You are on page 1of 9

LATERAL-ToRSIONAL STABILITY OF STEEL WEB-TAPERED I-BEAMS

By Dimos Polyzois 1 and Ioannis G. Raftoyiannis 2

ABSTRACT: This paper deals with the stability behavior of web-tapered beams subjected to bending loads.
According to the current AISC specifications, a tapered member is treated as prismatic with modified length and
cross section identical to that of the tapered member's smaller end. Modification factors are used to determine
the modified length and also to account for the moment gradient within the unbraced length and the restraining
effect for adjacent spans. In this paper, these factors are evaluated via a finite-element analysis and compared
to the ones proposed by the current specifications. The current design specifications for steel tapered beams are
reexamined, and modifications that cover a wider range of geometry and loading cases are proposed.
Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

INTRODUCTION beam theory. Wekezer (1985) derived numerical results for ta-
pered I-beams using the membrane-shell theory that were in
Tapered steel members with I-sections have become very good agreement with the results obtained by Kitipornchai and
popular in building construction. Tapered beams can resist a Trahair (1972). Yang and Yau (1987), Bradford and Cuk
maximum stress at a single location while in the rest of the (1988), and Chan (1990) developed and used tapered beam
member stresses are considerably lower. This results in appre- finite elements for their analyses. The lateral-torsional and lo-
ciable savings both in materials and in construction. The elas- cal buckling of web-tapered I-beams has been investigated by
tic analysis of tapered I-beams has been the focus of investi- Polyzois and Li (1993) using a finite-element method.
gation by numerous researchers. Krefeld et al. (1959), Butler In the theoretical investigation presented in this paper, the
and Anderson (1963), and Butler (1966) conducted experi- lateral-torsional buckling of tapered beams and girders is stud-
mental studies on I-shaped beams and channel sections ac- ied and the modification factor B used in the current AISC
cording to which no interaction exists between the flange and specification (Manual 1994) is redefined and separated into
web tapers. They also concluded that the interaction formula two factors: the stress gradient coefficient B for the analysis
suggested by Salvadori (1956) for uniform beam-columns was of laterally unbraced simply supported members, and the re-
applicable to a wide range of tapered I-sections. Boley (1963) straint factor R for the analysis of laterally braced members.
studied the behavior of tapered rectangular beams and con- Expressions for factors B and R are developed using the nu-
cluded that for members with angles of taper less than 15° the merical results from finite-element analysis.
usual approximation of structural mechanics can still be ap-
plied. Based on this finding, the analysis of tapered members REVIEW OF CURRENT DESIGN SPECIFICATIONS
is considerably simplified, and he suggested that design guide-
lines for prismatic members be extended to tapered members The current design specifications for web-tapered beams are
as well. Culver and Preg (1968) investigated tapered beam- covered in Appendix F3 of the AISC Specification (Manual
columns with unequal end moments. Prawel et al. (1974) con- 1994). It covers the design of doubly symmetric web-tapered
ducted experimental research on the inelastic stability of ta- beams the flanges of which have equal and constant area, and
pered I-shaped beams and columns. Lee et al. (1972) the depth of the web varies linearly as
investigated the behavior of tapered members extensively.
Their work has been adopted in Appendix F3 of the AISC
specification (Manual 1994). Kitipornchai and Trahair (1972,
d z = do (1 + 'Y ~) (1)
1975) studied the elastic stability of simply supported double- The taper ratio 'Y in (1) is defined as follows:
symmetric I-beams with either tapered flange or tapered web
and developed a general method for deriving the elastic lateral- d L - do L
'Y = - - - ~ 0.268 - ~ 6.0 (2)
torsional buckling load of tapered symmetric I-beams sub- do do
jected to various loading. The experimental results for tapered
aluminum I-beams were found to correlate very well with the where d z ' do, d L , and L in (1) and (2) are illustrated in Fig. 1.
predicted critical loads. Moreover, Kitipornchai and Trahair The basic approach followed in the current AISC specification
(1975) studied the behavior of monosymmetric I-beams and (Manual 1994) is that the critical moment Mer for a tapered
concluded that the out-of-plane bending and the torsion resis- beam subjected to end moments is the same as the critical
tances are independent of each other. Horne et al. (1979) de- moment of a prismatic beam with different length and the
veloped expressions for estimating the critical loads of tapered
members subjected to arbitrary bending moment distributions. r s r C

FI"·,,
A
Brown (1981) performed a numerical solution for tapered 1- r
shaped cantilever and simply supported beams. Numerical
techniques have also been employed by Kosko (1982) to in-
vestigate beams with varying cross sections using the classical
Ii L-A
-

2 ---tL-s L- c
I
'Prof.. Dept. of Civ. and Geol. Engrg., Univ. of Manitoba, Winnipeg, I- L II
Canada, R3T 2N2.
t,

~f""~1 1;1
2Postdoctoral Fellow, Dept. of Civ. and Geol. Engrg., Univ. of Mani-
toba, Winnipeg, Canada, R3T 2N2.
Note. Associate Editor: Amde M. Amde. Discussion open until March
1, 1999. To extend the closing date one month, a written request must
be filed with the ASCE Manager of Journals. The manuscript for this
paper was submitted for review and possible publication on November
21, 1997. This paper is part of the Journal of Structural Engineering,
l2
f-b--l
Section A-A
T
do
L
f-b--l
Section B-B
f- b--l
SectionC-C
Vol. 124, No. 10, October, 1998. ©ASCE, ISSN 0733-9445/98/0010-
1208-1216/$8.00 + $.50 per page. Paper No. 17054. FIG. 1. Geometry of Typical Tapered I-Beam

1208/ JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 1998

J. Struct. Eng. 1998.124:1208-1216.


Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

Fb2
Fb1 , - -_ _ ben_d_in_g_str_e_s_s_di_agram
_ _-,

cf------- t )
M
O

f--l ---t-- l --+--- l ~


M
I

I-- l --f--- l -----j

bracing 8><---~8--~8--~8 8><-----'l'r8-------1<8 bracing


(a) Case 1 (b) Case 2

Fb2~~
~ing stress diagram I FbI

I-- l --l---- l -----j


i - --
-l)
bracing ~8----~8--------'Io8 8 8 bracing
(c) Case 3 (d) Case 4
FIG. 3. Four Cases Considered in Current AISC Specification

same cross section as the one of the tapered beam correspond- and
ing to the smaller end (see Fig. 2). The factored bending re-
sistance of a tapered beam is hw = 1 + 0.OO385"(VLlrTO (8)

