[CIG-LONG] Error in _PCDVC_Calculate3D

Mark Brandon mark.brandon at yale.edu
Thu Sep 30 11:06:34 PDT 2010


  I wrote yesterday about a problem I am having but I realize now that I 
may have described it very well.
I am running a relatively simple 3D model on a linux machine with Gale 
1.5.0. The model has all faces as static,
and imposed velocities on three faces. A fluid of constant and uniform 
viscosity flows in through one face and
out another face, in a fashion similar to subduction zone with a 
box-like geometry.
I have set the inflow parameter to "true", as recommended in the Gale 
manual.

The computational grid is 90x100x240, with particlesPerCell set to 30. 
The model is run for 10 time steps.
I am able to run the model to completion without problems when I reduce 
the grid to 9x10x24. When I run
the model using the full grid, I get the following error during the last 
time step:

Error in _PCDVC_Calculate3D: Problem has an under resolved cell
(Cell Id= 125459), add more particles to your model
*** glibc detected *** ./Gale-1_5_0: double free or corruption (out): 
0x00000000a035d320 ***

I have searched the Gale archive and have found no discussion of this 
error. I have also read the source code
in PCDVC_Calculate3D, but that did not help either. The recommendation 
is that I need to "add more particles
to your model". I assume that this means that I need to increase the 
parameter value for "particlesPerCell"
from 30 to a larger number. I have no idea as to how much I should 
increase this value.

Has anyone else had this error? If so, how did you fix this problem? Is 
the "particlesPerCell" parameter relevant
to this problem" Is it possible to predict how values for this parameter 
should scale with other model parameters?

I have attached the xml and runlog files in case that information may 
prove useful.
Best,
Mark Brandon
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: runLog_inputPol_Okaya.txt
Url: http://geodynamics.org/pipermail/cig-long/attachments/20100930/c5a717de/attachment-0002.txt 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: inputPol_Okaya.xml
Type: text/xml
Size: 22368 bytes
Desc: not available
Url : http://geodynamics.org/pipermail/cig-long/attachments/20100930/c5a717de/attachment-0001.bin 
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: ATT00001..txt
Url: http://geodynamics.org/pipermail/cig-long/attachments/20100930/c5a717de/attachment-0003.txt 


More information about the CIG-LONG mailing list