core dump - engrid or machine limitation
Dear Bruno, dear Oliver,
for a mesh resolution study I would like to generate a mesh as fine as possible. At the moment I struggle to generate a boundary layer, where the surface mesh contains more than 2M surface cells (engrid 1.4). (Error: 3698 Speicherzugriffsfehler (Speicherabzug geschrieben)) In the old forum I read that only volume meshes of ~8M cells were generated in engrid so far (in 2010), so I ask myself if the limitation in engrid is reached or if my machine limits me in generating finer Meshes (32 GB RAM, no swap) As far as I see, the memory is not consumed at time of abortion... What Meshsize is known to be generated successfully? best regards Georg |
Hi Georg,
Are you positive that the engrid binary was built as a 64bit application? You can confirm this by running: Code:
file engrid Code:
file $(which engrid) Code:
engrid: ELF 64-bit LSB executable, x86-64 Bruno |
Dear Bruno,
yes, I can confirm that engrid was built as 64bit application. Got the output Code:
engrid: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.24, BuildID[sha1]=0x5a355b4b14a461551d3c213e1cea9fd2c66bf454, not stripped |
Hi Georg,
I'm on vacation these coming days, so I could use more specific information on the kind of test case I should be trying to do (number of surface mesh faces, volume mesh characteristics and so on). If you could share such a case would be great. Because if my guess is correct, there is a limitation in the number of faces count to 2 million or 4 million, but I need a test case to isolate this. Best regards, Bruno |
All times are GMT -4. The time now is 05:31. |