ELMER crash

The graphical user interface of Elmer
mzenker
Posts: 1999
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: ELMER crash

Post by mzenker »

Hi,

@Peter:
I don't understand your point either. In the BodyForce he says

Code: Select all

  Heat Source = 1
  Integral Heat Source = Variable Time;Real;0.0 0.0 ;5.0e-6 0.0 ;5.01e-6 60.0;15.0e-6 60.0;15.01e-6 22.0;1e-3 22.0;End;
I cannot see what is wrong with that. But maybe I overlook something.
The crash occurs after the heat solver, when resultoutputsolver attempts to save the timestep. So IMHO the problem must lie elsewhere.

@Antoine:
I would execute the ResultoutputSolver "after timestep", not "always".
If you run the exact same simulation, but without Resultoutput, do you still have a crash? You could examine your case130umCu20um.ep with ElmerPost to start with.

HTH,

Matthias
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: ELMER crash

Post by raback »

Hi

This is what is suspicious:

Code: Select all

HeatSolve: Heater Control Information
HeatSolve: Heater for body:                                1
HeatSolve: Heater type:                      Integral heater
HeatSolve: Heater Volume (m^3):                   0.0000E+00
HeatSolve: Heater Power (W):                             NaN
HeatSolve: Heater scaling:                               NaN
HeatSolve: Heater Power Density (W/kg):                  NaN
HeatSolve: Heater for body:                                2
HeatSolve: Heater type:                      Integral heater
HeatSolve: Heater Volume (m^3):                   1.5000E-12
HeatSolve: Heater Power (W):                      0.0000E+00
HeatSolve: Heater scaling:                        0.0000E+00
HeatSolve: Heater Power Density (W/kg):           0.0000E+00
The sif might look ok but if the keywords are applied to wrong body sections it might still fail. It is impossible to judge the reason for problems merely from the sif file.

-Peter
mzenker
Posts: 1999
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: ELMER crash

Post by mzenker »

Oh, I see. I was too quick browsing the log... :oops:

@Antoine: could you post your geometry / mesh so that it becomes clear to which bodies you apply your BodyForces?

Matthias
Tonio
Posts: 8
Joined: 12 Mar 2017, 20:35
Antispam: Yes

Re: ELMER crash

Post by Tonio »

Thank you so much to step into this issue.

I'm pretty sure I do Something wrong, but can't find what.

I did try to execute the ResultoutputSolver "after timestep", not "always". In such case, simulation starts and crashes after the first timestep.

If I use "Never" as keyword for ResultOutput so that this solver is never executed, the simulation runs properly, but of course I get no vtu output file to examine after simulation, I can only look to the result with ElmerPost or ElmerVTK, but without the postprocessing capability of ParaView.

I attach the geometry file as per your request. It is a .geo file, I use gmsh to generate the mesh for Elmer.

Thanks again for your help.

Antoine
Attachments
Chip0.25mm2_130um_Cu20.geo
(9.74 KiB) Downloaded 321 times
mzenker
Posts: 1999
Joined: 07 Dec 2009, 11:49
Location: Germany

Re: ELMER crash

Post by mzenker »

Hi,

you seem to apply the BodyForce to Body 5, which is a thin layer in the middle of two other volumes.
But it is not sure that the numbering in Elmer corresponds to the one in your geo file. So you should check that by re-exporting the Elmer mesh to gmsh: On the command line, cd into your mesh directory and say

ElmerGrid 2 4 mymesh

where mymesh is the name of your ELmer mesh files. Then examine the resulting mesh in gmsh to check the numbering of bodies and boundaries.

HTH,

Matthias
Post Reply