BUCKLING AND BRACING OF CANTILEVERS

By Sritawat Kitipomchai, 1 Peter F. Dux, 2 and Neville J. Richter 3

ABSTRACT: The elastic flexural-torsional buckling of cantilever I-beams is in-

Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

vestigated. The cantilevers have rigid translational or rotational restraints or

both at discrete points. The effect of the beam parameter, K, the load height,
the location of restraint positions along the beam, and the level at which the
restraint acts have been studied using the finite integral method. Results are
presented graphically as ratios of the increased critical load of the partially braced
beam and the corresponding critical load of the unbraced beam. The beam load
cases considered are concentrated loads and uniformly distributed loads. The
effectiveness- of the restraint locations and the types of restraint are investi-
gated. Experiments conducted using extruded high strength aluminium I-sec-
tion are reported. Test results obtained are in reasonable agreement with the
theoretical predictions.

INTRODUCTION

Questions which often arise in design are those related to the effec-
tiveness of the bracing system used to increase the buckling capacities
of the members. While most structural beams may be braced in different
ways, most arrangements can be represented by an idealized system
consisting of an elastic translational restraint acting at distance b above
the shear center of the beam cross section and an elastic rotational re-
straint (Fig. 1).
A number of studies (3,4,6-10) have been made on the effectiveness
of the various types of restraint and restraint stiffnesses. Mutton and
Trahair (8) investigated stiffness requirements for simply supported beams
and columns with mid-span rotational and translational restraint which
acted either at the top flange or the shear center. They calculated the
minimum restraint stiffnesses required to cause the member to buckle
in its second mode. Kitipomchai and Richter (6,7,10) studied the effec-
tivenesses of restraint location along the simply supported beam, and
the level of translational restraint within the beam cross section in re-
lation to the height of application of load. The loading cases considered
are end moments, point loads and uniformly distributed load. Optimum
braced locations for the various loading are given. They found that
translational restraint placed at the tension (bottom) flange level may be
effective for long shallow beams for which warping effects are of less
importance than those of uniform torsion. This conclusion is confirmed
by tests carried out by Roeder and Assadi (5,11).
Fewer studies have been made on the bracing of cantilever beams.
Nethercot (9) studied the effective length factors of cantilevers having
'Sr. Lect. in Civ. Engrg., Univ. of Q u e e n s l a n d , Q u e e n s l a n d , Australia.
2
Lect. in Civ. Engrg., Capricornia Inst, of A d v a n c e d Education, Q u e e n s l a n d ,
Australia.
3
System Programmer, Dept. of Mech. Engrg., Univ. of Q u e e n s l a n d , Q u e e n s -
land, Australia.
Note.—Discussion o p e n until February 1, 1985. To extend t h e closing date one
month, a written request m u s t be filed with the ASCE M a n a g e r of Technical a n d
Professional Publications. The manuscript for this p a p e r w a s submitted for re-
view a n d possible publication on September 16, 1983. This p a p e r is part of the
Journal of Structural Engineering, Vol. 110, No. 9, September, 1984. ©ASCE, ISSN
0733-9445/84/0009-2250/$01.00. Paper N o . 19157.

2250

J. Struct. Eng. 1984.110:2250-2262.

Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

' • & ) * ,

FIG. 1.—Idealiied Translational and Rotational Restraint

two restraint conditions at the end and under concentrated end load and
distributed load. He considered full restraint and translational restraint
at the shear center, and gave approximate expressions for buckling loads.
Although most cantilevers of practical dimensions would fail inelas-
tically, elastic buckling loads are of relevance during construction when
slendernesses are often greater than in completed structures. As well,
elastic buckling formulas are convenient for incorporation into design
rules in which theoretical elastic buckling loads are empirically reduced
either to inelastic capacities or to allowable loads. In this paper, the ef-
fectiveness of translational or rotational restraints or both on elastic can-
tilever beams is examined. The load cases considered are concentrated
end load and uniformly distributed load. Tests on high strength ex-
truded aluminium cantilever I-beams have been conducted to validate
the theoretical investigation.

