applied

sciences
Article
Experimental Study on the Static Behavior of
Reinforced Warren Circular Hollow Section (CHS)
Tubular Trusses
Wenwei Yang * , Jiankang Lin *, Ni-na Gao and Ruhao Yan
College of Civil and Hydraulic Engineering, Ningxia University, Yinchuan 750021, China;
may5023753@163.com (N.-n.G.); YRHendeavor@163.com (R.Y.)
* Correspondence: nxyangww@163.com (W.Y.); linhealth@163.com (J.L.)




Received: 18 October 2018; Accepted: 9 November 2018; Published: 13 November 2018


Abstract: For truss structures, the question of whether to weld hidden welds or not has been
controversial. In the actual construction process of truss structures, the members are usually spot
welded in place on the assembly platform, and then welded as a whole, while the hidden welds of
the truss are not welded, especially for small pipe diameter trusses. Furthermore, in this study, under
hidden weld unwelded conditions, two kinds of reinforcing method (adding a half outer sleeve
on each joint and filling concrete into the chord members) are adopted to achieve the purpose of
strengthening the truss. Therefore, this paper presents an experimental study on the static behavior
of four types of Warren tubular trusses made of CHS members. These four types are (1) T-HW:
The truss with hidden welds welded; (2) T-HN: The truss with hidden welds unwelded; (3) TS-AS:
The truss strengthened on the basis of T-HN by adding a half outer sleeve on each joint; (4) TS-FC:
The truss strengthened on the basis of T-HN by filling concrete into the top and bottom chord
members. The mechanical behavior, failure mode, bearing capacity, and load-displacement of all
specimens were investigated. The surface plasticity of the bottom chord member, the weld fracture
around tubular joints at the bottom chord member, and the bending deformation of the bottom
chord member were observed in the tests. Compared with the T-HW specimen, the load carrying
capacity of the T-HN specimen decreased by 18%. On the other hand, the T-HN specimen has better
deformability than the T-HW specimen. The reinforcing method of adding a half outer sleeve on each
joint and filling concrete into the chord members can effectively improve the load carrying capacity
and stiffness of the truss, thus reducing the overall deformation of the truss, but the reinforcing
method of filling concrete into the chord members is more efficient.

Keywords: reinforced warren circular hollow section (CHS) tubular truss; hidden weld; static
performance; finite element analysis

1. Introduction
Truss structures are popularly used in various engineering fields [1–4], and among them,
steel tubular trusses have been widely applied in large engineering structures, such as stadiums,
bridges, concert halls, and offshore platforms, because of their architecturally attractive shapes and
favorable structural properties [5]. Many investigations have been carried out related to steel tubular
truss structures [6–8]. In addition, a hybrid structural analysis and procedure were developed by
Khademi [9,10] to enhance the load rating of railway truss bridges. The common steel tubular truss
types are Warren trusses, Pratt truss, Fink truss, and Vierendeel truss, among which Warren trusses
generally provide the most economical solution due to their reasonable mechanical performance and
effective structural layout [11]. The joint form of Warren truss is usually K-joint and KK-joint, however,

Appl. Sci. 2018, 8, 2237; doi:10.3390/app8112237 www.mdpi.com/journal/applsci

Appl. Sci. 2018, 8, 2237 2 of 22

for these two types of joints, the overlapped condition arises when one brace (overlapping brace)
intersects the other brace (overlapped brace), requiring a weld between the braces. In addition, part of
the overlapped brace is hidden within the overlapping brace, and the hidden toe of the overlapped
brace may or may not be welded to the chord [12]. Nevertheless, the question whether hidden
weld of overlapped joints (K-joint and KK-joint) welded or unwelded is not clearly mentioned in the
current Chinese Standard for Design of Steel Structures (GB 50017-2017) [13]. The joints with hidden
welds welded may have an increased strength, but with reduction in joint ductility and hysteretic
performance [12,14–17]. It should be noted that the aforementioned studies mainly focus on single
joint, while the research on the performance of overall truss is very limited on joints with and without
hidden welds.
On other hand, steel tubular trusses made of circular hollow section (CHS) members usually
fail due to surface plasticity or punching shear of the chord members due to low stiffness of the
chord members. One of the commonly used reinforcing methods is to fill the chord members with
grout or concrete. Concrete-filled steel tubular (CFST) members, as compared to their pure steel or
concrete counterparts, have improved stiffness, load bearing capacity, buckling resistance [18–21],
and even improved impact resistance [22,23]. In addition, structural health monitoring of CFST
members [24–27] and assemblies [28,29] have attracted much attention. For CHS joints, extensive
research works have been carried out on the static performance [30,31], hysteretic performance [32–34],
impact performance [35,36], fatigue performance [37,38], and mechanical properties at elevated
temperature [39–41], among others. However, compared with CHS joints, little investigation has
been carried out on the performance of CFST joints. Packer et al. [42] analyzed the different failure
modes of CFST joints and compared them to CHS joints, indicating the increased resistance attained by
adding concrete in the chords of tubular joints. Sakai et al. [43], Huang et al. [44], and Feng et al. [45,46]
investigated the performance of CFST joints under static loading. An experimental investigation was
conducted by Yin et al. [47] on hysteretic behavior of tubular N-joints. Chen et al. [48] presented
an experimental and numerical investigation on double-skin circular CHS tubular X-joints under
axial compression. The failure mechanism was analyzed and new design equations were proposed.
Hou et al. [49] studied the static behavior of concrete filled double skin steel tubular (CFDST) chord
and CHS brace composite K-joints, and the effects of important parameters on joint strength were
discussed based on experiment results. Moreover, Xu et al. [50], Qian et al. [51], and Tong et al. [52]
carried out studies on the fatigue behavior of CFST joints.
The CFST structures have also been applied to arch trusses in bridges, as well as the truss girders
in steel structures [53–55]. The research on CFST truss has been widely concerned by scholars in recent
years. Experimental and analytical research was carried out by Xu et al. [56] on flexural behavior
of curved concrete filled steel tubular (CCFST) trusses with curved CFST chords and hollow braces.
Test results indicated that the stiffness and load-carrying capacity of CCFST trusses were larger than
those of CFST trusses. Zhou et al. [57] investigated the flexural behavior of 4 circular concrete-filled
stainless steel tubular truss girders. The influence of the location of concrete filling on the flexural
behavior of CHS stainless steel tubular truss girder was analyzed. Mujagic et al. [58] present analytical
and experimental findings pertaining to the design and behavior of composite truss members with
standoff screws as shear connectors. Composite beams constituted by a concrete-encased steel truss
welded to a continuous steel plate were analyzed using a nonlinear finite element formulation based on
Newmark’s classical model by Tullini et al. [59]. Huang et al. [60] tested the truss girders with different
web arrangements and investigated their behaviors, drawing the conclusion that the geometry of the
CHS girder joints was such that only chord face failure and punching shear failure could occur.
Although many researchers have studied the behavior of uniplanar CFST truss and CFST
members, there are few investigations being carried out on the performance of Warren CFST trusses.
In addition, the reinforcing method of adding a half outer sleeve on each joint was used for Warren CHS
tubular trusses and contrasted with the behavior of Warren CFST trusses in this study. Subsequently,
an experimental investigation was conducted in this paper on four types of Warren CHS tubular
Appl. Sci. 2018, 8, 2237 3 of 22
Appl. Sci. 2018, 8, 2237 3 of 21

trusses under static loading. The failure modes, load carrying capacity, overall deflection, and strain
deflection, and strain intensity of all test specimens were discussed. Finite element analysis (FEA)
intensity of all test specimens were discussed. Finite element analysis (FEA) software was used for
software was used for numerical modelling of Warren CHS tubular trusses and the results obtained
numerical modelling
by numerical analysisofwere
Warren CHS tubular
compared trusses
with those and
of the the results obtained
experimental results. by numerical analysis
were compared with those of the experimental results.
2. Experimental Study
2. Experimental Study
2.1. Test Specimens
2.1. Test Specimens
A total of four types of Warren CHS tubular trusses, including the truss with hidden welding
A total of four types of Warren CHS tubular trusses, including the truss with hidden welding
welded (T-HW), the truss with hidden welding unwelded (T-HN), the truss strengthened on the
welded (T-HW), the truss with hidden welding unwelded (T-HN), the truss strengthened on the basis
basis of T-HN by adding a half outer sleeve on each joint (TS-AS), and the truss strengthened on the
of T-HN by adding a half outer sleeve on each joint (TS-AS), and the truss strengthened on the basis of
basis of T-HN by filling concrete in top and bottom chord members (TS-FC), were designed
T-HN by filling concrete in top and bottom chord members (TS-FC), were designed according to the
according to the design guidelines given in the Chinese Standard for Design of Steel Structures
design guidelines given in the Chinese Standard for Design of Steel Structures (GB50017-2017) [13].
(GB50017-2017) [13]. The truss configuration of all specimens is Warren truss with symmetric
The truss configuration of all specimens is Warren truss with symmetric geometry, loading application,
geometry, loading application, and boundary conditions.
and boundary conditions.
The nominal dimensions of CHS members, including chord members, brace members, and
The nominal dimensions of CHS members, including chord members, brace members, and lateral
lateral bracings of all types of multiplanar tubular trusses, are identical for comparison, and the
bracings of all types of multiplanar tubular trusses, are identical for comparison, and the overall length,
overall length, width, and height are 2700 mm, 500 mm, and 450 mm, respectively. The effective
width, and height are 2700 mm, 500 mm, and 450 mm, respectively. The effective span between the end
span between the end supports of the bottom chord members is 2500 mm. The tubular joints are
supports of the bottom
equally spaced in 500 mm chord membersalong
increments is 2500themm.
trussThe tubular
span. The topjoints
and are equally
bottom chord spaced in 500
members ofmm
all
increments
specimens alongare CHS theoftruss span.with
Φ60×2.5 Theantop and diameter
outer bottom chord(D) ofmembers
60 mm and of all specimens
wall thicknessare (T)CHS
of 2.5of
Φ60
mm.×The 2.5 corresponding
with an outer diameter (D) of 60 mmthickness
chord diameter:chord and wall ratio
thickness
(2γ =(T)
D/T)of is
2.5equal
mm.toThe 12.corresponding
The diagonal
chord thickness ratio (2γ = D/T)
brace members of all specimens are CHS of Φ33 × 2 with an outer diameter (d) of 33members
diameter:chord is equal to 12. The diagonal brace mm and wall of all
specimens
thickness (t) areofCHS
2 mm.of Φ33
The × 2 with an outer
corresponding diameter
brace (d) of 33 mm
diameter:chord and wall
diameter ratiothickness
(β = d/D) (t)
andof brace
2 mm.
The corresponding brace diameter:chord diameter ratio (β = d/D) and brace thickness:chord
thickness:chord thickness ratio (τ = t/T) are equal to 0.58 and 0.8, respectively. Besides, for every joint thickness
ratio = t/T) arethe
in all(τspecimens, equal
angle tobetween
0.58 andbrace
0.8, respectively.
and chord (θ)Besides, for every
and the overlap joint
rate in all the
between specimens,
braces (Otheν)
angle between brace and chord (θ) and the overlap rate between the braces (O ) are equal to 60 ◦ and
are equal to 60° and 39.3%, respectively. The lateral bracings of all specimens are CHS of Φ22 × 2 ν
39.3%, respectively.
with outer diameterThe (d1)lateral
of 22 mm bracings of allthickness
and wall specimens (t1are
) of CHS
2 mm. Φ22welds
of The × 2 with outer diameter
connecting brace and (d1 )
ofchord
22 mm and walllateral
members, thickness (t1 ) ofand
bracing, 2 mm.chordThemembers
welds connecting brace and
were designed chord members,
according lateral
to the Chinese
bracing, and chord members were designed according to the Chinese national
national standard for the welding of a steel structure (JGJ81-2002) [61], using advanced welders with standard for the welding
offull
a steel structure and
penetration, (JGJ81-2002)
the weld [61], using
legs were advanced
twice the welders with full penetration,
wall thickness of the diagonal and the weldThe
brace. legs
were twice theofwall
dimensions thickness ofincluding
all specimens the diagonalchordbrace. The dimensions
members, of all specimens
brace members, and lateral including
bracingschord
are
members,
detailed inbrace members,
Figure and in
1 and listed lateral
Tablebracings
1. are detailed in Figure 1 and listed in Table 1.

