Analysis Programs (RISA vs Fastrac)

[eit09]

We currently use RISA Programs for all of our FEA needs. We recently had a CSC rep come and give a presentaion on Fastrac which we saw some features that seemed like they could save us some time (especially on the bigger structures). The first was the ease in applying geographical loadings and second the speed in which the program runs. Does any one have any experience with the Fastrac comparing it with RISA? Any pros or cons on Fastrac? I don't beleive We would ever get rid of RISA, but on larger structures we may need a different program that can run a large model with all its Load combinations quickly.

 

[JoshPlumSE]

EIT -

For what it's worth, you probably will not get much of a response on this forum. This is more geared towards pure FEM guys and less towards civil / structural types of FEM.

I'm obviously a biased observer (I work for RISA). But, I have a hard time seeing Fast Track as credible competition for RISA. They're just a relatively small bit-player in the US market. Maybe that will change in the near future. But, currently they are still a bit player.

RISA does tend to be a memory hog, though our solution times are quite fast for the FEM solution. I find it hard to believe that they can compete with the pure speed of our sparse solver.

That being said, some companies will sacrifice the ability to do non-linear analysis (P-Delta, tension only members, compression only springs et cetera) by doing a category solution (DL, LL, EL, WL, et cetera) and then using super position of results to quickly assemble member forces and stresses for a large number of load combinations. That certainly is a faster way to go if you can prove that it is appropriate to ignore non-linear behavior.

RISA doesn't take this short cut, so our solution could be slower in these cases (large numbers of LC's with little or no 2nd order effects). The theory is pretty straight forward. So, it wouldn't be all that difficult to change if we got significant user demand for the change.

Certainly, we've already had some user requests for modifying the way we're doing our solutions (especially related to P-Delta). From what I remember, however, those requests have NOT been based on complaints about solution speed. Rather they have been related to concerns about P-Little Delta effects. These effects usually (always?) get ignored in the category based super-position solution method. So, I'm not sure that this super position method has gotten much attention from us yet.

Just my $0.02

 

[eit09]

JoshPlum,

Thanks for the response worth more than $0.02 .