BUCKLING OF PARTIALLY BRACED CANTILEVERS

A cantilever I-beam under general loading and with an intermediate

restraint is shown in Fig. 2. Loads considered include a concentrated
load, P, acting at a distance, a, from the fixed end support and at a level
a above the shear center, a uniformly distributed load, w, acting at a
level w above the shear center line, and a point moment acting at a dis-
tance d, from the fixed end support. The translational restraint is applied
to the beam at a distance, b, from the fixed end support and at a level
b above the shear center, and provides a force, HA .
The differential equations of minor axis bending and torsion are

", = **,-& (1)

d_$ rf3cf)
(2)
dz dz
in which EIy, GJ and EIW are the minor axis b e n d i n g rigidity, the tor-
sional rigidity and the warping rigidity, respectively.
Vertical and horizontal forces Rt a n d Hx, major a n d minor axes fixed
end moments Mxl a n d Myl a n d torque reaction Tzl at the fixed e n d
support are

2251

J. Struct. Eng. 1984.110:2250-2262.

Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

(1
Mm
-V—U-IL.
L

ELEVATION
1
/
CROSS SECTION

"^A \ j r ~ S h e a r centre axis

L
PLAN
-- Distribufed load line

FIG. 2.—-Cantilever l-Beam under General Loading and Discrete Intermediate

Restraint

Rt^P + wL (3)
Hi = HA.... (4)
2
wL
Mxl = Pa + M (5)
2
Myl^HAb (6)

T zl = P{u„ + «()>„) + w[ \ udz + w <$>dz I + TA + HA1 (7)

in which HA and TA are the horizontal and torque reactions at the restraint.
The major and minor axis bending moment distribution are

Mx = RiZ = Mxl -P{z-a) wz M(z - d), (8)

and M y = Myl = Mx$ - Hxz + HA(z - b) (9)

in which the expressions inside Macauly brackets ( ) are taken to be zero
if their values are negative.
The axial torque distribution is
du
Tz = Tzl + Mx— - Rtu + P(u - u„ - a$a){z-a)
dz

+ w\uz~\ udz - w\ §dz) - TA{z-b) - HAb{z-b) . (10)

Jo Jo
fc«=Ofe

J. Struct. Eng. 1984.110:2250-2262.

 Combining Eqs. 1, 2 and 9 the governing differential equations of mi-
nor axis bending and torsion become
d2u
EIy-^ = M,J1-MA-Hiz + HA{z ~b) (11)
Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

d<b d2<$> du
and GJ-?--EIu-rj = Tx + Mx—-R1u + P{u - ua - a$a){z - b)
dz dz dz

+ w[uz- I udz-w\ <$>dz) -TA(z-b) - HAb(z - b) (12)

Jo Jo
The boundary conditions for Eqs. 11 and 12 are at the fixed end
du di>
2 = 0; u = 4> = — = ~r = 0 (13)
dz dz
at the free end
2 = L; — dhi = 0 (14)
dz
and at the restraint
H -
z = b; u = —A - b$h (15)

and <>
| =^ (16)

in which KH and KR are the translational and rotational stiffnesses of the

restraint.
The differential equations (Eqs. 11 and 12) together with the boundary
conditions (Eqs. 13-16) may be solved for the elastic critical load factor
using the method of finite integrals (2). A computer program has been
prepared and the solution technique is similar to that described in pre-
vious papers (1,7,8).

NUMERICAL RESULTS

General.—In practice, the common loadings for a cantilever beam are

concentrated load and uniformly distributed load. It is usual for con-
centrated load points to be also points of restraint and hence the height
of application of load does not affect beam buckling capacity. In a crane
runway beam, with discrete restraints along its length, the load may act
at any point. It is not obvious where the optimum restraint locations
should be, or whether translational restraint at the level of top flange is
fully effective. The uniformly distributed load case is common in roof
structures where the load may arise from wind or live loading. The load
may act at the top flange, shear center or bottom flange or at any level.
The translational restraint or rotational restraint or both may be applied
at any location along the beam.

