, and
platform jackets [1–3]. In many cases, the ultimate capacity of such
CHS structures is governed by their joint capacity, due to the premature
failure of the joints. A joint stiffened with external stiffeners welded to
the chord member and brace member at the brace-chord intersection is
called an external stiffener stiffened joint.
The bearing capacity and deformation of T-joints using external
stiffeners was explored by Zhu et al. [4,5]. The results illustrated that
the static strength and stiffness of stiffened joints was increased due to
the enhancement of the connection length of the brace-chord inter-
section zone. Extensive studies of stiffened tubular X-joints have been
performed, covering aspects such as the use of external stiffener [6],
external stiffening ring [7], internal stiffening ring [8,9], fully grouted
[10], double-skin grouting [11], longitudinal diaphragm [12], joint can
[13], and collar plate [14].
Limited research has focused on the behavior and strength of brace
axial tensile tubular X-joints. The ultimate strength of CHS X-joints
experiments [15]. Different approaches in the definition of joint
strength were discussed and the results of these approaches were
compared with the design recommendations in ISO and CIDECT design
codes. Further study of the effect of chord stresses on static strength was
also carried out by Choo et al. [16]. New design equations for CHS X-
joint ultimate strength have been included in the CIDECT design code
[17]. The existing codified guidance for the first crack loads and ulti-
mate loads of X-joints in tension was reviewed by Dier et al. [18]. The
ductile fracture of welded CHS-RHS X-joints was investigated by Ma
et al. [19] under brace tension, where the stress and strain distribution
of fillet welds were presented and the FE model was built to simulate
the weld actual mechanical behavior. Furthermore, the calculation of
load-deformation relationship for CHS joints were studied in [20,21]
with the consideration of weld properties. These authors proposed re-
vised mean and characteristic capacity formulations for predicting the
first crack. A parametric study of axial stiffness of CHS X-joints sub-
jected to brace tension was presented by Qiu et al. [22], who also in-
vestigated the factors influencing the axial stiffness and the formulae
used to calculate the joint axial stiffness. Experimental tests and sub-
sequent finite element (FE) calibration studies were conducted to study
the static strength of axially loaded (tension and compression) elliptical
hollow section X-joints [23,24].

⁎
Corresponding author.
E-mail address: zhulei@bucea.edu.cn (L. Zhu).

https://doi.org/10.1016/j.engstruct.2018.05.101
Received 27 January 2018; Received in revised form 28 May 2018; Accepted 29 May 2018
0141-0296/ © 2018 Elsevier Ltd. All rights reserved.
Y. Ding et al. Engineering Structures 171 (2018) 445–452

Nomenclature β ratio of brace to chord diameter d1/d0

γ ratio of chord diameter to twice the chord wall thickness
d1 brace diameter d0/(2t0)
d0 chord diameter τ ratio of brace wall thickness to chord wall thickness t1/t0
l0 chord length E0 Young’s modulus of chord
l1 brace length E1 Young’s modulus of brace
t0 chord wall thickness E2 Young’s modulus of stiffener
t1 brace wall thickness FFE joint load from numerical result
ts stiffener wall thickness Fu,exp ultimate strength obtained from test
ls stiffener length NC joint strength according to the Chinese design code
hs stiffener height NI joint strength according to ISO and CIDECT criteria
α ratio of chord length to radius 2l0/d0 NA joint strength according to AISC criterion

