Journal of Constructional Steel Research 71 (2012) 210–218

Contents lists available at SciVerse ScienceDirect

Journal of Constructional Steel Research

Cold-formed steel flexural member with edge stiffened

holes: Behavior, optimization, and design
Cheng Yu ⁎
Department of Engineering Technology, University of North Texas, Denton, TX, USA

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

Article history: Circular holes are commonly found on the web of cold-formed steel (CFS) flexural members for piping, elec-
Received 8 April 2011 tric-wiring, plumbing, or installing lateral bracing, etc. Traditional holes on CFS members are flat bunched
Accepted 5 September 2011 without edge lips. A new generation of CFS C-section flexural members with edge stiffened holes was recent-
Available online 2 October 2011
ly developed by the industry. However, research on the new generation C-section members is underdevel-
oped and available test results are limited. This paper presents finite element analyses to study on the
Keywords:
Cold-formed steel
stability of cold-formed steel thin plates and typical C-section members when edge stiffened circular holes
Flexural member are placed on those plates or members. Based on the elastic buckling analyses, the optimized profiles of
Finite element analysis the holes are obtained and then applied to standard C-section flexural members. The post-buckling finite el-
Web opening ement analysis is utilized to determine the flexural strength of those members. The results indicate that the
stiffened holes can significantly improve the flexural strength of CFS C-sections. New design provisions are
proposed to accurately predict the flexural strength of the new generation C-section flexural members
with the optimized hole profiles.
© 2011 Elsevier Ltd. All rights reserved.

1. Introduction
Holes can be seen in cold-formed steel (CFS) flexural members
such as joists/beams, studs for piping, electric-wiring, plumbing,
bracing, etc. In particular, large openings on the web of floor joists
are preferred by the contractors as the mechanical and electrical
work becomes more and more complex in today's building industry.
Traditional holes in the CFS members are flat bunched. The sizes of
the holes and the distance between holes are greatly restricted due to
the weakened flexural strength. To overcome those restrictions, a
new generation of profile for the web holes was developed by the in-
dustry. Fig. 1 shows one example of a new generation floor joists, the
web holes are stiffened by a continuous edge lip around the perimeter
of the hole. The new design allows oversize opening to be formed on
the web and at the same time the joists still provide satisfied flexural
strength.
The presence of the edge stiffener/lip around the holes leads to
changes in the boundary conditions of the web element as well as
the resultant stress distribution. Thus the elastic and post-buckling
behaviors of the whole member will also be potentially altered. In
the traditional C-sections, the area where the flat holes are located
is usually the weakest zone because the opening splits one stiffened
⁎ Corresponding author. Tel.: + 1 940 891 6891; fax: + 1 940 565 2666.
E-mail address: cyu@unt.edu.
element (the full web) into two unstiffened elements (one edge stiff-
ened and one edge free), as shown in Fig. 2(a). The classical solutions
from thin plate theory (e.g. Timoshenko and Gere 1961 [1]) indicate
that the stiffened element yields more than 9 time higher elastic
buckling stress than the unstiffened element (buckling coefficient
4.0 vs. 0.43) when the element is subject to compression. For the
new generation C-section as shown in Fig. 2(b), the edge stiffener
acts as edge supporter to the two flat portions of the web therefore
the opening splits the whole web element to two stiffened elements.
Compared to the traditional C-sections having two unstiffened ele-
ments at holes, the new generation C-sections is expected to generate
highly improved performance in both flexural and compression
strength. This paper focuses on the flexural strength of CFS C-sections
with stiffened circular holes.
The current North American Specification for the Design of Cold-
Formed Steel Structural Members (AISI S100 2007 [2]) does not pro-
vide design provisions for the C-sections with edge stiffened holes.
AISI S100 has specific provisions for C-section webs with flat holes
under stress gradient (Section B2.4, AISI S100 2007 [2]), the design
provisions were developed from 57 simple beam tests conducted at
University of Missouri-Rolla (Shan, LaBoube and Yu 1994 [3]), and
the specimens were C-sections beams with standard flat holes. How-
ever the new generation of C-section joists has shown significantly
improved performance due to the edge stiffened holes on the web
(NAHB 1999 [4]) and the current specification is no longer applicable
for those members. On the other hand, because the new products

