cancel
Showing results for 
Search instead for 
Did you mean: 

Nastran sparse FBS

Experimenter
Experimenter

Hello,

 

I have a problem with nastran sparse fbs.

I have two quite big SOL 101 Inertia Relief jobs with just few small geometric differences between them.

The calculation of the first of them takes approximately 7 hours, the second one needs however about 20 hours.

 

I looked into the .f04's und found out, that the fbs module is mainly responsible for this huge difference in computing time:

 

FBS METHOD S2 TIME ESTIMATE TO FORM UL1 - CPU= 6121, I/O= 493, TOTAL= 6614, PASSES= 1
10:39:22    44:02    873.1G    5.0    8414.6    0.0    FBS BEGN
11:02:33    67:13    1088.3G    215.2G    12095.5    3680.9    FBS END

 

for the first model and:

 

 

FBS METHOD S1 TIME ESTIMATE TO FORM UL1 - CPU= 24880, I/O= 497, TOTAL= 25377, PASSES= 1

17:54:32    45:12    873.4G    5.0    8780.7    0.0    FBS BEGN
7:43:37    874:17    1089.9G    216.5G    58390.8    49610.0    FBS END

 

for the second.

 

Could anyone tell me what FBS METHOD S1 and FBS METHOD S2 really are and if I can force using one them instead of letting nastran decide?

 

I use nastran 9.1 ILP-64 with 12 cores and following settings:

 

NASTRAN BUFFSIZE = 65537

NASTRAN SYSTEM(146) = 4
NASTRAN SYSTEM(206) = 72

 

 

Kind regards,

Wojciech