Sunteți pe pagina 1din 11

CSCE 2007 Annual General Meeting & Conference

Congrs annuel et assemble gnrale annuelle SCGC 2007

Yellowknife, Northwest Territories / Yellowknife, Territoires du nord-ouest


June 6-9, 2007 / 6 au 9 juin 2007

A Brief History of Beam-Column Design


I.J. MacPhedran, and G.Y. Grondin
Department of Civil & Environmental Engineering, University of Alberta, Edmonton, Alberta, Canada

Abstract: This paper follows the evolution of beam-column design in the Canadian steel construction
design standard S16, from simple strength considerations to the current strength and stability interaction
equations. Because the design of columns is intrinsically related to the method of analysis, there is also
consideration given to the advancement of analyses and stability theory over time, as well as changes in
the design environment and tools available to the designer. A brief review of American and European
design procedures is also included, and the current practise is examined. The general trend has been
toward a more complex model of the beam-column behaviour being adapted into design. As objective
design methods replace prescriptive design, there may be a return to earlier standards, where the
designer is given more freedom in choosing the design method.
While there has been considerable change, there are still aspects of the beam-column behaviour that are
not considered in the design standards. This reflects the lack of understanding we have of beam-column
behaviour. For example, beam-columns are still considered individually, rather than in the context of the
frame of which they are members. This discounts the interaction of members in the frame, a phenomenon
that is not yet well understood.

1. Introduction
Beam-columns are ubiqitous in structural design. Although the design of many structural members is
often simplified by assuming that they are either beams or columns, most structural elements are
subjected to a combination of flexure and axial load, making them effectively beam-columns.This paper
will cover the evolution of the design process for beam-columns in the Canadian design standard, and
recent changes adopted in other countries around the globe. As the analysis of the structure containing
the beam-columns is tightly coupled to the design of beam-columns, frame analysis considerations in the
design standards is included alongside the design equations.
Research in structural stability goes back to the 18th century, with work by investigators like Petrus van
Musschenbruch and Leonhard Euler on the elements of column buckling (Chen and Atsuta 1976).
Increased used of metal (iron and steel) construction in the 19th century pushed the need for work on
structural stability, including that of Johan Bauschinger, F. S. Jasinsky and others. The developments in
elastic stability continued into the early 20th century with work by Stephen Timoshenko, Ludwig Prandtl
and others (Timoshenko 1983). Theodore von Krmn laid the work for plastic instability of beamcolumns in the first decade of the 1900s (Chen and Atsuta 1976). Herbert Wagner's work on flexural
torsional buckling of open sections in 1929 and onward was the start of the investigations on the out of
plane instability in beam-columns.
Earlier papers have summarised the history of beam-column design. Massonnet (1976) wrote on the
research behind the design equations. Kennedy and Qureshi (1988) is of particular interest, in that it
covers the early history of beam-column design equations in the Canadian structural steel design

GC-147--1

standard S16 up to the S16.1-M84 standard. Sputo (1993) provides an American perspective up to a
similar time frame. However, those reviews are now some years old and do not include changes in the
Canadian standard, and in beam-column design elsewhere.
Beam-column design involves consideration of the combined effects of axial load and flexure, usually in
the form of an interaction equation, or rather inequality, compared to a limiting value. The first design
standards were based on allowable (or working) stress design (ASD or WSD) theory. In ASD, stresses in
the structure are calculated based on working, or service, loads. These working stresses are compared to
allowable stress limits that have been set by dividing a critical stress, such as yielding or buckling
stresses, by a factor of safety. As these stresses are all below the yield strength, this is referred to as an
elastic design philosophy. The current standards use a limit state design method that considers variability
of structural loads and resistances and allows consideration of yielding at the ultimate limit states.

