Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.
Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Convergence of Solver with Hyperelasticity - Problems
Posted 28.11.2012, 16:12 GMT-5 Low-Frequency Electromagnetics, MEMS & Nanotechnology, MEMS & Piezoelectric Devices, Geometry, Mesh, Studies & Solvers, Structural Mechanics Version 4.3 2 Replies
Please login with a confirmed email address before reporting spam
I've modeled a cantilever beam composed of Magneto-Rheological Elastomers (MAE) that bends when subjected to a magnetic field. The MAE used in the model currently uses a linear elastic material when in reality it follows hyperelastic behavior. When I try to incorporate the Arruda-Boyce hyperelasticity model into the materials (with ball-park inputs), the model cannot converge. Is there anything we can do such as changing the boundary conditions or solver configurations to make the solver converge on a solution?
Thanks,
Rob
Attachments:
Please login with a confirmed email address before reporting spam
I would highly recommend trying to obtain convergence with only the solid mechanics physics (with hyperelasticity) active. Use a dummy load and troubleshoot this relatively simple problem for convergence issues. In this way you can isolate this physics and have fewer "moving parts" to do deal with.
If it works, then add complexity by adding the magnetic fields and then the moving mesh.
If it doesn't work, then you know your hyperelastic material model is having issues. In this case, you could try solving for a less nonlinear material (ie adjust the material params to make the stress-strain response have a higher slope). Then, ramp up to the real material by sweeping using a Continuation parameter.
In my experience, hyperelastic material models often have convergence difficulties. You are not alone!
Regards,
Josh
Please login with a confirmed email address before reporting spam
I have recently had luck by changing my model to Mooney-Rivlin and simplifying my problem to a single-element cantilever beam. I plan on adding complexity as I solve the problem further but it is definitely easiest to experiment with a simple case and build from that. Starting off neck-deep in a hyperelastic material will almost always give you problems from what I've read.
Regards,
Rob