www.kssc.or.

kr

Steel Structures 6 (2006) 71-91

Stability Analysis and Design of Steel Building Frames

Using the 2005 AISC Specification
Donald W. White *, Andrea E. Surovek , Bulent N. Alemdar , Ching-Jen Chang ,
Yoon Duk Kim and Garret H. Kuchenbecker
1,

Structural Engineering, Mechanics and Materials, Georgia Institute of Technology, Atlanta, GA 30332-03551, USA
Civil and Environmental Engineering, South Dakota School of Mines and Technology, Rapid City, SD 57701, USA
3
RAM International, 2744 Loker Avenue West, Carlsbad, CA, 92010, USA
4
Structural Engineering, Mechanics and Materials, Georgia Institute of Technology, Atlanta, GA, USA
5
Hermanson Egge Engineering, Rapid City, SD, USA

1
2

Abstract
The 2005 AISC Specification reflects the latest advances in the stability analysis and design of structural steel buildings. The
new Specification defines the general requirements for stability analysis and design and gives engineers the freedom to select
or devise their own methods within these constraints. It also provides several specific procedures. This paper first gives an
overview of the elastic analysis and design procedures in AISC (2005) as well as specific second-order distributed plasticity
methods upon which, in part, these procedures are based. The relationship between the AISC elastic provisions and the refined
inelastic methods is explained. Secondly, the paper highlights one interpretation of the AISC inelastic analysis and design
provisions that greatly facilitates the application of elastic-plastic hinge methods of analysis. The paper closes by presenting
four basic examples selected to illustrate key characteristics of each of the methods.

Keywords: Advanced analysis, distributed plasticity analysis, direct analysis, effective length, plastic hinge analysis
1. Introduction

Chapter C of the AISC (2005a) Specification, Stability

Analysis and Design, states that any analysis and design
method that addresses the following effects on the overall
stability of the structure and its elements is permitted:
1. Flexural, axial and shear deformations (in members,
connections and other components),
2. Reduction in stiffness (and corresponding increases
in deformations) due to residual stresses and material
yielding,
3. P- effects, which are the effects of axial loads P
acting through the relative transverse displacements
of the member ends (see Fig. 1),
4. P- effects, which for initially straight members
loaded by bending in a single plane, are due to the
member axial load acting through the transverse
bending displacements relative to the member chord
(see Fig. 1), and
5. P-o and P-o effects, which are caused by the
member axial loads acting through unavoidable
initial o and o geometric imperfections (within
*Corresponding author
Tel: +404-894-5839, Fax: +404-894-2278
E-mail: dwhite@ce.gatech.edu

fabrication and erection tolerances) relative to the

ideal configuration of the structure.
The above Chapter C statement gives engineers the
freedom to select or devise methods that are best suited
for the various structure types encountered in practice. It
allows for innovation as long as there is proper
consideration of the physical effects that influence the
structural response.
Since the 1961 AISC Specification, when the column
effective length concept was first introduced by AISC,

Figure 1. Second-order P- and P- effects (from White

and Kim (2006)).

72

Donald W. White et al.

American analysis and design methods have addressed all

of the above effects in some fashion whenever they are
deemed to have an important influence on the structural
response. Also, there has always been implicit recognition
that engineers can use their professional judgment to
disregard specific effects (e.g., member shear deformations,
connection deformations, etc.) whenever they are considered
negligible. Member yielding, residual stress effects, and
geometric imperfection effects traditionally have been
addressed in the formulation of member design resistances,
and, with minor exceptions, have not been considered in
the analysis. In some cases, engineers have included a
nominal out-of-plumbness effect in the analysis of gravity
load combinations, particularly if the geometry and loading
are symmetric. Strictly speaking, this is not necessary for
the in-plane strength assessment of beam-columns in the
prior AISC Specifications. However, this practice is
necessary to determine -o effects on bracing forces,
beam moments, connection moments, and in-plane moments
for checking the out-of-plane resistance of beam-columns.
Also, plastic design procedures incorporate material
yielding into the analysis, typically by the use of idealized
plastic hinge models in adequately braced compactsection members. However, geometric imperfection and
residual stress effects traditionally are not included in this
type of analysis.
The 1961 AISC Specification, and other AISC
Specifications up until 1986, relied strictly on the
structural analysis
for calculation of geometrically
linear (first-order) forces and moments in the idealized
geometrically-perfect, nominally-elastic (or elastic-plastic)
structure. For elastic analysis and design, second-order
( - and -) effects were addressed very discreetly via
the following beam-column strength interaction equation:
P

only

fa
Fa

Cm fb

fa
F
Fe b

---- + -----------------------

(1)

1.0

1 -----
'

The expression
Cm
AF
fa
Fe

(2)

----------------- =

1 -----
'

in Eq. (1) amplifies the member flexural stresses b to

account for the - and - effects.
Generally, Eq. (2) gives only a coarse approximation of
the true second-order effects. LeMessurier (1977) and
others later addressed better ways to determine the
amplified bending stresses in general rectangular frames,
all within the context of Allowable Stress Design (ASD).
However, even today (2005), engineers using the AISC
(1989) ASD provisions can apply Eq. (2) in ways that
significantly under-estimate the physical second-order
effects in certain cases. The accuracy hinges largely on
the proper determination of e, the member axial stress at
f

incipient elastic buckling (considering the interaction of

the member with the rest of the structure). The parameter
e' is calculated by dividing this buckling stress by the
factor of safety for column elastic buckling, 23/12 = 1.92.
The axial stress e' is determined typically using the
equation
F

Fe

E
2
(KLb rb )

(3)

12
' = --------------------------23

where b b is the column effective slenderness ratio

in the plane of bending, and is the effective length
factor associated with the above buckling solution. Also,
this factor is used typically in calculating the column
axial resistance a. If desired, e' = e/1.92 and a can be
determined directly from the buckling analysis model.
The term m in Eqs. (1) and (2) is discussed subsequently.
AISC (1986) LRFD was the first American Specification
to refer explicitly to the calculation of second-order
moments from a structural analysis. This specification
introduced the following two-equation format for the
beam-column resistance:
KL /r

Pu
Mu

c P n b Mn

------------- + ----------2

Pu
c Pn

Mu

b Mn

8
---------- + -- ----------9

for

1.0

for

1.0

Pu
<
c Pn
----------

Pu

c Pn
----------

(4a)

0.2

(4b)

0.2

where u is defined as the maximum second-order

elastic moment along the member length. AISC (1986)
states that u may be determined from a second-order
elastic analysis using factored loads. However, it also
provides a procedure that uses amplification factors to
calculate the second-order elastic moments from a firstorder elastic analysis. This procedure is in fact an
approximate second-order analysis. The moments u in
Eqs. (4) are the second-order elastic moments in an
idealized model of the structure assuming initially perfect
geometry and completely linear elastic material response.
In all the AISC Specifications from AISC (1961)
through AISC (1989) ASD and AISC (1999) LRFD, the
influence of geometric imperfections and residual stresses
was addressed solely within the calculation of the
member resistances ( a and b in ASD and n and n in
LRFD). The new
provisions in AISC
(2005a) recognize that specific advantages can be
realized by moving an appropriate nominal consideration
of these effects out of the resistance side and into the
structural analysis side of the design equations. By
incorporating an appropriate nominal consideration of
these effects in the analysis, the resistance side of the
design equations is greatly simplified, and the accuracy of
the design checks is generally improved. These attributes
are discussed further in the subsequent sections.
All of the above analysis and design procedures are
based inherently on the use of second-order analysis
M

direct

analysis

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

(first-order elastic analysis with amplifiers being one type

of second-order elastic analysis). Elastic analysis does not
include the consideration of the member resistances in
itself. Therefore, all of the above elastic methods must
include member resistance equations. However, the
method of analysis and the equations for checking the
member resistances are inextricably linked. Changes in
the analysis calculation of the required strengths (e.g., a
and b m/(1 a/ e') in Eq. (1) or u and u in Eqs. (4))
can lead to simplifications in the calculation of the
member resistances (typically a in Eq. (1) or n in Eqs.
(4)). Specifically, if the structural analysis is configured
to provide an appropriate representation of the internal
member forces, the in-plane resistance of the structure
can be checked entirely on a cross-section by crosssection basis.
AISC (2005a) adopts the equations from AISC (1999)
LRFD as a base representation of the beam-column
resistances for all of its analysis and design procedures.
However, the notation is generalized such that the
equations apply to both ASD and LRFD:
f

f C

P M
-----r + 8---------r 1.0
P c 9 Mc

P M
-------r- + 8---------r 1.0
2 P c 9 Mc

P
P
P
----- < 0.2
P

for -----r 0.2

c

for

(5b, AISC H1-1a)

(5a, AISC H1-1b)

The terms in these equations are defined as follows:

r = the required axial compressive strength, determined
in ASD by analyzing the structure under 1.6 times the
ASD load combinations and then dividing the results by
1.6, or determined in LRFD by analyzing the structure
under the LRFD load combinations.
r = the required flexural strength, determined in ASD
by analyzing the structure under 1.6 times the ASD load
combinations and then dividing the results by 1.6, or
determined in LRFD by analyzing the structure under the
LRFD load combinations.
c = the allowable or design axial compressive strength,
given by n c in ASD or by c n in LRFD, where n is
the nominal compressive resistance determined in accordance
with Chapter E.
c = the allowable or design flexural strength, given by
n b in ASD or by b n in LRFD, where
n is the
corresponding nominal resistance determined in accordance
with Chapter F.
c and b = resistance factors for axial compression and
bending, both equal to 0.9.
c and b = factors of safety for axial compression and
bending, both equal to 1.67.
The above 1.6 factor for ASD is smaller than the
column safety factor of 1.92 in the AISC ASD (1989)
amplification of the flexural stresses (see Eqs. (1) through
(3)). However, ASD-H1 also states that m shall be taken

