Journal of Constructional Steel Research 121 (2016) 330–340

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

Shear tests of rivet fastened rectangular hollow flange channel beams

Poologanathan Keerthan, Mahen Mahendran ⁎
Queensland University of Technology (QUT), Brisbane, Australia

a r t i c l e i n f o a b s t r a c t

Article history: The intermittently rivet fastened Rectangular Hollow Flange Channel Beam (RHFCB) is a new cold-formed hollow
Received 25 November 2015 section proposed as an alternative to welded hollow flange channel beams. It is a monosymmetric channel sec-
Received in revised form 19 February 2016 tion made by intermittently rivet fastening two torsionally rigid rectangular hollow flanges to a web plate. This
Accepted 23 February 2016
process enables the end users to choose an effective combination of different web and flange plate sizes to
Available online 11 March 2016
achieve optimum design capacities. Recent research studies focused mainly on the shear behaviour of the most
Keywords:
commonly used lipped channel beam and welded hollow flange beam sections. However, the shear behaviour
Cold-formed steel beams of rivet fastened RHFCB has not been investigated. Therefore a detailed experimental study involving 24 shear
Rivet fastened hollow flange channel beams tests was undertaken to investigate the shear behaviour and capacities of rivet fastened RHFCBs. Simply support-
Shear capacity ed test specimens of RHFCB with aspect ratios of 1.0 and 1.5 were loaded at mid-span until failure. Comparison of
Experiments experimental shear capacities with corresponding predictions from the current Australian cold-formed steel de-
Design equations sign rules showed that the current design rules are very conservative for the shear design of rivet fastened
Direct strength method RHFCBs. Significant improvements to web shear buckling occurred due to the presence of rectangular hollow
flanges while considerable post-buckling strength was also observed. Such enhancements to the shear behaviour
and capacity were achieved with a rivet spacing of 100 mm. Improved design rules were proposed for rivet fas-
tened RHFCBs based on the current shear design equations in AISI S100 and the direct strength method. This
paper presents the details of this experimental investigation and the results.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction are very efficient as structural beams since they have the hollow flanges
away from the centre. In the past, the LSB has been highly researched
The use of cold-formed steel members in low rise building construc- due to its ability to provide capacities that are more typically associated
tion has increased significantly in recent times. It has been suggested with hot-rolled, than cold-formed steel members ([3–6,8,9,11–13,21,
that in the future more than 70% of the steel buildings will be construct- 22]). However, the OATM discontinued LSB production in 2012, mostly
ed using cold-formed steel. The cold-formed steel manufacturers have due to the expensive manufacturing cost associated with the dual elec-
continuously utilised thin high strength steels and new manufacturing tric resistance welding process.
technologies to develop advanced light weight sections that are more The LSB sections are efficient and attractive steel products, which
structurally efficient and cost-effective in order to improve the market can be used in floor and roof systems as well as in modular building sys-
share for cold-formed steel construction. Cold-forming process is sim- tems. Although the LSBs are no longer produced, there is a need to find
ple, efficient, economical and environmentally friendly, which is capa- an alternative manufacturing method to produce equivalent sections
ble of manufacturing very effective sections compared to the hot- due to their popularity and demand among architects, engineers and
rolled open steel sections. The Rectangular Hollow Flange Channel builders in recent times. In this research study, an alternative
Beam (RHFCB) is one such advanced light weight section introduced re- manufacturing method has been proposed based on a combined cold-
cently by cold-formed steel manufacturers and researchers. forming and rivet fastening process to produce RHFCBs (Fig. 2) thus
In 2005 OneSteel Australian Tube Mills [16] introduced the first eliminating the costly dual electric resistance welding process.
RHFCB section, known as the LiteSteel beam (LSB) primarily for use as The rivet fastened RHFCB shown in Fig. 2 is a mono-symmetric sec-
flexural members in residential and light commercial/industrial appli- tion where the rectangular hollow flanges are cold-formed first and
cations (see Fig. 1). LSB is manufactured from a single strip of high then connected to the web plate using rivet fastening. Due to this simple
strength steel using a combined cold-forming and dual electric resis- and flexible manufacturing process, the designers can effectively choose
tance welding process. The LSBs combine the stability of hot-rolled sec- different plate dimensions and thicknesses for web and flange elements
tions with the high strength to weight ratio of cold-formed sections, and to achieve the most efficient section to suit the clients' requirements. As
an example, selecting a thicker web plate element is likely to eliminate
⁎ Corresponding author. or delay the unique lateral distortional buckling observed in hollow
E-mail address: m.mahendran@qut.edu.au (M. Mahendran). flange beams [4]. Unlike for LSBs with only three hollow flanges

http://dx.doi.org/10.1016/j.jcsr.2016.02.016
0143-974X/© 2016 Elsevier Ltd. All rights reserved.
 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340 331

compared with the predicted shear capacities using the current design
rules. Suitable design rules are then developed based on the current
shear design equations in AISI S100 [2] and the direct strength method
(DSM).

