Hello,
Following the earlier discussions on implementing an algebraic equation into a structural mechanics model, I fortunately got away with writing out explicitly an analytical solution. Thanks a lot to Ivar on the helpful instructions so I could get around the ()^1/3 issue.
I am running a 2-D axisymmetric simulation of a cylinder expanding with time and eventually come into contact with a tube wall (parallel contact surfaces). When I assigned a fixed expansion rate, I could get this contact problem to work without too much difficulties (had to tune the penalty factor several times). However, once I included the denpendence of the expansion rate on the stress, the contact part of the problem simply could not converge. My guess is that on contact, the stress changes quite dramatically with time and this in turn changes the expansion rate. As the expansion rate also has a radial distribution, the problem becomes highly non-linear.
I have tried several things with the solver (I am using a Direct solver with automatic damping) including changing the time stepping tolerance factor. I also tried to vary the contact penalty factor and initial contact pressure value. None of these seemed to work. I wonder if there are certain strategies I could try to get pass this convergence issue? I am using a fast contact search.
Thank you very much!
Best regards,
Di
Following the earlier discussions on implementing an algebraic equation into a structural mechanics model, I fortunately got away with writing out explicitly an analytical solution. Thanks a lot to Ivar on the helpful instructions so I could get around the ()^1/3 issue.
I am running a 2-D axisymmetric simulation of a cylinder expanding with time and eventually come into contact with a tube wall (parallel contact surfaces). When I assigned a fixed expansion rate, I could get this contact problem to work without too much difficulties (had to tune the penalty factor several times). However, once I included the denpendence of the expansion rate on the stress, the contact part of the problem simply could not converge. My guess is that on contact, the stress changes quite dramatically with time and this in turn changes the expansion rate. As the expansion rate also has a radial distribution, the problem becomes highly non-linear.
I have tried several things with the solver (I am using a Direct solver with automatic damping) including changing the time stepping tolerance factor. I also tried to vary the contact penalty factor and initial contact pressure value. None of these seemed to work. I wonder if there are certain strategies I could try to get pass this convergence issue? I am using a fast contact search.
Thank you very much!
Best regards,
Di