2. The Canadian Design Standard


There have been seven releases of the standard for design of steel structures for working stress design
(published in 1924, 1930, 1940, 1954, 1961, 1965 and 1969) and six for limit state design (published in
1974, 1978, 1984, 1989, 1994 and 2001). Starting with the 1978 edition, the standards have used SI
(Systme International) units. Previous editions used the Imperial system of measure. While both
measurement systems will be used herein for comparison, the system used by the standard under
discussion will be given precedence to maintain consistency with the equations presented.
Standardisation in engineering design in Canada began with the formation of the Canadian Engineering
Standards Association (CESA)* in 1919 (McKenzie, 1928). In 1924 the CESA issued their 16th standard,
A16-1924 Standard Specification for Steel Structures in Buildings, the first design standard for steel
buildings in Canada, followed by an updated standard in 1930. These were both ASD guidelines that
compared the simple sum of the axial and flexural stresses with the allowable stress limits of 14 ksi
(95 MPa) (1924) and 15 ksi (105 MPa) (1930), the allowable stresses for compression in the respective
standards.
[1]

Fa P A Mc I

In the next standard, S16-1940, there was a recognition that the flexural and axial allowable stresses
represent different conditions and were given different values. The applied stresses were compared to the
respective allowable stresses as per Equation 2, directly giving a minimum design area, albeit with prior
knowledge of the radius of gyration and distance to extreme fibre. The allowable axial stress, p (Equation
3) and flexural stress f (Equation 4), are based on the slenderness ratio (l/r) of the member in
compression and the length of laterally unsupported compression flange, expressed as a ratio of the
flange width (l/b).
C My

p f r2

[2]

[3]

f y 1 .0015 l / r f e
f y 1 .0015 l / r f e
p 0.546

2
2

[4]

f 25,000 333 l / b 6 / 10 f y , for l / b 15

2E
2.86 x10 8
f y fe , fe

psi

l / r 2
l / r 2

In 1941, the first National Building Code of Canada was released. It carried similar design information to
the CESA standard, and used the S16-1940 interaction equation modified by dividing equation 2 by the
area, thus comparing the applied stresses (fa , fb) directly to the allowable stresses (Fa , Fb) (Equation 5).
*

CESA changed its name to Canadian Standards Association in 1944.

GC-147--2

The next release of the steel design standard, S16-1954, also used equation 5 (Kennedy and Qureshi,
1988).
[5]

fa
f
b 1.0
Fa Fb

Research on structural stability started in earnest at about this time. The Column Research Council
(CRC), originally formed in 1944, released the first edition of the The Guide to Design Criteria for Metal
Compression Members (now the Guide to Stability Design Criteria for Metal Structures) in 1960
(Johnston, 1981). The work in this guide was quickly adopted into design standards, and the preamble to
S16-1961 includes an acknowledgement of the CRC's work for new allowable limits for compressive
stresses. This column curve is defined by Equation 6, which gives the allowable stress in pounds per
square inch. The allowable stress for a very short column was 0.61fy, or 20 ksi (140 MPa) for the
commonly used steel grade of G40.4-1959 (fy of 33 ksi, or 230 MPa) at that time.
[6]

f y 145,000,000

(in psi)
Fa 20,000 70 KL / r
KL / r 2
33,000

The interaction equation in S16-1961 did not change from previous editions, but the designer was to
consider the interaction at various sections along the member. Points of lateral restraint were to be
checked with the full applied moments and with the maximum allowable compressive stress, and the
midpoint was to be checked with the allowable compressive stress considering slenderness and a
reduced moment, using Massonnets equivalent moment expression, Equation 7, (Massonnet, 1959) to
account for non-uniform moments. Figure 1 presents a graphical comparison of this equation with other
representations for equivalent moment factors. These factors were originally developed for lateral
torsional buckling, and then applied to P- effects.
[7]

M d 0.3 M 12 M 22 0.4 M 1 M 2

Plastic design was included for the first time in this edition, in view of requests received from users of the
Standard. (CSA 1961) The interaction equations for plastic design were somewhat more complex and
relied on a reduction in strength based on tabulated values. The design equations are presented below as
Equations 8 to 10. There are several other restrictions on the application of plastic design to ensure that a
plastic collapse mechanism can develop before other modes of failure. Member loads are effects caused
by factored loads, not specified loads.
[8]

2P
L

1 .0
Py 70 r

[9a]

Mo
P
1.18 1.18
Mp
Py

Double curvature, plastic hinges at both ends

[9b]

