You are on page 1of 6

EUROSTEEL 2014, September 10-12, 2014, Naples, Italy

PLASTIC MECHANISM STUDY OF COLD-FORMED STAINLESS STEEL


TUBULAR JOINTS SUBJECTED TO CHORD FACE FAILURE
Ran Fenga, Ben Youngb
a

Hefei University of Technology, School of Civil Engineering, Hefei, Anhui, China


r.feng@hfut.edu.cn
b
The University of Hong Kong, Department of Civil Engineering, Hong Kong, China
young@hku.hk

INTRODUCTION
The yield line mechanism analysis has been widely used to investigate the structural behaviour of
carbon steel members [1-5], such as plate, channel section, hat section, square and rectangular
hollow sections (SHS and RHS). This analysis was also successfully extended to welded tubular
joints. The existing design formulae given in various design specifications for welded tubular joints
subjected to chord face failure were all derived based on the yield line mechanism analysis. It
should be noted that the previous yield line mechanism analyses for welded tubular joints subjected
to chord face failure were all conducted on carbon steel tubular structures. There is little research
being carried out on theoretical analysis of stainless steel tubular joints. It is worth noting that the
yield line mechanism analysis was ever performed by Zhou and Young [6] on web crippling of
cold-formed stainless steel SHS and RHS. However, the concentrated load was applied by means of
steel bearing plates rather than welded brace members. It should be noted that the mechanical
properties of stainless steel sections are clearly different from those of carbon steel sections.
Stainless steel sections have a rounded stress-strain curve with no yield plateau and low
proportional limit stress compared to carbon steel sections. Hence, theoretical analysis should be
performed on cold-formed stainless steel tubular joints for the development of design rules.
1

SUMMARY OF EXPERIMENTAL AND NUMERICAL INVESTIGATIONS

1.1 Experimental work


A series of test program was conducted on a wide range of cold-formed stainless steel tubular Tand X-joints fabricated from SHS and RHS brace and chord members. A total of 22 T-joints and 11
X-joints was tested by applying axial compression force to the brace members using a wide range
of brace to chord width ratio ( = b1/b0); brace to chord thickness ratio ( = t1/t0), and chord width
to thickness ratio (2 = b0/t0). Three main failure modes were observed from the tests of coldformed stainless steel tubular T- and X-joints, namely chord face failure, chord side wall failure and
local buckling failure of brace, as shown in Figs. 1(a), 1(b) and 1(c), respectively. The failure loads
(Nf) were determined based on the joint deformation criteria defined specific to the different failure
modes. The joint strengths, failure modes as well as load-deformation curves of all test specimens
are detailed in Feng and Young [7, 8] for cold-formed stainless steel tubular T- and X-joints.
1.2 Numerical study
The general purpose finite element program ABAQUS was used for the numerical modelling of
cold-formed stainless steel tubular T- and X-joints by Feng and Young [9]. The material and
geometric nonlinearities have been taken into consideration in the finite element models. The
modelling of materials and welds, contact interaction between the T-joint specimens and the
supporting plate, applied load and boundary conditions were all carefully modelled. An extensive
parametric study was carried out to evaluate the effects of the main geometric parameters on coldformed stainless steel tubular joints. A total of 122 T-joints and 20 X-joints of cold-formed stainless
steel SHS and RHS tubes was analyzed in the parametric study. The joint strengths, failure modes
as well as load-deformation curves of the specimens were obtained from the numerical investigation
as detailed in Feng and Young [9].

a)

a)

b)

b)

c)

c)

Fig. 1. a) Chord face failure; b) Chord side wall failure; c) Local buckling failure of brace

PLASTIC MECHANISM MODELS FOR CHORD FACE FAILURE

2.1 CIDECT model


A simple plastic mechanism model for RHS tubular T-, Y- and X-joints was first derived by Jubb
and Redwood [1] and later summarized in the CIDECT code, as shown in Fig. 2(a). Plastic hinges
at the chord flange (mainly at the brace and chord intersection region) were considered, and two
plastic hinges were assumed on the corners of the chord member only. The joint strength can be
determined from the following design equation proposed based on this model [1].
N CIDECT

8M p h1

2 1

1 b0

where M p

f y0

b1
b0

f y 0 t 02
4

is the plastic moment of chord flange per unit length,


is the brace to chord width ratio,
is the yield stress of the chord.

