Civil Engineering Reference
In-Depth Information
(length by width of S4R element) ratio provides adequate accuracy in
modeling the web, while a finer mesh of approximately 25 75 mm was
used in the flange (see Figure 6.2 ) .
The hinged support of T1, shown in Figure 6.2 , was prevented from dis-
placement in the horizontal direction (direction 1-1 in Figure 6.2 ) and the
vertical direction (direction 3-3 in Figure 6.2 ) . On the other hand, the roller
support of T1, shown in Figure 6.2 , was prevented from displacement in the
vertical direction only (direction 3-3 in Figure 6.2 ) . To account for the lat-
eral restraints of the compression flange, the top compression flange was pre-
vented from lateral displacements, in direction 2-2 of Figure 6.2 , at the end
supports, which is identical to the test T1. The load was applied in incre-
ments as concentrated static load, which is also identical to the experimental
investigation. The nonlinear geometry was included to deal with the large
displacement analysis.
The stress-strain curve for the structural steel given in the EC3 [2.11] was
adopted in this study with measured values of the yield stress ( f ys ) and ulti-
mate stress ( f us ) used in the tests [ 6.27 ]. The material behavior provided by
ABAQUS [1.29] (using the PLASTIC option) allows a nonlinear stress-
strain curve to be used (see Section 5.4.2 of Chapter 5 ). The first part of
the nonlinear curve represents the elastic part up to the proportional limit
stress with Young's modulus of ( E ) 200 GPa and Poisson's ratio of 0.3 used
in the finite element model. Since the buckling analysis involves large inelas-
tic strains, the nominal (engineering) static stress-strain curves were con-
verted to true stress and logarithmic plastic true strain curves as detailed
in Section 5.4.2.
Previous investigations by the author have successfully modeled the ini-
tial geometric imperfections in steel beams [ 6.28 , 6.29 ] . Buckling of steel
beams depends on the lateral restraint conditions to compression flange
and geometry of the beams. Mainly two buckling modes detailed in [ 6.30 ,
6.31 ] could be identified as unrestrained and restrained lateral-distortional
buckling modes. The lateral-distortional buckling modes were successfully
predicted by the author [ 6.28 , 6.29 ] by performing eigenvalue buckling anal-
ysis (see Section 5.5.2 of Chapter 5 ) for the investigated steel beams with
actual geometry and actual lateral restraint conditions to the compression
flange. The same approach [ 6.28 , 6.29 ] was followed in this topic to model
initial geometric imperfections of the plate girder investigated T1. Figure 6.3
shows the buckling mode predicted from the eigenvalue buckling analysis
detailed in ABAQUS [1.29]. Only the first buckling mode (eigenmode 1)
is used in the eigenvalue analysis. Since buckling modes predicted by
Search WWH ::




Custom Search