Mo
P
BG
Mp
Py

Pin-based column (B and G tabulated)

[9c]
[10]

P
Mo
P
1.0 K
J
Py
Mp
Py

Mo
1.0
Mp

Single curvature (K and J tabulated)

The CRC became the Structural Stability Research Council (SSRC) in 1976

GC-147--3

1.2
AISC
Analytical LTB
Massonnet (1961)
1

Austin (1965)
Salvadori (1989)

Equivalent Moment Factor

EuroCode 3 Theory P/Pcr = 0.5


EuroCode 3 Theory P/Pcr = 0.0

0.8

0.6

0.4

0.2

0
-1

-0.5

0.5

Moment Ratio (positive = double curvature)

Figure 1. Equivalent moment factors. (Numbers in parentheses are dates of first inclusion of that
expression in S16.)
The standard released in 1965, S16-1965, expressed the separate checks of S16-1961 as two equations.
Equation 11 is a strength check, expressly using the allowable stress of 0.60fy. The new Equation 12 is a
stability check, with the allowable axial stress following the curve described in Equation 6. The applied
bending stress (f 'b) is calculated from the maximum bending moment, modified by Austins equivalent
moment factor (Equation 13), and a moment magnifier is included to account for the P- effect. The
column curve for S16-1965 was a three segment curve of allowable compressive stresses, including
factors of safety of: 1.67 for yielding; 1.92 for elastic buckling; and an interpolation between 1.67 and 1.92
for intermediate columns. A factor of safety of 1.92 is also used to calculate the moment magnifier for
beam columns.
[11]

fa
f
b 1.0
0.60 Fy Fb

[12]

fa

Fa

[13]

f 'b

1.0 , where Fe'

f
Fb 1 a
F 'e

M
C m 0.6 0.4 1 0.4
M2

149,000

KL r 2

2E

1.92KL r 2

S16-1969 introduced biaxial bending into the interaction equations. The allowable bending stress
(Equation 14) was increased by considering both the St Venant and warping components of torsion,
whereas S16-1965 only considered the larger of the two (Adams, 1970a).

GC-147--4

[14]
[15]
[16]

12000
149000
Fbe
L d / Af
L / r 2
t

f by
fa
f
bx
1.0
0.60 Fy Fbx Fby

1
f a Cmx f bx x Cmy f by y

1.0 , where
Fa
Fbx
Fby

f
1 a

F
'e

for the bending axis under consideration

The interaction equations for plastic design had also changed in the 1969 standard. The three equations
used for checking beam-columns in plastically designed frames are shown as Equations 1719. For
columns in braced frames, the effective length factor, k, was taken as 1.0.
[17]
[18]
[19]

My
Mx

1 .0
M px M py
Pf
Py

0.85M x 0.85M y

1.0
M px
M py

Pf

1.67 A Fa

Cmx M x x Cmy M y y

1.0
M px
M py

A new design philosophy based on reliability theory, called limit states design, was first introduced in
S16.1-1974. This brought design beyond the elastic range. Limit states design gradually became adopted
by other standards in Canada and worldwide. A new column curve was also brought in at this time the
SSRC column curve 2. The interaction relationships changed to three equations, Equation 20 checks the
strength of the member, in a fashion similar to the earlier allowable stress equations (Equation 21 is an
alternative formulation specifically for Classes 1 and 2 I-shaped members), Equation 22 checks the
stability of the member, and Equation 23 provides a check to ensure that the member has sufficient
moment capacity. In Equation 22, the term in parentheses accounts for the P- effect, and the term
shown in the numerator is the equivalent uniform moment factor calculated using Equation 13 proposed
by Austin. The plastic design equations from the previous standards were used as the basis for Equation
21.
[20]
[21]
[22]

[23]