The term rTO is the radius of gyration of the section at the


(3)
small end considering only the compression flange plus one
third of the compression web area (mm); Af is the area of the
where S; = section modulus of the tapered member at the point compression flange (mm2); L is the length of the section (mm);
of maximum stress; <Pb (=0.85) = bending resistance factor; and do is the depth of the section at the small end (mm). The
Fb'f = allowable lateral torsional buckling stress; and the factor term B in (4a) and (4b) is a modification factor that accounts
(5/3) is used to increase the allowable stress to a limit state for the moment (or stress) gradient and/or the restraining effect
stress. of adjacent spans of a member that is continuous over lateral
The allowable stress Fb'f is given by supports.
Four cases are covered in Appendix F3 of the AISC speci-
Fb'f = BVF;., + F~., for Fb.,:S Fyl3 (4a) fications (Manual 1994) that are summarized in the following:

• Case I: When the maximum moment M 2 in three adjacent


segments of approximately equal unbraced length is lo-
cated within the central segment, and M 1 is the larger
where Fy = yield stress (MPa); Fs'f = allowable Saint-Venant's moment at one end of the three-segment portion of the
torsional resistance (MPa); and F w., = allowable warping re- member [see Fig. 3(a)]:
sistance of the equivalent prismatic member (MPa). The terms
F.., and FW'f are defined as follows:
3
B = 1 + 0.37 (1 + Z:) + 0.50"( (1 + Z:) ~ 1.0 (9)
F,., = 83 X 10 (5)
hsLdolAf where M/M 2 is considered negative when the critical seg-
ment is bent in single curvature. If M/M2 is positive, the
3
F = 1,170 X 10
(6)
specification recommends this ratio be taken as zero.
w., (h w LlrTO)2 Case 2: When the largest computed bending stress F b2
occurs at the large end of two adjacent segments of ap-
where proximately equal unbraced length, and Fbi is the com-
puted bending stress at the smaller end of the two-seg-
h, =1 + 0.023"(VLdoIAf (7) ment portion of the member [see Fig. 3(b)]:
JOURNAL OF STRUCTURAL ENGINEERING I OCTOBER 1998 11209

J. Struct. Eng. 1998.124:1208-1216.


B =1 + 0.58 (1 + FbI) - 0.70", (1 + Fbi)
Fb2 Fb2
~ 1.0 (10)

where F b/Fb2 is considered negative when the two-seg-


ment portion of the member is bent in single curvature.
The ratio Fbl /Fb2 always takes nonzero values.
• Case 3: When the largest computed bending stress F b2
occurs at the smaller end of two adjacent segments of
approximately equal unbraced length, and FbI is the com- I---- L --+- L -+-- L - - l
puted bending stress at the larger end of the two-segment bracing
portion of the member [see Fig. 3(c)]: 8 8 8 8
FIG. 4. Effect of Adjacent Spans In Laterally Braced Beams
B =1 + 0.55 (1 + FbI) + 2.20", (1 + FbI) ~ 1.0 (11)
Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

Fb2 Fb2
critical segment due to restraints provided by adjacent seg-
• Case 4: When the computed bending stress at the smaller ments was about 35%. Based on these findings, relationships
end of a tapered member or segment is equal to zero [see between the buckling capacity of laterally restrained tapered
Fig. 3(d)]: members with various moment gradients and the taper ratio
were developed and incorporated in Appendix F3 of the cur-
B = 1.75 (12) rent AISC specifications (Manual 1994).
1 + 0.25Y:Y
DEVELOPMENT OF STRESS GRADIENT AND
In this case, '" is calculated for the unbraced length ad- RESTRAINT FACTORS
jacent to the point of zero bending stress.
Since the AISC expressions for the modification factor B
Eqs. (9)-(12) have been developed by Lee et al. (1972), include both the effect of stress gradient and the effect of con-
tinuity, it is desirable to quantify the level of importance of
Morrell and Lee (1974), and Lee and Morrell (1975). In their
original work these investigators referred to the term B in each parameter. Use of the recommended AISC values of B
factor implies that both parameters are equally important.
(9)-(12) as R~, a factor that accounts for both the stress gra-
dient and the restraint provided by adjacent spans on the crit- There may be situations in practice, however, where the effect
of continuity is drastically diminished, especially when the lat-
ical sections shown in Figs. 3(a)-(c). This term is defined as
eral supports have insufficient stiffness and/or are improperly
R = (IT,)LR applied to the tapered member. In this case, it may be wise to
(13) ignore the effect of continuity.
, (IT,)ssla=.
The finite-element analysis program Buckling Analysis of
where (lT~)LR = elastic buckling stress of a critical section in a Stiffened Plates (BASP) has been used to develop separate
laterally restrained tapered beam; and (lT~)ss = elastic buckling expressions for the stress gradient factor B and the restraint
stress of a simply supported tapered beam the dimensions of factor R of tapered beams. The BASP program was developed
which are identical to those of the critical section that is loaded at the University of Texas at Austin (Akay et al. 1977), and
with end moments producing a nearly uniform stress in the was used here to analyze the elastic stability of I-sections with
critical section. The term k is the ratio between the section one axis of symmetry perpendicular to the bending axis. In
modulus at the end and the section modulus at the large end this program, plate elements are used for the web and one-
(k = SO/SL); and a is the ratio between the end moment at the dimensional elements are used for the flanges and stiffeners.
small end and the end moment at the large end (a = Mo/ML ). The program accounts for the cross section distortion as well
Thus, a = k implies a constant stress, while a :¢: k implies a as the local buckling of the web and torsional buckling of the
stress gradient. flanges. Various loading and restraint conditions can be incor-
Factor B in cases (l) and (3) to include the effect of restraint porated in the program. The program first carries out a stress
provided by adjacent spans to the critical span. The effect of analysis in the plane of the web under the applied loading,
continuity over lateral supports has been well documented in which is termed "membrane analysis," then an out-of-plane
the literature (Nethercot 1983; Hartman 1967; Trahair 1977) analysis called "buckling analysis," which uses the stresses
and methods for evaluating the buckling capacity of laterally calculated under the applied loading. It considers the problem
restrained prismatic beams have been developed. The degree as a linear-elastic buckling problem, and inverse iteration is
of restraint provided by adjacent spans to the critical segment used to solve the resulting eigenvalue. The critical stress can
in a member continuous over a number of lateral supports then be obtained by multiplying the applied stress by the re-
depends on the stress level in these spans and the stiffness of sulting eigenvalue from the buckling analysis.
the lateral supports. The higher the stress level, the lower the Two-dimensional plane stress elements (quadrilateral ele-
restraining contribution. In the design of prismatic beams, a ments composed of four triangles) were used to model the web
conservative critical load is determined ignoring the restraint for the in-plane stress analysis. There are two degrees of free-
contribution of adjacent spans, and thus each unbraced seg- dom at each corner and interior node of the assembled quad-
ment is considered laterally and torsionally simply supported. rilateral. For the out-of-plane analysis, the quadrilateral ele-
To investigate the buckling behavior of tapered members ments were composed of triangular plate bending elements
and the effect of adjacent unbraced segments on the buckling where the corner nodes and the central interior node have three
capacity of a critical segment, Morrell and Lee (1974) used degrees of freedom. Beam elements were used to model the
the moment diagram shown in Fig. 4. This type of moment flanges and stiffeners. The beam element has a cubic displace-
distribution was chosen so that the center span would be crit- ment expansion for the translation normal to the web and lin-
ical and the stress would be nearly uniform in this span. The ear expansion for the torsional rotation. More detailed infor-
computed ratio of the critical stress of the laterally restrained mation about the BASP program is given by Akay et al.
member to the critical stress of the unrestrained segment was (1977).
almost constant for the range of taper ratios investigated by Three test examples, chosen from the work by Krefeld et
Morrell and Lee. The increase in the buckling capacity of the al. (1959), were used to check the accuracy of the BASP pro-
1210 I JOURNAL OF STRUCTURAL ENGINEERING I OCTOBER 1998

