See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/265524935

Extending Direct Strength Design to Cold-Formed Steel Beams with Holes

Article · January 2010

CITATIONS READS
13 560

2 authors:

Cristopher Dennis Moen Benjamin Schafer

Virginia Tech (Virginia Polytechnic Institute and State University) Johns Hopkins University
80 PUBLICATIONS   1,854 CITATIONS    284 PUBLICATIONS   9,778 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Cold-Formed Steel C-Section Joists with Unstiffened Web Holes View project

GOALI: Enabling Advanced Wind Turbine Tower Manufacturing with Reliability-Based Design View project

All content following this page was uploaded by Cristopher Dennis Moen on 31 January 2015.

The user has requested enhancement of the downloaded file.

Twentieth International Specialty Conference on Cold-Formed Steel Structures
St. Louis, Missouri, USA, November 2010

Extending Direct Strength Design to

Cold-Formed Steel Beams with Holes
Cristopher D. Moen1, Benjamin W. Schafer 2

Abstract

The extension of the American Iron and Steel Institute’s Direct

Strength Method (DSM) to cold-formed steel beams with holes is nearly in
place. DSM was first introduced to the AISI specification in 2004 as an
alternative to the effective width method, and is widely considered a major
advancement in cold-formed steel component design. In DSM, the beam elastic
buckling properties for a general cross-section are obtained with a computer
analysis utilizing the finite strip method. A disadvantage of the finite strip
method and DSM has been that discrete holes along the member length could
not be easily accounted for, although the recent development of simplified
elastic buckling approximations including holes has now alleviated the inherent
shortcoming. This paper provides an introduction to the DSM approach for cold-
formed steel beams with holes, where the critical elastic buckling moments for
local, distortional, and global buckling are calculated including the presence of
holes, and then input into strength prediction expressions modified to capture the
strength reduction from yielding at the net section. A DSM design example of a
joist with evenly spaced web holes is provided.

Introduction

Cold-formed joists are a popular structural component in the floor systems of

low and midrise buildings. These thin-walled structural steel flexural members
are manufactured by cold bending steel sheet into an open cross-section, most
commonly a C-section. The joists are provided with evenly spaced web holes to
accommodate the passage of electrical conduits, plumbing pipes, and HVAC
ducts. Hole sizes and shapes vary by manufacturer, and the hole edges can be
either unstiffened (Figure 1a) or stiffened (Figure 1b).
______________________
1
Assistant Professor, Virginia Tech, Blacksburg, VA, 24061, USA. ([email protected])
2
Professor, Johns Hopkins University, Baltimore, MD, 21218, USA. ([email protected])
 (a)

(b)
Figure 1 (a) Cold-formed steel joist with unstiffened punched web holes (b) C-section joists
with stiffened circular holes (photos courtesy of Don Allen)

The broad range of hole shapes, sizes, and spacings in cold-formed steel
construction today has exceeded the original scope of the American Iron and
Steel Institute (AISI) design equations developed for beams with holes over the
last four decades. The current AISI design equations were derived within the
context of the effective width method (Yu 2000), and address the influence of
unstiffened holes on local buckling dominated failures. The AISI provisions are
written specifically for unstiffened holes in C-section webs, and are limited to a
somewhat restricted range of hole sizes and spacings. For example, the effective
width equations are only applicable when unstiffened web holes are provided
with a centerline spacing of 457 mm (18 in.) or greater, and where the hole
depth is less than 63 mm (2.5 in.) regardless of the cross-section dimensions
(AISI-S100 2007, Section B2.4).

The AISI specification addresses the influence of unstiffened holes on local

buckling through the effective width method, however holes are not currently
considered for global buckling and distortional buckling limit states. When
unstiffened holes are present in a cold-formed steel beam, the critical elastic
flexural-torsional buckling load decreases relative to the same beam without
holes, which increases the global slenderness and decreases predicted strength
(Moen and Schafer 2009a). For distortional buckling, a form of buckling related
to intermediate and/or edge stiffeners commonly observed in open cross-
sections, the presence of unstiffened web holes decreases the stabilizing
influence of the web on the cross-section, reducing the critical elastic
distortional buckling moment and increasing the tendency for distortional
buckling to initiate at a hole (Kesti 2000; Moen and Schafer 2008; Moen and
Schafer 2009a). A more general design method which considers the influence of
holes across all cold-formed steel limit states is needed.