Loading point Loading point

Diagonal brace
Top chord
30
450
30

Bottom chord
End support End support

100 250 250 250 250 250 250 250 250 250 250 100
2700
(a)

Figure 1. Cont.
Appl. Sci. 2018, 8, 2237 4 of 22
Appl. Sci. 2018, 8, x FOR PEER REVIEW 22 of 22

Figure 1. Warren circular hollow section (CHS) tubular truss. (a) Front view of the tubular truss (mm);
(b) Top view of the tubular truss (mm); (c) Joints detail of the tubular truss.

Table 1. Geometric parameters of specimens.

Chord Brace Lateral Bracing Dimensionless Parameters

Specimen ID
D × T (mm) d × t (mm) d1 × t1 (mm) θ β γ τ Oν
T-HW 60 × 2.5 33 × 2 22 × 2 60o 0.58 12 0.8 39.3%
T-HN 60 × 2.5 33 × 2 22 × 2 60o 0.58 12 0.8 39.3%
TS-AS 60 × 2.5 33 × 2 22 × 2 60o 0.58 12 0.8 39.3%
TS-FC 60 × 2.5 33 × 2 22 × 2 60o 0.58 12 0.8 39.3%

2.2. Material Properties

All specimens were fabricated with Chinese Standard Q235 steel (nominal yield stress
fy = 235 MPa). Tensile coupon tests were conducted according to the test procedures given in the
Chinese Standard of Metallic Materials (GB/T 228-2002) [62] to determine the mechanical properties of
carbon steel CHS tubes. The material properties obtained from the tensile coupon tests are the elastic
modulus (E), tensile yield stress (fy ), ultimate tensile stress (fu ), and elongation. The concrete-filled
Warren CHS tubular trusses were fabricated by filling the concrete with nominal cube strength of
25 MPa in the compression top chord members and the intension bottom chord members. The material
properties of concrete cubes with the nominal length of 150 mm were prepared and tested based
on the recommendations of the Chinese Standard on Ordinary Concrete (GB/T 50081-2002) [63].
The measured concrete cube strength (fcu ) is 18.3 MPa and the measured elastic modulus (EC ) is
27.7 GPa. The obtained material properties data are listed in Table 2.
Appl. Sci. 2018, 8, 2237 5 of 21
Appl. Sci. 2018, 8, 2237 5 of 21

Appl. Sci. 2018, 8, 2237 Table 2. Material properties. 5 of 22

Table 2. Material properties.
Elastic Ultimate
Elastic Tensile Yield Ultimate Elongation δ
Steel Tube Modulus Table
Tensile2. Material
Yield properties.Tensile fy/fu Elongation δ
Steel Tube Modulus Stress f y (MPa) Tensile fy/fu (%)
E (N/mm 2)) E
2 Stress fy (MPa) Stress fu (MPa) (%)
Steel Tube E (N/mm
Elastic Modulus Tensile Yield Stress
Ultimate fu (MPa) f y /f u
Tensile
Elongation δ (%)
Φ60 × 2.5 2.071
(N/mm× 2105
) Stress f268.57
y (MPa) Stress f u 430.14
(MPa) 0.62 17.87
Φ60 × 2.5 2.071 × 105 268.57 430.14 0.62 17.87
Φ33
Φ60 ××2 1.957
2.071 ××105
105 306.21 446.04 0.620.69 19.29
Φ33 × 22.5 1.957 × 105 268.57
306.21 430.14
446.04 0.69 17.87
19.29
Φ33 ××22
Φ22 1.957 ××105
2.123 105 306.21
324.36 446.04462.35 0.690.70 19.29
15.62
Φ22
Φ22××22 2.123
2.123 ×× 105
105 324.36
324.36 462.35
462.35 0.700.70 15.62
15.62
elastic modulus
concrete type elastic
elastic modulus
modulus concrete cube strength (MPa)
concrete
concretetype
type
EcEE(N/mm
c (N/mm 2 ) 22) concrete cube
concrete cube strength
strength (MPa)(MPa)
c (N/mm )
Self-compacting
Self-compacting 2.77
2.77 × 104
× 104 18.3
18.318.3
Self-compacting 2.77 × 104

2.3. Test Program

2.3. Test
Test Program
A schematic sketch of a Warren CHS tubular truss under total vertical loading is shown in
A schematic
schematicsketch
sketchofofa Warren
a Warren CHS CHS tubular
tubular trusstruss
underunder total vertical
total vertical loadingloading
is shown is in
shown
Figurein2.
Figure 2. The
The test2. setup
Figure
test setup
includes
The test
includes
the reaction
setup includes
the reaction
the frame
reaction and frame and
supports,
frame
supports,
hydraulic
and supports,
hydraulic jack
jack (JSYZ-200,
hydraulic
(JSYZ-200,
jack (JSYZ-200,
Xi’an
Xi’an
Xi’an Oriental
Oriental
Metal
Oriental Metal Xi’an,
Structure,
Metal Structure,
China),
Structure, Xi’an,
Xi’an, China)
and load
China) , and load
measuring
, and load
measuring
system (DH3861N,
measuring
system (DH3861N, DONGHUA,
DONGHUA,
system (DH3861N,Jiangsu,DONGHUA,China).
Jiangsu,
The test China).
setup was The test
firmly setup was
connected firmly
to the connected
strong floor.to the
The strong
specimens floor. The
were specimens
simply were at
supported simply
their
Jiangsu, China). The test setup was firmly connected to the strong floor. The specimens were simply
supported
extremities toat their extremities to allow both free in-plane rotation and in-plane longitudinal
supported at allow
their both free in-plane
extremities to allowrotation
bothand in-plane
free longitudinal
in-plane rotation displacement.
and in-plane Compression
longitudinal
displacement.
force was appliedCompression force was applied
at steel blocks at steel blocks welded at one-fourth and three-fourths
displacement. Compression force welded at one-fourth
was applied and three-fourths
at steel blocks of the top
welded at one-fourth chord
and members
three-fourths
of the1000
with topmm chord
in membersaswith
distance the 1000 mm
loading in distance
points using as layers
two the loading
of points
spreader using Atwo
beams. layers
photo andofof
a
of the top chord members with 1000 mm in distance as the loading points using two layers
spreader beams.
schematicbeams. A
of the testphoto and
setupand a schematic
areashown of
in Figure the test setup are shown in Figure 3.
spreader A photo schematic of the3.test setup are shown in Figure 3.

Figure 2. Layout of the Warren CHS tubular truss.

Figure 2.
Figure Layout of
2. Layout of the
the Warren
Warren CHS tubular truss.

Figure 3. Test setup.

Figure 3. Test setup.
An
An initial
initialload was
load applied
was to the
applied to specimens to eliminate
the specimens any possible
to eliminate gaps between
any possible gapsthe specimens
between the
An initial load was applied to the specimens to eliminate any possible gaps between the
and the loading machine. All specimens were subjected to incremental monotonic static
specimens and the loading machine. All specimens were subjected to incremental monotonic static loading, which
specimens and the loading machine. All specimens were subjected to incremental monotonic static
was equally
loading, whichdivided into different
was equally dividedload
intolevels. During
different loadthe tests,During
levels. the applied loads
the tests, theinapplied
each load level
loads in
loading, which was equally divided into different load levels. During the tests, the applied loads in
were gradually increased to 10 kN when the materials were within the elastic range, while
each load level were gradually increased to 10 kN when the materials were within the elastic range, the applied
each load level were gradually increased to 10 kN when the materials were within the elastic range,
loads in each load level were gradually increased to 2 kN when material plasticity was reached.
The readings of the data acquisition system in each load level were recorded with a pause at the
Appl. Sci. 2018, 8, 2237 6 of 21
Appl. Sci. 2018, 8, 2237 6 of 22
while the applied loads in each load level were gradually increased to 2 kN when material plasticity
was reached. The readings of the data acquisition system in each load level were recorded with a
applied
pause atloading level for
the applied 2 min.level
loading The applied
for 2 min. loads
Thewere then loads
applied increased
weretothen
the next loading
increased tolevel and
the next
held in place for another 2 min and the readings of the data acquisition
loading level and held in place for another 2 min and the readings of the data acquisition system system were taken again.
This
weretest procedure
taken wastest
again. This repeated until was
procedure failure of the specimens.
repeated until failure of the specimens.
Five
Five displacement
displacementtransducers
transducersD1–D5
D1–D5(D050,(D050, Jing Ming
Jing Ming Technology,
Technology, Yangzhou,
Yangzhou,China) were were
China) used
to record the deflections of each specimen during the test. The midspan
used to record the deflections of each specimen during the test. The midspan deflection of the deflection of the bottom
chord
bottommembers was monitored
chord members by displacement
was monitored transducer
by displacement D1, andD1,
transducer theand
vertical deflections
the vertical of the
deflections
end tubular joints at the bottom chord members were monitored by displacement
of the end tubular joints at the bottom chord members were monitored by displacement transducers transducers D2
and D3.D3.
D2 and In Inaddition, thethe
addition, out-of-plane
out-of-planedistortions
distortionsofofspecimens
specimenswere weremonitored
monitored by by displacement
displacement
transducers D4 and D5, which were arranged on the both sides of the midspan midspan of the the top
top chord
chord
members. The arrangement of the displacement transducers is shown
displacement transducers is shown in Figure 4a. in Figure 4a.
Two
Two single
single element
element strain
strain gauges
gauges were
were attached
attached at at the
the middle
middle length
length of the the brace
brace and
and chord
chord
members
members at athalf
halfcircumference
circumferenceinterval
intervalforforright
righthalf
halfpart
partof of
thethe
specimens
specimens by taking
by takingadvantage
advantageof the
of
symmetry in geometry, loading application, and boundary conditions
the symmetry in geometry, loading application, and boundary conditions to obtain the strainto obtain the strain intensity of
each specimen
intensity of eachunder different
specimen load different
under levels. The arrangement
load levels. The of arrangement
single elementofstrain
single gauges
elementis shown
strain
in Figure 4b–d for the front view, top view, and bottom view, respectively.
gauges is shown in Figure 4b–d for the front view, top view, and bottom view, respectively. The The strain along the welded
connection
strain alongofthe chord and brace
welded members
connection were measured
of chord and braceby four multi-axial
members strain gauges
were measured by four (strain gauge
multi-axial
rosettes) attached
strain gauges around
(strain the weld
gauge toe of
rosettes) connection
attached joints.
around theThe arrangement
weld of multi-axial
toe of connection joints.strain
The
gauges is shown in Figure 4e–g for the joints of top chord, midspan joints of
arrangement of multi-axial strain gauges is shown in Figure 4e–g for the joints of top chord, midspanbottom chord, and end
joints
joints of
of bottom
bottom chord,
chord, respectively.
and end joints of bottom chord, respectively.

