Help needed for fundamentals of FEA

[saxena_dheeraj]

hi all,

I am a new user of pro-mechanica. I have been using pro-e for quiet
some time now but recently i have started working on the
pro-mechanica module.I am currently doing the Pro-e tutorials and try
to get some help from the web, So just wanted to know if any of you
guys can give me some webaddress or send me some tutorials files for
pro-mechanica.



Also i have one other Question for the Experts.



I have done my graduation in Mechanical Engg and have no experience
whatsoever on softwares such as Ansys or Lsdyna or Nastran ...but since
my college days i wanted to work in the feild of FEA.......also i did
some major projects but was unable to get a break in FEA......

so how good it is for getting in to the feild of FEA with pro-mechanica
without the knowledge of these dedicated softwares and programing
languages.



Hope to find some encouraging replies

Regards

Dheeraj

 

[TeQie]

working with FEA makes good engineer even better but poor engineer even poorer...To use FEA, you must know the rules and tricks and the most important is NEVER rely truly on FEA result as it comes with error somehow... So hand clac is also a need... Back to your question about programming stuff, you dont really need to know much (or perhaps) anything about programming to use FEA, they are all based on friendly GUI...but the most important is how you represent your case study into the program (e.g. Pro/Mechanica).


If ur asking ansys, solid works or LS-Dyna FEA, they are all basically have similarity in application...But to do the real thing, you would be advised to buy some books behind the FEA theory such as Linear, non-linear analysis, contact analysis,convergence, multi adapative, single adaptive, meshing, AutoGEM, etc... They are all important for you to know...Many books are available in Amazon for you to read...Tutorial is not at all helpful without knowing what you should know. Hope this helps.

 

[jabbadeus]

A few point...


If you are solving using Mechanica, its mesher and solver uses Geometric Element Analysis, versus Finite Element Analysis (as in NASTRAN, ANSYS, etc.). FEA uses h-type elements in which convergence and better results are obtained by decreasing element size (and hence increasing the number of nodes). GEA uses p-type elements in which Mechanica increases the polynomial order of the shape function that describes the displacement along element edges.


In my years doing FEA, we spent most of our time meshing the models. Mechanica eliminates this need. There is a FEM mode available in Mechanica, and this will allow you to output data decks for solving in a third party package like NASTRAN. The downside of doing FEA in Mechanica is that you lose the seamless integration between the design side and the analysis side, as well as the optimization capability.


But TeQie makes a number of good points. Where people screw up structural and thermal analysis is how they choose to simulate loads and constraints. This is where good engineering knowledge -- and experience -- come in. Simulation is both a skill and an art, and getting hands on, actually performing (and screwing up) analyses is how one becomes better. And before jumping in the software, one should always do some hand calcs and check Roark's Handbook to get some rough order of magnitude ballpark answers. Always look at the results and ask yourself if they make sense. Don't simply trust the software, because garbage in, garbage out.

 

[Miko]

Hi all .. I have to saythank you both for necessary basic words.


Myquestions: When are you working on the real project, what are you doing first? Is the calculations on the paper first or simulation with some software? If I want to be a engineer in Mechanica, should I know a calculate some or all problems on the paper first?


I would like to start with Mechanica too, but I really doubt about enough informations from some free tutorials.And from my country I cant buy some books.I hope, thetime willfix it.


Michal M.

 

[TeQie]

You necessarily need to have a knowledge in paper-base calculation such as basic solid mechanics, strength of material including mohr circle that's what i would suggest... You can run your simulation first and then check with your hand calc...or vise versa...You can't just trust either one, but you have to judge both.


I recently run a thermal analysis on mechanica as well as hand calc...but both results show big irrelevant...So which one should i trust? then i do experiment to judge...


I also run a structural test recently for heavy industry material handling equipment...With idealization and 3D it shows a slight difference...But i do know that beam idealization cannot simulate the error however i still run a paper-based calculation and see what's going on...


In my summary, you can just simply run mechanica to make a judgement, BUT you must understand to have a big tolerance or safety factor to overcome what may likely to occur. But for precision engineering, such as thermal expansion of aircraft propeller blade (transient temperature), tolerance is really tight and you need to be realistic and not just rely on one (mechanica).


I have a mechanica structure and thermal (PTC internal use only) from the training session. But i would like to swapif anyone have a pro/moldesign training manual from PTC. Any clue?

 

[shivdeep]

I think if you are an engineer and studied all the subjects carefully like strength of materials, fluid mechanics, thermal, heat transfer etc, then you are quite able to calculate things on your own. you can choose any of the field to expertise. then calculate using your knowledge. calculate on paper which gives you a actual picture. try to breakdown the structure so that it gives you a clear view. for me these fea softwares are the sort of confirmation to your results in case your problems are bit complex. FEA is very bad tool in the hands of moron.


trysome books on FEA, tutorials are sh*t! try to get some knowledge about FEA how it works and how it calculates so that when applying it to a problem you can accertain the percentage of error. you can eaily get these books in uni, libraries, amazon etc.


good luck

 

[Miko]

Thank you very much for reply! .. I expected answare like this. Thanks