🚀 go-pugleaf

I am posting this for someone who was not able to post on the net. I found
it very informative.

	Pushkar
-----------------------------original message-----------------------------

There are, as noted in the original posting, several commercially available 
Finite Element packages on the market.  In addition to ABAQUS, MSC/NASTRAN, 
and ANSYS, all of which are good packages, but with different capabilities, 
and markets, there are a considerable number of less well known packages.  
Some of these are specifically for the IBM-PC platform (such as ALGOR), and 
some have better dynamics capabilities, or buckling, or thermal capabilities 
than others.  There is even a P-version code out commercially, called RASNA, 
which is very good.

For the practitioner who needs to make a rational decision about what package 
to purchase, the first oder of business is deciding what type of problems 
will be solved with the package.  It is not really as easy as just deciding 
which is best, each of the packages has its good points and bad points.  
ABAQUS is an extremely good non-linear code, but is overkill for the more 
common linear elastic problems.  MSC/NASTRAN is very good at the airplane 
kinds of analysis, as one might guess (linear statics and dynamics).  ANSYS 
seems to be good for the power plant and automotive industries.  If a small 
commercial shop has AUTOCAD, then ALGOR might be the best answer.  

If I might make a suggestion, the Computers in Engineering Division of ASME 
puts on a conference each year, which deals in part with this issue.  There 
is also an exhibit with the conference with participation from most if not 
all of the larger commercial FE vendors.  I happen to be the Chairman of the 
conference this year, and would be delighted to see this whole issue brought 
up and fully discussed at the conference.  It will be in San Diego in August 
of 1993.  I can email anyone a copy of the call for papers that wants one, or 
I would be delighted to respond to any other requests for information.  
Alternatively, if anyone is interested, I have been in this business for a 
long time, and can help the uninitiated.  Those that have been around for 
some time I hope will understand that I am somewhat reticent to engage in 
what always tends to be a religious war.  That is, the choice of the "BEST" 
finite element package.  There is NO BEST ONE.  It all depends on the problem 
that is to be solved, and the analyst's abilities and proclivity.

Edward M. (Ned) Patton          npatton@swri.edu
Southwest Research Institute
San Antonio, TX  78228