2253

J. Struct. Eng. 1984.110:2250-2262.

 The effectiveness of a restraint may be measured by, c, the ratio of
the buckling load of the cantilever with the restraint arrangement and
the buckling load of a similarly unbraced cantilever. Thus the values of
c give an indication of the improvement in stability provided by the re-
straint. For doubly symmetric I-beams, buckling loads may be expressed
Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

in terms of the beam parameter (7),

D
•EI,.,
K= 4.47- (17)
GJL2

in which D/T = section depth to flange thickness ratio; and L/ry = slen-
derness ratio. Large values of K imply short beams or deep sections or
both of very thin-walled sections for which warping effects are impor-
tant, whereas small values of K are associated with long beams or shal-
low sections or both for which warping effects are of less importance
than those of uniform torsion.
Position of a Full Restraint Along Cantilever.—The influence of the
position of a full restraint is investigated. A full restraint is assumed to
be capable of preventing both lateral deflection and twisting of the braced
cross section. The critical load ratio, c, for values of the beam parameters
it j
14r *"~
'\
fi A^
12 r 1
I - • 1
10 Load at rop flange j Optimum
' restraint
8- &
h
= * 1,0 p- location

6_ /~—L°
4- ///
2 0,1

0 i i i 1 1

"6 Lo if at shear centre

o
i=5 78 = 0
< h
/ // \
"=4 _
o
<3 -
o ^K=a.o -V//r = ^ : ^=^sa
_ i w —'//
_, 2 0,3
0,1
7
'
2i I ! I I

0,2 0,4 0,6 0,8 0,2 0,4 0,6 0,6

RESTRAINT LOCATION a RESTRAINT LOCATION

FIG. 3.—Buckling Load for Cantilevers FIG. 4.—Buckling Load for Cantilevers
with Concentrated Tip Load and Full with Uniformly Distributed Load and Full
Restraint Restraint

2254

J. Struct. Eng. 1984.110:2250-2262.

 K = 0.1-3.0 are shown in Figs. 3 and 4 for cantilevers with a concen-
trated tip load and uniformly distributed load respectively. The loads
are applied at top flange, shear center and bottom flange. It can be seen
that the increases in the buckling load are greatest for large values of
the beam parameter, K, and more so for top flange loading. The maxi-
Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

mum value of c that may be achieved range from 3 for small values of
K-14 for large values of K. However, it is likely that in-plane bending
or inelastic buckling will govern the design for cantilevers with large
values of K.
The results show that for small values of K, the optimum restraint
location is near mid-span for a concentrated tip load and near 0.4 of the
length from the fixed end for uniformly distributed load. For higher val-
ues of K, the optimum restraint locations move towards the cantilever
tip as the height of load application moves toward the top flange. For
a concentrated tip load, the optimum location varies between 0.5 and
0.8 and for a uniformly distributed load, it varies between 0.4 and 0.7.
Effects of Translational or Rotational Restraint or Both.—The effec-
tiveness of the level of translational restraint is compared with that of
rotational restraint and of full restraint for top flange (2a/h = 1), shear
center (2d/h = 0) and bottom flange loading (2a/h = -1) for values of
K = 0.6 and 3.0. The results are shown in Figs. 5 and 6 for concentrated
tip load and in Figs. 7 and 8 for uniformly distributed load.
The various types of restraints have different effects on the critical
buckling load, depending on the level of load application (2a/h). In all
cases it can be seen that full restraint is by far the best for all K values.