2. Shear capacities of lipped channel beams and hollow flange beams

In the past the shear buckling capacities of conventional lipped

channel beams (LCBs) are computed by neglecting the effect of flange
rigidity. LaBoube and Yu [15] investigated the shear strength of cold-
formed steel LCBs by considering different web slenderness ratio, the
edge support conditions provided by the flanges with varying flat
width-to-thickness ratios and the mechanical properties of the steel.
Based on their experimental studies a new set of design rules was pro-
posed for the shear strength of cold-formed steel beams. These design
rules were adopted in AISI S100 [1] and AS/NZS 4600 [24]. However,
they were developed by assuming simply supported web elements,
thus the rigidity offered by the flange elements to the web was ignored
and the post-buckling shear strength was also not considered. The shear
Fig. 1. LiteSteel beam [16]. strength design capacity (Vv) equations are provided next.
sffiffiffiffiffiffiffiffi
(45 × 15 mm, 60 × 20 mm and 75 × 25 mm) the rivet fastened RHFCBs d1 Ekv
Vv ¼ Vy ¼ 0:6 f yw d1 tw for ≤ ð1Þ
can be produced with many different hollow flange sizes. In addition to tw f yw
these features, they have additional flange lips that are used for rivet
fastening, which may enhance their design capacities. However, limited sffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffi
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Ekv
investigation has been carried out on the structural behaviour and Ekv d1
Vv ¼ Vi ¼ 0:6t2w Ekv f yw for b ≤1:508 ð2Þ
strength of rivet fastened RHFCBs. f yw tw f yw
For the newly developed rivet fastened RHFCB sections to be used as
sffiffiffiffiffiffiffiffi
flexural members, their flexural and shear capacities need to be fully in-
kv π2 Et3w d1 Ekv
vestigated. In this research the shear behaviour and capacity of rivet fas- Vv ¼ Vcr ¼ for N1:508 ð3Þ
12ð1−ν2 Þd1 tw f yw
tened RHFCB were investigated using experimental studies.
The web shear buckling design is based on the assumption that the
web elements are simply supported, in which the effect of flange rigid- where Vy = shear yield capacity, Vi = inelastic shear buckling capacity,
ity is neglected in conventional lipped channel beams (LCBs). LaBoube Vcr = elastic shear buckling capacity, d1 = depth of the flat portion of
and Yu [15] calculated the ultimate strengths of LCBs by assuming that web measured along the plane of the web, tw = web thickness, fyw =
the web-flange joints are simply supported. However, Keerthan and web yield stress, E = Young's modulus, υ = Poisson's ratio and kv is
Mahendran [7,8] found that there is considerable fixity at the web to the elastic shear buckling coefficient, which is determined as follows.
flange juncture, in particular for welded hollow flange channel beams. For beams with transverse stiffeners
They also showed the presence of significant post-buckling shear
strength for welded hollow flange channel beams. However, intermit- 4 a
kv ¼ 5:34 þ 2
for ≥1 ð4Þ
tent rivet fastening instead of continuous welding of flanges to web el- ða=d1 Þ d1
ements may reduce these beneficial effects on the shear capacity. Hence
this research investigated the shear capacities of rivet fastened RHFCBs.
5:34 a
This paper presents the details of a series of primarily shear tests of rivet kv ¼ 4 þ for b1 ð5Þ
2 d1
fastened RHFCBs, and the results. Experimental shear capacities are ða=d1 Þ

Fig. 2. Rivet fastened rectangular hollow flange channel beam (RHFCB).

332 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340

where ‘a’ = shear panel length and (a/d1) = aspect ratio. kv is taken as might not have reached their full shear capacities due to premature fail-
5.34 for unstiffened webs. ure (combined shear and web crippling failure). Pham and Hancock also
Pham and Hancock [17] performed numerical analyses using an reported that LaBoube and Yu's [15] test specimens failed prematurely
isoparametric spline finite strip method to study the elastic buckling be- due to combined shear and web crippling.
haviour of unlipped and lipped channel sections subject to shear. The Keerthan and Mahendran [9] found that simply supported condi-
numerical analyses showed that the flanges improved the shear buck- tions are sufficient to develop post-buckling strength in web elements
ling capacity of thin-walled channel sections. Further, it was shown due to tension field action. They also found that web elements need to
that when there is inadequate lateral restraint for sections with narrow be fully supported at the ends to develop post-buckling strength in
flanges, they can buckle prematurely in twisting and lateral buckling shear. Therefore a similar test set-up used in Keerthan and Mahendran
mode. Keerthan and Mahendran [10] performed finite element analyses [7] was considered in the shear tests of rivet fastened RHFCBs.
of LCB sections and proposed simplified design equations for the shear
buckling coefficients, which include the effect of web-flange rigidity.
Pham and Hancock [18] conducted both experimental and numeri- 3. Shear tests of rivet fastened RHFCBs
cal studies to investigate the shear behaviour of high strength cold-
formed steel LCB sections. Suitable design equations for the shear capac- 3.1. Test specimens and test set-up
ity of LCBs were then proposed in Pham and Hancock [18]. These shear
design equations have been adopted in 2012 AISI S100 [2]. These equa- In this experimental study a series of 24 primarily shear tests of sim-
tions predict the shear capacities of LCBs which include their available ply supported rivet fastened RHFCBs subjected to a mid-span load was
post-buckling strength and the effect of additional fixity at the web- conducted (see Fig. 3 and Table 1). The RHFCB test sections were fabri-
flange juncture. cated by connecting two rectangular hollow flanges to a web plate with
Keerthan and Mahendran [14] investigated the shear behaviour of rivets spaced at 100 mm. Due to fabrication limitations at the University
high strength cold-formed steel LCBs and LSBs using experimental and Workshop, only one hollow flange of nominal dimensions, 51 mm
numerical studies. Based on their test and FEA results, they proposed flange width (bf), 16 mm flange depth (df) and 20 mm lip length (lf),
appropriate design equations for the shear capacities of LCBs and LSBs was used in all the tests with varying web height and web and flange
(Eqs. (6) and (7)). These design equations consider the effect of addi- thicknesses. The hollow flange width to depth ratio was approximately
tional fixity at the web-flange juncture as well as post-buckling three as for LSBs while a lip length of 20 mm was chosen to accommo-
strength. In these equations the DSM based nominal shear capacity date the required rivet.
(Vv) is proposed based on Vcr (elastic buckling capacity in shear) and Two rivet fastened RHFCB sections were bolted back to back using
Vy (shear yield capacity). three 30 mm thick T-shaped stiffeners between them and twelve
sffiffiffiffiffiffiffiffi 10 mm full height web side plates on both sides, which were located
d1 Ekv at the end supports and the loading point in order to eliminate any tor-
Vv ¼ Vy for ≤ ð6Þ
tw f yw sional loading of test beams and possible web crippling and flange bear-
ing failures. A 30 mm gap was included between the two rivet fastened
  n  n sffiffiffiffiffiffiffiffi RHFCB sections to allow the test beams to behave independently while
Vcr Vcr d1 Ekv
VV ¼ 1−0:15 Vy for N ð7Þ remaining together to resist torsional effects. There is a possibility of
Vy Vy tw f yw flange distortion failure during the shear tests of rivet fastened RHFCBs.
This failure mode was eliminated by restraining the flanges by using
where, n = 0.55 for LCBs and 0.50 for LSBs, Vy and Vcr are given by four small angle straps as shown in Fig. 3. In order to simulate a primar-
Eqs. (1) and (3), and kv is the enhanced elastic shear buckling coefficient ily shear condition, relatively short test beams were selected based on
of LCB and LSB sections and the equations to predict them are given in suitable aspect ratios (shear span a/clear web height d1) of 1.0 and
Keerthan and Mahendran [8]. 1.5. Table 1 presents the details of the rivet fastened RHFCB test speci-
Keerthan and Mahendran [6] performed research on the elastic mens used in this experimental study. It includes the measured web
shear buckling behaviour of LiteSteel beams (LSBs) and proposed im- thicknesses (tw), clear web heights (d1) and yield stresses of the web el-
proved web shear buckling coefficient (kv) considering the effect of ements (fyw) of tested rivet fastened RHFCBs. As seen in Table 1, RHFCB
web-flange rigidity. Their results showed that the presence of rigid hol- section sizes are given as dxbfxtfxtw, which are shown in Fig. 2. Young's
low flanges increased the level of web to flange rigidity to 87% fixity modulus (E) and Poisson's ratio were taken as 200 GPa and 0.3,
level. In subsequent experimental and numerical studies by Keerthan respectively.
and Mahendran [7] they developed appropriate shear capacity (Vv) The rivet fastened RHFCB sections were loaded using the central T-
equations by including the available post-buckling strength and the in- shaped stiffener that was attached to the back to back test beams and
creased shear buckling coefficient (kv). They also developed suitable the two web side plates with 4 M16 bolts at the mid-span loading
DSM based design equations for the shear capacity of hollow flange point. The use of T-shaped stiffeners was important as they avoided
channel beams (LSBs). bearing failures of the flanges. Such a loading method also eliminated
Web side plates are required to simulate the ideal simply supported eccentric loading and web crippling. Similar T-shaped stiffeners were
end conditions during shear tests. Keerthan and Mahendran [7] investi- also located at the supports, and were bolted to the two rivet fastened
gated the effect of different sizes of the web side plates and having only RHFCBs and the two full web side plates on either side of the web ele-
one web side plate. They showed that the shorter web side plate heights ment. It is assumed that such end support conditions enable the panel
as well as one side plate were not adequate to provide the required sim- to be considered as fully supported at the end supports and the loading
ply supported conditions in the shear tests. Hence full web side plates point in all the tests. Fig. 3 shows the experimental set-up used in the
were provided on both sides in their shear test set-up to simulate shear tests of this research.
ideal simply supported conditions (not fixed conditions) for LiteSteel As seen in Fig. 3 the test beam was provided with pinned supports at
beams in shear. They were used at the supports and loading point each end. The test beam was supported on round sections. The applied
to provide the required simply supported conditions, while also elimi- load at the mid-span of the test beam and associated test beam dis-
nating any web crippling and flange bearing failures. However, in placements were measured during the test until failure. Two laser dis-
LaBoube and Yu's [15] test set-up, the LCB web elements appear to be placement transducers were located on the test beam under the
not fully supported at the ends due to the use of only one web side loading point and on the web panel to measure the vertical and lateral
plate with reduced number of bolt rows. Hence their tested LCBs deflections, respectively, as seen in Fig. 3.
 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340 333

