CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   Bitching about fluent (http://www.cfd-online.com/Forums/main/4273-bitching-about-fluent.html)

Joel January 10, 2002 12:34

Bitching about fluent
 
I know that this may start a bit of a riot but here we go anyway: From looking around this forum for a fair few months now i have discovered one thing, no one likes fluent (fluent employees aside). Could someone in one distinct answer tell me why no one likes it. The pre-processor seems reasonably well liked (see discussion a few lines down) so what is wrong with the solver and the post processor? I am using Star at the mo and am getting on with it well but would like to know where star falls down and where other codes are better (CFX, Phoenics etc etc) and if star is considered more highly than fluent. I know this is dangerous but hey, should be interesting! P.S. Please no fluent/star/cfx/phoenics employees blowing their own trumpets please, i would like impartial views from users not sales men.

cfd user January 10, 2002 22:13

Re: Bitching about fluent
 
I don't think it's really ever a question of which code is better or worse. Currently, each code offers something different or slightly better than the other. Examples include chemistry and turbulence models, mesh generators, solver discretization, speed, etc. It also is somewhat political for the better Fluent users will not change to Star and vice-versa.

There are better ways in determining what code works best for your application. For example, the calculation of the moving piston/valve problem for an engine is easy to implement in Star but the chemistry is perhaps better in Fluent. Turbomachinery folks like CFX. You need to learn the codes and make the decision your self based on your needs and constraints.

The same situation holds true for mesh generators. Some use Pro-am, some use the Fluent or CFX generators. Others decide that, in addition to a CFD package, they need to use ICEM.

My take on this rather political debate is that an engineer should go to his trusty tool box and pick which tool works best for a given problem. In an ideal world, cost would not be the issue. Thus, the engineer must decide which code, in general, works best for a given set of problems that he will encounter while at his job.

I will say that I am a life long Star-CD user but realize the benefits of Fluent from experience. I have no experience with CFX. I have also reviewed papers to deal with the accuracy issue and can not come to a conclusion. With three main players out there - do I really have to? Each will get better from the competition.


All times are GMT -4. The time now is 07:24.