3 2
g

e D3
3 D2 D1 2

D4 D5

2-2 3-3

(a)
Strain gague Strain gague

Strain gague
Strain gague
(b) (c)

Figure 4. Cont.
Appl. Sci. 2018, 8, 2237 7 of 22
Appl. Sci. 2018, 8, 2237 7 of 21

Strain gague
Multi-axial strain gague

Multi-axial strain gague

Strain gague
(d) (e)
Multi-axial strain gague Multi-axial strain gague

Multi-axial strain gague Multi-axial strain gague

(f) (g)

Figure 4.4. Arrangement

Arrangement of of displacement
displacement transducers
transducers and strain gauges. (a) (a) Arrangement
Arrangement of of the
displacement transducers;(b)
displacement transducers; (b)Front
Front view
view of single
of single element
element strainstrain gauges;
gauges; (c) Top(c) Topofview
view singleofelement
single
element strain (d)
strain gauges; gauges;
Bottom(d)view
Bottom view of
of single single strain
element element strain(e)
gauges; gauges;
Joints (e) Joints
of top of top
chord; (f)chord;
Midspan (f)
joints of bottom
Midspan chord;
joints of (g)chord;
bottom End joints of bottom
(g) End chord.
joints of bottom chord.

3. Test
3. Test Results
Results
3.1. Failure Mode
3.1. Failure Mode
For the layout of all the specimens, as shown in Figure 2, the typical failure modes for the three
For the layout of all the specimens, as shown in Figure 2, the typical failure modes for the three
empty tubular trusses—T-HW, T-HN, and TS-AS—are basically the same, and the bottom chord
empty tubular trusses—T-HW, T-HN, and TS-AS—are basically the same, and the bottom chord has
has different degrees of bending along the span of the truss. For specimen T-HW, distinct concave
different degrees of bending along the span of the truss. For specimen T-HW, distinct concave
deformation occurred around the tubular joints ‘k’ and ‘p’ at the bottom chord member, and the
deformation occurred around the tubular joints ‘k’ and ‘p’ at the bottom chord member, and the
bending deformation occurred at the bottom chord member. The failure of this specimen resulted from
bending deformation occurred at the bottom chord member. The failure of this specimen resulted
the surface plasticity of the bottom chord member, and the bending of the bottom chord member, as
from the surface plasticity of the bottom chord member, and the bending of the bottom chord
shown in Figure 5. The failure of specimen T-HN was similar to specimen T-HW, while it had a greater
member, as shown in Figure 5. The failure of specimen T-HN was similar to specimen T-HW, while
degree of the bending deformation of the bottom chord member and surface plastic deformation
it had a greater degree of the bending deformation of the bottom chord member and surface plastic
around the tubular joints ‘k’ and ‘p’ at the bottom chord member, as shown in Figure 6. For specimen
deformation around the tubular joints ‘k’ and ‘p’ at the bottom chord member, as shown in Figure 6.
TS-AS, the concave deformation occurred around the tubular joints ‘k’ and ‘p’ at the bottom chord
For specimen TS-AS, the concave deformation occurred around the tubular joints ‘k’ and ‘p’ at the
member, and the weld fracture occurred around the tubular joint ‘n’ along the diagonal brace ‘fn’,
bottom chord member, and the weld fracture occurred around the tubular joint ‘n’ along the
while the local buckling of diagonal brace ‘bk’ around the tubular joint ‘k’ was observed. In addition,
diagonal brace ‘fn’, while the local buckling of diagonal brace ‘bk’ around the tubular joint ‘k’ was
the bottom chord bent along the truss span. It should be noted that the concave deformation around
observed. In addition, the bottom chord bent along the truss span. It should be noted that the
tubular joints ‘k’ and ‘p’ of TS-AS were less than the concave deformation of specimen T-HW and
concave deformation around tubular joints ‘k’ and ‘p’ of TS-AS were less than the concave
T-HN, while its concave deformation areas of tubular joints ‘k’ and ‘p’ were more extensive than that
deformation of specimen T-HW and T-HN, while its concave deformation areas of tubular joints ‘k’
of the T-HW and T-HN specimens. The failure of this specimen resulted from the local buckling of the
and ‘p’ were more extensive than that of the T-HW and T-HN specimens. The failure of this
diagonal brace member, the surface plasticity around the tubular joints at the bottom chord member,
specimen resulted from the local buckling of the diagonal brace member, the surface plasticity
and the bending of the bottom chord, as shown in Figure 7. For specimen TS-FC, when reaching the
around the tubular joints at the bottom chord member, and the bending of the bottom chord, as
maximum load the test setup could afford, it did not reach failure state and no obvious deformation
shown in Figure 7. For specimen TS-FC, when reaching the maximum load the test setup could
occurred at each member and joint with slight overall flexure.
afford, it did not reach failure state and obvious deformation occurred at each member and joint
It was shown from the tests that the failure modes of the trusses with hidden welds welded or
with slight overall flexure.
not are basically the same, which are the bending of the bottom chord and surface plasticity of the
It was shown from the tests that the failure modes of the trusses with hidden welds welded or
bottom chord member. In addition, it is shown from the test that TS-AS failed by the local buckling of a
not are basically the same, which are the bending of the bottom chord and surface plasticity of the
bottom chord member. In addition, it is shown from the test that TS-AS failed by the local buckling
of a diagonal brace member and the surface plasticity of the bottom chord member because each
Appl. Sci. 2018, 8, 2237 8 of 22
Appl.
Appl. Sci.
Sci. 2018,
2018, 8,
8, 2237
2237 88 of
of 21
21

diagonal
joint brace member and the surface plasticity of the bottom chord member because each joint was
joint was
was reinforced
reinforced with with aa half
half outer
outer sleeve
sleeve onon it.
it. The
The TS-FC
TS-FC specimen
specimen did did not
not fail
fail until
until the
the test
test
reinforced
setup with a half outer sleeve on it. The TS-FC specimen did not fail until the test setup reached
setup reached the maximum load it could impose, which demonstrated that the reinforcement of
reached the maximum load it could impose, which demonstrated that the reinforcement of
the maximum load
filling it couldtopimpose,bottom
which demonstrated that the reinforcement of fillingreinforce
concrete into
filling concrete
concrete into
into thethe top andand bottom members members made made aa greater
greater contribution
contribution to to reinforce thethe
the top and
Warren bottom members made a greater contribution to reinforce the Warren CHS tubular truss
Warren CHS CHS tubular
tubular trusstruss than
than adding
adding aa halfhalf outer
outer sleeve
sleeve toto each
each joint.
joint. It
It was
was shown
shown from
from thethe
than adding
comparison a half
that outer
the sleeve
failure to
modes eachof joint.
T-HW It was
and shown
T-HN from
are the comparison
different from the that
two the failure
specimens modes
with
comparison that the failure modes of T-HW and T-HN are different from the two specimens with
of T-HW and T-HN are different
reinforcement—TS-AS from the two specimens with reinforcement—TS-AS and TS-FC—and
reinforcement—TS-AS and and TS-FC—and
TS-FC—and the the latter
latter two
two specimens
specimens can can improve
improve the the mechanical
mechanical
the latter
behavior two
and specimens
integral can
stiffnessimprove
of the
trusses. mechanical behavior and integral stiffness of trusses.
behavior and integral stiffness of trusses.

Figure
Figure 5.
5. Failure
Failure modes
modes of
of specimen
specimen T-HW.
T-HW.
T-HW.

Figure
Figure 6.
6. Failure
Failure modes
modes of
of specimen
specimen T-HN.
T-HN.
T-HN.
Appl. Sci. 2018, 8, 2237 9 of 22
Appl. Sci. 2018, 8, 2237 9 of 21

Figure 7.
Figure 7. Failure modes of specimen TS-AS.

3.2.
3.2. Load
Load Carrying
Carrying Capacity
Capacity
The
The yield
yield loads
loads (F (Fyy)) which
which initiate
initiate the
the yielding
yielding of of truss
truss members
members and and the the peak
peak loadload (F(Fpp)) of
of all
all
specimens
specimens are are summarized
summarized in in Table
Table 3.3. It
It isis shown
shown from from thethe comparison
comparison that that the
the peak
peak loads
loads (F (Fpp)) of
of all
all
specimens
specimens werewere increased
increased roughly
roughly byby 10%
10% from from the the yield
yield loads
loads (F ), except
(Fyy), except for for the
the specimen
specimen TS-FC,TS-FC,
whose
whose peak load (Fpp) and yield load (Fyy) were not obtained because of the limit of test setup, which
peak load (F ) and yield load (F ) were not obtained because of the limit of test setup, which
shows
shows that
that yield
yield load
load (F (Fyy)) and
and peak
peak load
load (F (Fpp)) of
of specimen
specimen TS-FCTS-FC are are the
the largest
largest amongamong thesethese fourfour
specimens.
specimens. On the other hand, it can be seen from Table 3 that the yield loads (Fy) and peak loads (F
On the other hand, it can be seen from Table 3 that the yield loads (F y ) and peak loads (Fpp))
of
of specimens T-HN, T-HW,
specimens T-HN, T-HW, TS-AS,TS-AS, and
and TS-FC
TS-FC are are inin ascending
ascending order.
order. Compared
Compared to to the
the load
load carrying
carrying
capacity
capacity of of specimen
specimenT-HW, T-HW,the theload
loadcarrying
carrying capacity
capacity of specimen
of specimen T-HNT-HN waswas decreased
decreased by 18%,
by 18%,and
the load carrying capacity of specimen TS-AS increased by 36%,
and the load carrying capacity of specimen TS-AS increased by 36%, whereas the load carrying whereas the load carrying capacity of
specimen
capacity ofTS-FC increased
specimen TS-FC byincreased
more thanby60%. more It is
thanillustrated
60%. It from the comparison
is illustrated from thethat being hidden
comparison that
weld welded or not influences the load carrying capacity, and the reinforcing
being hidden weld welded or not influences the load carrying capacity, and the reinforcing method method of filling concrete
into the top
of filling and bottom
concrete into thechordtopmembers
and bottom canchordimprove the load
members cancarrying
improve capacity
the load of Warren
carryingCHS tubular
capacity of
trusses
Warrenbetter than adding
CHS tubular trusses a half outer
better than sleeve
adding on each
a halfjoint,
outerwhich
sleevemeanson each thejoint,
reinforcing
which method
means the of
filling concrete
reinforcing methodinto the top and
of filling bottominto
concrete chord the members
top and bottom is much moremembers
chord efficient is inmuch
practice.
more efficient
in practice.
Table 3. Load carrying capacity of specimens.
Table 3. Load carrying capacity of specimens.
Yield Load Peak Load Comparison
Specimen
Fy (kN)
Yield Load Peak FpLoad Fp /Fy
(kN) Comparison
Specimen
T-HW Fy100
(kN) Fp (kN)
110 Fp/Fy 1.1
T-HN
T-HW 80
100 11090 1.1 1.13
TS-AS
T-HN 140
80 90150 1.13 1.07
TS-FC
TS-AS —-
140 150—- 1.07 —-
TS-FC ---- ---- ----
3.3. Overall Deflection
3.3. Overall Deflection
Figure 8 shows the total vertical loading–deflection curves of all specimens. In Figure 8, the
Figureaxis
horizontal 8 shows the total
represents the vertical
vertical loading–deflection curves
deflection (w) of the bottomof all specimens.
chord memberInobtained
Figure 8,from
the
horizontal axistransducer
displacement representsD1theand
vertical
D3, anddeflection (w) of
the vertical therepresents
axis bottom chord member
the total obtained
vertical from
loading (P)
displacement
applied transducer
to the loading D1 at
points and D3, and and
one-fourth the three-fourths
vertical axis represents the total
of the top chord vertical
members loading
with 1000 mm(P)
applied
in to the loading points at one-fourth and three-fourths of the top chord members with 1000
distance.
mm in distance.
It is shown from the comparison that the midspan deflection of all specimens is greater than
that of the joint ‘p’, however, the changing trends of the initial stiffness and the vertical deflection of
Appl. Sci. 2018, 8, 2237 10 of 22

