[CIG-LONG] Gale in 3D

[Walter Landry]

Hi Rainer,

3D models work for me, but running the 2D thermal cookbook in 3D does
not.  You have to fix the boundary conditions for that to work.  Can
you run the 3D cookbook input/cookbook/extension3D.json?  If that
works, you probably just need to fix your input file a bit.

Cheers,
Walter Landry

Rainer Nerlich <rainer at simula.no> wrote:
> Hi Walter,
> 
> I had several things to do lately so I couldn't test the new repository version
> of Gale earlier. However, it's installed now (with PETSc 3.2), but unfortunately
> I cannot run 3D models anymore (and thus couldn't test my own input files). The
> cookbooks work fine in 2D but as soon as I switch them to 3D, I get a
> segmentation fault after the first time step.
> 
> In 2D:
> 
> (gdb) run gale/cookbook/thermal.json 
> The program being debugged has been started already.
> Start it from the beginning? (y or n) y
> Starting program: /home-3/rainer/Gale gale/cookbook/thermal.json
> [Thread debugging using libthread_db enabled]
> TimeStep = 0, Time = 0
> TimeStep = 1, Time = 4.78516e-05
> TimeStep = 2, Time = 9.5703e-05
> TimeStep = 3, Time = 0.000143554
> TimeStep = 4, Time = 0.000191406
> TimeStep = 5, Time = 0.000239257
> TimeStep = 6, Time = 0.000287108
> TimeStep = 7, Time = 0.000334959
> TimeStep = 8, Time = 0.00038281
> TimeStep = 9, Time = 0.000430661
> TimeStep = 10, Time = 0.000478512
> 
> Program exited normally.
> 
> In 3D:
> (gdb) run gale/cookbook/thermal.json 
> Starting program: /home-3/rainer/Gale gale/cookbook/thermal.json
> [Thread debugging using libthread_db enabled]
> TimeStep = 0, Time = 0
> 
> Program received signal SIGSEGV, Segmentation fault.
> 0x0000000000cbc818 in Triquadratic_EvalLocalDerivs (elementType=0x1a23280,
> localCoord=0x0, derivs=0x20aab10) at StgFEM/Discretisation/src/
> Triquadratic.cxx:265
> 265 double xi = localCoord[0], eta = localCoord[1], zeta = localCoord[2];
> (gdb) 
> 
> I get the same error message when I run my own models which worked with the old
> version (e.g. the files I had sent to you a while ago). As my models have to be
> in 3D, I hope there is a way to get around this. Do 3D models run on your system
> with the latest repository version (I got it yesterday)? If they work, it must
> be my installation again.
> 
> Thanks for your help and best regards,
> Rainer
>