as 0.85 in Eqs. (1) and (2) for frames subject to joint

translation. This m value typically underestimates the
sidesway amplification effects (Salmon and Johnson, 1996).
Nevertheless, the ASD moment amplifier summarized in
Eq. (2) is still conservative in many practical cases. This
is because the predominant second-order effects are often
associated solely with the structure sidesway. Equation
(1) applies a single amplifier indiscriminately to the total
flexural stresses from both sidesway and non-sway
displacements. The amplification factor procedure in
AISC (1999) LRFD and AISC (2005a) is more accurate,
but involves a cumbersome subdivision of the analysis
into separate no-translation (nt) and lateral-translation (lt)
parts. Kuchenbecker
. (2004) and White
. (2005a
& b) outline an amplified first-order elastic analysis
approach that provides good accuracy for rectangular
framing. This approach avoids the above cumbersome
attributes of the AISC (1999 & 2005a) amplification factor
procedure.
In many cases, Eqs. (5) provide a more liberal
characterization of the beam-column resistances than the
multiple beam-column strength curves in AISC ASD
(1989). Also, AISC (2005a) provides the following
alternative interaction equation to characterize the out-ofplane flexural-torsional buckling resistance of doublysymmetric I-section members subjected to axial compression
and major-axis bending moment:
C

et al

P /

M /

-------r + --------r 1.0

Pco Mcx

et al

(6, AISC H1-2)

where co is the out-of-plane column strength and cx

is the flexural strength with respect to lateral-torsional
buckling. White and Kim (2006) provide a detailed
discussion of the background to Eq. (6) and explain that
this equation should be applied only for doubly-symmetric
compact I-section members.
In the subsequent developments, it is useful to consider
the characterization of separate in-plane and out-of-plane
beam-column resistances. This can be accomplished by
using Eqs. (5) with different definitions of c and c, or
by using Eqs. (5) to characterize the in-plane strength and
Eq. (6), where it is applicable, to characterize the out-ofplane strength. The in-plane resistance is estimated with
Eqs. (5) by neglecting: (1) out-of-plane flexural, torsional
or flexural-torsional buckling in the calculation of c and
(2) lateral-torsional buckling in the calculation of c. The
out-of-plane resistance is estimated with Eqs. (5) by using
c = co and using the governing resistance from Chapter
for c (White and Kim, 2006).
For checking the in-plane and out-of-plane strength of
I-section members, Eqs. (5) must be used. However,
for doubly-symmetric compact I-section members subjected
to in-plane major-axis bending and large axial loads, Eq.
(6) provides a more liberal assessment of the out-of-plane
P

general

The AISC (2005a) equation numbers are denoted by AISC followed by the equation number.

73

74

Donald W. White et al.

flexural-torsional buckling strength. AISC (2005a) allows

the engineer to neglect out-of-plane moments whenever
Mry/Mcy is smaller than 0.05 in the out-of-plane direction,
where Mry and Mcy are the required moment and the
corresponding resistance in this direction. Otherwise, an
extended form of Eqs. (5) must be used that includes the
out-of-plane bending.

2. Overview of Stability Analysis and Design

Procedures
2.1. Design by distributed plasticity analysis

This section discusses the requirements that must be

satisfied for strength design using a refined second-order
inelastic frame analysis in which the spread of yielding is
tracked explicitly through the cross-section and along the
member length. This type of analysis is referred to in this
paper as a distributed plasticity analysis.
For hot-rolled compact I-section members, ASCE (1997),
Deierlein (2003), and Surovek-Maleck and White (2003
& 2004) have shown that distributed plasticity analysis
closely replicates the in-plane AISC LRFD beam-column
strengths based on an exact inelastic effective length, for
a comprehensive range of end conditions, when the
following nominal geometric imperfections, residual stresses
and material idealizations are included in the analysis:
A sinusoidal or parabolic out-of-straightness with a
maximum amplitude of o = L/1000, where L is the
unsupported length in the plane of bending.
An out-of-plumbness of o = L/500, the maximum
tolerance specified in the AISC (2005b) Code of Standard
Practice.
The Lehigh (Galambos and Ketter, 1959) residual
stress pattern shown in Fig. 2.
An elastic-perfectly plastic material stress-strain
response.
These results are not surprising, since Eqs. (5) were
originally developed in part based on calibration to results
from this type of analysis (ASCE, 1997; Surovek-Maleck
and White, 2004). Since the AISC (2005a) analysis and
design procedures do not make any distinction between
hot-rolled and general built-up members, the above
nominal geometric imperfections and residual stresses are
sufficient to capture the Specification requirements for all
compact I-section member types. Figure 3 defines two
simple cases that illustrate the above findings. These are
W10 60 cantilever beam-columns with L/r = 40 subjected
to axial and transverse loads at their free ends. In the first
case, the member is subjected to major-axis bending,
while in the second case it is subjected to minor-axis
bending. Figure 4 compares the nominal strength interaction
curves obtained from the above type of distributed
plasticity analysis to the corresponding AISC (2005a)
beam-column strength curve. The AISC curve is obtained
using the AISC (2005a) effective length method with K =
2, and is the same for both the minor and major-axis
bending examples, since their L/r values are the same in

Lehigh (Galambos and Ketter 1959) residual

stress pattern.
Figure

2.

Figure 3.

Example cantilever beam-columns.

the plane of bending. Martinez-Garcia (2002), SurovekMaleck and White (2003), and Deierlein (2003) summarize
the results from other more comprehensive studies.
If the above distributed plasticity analysis is to be used
in an AISC (2005a) LRFD context, the resistance factors
c = b = 0.9 must be included. One way of doing this is
to determine the nominal beam-column strength curves as
shown in Fig. 4, and then to multiply both the abscissa
and the ordinate by c = b = 0.9 to obtain the final
member design resistance. However, identical results are
obtained if both the yield strength Fy and the elastic
modulus E are factored by 0.9. If only the yield strength
Fy is factored by 0.9, the design strengths are overestimated
for highly slender columns that fail by elastic buckling.
The factoring of both E and Fy by 0.9 up front is
preferred, since this approach facilitates the general
inelastic analysis and design of structural systems. There
is no straightforward way of applying distributed plasticity
analysis, or any other form of inelastic analysis, in the
context of ASD. In as such, AISC (2005a) disallows the
use of inelastic analysis for this approach. Most of the

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

75

values producing the smallest total resistance should be

selected. For structures in which the postbuckling response
is asymmetric, with one direction corresponding to an
unstable postbuckling response and the other corresponding
to a stable one, the direction associated with the unstable
postbuckling response gives the smallest total resistance.
Although they can be programmed, the above out-ofstraightness considerations involve a level of complexity
that many engineers would find unacceptable. The r <
0.15 eL rule allows the engineer to disregard these
considerations in many practical situations.
The appropriate direction for the out-of-plumbness is
typically much easier to specify than the appropriate
direction for out-of-straightness. For the overall assessment
of building frames, it is sufficient in the vast majority of
cases to specify the out-of-plumbness in the total net
direction that the structure sways under the applied
loadings.
At the present time (2005), distributed plasticity analysis
has been applied most commonly in research studies
investigating the resistance of frames composed of
adequately braced compact-section members. However,
in design practice, the member strengths can be governed
by out-of-plane buckling, flange or web local buckling, or
combinations of these strength limits. These limit states
cannot be captured by a planar distributed plasticity
analysis. Although some progress has been made on 3D
distributed plasticity methods (e.g., see Pi and Trahair
(1994), Izzudin and Smith (1996), Teh and Clarke (1998),
Battini and Pacoste (2002) and Nukala and White (2004)
among others), the complexity and cost of the analysis is
significantly greater. Furthermore, the appropriate handling
of residual stresses, geometric imperfections, warping
continuity at beam-to-column joints, local-overall member
buckling interactions, restraint from and interaction with
floor slabs, and other important attributes that can influence
the 3D response have not been studied thoroughly at this
time. In fact, many of these effects are considered in
rather simplistic ways in ordinary design practices, e.g.,
out-of-plane and spatial beam-column resistances are
based typically on the assumption of unrestrained warping
at the member ends. Nevertheless, in frames subjected
predominantly to in-plane loading, the authors assert that
distributed plasticity analysis can be utilized to determine
an accurate estimate of the internal forces in the structure.
These forces then can be checked against Specification
member resistance equations corresponding to any of the
above 3D limit states not included in the analysis.
Various analysis refinements are possible relative to the
above procedures. For instance, a number of approaches
have been suggested for reducing the nominal out-ofplumbness relative to base values, e.g., see the Commentary
of AISC (2005), White
. (2003) and CEN (2003).
Also, other nominal residual stress distributions as well as
more complete stress-strain models (e.g., models including
strain-hardening) can be incorporated within the distributed
plasticity analysis. However, these considerations introduce
P

Nominal strength curves by distributed plasticity

analysis versus the AISC (2005a) effective length method
for the example W10 60 cantilever beam-columns.
Figure 4.

subsequent discussions in this paper are phrased in the

context of LRFD.
The distributed plasticity analysis solution captures the
in-plane design resistances of compact I-section members
completely when it includes the above attributes. Therefore,
the in-plane beam, column and beam-column resistance
checks are automatically satisfied for these member types
if the distributed plasticity analysis shows that the
structure supports the design loadings. The engineer does
not need to perform any separate evaluation of the inplane member resistances. The Australian Standard
AS4100 (SAA, 1990) was the first to explicitly permit
this type of analysis and design. This Standard coined the
term
, in a very specific context, to
denote an analysis that supersedes the Specification member
strength checks. Subsequently, various SSRC publications
(e.g., White and Chen (1993)) as well as other research
papers and reports have adopted this terminology.
In many practical steel building structures, member
out-of-straightness has little to no effect on the frame
resistance. White and Nukala (1997) explain that when
r is less than eL/7 0.15 eL, where
= 1.6 for ASD or 1.0 for LFRD and
using the moment of inertia I in the plane
eL =
of bending,
out-of-straightness effects can be neglected generally in
the distributed plasticity analysis (i.e., o can be taken equal
to zero). Otherwise, one must determine the appropriate
direction for the member out-of-straightness. Usually, the
strength is reduced the most due to member out-ofstraightness if o is specified in the direction of the
member deformations (relative to its chord) due to the
applied loads. However, in some cases, it is advisable to
specify the various member o values in the pattern of the
fundamental buckling mode obtained from an eigenvalue
buckling analysis. In these cases, the direction of the o
advanced analysis

EI/L

et al

76

Donald W. White et al.

additional complexities that must be addressed. The

above procedures fully satisfy the base requirements of
the AISC (2005a) Specification.