0,2 0,4 0,6 0,8 1,0

RESTRAINT LOCATION a

FIG. 5.-—Comparison of Restraint Types for Cantilevers with Concentrated Tip Load,
K = 0.6

2255

J. Struct. Eng. 1984.110:2250-2262.

Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

ni 1 1 1 1 1
u
0,2 0,4 0,6 0,8 1,0
RESTRAINT LOCATION a

FIG. 6.—Comparison of Restraint Types for Cantilevers with Concentrated Tip Load,
K = 3.0

The optimum positions have been discussed in the previous section. If

full restraint cannot be achieved, rotational restraint is the next best as
can be seen from Figs. 5-8, particularly if the restraint is placed within
0.4L from the fixed support.
For translational restraints only the critical buckling load ratios in-
crease slowly as the restraint moves towards the free end, irrespective
of the level of the restraint. Varying the value of K has only little effect
on the maximum value of c for the top flange and bottom flange load-
ings. However, for shear center loading and with top flange restraint,
the effect of increasing K shows a marked improvement in the value of c.
It is recommended that translational restraints be placed as close as
possible to the cantilever tip. The effectiveness increases as the level of
application of load moves towards the bottom flanges. In all cases if
translational restraint alone is used, it should be placed near the top
flange and as close as possible to the end of the cantilever. Translational
restraints placed less than 0.4L from the fixed end are practically useless
and therefore are wasted.

EXPERIMENTAL INVESTIGATIONS

General.—A series of cantilever I-beams with concentrated loads was

tested to verify the theoretical results obtained using the finite integral
2256

J. Struct. Eng. 1984.110:2250-2262.

 14 K =3,0 Full restraint
Rot
restraint
12 only
10
8
6
Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

- restraint
4 only -2i-.i r °
2
o0
0 f=6 r~ Full restraint
< SF Loading
°= 5-
li— Rot, restraint only
gs sc Loading
y — F u l l restrain! -Uj-Trans. restraint only
/ / — Rot. restraint only § 4
S4 h ^ — T r a n s , restraintonly o
§3 -*3
32 - < 2-
«M s ? = =:==-" -=====5^7 " ~ LJ ^^=±~
i i i i
t: 1 i I i i
E°5 cc
Full restraint
S BF Loading
^0
Full restraint Rot. restraint only
4
Rot. restraint only
5 Trans, restraint only
3 4
2 3
1 2
0 1
1,2 0,4 0,6 0,8 1,0 0,2 0,4 0,6 0,8
RESTRAINT LOCATION a 0 RESTRAINT LOCATION

FIG. 7.—Comparison of Restraint Types FIG 8.—Comparison of Restraint Types

for Cantilevers with Uniformly Distrib- for Cantilevers with Uniformly Distrib-
uted Load, K = 0.6 uted Load, K = 3.0

method of solution. The beams were high strength aluminium extru-

sions, similar to those used previously in the experimental investigation
of elastic simply supported beams (1,7,10). The beam cross sectional di-
mensions and the measured properties are given in Fig. 9. The experi-
mental program consisted of testing five different lengths of cantilevers
ranging from 1.0-3.0 meters with varying restraint conditions and lo-
cations. All beams were loaded at the level of the top flange and at the
shear center by means of a loading yoke (Fig. 10). The load was applied
as close to the tip as practicable. The restraint conditions were full trans-
lational restraint at the level of top flange (T), shear center (S), bottom
flange (B) or full restraint (F) against both translational and rotational
deformation.
Equipment and Procedure.—The test apparatus and procedure closely