J. Struct. Eng. 1998.124:1208-1216.


p
ber and the smallest computed bending stress Fbi occurs
at the small end [see Fig. 6(b)].
• Case l(c): Single span, where the largest bending stress
F b2 occurs at the small end of the tapered member and
the smallest computed bending stress Fbi occurs at the
large end [see Fig. 6(c)).
• Case I(d): Single span, where the computed bending mo-
ment at the small end of the tapered member is zero [see
Fig. 6(d)].

FIG. 5. Cantilevered Web-Tapered I-Beam (Krefeld et al.1959) Expressions for Band R for three cases of continuous ta-
pered I-beams were also developed. These are
gram. These involved cantilevered web-tapered I-beams, as il-
Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

lustrated in Fig. 5, with a vertical load at the free end. Lateral • Case 2(a): This case is similar to case 1 of the AISC
bracing was provided at the top flange of the free end. The specification (Manual 1994). It applies to a tapered 1-
dimensions of the cross section are shown in Table 1. The beam with three consecutive spans, where four lateral
modulus of elasticity and Poisson's ratio were E = 200,000 braces are placed at equal intervals. An interior moment
MPa and v = 0.3, respectively. M2 occurs at the central span, and M I is the larger of the
The finite-element mesh used in these examples consisted two moments occurring at the ends of the beam [see Fig.
of 4 X 20 plate elements in the web, and 40 bar elements for 3(a)].
the flanges. The results from BASP program are shown in • Case 2(b): This case is similar to case 2 of the AISC
Table 1, along with the results obtained by Krefeld et al. specification (Manual 1994). It applies to a tapered 1-
(1959). The results indicate that the BASP program yields con- beam with two spans, where three lateral braces are
servative but satisfactory results. placed at equal intervals. The largest bending stress F b2
In developing the expressions for the stress gradient factor occurs at the large end of the beam, and the smallest bend-
B and the restraint factor R, four cases of loading and restraint ing stress Fbi occurs at the small end of the beam [see
conditions for simply supported tapered I-beams were inves- Fig. 3(b)].
tigated. These cases are • Case 2(c): This case is similar to case 3 of the AISC
specifications (Manual 1994). It applies to a tapered 1-
• Case l(a): Single span, where the maximum moment M 2 beam with two spans, where three lateral braces are
occurs within the span and M 1 is the larger moment at placed at equal intervals. The largest bending stress F b2
ends [see Fig. 6(a)]. occurs at the small end of the beam, and the smallest
• Case 1(b): Single span, where the largest computed bend- bending stress Fbi occurs at the large end of the beam [see
ing stress F b2 occurs at the large end of the tapered mem- Fig. 3(c)].

TABLE 1. Dimensions of Cantilevered Web-Tapered I-Beam Used by Krefeld et al (1959)


Buckling Load
Beam L do dL bo bL t, tw Krefeld et al. BASP Difference
number (mm) (mm) (mm) (mm) (mm) (mm) (mm) 'Y (N) (N) (%)
(1 ) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11 ) (12)
1 3,658 254 50.8 85.7 85.7 9.53 6.35 0.80 12,365 10,502 -15.1
2 4,267 330.2 76.2 101.6 101.6 11.11 7.94 0.77 17,347 15,435 -11
3 4.877 406.4 101.6 146 146 7.94 6.35 0.75 27,133 25,910 -4.5

Fb2
FbI L-_ _b_en_d_in~g_s_tres_s_dl_'a~gram_--J

I----- L ---+- L ----+--- L --j


tf=- ==p
I------ L - - 1 - - - L ----..,
> t---------4<8
bracingg 8*-----------1<8 bracing
(a) Case l(a) (b) Case l(b)

bending stress diagram

i---
I------ L - - 1 - - - L ----1 L -----I
bracing 8*---------~8 8*-----------1<8 bracing
(c) Case l(c) (d) Case l(d)
FIG. 6, Loading Conditions for Development of Stress Gradient Factor B

JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 1998/1211

J. Struct. Eng. 1998.124:1208-1216.


Cases l(a), l(b), l(c), and l(d) were used to develop the and F y = 350 MPa. Since the main objective of this investi-
stress/moment gradient factor B. In these cases the beams were gation was to study the lateral-torsional buckling of tapered
simply supported and laterally unbraced. Cases 2(a), 2(b), and beams, the cross sections were chosen to have a height-to-
2(c) were used to develop the restraint factor R. These cases thickness ratio sufficient to prevent local buckling. The stress/
also involved simply supported beams but laterally braced at moment ratio varied from -0.2 to -0.8 and the taper ratio
equal intervals. In all cases, the expressions for factors Band varied from 0.0 to 3.0. These ranges cover the majority of
R are derived using curve-fitting techniques on the numerical
results from finite-element analysis. TABLE 4. Factor B Obtained from BASP for Case 1(b)
In cases l(a) to l(d), factor B represents the stress gradient
Fb ,IFb2
in a similar way as the moment gradient coefficient Cb in pris-
matic members. In these cases, factor B is defined as Section 1 Section 2
'Y -0.2 -0.4 -0.6 -0.8 -0.2 -0.4 -0.6 -0.8
B = (lJ'~>SSla~k (14a) (1 ) (2) (3) (4) (5) (6) (7) (8) (9)
Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

