Assumption 1) You want to write your own program from the ground up.<br>
<br>
Have you taken a theoretical FEA class? If not, then I highly recommend it. The math is very complex. If you do not understand the math then your program will not work. BTW, FEA theory classes tend to be Master's level and for good reason. I was a senior undergrad and the math was at my limit. Be prepared for working with partial differential equation matrices. (And if you didn't understand that, don't bother writing the program.)<br>
<br>
Assumption 2) You want to write a script or a couple of routines for a FEA engine/program such as strudl or ansys.<br>
<br>
I don't have any books to recommend. Mine is 10+ years old and very theoretical. There are several books on each of the major FEA programs. However, there is a lot more to understanding FEA than just typing the correct formats for the program's statements. For instance, which element types do you use for which geometry and stress case? You might be able to find a good book to cover this. I suggest taking a class at an engineering college. <br>
<br>
If anyone else knows a good book on FEA practice (not theory), I'd be interested in it also. <p> Imagineer<br><a href=mailto: > </a><br><a href= > </a><br>