Cf
Cr

Cf
Cr

Cf
Cr
M

fx

M rx

fx

M rx

M ry

0.85M fx
M rx

fy

1 .0

0.60 M fy
M ry

x M fx

Cf
M rx 1
C ex
M fy

1.0
M ry

1.0

y M fy
Cf
M ry 1
C ey

1.0 , 0.6 0.4

M2
0 .4
M1

The relevant provisions of the 1974 standard would remain unchanged for the first two S.I. releases of the
S16 standard (S16.1-M77 and S16.1-M84). In the next release, S16.1-M89, several changes were made.
(Kennedy, et al. 1990). Only class 1 sections could be used for the full plastification equation, as work by
Kulak and Dawe (1984) indicated that Class 2 sections would experience local buckling prior to full plastic
moment developing. The P- effects have been combined into one factor, U1, a number that includes the

GC-147--5

equivalent moment factor, 1. As detailed below, each main interaction equation (Equation 24 or 25) acts
as three equations. Equation 24 is strictly for Class 1 I-shaped sections. Equation 25 covers other
sections. The moment strength check from the previous standard (Equation 23) is still in force.
[24]
[25]

Cf
Cr
Cf
Cr

0.85U 1x M fx
M rx
U 1x M fx
M rx

0.60U 1 y M fy
M ry

U 1 y M fy
M ry

1.0

1.0

These equations perform three functions by using different definitions for the resistances. Taking Cr and
Mr as the capacity for a zero length column and laterally braced beam ( S Fy or Z Fy), the equations
check cross-section strength. With Cr based on buckling in the plane of bending (or about the weaker axis
for biaxial bending) with an effective length factor of 1.0, this gives the in-plane strength. And, with Cr
based on weak axis buckling and taking lateral-torsional buckling into account for Mr, these check the outof-plane strength. The parameter U1 is a moment magnifier that compensates for the P- effect, and
combines the 1 term and the softening term from the previous formulation. This cannot be greater than
unity during the check for out-of-plane instability, as this would create a situation whereby the equivalent
moment factors (the terms) would combine to produce an unconservative result. Trahair (1986), in his
analysis of the S16.1-M84 design provisions, points out the problems in using both terms. Massonnet
(1959) was somewhat more specific the same factor was used for lateral torsional buckling and for
moment magnification, when they are actually different. Kennedy et al. (1990) provide a more detailed
description of the problems with U1 and resolutions to them.
The 1994 standard (S16.1-94) made no substantive changes to the interaction equations. However, the
column strength curve was changed from five polynomial segments to one exponential expression, (Loov,
1996) and Appendix D was added for calculation of torsional buckling effects due to axial load.
The changes in the current version of the S16 standard, S16-01, are detailed by Essa and Kennedy
(2000). The moment magnifier factor, U1, is now considered to be 1.0 for members in sway frames, as the
P- effects are much larger than the P- effects. Cross-section strength need not be checked for sway
frames, as instability will govern for these beam-columns. Class 2 sections are once again included with
Class 1 sections. Post-buckling analysis showed that the cross-section strength was adequate for these
sections to perform as fully plastic, even with local buckling of the beam-column. Torsional and flexuraltorsional buckling strength calculations were incorporated into the standards body, rather than being
relegated to an appendix. Class 4 design was included. The most direct change to the interaction
equations is that the weak axis interaction equation was modified to correct unconservative results with
members that were governed by moments. Equation 24 was modified to replace the weak axis moment
modifier of 0.6 with the variable (Equation 27) to produce Equation 26.
[26]

[27]

Cf
Cr

0.85U 1x M fx
M rx

U 1 y M fy
M ry

= 0.6 + 0.4 1.0 , where =

1.0 , where

L
ry

Fy
E

Table 1 summarises the provisions of 40 years of Canadian standards for steel beam-column interaction
equations. The general trend is towards a more complex representation of the behaviour of beamcolumns in the design equations, reflecting an increasingly better understanding of the complex behaviour
of beam-columns. This is exclusive of the other complexities in the standard that are, in general, also
increasing, as the standard also incorporates better knowledge of other behaviours. A notable exception
is the simplification of the column design curve in 1994.