2.2. Elastic analysis and design methods in AISC

(2005a)

The AISC (2005a) Specification defines three specific

elastic analysis and design methods. These are:
1. The direct analysis method, detailed in Appendix 7,
2. The effective length method, detailed in Section
C2.2a and
3. The first-order analysis method, detailed in Section
C2.2b.
Table 1, summarizes the key attributes of each of these
methods. Within the restrictions specified on their usage,
and provided that effects such as connection rotations or
member axial and shear deformations are properly
considered in the analysis when these attributes are
important, each of the above methods is intended to
comprehensively address all of the effects listed at the
beginning of Section 1. The following subsections provide
an overview of these AISC (2005a) procedures. The
reader is referred to Deierlein (2004), Nair (2005a), Nair
(2005b) and White and Kim (2006) for additional
discussions.

2.2.1. Direct analysis method

The direct analysis method is the only one of the above

three procedures that is generally applicable to all types
of frames. This method involves two simple modifications
to the second-order elastic analysis: (1) the use of a
nominal reduced elastic stiffness and (2) the inclusion of
a nominal initial out-of-plumbness. These two devices are
adjustments to the analysis that approximate the internal
forces and moments from the type of distributed plasticity
analysis explained in the previous section. The reduced
elastic stiffness is taken as 0.8 of the nominal elastic
stiffness of the structure, except in members subjected to
large axial loads of P > 0.5P , where the member
flexural rigidity is taken as 0.8 times the column inelastic
stiffness reduction factor (see Table 1). The base
nominal out-of-plumbness is taken as = L/500, the
same value as discussed previously for distributed plasticity
analysis. However, the direct analysis provisions permit
the use of a smaller nominal out-of-plumbness where
justified. For instance, when the sidesway amplification
2nd/1st is smaller than 1.5, AISC (2005a) permits the
out-of-plumbness effect to be neglected when the
associated notional load (see the discussion below) is
smaller than the corresponding applied lateral load.
Many engineers may prefer to model the above out-ofplumbness effects by using notional lateral loads. As
shown in Fig. 5, if the framing above and below a given
vertical load elevation has the same /L, the out-ofplumbness effect can be represented accurately by
applying a notional lateral load of N = Y /L at the level
under consideration, where Y is the total vertical load
r

applied at this level. Table 1 emphasizes the use of

notional lateral loads. However, in general cases where
questions may arise about the appropriate calculation of
these loads, one can always use the more fundamental
out-of-plumb geometry. For example, the total base shear
due to any out-of-plumbness is always zero, and thus the
total base shear due to the notional loads also must be
zero. As shown in Fig. 5, the notional loads arise from the
sum of the P shear forces above and below each level.
The P shear at the base of the structure offsets the sum
of all the horizontal notional loads in the analysis model.
Explicit modeling of the out-of-plumbness also removes
the need to calculate different notional loads for different
load combinations.
The direct analysis method provides an improved
representation of the structures distributed plasticity
forces and moments at the strength limit of the most
critical member or members. Due to this improvement in
the calculation of the internal forces and moments, AISC
(2005a) bases its calculation of P , the column nominal
strength for checking the in-plane resistance in Eqs. (5),
on the actual unsupported length in the plane of bending.
In short, the need to calculate in-plane effective length
(K) factors is eliminated.
Interestingly, the use of P = P for members with
compact cross-section elements was considered in the
development of the direct analysis approach (Maleck,
2001). Although this is a viable option, it requires the
modeling of out-of-straightness in the analysis for members
subjected to large axial compression (to properly capture
in-plane limit states dominated by non-sway column
flexural buckling). The modeling of member out-ofstraightness adds an additional level of complexity to the
analysis, and in many steel structures, P based on the
actual unsupported length is only slightly smaller than P .
Therefore, AISC (2005a) uses P based on the actual
unsupported length (K = 1) to capture the influence of
potential in-plane non-sway column flexural buckling.
For certain member types, such as tapered members,
there are significant advantages to using the cross-section
axial strength rather than the nominal buckling strength
P as the axial strength term in the beam-column
interaction check (White and Kim, 2006). In cases where
the member axial loads are small and the cross-section is
compact, the column and beam-column resistances are
represented accurately using P = P , without the inclusion
of any member out-of-straightness in the analysis. This
simplification is appropriate whenever P < 0.1P . For
members with cross-section elements that are slender
under axial compression, White and Kim (2006) show
that P may be taken as QP when the above limit is
satisfied, where Q is the AISC form factor accounting for
local buckling effects with the web edge stress f taken as
F.
AISC (2005a) introduces a plethora of rules intended to
allow (and provide limits on) the use of various idealizations
and approximations (see Table 1). This characteristic is
o

ni

ni

ni

ni

ni

ni

ni

eL

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

77

Summary of specific AISC (2005a) elastic analysis-design procedures adapted from White and Kim (2006)
Direct analysis (Appendix 7) Effective length (Section C2.2a) First-order analysis (Section C2.2b)
Limitations on the
2nd/1st 1.5, Pr 0.5Py
None
2nd/1st 1.5
(See Note 5)
use of the method
First-order, B1 is applied to the
Type of analysis
Second-order (See Note 1)
Second-order (See Note 1)
member total moment
Structure geometry
Nominal (See Note 2)
Nominal
Nominal
used in the analysis
i Minimum if 2nd/1st
Notional load to be 0.002
1.5
Additive
0.002 i minimum
2.1(/ )Yi 0.0042 i additive
applied in the analysis (See Note 2) if 2nd/1st > 1.5
Table 1.

0.8 * Nominal, except eff =

0.8b when r > 0.5 y
(See Note 3)
b = 4[ r y(1 r y)]
Use of b = 1 is permitted in all
cases if additional notional loads
of 0.001 i are applied, additive
to other lateral loads
EI

EI

Effective stiffness
used in the analysis

P /P

Nominal

Nominal

P /P

is based on the unsupported

length in the plane of bending,
= 1)
i (i.e.,

Pni

in moment-frame columns is
based on a buckling analysis or
the corresponding effective
length ; ni in all other cases
is based on i = i (i.e., = 1).

Pni

KL P

KL

In-plane flexural
buckling strength ni

ni is based on the unsupported

If r < 0.15 eL, or if a member
length in the plane of bending, i
out-of-straightness of 0.001L or
the equivalent notional loading If 2nd/1st < 1.1, K may be taken
equal to one in all cases.
is included in the analysis, ni
may be taken equal to y
(See Note 4)
no is based on the unsupported length in the out-of-plane direction, o
Out-of-plane flexural
buckling strength no Alternatively, no may be based on an out-of-plane buckling analysis or the corresponding effective
length o (see Note 4)
General Note. 2nd/1stt is the ratio of the 2nd-order drift to the 1st-order drift (for rectangular frames, 2nd/1st may be taken as B2 calculated
by Section C2.1b). /L is the largest 1st-order drift from all the stories in the structure. In structures that have flexible diaphragms, the /L in
each story is taken as the average drift weighted in proportion to the vertical load, or alternatively, the maximum drift. All 2nd/1st and /L
ratios shall be calculated using the LRFD load combinations or using a factor of = 1.6 applied to the gravity loads in ASD. The factor a is
1.0 for LRFD and 1.6 for ASD. The term Y is the total gravity load applied at a given level of the structure. P is the member elastic
buckling resistance based on the actual unsupported length in the plane of bending, 2EI/L2 for prismatic members.
Note 1. Any legitimate method of second-order analysis that includes both P and P effects is permitted, including 1st-order analysis with
amplifiers. For P < 0.15P , a P-large delta (P-) analysis using one element per member generally provides an accurate solution for the
sidesway displacements and the corresponding internal second-order forces and moments. However, for members with P > 0.05P ,
either multiple elements must be used per member to obtain accurate second-order internal moments (unconservative error less than or
equal to 5%) in general from a P-large delta analysis, or a P-small delta amplifier must be applied to the element internal moments. Accurate
general P- analysis solutions may be obtained by maintaining P < 0.05P , where P = 2EI/l2 is the Euler buckling load in the plane of
bending based on the element length l. Second-order analysis methods that directly include both P- and P- effects at the element level
generally provide better accuracy than P-large delta analysis procedures. The target of 5% maximum unconservative error is based on the
original development of the AISC LRFD beam-column strength equations (ASCE 1997; Surovek-Maleck and White 2004a).
Note 2. A nominal initial out-of-plumbness of o/L = 0.002 may be used directly in lieu of applying 0.002Y minimum or additive notional
loads.
Note 3. The nominal stiffness and geometry should be employed for checking serviceability limit states. The reduced effective stiffness and
the notional loads or nominal initial out-of-plumbness are required only in considering strength limit states.
Note 4. AISC (2005) does not explicitly state this provision in the context of the direct analysis method. This provision is encompassed
within the Chapter C requirements for general stability analysis and design, which allow any method of analysis and design that addresses
the effects listed at the beginning of Section 1.
Note 5. The largest unconservative error associated with the limit P < 0.1P is approximately 5% and occurs for a simply-supported,
concentrically loaded column with zero moment and P = 0.1P = P . The target of 5% maximum unconservative error is based on the
original development of the AISC LRFD beam-column strength equations (ASCE 1997; Surovek-Maleck and White 2004a).
Note 6. The 1st-order analysis method does not account for the influence of significant axial compression in the rafters or beams of portal
frames. Therefore, this method strictly should not be applied for the analysis and design of the primary moment frames in these types of
structures.
P

KL

eL

eL

el

el

eL

eL

eL

78

Donald W. White et al.

Relationship between notional lateral loads and

nominal out-of-plumbness.
Figure 5.

typical of design Specifications and is not new to these

AISC provisions. One can avoid the need to consider all

of these caveats by using the direct analysis method,

calculating Pn as the applicable column strength based
on the actual unsupported length, in-plane or out-ofplane, including an elastic stiffness reduction factor of 0.8
or 0.8b in the analysis as applicable, and including a
uniform initial out-of-plumbness of 0.002L in the analysis
relative to the perfect geometry of the structure.