31,4
r I
Br- Ely = 1040Nm2
GJ = 21,15 Nm2
2,0 E L = 1,347Nm4
Weight = 0,87 kg/m

Tr-
FIG. 9.—-Test Beam Dimensions and Cross-Sectional Properties

2257

J. Struct. Eng. 1984.110:2250-2262.

Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

FIG. 10.—Restraint and Loading Device

followed that used in the previous investigations (1,7,10). The fixed end
support arrangement was similar to that used by Anderson and Trahair
(1). It allowed a single beam to be moved through the support for vari-
able cantilever length. Steel blocks machined to shape were fitted on
either side of the web between the flanges. With the beam in position
four bolts, two on each side of the beam, were then used to tightly clamp
a 20 mm thick plate overlying the top flange to a base support itself
rigidly connected to a substantial supporting frame.
Restraint at the cantilever tip was achieved by the use of one or two
long wires attached at their mid-points to the free end. Each wire had
a brass socket approximately 15 mm in length at its mid-point. Brass
fittings providing longitudinal pins of similar length at top flange, shear
center and bottom flange levels were mounted on the web at the can-
tilever tip. Connection between bracing wire and cantilever tip was ef-
fected by slipping the socket onto a pin as indicated in Fig. 10. The far
ends of each wire were connected to adjustable mountings which con-
trolled wire loading and permitted the wire to travel downwards with
the beam. This avoided the development of significant upwards re-
straining forces. A single wire provided restraint against lateral deflec-
tion at any of the attachment points. Full restraint was achieved by at-
taching wires at both top and bottom flange levels.
The test loads were applied to the top flange and shear center through
loading yokes supporting a bucket carrying lead shot. Fig. 10 shows the
yoke applying top flange loading. Loads at lower levels were applied by
hanging the load bucket on a wire passing through the web close to the
cantilever tip. A spreader above the bucket ensured that the wire did
not touch the cantilever bottom flange. The loading arrangements en-
sured vertical loading through the desired load points irrespective of any
lateral movement of the points. Load increments were applied by the
gradual addition of lead shot to the bucket. Increment magnitudes de-
2258

J. Struct. Eng. 1984.110:2250-2262.

 creased as expected cantilever capacity was approached. After each in-
cremental loading deflections were measured.
Tests were carried out on the longest cantilevers first in order to pre-
vent any effect on subsequent beams due to accidental damage to the
beam near the support. Each experiment was repeated a minimum of
Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

three times in order to ensure repeatability of results and variation was

less than 3% in all cases. The modified Southwell plot was used to obtain
the experimental critical loads from load and lateral deflection measure-
ments (1,7).
Lateral deflections of the beam were measured at the level of the shear
center at a location where maximum lateral deflections were anticipated.
These were 0.625, 0.375 and 0.7 of the length from the fixed end for b/L
= 0.25, 0.75 and 1.0, respectively. A micrometer connected into an elec-
trical circuit allowed very sensitive lateral deflection readings to be ob-
tained. The critical load for each test was assumed to be closely ap-
proximated by the slope of the line obtained by plotting lateral deflection

TABLE 1.—Comparison of Results

Posi-
Posi- tion Buckling Load, Per-
tion of re- Type in Newtons centage
Length, of straint of re- Experi- differ-
Beam in meters K load b/L straint ments Theory ence
d) (2) (3) (4) (5) (6) (7) (8) 0)
1 3.0 0.27 TF 1.0 F 181.5 188.3 -3.6
2 3.0 0.27 TF 1.0 T 104.0 112.7 -7.7
3 3.0 0.27 TF 1.0 S 89.8 97.1 -7.5
4 3.0 0.27 TF 1.0 B 80.4 85.0 -5.4
5 3.0 0.27 SC 0.5 F 297.0 284.8 +4.3
6 3.0 0.27 SC 0.5 T 103.9 100.9 +3.0
7 3.0 0.27 SC 0.5 B 76.6 74.4 +3.0
8 2.5 0.32 SC 0.5 F 439.4 436.4 +0.7
9 2.5 0.32 SC 0.5 T 158.6 158.6 0
10 2.5 0.32 SC 0.5 S 130.9 124.6 +5.1
11 2.5 0.32 SC 0.5 B 119.3 115.7 +3.1
12 2.5 0.32 SC 0.25 F 195.7 206.3 -5.1
13 2.5 0.32 SC 0.75 F 324.3 337.9 -3.8
14 2.0 0.40 SC 0.5 F 804.0 738.4 +8.9
15 2.0 0.40 SC 0.5 T 277.0 275.4 +0.6
16 2.0 0.40 SC 0.5 B 196.5 197.0 -0.3
17 1.5 0.54 TF 1.0 F 803.2 902.5 -11.0
18 1.5 0.54 TF 1.0 T 382.0 424.8 -10.1
19 1.5 0.54 TF 1.0 S 296.0 326.7 -9.4
20 1.5 0.54 TF 1.0 B 294.0 282.2 +4.2
21 1.0 0.79 TF 1.0 T 750.5 801.8 -6.4
22 1.0 0.79 TF 1.0 S 573.9 565.4 + 1.5
23 1.0 0.79 TF 1.0 B 519.9 520.2 0
Note: TF = load at top flange {2a/h = +1); SC = load at shear center {2a/h =
0); F = full restraint; T, S, B = translational restraint at level of top flange {2b/h
= +1), shear center {2b/h = 0) and bottom flange {2b/h - 1), respectively.