Fig. 3. Shear test set-up.

3.2. Test procedure was then positioned in the test rig and the three point loading was
applied.
Two rivet fastened RHFCB were fabricated and their sizes, in partic- The clear height of web (d1) of RHFCB is considered in the design in-
ular, the clear web height (d1) and web thickness (tw), were measured stead of the depth of the flat portion of web measured along the plane of
as they are critical to the shear capacity calculations (Table 1). Test spec- the web as defined in AS/NZS 4600 [24] for cold-formed channel
imens were cut 50 mm longer than their required span in order to allow sections. This consideration is due to the following reasons.
25 mm overhang at each end of the test beam. Suitable bolt holes were
then inserted at both the loading and support points to allow for effec- ❖ RHFCB has additional lips, which are likely to increase the shear
tive connections at these points. capacity.
Shear span (a) was taken as the distance between the centres of ❖ RHFCB has two rectangular hollow flanges, which are likely to
inner bolts on the web side plates, and the required test specimen increase the shear yielding capacity by framing action.
length was calculated based on the aspect ratio. For example, in the ❖ The corners between web and hollow flanges have short lips on both
case of 200 × 51 × 1.15 × 1.15 RHFCBs with d1 = 170 mm, shear span sides unlike in cold-formed channel sections.
was 255 mm corresponding to an aspect ratio of 1.5. Hence the speci- ❖ Buckling occurs within the clear height of web.
men length was 695 mm based on the spacing of bolts in the web side ❖ Although the use of d1 as the depth of the flat portion of web is
plates of 45 mm and the edge distance of outer bolts of 25 mm. Two conservative in estimating the shear yielding capacity, it is not safe
rivet fastened RHFCBs were then assembled as back to back rivet fas- in the case of elastic buckling.
tened RHFCBs. The assembled pair of rivet fastened RHFCB sections

