Finite Element Procedures by K.J.Bathe is a good book.
But suppose if you are new to this subject I would recommend you to also refer Theory and Applications with Ansys by Saeed Moaveni.
One more book which I feel is very good is Finite Element Handbook by Hayrettin Kardestuncer and Douglas Norrie.

Hi,
there are many threads, that have discussion on this thema.
Actually, you should realize that most books are written for different level of readers.
E.g. Finite Element Procedures by K.J.Bathe is one for advanced readers, i.s reader with strong basic FEM knowledge. So if you are one of them, you will find it as a nice book. Otherwise try another entry level book mentioned in those threads.
You can also learn FEM by using delivered manuals of any FEM software. ANSYS has a good FEM explanation in their manuals.

regards

Ya zuardy that's true.

R.D. Cook's books are good as well.

Introduction to the Finite Element Method, Hinton & Owen (Swansea Univ) is a very good text book for the beginner.

I need a book with more of structural engineering applications!!!
Anyone plz?

hajbann, the most FEM books, that i have read, discuss mainly on structural engineering applications, this is true, because the successful stroy of FEM began with its implementation in structural engineering, right? This is also to my know the reason why the K matrix always called stiffness matrix, even if it discuss about other discipline.
The most examples in K.J.Bathe are talking about structural mechanic problems. Am i misunderstand your statement?

cheers

Based on my experience, as a structural engineer, the books listed in this thread all deal with FEA and applications to structural analysis or structural/continuum mechanics.  That is, most theory and applications are developed based on structural apps while other examples (thermal, fluids)are introduced as appropriate.

I just wrote a reply to another question along this lines. I would recommend using FEA books related to theory as a reference. The best way to learn about any FEA code and to experiment with solutions is to create your own custom tutorial problems. Make up some simple problems with know solutions that you can work out for yourself using a pencil and paper, then create a few FEA model idealizations. The bottom line is "if you can't solve a problem with a pencil & paper, you will never be able to solve problems with a computer". As a 2nd approach look around on other FEA web sites and download their tutorial problems. Try to duplicate the results using your code. See this site for starters.

http://www.mscsoftware.com/support/online_ex/

David R. Dearth, P.E.
Applied Analysis & Technology
E-mail "AppliedAT@aol.com"

Thanks for all the book suggestions and input on FEM.  I have ordered the book by Cook.  

DRDearth's comments are right on the head, if we use a computer program like a black box solution to our engineering problems, without understanding the process behind it, how do we know it is correct?  How do you check it if you don't have a feel for the correct answer?  The computer is still just a "fast slide rule", you still have to know how to state the problem and the method used to solve it, for the answer to make sense.

Hi all,
i wonder why jheidt2543 & DRDearth have agreement about "use a computer program like a black box solution to our engineering problems, without understanding the process behind it", although this was fully discussed in Thread727-21832 : What Level of Theretical Knowledge do FEA Analysists Need?
Beside that, the original question is only about FEM Book Recomendations? So i think the responses should really only about FEM Book Titels. Otherwise the threads on this forums have no meaning, any suggestion?
I personally agree with the FEM Book titel suggestions in this thread. Especially the "Introduction to the Finite Element Method", Hinton & Owen (Swansea Univ), as you know the Zimkiwycs is from there, right?

cheers

Here's the deal guys…. The reference books suggested in previous posts are good. I used them in several courses I've taught at California State University (both undergraduate and graduate). The problem is… unless one is going to be writing FEA code from the ground up, the specific FEA process is (for the most part) transparent to the user. One can, of course, take the masochistic course and invert those big matrices by hand in order to learn the FEA solution mathematics, but in the long run... everyone just wants to know how to solve engineering problems using FEA techniques and simply wants to make sure they are doing things correctly. Ergo, it is far better to practice, practice, practice… on simple problems with known solutions. The same hoops and widgets that one has to jump through to solve little problems are the same hoop and widgets that one needs to jump through to solve the BIG problems. The only difference is typically the complexity of the geometry in the BIG (or real life) problems.

David R. Dearth, P.E.
Applied Analysis & Technology
E-mail "AppliedAT@aol.com"

Hi,
Finite Element Procedures by K.J.Bathe is originated also for educational purpose at MIT. It doesn't teach you "how to invert big matrices by hand", no, absolutely not!, but rather the methode to invert those big matrices and some consideration about numerical difficulty. Futhermore it explain the mathematical derivation of some elements and also about dynamic aspect.
May be you have just read the first paragraph about matrice formulation and solution of Bathe, don't you? I think, Bathe is not so stupid to teach you how to invert matrices by hand in the age of computational matrice softwares like MATLAB & co. Beside that, i solve simple problems not with a pencil and paper, but conveniently with mathematical software like MATHCAD.
Your propose to practice, practice, practice… on simple problems with known solutions was also proposed by zuardy before : "learn FEM by using delivered manuals of any FEM software". FEM manuals, e.g. NASTRAN, have always problems examples called verification problems, right?

cheers

I just asked for some suggestions on books, I wasn't trying to start an argument on the philosopy of computers and software in engineering.  As most of you, I've taken the structural analysis and design cousrses, matrix algebra and matrix mechanics and Fortran courses in school.  What I was looking for is a reference on FEM, which I didn't take in school.  We also didn't have the desktop computers that make life so easy now.  

I can have a secretary or my kids input the data into the FEM programs that are out there today!  The programs are user friendly, full of graphics and fun to use (a black box).  But, my secretary and kids can't interpret the structural meaning of the output.  That takes engineering intuition and experiance.  DRDearth's suggestion of working on small models with known solutions is a good one.  If you can understand the relationship between the program output and the simple, actual structural member it represents then, you can have some confidence in the answeres to the more complex problems.  "The longest journey begins with the first step" and proceeds one step at a time.

Thanks again to all for the thoughtful suggestions!

Hi all,
this is also common mistake in FEA. One is sure that all inputs to FEA program are correct. first, assume you just only input the element types and materials and use the free mesh, but free mesh is in most case not reliable, since it doesn't know your intend. Second, it is also common, that one has mistyped something, so it is a good habit to do a mesh/model checking prior to analysis/solving and of course a kid or secretary can't do it before he/she studies FEA first , right?

regards

So, FEA, pre- & postprocessing is not for dummy, but should be done by engineers, although it seems like an art  by modeling (preprocessing, just like CAD-modeler, it needs at least basic FEA knowledge, i.e through a good book and of course one should have engineering degree first, since there is no book titled with "FEA for dummys" like e.g "EXCEL for dummys"

cheers