With the simple free meshing of an area as shown by the input below,
ANSYS 8.1 aborts with a division by zero error:
fini
/clear
/plopts,info,1
/prep7
k,1, 0.00, -0.232, 0
k,2,-0.632E-01, -0.232, 0
k,3,-0.892E-01, -0.387, 0
k,4,-0.121 , -0.332, 0
k,5,-0.576E-01, -0.441, 0
k,6, 0.892E-01, -0.387, 0
k,7, 0.576E-01, -0.441, 0
k,8, 0.121 , -0.332, 0
k,9, 0.632E-01, -0.232, 0
l,2,1
l,1,9
l,9,8
l,8,6
l,6,7
l,7,5
l,5,3
l,3,4
l,4,2
al,1,2,3,4,5,6,7,8,9
et,1,shell63
lesize,all,,,2
amesh,all
*** FATAL ***
Divide by Zero in FEM_utNorm.
ALL CURRENT ANSYS DATA WRITTEN TO FILE NAME=
ansabort.db
FOR POSSIBLE RESUME FROM THIS POINT
Is this a known bug? The same input works fine in ANSYS 8.0.
Yes, Defects 22623, 22725, and 22750 were filed (independently with similar meshing tasks) and subsequently fixed in ANSYS 9.0 Preview 1. The fix has also been included in ANSYS 8.1 SP1. This service pack is expected to be released in the 4th week of June, 2004. As a workaround, you can issue /NERR,-1,-1,-1 to force the meshing by ignoring the error. |
||
|