0143-974X/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.jcsr.2011.09.008
 C. Yu / Journal of Constructional Steel Research 71 (2012) 210–218
Fig. 1. Floor joist with edge stiffened holes.
were newly developed, the industry has not established standards for
the geometric configurations. Therefore the research objectives of this
work are (1) to the study of behavior of the CFS plates and C-section
joists with edge stiffened circular holes, (2) to establish recommenda-
tions for optimized profiles for the holes for the new generation joists,
and (3) to develop design equations for the flexural strength of the
new generation C-sections with optimized circular hole profiles. Fi-
nite element (FE) analysis is employed for the research and the
study focuses on Steel Stud Manufacturers Association (SSMA) stan-
dard C-section joists.
2. Ealstic buckling of cold-formed steel thin plates with circular
holes
The plate stability normally determines the strength of build-up
members such as CFS C-section when those members fail by local
buckling. In the actual applications, the floor joists are usually lateral-
ly supported by bridging and the compression flange of floor joists is
commonly restrained by the decking. Therefore the lateral–torsional
buckling and distortional buckling failure are generally restricted in
practice, and the local buckling becomes the primary failure mode
for floor joists. Thus, the first step of the research is to examine the
stability of thin plates with stiffened or unstiffened holes.
2.1. Thin plates subject to uniform compression
The FE method using ABAQUS (2003 [5]) is employed to deter-
mine the elastic buckling load of CFS thin plates under uniform com-
pression. The four node shell element with reduced integration
(S4R in ABAQUS) is chosen for the FE model. The element sizes
(12.7 mm × 12.7 mm) are set to 5% of the plate dimensions
(b) C-section with edge
(a) C-section with flat hole stiffened hole
web
edge stiffener
hole
Fig. 2. C-sections with web holes.
211
(254 mm × 254 mm). Two plate thicknesses are considered,
2.540 mm and 0.508 mm which covers the thickness of the SSMA
standard floor joists ( varies from 33 mil to 97 mil), The material
properties are Young's modulus E = 203.4 GPa and Poisson's ration
μ = 0.3. Fig. 3 shows the FE mesh and the boundary conditions of a
thin plate with a flat circular hole subjected to uniform compression.
In order to single out the effect of the hole size and the stiffener
length to the plate stability as well as to simplify the derivations, a
sharp corner is modeled for all holes. All the analyzed plates are
square and the hole is centered in the plate. A uniform compression
stress is applied at both ends of the thin plate. The plate is simply sup-
ported on four edges. The out-of-plane displacement of all the nodes
on the edges is restricted. Due to the symmetry of the plate geometry,
the mid nodes on the two loaded edges are fixed in the traverse direc-
tion and the mid nodes on the two unloaded edges are fixed in the
loading direction so that the model becomes solvable in the FE
program.
The FE model is verified by comparing the results of elastic buck-
ling loads with the classical thin plate solutions on plates without
holes. The FE results of critical elastic bucking load are 47.49 KN
for a 254 mm × 254 mm × 2.54 mm plate, and 0.381 KN for a
254 mm × 254 mm × 0.508 mm plate, which are close to the theoreti-
cal solutions 47.40 KN and 0.379 KN respectively. Then the FE model
is applied to a total of 198 different thin plates with unstiffened or
stiffened holes under uniform compression. Fig. 4 shows the typical
buckling shapes of thin plates. The results are summarized in Tables 1
and 2, and only the ratios of FE vs. theoretical results are listed. It in-
dicates that the elastic buckling load decreases if a flat/unstiffened
hole is formed in the plate, and the decreasing is greater when the
hole diameter becomes bigger. On the other hand, thin plates with
stiffened hole demonstrate improved elastic stability. The buckling
load increases as the edge stiffener length or/and the hole diameter
get larger. Because the existence of the edge stiffener changes the
boundary conditions of the flat portions of the plate from free to sup-
ported, the larger edge stiffener provides more restraint. Therefore
the buckling coefficient k increases from 0.43 (one edge simply sup-
ported, one edge free) to a higher value up to 5.42 (one edge simply
supported, one edge fixed) (Timoshenko and Gere 1961 [1]). The
hole diameter also influences the plate stability greatly. The classical
solution of the elastic buckling load (Fcr) of a thin plate is
π2 Et 3
Fcr ¼ k , where t is the plate thickness, h is the plate
12ð1−μ 2 Þh
width of the loaded side, k is the bucking coefficient, and the other
parameters are constants not related to the geometric profile. When
a stiffened hole is formed in the plate, the width h is reduced at the
hole cross-section, which helps increase the buckling load. At the
same time, the buckling coefficient k will not be changed significantly
for the resultant two flat portions since the edge stiffener produces
stiffened edges. By a combined effect of the edge stiffener and the re-
duced flat width, the elastic buckling of a thin plate can be improved
Fixed in 3
Fixed in 2
Fixed in 2
1
2
3
Fixed in 3
Nodes on four edges fixed in 1
Fig. 3. FE mesh and boundary conditions for plates under uniform compression.
212 C. Yu / Journal of Constructional Steel Research 71 (2012) 210–218

Fig. 4. Buckling shape of 1st mode for thin plates under compression.

Table 1
Ratios of FE vs. theoretical results of elastic buckling load for 254 mm × 254 mm × 2.54 mm plates under compression.

