FEA, NISA vs ANSYS

[K.Hamilton]

I've been looking at several PC based FE packages. Both NISA and ANSYS
look to be good. Do you have any experience/comments on either or both
of these packages?

[SONOTRODE]

SDRC's Ideas is a good PC based analysis package (Works on NT 4.0
operating system, and you'll want a LOT of ram , disk storage, and fast
processors).

[John Schindler]

>SDRC's Ideas is a good PC based analysis package (Works on NT 4.0
>operating system, and you'll want a LOT of ram , disk storage, and fast
>processors).

Try Cosmos/M. It's very fast, easy to use, and powerful.

John Schindler, P.Eng. | SCHINDLER TECHNOLOGIES CORPORATION
President | Finite Element Analysis Specialists
jsch...@ica.net | Tel. 905-415-8106; Fax 905-415-9102
| http://schindler.ica.net

[Jim Hurdle]

K.Hamilton <kenneth....@gecm.com> wrote in article
<5k1vbm$m...@gcsin3.geccs.gecm.com>...

>
> I've been looking at several PC based FE packages. Both NISA and ANSYS
> look to be good. Do you have any experience/comments on either or both
> of these packages?
>
>

We used to have NISA. We now have ANSYS. It is extremely difficult to
compare FEA packages by reading the sales information. You really to use
it in "real life" situations to make a good evaluation. This, of course,
is difficult to do with multiple packages in an evaluation mode. Although
it depends somewhat on what type of analysis you want to do, in general,
you get what you pay for and you should go with ANSYS.

When we had NISA it was extremely buggy, poorly documented, contained
commands that did not work and had poor technical support. Their
integration of pre, post and solver was very poorly implemented. For
instance, if you used layered composite elements they were supported in the
solver but not the preprocessor. So when you built the model in the
preprocessor you had to call it a different element. Then you had to use a
system text editor to modify the input to include the data needed for the
composite elements. I haven't used NISA in a few years so maybe some of
these problems have been fixed but overall it was very unimpressive.

I have used ANSYS for nearly 10 years and have been very happy with it.
Let me know if you have any specific questions.

[bere...@caent11.esa.lanl.gov]

In article <01bc56b4$02635ea0$170e5bcf@mothership> "Jim Hurdle" <jh3...@navix.net> writes:
>
> When we had NISA it was extremely buggy, poorly documented, contained
> commands that did not work and had poor technical support. Their
> integration of pre, post and solver was very poorly implemented. For
> instance, if you used layered composite elements they were supported in the
> solver but not the preprocessor. So when you built the model in the
> preprocessor you had to call it a different element. Then you had to use a
> system text editor to modify the input to include the data needed for the
> composite elements. I haven't used NISA in a few years so maybe some of
> these problems have been fixed but overall it was very unimpressive.

As a current NISA user, I can tell you that these problems no longer
exist. Documentation is excellent, all of the commands work, and
integration is as good as any other package. The composite elements
are fully integrated and supported, material properties are easy to
input (tabular format), and you can get plots of ply orientation, etc.

Disclaimers: I am not affiliated with EMRC in any way except as a
satisfied user. This message should not be taken as an endorsement of
NISA.
--
Barry Berenberg
Los Alamos National Laboratory
ESA-DE
bere...@lanl.gov