GC-147--6

Table 1 Complexity of interaction equations for various CSA standards


Standard
Number of Equations Number of Parameters Number of Calculations
S16-1961
1
4
3
S16-1965
2
7
9
S16-1969
2
11
26
S16.1-1974
3
12
29
S16.1-M89
4
14
42
S16-01
4
16
44

3. Related Trends
The design of beam-columns cannot be considered limited to a strict evaluation of the interaction
equations listed above. The analysis of the structures in which the members perform must also be
considered jointly with the members, and the stability of the structure as a whole must be considered.
Complex analyses were available to designers early in the 20th century. Bleich (1952) outlines three
analysis techniques for frame analysis with buckling considered: slope deflection analysis with axial loads
taken into effect (P- effect); energy methods; and moment distribution with reduced stiffnesses to allow
for axial loads. Sputo (1993) shows similar recommendations had been made in the late 19 th century.
Starting with S16-1961, the standard strictly used the effective length concept for columns in sidesway
permitted or braced frames. This permitted the designer to simply increase or decrease the apparent
length of a member based on the restraint at its ends, using a nomograph approach to account for flexible
end restraint. Subsequent standards, such as S16-1969, allowed the designer to use either effective
length factors or explicit calculation of the P- effect (Adams, 1970b). A procedure for the calculation of
the P- effect was given in the non-mandatory Appendix J of S16-1974. A complete second order
elastic analysis (e.g. by computer) was also permitted, wherein the effect of displacement of the loads is
included in the final member forces. This was in effect for several releases of the standard.
Kennedy et al. (1990) outline several changes made in the 1989 standard. The use of effective lengths (K
factor) as a method to account for second order effects in sway permitted frames was removed in S16.1M89, and the consideration of sidesway load effects in analysis was made mandatory. A simple
procedure for calculation of these effects was included in the standard as a multiplier for moments due to
lateral loads, called U2. In addition, a minimum lateral load of 0.005 times the gravity load at a storey was
required as the shear at that storey. This pseudo lateral load was meant to account for cases of frames
that were not likely to sway on their own, such as frames with symmetric geometry and loading.
[28]

U2

1
C f f
1

V f h

There were no major changes in the 1994 standard to the frame analysis and stability. The value of U2
was limited to 1.4. If the value exceeded the limit, the designer had the option of either conducting a
second order elasto-plastic analysis on the frame, or the frame had to be made stiffer. The pseudo
lateral load of 0.005 gravity loads was renamed a notional load (Kennedy 1995). Notional loads are
intended to account for imperfections in the member, both in geometry (e.g. straightness, plumb) and
material (e.g. residual stresses which cause partial yielding). In S16-01, the notional load must be applied
to all load cases. This requirement permits the removal of the limit on U2, as using both the notional load
and the U1 factor balance the overly optimistic in-plane strength equation. Essa and Kennedy (2000) note:
the notional load should always be applied because it accounts for the stability effects by transforming
the sway-buckling problem into an in-plane bending strength problem. It is of note that Appendix J

GC-147--7

(iterative algorithm for the P- analysis) was removed in this release since by that time software capable
of performing a second order analysis was widely available.
Outside the self-constructed environment of the design standard, other trends were emerging in the
design of steel beam-columns. Significant developments in design tools have occurred, even over the
past few years. In the early 1900s, design tools consisted of pre-calculated tables and graphs, and the
slide rule was the arithmetic tool of choice. The first electronic desktop calculators became available
during the 1960s, and the first desktop computers in the 1970s. Now, designers have their own
computers, spreadsheets that simplify design tasks, and analysis software capable of solving complex
problems.

4. American and European Directions


Sputo (1993) provides a summary of the history of the AISC Specification provisions, which closely follow
the Canadian developments during the development of allowable stress design (ASD). A limit states
design philosophy, named Load and Resistance Factor Design (LRFD) was first adopted in 1986,
although both design approaches are used currently. The first joint LRFD / ASD Specification was
released 2005, and contains the interaction equation in two parts represented by Equation 29. Torsion is
considered in a separate equation to ensure that the sum of all stresses do not reach the yield stress.