At the start of the tests, a small load was applied on the test speci-
mens to allow the entire test set-up to evenly settle on the bearings.
Table 1 The data acquisition system was then reset to zero before the actual
Details of rivet fastened RHFCB specimens.
loading was initiated. The loading rate of the testing machine's cross-
Test no. RHFCB section (dxbfxtfxtw) d1 (mm) tw (mm) fy (MPa) a/d1 head was kept at a constant rate of 0.7 mm/min until the failure of the
1 150 × 51 × 0.95 × 0.95 117 0.91 290 1.0 test specimens. Two laser displacement transducers were positioned
2 150 × 51 × 1.15 × 1.15 117 1.10 310 1.0 and connected to the data acquisition system to measure the vertical
3 200 × 51 × 0.95 × 0.95 166 0.91 290 1.0 and lateral deflections as shown in Fig. 3.
4 200 × 51 × 1.15 × 1.15 166 1.10 310 1.0
5 200 × 51 × 1.15 × 0.95 166 0.91 290 1.0
6 250 × 51 × 1.15 × 1.15 220 1.10 310 1.0
3.3. Test results
7 250 × 51 × 1.15 × 0.95 220 0.91 290 1.0
8 150 × 51 × 1.25 × 1.25 115 1.25 352 1.0 The full-scale shear tests were conducted to determine the actual ul-
9 200 × 51 × 1.25 × 1.25 170 1.25 352 1.0 timate shear capacities of rivet fastened RHFCB sections. The shear force
10 250 × 51 × 1.25 × 1.25 219 1.25 352 1.0
on each beam was calculated based on the applied load (P) on the test
11 150 × 51 × 0.95 × 0.95 117 0.91 290 1.5
12 150 × 51 × 1.15 × 0.95 117 0.91 290 1.5 specimens divided by four since each test set-up consisted of two
13 200 × 51 × 0.95 × 0.95 166 0.91 290 1.5 RHFCB sections. Table 2 shows the shear capacities of rivet fastened
14 200 × 51 × 1.15 × 0.95 166 0.91 290 1.5 RHFCB as obtained from this experimental study.
15 250 × 51 × 1.15 × 0.95 220 0.91 290 1.5 The load-deflection curve obtained from the shear test of
16 150 × 51 × 1.25 × 1.25 120 1.25 352 1.5
250 × 51 × 1.15 × 1.15 RHFCB (aspect ratio = 1.0) is shown in Fig. 4.
17 200 × 51 × 1.25 × 1.25 167 1.25 352 1.5
18 250 × 51 × 1.25 × 1.25 217 1.25 352 1.5 It can be seen that the web started to deflect out of plane at Point 1, in-
19 150 × 51 × 1.15 × 0.70 117 0.71 303 1.5 dicating the elastic shear buckling load to be about 13.1 kN (applied load
20 200 × 51 × 1.15 × 0.70 166 0.71 303 1.5 of 52.4 kN/4). At Point 2 the beam reached its ultimate shear capacity of
21 150 × 51 × 1.15 × 1.50 117 1.55 545 1.0
28.5 kN (applied load of 114 kN/4). These observations confirm that
22 150 × 51 × 1.15 × 1.90 117 1.91 480 1.0
23 150 × 51 × 1.15 × 1.15 117 1.10 310 1.5 the slender rivet fastened RHFCBs have significant post-buckling
24 250 × 51 × 1.15 × 1.15 220 1.10 310 1.5 capacity. Fig. 5 (a) and (b) shows the typical shear failure modes of
Note: d = overall height, bf = flange width (51 mm), df = flange depth (16 mm), lf = lip
200 × 51 × 1.15 × 0.95 RHFCB with an aspect ratio (AR) of 1.0 and
length (20 mm), tf = flange thickness, tw = web thickness, a = shear span, d1 = clear 250 × 51 × 1.15 × 0.95 RHFCB with an aspect ratio (AR) of 1.5, respec-
height of web, fy = web yield stress. tively while Fig. 5 (c) and (d) shows the shear failure modes of
334 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340

Table 2
Shear capacities of rivet fastened RHFCBs.

Test no. RHFCB section a/d1 tf/tw Shear capacity (kN) Test/AS/NZS
(dxbfxtfxtw) 4600
Test AS/NZS 4600

1 150 × 51 × 0.95 × 0.95 1.0 1.0 17.00 10.89 1.56

2 150 × 51 × 1.15 × 1.15 1.0 1.0 23.75 18.64 1.27
3 200 × 51 × 0.95 × 0.95 1.0 1.0 20.08 7.67 2.62
4 200 × 51 × 1.15 × 1.15 1.0 1.0 27.00 13.55 1.99
5 200 × 51 × 1.15 × 0.95 1.0 1.2 21.75 7.67 2.84
6 250 × 51 × 1.15 × 1.15 1.0 1.0 28.50 10.23 2.79
7 250 × 51 × 1.15 × 0.95 1.0 1.2 23.25 5.79 4.02
8 150 × 51 × 1.25 × 1.25 1.0 1.0 31.73 25.64 1.24
9 200 × 51 × 1.25 × 1.25 1.0 1.0 35.45 19.42 1.83
10 250 × 51 × 1.25 × 1.25 1.0 1.0 38.08 15.08 2.53
11 150 × 51 × 0.95 × 0.95 1.5 1.0 14.25 8.30 1.72
12 150 × 51 × 1.15 × 0.95 1.5 1.2 15.00 8.30 1.81
13 200 × 51 × 0.95 × 0.95 1.5 1.0 16.25 5.85 2.78
14 200 × 51 × 1.15 × 0.95 1.5 1.2 18.80 5.85 3.21
15 250 × 51 × 1.15 × 0.95 1.5 1.2 19.38 4.41 4.39
16 150 × 51 × 1.25 × 1.25 1.5 1.0 26.95 20.98 1.28
17 200 × 51 × 1.25 × 1.25 1.5 1.0 29.50 15.07 1.96
18 250 × 51 × 1.25 × 1.25 1.5 1.0 31.98 11.60 2.76
19 150 × 51 × 1.15 × 0.70 1.5 1.6 13.75 3.94 3.49
20 200 × 51 × 1.15 × 0.70 1.5 1.6 15.50 2.78 5.58
21 150 × 51 × 1.15 × 1.50 1.0 0.7 41.25 Combined Combined
22 150 × 51 × 1.15 × 1.90 1.0 0.6 35.00 Combined Combined
23 150 × 51 × 1.15 × 1.15 1.5 1.0 18.25 Combined Combined
24 250 × 51 × 1.15 × 1.15 1.5 1.0 24.00 Combined Combined
Mean 2.58
COV 0.441

Note: Test Specimens 21 to 24 failed by combined bending and shear actions.

250 × 51 × 1.15 × 0.95 RHFCB (AR = 1.0) and 200 × 51 × 0.95 × 0.95
RHFCB (AR = 1.0), respectively. These RHFCBs were made of a thinner
web element (0.95 mm).
Fig. 6 (a) to (c) shows the shear failure modes of three RHFCBs with
thicker web elements, 150 × 51 × 1.25 × 1.25 RHFCB
200 × 51 × 1.25 × 1.25 RHFCB and 250 × 51 × 1.25 × 1.25 RHFCB
(AR = 1.5), respectively. Fig. 7 (a) shows the plot for applied load versus
lateral deflection for 200 × 51 × 1.25 × 1.25 RHFCB with an aspect ratio
of 1.0, and confirms the presence of significant reserve capacity beyond
elastic buckling for RHFCBs. Keerthan and Mahendran [9] found that
slender welded LSB web panels do not collapse when elastic buckling
stress is reached and still have considerable post-buckling strength.
Fig. 7 (b) shows the mechanics of rivet fastened RHFCB web panel
post-buckling behaviour in shear. It shows that there is no separation
between the two flange lips and the web of RHFCB in the elastic buck-
ling and ultimate failure stages (see Fig. 7 (b)) with a small separation
in the post-ultimate failure stage.
All the RHFCB test specimens were rivet fastened at 100 mm spacing
as seen in Figs. 5 to 7. This spacing was chosen based on preliminary