(lJ')SSla=k 0.0 1.566 1.382 1.230 1.104 1.543 1.371 1.225 1.103
0.2 1.482 1.331 1.202 1.093 1.441 1.307 1.190 1.089
where (lJ'~)SSla~k = elastic critical buckling stress of a simply 1.417 1.290
0.4 1.178 1.080 1.369 1.260 1.162 1.076
supported single span tapered member with stress gradient, 0.6 1.366 1.256 1.158 1.072 1.316 1.224 1.155 1.065
i.e., unequal end stresses; and (lJ')SSla=k = critical buckling 0.8 1.323 1.227 1.140 1.063 1.275 1.195 1.122 1.055
stress of the same member without stress gradient, i.e., equal 1.0 1.287 1.201 1.124 1.054 1.241 1.171 1.106 1.046
end stresses. In case l(d) the moment at one end of the tapered 1.2 1.255 1.178 1.108 1.044 1.213 1.151 1.092 1.038
beam is zero, i.e., ex = O. Thus, B is defined as 1.4 1.226 1.158 1.094 1.035 1.189 1.133 1.079 1.029
1.6 1.200 1.139 1.081 1.027 1.168 1.116 1.067 1.022
1.8 1.178 1.122 1.069 1.019 1.149 1.102 1.057 1.015
B = (lJ'~>Ssla=o (14b) 2.0 1.157 1.106 1.057 1.010 1.132 1.088 1.046 1.008
(lJ'~)SSla=k 2.2 1.138 1.091 1.046 1.004 1.116 1.076 1.038 1.001
2.4 1.120 1.077 1.035 0.999 1.102 1.065 1.028 0.997
In cases 2(a) to 2(c), where lateral restraints are imposed to 2.6 1.104 1.064 1.025 0.997 1.088 1.053 1.019 0.994
tapered sections, the restraint factor R is used to account for 2.8 1.088 1.051 1.015 0.994 1.076 1.043 1.012 0.992
the restraining effect. Factor R is defined as 3.0 1.073 1.039 1.006 0.992 1.064 1.033 1.005 0.989

R = (lJ'~)LRla=k (15) TABLE 5. Factor B Obtained from BASP for Case 1(c)
(lJ'~>Ssla=k
Fb ,IFb2
where (lJ'~)LRla=k = elastic critical buckling stress of the critical
span in a laterally restrained tapered beam with zero stress Section 1 Section 2
gradient. 'Y -0.2 -0.4 -0.6 -0.8 -0.2 -0.4 -0.6 -0.8
(1 ) (2) (3) (4) (5) (6) (7) (8) (9)
NUMERICAL RESULTS AND DISCUSSION 0.0 1.566 1.383 1.231 1.106 1.543 1.371 1.226 1.103
0.2 1.661 1.437 1.258 1.115 1.666 1.444 1.263 1.117
Two typical tapered beams were considered in this study. 0.4 1.747 1.483 1.279 1.122 1.779 1.505 1.292 1.126
The cross-sectional dimensions at the small end are given in 0.6 1.824 1.522 1.296 1.125 1.882 1.557 1.314 1.132
Table 2. The two sections are identical to sections I and III 0.8 1.894 1.555 1.309 1.127 1.975 1.599 1.330 1.135
originally used by Lee et al. (1972) and Morrell and Lee 1.0 1.959 1.583 1.319 1.128 2.058 1.635 1.342 1.136
1.2 2.016 1.606 1.325 1.126 2.134 1.665 1.352 1.136
(1974). In the current analysis, the BASP finite-element pro- 1.4 2.069 1.626 1.330 1.124 2.201 1.688 1.358 1.135
gram was used, where the web was divided into quadrilateral 1.6 2.118 1.643 1.333 1.121 2.259 1.706 1.360 1.131
elements and the flange was modeled with beam elements. The 1.8 2.162 1.657 1.335 1.118 2.313 1.722 1.363 1.129
length of the tapered beam used in the analysis was L = 3,658 2.0 2.240 1.667 1.323 1.099 2.359 1.732 1.362 1.124
mm. The material properties were E = 200,000 MPa; v = 0.30; 2.2 2.277 1.673 1.319 1.092 2.403 1.743 1.362 1.121
2.4 2.332 1.669 1.299 1.071 2.439 1.749 1.359 1.116
2.6 2.359 1.668 1.292 1.066 2.472 1.754 1.357 1.111
TABLE 2. Dimension of Sections Used in Analysis 2.384 1.667 1.285 1.062 2.501 1.756 1.353 1.106
2.8
Dimension Section 1 Section 2 3.0 2.403 1.662 1.276 1.059 2.526 1.757 1.349 1.092
(1 ) (2) (3)
do (mm) 152.4 152.4 TABLE 6. Factor B Obtained from BASP for Case 1(d)
b (mm) 101.6 101.6
t[ (mm) 6.35 19.05 'Y Section 1 Section 2
tw (mm) 2.54 6.35 (1 ) (2) (3)
0.0 1.785 1.737
0.2 1.656 1.588
TABLE 3. Factor B Obtained from BASP for Case 1(a) 0.4 1.561 1.488
0.6 1.487 1.415
MolM,
0.8 1.428 1.360
Section 1 Section 2 1.0 1.379 1.316
1.2 1.336 1.280
'Y -0.2 -0.8 -0.2 -0.8
1.4 1.299 1.249
(1 ) (2) (3) (4) (5) 1.6 1.266 1.222
0.0 1.019 1.092 1.028 1.095 1.8 1.237 1.199
0.5 1.005 1.046 1.009 1.047 2.0 1.179 1.177
1.0 0.997 1.018 1.006 1.022 2.2 1.154 1.159
1.5 0.998 1.007 1.011 1.016 2.4 1.098 1.142
2.0 0.995 1.001 1.003 1.013 2.6 1.075 1.125
2.5 0.986 1.017 1.002 1.011 2.8 1.053 1.110
3.0 0.989 1.014 1.001 1.009 3.0 1.034 1.096