It is
Appl. Sci. shown
2018, 8, 2237from the comparison that the midspan deflection of all specimens is greater 10 than
of 21
that of the joint ‘p’, however, the changing trends of the initial stiffness and the vertical deflection of
all
all specimens
specimens are are quite
quite similar.
similar. The
The initial
initial slopes
slopes inin the
the elastic
elastic stage
stageof of the
thetest
test curves
curvesof ofthe
theT-HN,
T-HN,
T-HW, TS-AS, and
T-HW, TS-AS, andTS-FC
TS-FCspecimens
specimensare areininascending
ascendingorder,order,which
whichmeans
means their
their integral
integral rigidities
rigidities areare
in
in ascending
ascending order
order too.too. In other
In other words,
words, thethe reinforcing
reinforcing method
method of adding
of adding a halfa half
outerouter
sleevesleeve on each
on each joint
joint and filling
and filling concrete
concrete intochord
into the the chord members
members can effectively
can effectively improve
improve stiffness
stiffness of truss.
of the the truss.
With With
the
the gradual
gradual increase
increase of total
of total vertical
vertical load,load, the deformation
the deformation of theof the develops
truss truss develops
from the from theoverall
initial initial
overall
deflectiondeflection to local deformation
to local deformation at the
at the joints. Thejoints. The deflections
deflections of the TS-AS,of the TS-AS,
T-HN, andT-HN,
T-HWand T-HW
specimens
specimens are in descending
are in descending order,that
order, meaning meaning
specimen that TS-AS
specimenhasTS-AS
the besthasdeformability,
the best deformability,
while specimenwhile
specimen
T-HN has T-HN has better deformability
better deformability than specimen thanT-HW.
specimen T-HW.specimen
Moreover, Moreover, specimen
TS-FC did notTS-FC
fail at did not
the end
fail
of theat the
test,end
andof thethe test,ofand
slope the slope
specimen of specimen
TS-FC TS-FC
in the elastic in the
stage elastic
is larger stage
than is of
that larger than that
specimen of
TS-AS,
specimen
which means TS-AS,
that which means that
the reinforcing the of
method reinforcing method
filling concrete intoofthe
filling
top andconcrete
bottom into the members
chord top and
bottom chord members
could constrain could constrain
the deformation at jointsthe
moredeformation
effectively.at joints more effectively.

180 180

160 160

140 140
120 120
P/kN
P/kN

100 100
80 80
60 60 T-HN
T-HN
40 T-HW 40 T-HW
TS-AS TS-AS
20 TS-FC 20 TS-FC
0 0
0 2 4 6 8 10 12 14 16 0 1 2 3 4 5 6
w/mm w/mm
(a) (b)

Figure
Figure 8.8. Total
Total vertical loading–deflection curves of specimens. (a)
(a) Midspan
Midspan (transducer
(transducer D1);
D1); (b)
(b) End
End
joint
joint ‘p’
‘p’ (transducer
(transducer D3).

3.4. Strain
3.4. Strain Intensity
Intensity
The total
The total vertical
verticalloading
loadingversusversusaxialaxialstrain
straincurves
curves ofofallall specimens
specimens areare plotted
plotted in Figure
in Figure 9, in9,
in which
which the the horizontal
horizontal axis axis represents
represents the axial
the axial strainstrain
(ε) of (ε) of joints,
joints, brace, brace,
and chord and members,
chord members, where
where positive indicates tensile strain and negative indicates
positive indicates tensile strain and negative indicates compressive strain, and the compressive strain, and thevertical
vertical axis
axis
represents the
represents the total
total vertical
vertical loading
loading(P) (P)applied
appliedtotothe
theloading
loadingpointspointsatatone-fourth
one-fourth andandthree-fourths
three-fourths of
the top chord members with 1000 mm in distance. It is illustrated
of the top chord members with 1000 mm in distance. It is illustrated from the comparison that the from the comparison that the axial
strains
axial of theoftop
strains thechord members
top chord membersin compression are much
in compression smaller
are much than the
smaller thanyield
the strain (1297 ×
yield strain 10−6×)
(1297
of−6the
10 ) ofsteel
the materials,
steel materials,whereas the axial
whereas the strains of the bottom
axial strains chord members
of the bottom chord membersin tension are smaller
in tension are
than the yield strain of the steel materials, and the axial strains
smaller than the yield strain of the steel materials, and the axial strains of the diagonal braceof the diagonal brace members are
close to the yield strain (1565 × 10 −6 ) of the steel materials, as shown in Figures 9 and 10 for top chord
members are close to the yield strain (1565 × 10−6) of the steel materials, as shown in Figures 9a–e and
members,
10 for top chordbottom chord member,
members, bottomand chord diagonal
member, brace
andmembers,
diagonal respectively.
brace members, respectively.
For the top chord members in compression,
For the top chord members in compression, as shown in as shown in Figure
Figure 9a–d,
9a–d, thethe initial
initial slopes
slopes ofof the
the
test curves
test curves ofofspecimens
specimens T-HW,
T-HW, T-HN,T-HN,and and
TS-AS are almost
TS-AS identical,
are almost which are
identical, smaller
which are tosmaller
specimen to
TS-FC. Furthermore, the axial strains of specimens T-HW, T-HN,
specimen TS-FC. Furthermore, the axial strains of specimens T-HW, T-HN, and TS-AS under the and TS-AS under the same load level
are almost
same load identical,
level arewhich almostareidentical,
larger than that of
which arethelarger
specimenthanTS-FC.
that ofFor thethespecimen
bottom chordTS-FC. member
For the in
tension,chord
bottom as shown memberin Figure 9e, the as
in tension, testshown
curvesinofFigure
these four
9e, thespecimens
test curves are ofnearly
theseidentical. It is shown
four specimens are
from the comparison that the filled concrete makes a greater
nearly identical. It is shown from the comparison that the filled concrete makes a greater contribution to the top chord members
in compression
contribution than
to the toptochord
the bottom
members chord member in tension.
in compression than to the Forbottom
the diagonal brace members,
chord member in tension. as
shown
For the in Figure 10,
diagonal bracethe members,
axial strains asof all specimens
shown in Figureunder 10, thedifferent load levels
axial strains of allare nearly identical.
specimens under
different load levels are nearly identical. In addition, the strains at crown positions onand
In addition, the strains at crown positions on the joint ‘p’ of the specimens T-HW theT-HN vary
joint ‘p’ of
greatly,
the which means
specimens T-HW there is a case
and T-HN varyof stress concentration,
greatly, which means whereas
there theis a strains
case ofatstress
the same position of
concentration,
specimens
whereas theTS-AS
strainsand TS-FC
at the same vary uniformly,
position as shownTS-AS
of specimens in Figureand11. It is vary
TS-FC shown from the as
uniformly, comparison
shown in
Figure 11. It is shown from the comparison that the reinforcing methods of adding a half outer sleeve
on each joint and filling concrete into the top and bottom chord members could help to improve the
force condition of joints. On the other hand, the strains at saddle positions on joint ‘n’ of the four
specimens vary uniformly.
Appl. Sci. 2018, 8, 2237 11 of 22

that the reinforcing methods of adding a half outer sleeve on each joint and filling concrete into the top
and bottom chord members could help to improve the force condition of joints. On the other hand, the
strains
Appl. Sci.at saddle
2018, 8, 2237positions on joint ‘n’ of the four specimens vary uniformly. 11 of 21

180 180
160 160
140 140
120 120

P/kN
100 100
P/kN

80 80
TS-FC
60 TS-FC
T-HW 60
T-HW
TS-AS
40 40 TS-AS
T-HN
T-HN
20 20
0 0
-600 -500 -400 -300 -200 -100 0 -600 -500 -400 -300 -200 -100 0
-6 -6
ε/10 ε/10
(a) (b)
180 180
160 160
140 140
120 120
100 100
P/kN
P/kN

80 80
TS-FC
60 TS-FC
60 T-HW T-HW
TS-AS 40 TS-AS
40
T-HN T-HN
20
20
0
0 -700 -600 -500 -400 -300 -200 -100 0
-500 -400 -300 -200 -100 0 -6
-6 ε/10
ε/10
(c) (d)
180
160
140
120
100
P/kN

80
60 TS-FC
T-HW
40 TS-AS
20 T-HN

0
0 200 400 600 800 1000 1200
-6
ε/10
(e)

Figure 9. 9. Total
Totalvertical
verticalloading–axial
loading–axialstrain
strain curves
curves of chord
of chord members.
members. (a) Top
(a) Top chordchord ‘hj’;
‘hj’; (b) Top(b)chord
Top
chord ‘fh’; (c) Top chord ‘gi’; (d) Top chord ‘eg’; (e) Bottom
‘fh’; (c) Top chord ‘gi’; (d) Top chord ‘eg’; (e) Bottom chord ‘no’. chord ‘no’.

180 180
160 160
140 140
120 120
100
P/kN

100
P/kN

80 80
TS-FC TS-FC
60 T-HW 60 T-HW
40 TS-AS 40 TS-AS
T-HN T-HN
20 20
0 0
-1400 -1200 -1000 -800 -600 -400 -200 0 0 200 400 600 800 1000 1200
-6 -6
ε/10 ε/10
(a) (b)
 0
0 200 400 600 800 1000 1200
-6
ε/10
(e)

Figure
Appl. Sci. 2018,9. Total vertical loading–axial strain curves of chord members. (a) Top chord ‘hj’; (b) Top
8, 2237 12 of 22
chord ‘fh’; (c) Top chord ‘gi’; (d) Top chord ‘eg’; (e) Bottom chord ‘no’.

180 180
160 160
140 140
120 120
100
P/kN

100

P/kN
80 80
TS-FC TS-FC
60 T-HW 60 T-HW
40 TS-AS 40 TS-AS
T-HN T-HN
20 20
0 0
-1400 -1200 -1000 -800 -600 -400 -200 0 0 200 400 600 800 1000 1200
Appl. Sci. 2018, 8, 2237 ε/10
-6
ε/10
-6 12 of 21
Appl. Sci. 2018, 8, 2237 12 of 21
(a) (b)
180 180
180 180
160 160
160 160
140 140
140 140
120 120
120 120
100 100
P/kN

P/kN
100 100
P/kN

P/kN
80 80
80 80 TS-FC
TS-FC TS-FC
60 TS-FC 60 T-HW
60 T-HW 60 T-HW
T-HW TS-AS
40 TS-AS 40 TS-AS
40 TS-AS 40 T-HN
T-HN T-HN
20 T-HN 20
20 20
0 0
0
-1400 -1200 -1000 -800 -600 -400 -200 0 00 200 400 600 800 1000 1200 1400
-1400 -1200 -1000 -800 -600
-6 -400 -200 0 0 200 400 600 800
-6 1000 1200 1400
ε/10 -6 ε/10 -6
ε/10 ε/10
(c) (d)
(c) (d)
Figure 10. Total vertical loading–axial strain curves of diagonal braces. (a) Diagonal brace ‘jp’; (b)
Figure 10. Total
Totalvertical
verticalloading–axial
loading–axialstrain
straincurves
curvesofof
diagonal
diagonal braces. (a) (a)
braces. Diagonal brace
Diagonal ‘jp’; ‘jp’;
brace (b)
Diagonal brace ‘jo’; (c) Diagonal brace ‘ip’; (d) Diagonal brace ‘io’.
Diagonal
(b) brace
Diagonal ‘jo’;‘jo’;
brace (c) Diagonal brace
(c) Diagonal ‘ip’;‘ip’;
brace (d) (d)
Diagonal brace
Diagonal ‘io’.‘io’.
brace

