XANSYS Message: 3813 [Go back to message list]
[bookmark on del.icio.us]
No rating yet
Rate item:

Subject: Re: SuperElement Expansion Pass Problem.
Author: John Swanson
Date: 1999-08-06 14:22:00

Make sure also that you have permission to use the file.
Sometimes ANSYS requires write permission to files when
it probably should only be reading them.
Actually, the procedure you are using is currently in favor
by large users of substructures, because you can avoid
writing the .tri file in the generation run, and also avoid having
to keep it until the stress pass.
You might want to try using the iterative (PCG) solver
for the stress pass, since the problem is well constrained
by the solutions at the master DOFs.
If we were to develop substructuring in today's computing
environment, we would not use the .tri file. ANSYS should
probably review the current implementation of this procedure
to see if it can be simplified, so it does not "involve a lot of
work".
John Swanson
Swanson Analysis Services, Inc

> -----Original Message-----
> From: Greg Zhang [SMTP:gzhang@b...]
> Sent: Friday, August 06, 1999 1:46 PM
> To: xansys@o...
> Subject: [xansys] SuperElement Expansion Pass Problem.

> From: "Greg Zhang"

> Dr. John Swanson:

> Thank you for your response. I do have the right .tri file in the
> directory. The first time we got this problem is about 4 years ago by
> another engineer. We are wondering how we can prevent it.

> Right now, our solution to it is:
> get the displacement of the superelement nodes in use pass;
> input the displacement into the nodes with master degree of
> freedom;
> run a static analysis to recover the stress.
> But it involves a lot of work. Is there any problem for this solution?
> THANKS

> BEST REGARDS

> GREG ZHANG
> DESIGN ENGINEER
> BUCYRUS INTERNATIONAL INC.

> From: John Swanson

> Is it possible that the .tri file from the SuperElement generation
> run is not available (wrong name, etc)
> for the stress pass? This is the file which has the information to
> expand from the
> reduced displacement set to the full displacement set.
> John Swanson
> Swanson Analysis Services, Inc

Posts possibly associated with message #3813AuthorDateScore
3809SuperElement Expansion Pass Problem.Greg Zhang1999/08/06 
3812SuperElement Expansion Pass Problem.Greg Zhang1999/08/06 
3813Re: SuperElement Expansion Pass Problem.John Swanson1999/08/06