(1)

2.2 Kato model


In 1980, Kato and Nishiyama [2] proposed a similar yield line model as CIDECT model, which
considered the finite size of the fillet welds (w) and the corner radius in the chord member. The
schematic sketch of the Kato model is shown in Fig. 2(b), where the inner corner radius of chord
member is assumed to be equal to t0. The joint strength can be determined from the design equation
proposed based on this model [2] as follows:
N kato

8M p h1'

2 1

'
1 b0

'
where b1'

b0

(2)

b1 2 w
, h1' h1 2w .
b0 4 1.5 2 t0

2.3 Modified Kato model


In 1991, Zhao and Hancock [3] recommended that the plastic hinges generally formed at the top of
the chord webs adjacent to the corners rather than at the center of the corners. This is because the
corners have a higher yield stress than the adjacent plates due to the strain hardening of the material.
A modified Kato model was developed to move the plastic hinges from the center of the corners to
the top of the chord webs adjacent to the corners as described in Zhao and Hancock [3]. The
schematic sketch of the modified Kato model is shown in Fig. 2(c). The joint strength can be
determined from the design equation proposed based on this model [3] as follows:
N zhao1

8M p h1'

2 1

1 b0

(3)

'
where b1 b1 2w , h1' h1 2w .

b0

b0

2.4 Davies and Packer model


Another similar plastic mechanism model was proposed by Davies and Packer [4], which assumed
that the plastic hinges were located around the edge of the weld as well as at the chord corners
along centre line of the tube. The schematic sketch of this yield line model is shown in Fig. 2(d).
The identical design equation as the Kato model [2] is given as follows:
N DP

8M p h1'

2 1

'
1 b0

(4)

'
where b1' b1 2w , h1' h1 2w .

b0 t 0

b0

b0

b0
b 0'
b 1'

b1
Fillet
weld

t0

b1

t1

Fillet
w eld

t0

Plastic hinges

t0

t1
w

Plastic hinges

t0

h0

h0

b)

a)
b0

b0
b 0'

b '1

b '1

b1
Fillet
weld

t0
t0

b1

t1

Fillet
weld

t0

Plastic hinges
t0

h0

c)

t1
w

Plastic hinges

h0

d)

Fig. 2. a) CIDECT model; b) Kato model; c) Modified Kato model; d) Davies and Packer model

2.5 Membrane mechanism model for welded tubular T-joint


The aforementioned plastic mechanism models did not consider the plastic hinges at the chord webs,
which were normally observed during the tests in the ultimate limit state. Therefore, the post-yield
response of welded tubular joint cannot be properly simulated. In 1991, a new plastic mechanism
model was developed by Zhao and Hancock [5] for RHS tubular T-joint under concentrated force.
This model considered the plastic hinges at the chord webs, the membrane force in the chord flange
and the strain hardening of the material in the analysis, as shown in Fig. 3. This model can be used
to predict the yield load, the post-yield response and the ultimate load of RHS tubular T-joint under
concentrated force. The joint strength can be determined from the design equation proposed based
on this model [5] as follows:
N zhao 2 Pm

k P
i

i 3, 5

(5)

b0 b1
1
,
t 0 f y 0 , sin y 1
2
1 y 2

where Pm 2S y sin y , S y

ki
Pi

is the total number of the ith plastic hinge,


is the partial yield load resulted from each ith plastic hinge.
b0
n

Chord
f

h1
e
t0

h0

3
1

4
1

b1

d,c

i,j

h0

8
7

o,p

h1

k,l
J

b0

a,b

6
6

m,n

b1

h,g

e,f
eh0

e,h

L0

f,g

a,d

b,c

h1

Fig. 3. Membrane mechanism model for welded tubular T-joint (Zhao and Hancock [5])

2.6 Membrane mechanism model for welded tubular X-joint (Proposed model)
Based on the membrane mechanism model [5] developed for RHS tubular T-joint, the joint strength
of RHS tubular X-joint can also be determined from Eq. (5) with a slight modification. For the RHS
tubular X-joint, the plastic hinges at the chord webs are formed at the mid-height of the webs due to
the symmetric loading and boundary conditions. Hence, the geometric parameter e, where e =
(n/h0), is equal to 0.5 for X-joint. The schematic sketch of the membrane mechanism model
proposed in this study for RHS tubular X-joint is shown in Fig. 4. The joint strength can be
determined from the design equation proposed based on this model as follows:
N zhao 2 2S y sin y