Further, the ultimate static strength of 13 double-skin grouted X-
joints in brace tension was tested by Feng [25], who discussed the
failure modes of joints and compared the ultimate strength obtained
from testing with the current design codes and then presented new
formulations. A study of the static capacity of fully grouted X-joints
under brace axial tension and in-plane bending was conducted by Chen
[26], and significant improvements over punching shear formulations
were gained according to the experimental results. Study of the static
behavior of internally ring-stiffened CHS DT-joints subjected to brace
axial compression or tension was reported by Wang et al. [27]. A large
number (about 800) of FE models were built to obtain a strength design
equation for stiffened joints. Furthermore, an extensive study of 1264
unstiffened and internally ring-stiffened DT-joints subjected to axial
tensile and compressive loading was conducted by Lan et al. [28]. The
failure mechanism and equations for predicting the stiffener strength
were presented.
External stiffener strengthening is an effective and convenient re-
inforcement method for CHS joints, with broad applicability. Still,
techniques for assessing the capacity of external stiffener stiffened
joints under brace tension are lacking. Further research is needed to
provide guidance for the application of external stiffeners, with the aim
of enriching understanding of reinforced X-joints subjected to brace
tension.
This paper investigates the static strength of external stiffener stif-
fened CHS X-joints under brace axial tensile load. Three different brace
to chord diameter ratios (β = 0.25, 0.50, and 0.73) were used in the
experimental study. The ultimate strength, failure mode, and load–-
displacement curves of the specimens were recorded in the experiment.
Furthermore, FE modeling was developed with experimental valida-
tion, to further understand the static behavior of the joints.
2.2. Specimen geometry and scenarios
The specimens’ dimensions are shown in Table 1, with corre-
sponding geometric variables illustrated in Fig. 2. The test series was
divided into three pair groups: three specimens were reinforced with
the stiffener (X-0.25-R, X-050-R, X-073-R) and the others were un-
reinforced (X-0.25, X-050, X-073), to serve as reference. Each specimen
had the nominal chord diameter (d0) of 300 mm and the nominal chord
length (l0) of 1800 mm, resulting in a ratio of chord length to radius (α)
of 12, which avoided any short chord influence. The height (hs) and
length (ls) of each external stiffener were designed to be twice those of
the brace diameter and the nominal thickness of the stiffener (ts) was
8 mm. Two 45 mm thick endplates were welded at two brace ends of the
specimen to facilitate specimen installation and load application.
The brace to chord diameter ratio (β) varied from 0.25 to 0.73 and
the wall thickness of brace was thick enough to avoid any premature
failure in the member. The diameter and wall thickness of each tubular
member were measured at three locations in a selected cross-section
and then the average was used. The chosen geometric parameters (α, β,
γ, and τ) correspond to typical values for X-joints in practical applica-
tions.
All braces and chords were hot-rolled seamless and low carbon steel
tubes. All six chords were cut from a single 12 m long tube in order to

2. Experimental program

2.1. Experimental setup and instrumentation

Fig. 1 shows the typical setup of the X-joint experiments. Each brace
end was provided with an endplate. One plate was bolted to a high-
stiffness short beam and the other to the strong base of the test rig,
which was fixed on the ground by a ground anchor. Loading was
achieved through two 100-ton actuators acting on both ends of the
short beam. The displacement-controlled actuator had a rated com-
pression capacity of 1000 kN and a stroke of ± 300 mm. For each test,
the load was applied at an initial stroke rate of 1.0 mm/min for the
linear load range, which was then decreased to 0.3 mm/min at higher
load ranges.
As shown in Fig. 1, two linear variable displacement transducers
(LVDT1 and LVDT2) were located at the bottom plate. Four LVDTs
(LVDT3 to LVDT6) were located at the top endplate to monitor the
vertical displacement difference between two brace end plates with
reference to the values measured by LVDT1 and LVDT2. Fig. 1. Experiment setup.

446
Y. Ding et al. Engineering Structures 171 (2018) 445–452

Table 1
Specimen parameters.
Specimen Type l0 (mm) d1 (mm) t0 (mm) t1 (mm) l1 (mm) ls (mm) hs (mm) β γ τ