Edge stiffener height (mm)

0 2.54 5.08 10.2 15.2 20.3 25.4 30.5 38.1 45.7 50.8

Hole diameter (mm) 12.7 0.99 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
25.4 0.96 0.99 1.02 1.03 1.03 1.03 1.03 1.03 1.03 1.03 1.03
50.8 0.88 0.93 1.08 1.13 1.13 1.13 1.13 1.13 1.13 1.13 1.13
76.2 0.80 0.87 1.13 1.29 1.30 1.30 1.30 1.30 1.30 1.30 1.30
102 0.75 0.82 1.17 1.52 1.56 1.56 1.56 1.56 1.56 1.56 1.57
127 0.72 0.78 1.20 1.83 1.96 1.97 1.97 1.97 1.97 1.97 1.98
152 0.70 0.76 1.22 2.25 2.54 2.60 2.61 2.61 2.62 2.63 2.63
178 0.69 0.74 1.21 2.77 3.40 3.55 3.61 3.66 3.72 3.74 3.76
203 0.74 0.76 1.20 3.27 4.90 5.42 5.61 5.72 5.87 5.97 6.01

Table 2
Ratios of FE vs. theoretical results of elastic buckling load for 254 mm × 254 mm × 0.508 mm plates under compression.

Edge stiffener height (mm)

0 2.54 5.08 10.2 15.2 20.3 25.4 30.5 38.1 45.7 50.8

Hole diameter (mm) 12.7 0.99 1.01 1.01 1.01 1.01 1.01 1.01 1.01 1.01 1.01 1.01
25.4 0.96 1.04 1.04 1.04 1.04 1.04 1.04 1.04 1.04 1.04 1.04
50.8 0.88 1.13 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15 1.15
76.2 0.81 1.26 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35 1.35
102 0.76 1.42 1.64 1.65 1.65 1.65 1.65 1.65 1.65 1.65 1.65
127 0.73 1.59 2.06 2.10 2.11 2.11 2.11 2.12 2.12 2.12 2.12
152 0.71 1.76 2.69 2.82 2.83 2.85 2.86 2.87 2.88 2.89 2.89
178 0.70 1.91 3.52 3.76 3.82 3.87 3.92 3.97 4.05 4.12 4.17
203 0.75 2.08 5.43 6.26 6.38 6.49 6.61 6.72 6.88 7.04 7.15

greatly. The FE results show that the increase can be as high as 6 times
by comparing plates with stiffened holes and the plates without
holes. The hole diameter is the most influential. The edge stiffener
does not change the buckling load dramatically when the edge stiff-
ener is larger than 15.2 mm for 2.54 mm thick plates and 12.7 mm
for 0.508 mm thick plates.
The FE results also indicate that the influence on the elastic stabil-
ity by the edge stiffened holes has little correlation with the plate
thickness. The analyses on two groups of plates with quite different
thickness yield similar results in terms of the increasing ratios. There-
fore it can be concluded that the edge stiffened hole increases the
elastic buckling load of thin plates, and the ratios of hole diameter
vs. plate width (d/h) and edge stiffener length vs. plate width (q/h)
control the influence. For the range of plate thickness between
0508 mm and 2.54 mm, the edge stiffener with q/h = 0.06 is
recommended.
buckling analyses. It is shown that the shape of the first buckling
mode of a solid thin plate changes from a single half-wave for uni-
form compression to a full wave at the compressive area for stress
gradient. If a flat hole is formed in the plate, the first buckling shape
becomes one half-wave. If the hole is stiffened, the buckling shape re-
mains the same shape as that for solid thin plates. Tables 3 and 4 sum-
marize the results for the stress gradient cases. Similar to the findings
from the compression cases, the thin plates under stress gradient
yield higher elastic buckling loads as the edger stiffener gets longer
and/or the hole becomes larger. The edge stiffener length does not
Fixed in 3
Fixed in 2

2.2. Thin plates subject to transverse stress gradient

The web of a C-section member in flexure is actually subject to a Fixed in 2

stress gradient in the transverse direction therefore the stability of a 1
2
thin plate with stiffened opening under stress gradient is directly re-
lated to the performance of the new generation joists. The same thin 3
plates analyzed in the previous section are investigated here but sub- Fixed in 3
Nodes on four edges fixed in 1
ject to a stress gradient on both ends. Fig. 5 shows the FE model and
Fig. 6 shows typical buckling modes obtained from the elastic Fig. 5. FE mesh and boundary conditions for plates under stress gradient.
 C. Yu / Journal of Constructional Steel Research 71 (2012) 210–218 213

Fig. 6. Buckling shape of 1st mode for thin plates under stress gradient.

Table 3
Ratios of FE vs. theoretical results of elastic buckling load for 254 mm × 254 mm × 2.54 mm plates under stress gradient.