2.2.2. Effective length method

The AISC (2005a) effective length method is the same

as the traditional AISC method of analysis and design,
but with the addition of a notional minimum lateral load
for gravity-only load combinations. This minimum lateral
load accounts for the influence of nominal geometric
imperfections on the brace forces, beam moments,
connection moments and in-plane moments used for outof-plane strength design of beam-columns. In actuality,
the effects of any physical out-of-plumbness are present
for all load combinations. However, these effects are
overwhelmed by the effects of the primary lateral loads in
all the ASCE 7 (ASCE, 2005) lateral load combinations,
as long as the structures sidesway amplification is not
excessive. Therefore, in the AISC (2005a) effective length
method, the notional lateral loads are specified solely as
minimum lateral loads in gravity-only load combinations.
AISC (2005a) does not allow the use of the effective

length method when the second-order amplification of the

sidesway displacements is larger than 1.5, i.e., 2nd/1st >
1.5 (based on the nominal elastic stiffness of the
structure). This is because the effective length method
significantly underestimates the internal forces and
moments in certain cases when this limit is exceeded
(Deierlein, 2003 & 2004; Kuchenbecker et al., 2004;
White and Kim, 2006). For structures with 2nd/1st > 1.5,
AISC (2005a) requires the use of the direct analysis
method. Correspondingly, when using the direct analysis
approach with structures having 2nd/1st < 1.5, AISC
(2005a) allows the engineer to apply the notional lateral
loads (or the corresponding nominal out-of-plumbness) as
minimum values solely in the gravity-only load combinations.
For column and beam-column in-plane strength assessment
in moment frames, the effective length approach focuses
on the calculation of the member axial stresses Fei at
incipient buckling of an appropriately selected model (the
subscript i is used to denote in-plane flexural buckling).
This buckling model is usually some type of subassembly
that is isolated from the rest of the structural system
(ASCE, 1997). Engineers often handle the elastic buckling
stresses (Fei) implicitly, via the corresponding column
effective lengths KLi. The effective length is related to the
underlying elastic buckling stress via the relationship
2
Fei = ------------E-------2(KLi ri)

(7a)

or
Ki =

( E ) ( Li r i )

-------------------------------

Fei

(7b)

In the effective length method, the effects of residual

stresses, P-o effects and P-o effects are addressed
implicitly by the calculation of Pni from the column
strength equations. These equations can be written in
terms of either KLi or Fei (AISC 2005a). Unfortunately,
the selection of an appropriate subassembly buckling
model generally requires considerable skill and judgment.
As a result, there is a wide range of different buckling
models and K factor equations. In certain cases, subtle
differences in the models can produce radically different
results (ASCE, 1997).
In particular, one should note that a rigorous buckling
analysis of the entire structure does not necessarily
provide an appropriate Fei or Ki (ASCE, 1997). Members
that have small axial stress Fei at the buckling limit
(relative to 2E/(Li/ri)2) tend to have large values for Ki
from Eq. (7b). In some cases, these large Ki values are
justified while in other cases they are not. If the member
is indeed participating in the governing buckling mode, a
large Ki is justified. If the member is largely undergoing
rigid-body motion in the governing buckling mode, or if
it has a relatively light axial load and is predominantly
serving to restrain the buckling of other members, a large

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

Ki value often is not justified. The distinction between

these two situations requires engineering judgment. Some

of the situations requiring the greatest exercise of
judgment to avoid excessively large K values include: (1)
columns in the upper stories of tall buildings, (2) columns
with highly flexible and/or weak connections and (3)
beams or rafters in portal frames, which may have
significant axial compression due to the horizontal thrust
from the base of the frame.
There is no simple way of quantifying the relative
participation of a given member in the overall buckling
of the structure or subassembly under consideration.
Quantifying the participation requires an analysis of the
sensitivity of the buckling load to variations in the
member sizes. Even if one conducted such an analysis,
there is no established metric for judging when Eq. (7b)
should or should not be used. Engineers typically base
their effective length calculations on story-by-story models
to avoid the first of the above situations. They idealize
columns with weak and/or flexible connections as pinended leaner columns with K = 1 to avoid the second of
the above situations. Lastly, many engineers utilize K = 1
for design of the beams or rafters in portal frames for the
axial compression effects, although Eq. (7b) may suggest
K > 1 based on the Fei from an eigenvalue buckling
analysis.
The direct analysis method provides a more straightforward
and accurate way of addressing frame in-plane stability
considerations. By including an appropriately reduced
nominal elastic stiffness, an appropriate nominal out-ofplumbness of the structure, and an appropriate out-ofstraightness (for members subjected to high axial loads)
in the analysis, the member in-plane length effects can be
removed entirely from the resistance side of the design
equations. The member in-plane column strength Pni may
be taken simply as Py for members that satisfy the
previously discussed caveats. In-plane stability is addressed
by estimating the actual required internal cross-section
strengths Pr and Mr directly from the analysis, and by
comparing these required strengths against appropriate
cross-section based resistances. Alternatively, to avoid the
need to include out-of-straightness effects in members
with large axial loads, Pni may be calculated using the
actual in-plane unsupported length Li (K = 1).

2.2.3. First-order analysis method

The first-order analysis method, summarized in Table

1, is implicitly a simplified conservative application of
the direct analysis approach, targeted at rectangular or
tiered building frames. White and Kim (2006) detail the
conservative assumptions invoked in the development of
this procedure. Although the first-order analysis method
can be useful for simplified analysis and design of some
types of frames, this method is really just a direct analysis
with a number of simplifying assumptions. There are
numerous other ways to apply direct analysis using an
approximate second-order analysis, many of which are

79

often significantly more accurate. Therefore, the firstorder analysis method is not considered further in this
paper.

2.2.4. General comments

Both the direct analysis and effective length methods

require a second-order elastic analysis. However, any
second-order elastic analysis procedure is sufficient,
including first-order analysis with amplifiers, assuming
that the procedure is sufficiently accurate or conservative.
The above methods differ in the way that they handle
geometric imperfection and distributed yielding effects in
the second-order analysis model and in the member
resistance equations.
The beam-column out-of-plane resistance check is the
same in both of the above methods, albeit with different
values of Pr and Mr. In AISC (2005a), the simplest outof-plane beam-column resistance check is given by Eqs.
(5) but with Pn = Pno, where Pno is the out-of-plane
flexural, torsional or flexural-torsional buckling strength
of the member as a concentrically-loaded column. Eq. (6)
generally provides a more liberal estimate of the out-ofplane flexural-torsional resistance of doubly-symmetric
compact I-section members.
2.3. Design by direct elastic-plastic hinge analysis

Since 1961, the AISC Specifications have permitted the

use of plastic analysis and design in cases where members
subjected to plastic hinging satisfy requirements that
ensure their ductility. However, the AISC Specifications
from 1969 through 1999 have also generally required the
engineer to supplement the plastic analysis by beamcolumn strength interaction checks in which the axial
resistance term is based on a member effective length.
This practice adds significant complexity to the AISC
plastic analysis and design procedures. Furthermore, at
best, the resulting beam-column interaction equations
provide only an approximate assessment of the frame
stability behavior under progressive plastic hinge formation.
In many cases, they overly restrict the forces and moments
in sway frame columns (Ziemian et al., 1992; McGuire,
1995).
The AISC (2005a) direct analysis method can be
extended to provide an attractive alternative to the above
procedures. The extension is very simple - for members
that satisfy separate requirements to ensure sufficiently
ductile response (i.e., sufficient rotation capacity), moment
redistribution is allowed based on the assumption of
elastic-perfectly plastic hinge behavior at the limit of the
member resistance from Eqs. (5). This extension satisfies
the AISC (2005a) Appendix I provisions for inelastic
analysis and design. The separate ductility requirements
in AISC (2005a) Appendix 1 include restrictions on:
The material yield strength Fy,
The flange and web slenderness values bf/2tf and hp/tw,
The member out-of-plane slenderness Lb/ry,
The magnitude of the axial force Pr ( = 1.0 for

80

Donald W. White et al.

LRFD loadings) and

The connection details.
These restrictions tend to preclude the formation of
plastic hinges in beams and beam-columns with noncompact
flanges or webs, in beams where the resistance is governed
by out-of-plane lateral-torsional buckling, and in beamcolumns with out-of-plane unbraced lengths large enough
such that the resistance is governed by Eq. (6). Therefore,
for members subjected to in-plane loading and where
inelastic redistribution is allowed, the strength either is
given or is accurately approximated by Eqs. (5) with n
= p and with n = ni calculated using the actual member
unsupported length in the plane of bending. Furthermore,
when r is less than 0.1 eL, ni = y is an acceptable
approximation. The approximation ni = y also can be
used for general I-section members as long as an appropriate
out-of-straightness is included in the second-order
analysis. Members that do not satisfy all the requirements
necessary to ensure ductile response must be designed
elastically in the manner discussed in Section 2.2.1.
The above type of analysis and design is referred to in
this paper as the
method. In
this method, the elastic stiffness of the structure is reduced
to account for distributed yielding effects neglected in the
elastic-plastic hinge idealization. Also, a nominal initial
out-of-plumbness is included to account for geometric
imperfection effects. These devices eliminate the need to
calculate and apply column effective lengths in the
context of inelastic design, as long as the second-order
effects are captured in the elastic-plastic hinge analysis.
Furthermore, these devices allow the engineer to take
advantage of second-order elastic-plastic hinge analysis
software. This kind of software is becoming more and
more available in engineering practice.
Other limits are possible within which the engineer may
be permitted to perform a classical rigid-plastic analysis
and design, e.g., see King (2001) and Davies and Brown
(1996). These limits are not addressed in this study.
M

direct elastic-plastic hinge

3. Illustrative Examples
This section provides a number of basic examples
aimed at illustrating the relative merits of the various
analysis and design procedures outlined in Section 2. The
first example addresses the strength predictions for one of
the major-axis bending cases of the Fig. 3 cantilever
beam-columns. Since this is a statically determinate
structure, inelastic analysis and design does not provide
any advantage. This example is representative of numerous
nonredundant stability critical benchmark problems
considered in the development of the direct analysis
method (Deierlein 2003; Surovek-Maleck and White
2003 & 2004). The second example is a nonredundant
portal frame previously posed by LeMessurier (1977).
This frame exhibits significant sway under gravity loadings,
due to lack of symmetry of its geometry. Also, the beam
governs its maximum resistance rather than the columns.

