Source for Articles on comparing FEA software: ANSYS, ALGOR, COSMOs 1

[jcarey]

I'm looking for independant reviews or comparisons ( in a magazine article/review/guide) on FEA software.<br>

The software I'm really looking at is ALGOR, ANSYS and COSMOS.

 

[lyletk]

You might try ASME's Mechanical Engineering Magazine. You can read it online at asme.org.

 

[KSE]

I recently took a 3-day course on using CosmosWorks 4.0 and GeoStar. I found Cosmosworks, an application that runs within SolidWorks, to be extremely user-friendly and straightforward (especially if you have gone through the course). Cosmosworks is mostly used for linear FEA analysis, though. If you own Solidworks (or are planning to buy it) and you're after a user-friendly interface, AND you'll be doing mainly linear analysis, I recommend Cosmosworks. Although I have not tried the other two packages you mentioned, Cosmosworks leaves me no desire to try anything else. Good luck to you.

 

[lyletk]

I am also a user of CosmosWorks from within Solidworks and I agree with KSE. It is a very user friendly FEA program. I have also used Cosmos/M 2.0 which is an expanded version of CosmosWorks. It is not as intuitive as CosmosWorks but has more capabilities. Also, Cosmos/M acts more like a DOS based program. A colleage of mine uses NASTRAN from within Autodesk's Mechanical Desktop. It has virtually the same capabilities as CosmosWorks but is more difficult to use. It has a wizard that likes to guide you through the FEA process but we have found that the results are not trustworthy (could be operator error). The results need to be verified manually (hand calculations) as much as possible. I'm with KSE, I recommend Solidworks with CosmosWorks.<br>

 

[coolbreeze]

I can tell you that NASTRAN linear static stress for Windows is about $10,000, COSMOS linear static stress is about $6000, and the basic ALGORE is about $3,000. Something to consider is the amount of time each has been on a PC. Originally NASTRAN was used on a UNIX platform, so their product had to be adapted to the PC as the PC became more and more powerful. ALGOR on the other has always been on the PC, and it's capabilities have grown with the PC's computing power. I have had the unfortunate experience of trying to adapt a UNIX based program into a WINDOW's environment and I know that it's not worth the effort. I am not certain NASTRAN took the time and effort to completely re-write the code, but if there is an interpreter running in the background, your in for trouble.

 

[Irwin]

I used COSMOS/M, ANSYS, I-DEAS, FEMAP and NASTRAN. My opinion is the next: COSMOS/M is good, but it is a very simple program. It can solve different problem, but for a big system estimation, this software is not the best. (e.g.: You can not define load cases with different constrain( restrain) conditions. You can modifie only the forces and pressures in this case) ANSYS- My opinion, this software is better, but the view of this software is still DOS. The connection between the PRO-Engineer is good. The automatic mesh is also good. The big advantege of this software the programabilty. (You can write routines, and you can put a buttom on a screen, therfore this software is very fast.) Disadvanteges: Everything base on geometry (very difficult to generate something from the element.) Measuring is very simple, modification is very simple (missing move to etc. command) I-DEAS - the best what I ever used. It has translator for NASTRAN and ANSYS etc. FEMAP - A cheap software with lot of translator. (NASTRAN for WINDOWS is also this) The best if you know the NASTRAN programming language. It has lot of advantage, but one big disadvantage of it the cancel and ok buttons. You need to use always there, therefore this reduces the efficiency. NASTRAN - It is very useful. I used only on UNIX operating system, and it was very good. For linear calculations this is the ace. My list:1. I-DEAS/NASTRAN 2.ANSYS 3.FEMAP/NASTRAN (Nastran for Windows) 4. COSMOS/M

 

[eks]

Take a look at &quot;Building Better Products with FINITE ELEMENT ANALYSIS&quot; by Vince Adams and Abraham Askenazi. I have just purchased but looks to be very practical and has full industry comparisons of pre and post-processors included.

 

[Anthonyr]

I have used ANSYS when it was primarily a &quot;mainframe&quot; style program. In recent years, I have used ALGOR and found it to be easier to use. The preprocessing and postprocessing tools are easy to learn. I feel you also need to review the element library of the programs to make sure that they will meet your needs, before you can make a final decision.