FEA training..where to start 2

1. Take some courses that cover FEA theory, not just how to use canned programs.
2. Read all the good books on the subject (Bathe and Gallagher are some easy to read authors).
3. Start with simple problems that you can also solve by hand and compare results using different element types and meshes.
4. Find a mentor.
5. Attend symposia to see how others solve different types of problems.
6. Whenever you get a chance, compare your analyses to real world test results. This is really important. You will learn a lot from comparing your analysis with test results (importance of boundary conditons and other assumptions) and it will teach you to be skeptical of the most realistic looking finite element models.

Nice recommendations by KSWpe.
Brad

Indyneer,

I agree with the post by KSWpe. It is very important to understand the limitations of the theory when using FEM. I have found that many people requesting analysis do not understand what they want and analysts and design engineers are driven to “just do an FEA” on a part. There are several users of FEA software that understand the theory no better.

I would recommend the following resources to chase down some additional information.

(Membership to NAFEMS may be cost prohibitive, but I would at least look into getting the magazine. It is very good.)


I hope this helps. Good luck!

Best regards,

Matthew Ian Loew
Sr. Engineer

I suggest you attend graduate school. It was'nt until I completed my Master's that I really understood what I was doing. Even better, if you attend while you are working or after a period of work, you will feel encouraged to work harder at the understanding and no so much the grade. My undergraduate knowledge (for doing FEA) was just pathetic compared to after grad school.

I had done some simple analysis before grad school and I now realize how poor of a job I did. Basically, garbage in = garbage out. To get garbage out in FEA very easy, and the problem is it can be convincing to yourself and others. This is a big problem with analysis today and many of the "press a button and get a pretty picture" codes.