[29]

M uy
Pu 8 M ux
+
+
Pn 9 b M nx b M ny

1.0 , for Pu 0.2 ,

Pn

M ux
M uy
Pu
+
+
2Pn b M nx b M ny

1.0 , for Pu < 0.2

Pn

Chapter C of the AISC 2005 specification includes detailed instructions on calculating second order
effects, whereas earlier editions (e.g. the 1986 LRFD specification) merely demand inclusion of P-
effects. Two factors are included, B1 and B2, to convert first order elastic analysis results into second order
values.
[30]

M f B1 M nt B2 M lt

[31]

B1

Cm
1.0
1 Pr Pe1

[32]

B2

1
1.0
1 Pnt Pe 2

One significant difference between the American approach and those of other national design
recommendations was that an effective length factor was exclusively used in the AISC Specification,
whereas notional loads were used in most other guidelines. An extensive review of effective length
factors versus notional loads was conducted (Hajjar et al., 1997), and resulted in the inclusion of an
appendix on using notional loads in the 2005 AISC Specification. The review recommended that the
notional loads be applied to both in-plane and out-of-plane directions, and this is reflected in the current
AISC Specification. Work by Surovek-Maleck and White (2004) indicated that a problem might be
experienced with notional loads as presented by Hajjar et al. Surovek-Maleck and White determined that
for beam-columns that are susceptible to elastic sidesway buckling and that are loaded with concentric
loads close to their buckling strength, the notional load method of AISC would result in factored
resistances close to the elastic buckling strength. To avoid this situation, the notional load method of
AISC incorporates a reduction factor for the frame stiffness of 0.8 (or lower) during the elastic analysis of
the frame, with notional loads of 0.002 times the gravity loads. A further stiffness reduction, b can be
avoided if the designer uses a higher notional load of 0.003 times the gravity loads. White et al. (2006)
presents this in great detail.
Structural steel design in Europe is primarily controlled by the EuroCode 3, which has recently been
revised and released as EN1993-1.1. Four methods of interaction equations are included. The existing

GC-147--8

methods known as the General Method and Stable Length Method are retained, and two new methods
were introduced (Greiner and Lindner, 2006). Method 1 (Boissannade et al., 2004) is based on second
order elastic theory, and was proposed by a French-Belgian team. Method 2 (Greiner and Lindner, 2006)
is based on empirical data and finite element analyses, and was proposed by an Austrian-German team.
A comparison between the AISC and EuroCode 3 provisions is made by Yong et al. (2006), who present
the interaction equations 33 (in-plane, or strong-axis, behaviour) and 34 (out-of-plane, or weak-axis,
behaviour) as the basis for the EuroCode 3 design.
[33]

M y , Ed
M z , Ed
N Ed
k yy
k yz
1 .0
y N Rk M 1
LT M y , Rk M 1
M z , Rk M 1

[34]

M y , Ed
M z , Ed
N Ed
k zy
k zz
1.0
z N Rk M 1
LT M y , Rk M 1
M z , Rk M 1

Values in these equations with the subscript Ed are applied design loads, including applicable second
order effects, and those with Rk are the member resistances accounting for full plastic capacity. The
values are reductions to the flexural and lateral-torsional buckling strengths. The k factors account for
interactions between loading, which include the effects of moment distribution along the length of the
beam (Cm) for both flexural and lateral-torsional buckling, the effect of plastic section interaction, and the
axial elastic buckling reductions, among other considerations. The interaction factors, k, are determined
by the design methods. The factors are resistance factors, called partial safety factors.
A summary of the two methods from Greiner and Lindner (2006) is presented in Table 2. The theoretical
Cm value for Method 1 (Boissannade et al., 2004) is also presented in Figure 1. Note that this value
depends on the ratio of the axial load to the buckling load, and uses a sinusoidal moment distribution as
the reference.