Fig. 5. Shear failure modes of rivet fastened RHFCBs with thinner web elements.

shear and bending tests of RHFCBs [19,20]. The shear tests conducted
in this study did not show any unacceptable separation between the
two flange lips and the web of RHFCB test specimens (see Fig. 7).
There was noticeable separation of flange lips and web element only
after the ultimate load was reached with test specimens retaining the
torsional rigidity provided by the hollow flanges. The lack of continuity
along the web to flange juncture did not appear to cause any major
changes to the expected shear buckling and failure modes as seen in
Fig. 4. Plot of applied load versus lateral deflection. (250 × 51 × 1.15 × 1.15 RHFCB, aspect Figs. 5 to 7. Their shear failure modes were similar to those observed
ratio = 1.0). for welded RHFCBs (LSBs) and included tension field action ( [7] and
 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340 335

hollow flange channel beams (LSBs) is less than 0.65. However, test
specimens 21 to 24 failed in combined bending and shear actions.
Thus, only the test specimens which failed under primarily shear action
were considered in the following sections. Fig. 8 (a) and (b) shows the
combined bending and shear failure modes of 150 × 51 × 1.15 × 1.50
RHFCB with an aspect ratio of 1.0 and 150 × 51 × 1.15 × 1.15 RHFCB
with an aspect ratio of 1.5, respectively. Test Specimens 21 to 24 failed
primarily due to bending with only minor evidence of shear failure.
Local buckling/yielding of the compression flange was much more pro-
nounced in these specimens, which failed primarily by bending at mid-
span.
Test specimens with aspect ratio of 1.0 generally failed in primarily
shear. However Test specimens 21 and 22 (AR = 1.0) failed in the com-
bined bending and shear failure mode. In Test specimens 21 and 22, the
web thicknesses and yield strengths were 1.5 mm and 545 MPa, and
1.9 mm and 480 MPa, i.e. higher than the flange thickness and yield
strength of 1.1 mm and 310 MPa. Hence these test specimens had a
lower section moment capacity in relation to their shear capacity, and
thus failed in the combined bending and shear failure mode.
The shear capacities from the experimental study are compared with
the predictions based on the current cold-formed steel design standard
AS/NZS 4600 [24] (Table 2). The mean value of the ratios of shear capac-
ities from test results and AS/NZS 4600 design rules is 2.58 with a coef-
ficient of variation of 0.441. This comparison clearly shows that the
shear capacities are significantly under-estimated by the current AS/
NZS 4600 design rules. This is because they do not include the post-
buckling strength observed in the shear tests and the effect of increased
fixity at the web to flange juncture of rivet fastened RHFCBs despite the
lack of continuous fastening/welding. Hence improved shear capacity
equations were proposed in the following sections based on the current
shear capacity design equations in AISI S100 [2] and experimental
results.

4. Proposed equations for the shear capacity of rivet fastened RHFCBs

Keerthan and Mahendran [7] developed new shear strength equa-

tions for welded LSBs based on the current shear strength design equa-
tions in AISI S100 [1], experimental and finite element analysis results.
Similar equations can be developed for rivet fastened RHFCBs subject
to primarily shear action. Eqs. (8) to (10) present the proposed shear
strength equations which include the available post-buckling strength
and the additional fixity at the web-flange juncture in the rivet fastened
RHFCBs tested in this study (100 mm rivet spacing).
Eqs. (9) and (10) include the experimentally observed post-buckling
strength in rivet fastened RHFCBs and the additional fixity at the web-
flange juncture. In these equations, a post-buckling coefficient of 0.45
is introduced to include the available post-buckling strength in rivet fas-
tened RHFCBs. They also include the effects of additional fixity at the
web-flange juncture of RHFCBs. This effect is included using an in-
Fig. 6. Shear failure modes of rivet fastened RHFCBs with thicker web elements. creased shear buckling coefficient (kRHFCB). For this purpose, finite ele-
(AR = 1.5).
ment elastic buckling analyses of rivet fasted RHFCBs were performed
with a 100 mm rivet spacing as used in the shear tests. Finite element
models of RHFCBs in shear were developed using the same principles
[8]). Shear test results also showed that there were no rivet fastener fail- described in Keerthan and Mahendran [9] for welded LSBs. In these
ures. However, the effect of rivet spacing should be further investigated models the rivets connecting the hollow flange lips to the web were
using numerical and experimental studies. not explicitly simulated. Instead they were simulated using perfect Tie
The shear tests were carried out on relatively short beams with as- MPCs as the test results showed that there were no rivet fastener fail-
pect ratios of 1.0 and 1.5. Although these test specimens were assumed ures, which makes all active degrees of freedom equal on both sides of
to be subjected to pure shear due to their shorter spans and lower as- the connection. Eq. (11) presents the proposed equation for kRHFCB
pect ratios, the beams were also subjected to a bending moment. How- based on the elastic shear buckling loads of RHFCBs obtained from finite
ever, it is safe to assume that this bending moment did not affect the element analyses. For the welded LSBs, the coefficient in Eq. (11) was
shear capacities computed from these tests because, the ratio of the ap- 0.87, indicating the benefit of web-flange continuity. The use of this co-
plied moment (M⁎) to the section moment capacity (ϕMs) is less than efficient as 0.80 is considered to provide safer predictions for rivet fas-
0.75 based on the design rules of AS 4100 [23] for combined bending tened RHFCBs. It also implies that the effect of rivet spacing on elastic
and shear actions. This was confirmed previously by Keerthan and shear buckling strength and thus on ultimate shear strength of RHFCBs
Mahendran [11,12], in which they found that this ratio for welded is small. However, further finite element analyses will be carried out to
336 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340

Fig. 7. Mechanics of RHFCB web panel post-buckling behaviour in shear. 200 × 51 × 1.25 × 1.25 RHFCB (AR = 1.0).