Edge stiffener height (mm)

0 2.54 5.08 10.2 15.2 20.3 25.4 30.5 38.1 45.7 50.8

Hole diameter (mm) 12.7 1.00 1.00 1.01 1.01 1.01 1.01 1.01 1.01 1.01 1.01 1.01
25.4 1.00 1.00 1.01 1.02 1.02 1.02 1.03 1.03 1.03 1.03 1.03
50.8 0.97 0.99 1.01 1.03 1.04 1.06 1.07 1.08 1.09 1.09 1.09
76.2 0.85 0.90 1.02 1.06 1.08 1.10 1.12 1.14 1.16 1.17 1.17
102 0.73 0.79 1.06 1.14 1.17 1.19 1.21 1.23 1.26 1.29 1.30
127 0.66 0.71 1.07 1.29 1.33 1.35 1.37 1.39 1.42 1.45 1.47
152 0.65 0.69 1.09 1.54 1.61 1.64 1.66 1.68 1.71 1.74 1.76
178 0.68 0.72 1.17 2.01 2.16 2.20 2.23 2.25 2.28 2.32 2.34
203 0.77 0.83 1.33 2.87 3.21 3.26 3.29 3.32 3.36 3.41 3.44

Table 4
Ratios of FE vs. theoretical results of elastic buckling load for 254 mm × 254 mm × 0.508 mm plates under stress gradient.

Edge stiffener height (mm)

0 2.54 5.08 10.2 15.2 20.3 25.4 30.5 38.1 45.7 50.8

Hole diameter (mm) 12.7 1.00 1.01 1.01 1.01 1.01 1.01 1.01 1.01 1.01 1.01 1.01
25.4 1.00 1.01 1.01 1.02 1.03 1.03 1.03 1.03 1.03 1.03 1.03
50.8 0.97 1.37 1.02 1.04 1.05 1.08 1.08 1.09 1.10 1.10 1.10
76.2 0.85 1.04 1.06 1.08 1.09 1.11 1.13 1.15 1.17 1.19 1.19
102 0.74 1.11 1.15 1.17 1.19 1.21 1.23 1.23 1.28 1.31 1.33
127 0.67 1.23 1.33 1.36 1.37 1.39 1.41 1.43 1.46 1.49 1.51
152 0.65 1.44 1.65 1.67 1.69 1.71 1.73 1.75 1.78 1.81 1.83
178 0.69 1.79 2.26 2.30 2.32 2.35 2.38 2.41 2.44 2.48 2.51
203 0.77 2.21 3.45 3.47 3.52 3.59 3.67 3.73 3.81 3.88 3.93

improve the thin plate stability significantly when the q/h ratio is
larger than 0.02 for 0.508 mm thick plates and 0.06 for 2.54 mm
thick plates. On the other hand, the elastic buckling load increases
continuously as the hole diameter gets larger. In summary, it can be
concluded that the optimized and most efficient length of edge stiff-
ener is 0.06 times of the plate width, and it is applicable to thickness
from 0.508 mm to 2.54 mm. The hole diameter has a positive relation
with the elastic buckling load of thin plates under a stress gradient.
3. Ealstic buckling of cold-formed steel flexural members with
stiffened holes
the joist length. The material properties are the same as those used
in the previous FE models for thin plates. The dimensions of the ana-
lyzed gross sections are chosen from two SSMA standard C-section
joists 1000S162-43 and 1000S250-97 (SSMA 2010 [6]). The addition-
al parameters considered in the analyses include the ratio of the hole
diameter vs. the web height (d/h) which varies from 0.2 to 0.8, the
ratio of the edge stiffener length vs. the web height (q/h) which var-
ies from 0.06 to 0.1, and the hole spacing which varies from 305 mm
to 1219 mm. The joist length is 6.10 m for all models.
Stress gradient applied to both ends

It was found from previous sections that the elastic buckling load
of thin plates with stiffened holes increases as the hole diameter be-
comes bigger and the influence by the edge stiffener tends to be
Nodes at end fixed in 1 and 2
unchanged as the q/h ratio is greater than 0.06. This section will pre-
sent the elastic buckling analyses on the CFS flexural members with
stiffened holes, and the emphasis is to study of the optimized diame-
ter for the circular holes as well as the hole spacing.
Fig. 7 shows the FE model for the elastic buckling analyses. To be Nodes at the mid length fixed in 3

consistent with the elastic buckling analyses of plates, a sharp corner

is the adopted in the FE models for members. Four node shell element 1
2
is chosen for the model and the element size is set to be
Nodes at end fixed in 1 and 2
12.7 mm × 12.7 mm. The joist is simply supported at both ends. The 3
mid-section of the joist is fixed in the longitudinal direction. Stress
gradient is applied to both ends to generate uniform moment along Fig. 7. FE model for elastic buckling analysis on joists with holes.
214 C. Yu / Journal of Constructional Steel Research 71 (2012) 210–218