X-025 Unreinforced 1801.0 76.0 7.75 5.87 372.5 0.25 19.35 0.76
X-025-R Reinforced 1803.3 76.0 8.03 5.68 375.1 152.0 152.0 0.25 18.68 0.71
X-050 Unreinforced 1799.5 151.0 8.11 8.75 752.6 0.50 18.50 1.08
X-050-R Reinforced 1797.8 152.0 8.38 8.69 749.5 302.0 302.0 0.51 17.90 1.04
X-073 Unreinforced 1797.1 219.0 7.95 8.45 1123.0 0.73 18.87 1.06
X-073-R Reinforced 1799.2 220.0 8.20 8.27 1122.5 438.0 438.0 0.73 18.29 1.01

stopped when weld cracks occurred at the brace-to-chord intersections,

as shown in Fig. 5. The other specimens failed by severe chord plasti-
fication at the chord midspan. Fig. 6 shows an overall view of the de-
formation of X-025-R.
In general, the failure could be classified into two types, chord
plastification and weld failure. All reinforced specimens and un-
reinforced specimens except X-073 failed due to severe chord plastifi-
cation that was much greater than 6%d0 deformation limit. For spe-
cimen X-025 and X-073-R, the 6%d0 deformation limit was achieved
before a weld crack appeared, but the contrary was the case for un-
reinforced specimen X-073. Hence, the failure mode for all the speci-
mens except X-073 was chord plastification, with weld failure occurring
only in X-073.

3.2. Load-displacement curves

Fig. 7 shows the load–displacement plots of all the tested specimens,

Fig. 2. Schematic arrangement of specimen.
obtain identical chord for each joint. Partial penetration fillet welding
was used to weld the brace, chord, and stiffener together and the weld
size was 11 mm determined according to the Eurocode 3 re-
commendation [29].
2.3. Material properties
Coupon testing of the chord, brace, and stiffener members was
conducted. Based on the requirements of the Chinese Code of Metallic
Materials [30], the coupon tests determined the Young’s modulus and
yield stress for each chord, brace, and stiffener. The recorded average
yield stress and Young’s modulus for chord members were 267.7 MPa
and 200.3 GPa and for the stiffener were 358.3 MPa and 246.5 GPa
respectively. For the specimens with β of 0.25, β of 0.50, and β of 0.73,
the yield stresses of brace members were 313.3 MPa, 295.0 MPa, and
317.7 MPa and the Young’s modulus of brace members was 222.3 GPa,
240.0 GPa, and 244.3 GPa respectively. The stress–strain curves are
shown in Fig. 3.
where the displacement is the difference between the displacements
recorded by the LVDTs at the top and bottom brace ends. The curves for
the specimens X-0.25-R, X-050-R, and X-050-R exhibit a continuous
rising trend and no defined peak occurs. For specimens X-025, X-073
and X-073-R, the curves reach their peak when the weld crack ap-
peared.
For specimen X-073, brittle crack occurred before the chord reached
6%d0 ovalization. However, the maximum load obtained from the curve
peak is not the ultimate load of X-073, because the weld failure is not
typical of the normal failure modes of X-joints such as chord plastifi-
cation or punching shear. Hence, no ultimate load was obtained from
experiment for specimen X-073.
A comparison of the reinforced and unreinforced joints with same β
value is shown in Fig. 7. The reference ultimate loads of the three re-
inforced specimens, as determined according to the 6%d0 deformation
limit, exceeded the ultimate loads of unreinforced specimens

3. Experimental results

In this study, the load carrying capacity for specimens was defined
as the first peak load in the load–displacement diagram if it occurred
before a chord ovalization of 6%d0, or the load at 6%d0 ovalization if
the peak load occurred after the 6%d0 ovalization, in accordance with
Ref. [31].

3.1. Failure modes

After each specimen was tested, a section or ring segment was cut
from the midspan of the chord by flame-cutting. The chord ovalization
was observed through these sections or rings. Fig. 4 (left) shows de-
tailed views of the chord slices (cut at the midspan) from each spe-
cimen. As can be seen, significant chord ovalization occurred at mid-
span of the specimens. For X-025, X-073, and X-073-R the tests were Fig. 3. Stress–strain curves.