Fig. 8. Combined bending and shear failure modes of rivet fastened RHFCBs.
 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340 337

investigate the effect of rivet spacing on the shear buckling coefficient of in rivet fastened RHFCBs. Therefore Eqs. (8) to (10) can be used to
rivet fastened RHFCBs. predict the accurate shear capacity of rivet fastened RHFCBs by consid-
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ering post-buckling shear strength and the web–flange fixity using a
d1 EkRHFCB post-buckling coefficient of 0.45 and a higher buckling coefficient
Vv ¼ Vy for ≤ ðShear yieldingÞ ð8Þ
tw f yw (Eq. (11)), respectively.
The recently developed direct strength method (DSM) is superior to
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the traditional effective width method that has been adopted as an al-
  EkRHFCB d1
Vv ¼ Vi þ 0:45 Vy −Vi b ternative design procedure in AISI S100 [1] and AS/NZS 4600 [24].
f yw tw
ð9Þ Hence the proposed Eqs. (8) to (10) are reorganized in the DSM format
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
EkRHFCB and are given as Eqs. (14) to (16). In these equations, the required shear
≤1:508 ðInelastic shear bucklingÞ slenderness (λ) was calculated using Eq. (17). Table 3 compares the
f yw
shear test capacities with the corresponding shear capacities predicted
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi by Eqs. (14) to (16). Comparison of test capacities with predicted capac-
  d1 EkRHFCB
Vv ¼ Vcr þ 0:45 Vy −Vcr N 1:508 ðElastic shear bucklingÞ ities from Eqs. (14) to (16) shows a good agreement with a mean of 1.06
tw f yw and a COV of 0.087 unlike in the case of AS/NZS 4600 design equations
ð10Þ (see Table 2).

where, Vy (shear yield capacity), Vi (inelastic shear buckling capacity) 4.1. Direct strength method
and Vcr (Elastic shear buckling capacity) are given by Eqs. (1) to (3),
respectively.
For rivet fastened RHFCBs
Vv
¼1 for λ ≤0:815 ð14Þ
kRHFCB ¼ kss þ 0:80ðksf −kss Þ ð11Þ Vy
 
5:34 2:31 a Vv 0:815 0:815
ksf ¼ þ −3:44 þ 8:39ða=d1 Þ for b1 ð12Þ ¼ þ 0:45 1− for 0:815bλ ≤1:23 ð15Þ
ða=d1 Þ
2 ða=d1 Þ d1 Vy λ λ
 
5:61 1:99 a Vv 1 1
ksf ¼ 8:98 þ − for ≥1 ð13Þ ¼ 2 þ 0:45 1− 2 for λN1:23 ð16Þ
ða=d1 Þ
2
ða=d1 Þ
3 d1 Vy λ λ

where kss and ksf are the shear buckling coefficients of plates with sim- where
ple–simple and simple–fixed boundary conditions, and kss is given by s
ffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi  s
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Eqs. (4) and (5). Vy d1 f yw
λ¼ ¼ 0:815 : ð17Þ
The AS/NZS 4600 [24] design equations do not include the post- Vcr tw EkRHFCB
buckling strength in shear, thus the design shear stress in the web is
limited by the elastic buckling capacity. However, Keerthan and New design equations were also proposed for the shear strength of
Mahendran's [7] experimental studies have shown that post-buckling rivet fastened RHFCBs in a similar manner to those of the section mo-
shear strength is present in LSBs, which can be included in their design ment capacity of beams subject to local buckling (Eqs. (18) and (19))
equations. Experimental results discussed in Section 3 have also shown using our experimental results. These equations are based on the DSM
that significant reserve shear strength beyond elastic buckling is present format. As in hot-rolled I-sections, only two regions based on shear

Table 3
Comparison of shear capacities of rivet fastened RHFCBs with proposed equations.

Test no. RHFCB section a/d1 tf/tw Shear capacity (kN) Test/Eqs. (14) to (16) Test/Eqs. (18) and (19)
(dxbfxtfxtw)
Test Eqs. (14) to (16) Eqs. (18) and (19)

1 150 × 51 × 0.95 × 0.95 1.0 1.0 17.00 15.53 14.66 1.09 1.16
2 150 × 51 × 1.15 × 1.15 1.0 1.0 23.75 21.64 20.48 1.10 1.16
3 200 × 51 × 0.95 × 0.95 1.0 1.0 20.08 17.22 17.37 1.17 1.16
4 200 × 51 × 1.15 × 1.15 1.0 1.0 27.00 24.81 24.35 1.09 1.11
5 200 × 51 × 1.15 × 0.95 1.0 1.2 21.75 17.22 17.37 1.26 1.25
6 250 × 51 × 1.15 × 1.15 1.0 1.0 28.50 27.45 27.85 1.04 1.02
7 250 × 51 × 1.15 × 0.95 1.0 1.2 23.25 19.75 19.83 1.18 1.17
8 150 × 51 × 1.25 × 1.25 1.0 1.0 31.73 28.61 27.04 1.11 1.17
9 200 × 51 × 1.25 × 1.25 1.0 1.0 35.45 33.85 32.86 1.05 1.08
10 250 × 51 × 1.25 × 1.25 1.0 1.0 38.08 36.62 37.08 1.04 1.03
11 150 × 51 × 0.95 × 0.95 1.5 1.0 14.25 14.83 14.06 0.96 1.01
12 150 × 51 × 1.15 × 0.95 1.5 1.2 15.00 14.83 14.06 1.01 1.07
13 200 × 51 × 0.95 × 0.95 1.5 1.0 16.25 16.40 16.63 0.99 0.98
14 200 × 51 × 1.15 × 0.95 1.5 1.2 18.80 16.40 16.63 1.15 1.13
15 250 × 51 × 1.15 × 0.95 1.5 1.2 19.38 19.13 18.96 1.01 1.02
16 150 × 51 × 1.25 × 1.25 1.5 1.0 26.95 28.02 26.48 0.96 1.02
17 200 × 51 × 1.25 × 1.25 1.5 1.0 29.50 31.62 31.22 0.93 0.95
18 250 × 51 × 1.25 × 1.25 1.5 1.0 31.98 34.85 35.39 0.92 0.90
19⁎ 150 × 51 × 1.15 × 0.70 1.5 1.6 13.75 9.88 9.97 1.39 1.38
20⁎ 200 × 51 × 1.15 × 0.70 1.5 1.6 15.50 11.81 11.74 1.31 1.32
Mean 1.06 1.08
COV 0.087 0.086
⁎ Tests 19 and 20 were not considered in the design equations as the flange to web thickness (tf/tw) ratio exceeds the limit of proposed design equations (tf/tw b 1.2).
338 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340

yielding, and elastic and inelastic shear buckling, were considered. In

these equations, a power coefficient of 0.3 was used based on the exper-
imental results of rivet fastened RHFCBs from this study.

