【讨论】如何控制nastran的计算输出

hptji

对于大模型，nastran 计算输出的中间文件将大的离谱，这对硬盘小的
兄弟来说，可苦了
  
  请问高手，如何修改 bdf 文件控制输出.............

fota

Quote from Nastran Q&A, you can add following cards in your bdf file:
  
INIT SCRATCH LOGICAL=(SCR1( 150MB ),SCR2( 100MB )) ,
SCR300=(SCRA( 250MB ),SCRB( 300MB ))
ASSIGN SCR1=’file_name.scr1’
ASSIGN SCR2=’file_name.scr2’
ASSIGN SCRA=’file_name.scra’
ASSIGN SCRB=’/mydisk1/file_name.scrb’
  
Above example creates the SCRATCH DBset with logical names of SCR1, SCR2, SCRA, and SCRB (Of course, you can have more such as SCR3, etc.). Two physical files, file_name.scr1 and file_name.scr2, are created with 150 and 100 megabytes, respectively. These two files are regular scratch files and reside in the current directory.
Two additional physical files, file_name.scra and file_name.scrb, are created with 250 and 300 megabytes, respectively. These two files are SCR300-type files. The SCR300 file file_name.scra resides in the current directory. The SCR300 file file_name.scrb is allocated to the file system (directory) /mydisk1.
  
But, if your hard disk is too small to store all DBset. You will get failure message.

hptji

不胜感激！！
  
  但如果用命令行 nastran jobname scratch=yes   
  
  是否同样有效

fota

no.
  
scr=yes means delete the *.master, *.dball after job finished.

FEM-Fan

不管怎样，NASTRAN计算时中间文件就是这么大，现在PC机的硬盘不是很贵

cuiyihua007

用不着这么麻烦的只要把你关心的节点单元建到一个组中
然后把在analysis里的output设为这个组就行了

hbin_tu

hehe

daniu2006

谢谢了，长见识！