Professional Documents
Culture Documents
Lateral-Torsional Buckling of Steel Beams PDF
Lateral-Torsional Buckling of Steel Beams PDF
D. Mateescu1, V. Ungureanu1
ABSTRACT
The slender members loaded by transversal loads or ended moments acting around the
major axis of inertia, may collapse by lateral-torsional buckling before reaching the full
plastic resistant moment, Mpl. The present paper presents a comparison between Mateescu
proposal [1], Eurocode 3-Part 1.1 [2] and the ECBL approach [3], used to calculate the
ultimate lateral-torsional buckling moment. The experimental database from Eurocode 3 –
Background Documentation, Chapter 5 October 1989 [4] was used to evaluate the theoretical
results.
1. INTRODUCTION
In the final version of Eurocode 3-Part 1.1 [2] there exist two different sets of LT-
buckling curves:
- in paragraph 6.3.2.2: Lateral-torsional buckling curves - General case, the column
buckling curves a, b, c, d are specified for cross-section groups h/b , 2 and h/b>2
of rolled and welded sections, with a plateau of LT 0.2 ;
- in paragraphs 6.3.2.3: Lateral-torsional buckling curves for rolled sections or
equivalent welded sections, specific LT-buckling curves b, c, d are given for the
groups h/b,2 and h/b>2 of rolled and welded sections, and in contrary to
paragraph 6.3.2.2 with a plateau of LT 0.4 . The LT-buckling curves given in
6.3.2.3 are based on numerical simulations of single span beams under uniform
moment with idealized end-fork conditions [5,6].
Mateescu has proposed a similar method with the second one of Eurocode 3-Part 1.1,
more than ten years before [1]. Consequently, a comparison of Mateescu proposal with the
two sets of LT-buckling curves from Eurocode 3-Part 1.1 is presented in this paper. In
addition, for comparison the LT-buckling curves obtained with the Erosion of Critical
Bifurcation Load (ECBL) approach, developed by Dubina is shown.
1
Romanian Academy, Timisoara Branch, Laboratory of Steel Structures, M. Viteazul 24, Timisoara, Romania
2. LATERAL-TORSIONAL BUCKLING OF BEAMS IN BENDING ACCORDING
TO EUROCODE 3-PART 1.1
The design buckling resistance moment of a laterally unrestrained beam should be taken
as:
M b,Rd = LT Wy f y / M1 (2)
where
1
LT = 2 1/ 2
1 (3)
2
LT +[ LT LT ]
2
LT = 0.5[1 + LT ( LT 0 .2 ) + LT ] (4)
LT = Wyf y / M cr (5)
where
Wy is the appropriate section modulus as follows:
– Wy = Wpl,y for Class 1 or 2 cross-sections;
– Wy = Wel,y for Class 3 cross-sections;
– Wy = Weff,y for Class 4 cross-sections;
LT is the reduction factor for lateral-torsional buckling;
Mcr is the elastic critical moment for lateral-torsional buckling of the gross cross-section;
LT is the imperfection factor.
For the reduced slenderness LT 0.2 (the case of short beams), lateral-torsional
buckling effects may be ignored and only cross-sectional checking apply.
As an alternative, for rolled or equivalent welded sections in bending the values of LT
for the appropriate non-dimensional slenderness may be determined from:
LT 1
1
LT = 2 2 1/ 2
but 1 (6)
LT + [ LT LT ] LT 2
LT
2
LT = 0.5[1 + LT ( LT LT,0 ) + LT ] (7)
The following values are recommended for rolled sections: LT,0 = 0.4 (maximum
value) and F = 0.75 (minimum value).
The recommendations for buckling curves are given in Table 3.
It is important to underline that the new values of LT coefficient have been evaluated
using the ECCS experimental database [4].
