[Archive] Exceptions

SPIS
#1

Message by julien:
Initially posted Fri Nov 26, 2004 3:27 pm by markelov, forwarded by juju.
___
Hello,
I’ve created a simplified geometry of SMART-1. It works but when I’ve change a number of tetrahedrals, I’ve met the following during “UI to Num”.

xyzBd step 4 >>>>>> Exception raised in task  >>>>>> BEGIN TASK EXCEPTION HANDLING Traceback (most recent call last): File "/home/markelov/Spis3.00.01rc1/SpisUI/../SpisUI/Bin/Tasks/TaskSpisNumInterface.py", line 78, in run\_task sharedNum<a class="resourceLink" href="/projects/spine/home/community/forumsPages/spis/7938d0bbc0a8001400176c51ed28f250/7938d13bc0a80014002b98602ddb18b5/'SNVolList'"><span class="error">! 'SNVolList' !, sharedNum<a class="resourceLink" href="/projects/spine/home/community/forumsPages/spis/7938d0bbc0a8001400176c51ed28f250/7938d13bc0a80014002b98602ddb18b5/'SNSurfList'"><span class="error">! 'SNSurfList' !, sharedNum<a class="resourceLink" href="/projects/spine/home/community/forumsPages/spis/7938d0bbc0a8001400176c51ed28f250/7938d13bc0a80014002b98602ddb18b5/'SNBndList'"><span class="error">! 'SNBndList' !, sharedNum<a class="resourceLink" href="/projects/spine/home/community/forumsPages/spis/7938d0bbc0a8001400176c51ed28f250/7938d13bc0a80014002b98602ddb18b5/'SNIdList'"><span class="error">! 'SNIdList' ! = Interface.BuildInterface(shared<a class="resourceLink" href="/projects/spine/home/community/forumsPages/spis/7938d0bbc0a8001400176c51ed28f250/7938d13bc0a80014002b98602ddb18b5/'MeshElmtList'"><span class="error">! 'MeshElmtList' !) File "/home/markelov/Spis3.00.01rc1/SpisUI/./Bin/SpisNumInterface.py", line 1115, in BuildInterface if len(s.SNVolList<a class="resourceLink" href="/projects/spine/home/community/forumsPages/spis/7938d0bbc0a8001400176c51ed28f250/7938d13bc0a80014002b98602ddb18b5/9"><span class="error">! 9 !<a class="resourceLink" href="/projects/spine/home/community/forumsPages/spis/7938d0bbc0a8001400176c51ed28f250/7938d13bc0a80014002b98602ddb18b5/i"><span class="error">! i !) != 2: IndexError: index out of range: 51527 >>>>>> END TASK EXCEPTION HANDLING task SpisNumInterface just terminatet terminate
When I've used a working geometry with original version of spis on Suse 8.1 and pressed "Run Solver", I've got 
Now, everything should be ready for simulation SPIS/NUM main is starting GL\_DATA\_PATH: /tmp/gmarkelospistmp1101401490/ >>>>>> Exception raised in task  >>>>>> BEGIN TASK EXCEPTION HANDLING Traceback (most recent call last): File "/usr/src/spis/Spis3.00.01rc1/SpisUI/../SpisUI/Bin/Tasks/TaskJyTop.py", line 71, in run\_task caller.Run() File "/usr/local/spis3/SpisUI/./Bin/SpisNumCaller.py", line 69, in Run sharedSolver<a class="resourceLink" href="/projects/spine/home/community/forumsPages/spis/7938d0bbc0a8001400176c51ed28f250/7938d13bc0a80014002b98602ddb18b5/'jytop'"><span class="error">! 'jytop' !.Run() File "/usr/local/spis3/SpisUI/./Bin/JyTop4.py", line 458, in Run self.simu.integrate() spis.Util.Exception.SpisRuntimeException: spis.Util.Exception.SpisRuntimeException: advance: particle not in the right tetrahedron >>>>>> END TASK EXCEPTION HANDLING task JyTop just terminate
Actually, "UI to Num" crashes everytime, if I don't wait for 10 secs at least after finishing "Fields". Regards, Gennady
#2

Message by julien:
Dear Gennady,
First, sorry to answer so late.
The reasons of such kind of error message is a bad definition of properties settings and fields deployement. The main source of trouble has been correct since your question. However, this may still append in the following cases:

  • The boundary conditions are not properly defined regarding the type of object. This is the case for exemple for the spacecraft surface (default spacecraft plasma), if you do not set any material properties or electrical node. The corresponding fields are then missing during the conversion UI to Num, that (hopefully) crash.
  • If you have a inner object without volume or surface, a thin wire only for example. This is a bug that we are currently fixing.
  • If the same surface is in the same time the inner and the external boundary. This is typically the case if your spacecraft touchs the external surface of the computationnal volume.
  • In some cases when you start from an existing project. In this case, the groups definition seems to be not always well saved into the common data bus. To solve this problem, you have just to call-back the group manager one more time, save and quit it, perform again the group conversion and the field mapping. The conversion UI to Num should be ok then.
Of course, generally speaking, if the conversion UI to Num is not correctly done, the execution of the simulation cannot be done and will crash. Best,ju.
#3