Complementary study material for FEM 1

[jd90]

Hello, as I come to end of my degree I feel like I have consumed so much knowledge some earlier more fundamental topics have inevitably been pushed out of my brain! Whilst I have an appreciation and understanding of general engineering concepts, such as mechanics, calculus, matrix algebra etc. etc. I feel that as I move into my career (which will involve FEA) I would like to re-visit some more fundamental topics and set my understanding in concrete.

My question is...besides studying FEM theory (numerical methods and solvers etc.), what would you recommend to accompany this. I assume materials and engineering mechanics (statics and dynamics) are probably the best place to revisit.

Any thoughts or recommendations are welcome. Also how do you keep on top of these things? Or if you have been working in this sector for a long time is it just second nature to you?

Thanks. J

 

[ShellsPlatesMeshes]

It depends on what exactly class of problems you are going to solve using the FEA. For example, if you intend to work with shells, a very important branch of mathematics to study would be the differential geometry of curved surfaces.

 

[Andries]

Hi jd90,

After 46 years at "the coal face" I can definitely recommend revisiting the fundamentals of strength of materials and really get to grips with issues such as free-body diagrams, boundary conditions, linear- or non-linear behaviour. If you are going to work in stress- or structural analysis you must be able to check your FEA results using "hand calculations".
I used FEA a lot in my career but always kept good strength of materials books handy. I would recommend that you buy Timoshenko's classics.

All the best.

Andries

 

[Illbay]

What these guys said. 33 years as a structural engineer, doing all sorts of computer analysis, and the big problem you have coming out of school - especially with a grad degree - is how slanted your view is toward the abstract stuff the profs deal with compared to the simple problem-solving skill set you really need when actually WORKING as an engineer.

Really, the theory of mechanics of materials, statics, dynamics, etc., haven't changed at all since Timoshenko and Gere et al wrote those books back in the 20s and 30s. He was an applied mathematician trying to catalogue as many closed-form solutions to solid mechanics as he could find, and the insights you get from his explanation is worth all the reams of FEA program output you can produce yourself.

If you don't understood what's actually going on vs. what the computer is telling you, then you will eventually get yourself into trouble, because the real-world doesn't care whether you coded the problem correctly or not.

I'm 58 and finished my graduate work 30 years ago; I can only IMAGINE how bad it's gotten since then (I can pick up any building code and tell that we're overthinking things a bit).

Go find the Timoshenko books, really. No one else ever did anything better to explain the theory, and I don't care how complex a formulation your finite element has in it, in the end they're just trying simulate the same thing Timoshenko was demonstrating. Sure, you can solve problems he couldn't but that's only a difference in computing power, not theory.

"No one is completely useless. He can always serve as a bad example." --My Dad ca. 1975

 

[LRJ]

Getting a good understanding of statistical methods would be my suggestion. More and more engineers are required to substantiate their 'engineering judgement' - much to the annoyance of older engineers who tend to have less knowledge on this stuff and seemingly believe that 'engineering judgement' is infallible - and this invariably requires a good understanding of the application of statistics in engineering practice. Statistical methods and FEM actually go quite well together too.

 

[jd90]

Thanks for the advice, this is exactly the response I was hoping for. I'll definitely search out the recommended books, and likely I will focus at a later stage. I will be working in a consulting environment so I will be covering a wide range of problems before I start to specialis in a specific area.