1212 I JOURNAL OF STRUCTURAL ENGINEERING I OCTOBER 1998

J. Struct. Eng. 1998.124:1208-1216.


TABLE 7 Factor R Obtained from BASP for Case 2(a)
M,/M2
Section 1 Section 2

"Y 0.0 -0.25 -0.50 -0.75 -1.0 0.0 -0.25 -0.50 -0.75 -1.0
(1 ) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11 )
0.00 1.363 1.316 1.258 1.185 1.120 1.354 1.292 1.242 1.171 1.088
0.25 1.361 1.319 1.268 1.206 1.136 1.362 1.307 1.234 1.158 1.068
0.40 1.357 1.316 1.267 1.209 1.142 1.386 1.329 1.262 1.183 1.098
0.50 1.333 1.299 1.257 1.205 1.144 1.396 1.342 1.279 1.206 1.125
0.57 1.337 1.299 1.255 1.201 1.138 1.393 1.345 1.288 1.219 1.144
0.63 - - - - - 1.379 1.339 1.287 1.225 1.153
0.67 - - - - - 1.349 1.319 1.276 1.219 1.153
Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

TABLE 8(a). Factor R Obtained from BASP for Case 2(b): TABLE 9(a). Factor R Obtained from BASP for Case 2(c):
Section 1 Section 1
Fb1 /Fb2 Fb1 /Fb2
"Y 1.0 0.75 0.50 0.25 0.0 -0.25 -0.50 -0.75 -1.0 "Y 1.0 0.75 0.50 0.25 0.0 -0.25 -0.50 -0.75
(1 ) (2) (3) (4) (5) (6) (7) (8) (9) (10) (1) (2) (3) (4) (5) (6) (7) (8) (9)
0.00 1.293 1.275 1.254 1.230 1.201 1.165 1.121 1.068 1.006 0.00 1.293 1.275 1.254 1.230 1.201 1.165 1.121 1.068
0.17 1.296 1.279 1.261 1.239 1.213 1.183 1.148 1.105 1.057 0.25 1.309 1.288 1.263 1.232 1.195 1.147 1.088 1.017
0.29 1.296 1.279 1.262 1.242 1.218 1.191 1.160 1.124 1.084 0.50 1.316 1.295 1.267 1.233 1.190 1.136 1.069 0.991
0.38 1.292 1.276 1.259 1.240 1.219 1.194 1.166 1.135 1.099 0.75 1.319 1.300 1.274 1.238 1.194 1.136 1.064 0.979
0.44 1.285 1.271 1.255 1.237 1.216 1.194 1.169 1.141 1.106 1.00 1.359 1.326 1.297 1.258 1.207 1.141 1.058 1.962
0.50 1.272 1.260 1.247 1.232 1.215 1.197 1.176 1.154 1.119 1.25 1.349 1.342 1.314 1.275 1.222 1.152 1.063 0.953
0.55 1.267 1.255 1.243 1.229 1.213 1.197 1.178 1.157 1.119 1.50 1.409 1.381 1.349 1.305 1.245 1.164 1.062 0.949
0.58 1.261 1.251 1.239 1.226 1.211 1.196 1.178 1.159 1.119 1.75 1.434 1.404 1.374 1.328 1.265 1.180 1.073 0.955
0.62 1.254 1.245 1.235 1.224 1.213 1.200 1.187 1.172 1.127 2.00 1.454 1.426 1.398 1.351 1.287 1.197 1.085 0.962
0.64 1.250 1.242 1.232 1.223 1.212 1.200 1.188 1.174 1.126 2.25 1.504 1.479 1.458 1.396 1.321 1.218 1.088 0.971
0.67 1.247 1.239 1.229 1.221 1.211 1.199 1.188 1.176 1.124 2.50 1.536 1.505 1.477 1.424 1.347 1.239 1.104 0.962
2.75 1.609 1.564 1.533 1.475 1.389 1.266 1.112 0.955
3.00 1.638 1.592 1.567 1.507 1.418 1.291 1.131 0.969
TABLE 8(b). Factor R Obtained from BASP for Case 2(b):
Section 2
TABLE 9(b). Factor R Obtained from BASP for Case 2(c):
Fb ,/Fb2 Section 2
"Y 1.0 0.75 0.50 0.25 0.0 -0.25 -0.50 -0.75 -1.0
Fb ,/Fb2
(1 ) (2) (3) (4) (5) (6) (7) (8) (9) (10)
"Y 1.0 0.75 0.50 0.25 0.0 -0.25 -0.50 -0.75
0.00 1.265 1.250 1.233 1.219 1.189 1.159 1.122 1.072 1.008
0.17 1.264 1.252 1.239 1.219 1.188 1.169 1.147 1.119 1.084 (1) (2) (3) (4) (5) (6) (7) (8) (9)
0.29 1.263 1.249 1.239 1.221 1.194 1.179 1.162 1.142 1.120 0.00 1.265 1.250 1.233 1.217 1.189 1.159 1.122 1.072
0.38 1.262 1.243 1.239 1.231 1.202 1.189 1.175 1.159 1.144 0.25 1.291 1.272 1.249 1.221 1.184 1.135 1.067 0.976
0.44 1.258 1.239 1.236 1.221 1.211 1.199 1.187 1.175 1.163 0.50 1.309 1.286 1.258 1.222 1.174 1.107 1.014 0.901
0.50 1.254 1.246 1.238 1.229 1.219 1.209 1.198 1.188 1.180 0.75 1.320 1.294 1.262 1.219 1.163 1.082 0.974 0.854
0.55 1.261 1.253 1.245 1.237 1.228 1.218 1.209 1.199 1.196 1.00 1.329 1.298 1.262 1.216 1.153 1.065 0.949 0.825
0.58 1.266 1.259 1.251 1.243 1.234 1.226 1.218 1.209 1.204 1.25 1.326 1.299 1.261 1.212 1.147 1.054 0.933 0.808
0.62 1.270 1.263 1.256 1.248 1.240 1.233 1.225 1.217 1.209 1.50 1.348 1.307 1.269 1.220 1.152 1.055 0.930 0.803
0.64 1.274 1.267 1.260 1.253 1.245 1.239 1.231 1.224 1.213 1.75 1.362 1.316 1.280 1.231 1.159 1.059 0.932 0.804
0.67 1.277 1.270 1.263 1.256 1.249 1.243 1.236 1.229 1.214 2.00 1.378 1.325 1.292 1.241 1.169 1.067 0.937 0.807
2.25 1.395 1.333 1.303 1.253 1.179 1.075 0.944 0.812
2.50 1.412 1.338 1.315 1.264 1.191 1.085 0.953 0.819
practical situations. The results for the factor B obtained 2.75 1.447 1.374 1.349 1.295 1.216 1.103 0.961 0.822
through the BASP program using (14a) and (14b) are listed 3.00 1.467 1.379 1.363 1.310 1.231 1.115 0.972 0.829
in Tables 3-6 for the four cases, respectively.
Regarding case I(a), one can see that the results obtained
where the factor Cf is defined as
from the BASP program were approximately equal to I, i.e.,
the factor was not affected by changes of the taper ratio. Thus,
it is recommended that, for case I(a), Cf =1 + FbI (19)
Fb2
B = 1.0 (16)
In (19), the term Fb /Fb2 is considered negative when the mem-
For cases I(b) and l(c), the regression method and curve- ber is bent in single curvature.
fitting technique (least-square method) were used on the results In case I(d), the moment gradient is equal to zero, i.e., a =
obtained from the BASP program, relating these values to the O. Factor B can be obtained using (17) for F b /Fb2 = O. Thus,
consistent values for prismatic members. The stress gradient for case I(d),
factor B for these cases is obtained as a function of the taper
ratio "/. For case I(b), B = 1.64 - 0.3823oy + 0.066oy2 2: 1.0 (20)