On the purpose of avoiding the discontinuity in the lateral-torsional buckling curve of
beams, as it was the case of ENV version of Eurocode 3-Part 1.1, Mateescu, at that time,
suggested the following formula for LT, but with the imperfection coefficient LT=0.27, for
hot-rolled I beams and LT=0.60, for welded I beams:
2
LT = 0.5[1 + LT ( LT 0 .4 ) + LT ] (8)
By using this formula, to calculate the LT factor, the jump for LT = 0.4 will be
eliminated and, evidently, LT = 1 will be obtained.
The Erosion of Critical Bifurcation Load (ECBL) approach, developed by Dubina [3], is
a method where the erosion of the critical bifurcation load of a steel member (owing to the
presence of imperfections as well as to the coupling of instability modes) is quantified by
means of an erosion factor, LT.
The non-dimensional moment M LT , given by equation (9) represents a solution of the
Ayrton-Perry formula, including the generalised imperfection coefficient,
LT = LT ( LT 0.4 ) :
2
1+ LT( LT 0.4) + LT 1 2 2 2
MLT = 2 2
[1 + LT( LT 0.4) + LT] 4 LT (9)
2 LT 2 LT
The formula which link LT factor with previously defined LT factor is:
2
= LT
(10)
0.6 (1 LT )
LT
Thus, by calibrating LT factor, the resultingLT values may be obtained for series of
particular steel sections.
The ECBL approach for lateral-torsional buckling of beams is similar to that of
Eurocode 3-Part 1.1, but in eqn. (4) is used a different generalised imperfection coefficient
instead of the related formula given in the code. It means the LT formula becomes:
2
LT = 0.5[1 + LT ( LT 0.4) + LT ] (11)
and LT should be calculated from eqn. (10) depending on LT erosion factor which has to be
evaluated by statistical processing of relevant test specimens.
There are two practical ways that can be used to evaluate the LT erosion factor: (1) the
experimental procedure; (2) the numerical approach. In the present paper the experimental
“mean” approach is used.
Given a specimen series characterized by the same nominal properties, the design value
of the erosion factor results from:
LT = m + 1.64s (12)
Mi , exp
in which s is the standard deviation related to LT,i = 1 Mi , exp , where M i, exp = and
Mi, pl
1 n
m = (1 M i, exp ) values for all n specimens.
n i =1
As an alternative to the “mean” approach, the Annex D of EN1990 [7] (former Annex Z
of Eurocode 3 in the ENV version) can be used for the experimental calibration of LT and
LT factors [8].
Table 6 shows the statistical results for hot-rolled and welded I beams. Figures 1 and 2
show the comparison between tests and the numerical results related to Eurocode 3 formulas,
Mateescu proposal and ECBL approach. For the case of ECBL approach, the interactive
slenderness range was assumed to be LT ± = LT ± 0.20 , and a scattering value of 50% is
usual for the experimental values in the field of structural engineering tests.
Table 6. Statistical values for hot-rolled and welded I-beams
I Hot-rolled Beams (144 Tests)
Method Eurocode 3 Part 1.1 Eurocode 3 Part 1.1
Mateescu ECBL
Method 1 Method 2
Statistical
LT=0.21 LT=0.34 LT=0.34 LT=0.49 LT=0.27 LT=0.185
parameters
m 1.178 1.286 1.106 1.190 1.165 1.108
s 0.100 0.132 0.077 0.110 0.116 0.096
m-1.64s 1.014 1.070 0.979 1.009 0.975 0.952
v 0.085 0.103 0.070 0.093 0.100 0.086
0.963 0.944 0.959 0.939 0.958 0.968
I Welded Beams (71 Tests)
LT=0.49 LT=0.76 LT=0.49 LT=0.76 LT=0.60 LT=0.583
m 1.188 1.313 1.055 1.142 1.150 1.144
s 0.200 0.279 0.125 0.182 0.201 0.197
m-1.64s 0.860 0.855 0.850 0.844 0.819 0.820
v 0.168 0.213 0.119 0.159 0.175 0.172
0.876 0.857 0.893 0.890 0.892 0.893
1.0
Mb,Rd-EC3_vers2 ( LT=0.34)
Mb,Rd-EC3_vers1 ( LT=0.21)
0.6
experiments
Mb,Rd-EC3_vers1 ( LT=0.34)
0.4
Mb,Rd-EC3_vers2 ( LT=0.49)
0.2
0.0
0.00 0.50 1.00 1.50 2.00 2.50
Fig. 1. Numerical/Experimental comparison for hot-rolled I beams
7. CONCLUDING REMARKS
For the case of hot-rolled I-beams, it can be seen from Table 6 that good correlation
values were obtained for all methods. However, Georgescu & Dubina shown in [8] the studied
hot-rolled profiles frame all on the buckling curve “a” LT max
(
= 0.21 , which does not comply)
with the classification proposal for hot-rolled profiles used by second method of Eurocode 3-
Part 1.1, presented in the last column of Table 3. For the case of welded I-beams, correlation
values are still good (see Table 6).