The third example is a hypothetical industrial building

frame that meets representative service drift requirements.
However, the behavior of the lateral load resisting system
is sensitive to stability considerations to the extent that
the structures limit load is reached when the first column
plastic hinge forms in the direct elastic-plastic hinge
analysis. Lastly, a fourth example is provided in which the
lateral load resisting columns are subjected to substantial
non-sway gravity load moments. This frame exhibits
some reserve strength beyond the first plastic hinge for its
critical load combination. All of these examples are
individual members or single-story rectangular structures
with either fully-restrained (FR) or simple connections.
The reader is referred to Maleck (2001), Martinez-Garcia
(2002), Deierlein (2003), Surovek
(2005), White
. (2005a & b) and White and Kim (2006) for consideration
of a broader range of structure types including multi-story
frames, braced and combined framing systems, trussed
framing, PR frames, gabled frames, and frames with slender
element section members, singly-symmetric members,
nonprismatic members, and/or members with nonconstant axial load along their lengths.
In all the following examples, the appropriate stiffness
reduction factor in the direct analysis is 0.8. That is, y
is always less than 0.5, and therefore b = 1.0. The
distributed plasticity analysis is conducted using a factor
of 0.9 on the elastic stiffness and the strength y in all
cases. The axial force is close to or smaller than 0.1 eL
in all the above problems. Therefore, the column out-ofstraightness is neglected and, to illustrate the validity of
this option, ni is taken equal to y in the direct analysis
solutions. A nominal column out-of-straightness is included
in the direction of the member curvature due to the applied
loadings in all of the distributed plasticity solutions, to
demonstrate that the direct analysis solutions give accurate
predictions for these cases with out-of-straightness
neglected. All of the direct and distributed plasticity
analyses are conducted using an out-of-plumbness of
o = 0.002 , specified in the direction of the drift under
the applied loads. For the effective length method, a
second-order elastic analysis is conducted using the nominal
elastic stiffness and the initially perfect geometry.
The second-order elastic and distributed plasticity
solutions are conducted using the GT-Sabre software system
(Chang 2005). A flexibility-based element (Alemdar and
White 2005), which is based on an equilibrium-based
distribution of moments and a correspondingly accurate
representation of inelastic curvatures (including the influence
of transverse distributed loads), is used for the beams.
The mixed element developed by Alemdar and White
(2005), which is capable of accurate modeling of the
second-order moments and inelastic curvatures, is used
for the beam-columns. A small modulus of 0.001 is
assumed for the yielded material. The direct elasticplastic hinge solutions are conducted using the Mastan2
software system (McGuire
., 2000). For major-axis
bending, the Mastan2 yield surface, with axial force and
et al.

et

al

P/P

et al

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

81

Ratio of the effective moment of inertia Ie at

the distributed plasticity limit load to the elastic crosssection moment of inertia I, example cantilever beamcolumn subjected to major-axis bending.
Figure 7.

Comparison of effective length and direct analysis

method beam-column strength interaction calculations to
distributed plasticity analysis, example cantilever beamcolumn subjected to major-axis bending, adapted from
White and Kim (2006).
Figure 6.

bending moment anchor points of 0.9Py and 0.9Mp, is

slightly more liberal than the yield surface corresponding
to Eqs. (5) with Pc = 0.9Py and Mc = 0.9Mp.
At the strength design loads, / , which is
equivalent to the sidesway amplifier B in AISC (2005a),
is larger than 1.5 in all of these examples. Therefore,
AISC (2005a) disallows the use of the effective length
method for all of these structures. The effective length
method is applied in the examples to illustrate the
implications of its use for these types of frames.
2nd

1st

3.1. Cantilever column

Figure 6 compares the results predicted by the direct

analysis and the effective length methods to the results
from distributed plasticity analysis for the Fig. 3 beamcolumn subjected to major-axis bending and the specific
loading of H = 0.01P. The direct (elastic) and direct
elastic-plastic hinge analysis methods are one and the
same up to the maximum strength limit for this example,
since the structure is statically determinate. Figure 6
shows the force-point traces at the column base (i.e., the
variation of the member base moment and axial force for
increasing levels of the applied loads) determined by the
direct analysis, effective length and distributed plasticity
analysis solutions.
Figure 6 also shows the factored beam-column strength
curves for the effective length and the direct analysis
methods (with = = 0.9). In this problem, the member
in-plane strength governs in both the direct analysis and
the effective length solutions if the member is braced at
its top and bottom in the out-of-plane direction, K = 0.7 is
used for the calculation of Pno, and Eq. (6) is used for the
out-of-plane strength assessment. Therefore, Fig. 6 focuses
only on the in-plane resistance. The direct analysis inplane strength curve is based on cPni = 0.9Py (although P
is slightly larger than 0.1PeL). The corresponding axial
strength for the effective length method is cPni = 0.9
c

(0.626Py) based on K = 2, or KL/rx = 80. The anchor point

on the horizontal axis is bMn = 0.9Mp for both of the
above strength curves.
The direct analysis provides a reasonably accurate
estimate of the distributed plasticity internal force and
moment up to the predicted maximum resistance. The
force point trace from the distributed plasticity analysis
indicates a larger moment than determined by the secondorder elastic analysis of the nominally-elastic initiallyperfect structure by the effective length method. This is
due to the stiffness reduction factor of c = 0.9 in the
distributed plasticity analysis as well as the effects of
distributed yielding as the maximum strength limit is
approached. Figure 7 shows the ratio of the effective
moment of inertia Ie for the beam-column (i.e., the
moment of inertia of its elastic core) to the elastic crosssection moment of inertia at the distributed plasticity
analysis limit load. The member experiences significant
yielding at the strength limit, but a full plastic hinge has
not yet formed at its base. The maximum value of P on
the force-point trace from the distributed plasticity
analysis corresponds to the limit load in the refined
inelastic solution. Correspondingly, the design strength
for the effective length and the direct analysis methods is
defined by the intersection of their force-point traces and
the corresponding beam-column strength curves. Both the
effective length and direct analysis provisions are calibrated
such that these intersection points give an accurate to
conservative estimate of the actual maximum strength
represented by the distributed plasticity solution.
The direct analysis method accounts for all the key
attributes that influence the in-plane system stability effects
directly within the analysis. Hence, its force point trace
may be compared against the cross-section strength limit
given by Eqs. (5) with Pni = Py and Mn = Mp. Conversely,
the effective length method accounts for the in-plane
system stability effects by reducing the member axial
strength Pni via the effective length KL, or the elastic
buckling stress Fei, obtained implicitly or explicitly from
an appropriately configured buckling analysis.
Table 2 summarizes the results at the calculated strength

82

Donald W. White et al.

Summary of calculated design strengths, cantilever beam-column example

Pmax (kN)
Mmax (kN-m)
Mmax/HL
Pmax/Pmax
Traditional effective length
1590
119
1.68
0.96
Direct analysis
1730
217
2.79
1.05
Distributed plasticity analysis
1650
198
2.68
Table 2.

(Distributed Plasticity)

limits by each of the above methods. The ratios of the

base moments Mmax = HL + P( + o) to the primary
moment HL indicate the magnitude of the second-order
effects. The axial load at the direct analysis strength limit,
which is representative of the strength in terms of the
total applied load, is 5% higher than that obtained from
the distributed plasticity analysis. Conversely, the axial
load at the effective length method beam-column strength
limit is 4% smaller than that obtained from the distributed
plasticity solution. Both of these estimates are within the
targeted upper bound of 5% unconservative error relative
to distributed plasticity analysis established in the original
development of the AISC LRFD beam-column strength
equations (ASCE 1997; Surovek-Maleck and White 2004).
The difference in the calculated internal moments is
much larger. This difference is expected since the effective
length approach compensates for the underestimation of
the actual moments by reducing the value of the axial
resistance term Pni, whereas the direct analysis method
imposes additional requirements on the analysis to obtain
an improved estimate of the actual internal moments.
This more accurate calculation of the internal moments
also influences the design of the restraining members and
their connections. For instance, in this example, the direct
analysis base moments are more representative of the
actual moments required at this position to support the
applied loads at the limit of the member resistance. In
other words, the direct analysis method provides a direct
calculation of the required strengths for all of the structural
components. Conversely, the effective length approach
generally needs supplementary rules for calculation of the
required component strengths. In AISC (2005a), these
supplementary rules are: (1) A minimum notional lateral
load is applied in gravity-only load combinations, and (2)
The effective length method is limited to frames having
2nd/1st < 1.5 (see Table 1).

3.2. LeMessuriers (1977) Example 3

The structure shown in Fig. 8 is a market shed deliberately

designed to have slender columns. LeMessurier (1977)
presents this frame as his Example 3. He explains that
column CD is on an open side and is slender for architectural
reasons. Column AB is selected to limit the first-order
drift under nominal wind to a maximum of 25.4 mm (1
in). The discussions below focus on checking the maximum
resistance of this frame by LRFD using the various
analysis and design procedures. The girder is assumed to
be adequately braced in the out-of-plane direction such
that Mn = Mp. Column AB is braced at its ends in the outof-plane direction. This bracing is sufficient such that its

Figure 8.

LeMessuriers (1977) Example 3.

Design load fraction versus the story drift for

LeMessuriers Example 3.

Figure 9.

resistance is governed by in-plane limit states. A live-todead load ratio of 3.0 is assumed for the LRFD
calculations. The loading 1.2 + 1.6Lr + 0.8W, with the
wind applied in the direction of the sway under the
gravity load, is the most critical of the ASCE 7-05 load
combinations for this structure. The following discussions
focus on the behavior of the frame under this load
combination.
Figure 9 shows the fraction of the design load versus
the drift /L from the different analysis methods, Fig. 10
shows the moment diagrams and the diagrams of the
effective moment of inertia Ie at the distributed plasticity
analysis limit load, Fig. 11 shows the deflected shape, the
location of the single plastic hinge and the design load
fraction at the formation of the plastic hinge from the
direct elastic-plastic hinge analysis, Fig. 12 shows the
applied fraction of the design load versus the girder
maximum bending moment from the different analysis
methods, and Fig. 13 shows the force-point trace at the

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

Diagrams of member moments and effective moment of inertia

load, LeMessuriers Example 3.
Figure 10.

Ie

83

at the distributed plasticity analysis limit

Deflected shape, location of plastic hinge, and

design load fraction at the formation of the plastic hinge
from direct elastic-plastic hinge analysis, LeMessuriers
Example 3.
Figure 11.

Force-point trace at the top of column AB

obtained from the different analysis-design methods,
LeMessuriers Example 3.
Figure

Applied fraction of the design load versus the

girder maximum bending moment obtained from the different
analysis-design methods, LeMessuriers Example 3.
Figure 12.

top of column AB from the different analysis methods.

