CATIA V5 2

[robcp]

Hello,
Do you recommand Generative Structural Analysis that CATIA V5 has for FEA tests?
Thank you

 

[GregLocock]

What sort of tests? Can you easily pull in the real world data you must have for a correlated model?



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[johnhors]

The Catia V5 GPS (Generative Part Stress) module is much like any other CAD embedded linear analysis FE system. It is quick and easy to use (IMHO it is damn far too easy!). I have checked it by replicating the same BC's using the same geometry in more traditional software and got more or less the same result, with any differences being explained by different mesh patterns and refinement. You couldn't expect anything else!

The problem as I see it, others will surely disagree, is that anyone who can drive Catia V5 can run analyses using the GPS module, regardless of how little or how much they know about stress analysis, and the correct procedures for setting up a meaningful analysis. With probably 99% certainty a typical GPS analysis will have loads applied at one end of a model with the other end fully clamped. Thus every component has become a cantilever! Reverse the loaded and clamped ends to see a different set of results!

Try explaining free body diagrams of applied loads, fully load balanced analyses with minimal supports and again with 99% certainty you'll get a blank stare back from the typical CAD jockey.

Sadly though very few people seem to be aware that once you include the additional advanced modules of Catia V5 it becomes a truly world class FEA system that allows the user to apply complex variable pressure loadings, fully balance their loads (as per the free body diagram) and apply the minimal 3-2-1 support method. The GPS module on it's own is too simplistic, probably for commercial reasons where ease of use is the paramount objective to the detriment of reality as it will be wrongly compared against rival products on this point alone by many prospective users.

 

[rb1957]

i would consider GPS to be very crude analysis. i believe it can handle only one load case at a time. i believe the iteration approach is laborious, but probably convergent. i forget how much control you have on the elements, hopefully you can avoid those horrid tet4s. i think it's very limited in its constraint assumptions. i doubt that most of the deigners using it really know what they're doing (FEA is a loaded gun), 'cause if they were that specialised they'd probably be working as stress engineers.

i touched GPS some years back, and maybe they've improved some of these features. you can always address some of the concerns by doing patch tests on your own, and maybe using a demo version to see how it fits your needs.

good luck

 

[enthu1980]

I think rb1957 has probably not seen the catia version v5r16. It has the capacity to use thousands of loads having multi-load case,combined case,transfer of displacements as well as loads from sub-analysis to main-analysis.For a rough analysis at designer level, it is emerging as a very useful CAD integrated fea system.I agree with johnhors that using EST licensce, it becomes a very powerful tool.

 

[johnhors]

When I saw what the full suite of FEA modules in Catia V5 could do, it blew my socks off! As for multiple load cases, I've seen guys use macros to extract load and moment data at many points from in-house beam models for hundreds of cases and put these into a design table, catia then duly reads the design table and runs all the cases automatically. The designer (with an FEA man sat alongside him!) was then able to run several design iterations on a large complex design and optimise his structure accordingly, all within an afternoon!

Like any other FE software you can (and should always) use tet10 elements. The only limitation on constraints and loading is the knowledge and experience of the guy driving the program, this is where a good understanding of free body diagrams and how to correctly apply BC's to a FE model is essential, otherwise you get the "every component I ever model is a cantilever" approach!

For optimising designs where a linear static analysis is adequate, then GPS (+EST, FMD, FMS and other modules) is probably the most advanced and powerful cad embedded FEA system available and dare I say it superior to many standalone FE systems. Unfortunately it comes at a price.

 

[PerKr]

It's way better than not having any FE analysis at all. As long as you know what you are doing with your loads and boundary conditions, it should work perfectly for at least simple analysis. But if you are going to use it professionally, it's probably a good idea to receive some training, because it's easy to make a mess when you aren't sure what you are doing.

A general question to all of you: How often would the CAD engineer really need to do an analysis with multiple load cases, strangely distributed forces, material imperfections and weird boundary conditions?

 

[Stringmaker]

"A general question to all of you: How often would the CAD engineer really need to do an analysis with multiple load cases, strangely distributed forces, material imperfections and weird boundary conditions?"

This could open a big can of worms concerning where the line is drawn between designers and analysts. My answer would be not often! Companies who need analysts have them. If they don't they will shortly after they make have product failures.

 

[GregLocock]

To put /that/ in perspective we have almost as many CAE engineers as CAD guys. Boundary conditions and real world correlation are the two biggest issues, in my opinion (and they are closely linked of course).



Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[123ash]

i dont know about iteration approach
Can anyone tell me