Table 5
Ratios of elastic buckling moments of 1000S162-43 with holes vs. no holes.

q/h Hole spacing Lateral–torsional buckling Local buckling

(mm)
d/h d/h

0.2 0.4 0.5 0.6 0.8 0.2 0.4 0.5 0.6 0.8

0.06 305 1.01 0.98 0.96 0.92 0.79 1.03 1.17 1.30 1.46 2.07
610 1.01 0.99 0.98 0.95 0.87 1.01 1.03 1.05 1.07 1.12
914 1.02 1.00 0.99 0.97 0.89 1.00 1.02 1.03 1.03 1.06
1219 1.02 1.00 1.00 0.99 0.94 1.00 1.01 1.01 1.01 1.02
0.08 305 1.02 0.99 0.96 0.92 0.80 1.04 1.17 1.30 1.47 2.10
610 1.02 1.00 0.98 0.96 0.87 1.01 1.04 1.05 1.07 1.12
914 1.01 1.00 0.99 0.97 0.90 1.00 1.02 1.03 1.03 1.06
1219 1.01 1.01 1.00 0.98 0.94 1.00 1.01 1.01 1.01 1.02
0.10 305 1.02 0.99 0.96 0.93 0.80 1.04 1.18 1.31 1.48 2.12
610 1.01 1.00 0.98 0.96 0.87 1.01 1.04 1.05 1.07 1.12
914 1.02 1.00 0.99 0.97 0.90 1.00 1.02 1.03 1.03 1.06
1219 1.02 1.01 1.00 0.99 0.94 1.00 1.01 1.01 1.01 1.02

Table 6
Ratios of elastic buckling moments of 1000S250-97 with holes vs. no holes.

q/h Hole spacing Lateral–torsional buckling Distortional buckling

(mm)
d/h d/h

0.2 0.4 0.5 0.6 0.8 0.2 0.4 0.5 0.6 0.8

0.06 305 1.00 0.96 0.93 0.89 0.77 1.01 1.04 1.08 1.13 1.24
610 1.00 0.98 0.96 0.94 0.84 1.00 1.02 1.03 1.04 1.07
914 1.00 0.98 0.97 0.95 0.87 1.00 1.01 1.02 1.03 1.04
1219 1.00 0.99 0.98 0.97 0.92 1.00 1.01 1.02 1.03 1.03
0.08 305 1.00 0.97 0.94 0.91 0.78 1.01 1.04 1.08 1.13 1.23
610 1.00 0.98 0.97 0.94 0.86 1.00 1.02 1.03 1.04 1.07
914 1.00 0.99 0.98 0.96 0.88 1.00 1.01 1.02 1.03 1.04
1219 1.00 0.99 0.99 0.97 0.92 1.00 1.01 1.02 1.03 1.03
0.10 305 1.00 0.97 0.95 0.92 0.80 1.01 1.05 1.08 1.14 1.26
610 1.00 0.99 0.97 0.95 0.87 1.00 1.02 1.03 1.05 1.07
914 1.00 0.99 0.98 0.96 0.89 1.00 1.01 1.02 1.03 1.04
1219 1.00 0.99 0.99 0.97 0.93 1.00 1.01 1.02 1.03 1.04

A total of 240 member configurations were analyzed, and compar-
ison is made between the joists with web stiffened holes and the
joists without web holes in terms of the elastic buckling moments
of the lateral–torsional mode and the first localized mode (local buck-
ling for 1000S163-43, distortional buckling for 1000S250-97). The re-
sults are reported in Tables 5 and 6. The results indicate that
adding stiffened holes on the web will not greatly improve the later-
al–torsional buckling (LTB) moment. For most analyzed joists, the LTB
loads decrease slightly, and the LTB reduction increases as the hole di-
ameter gets larger. On the other hand, both the local buckling (for
1000S162-43) and the distortional buckling (for 1000S250-97) mo-
ment are improved when the joists have stiffened holes in the web
and the increasing grows as the hole diameter becomes bigger.
Therefore the optimized hole diameter should balance the two oppo-
site influences on the two different buckling modes. By considering
that lateral bridging is required by SSMA for up to every 2.13 m
length of the joists, it is chosen that the hole diameter vs. web height
ratio (d/h) of 0.5 be the optimized value for SSMA joists.
With respect to the effect by the hole spacing, the results show
that the shorter distance between holes the higher buckling moment
can be achieved for both the LTB and the localized buckling modes.
When the spacing is large enough, the original lower buckling mode
will occur at the solid area between holes and the joist will demon-
strate the same performance as the normal joists without holes.
Therefore 305 mm hole spacing is the choice for optimal perfor-
mance. For the edge stiffener length, Tables 5 and 6 indicate that

Lateral-torsional mode 1st localized buckling mode 1st localized buckling mode
for both cases of 1000S162-43 of 1000S250-97

Fig. 8. Observed elastic buckling shape for joists without holes.

