FEA package choice 1

[jkstoib]

We have been using Abaqus as an FEA package for about two years and have attended some training classes. Compared to other CAE programs we use, Abaqus is not very user-friendly or intuitive for the occasional user. Also, since we almost always run simulations on virtual parts that will never see production in any form, benchmarking the results of the simulation against physical tests is impossible. Many times the results are not what experience or hand calculation would confirm as reliable. We feel that it might take us more time to learn all the ins and outs of this software package than we have time to invest. Does anyone have any personal experience using Abaqus compared to other codes such as Ansys?

 

[xerf]

- I think ABAQUS/CAE is more intuitive than Ansys.

- Usually the models (or better said formulas) used in hand calculation are based on a rougher assumptions than the FEM underlying theory.

- If you cannot reproduce by FEM the results you are able to obtain by hand calculation probably your FE model is wrong.

 

[corus]

Personally I've not used ANSYS in many years, but have used Solidworks and a few others as well as Abaqus. Since they brought in Abaqus CAE I find it very easy to use compared to the others. It's what you get used to I guess. It's obviously not a click and go package but does require a bit of thought. Sadly a bit of thought doesn't seem to be required in a lot of packages these days and the results can be rubbish.
I agree with everything xerf has said, particularly the 3rd comment.

corus

 

[prost]

I've used a few different FEA programs, and I'd say none of them are intuitive to everyone--programs I thought were intuitive really were just very familiar to me because of my experience with them.

I dunno, sounds like you could be putting the cart before the horse. It appears you are trying to figure out what FEA program to use because you don't like what you have, and are worried about how much it will cost to get up to speed on it, when perhaps you should be worrying more about what you're doing with the pretty pictures and the 9 sig. fig. numbers that are produced by it. You said you are having difficulty reproducing hand calcs; this I would consider unacceptable, an indicator of FEA modeling problems. Ask yourself, do you have the mechanics and FEA background, both experience and theoretical, to properly interpret what you are getting out of the FEA? My point is that you seem to be worried about a relatively small expense, model construction and training, while you could be overlooking real costs, that danger that engineers unfamiliar with FEA are trying to interpret those results.

Of course I could be just all wet!

 

[GBor]

prost,

You may be wet about the specifics of jkstoib's situation, but your comments are spot on! A star for you...