B = 1 + O.64Cf - 0.3823,,/Cf + 0.066,,/2Cf 2: 1 (17) For the development of the restraint factor R, the same two
typical sections were used. The length of the tapered beams
For case I(c), was L = 7,315 mm. The material properties were E = 200,000
MPa; v = 0.30; and Fy = 350 MPa. The height-to-thickness
B= 1 + 0.55Cf + (0.175 + OA54,,/)C; - 0.0425,,/2Cf 2: 1 (18) ratio was adjusted to prevent local buckling. The stresslmo-
JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 1998/1213

J. Struct. Eng. 1998.124:1208-1216.


ment ratio varied from -0.2 to 0.8 and the taper ratio varied For case 2(a),
from 0 to 3. Again, these ranges cover the majority of practical
situations. The results for the factor R obtained through the R:l + 0.403Df + (0.300 - 0.223Df h ' - 0.331·/Dj 2: 1 (21)
BASP program using (IS) are listed in Tables 7, 8(a), 8(b), For case 2(b),
9(a), and 9(b) for the three cases discussed earlier.
For cases 2(a), 2(b), and 2(c), the regression method and R: 1.016 + -y(0.29 - 0.181-y) + CiO.207 - 0.031Cj )
curve-fitting technique (least-square method) were used on the
results obtained from the BASP program. The restraint factor + -yCi-O.15 + 0.016-yCj ) 2: 1 (22)
R for these cases was also obtained as a function of the taper
ratio 'Y and is given by (21), (22), and (23), respectively, as For case 2(c),
follows. R: 0.94 + -y(-0.267 + 0.053-y) + Cf (0.371 - 0.109Cf )

1.7 . , . . . . . - - - - - - - - - - - - - - - - - - - . + -yCf (0.176 - 0.018-yCf ) 2: 1 (23)


Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

-Eq.(17)
a:l
j 1.5
• FbltFb2=-o.2 (BASP)
... FbltFb2=-OA (BASP)
• FbltFb2~0.8 (BASP)
where Cf is computed from (19), and Df is defined as

I
M
Df : 1 +-1 (24)
M2
1.3
Figs. 7-9 show the numerical results for the stress gradient
.~
v,
I.l f:::~0r-~:::::;:~·:=:·~.~=t
Fb 11Fb2=-0.8 • • • •
factor B obtained through the BASP program as well as the
proposed theoretical values versus the taper ratio 'Y, while Figs.
0.9 L=::;:::~_- .........- _ - -__l 10-12 show the numerical and proposed values for restraint
o 0.5 1.5 2 2.5 3
Taper Ratio
1.8 ~-----------------____,
(a)
• BASP , Section #1

1.7 -r-----------------__, -Eq.(17) a:l 1.6


• BASP, Section #2

FbI1Fb2-0.2 • FbltFb2-0.2 (BASP)


!
l:Q

1.5
.. Fb ltFb2=-o.4 (BASP)
• Fb lIFb2-.Q.8 (BASP)
i
~
AISC-LRFD (1994)

-=
~ 1.3 ~1.4

j l.l f:.-::.~.-."..r-,r~=:::::=~~=;:~~ B1.2


< Il
FblIFb2=-0.8 • • • •
0.9 ~==:;::~-,__--r--__,--..,......-~ • •
o 0.5 1.5 2 2.5 • •
Taper Ratio II-------,.--......-----,----...---.,...---!
(b) o 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
Taper Ratio
FIG. 7. Stress Gradient Factor Bfor Case 1(b): (a) Section 1;
(b) Section 2 FIG. 9. Stress Gradient Factor Bfor Case 1(d)

. • . · •
1.7 1.8

1.6
FbI/Fb2=-0.4
· •· 1.7 ~(21)
M 1M2-.Q.25 reASP)
MIIM2-.Q.5 ~ ASP~
l:Q
- Eq.(18)
ll'i
1.6
"• MIIM2-.Q.7 (BAS )

!il 1.5
·.
FbI/Fb2~OA (BASP)
FbltFb2-0.6 (BASP) .9 1.5
MIIM2=-o.5
MIIM2=-o.25

g
1.4

1.3
FbltFb2=-0.8 (BASP)

• • •
~
'j
1.4
1.3
MIIM2=-0.75

I .. ..
\ /
j 1.2
FbI1Fb2=-0.8
llI:
1.2
l.l •
'" l.l .. . .. .. .. 1
1 0.9
2 0 0.1 0.2 0.3 0.4 0.5 0.6
0 0.5 1.5 2.5 3
Taper Ratio ofthe Critical Span
Taper Ratio
(a) (a)