 C. Yu / Journal of Constructional Steel Research 71 (2012) 210–218 215

Lateral-torsional mode 1st localized buckling mode 1st localized buckling mode
for both cases of 1000S162-43 with holes of 1000S250-97 with holes

Fig. 9. Observed elastic buckling shape for joists with stiffened holes.

Fig. 10. FE model for post-buckling analysis on joists with stiffened holes.

the analyzed ratios (q/h = 0.06, 0.08 and 0.1) do not have much dif-
ference in increasing buckling load. Thus the most efficient edge stiff-
ener length vs. web height ratio is 0.06 for SSMA joists.
Figs. 8 and 9 respectively show the three buckling modes (LTB,
local buckling, distortional buckling) for joists without holes and
with the optimized stiffened holes. It can be observed that the first lo-
calized buckling mode of 1000S162-43 changes from local buckling to
distortional buckling after the stiffened hole is formed on the web and
the moment is increased significantly (a 30% increase). For 1000S250-
97, adding stiffened holes does not change the lowest localized buck-
ling mode (still distortional buckling), and the buckling moment is
not improved significantly. Those observations indicate that the stiff-
ened hole can boost the elastic local buckling mode greatly but would
only slightly increase the elastic distortional buckling moment. It is of
importance to enforce the local buckling be the failure mode so that
the joists with edge stiffened holes will exercise its maximum
advantages.
4. Flexural strength of cold-formed steel joists with stiffened holes
The optimized dimensions for the stiffened holes are developed in
the previous section by elastic buckling analyses. This section is to de-
termine the flexural strength of the C-section joists with the opti-
mized stiffened holes. Non-linear post-buckling FE is employed to
simulate the behaviors of the cold-formed steel joists subject to typi-
cal loading and boundary conditions. Fig. 10 shows the post-buckling
FE model which simulates a typical four point bending test. The
6.10 m long joist is simply supported at both ends and loaded at

Table 7
Geometries and yield stress of the analyzed joists.

Member h (mm) a (mm) b (mm) θ (deg) t (mm) r (mm) d (mm) q (mm) fy (MPa)

600S162-33 152.4 41.3 12.7 90 0.8788 2.82 76.2 9.1 228

600S250-97 152.4 63.5 15.9 90 2.5832 6.45 76.2 9.1 345
800S162-33 203.2 41.3 12.7 90 0.8788 2.82 101.6 12.2 228
800S250-97 203.2 63.5 15.9 90 2.5832 6.45 101.6 12.2 345
1000S162-43 254.0 41.3 12.7 90 1.1455 2.95 127.0 15.2 228
1000S250-97 254.0 63.5 15.9 90 2.5832 6.45 127.0 15.2 345
1200S162-54 304.8 41.5 12.7 90 1.4376 3.61 152.4 18.3 228
1200S250-97 304.8 63.5 15.9 90 2.5832 6.45 152.4 18.3 345
216 C. Yu / Journal of Constructional Steel Research 71 (2012) 210–218

a is chosen to be the first localized elastic buckling mode obtained by

r r b tion is set to 0.034 times of the material thickness, which is corre-
θ sponding to an average imperfection value based on a statistical
q of forming, both increased yield stress and increased residual stresses
h d
t
Fig. 11. Definitions of member dimensions for C-section joists.
two 1/3 points. The joist is laterally supported by fixing the move-
ment of top corner nodes in direction 2 for every 305 mm. The corner
nodes at the other side of the top flange are forced to have the same
displacement in direction 1 as that of the laterally supported nodes.
This setup intends to reflect the restraint by the decking in real appli-
cations. Additionally, two bottom corner nodes at the 1/3 points are
also fixed in direction 2 to simulate the lateral bridging required by
SSMA in the real applications. The distance between the top and the
bottom flange at both ends end is kept unchanged during the FE sim-
ulation to prevent web crippling from occurring. The FE mesh and el-
ement type are the same as those used in the elastic buckling model
for thin plates. Two yield non-linear stress vs. strain curves are con-
sidered in the analyses, one curve with 227 MPa yield stress,
310 MPa tensile strength, and the other with 345 MPa yield stress,
448 MPa tensile strength. Both curves were obtained from actual ten-
sile tests conducted by Yu (2005 [7]). Geometric imperfections
are also considered in the post-buckling FE model. Since the lateral–
torsional buckling mode is restricted, the initial imperfection shape
the elastic FE model. The magnitude of the initial geometric imperfec-
study done by Schafer and Peköz (1998 [8]). Due to the cold-work
and strains exist at the corners. The two effects offset one another, it
is therefore a common practice to ignore the residual stresses and in-
creased yield stress at the corners in the FE analysis for CFS members
(Moen et al., 2008 [9], Schafer et al., 2010 [10]).
The post-buckling FE analyses cover both the regular C-section
joists and the joists with the optimized stiffened holes. Table 7 lists
the geometries and yield stresses of the studied joists. The selected
geometries include all the standard heights and the minimum/maximum
thickness of the SSMA (2010 [6]) floor joists. The notations used to define
the dimensions are shown in Fig. 11, and all the dimensions are out-to-
out measurements.
The results of post-buckling analyses are provided in Table 8. Five
normal joists (without web holes) failed in local buckling, and the
rest five joists failed in distortional buckling. Fig. 12(a) shows a local
buckling failure of the normal joist 1000S162-43 and Fig. 13(a)
shows a distortional buckling failure observed for a normal joist
1200S250-97. The flexural strengths of the normal joists obtained by
the FE model agree with the predictions by the Direct Strength Method
(DSM, Appendix 1 of AISI S100 [11]). The average FE vs. predicted ratio
is 1.034 with a standard deviation of 0.051. When the optimized stiff-
ened holes are formed on the web of the identical normal joists, the
flexural strengths are increased by an average of 12.9% (MfeH/MfeN
in Table 8). Fig. 12(b) and 13(b) shows the failure modes of the new
generation joists. It is found that all the analyzed new joists fail in a sim-
ilar mode, the modified distortional buckling mode, in which the flat
web portions between holes deform and the adjacent top flange rotates
at a wavelength of two hole spacing (610 mm).