An AISI research program was recently completed that capitalizes on advances

in cold-formed steel strength prediction, and specifically the AISI Direct
Strength Method (DSM) (AISI-S100 2007, Appendix 1), to deliver a more
general design approach for cold-formed steel beams with holes. DSM
represents an important advancement in cold-formed steel design because it
provides engineers and cold-formed steel manufacturers with the tools to predict
member strength for a general cross-section. With the design approach
summarized herein, DSM can now safely predict the strength of cold-formed
steel flexural members with the ever expanding range of cross-section types,
hole sizes, shapes and spacings common in industry. Note that this paper
focuses on flexural strength prediction for beams with unstiffened holes.
However, the DSM approach is also applicable to beams with stiffened holes,
and work is underway to formalize the design equations and elastic buckling
framework (Moen and Yu 2010).

The AISI Direct Strength Method

DSM for beams without holes

The AISI Direct Strength Method employs the elastic buckling properties of a
general cold-formed steel cross-section to predict strength. For members without
holes, the elastic buckling properties are obtained from an elastic buckling curve
generated with freely available software, for example CUFSM (Schafer and
Ádàny 2006) and GBTUL (Bebiano et al. 2008), which perform a series of
eigen-buckling analyses over a range of buckled half-wavelengths. An example
of an elastic buckling curve is provided in Figure 2 for a cold-formed steel C-
section beam, highlighting the three categories of elastic buckling considered in
DSM – local, distortional, and global buckling – where Mcrl, Mcrd, and Mcre are
the respective elastic buckling moments.
 Figure 2 Elastic buckling curve for a cold-formed steel beam without holes

Flexural capacity is calculated with DSM considering three limit states – global
buckling, local-global buckling interaction, and distortional buckling (AISI-
S100 2007, Appendix 1). The global strength of an unbraced beam span, Mne, is
determined with the global slenderness, λc=(My/Mcre)0.5; Mnl is calculated with
the local slenderness, λl=(Mne/Mcrl)0.5; and Mnd is obtained with the distortional
slenderness, λd=(My/Mcrd)0.5. When slenderness is high for global or distortional
buckling limit states, i.e. when Mcre or Mcrd is small relative to the yield moment
of the beam, My=SfFy, flexural strength is limited by elastic buckling. (Note that
Sf is the section modulus referenced to the outer fiber that yields first and Fy is
the steel yield stress.) When λc or λd is low, the flexural strength is controlled
by inelastic buckling and yielding. Considering the local-global buckling
interaction limit state for unbraced beams, the flexural-torsional buckling
capacity is reduced from Mne to Mnl to account for local buckling along the beam
span. The minimum strength from the three limit states is taken as the beam’s
flexural capacity, i.e. Mn=min(Mne, Mnl, Mnd).

Strategy for extending DSM to beams with holes

A logical extension of the Direct Strength approach to cold-formed steel beams

with holes is to maintain the assumption that elastic buckling properties can be
used to predict strength. For a beam with holes, this means that the elastic
buckling moments Mcrl, Mcrd, and Mcre, are calculated including the influence of
holes. A suite of simplified methods for obtaining these elastic buckling
moments was recently developed as an alternative to cumbersome thin-shell
finite element eigen-buckling analysis. The elastic buckling moments, including
the influence of holes, can be calculated with finite strip analysis or hand
calculations derived from classical buckling solutions (Moen and Schafer 2009a;
Moen and Schafer 2009b). The simplified elastic buckling prediction methods
are demonstrated in an example at the end of this paper.

It was concluded in the AISI research program that the elastic buckling moments
including the influence of unstiffened holes are viable parameters for predicting
capacity in a Direct Strength approach (Moen 2008). However, when yielding
controls strength, modifications to the existing DSM design expressions for
beams without holes were needed to limit flexural capacity to that of the net
section, i.e. Mynet=SfnetFy, where Sfnet is the section modulus at the net section.
Furthermore, the AISI research program concluded that inelastic buckling and
collapse at a hole may control flexural strength with intermediate slenderness
ranges (Figure 3), requiring a transition from the elastic buckling regime to the
net section limit (Moen 2008). DSM distortional buckling design expressions
presented in the following section have been modified to provide this transition.
For local-global buckling interaction, Mnl is capped at Mynet, imposing the net
section strength limit when flexural capacity is governed by inelastic buckling
and yielding, i.e. when λl and λc are both low.

Figure 3 DSM distortional buckling curve for beams with holes

DSM design expressions for beams with holes

The nominal strength of a cold-formed steel beam with holes shall be the
minimum of Mne, Mnl, and Mnd as given in the following sections.

