Nagi Elabbasi
Facebook Reality Labs
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
22 mar 2011, 17:00 GMT-4
Hi Kevin,
You do not need to start with the two objects not in contact. If the initial overlap is too much convergence may be a problem, but you have them initially forming a line contact which is fine.
There are many reasons why contact problems do not converge. I would start by examining the error message that COMSOL reports. If that doesn’t provide any clues then try disabling contact and checking that the model otherwise works. Then you can try reducing the nonlinearity of the model (temporarily) by applying a smaller load, removing or reducing friction if you have any. I notice also that the slab is very thin, so ideally it should be defined as the destination not the source of the contact pair (since it is less stiff).
Nagi Elabbasi
Veryst Engineering
Hi Kevin,
You do not need to start with the two objects not in contact. If the initial overlap is too much convergence may be a problem, but you have them initially forming a line contact which is fine.
There are many reasons why contact problems do not converge. I would start by examining the error message that COMSOL reports. If that doesn’t provide any clues then try disabling contact and checking that the model otherwise works. Then you can try reducing the nonlinearity of the model (temporarily) by applying a smaller load, removing or reducing friction if you have any. I notice also that the slab is very thin, so ideally it should be defined as the destination not the source of the contact pair (since it is less stiff).
Nagi Elabbasi
Veryst Engineering
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
23 mar 2011, 10:33 GMT-4
Nagi,
Thank you for the response. I have simplified my model by removing the piezoelectric component and now just have an Aluminum slab with a cylinder above it. The model is solving, but the cylinder is passing through the slab. One thing of note, I had to add solid.hmin_dst as a parameter in the global definitions. In the contact node under solid mechanics, the expression for penalty factor was not evaluating. I have search the documentation but can not find anywhere where it defines the solid.hmin_dst variable.
I have uploaded my file and would appreciate if you could take a quick look. Only takes a few moments to solve.
Thank you,
Kevin
Nagi,
Thank you for the response. I have simplified my model by removing the piezoelectric component and now just have an Aluminum slab with a cylinder above it. The model is solving, but the cylinder is passing through the slab. One thing of note, I had to add solid.hmin_dst as a parameter in the global definitions. In the contact node under solid mechanics, the expression for penalty factor was not evaluating. I have search the documentation but can not find anywhere where it defines the solid.hmin_dst variable.
I have uploaded my file and would appreciate if you could take a quick look. Only takes a few moments to solve.
Thank you,
Kevin
Nagi Elabbasi
Facebook Reality Labs
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
29 mar 2011, 23:37 GMT-4
Kevin,
The cylinder is passing through the slab because the applied cylinder displacement is too big for one step. If you apply the displacement gradually (over 3 steps) using a Parametric Sweep with values 0.1, 0.5 and 1.0 times the full displacement it works. Also there is no need to define the solid.hmin_dst variable since it is an pre-defined COMSOL variable related to element size.
Nagi Elabbasi
Veryst Engineering
Kevin,
The cylinder is passing through the slab because the applied cylinder displacement is too big for one step. If you apply the displacement gradually (over 3 steps) using a Parametric Sweep with values 0.1, 0.5 and 1.0 times the full displacement it works. Also there is no need to define the solid.hmin_dst variable since it is an pre-defined COMSOL variable related to element size.
Nagi Elabbasi
Veryst Engineering