180 180
180 180
160 160
160 160
TS-FC
140 TS-FC 140
140 T-HW 140
T-HW
120 TS-AS 120
120 TS-AS
T-HN 120
P/kN

100 T-HN 100

P/kN
P/kN

100 100
P/kN

80 80
80 80 TS-FC
TS-FC
60 60 T-HW
60 60 T-HW
TS-AS
40 40 TS-AS
40 40 T-HN
T-HN
20 20
20 20
0 0
0 -3000 -2000 -1000 00 200 400 600 800 1000 1200
0 1000 2000 3000 4000 0 200 400 600-6 800 1000 1200
-3000 -2000 -1000 0 -6 1000 2000 3000 4000 ε/10 -6
ε/10 -6 ε/10
ε/10
(a) (b)
(a) (b)
Figure 11. Total vertical
vertical loading–axialstrain
straincurves
curvesofofthethe crown and saddle position for joints. (a)
Figure 11. Total
Figure Total verticalloading–axial
loading–axial strain curves of crown
the crownand saddle
and position
saddle for joints.
position (a) The
for joints. (a)
The crown
crown position
position of joint
of joint ‘p’;The
‘p’; (b) (b) The saddle position of joint
‘n’. ‘n’.
The crown position of joint ‘p’; (b) saddle position
The saddle of joint
position of joint ‘n’.
4. Finite
4. Finite Element Analysis
Analysis
4. Finite Element
Element Analysis
4.1. General
4.1.
4.1. General
General
The general purpose finite element program ANSYS [64] (ANSYS 10.0, ANSYS-Inc.,
The
The general
Pittsburgh, general
PA, USA,
purpose
purpose finite element
finiteused
2006) was element program ANSYS
for theprogram
numericalANSYS modelling[64]of(ANSYS
[64] (ANSYS
Warren CHS
10.0, ANSYS-Inc.,
10.0,tubular
ANSYS-Inc.,
trusses.
Pittsburgh,
Pittsburgh, Penn,
Penn, America,
America, 2006)
2006) was
was used
used for
for the
the numerical
numerical modelling
modelling of
of Warren
Warren CHS
CHS tubular
tubular
The load–displacement nonlinear analysis was performed by using the full Newton–Raphson iteration
trusses. The
The load–displacement
trusses. available load–displacement nonlinear
nonlinear analysis
analysis was
was performed
performed by
by using
using the full
the taken
full
method in the ANSYS library. Both material and geometric nonlinearities have been
Newton–Raphson
Newton–Raphson iteration method available in the ANSYS library. Both material and geometric
into account in theiteration method
finite element available
models. Thein the ANSYS
failure modes,library. Both material
load carrying andand
capacities, geometric
overall
nonlinearities
nonlinearities have
have been
been taken
taken into
into account
account in
in the
the finite
finite element
element models.
models. The
The failure
failure modes,
modes, load
load
deflections were obtained from the numerical analysis. The finite element analysis includes various
carrying capacities, and overall deflections were obtained from the numerical
carrying capacities, and overall deflections were obtained from the numerical analysis. The finite analysis. The finite
element
element analysis
analysis includes
includes various
various important
important factors,
factors, such
such asas modelling
modelling of of materials,
materials, finite
finite element
element
type
type and mesh size, weld model, contact interaction between the steel tube and concrete infill,
and mesh size, weld model, contact interaction between the steel tube and concrete infill,
contact interaction between brace members and chord members, contact interaction
contact interaction between brace members and chord members, contact interaction between welds between welds
and
and steel
steel tubes,
tubes, and
and loading
loading and
and boundary
boundary conditions.
conditions.
Appl. Sci. 2018, 8, 2237 13 of 22

important factors, such as modelling of materials, finite element type and mesh size, weld model,
contact interaction between the steel tube and concrete infill, contact interaction between brace
members and chord members, contact interaction between welds and steel tubes, and loading and
boundary conditions.

4.2. Material Modelling

The bilinear isotropic kinematic hardening model was used for the material modelling of steel
tubes. The bilinear stress–strain curve of the steel tube is shown in Figure 12, which includes an
elastic stage and a strengthening stage, and the modulus magnitude of the strengthening stage
is 1% of that of the elastic stage, in accordance with Shao et al. [65]. The multilinear isotropic
hardening plasticity model was adopted for the material modelling of concrete infill, in which the
Appl.Sci.
Sci.
initial
Appl. 2018,of
part
2018, 8,8,2237
2237multi-linear stress–strain curve represents the elasticity with the measured13
the 13 of 21
elastic
of 21
modulus (Ec ) of 27.7 GPa and Poisson’s ratio (ν) of 0.2, while the tension and compression stress–strain
relationshipwas
relationship
relationship wasobtained
obtainedfrom
obtained fromthe
from theprovisions
the provisionsof
provisions ofthe
of theChinese
the ChineseCode
Chinese Codefor
Code forDesign
for Designof
Design ofConcrete
of ConcreteStructures
Concrete Structures
Structures
(GB50010-2010) [66],
(GB50010-2010) [66], as
as shown
shown in
in Figure
Figure 13.
13.

σσ
ffuu
ffy
y
0.01E
0.01E
11

εε
εεyy εεuu
Figure12.
Figure
Figure 12.Bilinear
12. Bilinearstress–strain
Bilinear stress–straincurve
stress–strain curveof
curve ofsteel
of steeltube.
steel tube.
tube.

σσ
ffc,rc,r

0.5ffc,r
0.5 c,r

εεt,rt,r
εεc,rc,r εεcucu εε
fft,rt,r

Figure13.
Figure
Figure 13.Stress–strain
13. Stress–strainmodel
Stress–strain model used
model usedfor
used for confined
for confinedconcrete.
confined concrete.
concrete.

4.3. Finite
4.3. Finite Element
Finite Element
Element TypeType and
Type and Mesh
and Mesh Size
Mesh Size
Size
4.3.
In this
In this study,
this study,
study, thethe Shell181
the Shell181 element
Shell181 element available
element available
available in in the
in the ANSYS
the ANSYS library
ANSYS library was
library was used
was used
used toto model
to model
model the the steel
the steel
steel
In
tube of Warren
Warren CHS
CHS tubular
tubular trusses
trusses since
since this
this element
element is
is suitable
suitable for
for analyzing
analyzing thin to moderately
moderately
tube of Warren CHS tubular trusses since this element is suitable for analyzing thin to moderately
thick shell
thick shell structures.
shell structures.
structures. The The Shell181
The Shell181 element
Shell181 element
element is is
is aaa 4-node
4-node element
4-node element
element withwith six
with six degrees
six degrees
degrees of of freedom
of freedom
freedom at at each
at each
each
thick
node: translations
translations ininthethex, x,
y, and
y, z axes.
and z In addition,
axes. In Shell181
addition, is well-suited
Shell181 is for linear,
well-suited for larger
linear,rotation,
larger
node: translations in the x, y, and z axes. In addition, Shell181 is well-suited for linear, larger
and largeand
rotation, strain
largenonlinear applications.
strain nonlinear
nonlinear applications.
rotation, and large strain applications.
The
The solid
solid element
element has beenbeen
has used by usedmanybyresearchers for finite element
many researchers
researchers for finiteanalysis
finite elementof concrete-filled
analysis of of
The solid element has been used by many for element analysis
tubular trusses.tubular
concrete-filled In this trusses.
study, the In Solid65
this element
study, the available
Solid65 in theavailable
element ANSYS library
in the was used
ANSYS to model
library was
concrete-filled tubular trusses. In this study, the Solid65 element available in the ANSYS library was
the
used concrete
to model
modelinfill
thesince this element
concrete infill sinceis used
since for the 3-D
this element
element modeling
is used
used for theof 3-D
the solids withoutof
modeling reinforcing
solids withoutbars,
without
used to the concrete infill this is for 3-D modeling of solids
and the solid
reinforcing is capable
bars, and thetheofsolid
cracking
solid in tension
is capable
capable and crushing
of cracking
cracking in compression.
in tension
tension and crushing
crushing Thein element
compression.Solid65 is
The
reinforcing bars, and is of in and in compression. The
element Solid65 is defined by eight nodes with three degrees of freedom
element Solid65 is defined by eight nodes with three degrees of freedom at each node: translations in at each node: translations in
the nodal
the nodal x, x, y,
y, and
and zz directions.
directions.
In order to guarantee
In order to guarantee the the mesh
mesh quality,
quality, aa sub-zone
sub-zone mesh mesh generation
generation method
method was was used
used during
during
the finite
the finite element
element modeling.
modeling. In In this
this method,
method, the the entire
entire Warren
Warren CHS CHS tubular
tubular truss
truss was
was divided
divided intointo
two different zones. One is the mesh dense zone at every joint, meshed
two different zones. One is the mesh dense zone at every joint, meshed by free meshing. The other is by free meshing. The other is
 Appl. Sci. 2018, 8, 2237 14 of 22

defined by eight nodes with three degrees of freedom at each node: translations in the nodal x, y, and
z directions.
In order to guarantee the mesh quality, a sub-zone mesh generation method was used during the
finite element modeling. In this method, the entire Warren CHS tubular truss was divided into two
different zones. One is the mesh dense zone at every joint, meshed by free meshing. The other is the
mesh loose zone between tubular joints, meshed by mapped meshing. The convergence studies were
carried out to obtain the optimum finite element mesh size. It was found that the smart size of 10 was
appropriate for the mesh dense zone. The ratio of the long side to short side of the mesh element of
about 16 could achieve accurate results with minimum computational time. The typical finite element
Appl. Sci. 2018, 8, 2237 14 of 21
mesh of a Warren CHS tubular truss is shown in Figure 14.

Figure
Figure14.
14.Finite
Finiteelement
elementmesh
meshofofa aWarren
WarrenCHS
CHStubular
tubulartruss.
truss.

4.4.Weld
4.4. WeldModeling
Modelingand
andSteel
SteelTube–Concrete
Tube–ConcreteInfill
InfillInterface
InterfaceModel
Model
Weldmodeling
Weld modelingisisseldom seldom taken
taken into
into consideration
consideration in finite
in finite element
element analysis.
analysis. However,
However, it wasit
was found that if the effect of welding is not considered, the numerical simulations
found that if the effect of welding is not considered, the numerical simulations of the load carrying of the load
carryingand
capacity capacity and joint
joint rigidity arerigidity
smallerare thansmaller than
those of those of experimental
experimental results [67]. results [67]. Corigliano
In addition, In addition,
Corigliano et al. [68,69] pointed out that in the nonlinear finite element
et al. [68,69] pointed out that in the nonlinear finite element analysis of welded joints, analysis of welded joints,the
the
mechanical properties of weld and steel parent metal should be considered differently.
mechanical properties of weld and steel parent metal should be considered differently. In this study, In this study,
Shell181was
Shell181 wasused
usedfor forweld
weldmodeling
modelingby byadding
addingaacircle
circleofofShell181
Shell181around
aroundthetheintersection
intersectionofofbrace
brace
and chord members, and the previous test results of weld metal under
and chord members, and the previous test results of weld metal under monotonic loading weremonotonic loading were used
for the material properties of welds (see reference [35] for details). It should be noted
used for the material properties of welds (see reference [35] for details). It should be noted that the that the contact
element
contact was used
element wasforusedsimulating the casethe
for simulating of hidden
case of weld
hidden unwelded. The weldThe
weld unwelded. modeling is shownisin
weld modeling
Figure 15.
shown in Figure 15.
The contact interaction between the steel tube and concrete infill was simulated by setting contact
pairs with the contact elements. t The steel tube was defined to be the target surface and the outer
surface of concrete infill was defined to be the contact surface. In addition, the friction coefficient (µ)
should be defined to characterize Brace the friction
member
behaviors of the contact interaction between the steel
tube and concrete infill.
Shell element The friction coefficient (µ) was suggested to be from 0.2 to 0.6, and 0.4 was
selected in this study.

