Ansys Vs ProMechnica

[Twoballcane]

Hello,

Just wanted your opinion on "what is the difference between Ansys and ProM?" and "which do you feel is better?".

I have to admit that I like ProM better because it is easier to use and I have correlated with hand calcs and emperical data so I feel comfotable with the answers. ProM is not as in-depth on what bricks to use and so forth, but analysis is half the answer until you test and correlate.

So what do you think?

Tobalcane
"If you avoid failure, you also avoid success."

 

[cbrn]

Hi,
I completely disagree, but of course I'm biased towards Ansys.
- correlation: depends on "how-to"s; where I work we managed to get within 1% error in displacement field and 5% error in stress field wrt tests even with the most complicated non-linear analyses. Correlation with hand-calcs is so full of examples within 0.1% relative error that I won't spend a word on it (remember hand-calcs are NOT the bible and that they are full of simplifying assumptions)
- ease of use: yes, the p-elem method is easier "per se", but afaik not so flexible. You can be extremely fast with Ansys when you know how to use it (you can also batch-execute and program it as you like, things that I doubt you can do with Pro/Mech), but I recognize the learning curve is very steep.
- general: don't consider a radical change in the FEA you use unless you feel your current one is limiting you in some way (knowing Ansys and a little bit of Mechanica, I can believe you may feel limited by Mechanica...)

Regards

 

[Twoballcane]

cbrn,

Thanks for responding

"you can also batch-execute and program it as you like, things that I doubt you can do with Pro/Mech"

If you mean run different analysis one after another like static, model, dynamic time, then random one after another in one shot, you can do that.

"but I recognize the learning curve is very steep."

Yes it is, actully Im trying to learn it now and WOW, its not tough because I now what Im looking for, but to look up the different elements for 2D and 3D and try to decide what to use where is kinda intimidating.

"don't consider a radical change in the FEA you use unless you feel your current one is limiting you in some way (knowing Ansys and a little bit of Mechanica, I can believe you may feel limited by Mechanica...)"

Its not that I feel limited in ProM, its that ANSYS is more respected among Analysis Engineers than ProM. And I agree, because Ansys is more in-depth in its use.


Tobalcane
"If you avoid failure, you also avoid success."

 

[cbrn]

Hi,
programmability: it's not only a matter of running several "studies" one after the others (it's something I call "queing"); ANSYS has its parametric programming language in the same way you can develop routines in VB, for example.
elem types: yes, it's intimidating, but you will realize how much mathematical rubustness there is here underlying, since each elem type is "tuned" for particular application(s). That also means "eficiency", because there is no point in using complicated formulation if your analysis needs only a more simplified one (one example which comes to mind is Timoschenko vs St Venant beams)...

Regards

 

[jagad5]

The major differences:
1. Pro/Mechanica uses p-elements while ANSYS generally uses h-elements. Pro/Mechanica increases the order of the shape functions to "converge". The shape functions of an h-element mesh are set when you mesh (usually linear, 2nd or sometimes 3rd order polynomials). To converge, you have to increase the number of elements.

2. Pro/Mechanica can only solve linear-elastic material deformation problems. It can handle geometric non-linearity (large deformations) but not yielding. We have not tested it lately, but Mechanica used to do a miserable job of computing contact stresses. We would use the contact features only for transfering loads, not predicing stresses.

3. ANSYS is capable of doing every type of FEA analysis needed by 95% of the world's FEA analysts. This includes material and geomertic linear and nonlinear problems, large deformation, plasticity, cyclic plasticity (if you've got the right data) creep, random and sine vibe.

There are other programs that can do certain types of analysis better than ANSYS. Abaqus is widely accepted as the superior tool for any non-linear problem, especially plasticity and contact. NASTRAN has traditionally excelled at dynamic problems (modal and vibe response) but a few years ago ANSYS caught up. I haven't read any comparisons in the last 3-4 years, though.

If you can only have one package and must solve these types of problems, then Pro/Mechanica is not the tool to have. Use ANSYS. If you can only have one, then ANSYS is the one to have. We have and use both but are planning to stop updating our Mechanica licenses because ANSYS will solve every problem we have while Mechanica can not.

All this is just my humble opinion.

Doug

 

[Twoballcane]

Thanks Doug for your input. You kinda gave me a nudge towards ANYSIS. I like ProM just because it is an easier transition from ProE, and the GUI is much easier to use.

Tobalcane
"If you avoid failure, you also avoid success."

 

[jagad5]

Look at Workbench. I find it even easier to use than Mechanica. Our designers use PRO/E and the interface from PRO/E to Workbench works for most geometries without difficulty. Workbench is the first program we've had besides Mechanica that could even mesh our components (jet engine fuel controls - 5000+ dimensioned features) without days-weeks of defeaturing.


Doug