Menu

Discussion Forum

New Discussion
Filter

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.

Viscoplastic creep in solder joints-Failed to find consistent initial values.

Posted Dec 7, 2011, 5:11 p.m. EST 5 Replies

Radu Florea
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Dear all,
I got the following error messages:
Failed to find consistent initial values.
Last time step is not converged.
Inverted mesh element near coordinates (7.22712e+011, 2.21891e+010, -6.82633e+011).

What I did.I replaced Anand model (tutorial Viscoplastic Creep in Solder Joints with my own).For few days I struggle with the same message.
Please help debugging this if possible.As an attachement you have the file with the errors.

Thank you very much,
Radu
5 Replies Last Post Feb 19, 2012, 7:34 p.m. EST
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 11, 2011, 8:44 a.m. EST
Hi

I'm not sure its only the material chose, when mixing physics you need to ensure you have enough BC to fix all values in ALL physics. I.e for your solid domain, you use a 2D boundary roller (=one direction and two rotations fixed), then you fix 1 direction on one point (=1 Dof fixed) you still have a free rotation that will give difficulties for the solver to resolve, try fixing another point on the opposite side along "w" too ;) But this is not enough

Another issue to be checked (might be OK, but COMSOL does not help us to check, is the thermal diffusivity linking mesh size and time steps (see the FORUM and Nagy's comments) respect the h<=sqrt(alpha*delta_t) where alpha is the diffusivity k/rho/Cp for each materials, h the mesh size (perpendicular to the heat flux direction) and delta_t is the solver stepping (the one the solver uses not yours in free mode)

Then I'm not familiar by using "eps" as disturbance (I would rather use sqrt(eps) as many formulas rely on the square of a number and eps^2=0 in most cases

I suspect that your two additional PDE do not have enough BC defined. Not sure, try to simplify your model, then add the PDEs one by one

--
Good luck
Ivar
Hi I'm not sure its only the material chose, when mixing physics you need to ensure you have enough BC to fix all values in ALL physics. I.e for your solid domain, you use a 2D boundary roller (=one direction and two rotations fixed), then you fix 1 direction on one point (=1 Dof fixed) you still have a free rotation that will give difficulties for the solver to resolve, try fixing another point on the opposite side along "w" too ;) But this is not enough Another issue to be checked (might be OK, but COMSOL does not help us to check, is the thermal diffusivity linking mesh size and time steps (see the FORUM and Nagy's comments) respect the h
Radu Florea
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Dec 19, 2011, 10:19 a.m. EST
Thank you,
I will work one by one PDE's.
Radu
Thank you, I will work one by one PDE's. Radu
Guy Clatterbaugh
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Feb 1, 2012, 11:06 a.m. EST
Radu,

Did you ever fix your problem. I have similar ones. If I have any degree of stress concentration my solutions donot converge. guy
Radu, Did you ever fix your problem. I have similar ones. If I have any degree of stress concentration my solutions donot converge. guy
Radu Florea
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Feb 19, 2012, 7:33 p.m. EST
Ivar,
I get the same message.I will try one more day and I will sent the model to you.
Thank you,
Radu
Ivar, I get the same message.I will try one more day and I will sent the model to you. Thank you, Radu
Radu Florea
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Feb 19, 2012, 7:34 p.m. EST
Ivar,
I get the same message.I will try one more day and I will sent the model to you.
Thank you,
Radu
Ivar, I get the same message.I will try one more day and I will sent the model to you. Thank you, Radu

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.

Suggested Content