From Figures 1 and 2 it can be seen that all curves fit well the experimental values.
However, it can be seen from Figure 1 that for short hot-rolled I beams the first method of
Eurocode 3 cover safety the range, using a safety factor M1=1. For the other curves, using a
safety factor M1=1.1 the range of short beams is in the safe side. In what concern the range of
long and medium length it can be seen that the first method of Eurocode 3 is too conservative,
while the curves obtained with Mateescu proposal, second method of Eurocode 3 and ECBL
one, cover well the whole range. Also, the curve obtained with Mateescu proposal fit very
well with the second method of Eurocode 3 (using an imperfection factor LT=0.34). The
curve obtained with the ECBL approach covers very well the range of medium length.
From Figure 2, for welded I beams the curves obtained with the first method of
Eurocode 3 is too conservative. The curves obtained with Mateescu proposal, the second
method of Eurocode 3 (with LT=0.76) and ECBL one, fit well the experimental results. The
curve obtained with the second method of Eurocode 3 (with LT=0.49) need to be affected
with a safety factor M1=1.1.
1.2
MMateescu ( LT=0.60)
1.0
MECBL ( LT=0.442, LT=0.583)
0.8
Mb,Rd-EC3_vers2 ( LT=0.49)
0.4
Mb,Rd-EC3_vers1 ( LT=0.76)
0.0
0.0 0.5 1.0 1.5 2.0 2.5
Fig. 2. Numerical/Experimental comparison for welded I beams
REFERENCES
[1] Mateescu, D.: Considerations on the value of reduction factor of lateral-torsional buckling
of beams in bending. Thin-Walled Structures, Vol. 20 (No. 1-4), 1994, 265-277.
[2] Eurocode 3: Design of steel structures. Part 1-1: General rules and rules for buildings
(EN1993-1-1). European Committee for Standardisation. 19 May 2003.
[3] Dubina, D.: The ECBL Approach for interactive buckling of thin-walled steel members.
Steel and Composite Structures, Vol. 1, no.1, 2001, 76-96.
[4] Eurocode 3: Background Documentation (1989)-Chapter 5, Document 5.03: Evaluation of
the test results on beam with cross-sectional classes 1-3 in order to obtain strength
functions and suitable model factors. October 1989.
[5] Greiner, R., Salzgeber, G., Ofner, R.: New lateral-torsional buckling curves LT –
numerical simulations and design formulae, ECCS Report 30. June 2000.
[6] Greiner, R., Kaim, P.: Comparison of LT-buckling curves with test-results. Supplementary
report. ECCS TC 8, TC 8-2003, May 2003, Graz University of Technology.
[7] EN 1990: Eurocode – Basis of structural design. European Committee for Standardisation.
July 2001.
[8] Georgescu, M., Dubina, D.: Lateral-torsional buckling of steel beams: A proposal to
calibrate the coefficients in the ECCS TC8 formula. Proceedings of the 1st International
Conference on STEEL & COMPOSITE STRUCTURES, 14-16 June 2001, Pusan,
KOREA, Vol. 1, 623-630.