Weld
Chord member
wh
T

ww
ww = t
wh = 0.5T+t
mechanical properties of weld and steel parent metal should be considered differently. In this study,
Shell181 was used for weld modeling by adding a circle of Shell181 around the intersection of brace
and chord members, and the previous test results of weld metal under monotonic loading were
used for the material properties of welds (see reference [35] for details). It should be noted that the
contact element was used for simulating the case of hidden weld unwelded. The weld modeling is
Appl. Sci. 2018, 8, 2237 15 of 22
shown in Figure 15.

Brace member
Shell element

Weld
Chord member
wh
T

ww
ww = t
wh = 0.5T+t

Figure
Figure 15.
15. Weld
Weld modelling.

4.5. Loading and Boundary

The contact Conditions
interaction between the steel tube and concrete infill was simulated by setting
contact pairs with the contact elements.
In order to make the boundary The steel of
conditions tube waselement
finite definedmodel
to be the
thetarget
same surface
as thoseandof the
the
outer surface of concrete infill was defined to be the contact surface. In addition,
experimental specimens, the end supports at the bottom members and loading points were restrainedthe friction
coefficient
against all (μ) should
degrees of be definedexcept
freedom, to characterize the friction at
for the displacement behaviors
the end of the contact
supports in theinteraction
direction
of the axis of the bottom chord member and at the loading points in the direction of the applied
load. The joints other than the loading points and end supports were free to translate and rotate in
any directions.

4.6. Verification of Finite Element Models

The midspan vertical deflections under different load levels obtained from the finite element
analysis (FEA) of each specimen were compared with the corresponding test results, as shown in
Figure 16a–d for the specimens T-HW, T-HN, TS-AS, and TS-FC, respectively. It is shown in Figure 16
that the initial stiffness of all specimens and the change trends in the load–midspan vertical deflection
curves showed good agreement. Furthermore, the load carrying capacity of the test for specimen
T-HW, T-HN, and TS-AS are 110 kN, 90 kN, and 150 kN, respectively, whereas the load carrying
capacity of the FEA for the corresponding specimens are 116 kN, 94 kN, and 158 kN, respectively.
For specimen TS-FC, since it did not reach failure, the complete load–midspan vertical deflection curve
was not obtained and it was basically in the elastic stage. The load–midspan vertical deflection curve
is basically consistent with the corresponding FEA result in the elastic stage, as shown in Figure 16d.
Overall, the load carrying capacities of the FEA are slightly higher than those of the experimental
test, and the highest difference value between the test and FEA is 5.4%, which meets the requirements
of practical engineering. Furthermore, the axial strains of all specimens from the test and FEA were
compared in Figure 17, in which the vertical axis represents the axial strain (ε) of chord and brace
members at the end of the experimental loading. It can be seen from Figure 17 that the axial strains of
the FEA are slightly smaller than the corresponding test, but the difference values between them are
relatively small (the largest difference value is 6.1%). It is shown from the comparison that the FEA
results of specimens agreed well with the corresponding test results.
 meets the requirements of practical engineering. Furthermore, the axial strains of all specimens from
the test and FEA were compared in Figure 17, in which the vertical axis represents the axial strain
(ε) of chord and brace members at the end of the experimental loading. It can be seen from Figure
17 that the axial strains of the FEA are slightly smaller than the corresponding test, but the
difference values between them are relatively small (the largest difference value is 6.1%). It is shown
from
Appl. Sci. the8, 2237
2018, comparison that the FEA results of specimens agreed well with the corresponding test 16 of 22
results.

140 120

120 100
100
80
80

P/kN
P/kN

60
60
40
40

Test results 20 Test results

20
FEA results FEA results
0 0
0 1 2 3 4 5 6 7 8 0 2 4 6 8 10 12
w/mm w/mm
Appl. Sci. 2018, 8, 2237 16 of 21
(a) (b)
180 240

160 210
140 180
120
150
100
P/kN

P/kN
P/kN

80
P/kN 120

90
60

40 60
Test results Test results
20 30
FEA results FEA results
0 0
0 2 4 6 8 10 12 14 16 0 2 4 6 8 10 12 14
w/mm w/mm
(c) (d)

Figure
Figure 16. Comparison
16. Comparison of of experimentaland
experimental and finite
finite element
elementanalysis vertical
analysis load–midspan
vertical vertical
load–midspan vertical
deflection curves. (a) T-HW; (b) T-HN; (c) TS-AS; (d)
deflection curves. (a) T-HW; (b) T-HN; (c) TS-AS; (d) TS-FC. TS-FC.

0 0

-100 -100
Test resultes
Test resultes
FEA resultes
-200 FEA resultes
-200
-6
-6

-300
-6
-6

/10
ε/10
/10
ε/10

-300
-400
-400
-500

-500 -600

-600 -700
T-HW T-HN TS-AS TS-FC T-HW T-HN TS-AS TS-FC
Specimens Specimens
(a) (b)
0 1400

-200 Test resultes 1200

FEA resultes
-400 1000
-6

-600 800
-6

-6
-6

/10
/10

ε/10
ε/10

-800 600

-1000 400
Test resultes
-1200 200 FEA resultes

-1400 0
T-HW T-HN TS-AS TS-FC T-HW T-HN TS-AS TS-FC
Specimens Specimens
(c) (d)

Figure
Figure 17. Comparison
17. Comparison of of experimental and
experimental and finite
finite element
elementanalysis axial
analysis strains.
axial (a) Top
strains. chordchord
(a) Top ‘hj’; ‘hj’;
(b)Top chord ‘fh’; (c) Diagonal brace ‘jp’; (d) Diagonal brace
(b)Top chord ‘fh’; (c) Diagonal brace ‘jp’; (d) Diagonal brace ‘jo’. ‘jo’.

On the other hand, the failure modes obtained from the finite element analysis of specimen
T-HW were compared with the corresponding test results, as shown in Figures 18 and 19. It is shown
from the comparison that good agreement between the experimental and finite element analysis
results was achieved, and the main cause of the experiment and FEA failures were the surface
plasticity of the bottom chord member in addition to bending deformation along the truss span.
Therefore, the developed finite element models were verified to be accurate and reliable. In addition,
Appl. Sci. 2018, 8, 2237 17 of 22

On the other hand, the failure modes obtained from the finite element analysis of specimen T-HW
were compared with the corresponding test results, as shown in Figures 18 and 19. It is shown from
the comparison that good agreement between the experimental and finite element analysis results
was achieved, and the main cause of the experiment and FEA failures were the surface plasticity of
the bottom chord member in addition to bending deformation along the truss span. Therefore, the
developed finite element models were verified to be accurate and reliable. In addition, it is observed
from Figures 18 and 19 that the stress distribution law for specimen T-HW at the joint ‘k’ and joint
‘p’ are
Appl.
Appl. basically
Sci.
Sci. the same. The areas of chord surface at the brace–chord intersection became 17
2018,8,8,2237
2018, 2237 plastic,
17 of
of 21
21
because the maximum stresses shown in Figures 18 and 19 are 369 MPa and 358 MPa, respectively,
MPa and
MPa
which and 358 MPa,
358
exceed MPa,
the respectively,
respectively,
yield which
tress of 268which exceed the
MPa. exceed
However, thethe
yield
yield tress
tress
other of 268
of
areas 268 MPa.members
MPa.
of chord However,
However, the
dothe other
notother areas
yield,areas
and
of
of chord
chord members
members
are in the elastic state. do
do not
not yield,
yield, and
and are
are in
in the
the elastic
elastic state.
state.

(a)
(a) (b)
(b)
Figure18.
Figure 18.Comparison
18. Comparison
Comparisonofof experimental
ofexperimental and
experimentaland
andFEA
FEA
FEA failure
failure mode
mode
failure forfor
for
mode specimen
specimen T-HW
T-HW
specimen atthe
at
T-HW the joint
joint
at the ‘k’. ‘k’.
‘k’.
joint
(a)Experimental;
(a) Experimental;(b)
(b)FEA
FEAanalysis.
analysis.

(a)
(a) (b)
(b)
Figure19.
Figure 19.Comparison
19. Comparison
Comparisonof of experimental
ofexperimental and
experimentaland
andFEA
FEA
FEA failure mode
failure
failure mode
mode forfor
for specimen T-HW
specimen
specimen atthe
T-HW
T-HW at the joint
at the ‘p’.‘p’.
joint
joint ‘p’. (a)
(a)
(a) Experimental;
Experimental;
Experimental; (b)(b)
(b) FEA
FEA FEA analysis
analysis
analysis

5. Conclusions
5. Conclusions
Conclusions andand Future
and Future Work
Future Work
Work
5.
An experimental
An experimental investigation
experimental investigation and
investigation and corresponding
and corresponding numerical
corresponding numerical investigations
numerical investigations were
investigations were performed
were performed
performed inin
in
An
this study
this study on
study onfour
on four different
four different Warren
different Warren CHS
Warren CHS tubular
CHS tubular trusses, including
tubular trusses, a
trusses, includingtruss with
including aa truss its hidden
truss with
with its welds
its hidden welded
hidden welds
welds
this
(T-HW), a truss with
welded (T-HW),
(T-HW), its hidden itswelds unwelded (T-HN), a (T-HN),
truss with truss
its hidden welds unwelded
welded aa truss
truss with its
with hidden
hidden welds unwelded
welds unwelded (T-HN), aa truss with its
with its hidden
hidden welds
welds
but reinforced
unwelded by adding by
but reinforced
reinforced a half
addingouter sleeve on sleeve
each joint (TS-AS), and a truss
andwith its with
hidden
unwelded but by adding aa half
half outer
outer sleeve onon each
each joint (TS-AS),
joint (TS-AS), and aa truss
truss its
with its
welds
hidden unwelded
welds but reinforced
unwelded but by fillingby
reinforced concrete
filling into the top
concrete intoand
the bottom
top and chord
bottommembers
chord (TS-FC).
members
hidden welds unwelded but reinforced by filling concrete into the top and bottom chord members
Some conclusions
(TS-FC). can be drawn
Some conclusions
conclusions can from the experimental
be drawn
drawn from the and numerical
the experimental
experimental investigations
and as follows: as
numerical investigations
investigations
(TS-FC). Some can be from and numerical as
follows:
follows:
(1) The typical failure modes for three empty tubular trusses—T-HW, T-HN, and TS-AS—are basically
(1) the same,
(1) The
The which
typical
typical resulted
failure
failure modes
modes from for
for the surface
three
three plasticity
empty
empty of the
tubular
tubular bottom chord member,
trusses—T-HW,
trusses—T-HW, T-HN, and
T-HN, the
andweld fracture
TS-AS—are
TS-AS—are
around tubular
basically
basically the same,
the joints
same, at the
which
which bottom
resulted
resulted chord
from
from themember,
the and bending
surface plasticity
surface plasticity deformation
of the
of the bottom chord
bottom occurring
chord member,
member, at the
the
the
bottom chord.
weld fracture
weld It
fracture aroundshould
around tubularbe noted
tubular joints that
joints at the
at theconcave
the bottom deformation
bottom chord
chord member, at the
member, and end joints
and bending of TS-AS
bending deformation were
deformation
slighter than
occurring
occurring atthe
at thebottom
the concave
bottom deformation
chord.
chord. ItItshould
should ofbeT-HW
be noted
noted and T-HN,
that
that while its
the concave
the concave concave deformation
deformation
deformation atthe
at the end areas
end joints
joints
at
of the
TS-AS end joints
were were
slighter more
than extensive
the than
concave for T-HW
deformation and ofT-HN.
T-HW
of TS-AS were slighter than the concave deformation of T-HW and T-HN, while its concave and T-HN, while its concave
deformation areas
deformation areas at
at the
the end
end joints
joints werewere more
more extensive
extensive than
than for
for T-HW
T-HW andand T-HN.
T-HN.
(2) The
(2) The TS-FC
TS-FC specimen
specimen did did not
not fail
fail until
until reaching
reachingthe the maximum
maximumload load the
the test
test setup
setup could
could produce,
produce,
and there
and there was
was no no visible
visible change
change in in appearance.
appearance. Furthermore,
Furthermore, the the load
load carrying
carrying capacity
capacity andand
integral rigidity
integral rigidity of of specimen
specimen TS-FC TS-FC are are larger
larger than
than for
for specimen
specimen TS-AS,
TS-AS, which
which illustrated
illustrated that
that
the reinforcing
the reinforcing method
method of of filling
filling concrete
concrete intointo the
the top
top and
and bottom
bottom chord
chord members
members can can improve
improve
Appl. Sci. 2018, 8, 2237 18 of 22