2.1
1.8 ...-_-=---",.,..,..- -,
Eq.(18) - &l.(21)
til 1.9 • FbI1Fb2=-o.4 ~BASP~
FbI1Fb2=·0.6 BASP FbI1Fb2=·0.4
1.7 •
...
M11M2~0.2S (BASP)
MIIM2=-0.5 (BASP)

j 1.7
A.
.•
FbI1Fb2--0.8 (BASP)
··• ll'i
.9
1.6
1.5
• MIIM2=-o.7S (BASP)
M11M2~0.75 MllM2-.Q.5 MllM2=-o.25

ij 1.5

1.3 .• .•
Fb IIFb2-.Q.6
• · ~ 1.4t===\~~\~-~·~'\~·E·:i
.~ 1.3
~ 1.2
\
~
.. ..

llI: 1.1
l.l
I
0.9 +---,---.,.....--.-----,..------.---1 0.9 +--.....,..--,--.....,..--.,.....-........--..---1
o 0.5 1.5 2 2.5 o 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Taper Ratio Taper Ratio ofthe Critical Span
(b) (b)

FIG. 8. Stress Gradient Factor B for Case 1(c): (a) Section 1; FIG. 10. Restraint Factor R for Case 2(a): (a) Section 1;
(b) Section 2 (b) Section 2

1214/ JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 1998

J. Struct. Eng. 1998.124:1208-1216.


1.6 3.926N/mm
_ Eq.~! 17 $ Z $2 S 257 S 252 S 7$2 S? \? S? V $1
1.5 • Flil ~.5 rnAS~
• Fbl1Fb2-o.0 BAS
• Fb11Fb2=-O.5 (BASP)
Ill: 1.4 Fb11Fb2=O.5
j 1.3 /
Fb11Fb2=O.0

--"7
Fb11Fb2=-O.5

'1
Ill:
1.2 •
1.1 >-

I
o 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Taper Ratio of the Critical Span
(a)
1,,4_ _9_144_IDID -._I+4_ _9_144_1DID -+l_1
Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

1.7 FIG. 13. Tapered Frame and Its Dimensions


Eq. (22)
• Fb11Fb2=O.5 (BASP~

Ij
1.6

~rf- j--
• FbI1Fb2~.0 ffiASP
~ FbI1Fb2--o.5 (BAS )
Ill: 1.5

i
I><

·1
1.4
1.3
Fb11Fb2=O.5

/• /
Fb11Fb2=O.0 FbI1Fb2--Q.5
...--
- ..
u_-j
~ 1.2
1.1
* •
• • ·· 1~27~1
I I I I 1~27~1
I 2 3 4 5 6
1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 I_ 5@1981.2mm .\
Taper Ratio of the Critical Span

~~
(b)

FIG. 11. Restraint Factor R for Case 2(b): (a) Section 1; MomentDiagram
(b) Section 2
FIG. 14. Tapered Roof I-Beam and Its Moment Diagram
2~----------------.....,
_ Eq.(23)
1.8 •• FbI/Fb2=O.5 (BASP)
Fb11Fb2=O.0 (BASP)
span (braced point 5 to 6) will be considered (see Fig. 14).
The taper ratio for this span is 'Y = 0.09, and the radius of
Ill: ~ FbI1Fb2~0.5 (BASP)
1.6
gyration rro = 26.0 mm at point 5. Since this span is at the
~
Fb11Fb2=O.0
FbI/Fb2=O.5
• •
I>< 1.4 • end, case 2(b) or 2(c) applies for the determination of the
• restraint factor R. First, the following quantities are computed:
'1
Ill:
1.2
F b1 lFb2 = 2.11(176.0) = 0.012
FbI/Fb2=-0.5 .--r 0: = M4 1M6 = (10.68)11,270.92 = 0.008
0.8
0 0.5 I 1.5 2 2.5 3
Taper Ratio of the Critical Span k = 8 4 186 = 4.75 x 10'/6.11 x 10' = 0.78
(a) Since 0: < k, case 2(b) applies. According to (19) and (22)
2 Cf =1 + 0.012
- Eq.(23)
1.8 •• Fb I/Fb2=0.5 ~BASP~
Fbl/Fb2=O.0 HASP R = 1.016 + 0.29 X 0.09 + 0.207 X 1.012 - 0.15 X 1.012
Ill: ~ Fbl/Fb2=-0.5 (BAS)
1.6 X 0.09 - 0.181 X (0.09)2 - 0.031 X (1.012)2 + 0.016
§ 1.4
Fb11Fb2=O.5 Fbl/Fb2=O.0 Fb l/Fb2=·0.5

X (0.09)2 X (1.012)2 = 1.205


'1
l>:i
1.2
To compute Fb'Y the following properties are needed:
Llrro = 1,981.2/26.0 = 76.2
0.8 +---..----..----r----,..---.......- - - - 1
o 0.5 I I.S 2 2.5 3 hw = 1 + 0.00385 X 0.09 X (76.2)112 = 1.003
Taper Ratio of the Critical Span
(b) LdolAf =(1.981.2 X 517.5)/(4.76 X 127) = 1,696.0
FIG. 12. Restraint Factor R for Case 2(c): (a) Section 1;
(b) Section 2 hs = 1.0 + 0.023 X 0.09 X 1,696.01/2 = 1.085
3
factor R. The use of the proposed factors Band R is demon- 83 X 10
strated in the following example. Fs'Y = 1.085 X 1,696.0 = 45.1 MPa

1,170 X 103
Design Example for Use of Restraint Factor R Fw'Y = (1.003 X 76.2)2 = 200.3 MPa
Consider the frame shown in Fig. 13, the members of which
are build-up tapered I-shapes with Fy = 300 MPa. Purlins are Thus
attached to the top flange and additional braces are dropped Fb'Y = RVF~'Y + F~'Y = 1.205V45.1 2 + 200.3 2 = 247.4 MPa
from the purlins to support the compression bottom flange. For
purposes of illustration, only the bending stress for the deepest Since Fb'Y > 1I3Fy , (4b) applies. That is,
JOURNAL OF STRUCTURAL ENGINEERING / OCTOBER 1998 /1215