Global Buckling

The nominal flexural strength, Mne, for lateral-torsional buckling is:

M ne = M cre for M cre < 0.56 M y
10 ⎛ 10 M y ⎞
= M y ⎜⎜1 − ⎟ for 2.78M y ≥ M cre ≥ 0.56 M y
⎟
9 ⎝ 36 M cre ⎠
= M y for M cre > 2.78M y (1)
where Mcre includes the influence of hole(s).

Local Buckling Interaction

The nominal flexural strength, Mnl, for local-global buckling interaction is:

M nl = M ne ≤ M ynet for λl ≤ 0.776

0.4
⎡ ⎛ M ⎞⎤
0 .4
⎛ M crl ⎞
= ⎢1 − 0.15⎜⎜ crl ⎟⎟⎥ ⎜ ⎟
⎜ M ⎟ M ne for λl > 0.776 , (2)
⎣⎢ ⎝ M ne ⎠⎦⎥ ⎝ ne ⎠
where λl=(Mne/Mcrl)0.5, and Mcrl includes the influence of hole(s).

Distortional Buckling
The nominal flexural strength, Mnd, for distortional buckling is:

M nd = M ynet for λd ≤ λd1

⎛ M ynet − M d 2 ⎞
= M ynet − ⎜⎜ ⎟(λ d − λ d 1 ) for λd1 < λd ≤ λd 2
⎟
⎝ λd 2 − λd1 ⎠
⎛ ⎛ M ⎞ ⎞⎟⎛ M ⎞
0.6 0 .6
⎜
= ⎜1 − 0.22⎜ crd ⎟ ⎟⎜ crd ⎟ M y for λd > λd 2 , (3)
⎜ ⎜ M y ⎟ ⎟⎜ M y ⎟
⎝ ⎝ ⎠ ⎠⎝ ⎠
where λd=(My/Mcrd)0.5, λd1=0.673(Mynet/My),  λd2=0.673(1.7(My/Mynet)1.7‐0.7),  Mcrd
includes the influence of hole(s), and

(
M d 2 = 1 − 0.22 (1 λ d 2 )
0.5
)(1 λ d2 )0.5 M y . (4)

Design example

The DSM design approach outlined in Eq. (1) to Eq. (4) is employed to calculate
the capacity of a perforated cold-formed steel joist (Figure 4) with an SSMA
550S162-33 lipped C-section (SSMA 2001), where Fy=55 ksi.

Joist geometry, boundary conditions, and loading

The joist carries a uniform vertical load and is assumed to be fully braced
against lateral-torsional buckling (Figure 4). Distortional buckling and local
buckling are viable strength limit states.

Figure 4 Column dimensions and boundary conditions

Gross and net section properties

The gross section and net section properties (Table 1) are calculated with the
section property calculator in CUFSM. To determine the net section properties
in CUFSM, assign a thickness of zero to the elements at the location of the
perforations, but do not delete them. Assuming 55 ksi steel, My=29.15 kip·in.
and Mynet=28.95 kip·in.
 Table 1 (a) gross section properties, (b) net section properties

(a)
Ag Ix Iy rx ry J Cw xo yo I xo I yo r xo r yo ro
in.2 in.4 in.4 in. in. in.4 in.6 in. in. in.4 in.4 in. in. in.
0.327 1.46 0.11 2.11 0.59 0.000130 0.682 -1.11 0.00 1.86 0.11 2.39 0.59 2.46
(b)
A net I xnet I ynet r xnet r ynet J net C wnet x onet y onet I xonet I yonet r xonet r yonet r onet
2 4 4 4 6 4 4
in. in. in. in. in. in. in. in. in. in. in. in. in. in.
0.275 1.45 0.10 2.29 0.61 0.000110 0.677 -1.20 0.00 1.84 0.10 2.59 0.61 2.66

Elastic buckling analysis

Local buckling

Local buckling in a cold-formed steel beam with holes is assumed to occur as

either buckling in the gross cross-section between holes (Mcrlnh) or buckling of
the compressed strip adjacent to a hole (Mcrlh). The buckled mode shape with the
lowest critical buckling load defines Mcrl, i.e. Mcrl=min(Mcrlnh, Mcrlh). The elastic
buckling curve for the gross cross-section (generated with CUFSM, see Figure
5) is used to obtain Mcrlnh. Taking the first minimum on the elastic buckling
curve, Mcrlnh= 17.61 kip·in. at a half-wavelength Lcrlnh=3.0 in.