One can observe that the direct elastic-plastic hinge
analysis provides a reasonable estimate of the distributed
plasticity solution. Due to the use of the smaller stiffness
reduction factor in the plastic hinge analysis (0.8 versus
0.9), the elastic deflections are slightly larger than those
predicted by the distributed plasticity analysis. However,
the sway deflection at the limit load is slightly smaller in
the plastic hinge analysis. The maximum resistances in all
of the analysis solutions correspond to the formation of a
plastic hinge in the girder. The second major slope

13.

discontinuity in the direct analysis load-deflection curve

(Fig. 9) occurs when a plastic hinge forms at the top of
column AB. The resulting overall shape of the direct
analysis load-deflection curve matches quite well with the
distributed plasticity analysis results.
Due to the smaller drift predicted in the second-order
elastic analysis by the effective length method, the
maximum girder moments are slightly smaller for a given
design load fraction. As a result, the effective length
method predicts a maximum strength of the structure (at
the formation of a plastic hinge with a resistance of 0.9Mp
in the girder) at 0.987 of the loading 1.2 + 1.6Lr + 0.8W.
The plastic hinge and distributed plasticity analysis methods
predict corresponding strength limits of 0.975 and 0.959
of the design loading. Although there is substantial
distributed yielding along the length of the girder at the
distributed plasticity strength limit (see Fig. 10), this
yielding has a small impact on the girder maximum
second-order moment.

84

Donald W. White et al.

Figure 14.

Malecks (2001) 11-bay industrial building frame.

3.3. Malecks (2001) Industrial Building Frame

Figure 14 shows an example 11-bay industrial building

frame developed by Maleck (2001) and studied further in
Martinez-Garcia (2002), Deierlein (2003) and SurovekMaleck and White (2004). Five gravity columns are
located on each side of the interior moment frame. In
practice, the exterior bays of this type of frame might be
designed with simply-supported joist girders, hence the
large number of gravity columns. Large gravity loads are
specified to simulate conditions in some single-story industrial
buildings, such as automobile plants (Springfield, 1991),
in which large equipment loads are located on the roof.
Significant - effects are transmitted to the moment
frame from the large number of gravity (leaner) columns.
However, the design lateral loadings from wind, etc. are
relatively small.
The same girder size (W27x84) is used throughout the
structure in Fig. 14. This size is selected to resist the
moments in the simply-supported exterior bays of the
frame. Although a 24 84 has sufficient strength to
withstand the design gravity loads, the deeper 27 84
profile is selected to reduce the lateral drift. The tops of
the two 10 49 interior columns are rigidly connected
to the middle 27 84 girder. The middle girder is
continuous over the tops of these columns, and the
columns are attached to the bottom of the girder by endplate connections. Base restraint is provided for the
lateral-load resisting columns to limit the drift under a
service loading of 1.0 + 0.5 + 0.7 to /400. The service
drift is calculated on the nominally-elastic geometricallyperfect frame, that is, no stiffness reduction or geometric
imperfections are applied for the service load analysis.
The strength design of the girders and the columns in
the above structure is governed by the LRFD gravity load
combination 1.2 + 1.6 . The ASCE7-05 required inclusion
of 0.8W with this combination is neglected to accentuate
the frame stability effects. The following discussions
focus on the frame behavior under this load combination.
The columns are again governed by the frame in-plane
stability limit states when Eq. (6) is used for the out-ofplane strength checks. The reader is referred to SurovekMaleck and White (2004) and to Kuchenbecker
.
(2004) for further details of the strength and service load
P

Lr

Lr

et al

Design load fraction versus the story drift for

Malecks frame.
Figure 15.

calculations.
Figures 15 through 19 present the same results as Figs.
9 through 13 for LeMessuriers (1977) Example 3
structure, but illustrate the responses for Malecks (2001)
frame. Figure 19 shows the force-point trace at the top of
the left-hand lateral load resisting column. This location
has the most critical combination of column axial force
and bending moment. Figure 18 shows the girder negative
bending moment versus the design load fraction just to
the left of the left-hand column. This is the location of the
largest girder bending moment.
Again, the direct elastic-plastic hinge analysis provides
a reasonable estimate of the distributed plasticity analysis
solution. Due to the smaller stiffness reduction factor in
the plastic hinge analysis (0.8 versus 0.9), the elastic
deflections are slightly larger than those predicted by the
distributed plasticity analysis. Also, the sway deflection
at the limit load is slightly larger in the plastic hinge
analysis. The maximum resistance in the plastic hinge
analysis (1.155 of the design load level) corresponds to
the formation of a plastic hinge at the top of the left-hand
column. Subsequently, plastic hinges form at the other
ends of the lateral load resisting columns within the postpeak range of the response at 1.115, 1.075 and 0.930 of
the design load level (see Fig. 17). If the example frame
were not so sensitive to stability effects, there would be

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

Diagrams of member moments

load, Malecks frame, 1.2 + 1.6L .
Figure 16.

and effective moment of inertia

Ie

85

at the distributed plasticity analysis limit

Deflected shape, plastic hinge locations, and

design load fraction at the formation of the plastic hinges
predicted by direct elastic-plastic hinge analysis, Malecks
frame, 1.2 + 1.6L .
Figure

17.

some reserve strength associated with the redistribution

of the left-hand column moments from its top to its base,
as well as the redistribution of moments from the lefthand column to the right-hand column. Of course, elastic
analysis and design discounts this reserve system strength.
The distributed plasticity analysis predicts a similar
response; however, at the maximum load level (at 1.158
of the design loading), there are no fully-formed plastic
hinges in the structure (see Fig. 16). The limit load is
reached due to a combination of P- effects along with
the progressive softening of the columns and girders due
to the spread of plasticity through their cross-sections and
along their lengths. At the distributed plasticity analysis
limit load, the girders are significantly yielded on the lefthand side of each of the columns (see Fig. 16). However,
this yielding is highly localized due to the moment
gradient in the girders at these positions.
Again, the second-order elastic analysis of the perfect
nominally-elastic structure by the effective length method
results in significantly smaller overall drift at the predicted
strength limit (see Fig. 15). The corresponding forcepoint trace for the columns intersects the column in-plane
strength curve at a design load fraction of 0.946. The
effective length method beam-column strength for this
frame is based on a column effective length factor of K =

Applied fraction of the design load versus the

maximum girder bending moment obtained from the
different analysis-design methods, Malecks frame, 1.2 +
1.6L .
Figure 18.

2.12 using Eq. (C-C2-5) of the AISC (2005a) Commentary.

The conservatism of this strength check is due to two
causes: (1) a substantial fraction of the column internal
moments are due to non-sway gravity loading and (2) the
AISC (2005a) effective length method requires the inclusion
of a minimum notional lateral load with gravity-only load
combinations. The original calibration of the effective
length method in AISC LRFD (1986) did not include any
sway frames that were subjected to gravity loads causing
non-sway member moments. The loadings in the original
benchmark solutions were all applied directly at the
beam-column joints such that the non-sway moments
were zero (ASCE 1997; Surovek-Maleck and White
2003; Deierlein 2003). The above conservatism of the
AISC (2005a) effective length method strength checks,

86

Donald W. White et al.

AISC Specification Committee Design Problem

13 (DP-13).
Figure 20.

Force-point trace at the top of left-hand lateralload resisting column obtained from the different analysisdesign methods, Malecks frame, 1.2 + 1.6L .
Figure 19.

due to the significant non-sway column moments, is

much larger in the next example (see Section 3.4). Also,
the original AISC (2005a) beam-column strengths were
calibrated to give accurate to conservative predictions of
the results from distributed plasticity analysis without the
inclusion of any notional lateral loads. However, even if
the beam-column strength checks in the example frame
are conducted with zero notional lateral load, a conservative
estimate of 1.119 of the design loading still is obtained
for the maximum frame resistance.
Although the above effective length method checks are
conservative for the in-plane strength of the columns of
Malecks frame, the moments for checking the out-ofplane strength of these members as well as the design of
the end-plate connections and column bases are dramatically
underestimated. Figure 19 shows that the maximum
moment at the top of the left-hand column is only 54.6
kN-m when the beam-column interaction equation is
intersected by the force-point trace of the effective length
method using zero notional load. However, this moment
is 186.9 kN-m at the distributed plasticity analysis limit
load (242% larger). The minimum lateral load requirement
in the AISC (2005a) effective length method increases the
predicted value to 100 kN-m when the beam-column
strength condition is reached. This moment is also
significantly smaller than the moment predicted by the
distributed plasticity analysis. However, in this case, the
conservatism of the in-plane strength check prevents the
beam-columns from reaching a state in which their inplane bending moments increase substantially with only a
minor increase in the applied load. The force-point trace
for the AISC (2005a) effective length method is a reasonable
approximation of the result predicted by the distributed
plasticity analysis up to the point at which its in-plane
beam-column strength condition is reached. The direct
elastic-plastic hinge analysis gives a conservative estimate
of the connection moment at the top of the left-hand
column of 231.7 kN-m at its limit load (24% larger than
the value from the distributed plasticity analysis). However,

this method does a good job of estimating the maximum

overall capacity of the frame. This is evidenced by the
similarity of the peak ordinate values for the direct and
distributed plasticity analysis methods in Figs. 15 and 19.

3.4. AISC Specification Committee (2004) Design

Problem 13 (DP-13)

Figure 20 shows a single story rectangular frame posed

as one of many design problems during the final Specification
Committee validation and checking of the AISC (2005a)
provisions. The interior columns in this structure are all
leaner columns. All the beam-to-column connections are
simple except for the connections to the exterior columns,
which are fully-restrained. The exterior columns are
subjected to relatively light axial loads, whereas they
experience substantial gravity load moments in addition
to the wind load moments. Also, the frame has significant
second-order effects, i.e., amplification of the member
sidesway internal bending moments. The columns are
braced in the out-of-plane direction at their base and at
the roof height. Simple base conditions are assumed in
both the in-plane and out-of-plane directions, and simple
connections are assumed in the out-of-plane direction at
the column tops. This problem is considered using the
following LRFD load combinations:
Load Case 1 (LC1): 1.2D + 1.6S
Load Case 2 (LC2): 1.2D + 0.5S + 1.6W
Figures 21 and 22 show the applied fraction of the
design loads versus the story drift by the distributed
plasticity analysis, direct elastic-plastic hinge analysis and
effective length methods for these load combinations. The
distributed plasticity analysis gives a maximum in-plane
capacity of 1.13 times the factored design load level for
LC1 and 1.06 of the factored design load level for LC2.
Each of the figures shows two curves determined from a
direct elastic-plastic hinge analysis, one in which a
stiffness reduction factor (SRF) of only 0.9 (the same as
the value used in the distributed plasticity analysis) is
employed, and the other in which a SRF of 0.8 (derived
from the AISC (2005a) direct analysis provisions) is
employed. For LC2, the direct analysis matches the

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

Design load fraction versus the story drift for

DP-13 load case 1 (LC1).
Figure 21.