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
d1 EkRHFCB
Vv ¼ Vy for ≤0:86 ðShear yieldingÞ ð18Þ
tw f yw

"  0:30 # 0:30 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

Vcr Vcr d1 EkRHFCB
Vv ¼ 1−0:15 Vy : for N0:86
Vy Vy tw f yw ð19Þ
ðElastic and inelastic shear bucklingÞ

Comparison of test capacities with predicted capacities from the

DSM equations (Eqs. (18) and (19)) in Table 3 shows a good agreement
with a mean of 1.08 and a COV of 0.086 unlike in the case of AS/NZS
4600 design equations (see Table 2).
Experimental ultimate shear capacity results were also calculated Fig. 9. Comparison of experimental shear capacities of rivet fastened RHFCBs with DSM
based on the DSM format and compared with the proposed DSM based design equations.
based design equations. The required Vy, Vcr and λ values of test speci-
mens are provided in Table 4. Fig. 9 shows the non-dimensional shear design equations can be used in designs as their predictions are conser-
capacity curves for rivet fastened RHFCBs and the experimental results. vative for higher tf/tw ratios.
In this figure, the shear capacities are plotted in a non-dimensional for- Fig. 10 shows the non-dimensional shear capacity curves for rivet
mat of Vv/Vy versus λ, where λ is (Vy/Vcr)0.5. It can be seen that shear de- fastened RHFCBs and the experimental results. The ultimate shear ca-
sign equations without post-buckling strength (Eqs. (1) to (3)) are pacities of welded LiteSteel beams from experiments were also included
significantly underestimating the shear capacities of rivet fastened in Fig. 10. The level of fixity along the web to flange juncture was similar
RHFCBs. However, Eqs. (14) to (16) and (18) and (19) safely predict for both welded and rivet fastened beams based on the elastic buckling
the shear capacities while also including the available post-buckling finite element analyses. These analyses showed a small reduction in fix-
strength and additional fixity at the web–flange juncture reasonably ity for rivet fastened RHFCBs as reflected by its shear buckling coeffi-
well (see Fig. 9). cients kRHFCB (0.80 versus 0.87 in Eq. (11)). However, Fig. 10 shows
In order to investigate the effect of flange thickness to web thickness that the shear capacities of rivet fastened RHFCB are slightly higher
(tf/tw) ratio on the ultimate shear capacities of rivet fastened RHFCBs, it than the shear capacities of LSBs, which is likely to be due to the contri-
was varied from 1.0 to 1.6 in the shear tests as seen in Table 2. However, bution from the additional lips in rivet fastened RHFCBs.
when it was increased from 1.2 to 1.6 in Tests 19 and 20, the shear ca- When the flange thickness to web thickness (tf/tw) ratio was in-
pacities increased considerably more than in other tests with smaller creased from 1.2 to 1.6 in Tests 19 and 20, the shear capacities increased
tf/tw ratios. Hence the design equations proposed in this section were considerably more than in other tests with smaller tf/tw ratios (more
limited to tf/tw ratio of 1.2, i.e., Tests 19 and 20 were not considered in than 30% increment in shear capacity). One of the main contributors
developing the design equations (Eqs. (14) to (16), (18) and (19)). could be the two additional thicker lips. In the rivet fastened RHFCBs,
The proposed design rules are conservative for Tests 19 and 20 as seen shear buckling is likely to occur within the top and bottom rivet levels.
in the comparisons in Tables 3 and 4. Detailed numerical studies will Therefore the effective clear height (deff) will be less than clear height
be conducted to investigate the effect of tf/tw ratio on the ultimate of web (d1). This can also lead to the increase in shear capacity of
shear capacities of rivet fastened RHFCBs. However, the proposed RHFCBs. Further detailed numerical studies will be conducted to extend
the available shear capacity data for rivet fastened RHFCBs and to con-
Table 4 firm the accuracy of the developed shear capacity equations.
Shear yielding and buckling capacities of rivet fastened RHFCBs. Fig. 11 shows the shear failure mechanism of rivet fastened RHFCBs.
Test no. RHFCB section Test Vv Vy Vcr VV/Vy λ
There is small separation between the two flange lips and the web of
(dxbfxtfxtw) (kN) (kN) (kN)

1 150 × 51 × 0.95 × 0.95 17.00 18.53 13.92 0.92 1.15

2 150 × 51 × 1.15 × 1.15 23.75 23.94 24.59 0.99 0.99
3 200 × 51 × 0.95 × 0.95 20.08 26.28 9.81 0.76 1.64
4 200 × 51× 1.15 × 1.15 27.00 33.96 17.33 0.79 1.40
5 200 × 51 × 1.15 × 0.95 21.75 26.28 9.81 0.83 1.64
6 250 × 51 × 1.15 × 1.15 28.50 45.01 13.07 0.63 1.86
7 250 × 51 × 1.15 × 0.95 23.25 34.83 7.40 0.67 2.17
8 150 × 51 × .25 × 1.25 31.73 30.36 36.70 1.04 0.91
9 200 × 51 × 1.25 × 1.25 35.45 44.88 24.83 0.79 1.34
10 250 × 51 × 1.25 × 1.25 38.08 57.76 19.29 0.66 1.73
11 150 × 51 × 0.95 × 0.95 14.25 18.53 11.80 0.77 1.25
12 150 × 51 × 1.15 × 0.95 15.00 18.53 11.80 0.81 1.25
13 200 × 51 × 0.95 × 0.95 16.25 26.28 8.32 0.62 1.78
14 200 × 51 × 1.15 × 0.95 18.80 26.28 8.32 0.72 1.78
15 250 × 51 × 1.15 × 0.95 19.38 34.83 6.27 0.56 2.36
16 150 × 51× 1.25 × 1.25 26.95 31.57 29.91 0.85 1.03
17 200 × 51 × 1.25 × 1.25 29.50 44.09 21.42 0.67 1.43
18 250 × 51× 1.25 × 1.25 31.98 57.39 16.46 0.56 1.87
19⁎ 150 × 51 × 1.15 × 0.70 13.75 15.10 5.60 0.91 1.64
20⁎ 200 × 51× 1.15 × 0.70 15.50 21.43 3.95 0.72 2.33
⁎ Tests 19 and 20 were not considered in the design equations as the flange to web Fig. 10. Comparison of experimental shear capacities of rivet fastened RHFCBs and LSBs
thickness (tf/tw) ratio exceeds the limit of proposed design equations (tf/tw b 1.2). with DSM based design equations.
 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340 339

Fig. 11. Shear failure mechanism of rivet fastened RHFCBs.