J. Struct. Eng. 1998.124:1208-1216.


~ (I - 6RYF~ + F~J Fy = ~ (I - 6 :~7.4) 300
Hartman, A. J. (1967). "Elastic lateral buckling of continuous beams."
Fb-y = J. Struct. Div., ASCE, 93(4), 11-28.
Horne, M. T., Shakir-Khalil, H., and Akhtar, S. (1979). "Tests on tapered
and haunched beams." Proc., Inst. Civ. Engrg., 67(2), 846-850.
= 159.6 MPa Kitipornchai, S., and Trahair, N. S. (1972). "Elastic stability of tapered
I-beams." J. Struct. Div., ASCE, 98(3), 713-728.
Since Fb-y < O.6Fy , the allowable bending stress is 159.6 MPa. Kitipornchai, S., and Trahair, N. S. (1975). "Elastic behavior of tapered
According to the AISC specification (Manual 1994), nonsymmetric I-beams." J. Struct. Div., ASCE, 101(8), 1661-1678.
Kosko, E. (1982). "Uniform element modeling of tapered frame mem-
B = I + 0.58 X (I + 0.012) - 0.70 X 0.09 X (I + 0.012) bers." J. Struct. Div., ASCE, 108(1),245-264.
Krefeld, W. J., Butler, D. J., and Anderson, G. B. (1959). "Welded can-
= 1.523 ~ 1.0 tilever wedge beams." Welding Res. J. Supplement, 38(3), 1-97.
Lee, G. C., and Morrell, M. L. (1975). "Application of AISC design
Thus provisions for tapered members." AISC Engrg. J., 12(1), 1-13.
Lee, G. c., Morrell, M. L., and Ketter, R. L. (1972). "Design of tapered
Fb-y = BYF:-y + F~-y = 1.523Y45.1 2 + 200.3 2 = 312.7 MPa
Downloaded from ascelibrary.org by Univ. of Alabama At Birmingham on 10/04/13. Copyright ASCE. For personal use only; all rights reserved.

members." Welding Res. Council Bull., No. 173, June, 1-32.


Since F b > l/3Fy , (4b) applies: Manual of steel construction-LRFD. (1994). 2nd Ed., American Insti-
tute of Steel Construction, Chicago, Ill.

F - ~
boy - 3
(I - Fy
6BYF:-y + F~-y
) F - ~
y - 3
(I _ 300 ) 300
6 X 312.7
Morrell, M. L., and Lee, G. C. (1974). "Allowable stress for web-tapered
beams with lateral restraints." Welding Res. Council Bull., No. 192,
February, 1-10.
Nethercot, D. A. (1983). "Elastic lateral buckling of beams." Beams and
= 168.0 MPa Beam Columns-Stability in Strength, R. Narayanan, ed., Applied Sci-
ence Publishers, Barking, U.K., 17-21.
Since Fb-y < O.6Fy , the allowable bending stress according to Polyzois, D., and Li, Q. (1993). "Stability of web-tapered beams." Proc.,
the AISC specifications is 168 MPa, that is, 5.3% higher than Struct. Stability Res. Council, 179-192.
the proposed value. Prawel, S. P., Morrell, M. L., and Lee, G. C. (1974). "Bending and buck-
ling strength of tapered structural members." Welding Res. J. Supple-
ment, 53(2), 1-75.
CONCLUSIONS Salvadori, M. G. (1956). "Lateral buckling of eccentrically loaded 1-
columns." Trans. ASCE, 121, 1163-1178.
In this paper, the current design specifications for web-ta- Trahair, N. S. (1977). "Lateral buckling of beams and beam-columns."
pered beams were reviewed and the background to these Theory of Beam-Columns, Vol. 2, W. F. Chen and T. Atsuta, eds., Mc-
guidelines was investigated. Six specific cases of loading and Graw-Hili Book Co. Inc., New York, 71-157.
restraining conditions are presented. The use of one modifi- Wekezer, J. W. (1985). "Instability of thin walled bars." J. Engrg. Mech.,
cation factor to account for both the moment or stress gradient ASCE, 111(7), 923-935.
Yang, Y. B., and Yau, J. D. (1987). "Stability of beams with tapered 1-
and the restraint provided by adjacent segments in laterally sections." J. Engrg. Mech., ASCE, 113(9), 1337 -1357.
braced beams was questioned. Based on the results obtained
from a thorough finite-element analysis, separate expressions APPENDIX II. NOTATION
for the stress gradient factor B and the restraint factor R were
The following symbols are used in this paper:
developed and recommendations for design considerations
were proposed. Af = flange area;
B = stress gradient factor;
ACKNOWLEDGMENT Cf = end stress ratio factor;
Df = end moment ratio factor;
The writers would like to express their appreciation to Qing Li for d = cross-sectional depth;
carrying out the major part of the computer analysis. h = cross-sectional height;
k = section modulus ratio;
APPENDIX I. REFERENCES L = beam length;
Fb-y = allowable lateral torsional buckling stress;
Akay, H. U., Johnson, C. P., and Will, K. M. (1977). "Elastic and local
buckling of beams and frames." J. Struct. Div., ASCE, 103(9),1821-
F,-y = allowable Saint-Venant's torsional resistance;
1832. Fw-y allowable warping resistance;
Boley, B. A. (1963). "On the accuracy of the Bernoulli-Euler theory for Fy yield stress;
beams of variable section." J. Appl. Mech., 30, 373-378. M = end moment;
Bradford, M. A., and Cuk, P. E. (1988). "Elastic buckling of tapered R = restraint factor;
monosymmetric I-beams." J. Struct. Engrg., ASCE, 114(5),977 -996. r = radius of gyration;
Brown, T. G. (1981). "Lateral-torsional buckling of tapered I-beams." J. rTo = radius of gymtion of section at small end, considering only
Struct. Div., ASCE, 107(4),689-697. compression flange plus 113 of web in compression;
Butler, D. J. (1966). "Elastic buckling tests on laterally and torsionally Sx section modulus;
braced tapered beams." Welding Res. J. Supplement, 45(1), 1-41.
IX = moment ratio;
Butler, D. J., and Anderson, G. B. (1963). "The elastic buckling of ta-
pered beam-columns." Welding Res. J. Supplement, 42(1), 1-29. "Y = taper ratio;
Chan, S.-L. (1990). "Buckling analysis of structures composed of tapered V Poisson's ratio;
members." J. Struct. Div., ASCE, 116(7), 1893-1906. (CT-y)LR buckling stress of laterally restrained tapered beam;
Culver, C. G., and Preg. S. M. (1968). "Elastic stability of tapered beam- (CT-y>Ss = buckling stress of simply supported tapered beam; and
columns." J. Struct. Div., ASCE, 94(2), 455-470. <l>b = bending resistance factor (=0.85).

1216/ JOURNAL OF STRUCTURAL ENGINEERING I OCTOBER 1998

J. Struct. Eng. 1998.124:1208-1216.

You might also like