Design load fraction versus the story drift for

DP-13 load case 2 (LC2).
Figure 22.

distributed plasticity solution closely when the larger SRF

of 0.9 is employed. However, for LC1, the direct analysis
over-predicts the maximum load capacity by 12% when
the SRF is taken as 0.9. Use of the recommended SRF of
0.8 results in a predicted maximum load capacity of 1.19
of the design load level for LC1 (5% larger than the
distributed plasticity analysis solution, which is equal to
the acceptable tolerance on the unconservative error
established in the original development of the LRFD
beam-column strength interaction equations (ASCE 1997;

87

Surovek-Maleck and White 2004)).

The load-drift curves for the effective length method
analysis indicate a stiffer load deflection response. These
curves end at the load level at which the right-hand
exterior beam-column force-point trace intersects the
effective length method strength envelope. As indicated
in the previous example, the original AISC effective length
method (no notional minimum lateral load) tends to be
conservative for beam-columns in which a large fraction
of the bending moments are associated with non-sway
gravity loading. To isolate the conservatism associated with
the original effective length calculations, the effective
length solution for LC1 is conducted without including
the minimum notional load required by AISC (2005a).
Since LC2 includes a nonzero lateral load, there is no
additional notional lateral load requirement for this load
case in AISC (2005a). The original effective length
procedure, which does not include any notional lateral
load, suggests that only 0.78 and 0.82 of the design loads
can be applied to the DP-13 frame for LC1 and LC2
respectively. That is, the effective length method gives a
capacity of only 0.78/1.13 = 0.69 and 0.82/1.06 = 0.77 of
the in-plane capacity from the distributed plasticity
analyses for LC1 and LC2. If the AISC (2005a) minimum
lateral load is included for LC1, the predicted capacity is
reduced only a slight additional amount to 0.76 of the
design load level.
Table 3 summarizes the fractions of the design loads
giving a value of 1.0 for the in-plane strength check from
Eqs. (5) for the AISC (2005a) direct (elastic) and effective
length methods, as well as the maximum capacities
predicted by the direct elastic-plastic hinge analysis using
Mastan2, and the distributed plasticity analysis using GTSabre. The direct (elastic) analysis method gives a slightly
smaller maximum strength than the direct elastic-plastic
hinge analysis using Mastan2 for LC1 (1.17 versus 1.19).
This is due to the slightly more liberal cross-section plastic
hinge strength assumed in Mastan2 compared to Eqs. (5).
For LC2, the direct (elastic) analysis force-point trace
reaches Eqs. (5) in the right-hand exterior column at a
design load fraction of 0.95. This value is comparable to
the first plastic hinge formation at the top of the righthand column at 0.96 of the design loading in Mastan2.
However, as indicated in Table 3 and illustrated by the

Applied fraction of the design loads giving a value of 1.0 for the in-plane strength check from Eqs. (5) for the
AISC (2005a) direct (elastic) and effective length methods with and without minimum lateral loads, maximum capacities
predicted by direct elastic-plastic hinge analysis using Mastan2, and maximum capacities predicted by distributed plasticity
analysis using GT-Sabre, DP-13 LC1
LC1
LC2
Direct (elastic) analysis
1.17
0.95
AISC (2005a) effective length
0.76
0.82
Effective length with zero N
0.78
0.82
Direct elastic-plastic hinge analysis first-hinge strength, Mastan2
1.19
0.96
Direct elastic-plastic hinge analysis limit load, Mastan2
1.19
1.02
Distributed plasticity analysis
1.13
1.06
Table 3.

88

Donald W. White et al.

Plastic hinges and deflected shape from direct elastic-plastic hinge analysis and effective moment of inertia
at the distributed plasticity analysis limit load, DP-13 LC2.

Ie

Plastic hinges and deflected shape from direct elastic-plastic hinge analysis and effective moment of inertia
at the distributed plasticity analysis limit load, DP-13 LC1.

Ie

Figure 23.

Figure 24.

direct elastic-plastic hinge load-drift curve in Fig. 22, the

maximum capacity of the frame is not reached under the
LC2 loading until a design load fraction of 1.02 by the
direct elastic-plastic hinge analysis (versus 1.06 by the
distributed plasticity solution).
Figure 23 illustrates the behavior at the strength limit
predicted by the direct elastic-plastic hinge and distributed
plasticity solutions for this load combination. The effective
moment of inertia Ie is reduced to zero at the top of the
right-hand column at the limit load in the distributed
plasticity analysis. Also, the bending rigidities are
significantly reduced in the positive bending region of the
left-most girder and at the top of the left-hand column at
this load level. However, significant portions of these
cross-sections are still elastic. The direct-elastic plastic
hinge analysis predicts the formation of a sway mechanism,
by plastic hinging at the tops of each of the exterior
columns, at its predicted capacity of 1.02 of LC2.
Figure 24 shows the corresponding predictions for
LC1. In this case, no plastic hinges have fully formed at
any location in the frame (i.e, Ie/I > 0 everywhere) at the
predicted capacity of 1.13 of the design loading in the
distributed plasticity analysis. The direct elastic-plastic
hinge solution reaches its limit load (at 1.19 of the design
loading) when a hinge forms at the top of the right-hand
column.
The key reason for the underestimation of the load
capacities by the effective length method is illustrated in

Force-point traces for the right-hand beamcolumn by the original effective length method, the direct
analysis method and the distributed plasticity analysis
method versus the effective length and direct analysis
strength curves, DP-13 LC1.
Figure

25.

Fig. 25. This figure compares the force-point trace at the

top of the right-hand column predicted by the original
effective length method (no notional minimum lateral
load) and the AISC (2005a) direct (elastic) analysis
method to the corresponding force-point trace from the
distributed plasticity analysis for LC1. The AISC (2005a)
strength envelopes (Eqs. (5)) are shown as dashed curves
in the figure. The significant in-plane stability effects in

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

this example result in a P of only 0.098P for LC1.

The corresponding effective length factor for the righthand column is 5.02 using a rigorous sidesway buckling
analysis (comparable values are obtained using the
equations in the AISC (2005a) Commentary, as long as
the leaner column effects and the effect of the simple
connections at the inside ends of the exterior girders are
correctly incorporated).
The direct analysis method gives a substantially larger
estimate of the in-plane resistance because it focuses on
a more realistic estimate of the internal moments and the
corresponding member resistances. In determining the
anchor point P for its beam-column strength envelope,
the effective length method overemphasizes the response
of the structure to idealized loads causing uniform axial
compression in all of the columns. The physical strength
of the DP-13 frame is dominated by its amplified internal
moments, a large fraction of which is not related to the
sidesway of the structure. The limit of the DP-13 frame
resistance is tied essentially to the flexural resistance of
the right-hand beam-column and the amplified internal
moments in this member, not by a column failure under
concentrically-applied axial loads.
When the effective length method is applied to the DP13 frame, the design resistance is governed by the inplane strength of the leeward beam-column (Eqs. (5)) for
both LC1 and LC2 (rather than the out-of-plane strength
as represented by Eq. (6)). However, the out-of-plane
strength based on Eq. (6) governs slightly relative to the
in-plane strength in both the distributed plasticity analysis
as well as the direct analysis methods. Eq. (6) is intersected
by the direct (elastic) analysis force-point trace for the
right-hand column at 1.16 and 0.94 of LC1 and LC2
respectively. The right-hand column does not satisfy the
AISC (2005a) unbraced length requirements for inelastic
design using braces only at its ends. Additional out-of-plane
bracing would be required to satisfy these requirements
such that the inelastic reserve strength illustrated by Fig.
22 can be utilized based on the AISC requirements.
n(KL)

n(KL)

4. Conclusions

This paper explains the AISC (2005a) stability analysis

and design provisions with an emphasis on their relationship
to refined second-order inelastic analysis methods and on
how they can facilitate the use of second-order inelastic
analysis in practical design.
First, an appropriate application of second-order distributed
plasticity analysis is discussed for refined assessment of
adequately braced compact I-section members and structural
systems. The base AISC (2005a) beam-column strength
interaction equations have been developed in part by
calibration against the results from refined inelastic
analyses of this type. The paper recommends that for
distributed plasticity analysis, both the material elastic
stiffnesses (E) and strengths (F ) should be reduced up
front by the factors b = c = 0.9. This is necessary to
y

89

obtain the same result as determined by analyzing a

member or structure with nominal E and F and
subsequently factoring the abscissa and ordinate of the
resulting strength curves by and . The appropriate
nominal residual stress, geometric imperfection, and
material stress-strain idealizations for use of distributed
plasticity analysis in a design context are discussed. Also,
a simple limit on the magnitude of the axial force is
provided within which member out-of-straightness may
be neglected in a distributed plasticity analysis.
Secondly, three specific elastic methods of analysis and
design detailed in AISC (2005a) are discussed in the
context of the above refined inelastic analysis and design
procedure. These are the direct analysis, the effective
length and the first-order analysis methods. The new
AISC (2005a) direct analysis method provides a better
approximation of the results from refined distributed
plasticity analysis. By improving the accuracy of the
analysis, the direct analysis method allows for greater
simplicity in the design checks by eliminating the need
for effective length factors. Also, a more accurate
characterization of the overall strength is attained in
general.
Thirdly, the paper presents a basic extension of the
direct (elastic) analysis method that satisfies the intent of
the Specification and captures the beneficial effects of
inelastic redistribution from compact adequately-braced
members after the individual member resistances are
reached. This extension, termed the direct elastic-plastic
hinge analysis method, is obtained simply by approximating
the response of these types of members by an elasticperfectly plastic hinge model at the Specification limit of
the member resistance.
Lastly, four examples are provided that illustrate key
characteristics of each of the above analysis and design
methods. The direct elastic and elastic-plastic hinge analysis
methods provide accurate elastic and inelastic approximations
of the refined distributed plasticity solutions. The effective
length analysis and design method is shown to give
accurate to conservative solutions for the in-plane beamcolumn resistances in all of the example cases. However,
the effective length method significantly underestimates
the internal moments that the beam, connection, and outof-plane beam-column resistances must accommodate in
certain cases. Furthermore, the effective length method is
shown to give conservative solutions for beam-columns
in sway frames where a large percentage of the moments
are due to non-sway gravity loading. AISC (2005a)
requires the use of a notional minimum lateral load in
gravity-only load combinations, and limits the application
of the effective length method to frames having 2nd/1st
< 1.5 to control these errors. The AISC (2005a) direct
elastic analysis method is generally applicable to all types
of frame structures. The direct elastic-plastic hinge method
is a useful extension of this method that facilitates
simplified inelastic analysis and design.
y

90

Donald W. White et al.