6. Future Directions and Conclusions


The 2005 AISC Specification notes in Appendix 7: Direct Analysis Method that the notional loads should
be applied independently in two orthogonal directions. The Australian design standard requires that
regular structures be analysed in each of two directions at right angles ( 4.3.1). The Canadian standard
is silent on the issue of direction of notional load application for frame analysis. Explicit mention should be
made of application of notional loads on both the in-plane and out-of-plane directions of the frame, to
provide the designer guidance on full second order analysis of frames. Specific guidance on the direction
of notional loads would also be of benefit for the analysis of structures subjected to torsional effects of
lateral loads, though this can be taken from the current requirement that the notional load is applied in the
direction of sway.
It is likely that interaction equations will be extended to encompass ever greater complexity. However, this
Table 2 Comparison of EuroCode 3, methods 1 and 2
Method 1
Method 2
Complex Cm expressions
Austins expression for Cm
Equivalent sinusoidal moments
Equivalent uniform moments
Lateral torsional buckling incorporated in Cm for
Lateral torsional buckling separate check
strong axis bending
Smooth transition from flexural to lateral
Separate treatment of torsionally stiff and
torsional buckling
flexible members
More complex formulae
Simpler formulae
Tuned for out-of-plane behaviour
Tuned for in-plane behaviour

GC-147--9

greater complexity could reach the point where it is no longer advantageous. If the design check
calculations require significant computing effort on the part of the designer or design aids, an advanced
analysis technique coupled with direct design will be at least as simple. Direct design techniques are not
new. For example, plastic design of frames is a direct design process. Extending that direction towards
advanced analysis will provide a method of simplifying the design process and capturing the frame
behaviour as well as looking at the strengths of the members outside the context of the frame.
Most current standards still separate the element from the frame during the design check process. This
can discount the interaction of members. An example of this is illustrated in MacPhedran and Grondin
(2007) wherein frames with a nominal resistance exactly equal to the applied loads exhibit interaction
equations that present some members being highly overstressed and others highly understressed. The
interactive out-of-plane buckling is not handled well by the equations, which look at the members as
individual components, rather than as part of the whole structure. As Massonnet (1976) indicated three
decades ago, a major problem with beam column design is the concept of isolating each member during
design. In his words, the isolated beam-column ... only exists in theoretical models and testing
machines.
7.

Acknowledgements

The authors gratefully acknowledge the financial support of the Alberta Ingenuity Fund in supporting the
work for this paper.

8. References
Adams, P.F. 1970a. Lecture III: Member stability provisions in CSA-S16-1969 Steel Structures for
Buildings. in Structural Engineering Report No. 32, (CSA-S16-1969 Steel Structures for Buildings
Seminar Notes), University of Alberta, November 1970.
Adams, P.F. 1970b. Lecture VI: Overall stability provisions in CSA-S16-1969 Steel Structures for
Buildings. in Structural Engineering Report No. 32, (CSA-S16-1969 Steel Structures for Buildings
Seminar Notes) University of Alberta, November 1970.
Bleich, F. 1952. Buckling Strength of Metal Structures, Engineering Society Monographs. (with Ramsey,
L) Bleich, H. editor
Boissonnade, N., Jaspart, J.-P., Muzeau, J.-P., and Villete, M. 2004. New interaction formulae for beamcolumns in Eurocode 3: the French-Belgian approach. Journal of Constructional Steel Research, Vol.
60, pp 421-431.
Chen, W.-F. and Atsuta, T. 1976. Theory of Beam-Columns: Volume 1 In-plane behaviour and design.
McGraw-Hill, New York, NY, USA.
Dawe, J.L. and Kulak, G.L. 1984. Local buckling behavior of beam-columns. ASCE Journal of the
Structural Division, Vol. 112, No. 11.
Essa, H.S. and Kennedy, D.J.L. 2000. Proposed provisions for the design of steel beam-columns in S162001. Canadian Journal of Civil Engineering, Vol. 27, No. 4, pp 610-619.
Greiner, R. and Lindner, J. 2006. Interaction formulae for members subjected to bending and axial
compression in EUROCODE 3 the method 2 approach. Journal of Constructional Steel Research,
Vol. 62 pp 757-770.
Hajjar, J.F., Bridge, R.Q., Clarke, M.J., Leon, R.T., Lui, E.M., Sheikh, T.M. and White, D.W. 1997.
Effective Length and Notional Load Approaches to Assessing Frame Stability: Implications for
American Steel Design. ASCE, New York, NY, USA.
Johnston, B.G. 1981. History of Structural Stability Research Council. ASCE Journal of the Structural
Division, Vol. 107, No. ST8, August 1981, pp 1529-1550.
Kennedy, D.J.L. 1995. Limit states design of beam-columns in CSA S16.1-94. Structural Stability and
Design. Kitipornchai, Hancock and Bradford (Editors)
Kennedy, D.J.L., Picard, A., and Beaulieu, D. 1990. New Canadian provisions for the design of steel
beam columns. Canadian Journal of Civil Engineering, Vol 17, No 6, pp 837-893.
Kennedy, D.J.L. and Qureshi, S. 1988. Design strengths of steel beam-columns: Discussion, Canadian
Journal of Civil Engineering, Vol. 15, No. 1, pp. 136-140.

