Hello!
Im a total beginner in this field, but Im looking in direction of writing something on my own.
Can you recommend an good book to FEM modeling, matrix structural analysis, even maybe how to start writing your own program.
Thank you in advance
Honestly, if you really are completely new to this stuff, I'd recommend taking a class at a local university. I really can't think of any subject that would be harder to self-teach than matrix structural analysis and FEA.
Yeah I agree with you, but the problme is that I had listen to the class on FEM , but the man was a, complete idiot and personaly I think he didnt have a jack clue what he was talking about. So its very sad taht we got cut off prety much in that field.
So maybe for a completly new begining I though of geting something like this:
So it sounds like you actually have had at least one class in FEA, which isn't much different from matrix structural analysis. In that case, you probably can teach yourself.
Matrix Analysis of Structures by Sennett might be a good place to start also.
McGuire's Matrix Structural Analysis is a pretty good second book on the subject, but totally unsuitable for an intro. It has the best virtual work description that I've ever seen, starting from the very conception of the subject (i.e. not the stripped down version that's in most structural analysis books).
"Computer Analysis of Structures" Holzer
"Finite Element Modeling" Spyrakos
Most FEM books are general and not specifically geared to structural engineering. The Holzer book is very good for matrix analysis and some FEM. The Spyrakos book is good for descriptions of the elements and some examples. Spyrakos also has another book that is larger and more complete. I don't know the titles. Both of the books I listed are for structural engineers. I'd get the books and take a class. I've found the courses to be easy and ifyou do an FEM problem by hand, look out... It is tedeous and boring. But learning the different elements and how they work is helpful when calling the computer program tech support so you know what they are talking about.
Vincent, there is an unfortunate downside to the Holzer book. (especially for me because I have the book too, LOL) Dr. Holzer uses nomenclature and a matrix assembly method that is completely different from other folks. After learning it his way, it's very difficult to pick up another book or class and know what's going on. His assembly method is very computationally inefficient also, so it's important to be able to look elsewhere for at least that part.
Otherwise, I totally agree. It's a good intro otherwise.
I learned from Holzer's apprentice. He taught both ordinary structural analysis, matrix analysis, FEM, conceptual structures, and wood design. My professor was the best professor I ever had. He actually cared about the students learning. He tried to get the other professors to take continuing education on teaching but none would (which made him angry). We didn't even have an FEM book. My notebook is better than any FEM book I've ever seen. I've taken two other FEM courses and learned to build the stiffness matrices in three different ways. FEM is very simple. The stiffness matrix is not a hard concept. There is nothing especially difficult to "grasp". There are only two things that you really need to remember from FEM; convergence and element DOF. Other than that, it is just simple mathmatics and basic structures. Matrix analysis is extremely simple. They are mysterious because they are computer applications. They hide in a black box somewhere. They are not rocket science. The only reason you need a computer to do the problem is that they are very computationally heavy, especially FEM. It takes 12 pages to calculate by hand the reactions of a 3 node cantilever beam using FEM. I know. I've done it.
271828,
I don't know about the symbology Holzer used because it is the only symbology I know. I have looked at other books and taken other classes. It wasn't that difficult to change. Maybe I was just young when I took the courses. now I might have a harder time. The brain erodes over time. 10 years ago I learned another language in a year. Now it is taking me mnore than 2 years to do the same thing. Amazing.
You might be right, Vincent. One thing's for sure: Dr. Holzer took his teaching seriously. I had him for those classes also and then had them again 10-11 years later, so perhaps my problem was that my mind was less malleable at that point.
Obviously this is a personal taste issue. Here is 2¢ worth of my taste buds.
You asked for an introductory book, and one that covers both matrix methods and FEM. Take a look at "Computational Methods for the Solution of Engineering Problems", by C.A.Brebbia & A.J.Ferrante.
To give you the flavour (to keep the taste metaphor running) of the book , here is the list of chapter headings. The menu if you like.
» Computers in Engineering
» Matrix Algebra by Computers
» Matrix Analysis of Simple Structural Systems
» Solid Mechanics
» Approximate Methods of Solution
» The Finite Element Technique
» Fluid Mechanics
(The application of FEM to fluid mechanics was Carlos Brebbia's specialty, but - apart from the final chapter - this book is oriented to structural analysis.)
The book also gives FORTRAN listings of programs and subroutines related to various bits of the text.
It might be a bit hard to get this book. It was originally published in 1978 by Pentech Press of England, and simultaneously published in USA by Crane, Russak & Co. of New York. Hopefully someone has seen fit to re-publish it more recently.
If you go with the Bathe book, Finite Element Procedures, I think if you go to ADINA's website and buy the book there, you can get a limited copy of ADINA when you buy the book. The license will expire in a year, but it gives you something you can play with and test against whatever code you are developing.