The net-section elastic buckling curve is generated in CUFSM by modifying the

gross section node and element geometry such that one finite strip element with
t=0 spans across the hole (Figure 6). A reference moment of 1 kip·in. is applied
to the cross section and CUFSM (Properties screen) is used to calculate the
corresponding stress distribution. The zero thickness element is then deleted,
and the two corners of the cross-section in compression are restrained in the
CUFSM z-direction. The resulting mode shape and elastic buckling curve is
provided in Figure 6. The lowest buckling load of the unstiffened strip occurs at
a half-wavelength less than the length of the perforation (Lcrlh =4.25 in. versus
Lhole=4.5 in.) meaning that the buckled half-wave can form within the length of
the hole, and therefore Mcrlh=10.51 kip·in. (Note that Mcrlh could be tabulated for
standard punchout sizes and shapes as a convenience to the engineer!)

Local buckling is predicted to occur in the net cross section since Mcrlh<Mcrlnh
and therefore Mcrl= 10.51 kip·in. The local buckling moment is 40% lower at a
hole, which means that buckling will tend to occur as unstiffened strip buckling
rather than in the web of the gross cross-section between holes.
Figure 5 Elastic buckling curve for gross cross-section

Figure 6 Local buckling curve for net cross-section

Distortional buckling

The critical elastic buckling moment for distortional buckling, including the
influence of web holes, is calculated by first obtaining the distortional buckling
half-wavelength from a finite strip analysis of the gross cross-section (Lc rd =16.6
in., see Figure 5). The reduced web bending stiffness caused by a hole over one
distortional half-wavelength is simulated by reducing the cross-section thickness
of the web (Moen and Schafer 2009a):
1/ 3
⎛ L ⎞
t r = ⎜⎜1 − h ⎟⎟ t . (5)
⎝ Lcrd ⎠
For Lh=4.5 in. and t=0.0346 in., tr =0.0311 in. which is then implemented in a
second finite strip analysis (Figure 5) performed just at Lcrd=16.6 in., resulting in
M crd =20.45 kip·in. The presence of perforations reduces Mcrd by 13% when
compared to a distortional buckling moment of 23.43 kip·in for a beam without
holes (Figure 5). Note that the beneficial influence of the moment gradient on
M crd (Yu 2005) is negligible and not considered because the beam’s span length
is much longer than Lcrd .

Ultimate Strength Calculation

Inputs from the elastic buckling analysis include:
M y := 29.15 ⋅ kip ⋅ in M crL := 10.51 ⋅ kip⋅ in
M ynet := 28.95 ⋅ kip⋅ in M crd := 20.45 ⋅ kip ⋅ in

DSM global buckling strength Eq. (1)

M ne := M y beam is fully braced against lateral-torsional
buckling

M ne = 29.15 ⋅ kip ⋅ in

DSM local buckling strength, Eq. (2)

M ne local slenderness
λL := λL = 1.6654 (including influence of
M crL
(subscript "L" = "l") holes)
M nL := (
min M ne , M ynet ) if λL ≤ 0.776

⎡⎡ ⎛ McrL ⎞ ⎤⎥ ⎛ McrL ⎞
0.4 0.4 ⎤
⎢⎢ ⎥
⎢⎢1 − 0.15⋅ ⎜ M ⎟ ⎥ ⎜ M ⎟ ⋅ Mne⎥ if λL > 0.776
⎣⎣ ⎝ ne ⎠ ⎦ ⎝ ne ⎠ ⎦
M nL = 17.45 ⋅ kip⋅ in

DSM distortional buckling strength, Eq. (3)

My distortional
λd := λd = 1.1939 slenderness
M crd (including influence
of holes)

⎛ Mynet ⎞
λd1 := 0.673 ⋅ ⎜ ⎟ λd1 = 0.6684
⎝ My ⎠

⎡ ⎛ M ⎞ 1.7 ⎤
⎢ y ⎥
λd2 := 0.673 ⋅ 1.7 ⎜ ⎟ − λd2 = 0.6865
⎢ 0.7
⎥
⎣ ⎝ M ynet
⎠ ⎦

⎡⎢ 0.5⎤
⎥ ⋅⎛ 1 ⎞
0.5
M d2 := 1 − 0.22⋅ ⎛⎜ ⎞
1
⋅ My M d2 = 25.8⋅ kip ⋅ in
⎢ λd2 ⎟ ⎥ ⎜ λd2 ⎟
⎣ ⎝ ⎠ ⎦⎝ ⎠

M nd := M ynet if λd ≤ λd1

⎡ ⎛ Mynet − Md2 ⎞ ⎤
⎢Mynet − ⎜ ⎟ ⋅ ( λd − λd1)⎥ if λd1 < λd ≤ λd2
⎣ ⎝ λd2 − λd1 ⎠ ⎦
⎡⎡ ⎛ Mcrd ⎞ ⎤⎥ ⎛ Mcrd ⎞
0.6 0.6 ⎤
⎢⎢ ⎥
⎢⎢1 − 0.22⋅ ⎜ M ⎟ ⎥ ⎜ M ⎟ ⋅ My⎥ if λd > λd2
⎣⎣ ⎝ y ⎠ ⎦⎝ y ⎠ ⎦
M nd = 19.4⋅ kip ⋅ in
Predicted flexural capacity (including holes):