447
Y. Ding et al. Engineering Structures 171 (2018) 445–452

Fig. 4. Deformed shapes of specimens after loading: (a) X-025; (b) X-025-R; (c) X-050; (d) X-050-R; (e) X-073; (f) X-073-R.

Fig. 5. Detailed view of welded regions: (a) X-025; (b) X-073; (c) X-073-R.

448
Y. Ding et al. Engineering Structures 171 (2018) 445–452

Fig. 6. Deformation of reinforced specimen X-025-R.

moderately-thick shell structures and large strain nonlinear structures,

the element SHELL181, a 4-node element with six degrees of freedom at
each node, was used to model the joints.
To build an accurate and reliable FE model, optimum mesh size as
well as reasonable definitions of material properties, loading applica-
tions, and boundary conditions are necessary. To save computing cost,
only half of the overall geometry was modeled, considering the sym-
metry in geometry and loading conditions. Fig. 8 shows the mesh of
specimen X-050-R, in which fine mesh was used in the welding area
along the joint interaction region; relatively coarse mesh was used in
locations away from the critical position, such as chord and brace ends.
The material property of the joint was assumed to be the elastic-
perfect plastic model. The Young’s modulus and the yield stresses were
aligned with experimental measurements, the Poisson’s ratio was 0.3 in
accordance with Chinese design code [32], and the Von Mises yield
criterion was applied.
The joint dimensions were taken in exactly the same way as in the
experiments. Chord, brace, stiffeners, end plates, and weld were mod-
eled for each FE model. The weld geometry is shown in Fig. 9 with an
area defined by a and b, where a = t1 and b = 0.5 t0 + t1, and the
thickness of the weld element is 0.5 t1. The weld fracture was not
Fig. 7. Experimental load–displacement curves of X-025 to X-073-R.
considered in the weld numerical model.
The FE model constrained the X and Z displacements of the brace
respectively by 3.7% for β of 0.25 and 16.6% for β of 0.50. Further- end to prevent out-of-plane displacement of the brace. The surface on
more, the joint local stiffness, defined as the slope of the straight part of the plane of symmetry was set with symmetric boundary conditions, as
the load–displacement curve, also increased. The initial stiffness of the shown in Fig. 8. Uniform pressure was used to apply axial tension to the
unreinforced joints was increased by 12.5% for β of 0.25, 9.6% for β of brace endplate. The large displacement static option was chosen and
0.50, and 15.1% for β of 0.73. It is evident from Fig. 7 that the external numerical data were recorded from every substep. Meanwhile, the so-
stiffener strengthening increased both the joint bearing capacity and lution of the equations was calculated using the arc-length algorithm.
joint stiffness for all the strengthened specimens.
Table 2 summarizes the experimental results and joint strength
5. Comparison and discussion
calculated by different formulae, where NC is calculated according to
the Chinese design code (Chinese standard GB 50017-2003) [32], NI is
The deformed shapes obtained by FE analysis of all the specimens
calculated according to the formulae from ISO 14346 [33] and CIDECT
Design Guide [17], and NA is calculated according to the formulae from
Table 2
the AISC standard [34]. It can be seen that the joint strengths of X-025
Joint strength from experimental and FE studies.
and X-050 compare well (with the difference less than 6%) with the
estimations from the Chinese design code; and the joint strengths cal- Specimen Fu,exp (kN) FFE (kN) FFE/Fu,exp NC (kN) NI (kN) NA (kN)
6%d0 6%d0 Chinese ISO 14346 AISC
culated from ISO and AISC are underestimated (by about 40%) in
Code
comparison to the experiment results.
X-025 188.2 176.1 0.94 178.06 98.78 114.92
X-025-R 194.7 201.0 1.03 189.81 105.49 123.37
4. FE modeling X-050 265.3 281.8 1.06 259.01 163.65 168.67
X-050-R 308.1 333.7 1.08 278.51 176.93 182.58
X-073 – 455.7 – 363.77 241.80 235.97
The FE program ANSYS 12.0 was used for the nonlinear numerical X-073-R 543.1 527.1 0.97 384.60 256.05 251.04
analysis of the joints. Because of its excellent ability to analyze thin to

449
Y. Ding et al. Engineering Structures 171 (2018) 445–452

Fig. 11.
In conclusion, for the joints with β greater than or equal to 0.5,
external stiffener strengthening method can lead to an obvious (more
than 16%) enhancement of ultimate strength.

