Tension buckling in ABAQUS

[pfsm01]

Hello,

I am trying to obtain the tension buckling modes in an edge-cracked plate using ABAQUS. If I run a regular stress analysis there are certainly compressive stresses in the direction perpendicular to the applied tensile in the region adjacent to the edge crack. However when I run a linear perturbation step, ABAQUS outputs the compression buckling modes (negative eigenvalues). Is there any step I'm missing?

Any help would be much appreciated.

 

[rb1957]

how does something buckle in tension ?

are you looking at the far (uncracked) edge of the plate ? looking at the compression side of the bending in the plate ??

it'd be an odd geometry that'd have a stable crack and compression buckling (of the far edge) ...

 

[pfsm01]

tension buckling will only occur in the presence of some type of flaw such as a crack. usually appears as complex wrinkling.

i'm looking at a tension flange stem in an I-shaped plate girder, where fatigue cracks can form at the fillet weld and propagate towards the outer tip of the flange stem. The edge with the crack is rotationally restrained and can only translate in the direction parallel to the tensile load. The compressive stresses appear in the region adjacent to the crack. I've seen it done in ABAQUS before but I can't figure out how to obtain the positive eigenvalues.

 

[rb1957]

ok, you're modelling a crack in a fillet rad that can only (?) grow along the beam (and not through the thickness ??).

not sure what you mean by "the crak is rotationally restained" ? or why you'd do that.

i assume you're modelling the crack as a duplication in the mesh ?

 

[pfsm01]

I mean to say that the edge of the plate with the crack is rotationally restrained.

ive been modelling the crack by assigning a seam to a line partition. i'm assuming the crack is already through-thickness

 

[rb1957]

confused again ... i thought we had a plate with an edge crack (growing across the width), in which case your constraint models the loaded ends of the plate moving parallel apart (under tension) ... the off-set moment at the cracked section is reacted by the constraint.

but you say (i think) you've got a crack in a fillet weld, growing along the legth of the plate ?

 

[pfsm01]

the crack is growing across the width, it just initiates at the fillet weld

 

[hydtools]

Post sketch to illustrate?

Ted

 

[pfsm01]

see attached. the tension load is being applied to the far ends. the image is displaying the transverse stresses (S22). notice the compression stresses in the vicinity of the crack. when I do a buckling analysis, the buckling does not seem to occur in that area.