JOJOMO
Mechanical
- Nov 14, 2006
- 29
Hi Folks,
I'm running a contact analysis where I slide elasto plastic spheres across each other and have had pretty good results so far. I'm running a parametric study and changing the interference for each run. I'm on one of my last cases now and cannot get the darn thing to converge.
I'm using/have used standard right now. I was thinking I may need to try this in explicit. I don't really know what else to try in standard. The problem is that the upper most elements on one of the spheres becomes highly distorted when they slide. It almost looks like a shear strain unit cell you'd draw in the intro mechanics course. Then the stress relaxes at the surface while still in contact. Physically, this doesn't make sense to me. I haven't seen this in any of the smaller interference problems. I don't think this is a physical phenomenon; I think it is a numerical issue due to the distortion of the surface elements.
I haven't run an explicit analysis before, so my question is : what do I need to do to change my standard job to an explicit job? If I want to run a quasistatic analysis, do I need a density?
Right now the standard job is considered as quasistatic. Basically, I start with the spheres just touching and then slide one across another until they come out of contact. I've tried breaking it up into very tiny steps. I've got 140 steps to complete sliding (at 60, the spheres are aligned vertically).
I'm running a contact analysis where I slide elasto plastic spheres across each other and have had pretty good results so far. I'm running a parametric study and changing the interference for each run. I'm on one of my last cases now and cannot get the darn thing to converge.
I'm using/have used standard right now. I was thinking I may need to try this in explicit. I don't really know what else to try in standard. The problem is that the upper most elements on one of the spheres becomes highly distorted when they slide. It almost looks like a shear strain unit cell you'd draw in the intro mechanics course. Then the stress relaxes at the surface while still in contact. Physically, this doesn't make sense to me. I haven't seen this in any of the smaller interference problems. I don't think this is a physical phenomenon; I think it is a numerical issue due to the distortion of the surface elements.
I haven't run an explicit analysis before, so my question is : what do I need to do to change my standard job to an explicit job? If I want to run a quasistatic analysis, do I need a density?
Right now the standard job is considered as quasistatic. Basically, I start with the spheres just touching and then slide one across another until they come out of contact. I've tried breaking it up into very tiny steps. I've got 140 steps to complete sliding (at 60, the spheres are aligned vertically).