6. Conclusions

Experimental and numerical investigations were conducted to un-

derstand the static behavior of external stiffener reinforced CHS X-
joints. Three pairs of unreinforced and reinforced specimens with dif-
ferent values of β were examined under brace tensile loading. ANSYS
12.0 was used to establish FE models for the X-joint specimens.
Comparison of the numerical and experimental results verified the ac-
Fig. 8. Typical FE model meshes. curacy of the numerical analysis in describing the static behavior of the
joints. From the presented results, several conclusions can be drawn as
follows:
are shown in Fig. 4 (right). Comparisons show that sidewall deforma-
tion was a key component of chord ovalization. The tensile load
(1) Two failure modes identified from the experimental investigations
transferred from the brace flattened the chord sidewall and deformation
were the weld failure and chord plastification. The former was
became obvious near the chord midspan area. The load–displacement
observed only for the unreinforced specimen with β of 0.73; the
curves from FE and the experimental results are plotted in Fig. 10 for
latter was observed for all the other specimens.
each specimen. The lines drawn in the graphs represent the 6%d0
(2) The experimental results reported in this study evidenced a sig-
ovalization criterion line. For specimen X-073, premature failure of
nificant ultimate load enhancement for X-joints reinforced by an
weld resulted in the lack of ultimate load. However, as can be observed
external stiffener compared to the corresponding unreinforced X-
from Fig. 10, the numerical and experimental results match very well.
joints. The enhancement of the ultimate strength of the reinforced
Based on that, the load from the numerical results for X-073 corre-
X-joints increased with the increase of β. For the specimens with β
sponding to 6%d0 limit was chosen to substitute the ultimate load from
of 0.25, β of 0.50, and β of 0.73, the results revealed ultimate load
test. Hence, the external stiffeners enhanced about 19.3% of the ulti-
enhancement by 3.7%, 16.6%, and 19.3% respectively. The initial
mate load in comparison to the unreinforced joint for β of 0.73. The
stiffness also increased by 12.5%, 9.6%, and 15.1% respectively.
experimental and FE results are together plotted in Table 2.
(3) FE modeling using SHELL181 element precisely simulated the load-
A comparison between the test and FE analysis results was carried
deformation behavior and major structural characteristics such as
out to verify the FE model. Good agreement between the numerical and
deformed shape of all the six joints, with the difference between
experimental load–displacement curves was achieved, with the max-
numerical and experimental results of less than 10%.
imum differences less than 10%, not only with respect to the initial
(4) In practical application, external stiffeners are recommended to use
stiffness but also to the elastic–plastic stage development tendency.
in the joints with β greater than or equal to 0.5.
As can be seen in Fig. 10, both experimental and numerical results
show an increase in initial stiffness and ultimate load with the increase
of β. Fig. 11 further shows the enhancement efficiency due to the ex- Acknowledgment
ternal stiffener in relation to β. The ultimate load enhancement effi-
ciency (i.e. the ultimate load increase ratio of the reinforced joint to the The authors gratefully acknowledge support from the Beijing Haiju
unreinforced joint) of the external stiffener becomes more obvious in Program, Beijing Advanced Innovation Center for Future Urban Design
the case of greater β values, and the initial stiffness enhancement effi- (Grant No. UDC2016030200), National Natural Science Foundation of
ciency (i.e. the initial stiffness increase ratio of the reinforced joint to China (No. 51778035), the JiandaJieqing Plan, and Beijing Cooperative
the unreinforced joint) becomes higher when β is 0.73, as observed in Innovation Research Center on Energy Saving and Emission Reduction.