2259

J. Struct. Eng. 1984.110:2250-2262.

 versus lateral deflection over load as abscissa. Although experimental
loads approached buckling loads, the recovery to initial readings of gages
for both lateral and vertical deflection indicated that no yielding occurred.
Results.—Results are summarized in Table 1. Experimental values in
this table represent the average of results of the several tests conducted
Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

on each beam. Also shown in the table are the theoretical results from
using the finite integral methods (1,7,10). The predictions have allowed
for self-weight of the beams and also for the fact that the major axis
flexural rigidity EIX is not infinitely larger than the other rigidities (12).
It was found the effect of neglecting both beam self-weight together with
major axis curvature is for one to approximately cancel the other. The
experimental results are in reasonable agreement with the theoretical
predictions. The results confirm the theoretical findings that transla-
tional restraint at the top flange level is more effective than at other
levels, but is not as effective as rotational or full restraint.

CONCLUSIONS

The elastic buckling of cantilever I-beams under general loading and

with a variety of restraint conditions is investigated. The governing dif-
ferential equations together with appropriate boundary conditions are
derived. The elastic buckling loads are obtained by solving the differ-
ential equations numerically using the method of finite integrals.
The influence of restraint location along the beam, the height of ap-
plication of load and the types of restraint are studied for the several
values of beam parameter, K. The load cases considered are concen-
trated tip loads and uniformly distributed load. It is found the optimum
location of a full restraint for most cases varies between 0.4-0.7 from
the fixed end support. For beams with a simple translational restraint,
the restraint is best placed near the top (tension) flange level. However,
this arrangement is not as effective as a rotational restraint or a full
restraint.
Experiments on extruded high strength aluminium cantilever I-beams
are reported. Twenty-three beams were tested with lengths varying from
1.0-3.0 m. The cantilevers were loaded with concentrated tip load. Re-
straints placed along the test beams were either a translational restraint
at the level of top flange, shear center, bottom flange or a full restraint.
Experimental buckling loads were generally lower than the theoretical
predictions. However, results confirm the conclusions from theoretical
studies on the order of effectiveness of the different types of restraint.

ACKNOWLEDGMENTS

Experimental investigations described in this paper were carried out

in the Department of Civil Engineering at the University of Queensland.
The assistance given by B. Johnson, P. D. Bolton-Hall and P. Moodie is
gratefully acknowledged.

APPENDIX I.—REFERENCES

1. Anderson, J. M., and Trahair, N. S., "Stability of Monosymmetric Beams and

22G0

J. Struct. Eng. 1984.110:2250-2262.

 Cantilevers/' Journal of the Structural Division, ASCE, Vol. 98, No. ST1, Paper
8648, 1972, pp. 269 286.
2. Brown, P. T., and Trahair, N. S., "Finite Integral Solution of Differential
Equations," Civil Engineering Transactions, Institution of Engineers, Australia,
Vol. CE10, No. 2, 1968, pp. 193-196.
3. Dux, P. F., and Kitipornchai, S., "Elastic Buckling of Laterally Continuous
Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