Acknowledgments
Many of the concepts discussed in this paper have
benefited greatly from discussions with the members of
the AISC Technical Committee 10 (TC10), the Structural
Stability Research Council (SSRC) Task Group 4 on
Frames, and the former SSRC Task Group 29 on Inelastic
Analysis for Frame Design. Professor J. Yura of the
University of Texas at Austin, Dr. S. Nair of Teng and
Associates, Inc. and Prof. G.G. Deierlien of Stanford
University are thanked for their guidance of the TC10
efforts toward the implementation of the AISC (2005a)
provisions. Various portions of this work were funded by
the National Science Foundation, the Georgia Institute of
Technology, the American Society of Civil Engineers,
and the Metal Building Manufacturers Association. The
funding by these organizations is gratefully acknowledged.
The opinions, findings and conclusions expressed in this
paper are the authors and do not necessarily reflect the
views of the above individuals, groups and organizations.

References
AISC (2005a). Specification for Structural Steel Buildings,
American Institute of Steel Construction, Chicago, IL.
AISC (2005b). Code of Standard Practice for Steel Buildings
and Bridges, American Institute of Steel Construction,
Inc., Chicago, IL.
AISC (1999). Load and Resistance Factor Design
Specification for Structural Steel Buildings, American
Institute of Steel Construction, Chicago, IL.
AISC (1989). Specification for Structural Steel Buildings:
Allowable Stress Design and Plastic Design, 9 Ed.,
American Institute of Steel Construction, Chicago, IL.
AISC (1986). Load and Resistance Factor Design
Specification for Structural Steel Buildings, American
Institute of Steel Construction, Chicago, IL.
AISC (1969). Specification for the Design, Fabrication and
Erection of Structural Steel for Buildings, American
Institute of Steel Construction, Chicago, IL.
AISC (1961). Specification for the Design, Fabrication and
Erection of Structural Steel for Buildings, American
Institute of Steel Construction, New York, NY.
Alemdar, B.N. and White, D.W (2005), Displacement,
Flexibility and Mixed Beam-Column Finite-Element
Formulations for Distributed Plasticity Analysis, Journal
of Structural Engineering, ASCE, 131(12), pp. 18111819.
ASCE (2005). Minimum Design Loads for Buildings and
Other Structures, SEI/ASCE 7-05, ASCE, Reston, VA.
ASCE (1997). Effective Length and Notional Load Approaches
for Assessing Frame Stability: Implications for American
Steel Design, Task Committee on Effective Length,
Technical Committee on Load and Resistance Factor
Design, Structural Engineering Institute, American
Society of Civil Engineers, pp. 442.
Battini, J.-M. and Pacoste, C. (2002). Plastic Instability of
Beam Structures Using Co-rotational Elements, Computer
Methods in Applied Mechanics and Engineering, 191, pp.
th

5811-5831.
CEN (2003). Eurocode 3: Design of Steel Structures - Part
1-1: General Rules and Rules for Buildings, Final Draft
prEN 1993-1-1:2003 E, European Committee for
Standardization.
Chang, C.-J. (2005). GT-Sabre User Manual, School of Civil
and Environmental Engineering, Georgia Institute of
Technology, Atlanta, GA.
Davies, J.M. and Brown, B.A. (1996). Plastic Design to BS
5950, The Steel Construction Institute, Blackwell
Science, U.K., pp. 326.
Deierlein, G. (2003). Background and Illustrative Examples
on Proposed Direct Analysis Method for Stablity Design
of Moment Frames, Report on behalf of AISC TC10,
July 13 2003, pp. 17.
Deierlein, G. (2004). Stable Improvements: Direct Analysis
Method for Stability Design of Steel-Framed Buildings,
Structural Engineer, November, pp. 24-28.
Galambos, T.V. and Ketter, R.L. (1959). Columns Under
Combined Bending and Thrust, Journal of the
Engineering Mechanics Division, ASCE, 85(EM2), pp.
135-152.
Izzudin, B.A., and Smith, D.L, (1996) Large-Displacement
Analysis of Elasto-plastic Thin-Walled Frames.
I:Formulation and Implementation, Journal of structural
Engineering, Vol. 122, No. 8, pp. 905-914
King, C. (2001). In-Plane Stability of Portal Frames to BS
5950-1:2000, SCI Publication P292, The Steel Construction
Institute, Berkshire, U.K., pp. 213.
Kuchenbecker, G.H., White, D.W. and Surovek-Maleck,
A.E. (2004). Simplified Design of Building Frames
using First-Order Analysis and K = 1.0, Proceedings,
SSRC Annual Technical Sessions, April, pp. 20.
LeMessurier, W. J. (1977). A Practical Method of Second
Order Analysis. Part 2: Rigid Frames, Engineering
Journal, AISC, 13(4), pp. 89-96.
Maleck, A.E. and White, D.W. (2003). Direct Analysis
Approach for the Assessment of Frame Stability:
Verification Studies, Proceedings, SSRC Annual
Technical Sessions, pp. 18.
Maleck, A.E. (2001). Second-Order Inelastic and Modified
Elastic Analysis and Design Evaluation of Planar Steel
Frames, Ph.D. Dissertation, Georgia Institute of
Technology, pp. 579.
Martinez-Garcia, J.M. (2002). Benchmark Studies to
Evaluate New Provisions for Frame Stability Using
Second-Order Analysis, M.S. Thesis, School of Civil
Engineering, Bucknell Univ., pp. 241.
McGuire, W. (1995). Inelastic Analysis and Design in
Steel, A Critique, Restructuring America and Beyond,
Proceedings of Structures Congress XIII, M. Sanayei
(ed.), ASCE, pp. 1829-1832.
McGuire, W., Gallagher, R.H. and Ziemian, R.D. (2000).
Matrix Structural Analysis, with Mastan2, Wiley, New
York.
Nair, R.S. (2005a). Stability and Analysis Provisions of the
2005 AISC Specification for Steel Buildings,
Proceedings, Structures Congress 2005, ASCE, pp. 3.
Nair, R.S. (2005b), Stability and Analysis, Modern Steel
Construction, September 2005.
Nukala, P.K.V.V. and White, D.W. (2004). A Mixed Finite

Stability Analysis and Design of Steel Building Frames: The AISC (2005) Specification and Beyond

Element Formulation for Three-Dimensional Nonlinear

Analysis of Frames, Computer Methods in Applied
Mechanics and Engineering, 193, pp. 2507-2545.
Pi, Y.L. and Trahair N.S. (1994), Nonlinear Inelastic
Analysis of Steel Beam-columns Theory., J. Struct.
Engrg., ASCE, 120(7), pp. 2041-2061.
SAA (1990), Steel Structures, AS4100-1990, Standards
Association of Australia, Australian Institute of Steel
Construction, Sydney, Australia.
Salmon, C.G. and Johnson, J.E. (1996). Steel Structures,
Design and Behavior, 4th Ed., Prentice Hall, NJ, pp. 1024.
Springfield, J. (1991). Limits on Second-Order Elastic
Analysis, Proceedings, Annual Technical Session,
Structural Stability Research Council, Chicago, IL, 89-99.
Surovek-Maleck, A.E., White, D.W. and Leon, R.T. (2005).
Direct Analysis for Design Evaluation of PartiallyRestrained Steel Framing Systems, Journal of Structural
Engineering, ASCE, to appear.
Surovek-Maleck A.E. and White, D.W. (2004). Alternative
Approaches for Elastic Analysis and Design of Steel
Frames. I: Overview, Journal of Structural Engineering,
ASCE, 130(8), pp. 1186-1196.
Surovek-Maleck, A. and White, D.W. (2003). Direct
Analysis Approach for the Assessment of Frame
Stability: Verification Studies, Proceedings, SSRC
Annual Technical Sessions, pp. 18.
Teh, L., and Clarke, M.J., (1998), Plastic-Zone Analysis of
3D Steel Frames Using Beam Elements, Journal of
Structural Engineering, Vol. 125, No. 11, pp. 1328-1337
White, D.W., Surovek-Maleck, A.E. and Kim, S.-C. (2005a).
Direct Analysis and Design Using Amplified First-Order
Analysis, Part 1 - Combined Braced and Gravity Framing

91

Systems, Engineering Journal, AISC, to appear.

White, D.W., Surovek-Maleck, A.E. and Chang, C.-J.
(2005b). Direct Analysis and Design Using Amplified
First-Order Analysis, Part 2 - Moment Frames and
General Rectangular Framing Systems, Engineering
Journal, AISC, to appear.
White, D.W. and Kim, Y.D. (2006). A Prototype
Application of the AISC (2005) Stability Analysis and
Design Provisions to Metal Building Structural Systems,
Report to Metal Building Manufacturers Association,
January 2006, pp. 156.
White, D.W., Surovek-Maleck, A.E. and Kim, S.-C. (2003).
Direct Analysis and Design Using Amplified First-Order
Analysis, Part 1 - Combined Braced and Gravity Framing
Systems, Structural Engineering, Mechanics and Materials
Report No. 42, School of Civil and Environmental
Engineering, Georgia Institute of Technology, Atlanta,
GA., pp. 34.
White, D.W. and Nukala, P.K.V.V. (1997). Recent
Advances in Methods for Inelastic Frame Analysis:
Implicatinos for Design and a Look Toward the Future,
Proceedings, National Steel Construction Conference,
American Institute of Steel Construction, pp. 43-1 to 4524.
White, D.W. and Chen, W.F (1993). Plastic Hinge Based
Methods for Advanced Analysis and Design of Steel
Frames - An Assessment of the State-of-the-Art, Structural

Stability Research Council, University of Missouri-Rolla,

Rolla, MO, pp. 299.
Ziemian, R.D., McGuire, W. and Deierlein, G.G. (1992).
Inelastic Limit States Design. Part I: Planar Frame
Studies, 118(9), pp. 2532-2549.