Fig. 9. Weld modeling.

450
Y. Ding et al. Engineering Structures 171 (2018) 445–452

Fig. 10. Comparisons of experimental and numerical load-deformation relationships.

Fig. 11. Enhancement efficiency of external stiffener corresponding to β.

Appendix A

d 0.2 pj
NC = 0.78 ⎛ ⎞ NCX
⎝t ⎠ (1)

pj 5.45
NCX = ψ t 2f
(1−0.81β )sinθ n

451
Y. Ding et al.
2
σ σ
where ψn = 1−0.3 f −0.3 ⎛ f ⎞
y ⎝ y⎠
NA =
( 5.7
Fy t 2 1 − 0.81β Qf )
sinθ
where Qf = 1.0
Fy0 t02
NI = Qu Qf
sinθ
1 + β ⎞ 0.15
Qu = 2.6 ⎜⎛ ⎟γ
⎝ 1−0.7β ⎠
Qf = (1−n)C1, C1 = 0.20
N0 M0
n= +
Npl,0 Mpl,0
References
[1] Zhao X-L, Wilkinson T, Hancock GJ. Cold-formed tubular members and connections:
structural behaviour and design. Elsevier; 2005.
[2] Liew JYR, Yan J-B, Huang Z-Y. Steel-concrete-steel sandwich composite structures-
recent innovations. J Constr Steel Res 2017;130:202–21.
[3] Yang X, Bai Y, Luo FJ, Zhao X-L, Ding F. Dynamic and fatigue performances of a
large-scale space frame assembled using pultruded GFRP composites. Compos Struct
2016;138:227–36.
[4] Zhu L, Zhao Y, Li S, Huang Y, Ban L. Numerical analysis of the axial strength of CHS
T-joints reinforced with external stiffeners. Thin-Walled Struct 2014;85:481–8.
[5] Zhu L, Song Q, Bai Y, Wei Y, Ma L. Capacity of steel CHS T-Joints strengthened with
external stiffeners under axial compression. Thin-Walled Struct 2017;113:39–46.
[6] Li W, Zhang S, Huo W, Bai Y, Zhu L. Axial compression capacity of steel CHS X-
joints strengthened with external stiffeners. J Constr Steel Res 2018;141:156–66.
[7] Zhu L, Yang K, Bai Y, Sun H, Wang M. Capacity of steel CHS X-joints strengthened
with external stiffening rings in compression. Thin-Walled Struct 2017;115:110–8.
[8] Mei Q, Gong J, Pang D, Yuan Z. Effect of internal ring-stiffened spacing on behavior
for X-joints. Steel Construct 2011;26(5):11–6. (in Chinese).
[9] van der Vegte GJ, Leray DH, Choo YS. The axial strength of uniplanar x-joints re-
inforced by T-shaped ring-stiffeners. In: The seventh international offshore and
polar engineering conference, International Society of Offshore and Polar Engineers;
1997.
[10] Chen Y, Zhang Z. Experimental research on compressive capacity of concrete filled
chord circular section X-joints. Jianzhu Jiegou Xuebao (J Build Struct)
2012;33(1):72–80 [in Chinese].
[11] Chen Y, Feng R, Xiong L. Experimental and numerical investigations on double-skin
CHS tubular X-joints under axial compression. Thin-Walled Struct
2016;106:268–83.
[12] Choo YS. Strength evaluation of X-joints internally stiffened with longitudinal
diaphragms. In: The fourth international offshore and polar engineering conference,
International Society of Offshore and Polar Engineers; 1994. p. 21–9.
[13] Van der Vegte GJ, Puthli RS, Wardenier J. The static strength of uniplanar tubular
steel X-joints reinforced by a can. Int J Offshore Polar Eng 1992;2(01).
[14] Nassiraei H, Zhu L, Lotfollahi-Yaghin MA, Ahmadi H. Static capacity of tubular X-
joints reinforced with collar plate subjected to brace compression. Thin-Walled
Struct 2017;119:256–65.
[15] Choo YS, Qian XD, Liew JYR, Wardenier J. Static strength of thick-walled CHS X-
joints—Part I. New approach in strength definition. J Construct Steel Res
2003;59(10):1201–28.