Beams," Journal of the Structural Division, ASCE, Vol. 108, No. ST9, Paper
17320, 1982, pp. 2099-2116.
4. Hancock, G. J., and Trahair, N. S., "Finite Element Analysis of the Lateral
Buckling of Continuously Restrained Beam Columns," Civil Engineering
Transactions, Institution of Engineers, Australia, Vol. CE20, No. 2, 1978, pp.
120-127.
5. Kitipornchai, S., discussion of "Lateral Stability of I-Beams with Partial Sup-
ports," by C. W. Roeder and M. Assadi, Journal of Structural Engineering,
ASCE, Vol. 110, No. ST1, 1984, Paper 18477, pp. 178-181.
6. Kitipornchai, S., and Richter, N. J., "Lateral Buckling of Beams with Discrete
Braces," Proceedings, Conference on Metal Structures, Institution of Engi-
neers, Australia, Perth, Australia, 1978, pp. 54-59.
7. Kitipornchai, S., and Richter, N. J., "Elastic Lateral Buckling of I-Beams with
Discrete Intermediate Restraints," Civil Engineering Transactions, Institution of
Engineers, Australia, Vol. CE20, No. 2, 1978, pp. 105-111.
8. Mutton, B. R., and Trahair, N. S., "Stiffness Requirements for Lateral Brac-
ing," Journal of the Structural Division, ASCE, Vol. 99, No. ST10, Paper 10086,
1973, pp. 2167-2182.
9. Nethercot, D. A., "The Effective Lengths of Cantilevers as Governed by Lat-
eral Buckling," The Structural Engineer, Vol. 51, No. 5, 1973, pp. 161-168.
10. Richter, N. J., "Application of the Finite Integral Method to Lateral Buckling
of I-Beams," Thesis presented to the University of Queensland, at Queens-
land, Australia, in 1978, in partial fulfillment of the requirements for the de-
gree of Master of Engineering Science.
11. Roeder, C. W., and Assadi, M., "Lateral Stability of I-Beams with Partial
Supports," Journal of the Structural Division, ASCE, Vol. 108, No. ST8, Paper
17279, 1982, pp. 1768-1780.
12. Vacharajittiphan, P., Woolcock, S. T., and Trahair, N. S., "Effect of In-Plane
Deformation on Lateral Buckling," Journal of Structural Mechanics, Vol. 3, 1974.

APPENDIX II.—NOTATION

The following symbols are used in this paper:

a = location of concentrated load;

a = height of point of application of load above shear center;
b = location of restraint along beam;
b = height of translational restraint above shear center;
c = ratio of critical load of restrained b e a m a n d similar unbraced
beam;
D = depth of beam;
d = location of point moment;
E = Young's m o d u l u s of elasticity;
G = shear m o d u l u s of elasticity;
HA = horizontal reaction at restraint;
Hi = horizontal reaction at fixed end support;
h = distance b e t w e e n flange centroids;
lx,ly = major a n d minor second m o m e n t of area;
/», = warping section constant;
/ = torsion section constant;
2261

J. Struct. Eng. 1984.110:2250-2262.

 K = beam parameter = VIT 2 El^/GJL 2 (See Eq. 17);
KH = lateral restraint stiffness;
KR = rotational restraint stiffness;
L = length of beam;
M = applied point moment;
Downloaded from ascelibrary.org by HUNT LIBRARY ACQ DEPT on 01/22/15. Copyright ASCE. For personal use only; all rights reserved.

Mx = major axis bending moment;

Mxl = major axis moment reaction at fixed end support;
My = minor axis bending moment-
Myl = minor axis moment reaction at fixed end support;
P = concentrated load;
Ri = vertical reaction at fixed end support;
r
y = minor axis radius of gyration;
T = flange thickness;
TA = torque reaction at restraint;
Tz = torque distribution along beam;
Ttl = torque reaction at fixed end support;
u = lateral deflection of shear center;
Ua = lateral deflection at distance z = a;
w = uniformly distributed load;
w = height of distributed load above shear center;
z = I centroidal axis with origin at fixed end support;
* = angle of twist;
<!>„ = angle of twist at z = a; and
<bb = angle of twist at z = b.

2262

J. Struct. Eng. 1984.110:2250-2262.