(2) The TS-FC specimen did not fail until reaching the maximum load the test setup could produce,
and there was no visible change in appearance. Furthermore, the load carrying capacity and
integral rigidity of specimen TS-FC are larger than for specimen TS-AS, which illustrated that the
reinforcing method of filling concrete into the top and bottom chord members can improve the
mechanical behavior and integral stiffness of trusses better than adding a half outer sleeve on
each joint.
(3) Compared with the load carrying capacity of specimen T-HW, the load carrying capacity of
specimen T-HN decreased by 18%, and the load carrying capacity of specimen TS-AS increased
by 36%, whereas the load carrying capacity of specimen TS-FC increased by more than 60%.
The deflections of the specimens TS-AS, T-HN, and T-HW are in descending order, meaning that
specimen TS-AS has the best deformability, while specimen T-HN has better deformability than
specimen T-HW.
(4) The axial strains of the top chord members in compression are much smaller than the yield strain
of the steel material, whereas the axial strains of the bottom chord members in tension are smaller
than the yield strain of steel materials, and the axial strains of diagonal brace members are close
to the yield strain of the steel materials. In addition, the strains at the crown positions on joints of
the T-HW and T-HN specimens vary greatly, which means there is a case of stress concentration,
whereas the strains at the same position of specimens TS-AS and TS-FC vary uniformly. It is
shown that the reinforcing methods of adding a half outer sleeve on each joint and filling concrete
into the top and bottom chord members could help to improve the force condition of joints of
warren CHS tubular trusses.
(5) The midspan vertical deflections under different load levels, the failure modes, and the load
carrying capacity obtained from the finite element analysis of each specimen were compared with
the corresponding test results. Good agreement between the experimental and finite element
analysis results was achieved and the finite element models were accurate and reliable to be the
basis of parametric analysis.

With the recent development of piezoceramic transducer-based structural health monitoring and
damage detection in civil structures [70–74], especially in concrete-filled tubular structures [75–80],
as a future work, the authors will explore health monitoring and damage detection of Warren CHS
tubular trusses using easily installed piezoceramic patch transducers.

Author Contributions: W.Y. incepted the original idea and designed the experiment. N.G. analyzed the data.
N.-n.G., J.L., and R.Y. wrote the paper. W.Y. revised and finalized the paper.
Funding: This research was supported in part by the National Natural Science Foundation of China (No. 51468054)
and in part by the Project of Key Research and Development of Ningxia Province (No. 2016KJHM38), also in part
by the Funding Project of First-Class Discipline Construction of Universities in Ningxia (Domestic First-Class
Discipline Construction) under Grant NXYLXK2017A03.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Tiainen, T.; Mela, K.; Jokinen, T.; Heinisuo, M. The effect of steel grade on weight and cost of warren-type
welded tubular trusses. Proc. Inst. Civ. Eng. Struct. Build. 2017, 170, 855–873. [CrossRef]
2. Gillman, A.; Fuchi, K.; Buskohl, P.R. Truss-based nonlinear mechanical analysis for origami structures
exhibiting bifurcation and limit point instabilities. Int. J. Solids Struct. 2018, 147, 80–93. [CrossRef]
3. Jiang, T.; Wu, Q.; Wang, L.; Huo, L.; Song, G. Monitoring of Bolt Looseness-induced Damage in Steel Truss
Arch Structure using Piezoceramic Transducers. IEEE Sens. J. 2018, 18, 6677–6685. [CrossRef]
4. Song, G.; Vlattas, J.; Johnson, S.E.; Agrawal, B.N. Active vibration control of a space truss using a lead
zirconate titanate stack actuator. Proc. Inst. Mechan. Eng. Part G J. Aerosp. Eng. 2001, 215, 355–361. [CrossRef]
5. Wardenier, J.; Packer, J.A.; Zhao, X.L.; Vegte, G.J.V.D. Hollow Sections in Structural Applications; CIDECT:
Geneva, Switzerland, 2010.
Appl. Sci. 2018, 8, 2237 19 of 22

6. Yang, W.W.; Wang, X.L.; Wang, D.D. Design of long-span and wave-shaped steel canopy structure of ningxia
helan mountain stadium. Adv. Mater. Res. 2011, 243–249, 72–78. [CrossRef]
7. Kaveh, A.; Khayatazad, M. Ray optimization for size and shape optimization of truss structures.
Comput. Struct. 2013, 117, 82–94. [CrossRef]
8. Fenton, M.; McNally, C.; Byrne, J.; Hemberg, E.; McDermott, J.; O’Neill, M. Automatic innovative truss
design using grammatical evolution. Autom. Constr. 2014, 39, 59–69. [CrossRef]
9. Khademi, F. Load Rating of Railway Bridges by Analysis and Testing; Dissertations & Theses—Gradworks;
Illinois Institute of Technology: Chicago, IL, USA, 2015.
10. Khademi, F. Enhancing Load Rating of Railway Truss Bridges through a Hybrid Structural Analysis and
Instrumentation Procedure. Ph.D. Thesis, Illinois Institute of Technology, Chicago, ON, USA, 2017.
11. Wardenier, J.; Kurobane, Y.; Packer, J.A.; Vegte, G.J.V.D.; Zhao, X.L. Design Guide for CIRCULAR Hollow
Section (CHS) Joints under Predominantly Static Loading; CIDECT (Ed.) and Verlag TÜV Rheinland:
Cologne, Germany, 2008.
12. Zhao, X.; Liu, J.; Xu, X.; Sivakumaran, K.S.; Chen, Y. Hysteretic behaviour of overlapped tubular k-joints
under cyclic loading. J. Constr. Steel Res. 2018, 145, 397–413. [CrossRef]
13. Chinese Code. GB/50017-2017 Standard for Design of Steel Structures; Standards Press of China: Beijing,
China, 2017.
14. Wang, X.L.; Yang, W.W.; Zou, L. Finite Element Analysis for Unstiffened Overlapped CHS K-Joints Welded
in Different Ways. Adv. Mater. Res. 2011, 163–167, 299–306.
15. Wang, X.L.; Chen, P.; Yang, W.W. Finite Element Analysis for Unstiffened Overlapped CHS KK-Joints Welded
in Different Ways. Adv. Mater. Res. 2011, 446–449, 533–536. [CrossRef]
16. Yang, W.; Wang, X. Numerical analysis of hysteretic behavior of unstiffened overlapped CHS K-Joints in
steel pipe structures by finite element method. J. Harbin Eng. Univ. 2012, 12, 9.
17. Yang, W.; Wang, X. Experimental research on seismic behavior of unstiffened overlapped CHS K-Joints.
J. Build. Struct. 2013, 34, 85–92.
18. Xue, J.Q.; Briseghella, B.; Chen, B.C. Effects of debonding on circular CFST stub columns. J. Constr. Steel Res.
2012, 69, 64–76. [CrossRef]
19. Han, L.H.; Zheng, L.Q.; He, S.H.; Tao, Z. Tests on curved concrete filled steel tubular members subjected to
axial compression. J. Constr. Steel Res. 2011, 67, 965–976. [CrossRef]
20. Ding, F.X.; Luo, L.; Wang, L.; Cheng, S.; Yu, Z.W. Pseudo-static tests of terminal stirrup-confined
concrete-filled rectangular steel tubular columns. J. Constr. Steel Res. 2018, 144, 135–152. [CrossRef]
21. Qian, X.D.; Choo, Y.S.; Liew, J.Y.R.; Wardenier, J. Static Strength of Thick-Walled CHS X-Joints Subjected to
Brace Moment Loadings. J. Struct. Eng. 2007, 133, 1278–1287. [CrossRef]
22. Du, G.; Zhang, J.; Zhang, J.; Song, G. Experimental Study on Stress Monitoring of Sand-Filled Steel Tube
during Impact Using Piezoceramic Smart Aggregates. Sensors 2017, 17, 1930. [CrossRef] [PubMed]
23. Du, G.; Li, Z.; Song, G. A PVDF-Based Sensor for Internal Stress Monitoring of a Concrete-Filled Steel
Tubular (CFST) Column Subject to Impact Loads. Sensors 2018, 18, 1682. [CrossRef] [PubMed]
24. Huang, Q.; Xu, B.; Li, B.; Song, G.; Teng, J. Monitoring for Large Cross-Section CFSTs of a Super High-Rise
Building with Piezoceramic Actuators and Sensors. Adv. Mater. Res. 2011, 163–167, 2553–2559. [CrossRef]
25. Xu, B.; Zhang, T.; Song, G.; Gu, H. Active interface debonding detection of a concrete-filled steel tube with
piezoelectric technologies using wavelet packet analysis. Mech. Syst. Signal Process. 2013, 36, 7–17. [CrossRef]
26. Liang, Y.; Li, D.; Parvasi, S.M.; Kong, Q.; Lim, I.; Song, G. Bond-slip detection of concrete-encased composite
structure using electro-mechanical impedance technique. Smart Mater. Struct. 2016, 25, 095003. [CrossRef]
27. Feng, Q.; Kong, Q.; Tan, J.; Song, G. Grouting compactness monitoring of concrete-filled steel tube arch
bridge model using piezoceramic-based transducers. Smart Struct. Syst. 2017, 20, 175–180.
28. Zhang, J.; Li, Y.; Zheng, Y.; Wang, Z. Seismic Damage Investigation of Spatial Frames with Steel Beams
Connected to L-Shaped Concrete-Filled Steel Tubular (CFST) Columns. Appl. Sci. 2018, 8, 1713. [CrossRef]
29. Zhang, J.; Xu, J.; Guan, W.; Du, G. Damage Detection of Concrete-Filled Square Steel Tube (CFSST) Column
Joints under Cyclic Loading Using Piezoceramic Transducers. Sensors 2018, 18, 3266. [CrossRef] [PubMed]
30. Gho, W.M.; Yang, Y.; Gao, F. Failure mechanisms of tubular CHS joints with complete overlap of braces.
Thin-Walled Struct. 2006, 44, 655–666. [CrossRef]
Appl. Sci. 2018, 8, 2237 20 of 22