Table 8
Results of post-buckling FE analysis on joists.

Member My (kip-in) McrL (kip-in.) McrD (kip-in.) MfeN (kip-in) MfeH (kip-in) MfeH/MfeN MfeN/MDSM MfeH/Mnew

600S162-33^ 2.23 1.92 2.78 1.80 2.17 1.208 1.000 1.077

600S250-97* 12.24 51.22 25.13 11.78 12.65 1.074 0.981 1.034
800S162-33^ 3.34 1.67 2.83 2.53 2.89 1.145 1.124 1.147
800S250-97* 18.07 48.72 31.17 16.59 18.38 1.108 0.983 1.018
1000S162-43^ 5.99 3.30 4.43 4.54 5.19 1.142 1.092 1.112
1000S250-97* 24.66 43.74 35.32 22.36 24.98 1.117 1.028 1.013
1200S162-54^ 14.81 5.86 6.35 9.52 10.25 1.076 1.039 0.993
1200S250-97* 32.03 39.40 35.79 26.66 30.91 1.159 1.026 0.965
Average 1.129 1.034 1.045
St. Dev. 0.045 0.051 0.062

Note: ^ — normal joist (without holes) failing in local buckling; * — normal joist failed in distortional buckling; My— yield moment; McrL— elastic local buckling moment; McrD—
elastic distortional buckling moment; MfeN— flexural strength of normal joists; MfeH— flexural strength of joists with stiffened holes; MDSM— flexural strength Direct Strength
Method; Mnew— flexural strength by the proposed design method for joists with stiffened holes.

a b

Fig. 12. Failure modes for 1000S162-43 joists without holes or with stiffened holes.
 C. Yu / Journal of Constructional Steel Research 71 (2012) 210–218 217

a b

Fig. 13. Failure modes for 1200S250-97 joists without holes or with stiffened holes.

Since the stiffened holes improve the flexural strength of the conservative predictions for the new generation joists. The new de-
SSMA joists by an average of 12.9%, the design method, DSM which sign curve provides reasonable and conservative predictions for the
is adopted by North American Specification is no longer accurate new generation C-section joists. The average FE vs. predicted ratio
and cannot account for the advantages of the new generation joists. for the new curve is 1.045 with a standard deviation of 0.062 (as
A new design procedure based on the DSM concept is developed to listed in Table 8).
predict the flexural strength of the SSMA C-section joists with opti- The resistance factors,ϕ for LRFD design method can be deter-
mized edge stiffened holes on the web. The new provisions are pro- mined in accordance with Chapter F of AISI S100 (2007 [2]) with a
posed as follows. target reliability index, β, of 2.5. The resistance factors, ϕ, can be de-
termined by the following Equation.
for λ‘ ≤ 0:925 Mn‘ ¼ My
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi
−β 2
VM þVF þCP VP þVQ
2 2 2
!0:35 ! !0:35 ϕ ¼ Cϕ ðMm Fm Pm Þe
McrL McrL
for λ‘ N 0:925 Mn‘ ¼ 1−0:05 My
My My
where:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where λ‘ ¼ My =Mcr‘ , Mcr‘ is the elastic local buckling moment of the Cϕ calibration coefficient (1.52 for LRFD);
joist without considering the holes and My is the yield moment. Mm mean value of material factor;
The proposed design procedure is applicable for the laterally- Fm mean value of fabrication factor;
braced C-section joists with the optimized edge stiffened holes. The Pm mean value of professional factor;
hole spacing shall not be larger than 305 mm and the edge stiffener β target reliability index;
length vs. web height ratio shall not be less than 0.06. The hole diam- VM coefficient of variation of material factor;
eter is half of the web height. And the joist should be laterally sup- VF coefficient of variation of fabrication factor;
ported to prevent lateral–torsional buckling. Fig. 14 shows a Cp coefficient factor;
comparison of the FE results with the design curves. It illustrates VP is coefficient of variation of test results;
that DSM gives good agreement with both local and distortional VQ coefficient of variation of load factor (0.21 for LRFD).
buckling strength of the normal C-section joists, but yields over
The values of Mm VM Fm and VF are taken from Table F1 in AISI
S100 (2007)[2] using Flexural Members–Bending Strength as the
type of component. Table 9 summarizes the adopted factors. The cal-
1.1 culated resistance factor for the proposed provision is 0.93 which is
Proposed Curve for Joists with Stiffened Holes higher than the AISI S100 specified value for DSM provisions in
1 beam design (0.90). It is recommended to use the AISI S100 specified
resistance factor and safety factor for the proposed design method.
0.9