1 8
k i Pi
2 i 3, 5

(6)
b0
n

h1
e
t0

h0

3
1

b1

d,c

b1

1
3

a,b

eh0

b0

h,g

e,f

Chord
f

m,n

b1

o,p

h1

i,j

k,l
n

J
L0

e,h

f,g

a,d

h1

Fig. 4. Membrane mechanism model for welded tubular X-joint

b,c

h0

2.7 Comparison of experimental and numerical results with current design strengths
The joint strengths obtained from the experimental and numerical investigations of cold-formed
stainless steel tubular T- and X-joints subjected to chord face failure were compared with the design
strengths calculated using the existing design Eqs. (1)-(6), as shown in Table 1. The mean values of
failure load-to-design strength ratios Nf/NCIDECT, Nf/Nkato, Nf/Nzhao1, Nf/NDP and Nf/Nzhao2 are 0.92,
0.58, 0.68, 0.63 and 0.59, with the corresponding coefficients of variation (COV) of 0.306, 0.296,
0.264, 0.270 and 0.366. It can be generally concluded from the comparison that the CIDECT model
is somewhat unconservative, but gives high values of COV; all other yield line models are quite
unconservative; however, the modified Kato model, and Davies and Packer model give relatively
low values of COV.
Table 1.

Comparison of experimental and numerical results with design strengths of stainless steel tubular joints
Specimen
Comparison
(A total of 32 T-joints = b1/b0
Nf /NCIDECT Nf /Nkato Nf /Nzhao1 Nf /NDP Nf /Nzhao2 Nf /Np1 Nf /Np2
and 14 X-joints) [7-9]
Mean, Pm
0.92
0.58
0.68
0.63
0.59
1.00
1.00
COV, Vp
0.306
0.296
0.264
0.270
0.366
0.264 0.271
<0.80
0.65
0.65
0.65
0.65
0.65
0.65
0.65
Resistance factor,
Reliability index, 0
2.19
1.05
1.56
1.33
0.93
2.63
2.58

PROPOSED DESIGN FORMULAE FOR COLD-FORMED STAINLESS STEEL


TUBULAR JOINTS SUBJECTED TO CHORD FACE FAILURE

3.1 Proposed design formulae


Based on the experimental and numerical results of stainless steel tubular joints [7-9] as well as the
recommended range of brace to chord width ratio given in the previous literature for chord face
failure in yield line mechanism analysis for carbon steel tubular joints, it is therefore recommended
in this study that the chord face failure governs for stainless steel tubular joints with < 0.80. The
comparison of the experimental and numerical results with design strengths determined from
different yield line models are quite unconservative, as shown in Table 1. In this study, the
proposed design formulae for cold-formed stainless steel tubular T- and X-joints are based on the
modified Kato model [3], and Davies and Packer model [4] as shown in Eqs. (3) and (4). In these
two plastic mechanism models, both the rounded corners of SHS and RHS sections and the size of
the fillet welds were taken into account. The design formulae proposed in this study are as follows:
N p ANd
(7)
where A = 0.68
is the reduction factor for the modified Kato model,
A = 0.63
is the reduction factor for the Davies and Packer model,
8M p h1'

2 1
Nd

'
1 b0

where M p

f y 0 t 02
4

b1'
, b1' b1 2w , h1' h1 2 w
'
b0

b b0

is the effective chord width for the modified Kato model,

b0' b0 t 0
fy0

is the effective chord width for the Davies and Packer model,
is the yield stress (0.2% proof stress for stainless steel) of the chord.

'
0

(8)

3.2 Comparison of experimental and numerical results with proposed design strengths
The failure loads (Nf) obtained from the experimental and numerical investigations were compared
with the proposed design strengths (Np1 and Np2), which were calculated using Eq. (7) with the
reduction factor (A) of 0.68 and 0.63, respectively. Good agreements were achieved for the
comparison of stainless steel tubular T- and X-joints, with the mean values of failure load-toproposed design strength ratios (Nf /Np1 and Nf /Np2) of 1.00 and 1.00, with the corresponding COV
of 0.264 and 0.271, as shown in Table 1.

