Lescombes
New member
- May 6, 2002
- 25
Hello everyone,
I've been having some fun recently developing some test models of bolted lap joints that have included a solid bolt and contact between relevant surfaces; for the purposes of learning i'm leaving friction out of the equation for the time being. So far i've found a way to get the solution to converge is to apply a small displacement in the first analysis step to close any gaps, and have a second analysis step where I apply the load of interest (which is one of the solutions proposed in my software package's training material to aid contact convergence issues).
This method, however, appears less effective as the number of fasteners (and hence contact surface pairs) increases. For instance, once I've got 4 fasteners my particular solution will only converge when loading is applied by displacement.
My question is - if the only way the model will converge is with displacement loading, is it just a matter of iterating until I get the displacement that corresponds with the internal load I'm wanting to analyse? Is this valid? Or is it more likely that my contact problem is just ill defined and I'm band-aiding over this problem by applying the displacement loading?
Thank you,
Greg
I've been having some fun recently developing some test models of bolted lap joints that have included a solid bolt and contact between relevant surfaces; for the purposes of learning i'm leaving friction out of the equation for the time being. So far i've found a way to get the solution to converge is to apply a small displacement in the first analysis step to close any gaps, and have a second analysis step where I apply the load of interest (which is one of the solutions proposed in my software package's training material to aid contact convergence issues).
This method, however, appears less effective as the number of fasteners (and hence contact surface pairs) increases. For instance, once I've got 4 fasteners my particular solution will only converge when loading is applied by displacement.
My question is - if the only way the model will converge is with displacement loading, is it just a matter of iterating until I get the displacement that corresponds with the internal load I'm wanting to analyse? Is this valid? Or is it more likely that my contact problem is just ill defined and I'm band-aiding over this problem by applying the displacement loading?
Thank you,
Greg