Comparing Abaqus/Patran/Nastran for FEM Analysis

[theOrange]

I need to decide on whether to use Abaqus or Patran/Nastran for FEM analysis. However I am not sure what the advantages/disadvantages are of each FEM package. Also, searching on google has not come up with a list of situations where one is better compared to the other. I'm just looking for a generalized list.

I'll have a month (February) to try out both packages, and do some tutorials with each to get a feel for them.

 

[timthereaper]

Abaqus is the best nonlinear FEA solver out there. For linear problems, I personally think it's about sixes. I think it all comes down to how you like the "feel" of the package and what your company is currently using.

 

[theOrange]

I have the possibility of using either at the company I work for. What do you mean "I personally think it's about sixes."?

Yeah I found that Abaqus is best for nonlinear, and February I'll be able to test both.

 

[timthereaper]

Sorry about my phrase. I think the correct cliche is "six of one, half a dozen of the other". It means it's really about the same. I don't know that either program is "better" for linear problems.

Let us know what your opinion is after you've tested them!

 

[theOrange]

Sorry about my phrase. I think the correct cliche is "six of one, half a dozen of the other". It means it's really about the same. I don't know that either program is "better" for linear problems.

Let us know what your opinion is after you've tested them!

Oh alright got it. Sure, I'll post it later, and a short description of the project I'm working on.

 

[theOrange]

Just to be clear if anyone else reads this. I am not asking for the best one, I understand that is very difficult. I am just asking in what areas are each good at, so advantages/disadvantages of each FEM package (patran/nastran, and abaqus).

 

[CFDFEAGURU]

What type of solutions are you intending to solve?

 

[theOrange]

What type of solutions are you intending to solve?

Well at the moment I'm just looking for a general comparison between the two. What applications is Patran better than Abaqus, and what application abaqus is better than Patran.

As for my project: I'll be deforming a CAD model of a flat plate, towards a 3d scan. The 3d scan is of the plate but then in a deformed position. Anyways, the specifics of my project aren't important right now. At the moment I would just like to know in general where both Abaqus and Patran excel. I haven't been able to find a list of pros/cons or advantages/disadvantages of either.

 

[CFDFEAGURU]

I don't have any knowledge of Patran but I do no that for this type of non-linear behavior, Abaqus is your best software.

 

[theOrange]

Yep, I have read that Abaqus is stronger when concerning a non-linear problem.

For my project we aren't sure whether it is a linear or non-linear problem. I'm planning on running a test to see if it is linear/non-linear.

 

[jh0]

From your description it seems that the simulation involves plastic deformation. If this is the case, I'd say the problem is non-linear.

 

[theOrange]

From your description it seems that the simulation involves plastic deformation. If this is the case, I'd say the problem is non-linear.

No, the deformation is elastic. Therefore we don't know for sure if it is linear or non-linear.

Hopefully this week or next week I'll be able to do a test in Patran to see if it is linear/non-linear.

 

[NumericalFEA]

This is not as much about the actual software; the critical issues are: (a) accurate input data preparation; (b) sensible mechanical interpretation of obtained numerical results.
Below are couple of examples devoted to the mechanical interpretation of FEA results for some simple shells:

http://members.ozemail.com.au/~comecau/quad_shell.htm

http://comecau.com/quad_shell_wtank_ir.htm [Broken]

 

[theOrange]

This is not as much about the actual software; the critical issues are: (a) accurate input data preparation; (b) sensible mechanical interpretation of obtained numerical results.
Below are couple of examples devoted to the mechanical interpretation of FEA results for some simple shells:

http://members.ozemail.com.au/~comecau/quad_shell.htm

http://comecau.com/quad_shell_wtank_ir.htm [Broken]

I feel like people are misunderstanding the original question. I am not asking which software is the best. I am asking which one has an advantage over the other for specific situations. If such a list isn't possible to make, then so be it.

 

[timthereaper]

Well, it's hard to say one program has an advantage over another without saying "this program is better than this one for this problem". Most of the errors you will encounter are model-based (read: user error) and the only "advantages" are user's preference. Numerically, I think they're all about the same although you may agree with the results of one solver over another.

Either way, I think you have your answer. Summing up what's been said here, there are no clear advantages either way it seems except that Abaqus is better for nonlinear problems.

 

[NumericalFEA]

I feel like people are misunderstanding the original question. I am not asking which software is the best. I am asking which one has an advantage over the other for specific situations. If such a list isn't possible to make, then so be it.

Unless you present a very concrete list of those specific situations, your question is senseless. What kind of problems are you going to solve ? Static, dynamic ? Natural frequencies ? Buckling ? Thermal expansion/contraction ? Are your models thin-walled shells ? Or Mindlin plates ? Or perhaps they are multi-layered composite structures ? Be concrete, and you may get a concrete answer.

 

[theOrange]

Well, it's hard to say one program has an advantage over another without saying "this program is better than this one for this problem". Most of the errors you will encounter are model-based (read: user error) and the only "advantages" are user's preference. Numerically, I think they're all about the same although you may agree with the results of one solver over another.

Either way, I think you have your answer. Summing up what's been said here, there are no clear advantages either way it seems except that Abaqus is better for nonlinear problems.

Yep, this is basically what my conclusion is so far. Thanks for the help.

Unless you present a very concrete list of those specific situations, your question is senseless. What kind of problems are you going to solve ? Static, dynamic ? Natural frequencies ? Buckling ? Thermal expansion/contraction ? Are your models thin-walled shells ? Or Mindlin plates ? Or perhaps they are multi-layered composite structures ? Be concrete, and you may get a concrete answer.

That is exactly my question. In what problems does one have an advantage over the other (if it exists). For example, abaqus is a bit stronger is non-linear problems.

Maybe, for example, patran gives better results for buckling (or maybe not).

 

[theOrange]

I've decided to continue with Patran. I have a question about it though, is it possible to export the geometry of the deformed part?