3.3 Reliability analysis


A reliability analysis was performed to assess the existing and proposed design rules for coldformed stainless steel tubular joints. The reliability of the design rules is measured by a reliability
index (0). In this study, a target reliability index of 2.5 was used. The design rules are considered
to be reliable if the reliability index is greater than 2.5. A resistance factor () of 0.65 was used to
calculate the reliability index (0) for the existing and proposed design rules. The statistical
parameters Pm and VP required for the reliability calculation are the mean value and COV of testedto-predicted load ratio, as shown in Table 1 for stainless steel tubular T- and X-joints subjected to
chord face failure. The reliability analysis is detailed in the Commentaries of the ASCE
Specification [10] and NAS Specification [11].
4

CONCLUSIONS

Plastic mechanism study of cold-formed stainless steel SHS and RHS tubular T- and X-joints
subjected to chord face failure have been presented in this paper. Different types of yield line
mechanism models for carbon steel tubular joints in the literature were used for stainless steel
tubular joints. The rounded corners of SHS and RHS sections as well as the size of the fillet welds
have been taken into account in the yield line mechanism analyses. The design formulae for
stainless steel tubular T- and X-joints subjected to chord face failure are proposed by modifying
from the existing design rules. The joint strength reduction factor (A) is recommended for the
chord face failure. It is shown that the experimental and numerical results compared well with the
design strengths calculated using the proposed design formulae. It is also clearly shown that the
design strengths determined from the proposed design formulae for stainless steel tubular T- and Xjoints subjected to chord face failure are generally more accurate and reliable than those calculated
using the existing design formulae. The proposed design formulae are capable of producing reliable
limit state designs when calibrated with the resistance factor of 0.65 for cold-formed stainless steel
tubular T- and X-joints subjected to chord face failure.
REFERENCES
[1] Jubb J.E.M., Redwood R.G., 1966. Design of joints to box sections, Conference on Industrial
Building and the Structural Engineer, Institution of Structural Engineers, UK.
[2] Kato B., Nishiyama I., 1980. T-joints made of rectangular tubes, Proceedings of the Fifth
International Specialty Conference on Cold-Formed Steel Structures, St. Louis, MO, USA, pp. 663-679.
[3] Zhao X.L., Hancock G.J., 1991. T-joints in rectangular hollow sections subject to combined actions.
Journal of Structural Engineering, ASCE, Vol. 117, No. 8, pp. 2258-2277.
[4] Davies G., Packer J.A., 1982. Predicting the strength of branch plate-RHS connections for punching
shear. Canadian Journal of Civil Engineering, Vol. 9, No. 3, pp. 458-467.
[5] Zhao X.L., Hancock G.J., 1991. Plastic mechanism analysis of T-joints in RHS under concentrated
force. Journal of the Singapore Structural Steel Society, Vol. 2, No. 1, pp. 31-44.
[6] Zhou F., Young B., 2006. Yield line mechanism analysis on web crippling of cold-formed stainless
steel tubular sections under two-flange loading. Engineering Structures, Vol. 28, No. 6, pp. 880-892.
[7] Feng R., Young B., 2008. Experimental investigation of cold-formed stainless steel tubular T-joints.
Thin-Walled Structures, Vol. 46, No. 10, pp. 1129-1142.
[8] Feng R., Young B., 2010. Tests and behaviour of cold-formed stainless steel tubular X-joints. ThinWalled Structures, Vol. 48, No. 12, pp. 921-934.
[9] Feng R., Young B., 2011. Design of cold-formed stainless steel tubular T- and X-joints. Journal of
Constructional Steel Research, Vol. 67, No. 3, pp. 421-436.
[10] American Society of Civil Engineers (ASCE). Commentary on Specification for the Design of ColdFormed Stainless Steel Structural Members. SEI/ASCE-8-02, Reston, USA, 2002.
[11] North American Specification (NAS). Commentary on North American Specification for the Design of
Cold-Formed Steel Structural Members, American Iron and Steel Institute, AISI S100-12-C, 2012
Edition, 2013.

You might also like