Entry++ level commercial FEA application

[4Pipes]

Hi all,

Please could somebody recommend an entry++ level commercial FEA appliation. Which would you chose? I am looking for a general purpose FEA package, mainly at lower end static stress for general application and pressure parts plus as many other features I could get for a price. There are many apps that would match my requirements. Choosing between them is more of a problem.

I would be most grateful for any suggestions or questions that I should be asking.
(I am not interested in template driven apps because I need the general purpose FEA. Secondly, models are also pretty easy to create which means that templates have now lost their earlier attractiveness).

 

[GregLocock]

Have you searched this forum for answers to that question?

What sort of elements do you need?




Cheers

Greg Locock

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

 

[GBor]

Are you looking for freeware? Or do you have a price limit? As Greg asked, what element types are you looking for?

Freeware: Calculix, but the GUI isn't very good
Minimual cost: Roshaz, which has some built-in solver capability, but also serves as a GUI for Calculix
Moderate to High Cost: The list is long and glorious! (NOT an exhautive list and not in order of personal preference) Abaqus, Algor, AMPS, Ansys, Cosmos, MSCNastran, NEiNastran, NXNastran, etc. etc.

Do the search on the forum; this discussion has been held several times.

Garland E. Borowski, PE
Engineering Manager
Star Aviation

 

[4Pipes]

Thanks,
I had looked around the formum and came up with some very useful data. I am looking a commerical rather than cheap or free. Something that has a a known and accepted reputation by both users and clients in process piping, petrochem, vessels industries. i.e. in GBor's "Long and Glorious" list. That's my problem. Most will serve my purpose of a general purpose FEA for both pressure parts and more general stress requirements. Shell, welded parts,bolted joints. i.e. Not FEPipe or similar template driven apps. Figuring out the ones that don't serve my purpose is the easy bit.
I have used Algor which I am more than happy with technically. NEI Fusion (Nastran Modules L1 and L2) also looks good on paper. I am probably safe with most. Its really down to the current price offers (generally) and the "bells and whistles" features that are included in the particular application. Mesh optimization etc. Hence my new post.
There are some very good offers on at the moment - I am the kid in the sweet shop. Without a choice I would be happy to pig out on the first thing I see. The problem is the choice.

 

[GBor]

I have long been an advocate for Algor and have used it for many years. I would never discourage it as a choice. Considering your particular situation, I suspect that you should look for the "low bidder". Algor has a well-developed interface and is pretty easy to use. NEiFusion is also great software, and generally comparably priced. AMPS, Ansys and Abaqus are probably more software than you need and their price tends to be higher for this type of application. Roshaz is running a special for $595.00. Calculix, while free, is an Abaqus knock-off and would give you more than enough power to grow into different areas of analysis. Roshaz is much easier to use than the Calculix GUI, but perhaps not as intuitive as some of these other packages...at $595 together, they will be hard to beat.

I'm passionate about FEA and could ramble for quite a while. It might be less painful if you could narrow the field for us a bit. If all you are trying to do and think you ever will do is static stress, any product that has been on the market for more than a few years probably does it sufficiently...otherwise, they'll be out of the business soon.

 

[xerf]

Before you choose a software you have to decide what types of problems you want to solve:
-stress analysis/heat transfer/coupled problems/fluid flow
-static/dynamic
-linear/non-linear
-material behaviors (elastic, hyperelastic, viscoelastic, various plasticity and hardening models, creep etc.)
-contact modeling calabilities
-buckling/frequency
-etc.

You have to look also at the pre-processing capabilities(how easy or time consuming is to develop a model) and post-processing capabilities (plots, data extraction, reports).

In general, the FEA software do not have CAD level geometric modeler (although some have decent capabilities). However, if you are interested in solving complex geometries, you might need a software which can import popular CAD formats.

Other factors to consider:
-how popular is the software in your field, can you exchange models with partners working in the same field or even on the same project
-how compatible the software is with any other software that you use
-available documentation
-hardware requirements and performance for the available hardware

Best.

 

[4Pipes]

Thanks again for your replies.
The most exciting change for me in FEA in the last couple of years has been in model creation in a reasonable time frame with practicable levels of complexity and at reasonable cost. (And without the endless bowing and scraping with the design office CAD dept to get that dimension modified before 3 weeks on Thursday). Geometry creation has changed from becoming one the biggest headaches for the end user to being the easiest and cheapest. (A bit like statistics when spreadsheets appeared). Even if a decent modeller is not included, you always have a fall back position with Alibre. Fall back is not the best description because you don't fall very far - if at all. Alibre is very competitively priced. Even the free version is very servicable. Algor has Alibre as one its many interfaces. I am not sure how it works with the free version but you always have IGES and STEP files.

 

[IFRs]

For basic structures, what is Cadre's reputation?

 

[spongebob007]

If you already have a CAD program, why not look into basic FEA that plugs in to your CAD software. ANSYS and COSMOS to name two, make lite FEA programs that work with Solid Works, Pro/E and the like.

 

[rb1957]

why not talk to some sales guys and get a trial license to see how you like the different packages ?

pretty much all the packages do pretty much the same job. some are easier to work with, but that is probably a personal perference thing (what intuitive to you maybe isn't to me).

what sort of problems do you think you'll need to address ... linear stress, non-linear, impact, ...

how scalable do you need to be ... adding modules later, adding seats (as your empire expands) ...

 

[4Pipes]

rb1957's advice about getting a trial license is spot on. I've just got a month of full license on a couple of packages. This far better than just a demo disc. I am expecting a month is the minimum to get a decent trial if comparing side by side alongside the day job. The other key advantage to the potential buyer is that the the price offers are exceptionally good at the moment.

 

[sjbjcl]

I have been using CADRE Pro to evaluate static stress in plates, beams, bolted joints for about 3 years. It is really easy to learn, runs really fast with minimum hardware, and does not cost very much. It is well worth evaluating for using for simple to moderate level structural analysis.