0.8
Mfe/My

DSM Distortional Curve Table 9

Resistance factor for proposed design methods.
0.7
Quantity 8
DSM Local Curve
Mean 1.045
0.6 St. dev. 0.062
COV 0.0593
Normal Joists Failed in Local Buckling Mm 1.10
0.5 Vm 0.10
Normal Joists Failed in Distortional Buckling
Joists with Stiffened Holes Fm 1.00
Pm 1.045
0.4
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Vf 0.05
β (LRFD) 2.5
(My/Mcr)0.5 VQ 0.21
ϕ (LRFD) 0.93
Fig. 14. Comparison of FE results with design curves.
218 C. Yu / Journal of Constructional Steel Research 71 (2012) 210–218
5. Conclusions
The elastic buckling analyses on CFS thin plates and C-section flex-
ural members with edge stiffened holes are analyzed by FE models in
ABAQUS. The results show that the stiffened holes can greatly in-
crease the critical loads of those CFS members. A parameter study is
performed to obtain optimized dimensions of the hole profiles for
SSMA C-section joists. And those new generation joists with opti-
mized holes are analyzed by post-buckling FE model to determine
the ultimate flexural strength. The results indicate that an average
12.9% increase in the flexural strength can be achieved by forming
the optimized stiffened holes on the web of SSMA joists. A DSM
based new design method for the CFS C-section flexural members
with optimized edge stiffened holes is developed. The new method
has shown a good agreement with the results of the FE analyses and
demonstrates the same level of reliability as the DSM.
References
[1] Timoshenko SP, Gere JM. Theory of elastic stability. New York, U.S.A: McGraw-Hill
Book Company; 1961.
[2] AISI S100. North American specification for the design of cold-formed steel structural
members, 2007 Edition. Washington, DC: American Iron and Steel Institute; 2007.
[3] Shan MY, LaBoube RA, Yu WW. Behavior of web elements with openings sub-
jected to bending, shear and the combination of bending and shear. Final Report,
Civil Engineering Series 94–2, Cold-Formed steel Series, Department of Civil
Engineering, University of Missouri-Rolla; 1994.
[4] NAHB. “innovative residential floor construction: structural evaluation of steel
joists with pre-formed web openings”, Research Report H-21134CA. Upper Marl-
boro, MD: NAHB Research Center, Inc.; 1999.
[5] ABAQUS. ABAQUS Version 6.4. RI, USA: SIMULIA, Providence; 2003.
[6] SSMA. Product technical information ICBO ER-4943P, 2010 Edition. Chicago, IL:
Steel Stud Manufacturers Association; 2010.
[7] Yu, C. (2005). “Distortional buckling of cold-formed steel beams”, Ph.D. Disserta-
tion, The Johns Hopkins University, Baltimore, MD.
[8] Schafer BW, Peköz T. Computational modeling of cold-formed steel: characterizing
geometric imperfections and residual stresses. J Constr Steel Res, 47 (3. Elsevier;
1998. p. 193–210.
[9] Moen CD, Igusa T, Schafer BW. Prediction of residual stresses and strains in cold-
formed steel members. Thin-Walled Struct, 46. Elsevier; 2008. p. 1274–89.
[10] Schafer BW, Li Z, Moen CD. Computational modeling of cold-formed steel. Thin-
Walled Struct, 48. Elsevier; 2010. p. 752–62.
[11] DSM. Appendix 1 of AISI S100 design of cold-formed steel structural members
using the direct strength method — 2007 Edition. Washington, DC: American
Iron and Steel Institute; 2007.