[16] Choo YS, Qian XD, Liew JYR, Wardenier J. Static strength of thick-walled CHS X-
joints—Part II. Effect of chord stresses. J Construct Steel Res 2003;59(10):1229–50.
[17] Wardenier J, Kurobane Y, Packer JA, van der Vegte GJ, Zhao XL. Design guide for
circular hollow section (CHS) joints under predominantly static loading. CIDEC
452
Engineering Structures 171 (2018) 445–452
(2)
(3)
series: Construct Hollow Steel Section; 2008.
[18] Dier AF, Smedley P, Solland G, Bang H. New data on the capacity of X-joints under
tension and implications for codes. In: ASME 27th international conference on
offshore mechanics and arctic engineering, American Society of Mechanical
Engineers; 2008. p. 557–65.
[19] Qian X, Zhang Y. Translating the material fracture resistance into representations in
welded tubular structures. Eng Fract Mech 2015;147:278–92.
[20] Gu B, Qian X, Ahmed A. A toughness based deformation limit for X- and K-joints
under brace axial tension. Front Struct Civ Eng 2016;10(3):345–62.
[21] Ma X, Wang W, Chen Y, Qian X. Simulation of ductile fracture in welded tubular
connections using a simplified damage plasticity model considering the effect of
stress triaxiality and Lode angle. J Constr Steel Res 2015;114:217–36.
[22] Qiu G, Gong J, Zhao J. Parametric formulae for axial stiffness of CHS X-joints
subjected to brace axial tension. J Zhejiang Univ-Sci A 2011;12(2):121–30.
[23] Shen W, Choo YS, Wardenier J, Packer JA, van der Vegte GJ. Static strength of
axially loaded elliptical hollow section X joints with braces welded to wide sides of
chord. I: Numerical investigations based on experimental tests. J Struct Eng
2014;140(1):04013035.
[24] Shen W, Choo YS, Wardenier J, Packer JA, van der Vegte GJ. Static strength of
axially loaded elliptical hollow section X joints with braces welded to wide sides of
chord. II: Parametric study and strength equations. J Struct Eng
2014;140(1):04013036.
[25] Feng WY. Static strength of double-skin grouted X-joint infilled with high strength
grout. PhD thesis. National Universtiy of Singapore; 2013.
[26] Zhou C. Static strength of tubular X-joint with chord fully infilled with high strength
grout. PhD thesis. National Universtiy of Singapore; 2010.
[27] Wang F, Lan X, Pan X, Ning C, Xu X, Liu D, et al. Strength calculation methodology
for internally ring-stiffened DT-joints. J Southeast Univ (English Ed)
2016;32(1):67–72.
[28] Lan X, Wang F, Ning C, Xu X, Pan X, Luo Z. Strength of internally ring-stiffened
tubular DT-joints subjected to brace axial loading. J Constr Steel Res
2016;125:88–94.
[29] B. EN, 1-8: 2005. Eurocode 3: Design of steel structures-Part 1–8: Design of joints.
British Standards Institution, United Kingdom; 2005.
[30] GB/T 228.1-2010 Metallic materials-Tensile testing-Method of test at ambient
temperature; 2010.
[31] Choo YS, van der Vegte GJ, Zettlemoyer N, Li BH, Liew JYR. Static strength of T-
joints reinforced with doubler or collar plates. I: Experimental investigations. J
Struct Eng 2005;131(1):119–28.
[32] Chinese standard GB 50017-2003. Code for design of steel structures; 2003 [in
Chinese].
[33] ISO 14346. Static Design Procedure for Welded Hollow-section Joints –
Recommendations, 1st ed. Geneva, Switzerland; 2013.
[34] AISC. Specification for structural steel buildings, Chicago; 2010.