RHFCB test specimens in the diagonal compression region and no sepa- Palhanoand and Thiago Silva for their assistance in performing the
ration between the two flange lips and the web of RHFCB test specimens shear tests.
in the diagonal tension region as shown in Fig. 11 (Post-failure Stage).
Post-buckling strength of rivet fastened RHFCBs mainly depends on References
the tension field action (TFA). It will not be reduced significantly due
[1] American Iron and Steel Institute (AISI), North American Specification for the De-
to the small separation in the diagonal compression region in the sign of Cold-formed Steel Structural Members, AISI S100, Washington, DC, USA,
post-failure stage. 2007.
[2] American Iron and Steel Institute (AISI), North American Specification for the De-
sign of Cold-formed Steel Structural Members, AISI S100, Washington, DC, USA,
5. Conclusions 2012.
[3] T. Anapayan, M. Mahendran, D. Mahaarachchi, Section moment capacity tests of
LiteSteel beams, Thin-Walled Struct. 49 (2011) 502–512.
This paper has presented the details of an experimental investiga- [4] T. Anapayan, M. Mahendran, Improved design rules for hollow flange sections sub-
tion into the shear behaviour and capacity of rivet fastened rectangular ject to lateral distortional buckling, Thin-Walled Struct. 50 (2012) 128–140.
[5] T. Anapayan, M. Mahendran, Numerical modelling and design of LiteSteel beams
hollow flange channel beams (RHFCBs). The ultimate shear capacities
subject to lateral buckling, J. Constr. Steel Res. 70 (2012) 51–64.
measured from the tests showed that AS/NZS 4600 [24] design equa- [6] P. Keerthan, M. Mahendran, Shear behaviour of LiteSteel beams, Proceedings of the
tions are very conservative for the shear design of rivet fastened 5th International Conference on Thin-walled Structures, Brisbane, Australia 2008,
RHFCBs. Rivet fastened RHFCBs exhibited considerably enhanced web pp. 411–418.
[7] P. Keerthan, M. Mahendran, Experimental studies on the shear behaviour and
shear buckling and post-buckling capacity when subjected to shear. strength of LiteSteel beams, Eng. Struct. 32 (2010) 3235–3247.
New shear capacity equations were proposed based on the current [8] P. Keerthan, M. Mahendran, Elastic shear buckling characteristics of LiteSteel beams,
shear strength design equations in AISI S100 [2] and experimental re- J. Constr. Steel Res. 66 (2010) 1309–1319.
[9] P. Keerthan, M. Mahendran, New design rules for the shear strength of LiteSteel
sults to incorporate these enhancements in shear capacity. Suitable beams, J. Constr. Steel Res. 67 (2011) 1050–1063.
shear design rules were also developed under the direct strength meth- [10] P. Keerthan, M. Mahendran, Shear buckling characteristics of cold-formed steel
od format. This experimental study was limited to 100 mm rivet spacing channel beams, Int. J. Steel Struct. 13 (2013) 385–399.
[11] P. Keerthan, D. Hughes, M. Mahendran, Experimental studies of hollow flange chan-
and hence further research is currently under way to investigate the ef- nel beams subject to combined shear and bending actions, Thin-Walled Struct. 77
fect of different rivet spacings and flange to web thickness ratios on the (2014) 129–140.
shear capacities of RHFCBs. [12] P. Keerthan, M. Mahendran, D. Hughes, Numerical studies and design of hollow
flange channel beams subject to combined bending and shear actions, Eng. Struct.
75 (2014) 197–212.
[13] P. Keerthan, M. Mahendran, E. Steau, Experimental study of web crippling behaviour
Acknowledgements
of hollow flange channel beams under two flange load cases, Thin-Walled Struct. 85
(2014) 207–219.
The authors would like to thank Australian Research Council for [14] P. Keerthan, M. Mahendran, Experimental investigation and design of lipped chan-
their financial support (Grant LP120200650) and the Queensland Uni- nel beams in shear, Thin-Walled Struct. 86 (2015) 174–184.
[15] R.A. LaBoube, W.W. Yu, Cold-formed steel beam webs subjected primarily to shear,
versity of Technology for providing the necessary facilities and support Research Report, American Iron and Steel Institute, University of Missouri-Rolla,
to conduct this research project. They would also like to thank Diego Rolla, USA, 1978.
340 P. Keerthan, M. Mahendran / Journal of Constructional Steel Research 121 (2016) 330–340

[16] OneSteel Australian Tube Mills, (OATM), Design of LiteSteel Beams, Brisbane, Conference on Cold-formed Steel Structures, St Louis, Missouri, USA 2014,
Australia, 2008. pp. 259–276.
[17] C.H. Pham, G.J. Hancock, Shear buckling of thin-walled channel sections, J. Constr. [21] J.K. Seo, M. Mahendran, Plastic bending behaviour and section moment capacities of
Steel Res. 65 (2009) 578–585. mono-symmetric LiteSteel beams with web openings, Thin-Walled Struct. 49
[18] C.H. Pham, G.J. Hancock, Direct strength design of cold-formed C-sections for shear (2011) 513–522.
and combined actions, J. Struct. Eng. ASCE 138 (2012) 759–768. [22] J.K. Seo, M. Mahendran, Member moment capacities of mono-symmetric LiteSteel
[19] R. Siahaan, P. Keerthan, M. Mahendran, Section moment capacity tests of rivet fas- beam floor joists with web openings, J. Constr. Steel Res. 70 (2012) 153–166.
tened rectangular hollow flange channel beams, Proc. of 22nd International Special- [23] Standards Australia, Australian Standard AS 4100 Steel Structures, Sydney, Australia,
ty Conference on Cold-formed Steel Structures, St Louis, Missouri, USA 2014, 1998.
pp. 277–294. [24] Standards Australia/Standards New Zealand, Australia/New Zealand Standard
[20] R. Siahaan, P. Keerthan, M. Maendran, Numerical studies of rivet-fastened rect- AS/NZS4600 Cold-formed Steel Structures, Sydney, Australia, 2005.
angular hollow flange channel beams, Proc. of 22nd International Specialty