31. Choo, Y.S.; Qian, X.D.; Wardenier, J. Effects of boundary conditions and chord stresses on static strength of
thick-walled CHS K-Joints. J. Constr. Steel Res. 2006, 62, 316–328. [CrossRef]
32. Shao, Y.; He, S.; Zhang, H.; Wang, Q. Hysteretic behavior of tubular T-joints after exposure to elevated
temperature. Ocean Eng. 2017, 129, 57–67. [CrossRef]
33. Wang, W.; Chen, Y.-Y. Hysteretic behaviour of tubular joints under cyclic loading. J. Constr. Steel Res. 2007,
63, 1384–1395. [CrossRef]
34. Suo, Y.; Yang, W.; Chen, P. Study on Hysteresis Model of Welding Material in Unstiffened Welded Joints of
Steel Tubular Truss Structure. Appl. Sci. 2018, 8, 1701. [CrossRef]
35. Liu, K.; Liu, B.; Wang, Z.; Wang, G.; Guedes Soares, C. An experimental and numerical study on the
behaviour of tubular components and T-joints subjected to transverse impact loading. Int. J. Impact Eng.
2018, 120, 16–30. [CrossRef]
36. Qu, H.; Hu, Y.; Huo, J.; Liu, Y.; Jiang, Y. Experimental study on tubular K-joints under impact loadings.
J. Constr. Steel Res. 2015, 112, 22–29. [CrossRef]
37. Li, T.; Lie, S.T.; Shao, Y.B. Fatigue and fracture strength of a multi-planar circular hollow section TT-joint.
J. Constr. Steel Res. 2017, 129, 101–110. [CrossRef]
38. Lee, C.K.; Chiew, S.P.; Lie, S.T.; Sopha, T. Comparison of fatigue performances of gapped and partially
overlapped CHS -joints. Eng. Struct. 2011, 33, 44–52. [CrossRef]
39. Azari Dodaran, N.; Ahmadi, H.; Lotfollahi-Yaghin, M.A. Static strength of axially loaded tubular KT-joints
at elevated temperatures: Study of geometrical effects and parametric formulation. Mar. Struct. 2018, 61,
282–308. [CrossRef]
40. He, S.-B.; Shao, Y.-B.; Zhang, H.-Y.; Yang, D.-P.; Long, F.-L. Experimental study on circular hollow section
(CHS) tubular K-joints at elevated temperature. Eng. Fail. Anal. 2013, 34, 204–216. [CrossRef]
41. Lan, X.; Huang, Y.; Chan, T.-M.; Young, B. Static strength of stainless steel K- and N-joints at elevated
temperatures. Thin-Walled Struct. 2018, 122, 501–509. [CrossRef]
42. Packer, J.A. Concrete-Filled HSS Connections. J. Struct. Eng. 1995, 121, 458–467. [CrossRef]
43. Sakai, Y.; Hosaka, T.; Isoe, A.; Ichikawa, A.; Mitsuki, K. Experiments on concrete filled and reinforced tubular
K-joints of truss girder. J. Constr. Steel Res. 2004, 60, 683–699. [CrossRef]
44. Huang, W.; Fenu, L.; Chen, B.; Briseghella, B. Experimental study on K-joints of concrete-filled steel tubular
truss structures. J. Constr. Steel Res. 2015, 107, 182–193. [CrossRef]
45. Feng, R.; Young, B. Tests of concrete-filled stainless steel tubular T-joints. J. Constr. Steel Res. 2008, 64,
1283–1293. [CrossRef]
46. Feng, R.; Chen, Y.; Wei, J.; He, K.; Fu, L. Behaviour of grouted stainless steel tubular X-joints with CHS chord
under axial compression. Thin-Walled Struct. 2018, 124, 323–342. [CrossRef]
47. Yin, Y.; Han, Q.H.; Bai, L.J.; Yang, H.D.; Wang, S.P. Experimental Study on hysteretic behaviour of tubular
N-joints. J. Constr. Steel Res. 2009, 65, 326–334. [CrossRef]
48. 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, 268–283. [CrossRef]
49. Hou, C.; Han, L.-H.; Mu, T.-M. Behaviour of CFDST chord to CHS brace composite K-joints: Experiments.
J. Constr. Steel Res. 2017, 135, 97–109. [CrossRef]
50. Xu, F.; Chen, J.; Jin, W.L. Experimental investigation of SCF distribution for thin-walled concrete-filled CHS
joints under axial tension loading. Thin-Walled Struct. 2015, 93, 149–157. [CrossRef]
51. Qian, X.; Jitpairod, K.; Marshall, P.; Swaddiwudhipong, S.; Ou, Z.; Zhang, Y.; Pradana, M.R. Fatigue and
residual strength of concrete-filled tubular X-joints with full capacity welds. J. Constr. Steel Res. 2014, 100,
21–35. [CrossRef]
52. Tong, L.W.; Xu, G.W.; Yang, D.L.; Mashiri, F.R.; Zhao, X.L. Fatigue behavior and design of welded tubular
T-joints with CHS brace and concrete-filled chord. Thin-Walled Struct. 2017, 120, 180–190. [CrossRef]
53. Wu, Q.; Yoshimura, M.; Takahashi, K.; Nakamura, S.; Nakamura, T. Nonlinear seismic properties of the
Second Saikai Bridge: A concrete filled tubular (CFT) arch bridge. Eng. Struct. 2006, 28, 163–182. [CrossRef]
54. Chen, B.C.; Wang, T.L. Overview of Concrete Filled Steel Tube Arch Bridges in China. Pract. Period. Struct.
Des. Constr. 2009, 14, 70–80. [CrossRef]
Appl. Sci. 2018, 8, 2237 21 of 22

55. Han, L.H. Some Recent Developments of Concrete Filled Steel Tubular (CFST) Structures in China.
In Proceedings of the International Conference on Steel & Composite Structures, Sydney, Australia,
21–23 July 2010; pp. 41–52.
56. Xu, W.; Han, L.-H.; Tao, Z. Flexural behaviour of curved concrete filled steel tubular trusses. J. Constr.
Steel Res. 2014, 93, 119–134. [CrossRef]
57. Zhou, W.; Chen, Y.; Wang, K.; Han, S.; Palacios Galarza, F. Experimental research on circular concrete filled
stainless steel tubular truss. Thin-Walled Struct. 2017, 117, 224–238. [CrossRef]
58. Mujagic, J.R.U.; Easterling, W.S.; Murray, T.M. Design and behavior of light composite steel–concrete trusses
with drilled standoff screw shear connections. J. Constr. Steel Res. 2010, 66, 1483–1491. [CrossRef]
59. Tullini, N.; Minghini, F. Nonlinear analysis of composite beams with concrete-encased steel truss. J. Constr.
Steel Res. 2013, 91, 1–13. [CrossRef]
60. Huang, W.; Fenu, L.; Chen, B.; Briseghella, B. Experimental study on joint resistance and failure modes of
concrete filled steel tubular (CFST) truss girders. J. Constr. Steel Res. 2018, 141, 241–250. [CrossRef]
61. Chinese Code. Jgj81-2002 Technical Specification for Welding of Steel Structure of Building; Standards Press of
China: Beijing, China, 2003.
62. Chinese Code. Gb/T 228-2002 Metallic Materials-Tensile Testing at Ambient Temperature; Standards Press of
China: Beijing, China, 2002.
63. Chinese Code. Gb/T50081-2002 Test Method of Mechanical Properties on Ordinary Concrete; Standards Press of
China: Beijing, China, 2003.
64. ANSYS-Inc. ANSYS Structural Analysis Guide. 2004. Available online: http://www.ansys.stuba.sk/html/
guide_55/g-str/GSTRToc.htm (accessed on 1 November 2018).
65. Shao, Y.-B.; Li, T.; Lie Seng, T.; Chiew, S.-P. Hysteretic behaviour of square tubular T-joints with chord
reinforcement under axial cyclic loading. J. Constr. Steel Res. 2011, 67, 140–149. [CrossRef]
66. Chinese Code. Gb/50010-2010 Design of Concrete Structures; Standards Press of China: Beijing, China, 2010.
67. Vegte, G.J.V.D. The Static Strength of Uniplanar and Multiplanar Tubular T- and X-Joints; Delft University Press:
Delft, The Netherlands, 1995.
68. Corigliano, P.; Crupi, V.; Guglielmino, E.; Fricke, W. FE analysis of cruciform welded joints considering
different mechanical properties for base material, heat affected zone and weld metal. Frattura ed
Integrità Strutturale 2014, 8, 304–310. [CrossRef]
69. Corigliano, P.; Crupi, V.; Fricke, W.; Friedrich, N.; Guglielmino, E. Experimental and numerical analysis of
fillet-welded joints under low-cycle fatigue loading by means of full-field techniques. Proc. Inst. Mech. Eng.
Part C J. Mech. Eng. Sci. 2015, 229, 1327–1338. [CrossRef]
70. Zhao, J.; Bao, T.; Chen, S.; Kundu, T. Smart Aggregate-Piezoceramic Patch Combination for Health
Monitoring of Concrete Structures. J. Sens. 2016, 2016, 1–7. [CrossRef]
71. Kaur, N.; Bhalla, S. Combined Energy Harvesting and Structural Health Monitoring Potential of Embedded
Piezo-Concrete Vibration Sensors. J. Energy Eng. 2015, 141, D4014001. [CrossRef]
72. Kong, Q.; Robert, R.; Silva, P.; Mo, Y. Cyclic Crack Monitoring of a Reinforced Concrete Column under
Simulated Pseudo-Dynamic Loading Using Piezoceramic-Based Smart Aggregates. Appl. Sci. 2016, 6, 341.
[CrossRef]
73. Xu, K.; Deng, Q.; Cai, L.; Ho, S.; Song, G. Damage Detection of a Concrete Column Subject to Blast Loads
Using Embedded Piezoceramic Transducers. Sensors 2018, 18, 1377. [CrossRef] [PubMed]
74. Du, G.; Kong, Q.; Zhou, H.; Gu, H. Multiple Cracks Detection in Pipeline Using Damage Index Matrix Based
on Piezoceramic Transducer-Enabled Stress Wave Propagation. Sensors 2017, 17, 1812. [CrossRef] [PubMed]
75. Xu, B.; Li, B.; Song, G. Active Debonding Detection for Large Rectangular CFSTs Based on Wavelet Packet
Energy Spectrum with Piezoceramics. J. Struct. Eng. 2013, 139, 1435–1443. [CrossRef]
76. Shi, Y.; Luo, M.; Li, W.; Song, G. Grout compactness monitoring of concrete-filled fiber-reinforced polymer
tube using electromechanical impedance. Smart Mater. Struct. 2018, 27, 055008. [CrossRef]
77. Yan, S.; Fu, J.; Sun, W.; Qi, B.; Liu, F. PZT-Based Detection of Compactness of Concrete in Concrete Filled
Steel Tube Using Time Reversal Method. Math. Probl. Eng. 2014, 2014, 1–11. [CrossRef]
78. Luo, M.; Li, W.; Hei, C.; Song, G. Concrete Infill Monitoring in Concrete-Filled FRP Tubes Using a PZT-Based
Ultrasonic Time-of-Flight Method. Sensors 2016, 16, 2083. [CrossRef] [PubMed]
Appl. Sci. 2018, 8, 2237 22 of 22

79. Zhang, J.; Li, Y.; Du, G.; Song, G. Damage Detection of L-Shaped Concrete Filled Steel Tube (L-CFST) Columns
under Cyclic Loading Using Embedded Piezoceramic Transducers. Sensors 2018, 18, 2171. [CrossRef]
[PubMed]
80. Xu, Y.; Luo, M.; Hei, C.; Song, G. Quantitative evaluation of compactness of concrete-filled fiber-reinforced
polymer tubes using piezoceramic transducers and time difference of arrival. Smart Mater. Struct. 2017,
27, 035023. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).