((
M n := min M ne M nL M nd )) LRFD (prequalified section)
ϕb := 0.90
M n = 17.45 ⋅ kip⋅ in (MnL controls)
ϕb ⋅ M n = 15.7⋅ kip ⋅ in

Local buckling at a hole is predicted as the governing failure mode, with a

decrease in flexural strength of 15% when compared to the same beam without
holes. This result is contrary to the AISI Main Specification Section B2.4,
which states that when dh/h<0.38, holes do not influence local buckling
capacity. It will be difficult to make definitely conclusions on the validity of
the Main Specification versus DSM until more experimental data is generated
for cold-formed steel joists with unstiffened holes. The elastic buckling
prediction of the unstiffened strip employed in DSM is certainly more
representative of the actual buckling behavior when compared to the Main
Specification as the net section finite strip approach (see Figure 6) considers
cross section interaction.

Conclusions

The AISI Direct Strength Method (DSM) for cold-formed steel beams with
holes utilizes the critical elastic buckling loads of a beam, including the
influence of holes, to predict strength. The elastic buckling predictions are
obtained with a suite of recently developed simplified methods that employ
finite strip analysis and hand calculations derived from classical buckling
solutions. The existing DSM design expressions for beams without holes have
been modified to limit flexural capacity to the strength of the net cross section,
and in the case of distortional buckling, a transition from the net section capacity
to the elastic buckling regime was added to predict flexural strength influenced
by inelastic buckling at the net cross section. DSM provides an accessible
design approach for cold-formed steel beams that can account for holes across
global, local, and distortional buckling limit states with improved accuracy and
generality when compared to existing strength prediction methods.

Ongoing research

The DSM approach presented in this paper has been developed and validated
primarily with nonlinear finite simulations (Moen 2008) in part because of the
lack of experimental data. An experimental program was recently completed by
the first author considering cold-formed steel joists with unstiffened holes which
will be used to supplement the ongoing validation effort.
 References

AISI-S100. (2007). North American Specification for the Design of Cold-

Formed Steel Structural Members, American Iron and Steel Institute,
Washington, D.C.
Bebiano, R., Pina, P., Silvestre, N., and Camotim, D. (2008). "GBTUL -
Buckling and Vibration Analysis of Thin-Walled Members."
DECivil/IST, Technical University of Lisbon
(http://www.civil.ist.utl.pt/gbt).
Kesti, J. (2000). "Local and Distortional Buckling of Perforated Steel Wall
Studs," Dissertation/Thesis, Helsinki University of Technology, Espoo,
Finland.
Moen, C. D. (2008). "Direct Strength Design for Cold-Formed Steel Members
with Perforations," Ph.D. Thesis, Johns Hopkins University, Baltimore.
Moen, C. D., and Yu, C. "Elastic buckling of thin-walled structural components
with stiffened holes." 51th AIAA/ASME/ASCE/AHS/ASC Structures,
Structural Dynamics and Materials Conference, April 12, 2005 - April
15, 2010, Orlando, FL.
Moen, C. D., and Schafer, B. W. (2008). "Experiments on cold-formed steel
columns with holes." Thin-Walled Structures, 46, 1164-1182.
Moen, C. D., and Schafer, B. W. (2009a). "Elastic buckling of cold-formed steel
columns and beams with holes." Engineering Structures, 31(12), 2812-
2824.
Moen, C. D., and Schafer, B. W. (2009b). "Elastic buckling of thin plates with
holes in compression or bending." Thin-Walled Structures, 47(12),
1597-1607.
Schafer, B. W., and Ádàny, S. "Buckling analysis of cold-formed steel members
using CUFSM: conventional and constrained finite strip methods."
Eighteenth International Specialty Conference on Cold-Formed Steel
Structures, Orlando, FL.
SSMA. (2001). Product Technical Information, ICBO ER-4943P, Steel Stud
Manufacturers Association, <www.ssma.com>.
Yu, C. (2005). "Distortional buckling of cold-formed steel members in
bending," Ph.D. Thesis, Johns Hopkins University, Baltimore.
Yu, W. W. (2000). Cold-Formed Steel Design, John W. Wiley & Sons, New
York, NY.

View publication stats