GC-147--10

Loov, R 1996. A simple equation for axially loaded steel column design curves. Canadian Journal of Civil
Engineering, Vol. 23, No. 1, pp 272-276.
MacPhedran, I. and Grondin, G.Y. 2007. Out-of-plane instability considering frame-member interaction.
Proceedings of the SSRC 2007 Annual Stability Conference, New Orleans, LA.
Massonnet, C. 1976. Forty years of research on beam columns in steel. Solid Mechanics Archives, Vol. 1,
No. 1, pp. 27-157.
Massonnet, C. 1959. Stability considerations in the design of steel columns. ASCE Journal of Structural
Engineerin., ASCE, Vol. 85, No. 7, Sep 1959, pp 75-111.
McKenzie, B. S. 1928. Canadian Engineering Standards Association (in Standardization Programs in
Industry). Annals of the American Academy of Political and Social Science, Vol. 137, Standards in
Industry. (May, 1928), pp. 17-24.
Sputo, T. 1993. History of steel beam-column design. ASCE Journal of Structural Engineering. Vol. 119,
No. 2, Feb 1993, pp 547-557.
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, Vol. 103, No. 8, pp. 1186-1196.
Timoshenko, S. P. 1983. History of Strength of Materials. Dover Publications Inc, New York, NY, USA
Trahair, N. 1986. Design strengths of steel beam-columns. Canadian Journal for Civil Engineering, Vol.
13, No. 6, pp 639-645.
White, D.W., Surovek, A.E., Alemdar, B.N., Chang, C.-J., Kim, Y.D. and Kuchenbecker, G.H. 2006.
Stability analysis and design of steel building frames using the 2005 AISC Specification. Steel
Structures. Vol 6, pp 71-91.
Yong, D.J., Lpez, A., and Serna, M.A. (2006) A comparative study of AISC-LRFD and EC3 approaches
to beam-column buckling resistance. Presented at Stability and Ductility of Steel Structures, Lisbon
Portugal, September 2006.

8.1 Design Standards and Codes


American Institute of Steel Construction (AISC) (1986) Load and Resistance Factor Design Specification
for Structural Steel Buildings, Sept. 1986, American Institute of Steel Construction, Inc.
AISC (1989) Specification for Structural Steel Buildings Allowable Stress Design and Plastic Design, June
1989
AISC (2005) ANSI/AISC 360-05, Specification for Structural Steel Buildings, March 2005, American
Institute of Steel Construction, Inc.
CSA (1961) S16-1961, Steel Structures for Buildings, Canadian Standards Association, Rexdale, ON
CSA (1969) S16-1969, Steel Structures for Buildings
CSA (1984) S16.1-M84, Steel Structures for Buildings
CSA (1989) S16.1-M89, Steel Structures for Buildings
CSA (1994) S16.1-94, Steel Structures for Buildings
CSA (2001) S16-01, Steel Structures for Buildings
CISC (1965) Structural Steel Data a synopsis based on CSA Standard S16-1965, CISC, Toronto, ON
National Research Council (NRC) (1941) National Building Code, 1st edition, Codes and Specifications
Section, National Research Council of Canada
NRC (1960) National Building Code of Canada, 3rd edition, Associate Committee on the National Building
Code, National Research Council of Canada
Standards Association of Australia (SAA) (1998) AS 4100-1998, Steel Structures, Standards Association
of Australia.

GC-147--11

S-ar putea să vă placă și