Combining various elements in FEA
Vložit
- čas přidán 13. 07. 2024
- Learn more about FEA! Check my FREE online course: enterfea.com/10xfea/
It's unreasonable to model everything with 3D mesh. 2D and 1D elements offer a lot of flexibility, but you will often need to connect all of them together! This is what I discuss in this video! - Věda a technologie
Dude! Your channel and blog are a gold mine! I'm so grateful for your amazing education. Thank you!
Thank you so much Nicolas, you are very kind! I'm happy that you find this useful :)
unbelievable !!! this is the stuff i was looking for .....fantastic !!!
I'm really glad that you like it :)
Fantastic video! I literally searched for this answer for hours before finding your channel.
Subscribed! Thanks for the wealth of knowledge.
Thank you, Stuart! I'm really glad that you found this useful :)
Very good, I was looking for how to mix plate with solid elements!
Thanks Mate! I'm glad that you found this useful :)
Great 🎉 You got a new follower
Awesome :) Welcome onboard :)
You are amazing sir
Thank you Sanjeev, I'm glad that you like the video. I think FEA is far more amazing than I am :)
I'm more clear on connections than before. Is there any blog where i can get more on it.
Thank you łukasz skotny sir for valuable knowledge
Hey! Thank you for the kind words!
I think you may be interested in this: enterfea.com/connection-rigidity-5-things-to-know/
2 questions: a) 17:14 both surfaces are connected with common nodes, that's fine. But how is the small surface parallel to the solid supposed to be connected to the solid? A specific contact type? How can we also make it rigid differently than to assign a material with high E to it? b) 19:52 RBE2 connects a point to an edge or surface. So in the previous case of connecting a surface to the solid how can this be employed to do the trick?
A note to the "a" question - the small surface needs to have thickness assigned to it. I gather that it should be as thin as possible but infinitely rigid. I can't think of a way to make it both thin and rigid in any other way than to assign a huge young's modulus to it but there's still a limit to that in Nastran for example.
Hey!
This will depend on how accurate you need to be, and what you are aiming at. if the interface place is the most important thing... just do everything with solid or shell, and it will work better. You will indeed need a relatively thin shell, but I wouldn't necessarily go crazy thin, just to avoid huge E - all you need is for the surface to be able to transfer bending as a pair of forces - this is relatively easy to calculate by hand to check the required capacity. As for the connection to the solid... just make HEX elements small enough, so the surface elements and solid elements have common nodes :)
The RBE2 won't work here, simply because you would make the connection infinitely rigid in the perpendicular direction... you would need RBE2 elements on a few vertical nodes of solid connected to the single node on the shell (in a vertical plane), and then add those RBE2 elements on each node (a lot of work, and mesh has to align with Solid and Shell... while I did that once in a model where I needed it... I wouldn't really recommend that route.
All the best!
Ł
it would be interresting to know, what you think about 'web spline fea' ...
and where the differences is to the 'normal' fea.
it is a very nice topic, but i have a problem to understand the numerical mathematics behind the conventional fea elements and the new web spline mesh ...so i dont understand why this new method is better.
but it seems that this new ' web spline mesh thing' is the future...
thanks in advance👍
and ofc a nice vid. i subscribed 😁👍
Thank you for the kind words, I'm really glad that you like the video :)
As for your question... I have never even heard about this web spline fea stuff that you mentioned... so I can't say a thing about it I'm afraid :(
I regret that I couldn't help you with that...
@@Enterfea no problem 😆 thank you anyway. 😊👍
but you should read a little bit about this 'web spline fea'.
it seems like you have much better convergence with this aproach than with quadratic-node finite elements... it is much faster and you have no more singularitiy problems.... (they say 🤔)
like the conventional fea this approach comes from the aero engineers
Thanks and bye 👍
@@kolavithonduraski5031 perhaps one day... for now, I have waaaay to much stuff on my plate. But thank you for the suggestion :)
All the best
I follow your blog.. I must say your tutorials and blogs are very helpful in giving clear understanding on various topics..
I work as a FEA Engineer and I am looking for some books for thorough learning of tool.. So I will be grateful if you could suggest me any book?
Hey Amba... At this point, I feel I really need to create a copy-paste answer for the "book question"... I'm afraid I will disappoint you. All FEA books I saw/read were awfully theoretical and not useful. I mean, sure I've learned a bit here or there, but from the practical standpoint - it was my job and experiments that taught me :(
If you are searching for a book that will teach you the practical aspects of doing FEA (and not writing your own solver), that may be a long and frustrating search I'm afraid. If you will find such a book, definitely let me know. I will buy it and read it, and if it is any good I will recommend it to others as well :)
All the best!
@@Enterfea No issue! Thanks for your honest remark.Even I experienced the same problem with the FEA books..
@@ambasinha4993 Yea, this is tough. No worries though - I'm working on my FEA book - it should be out in months (but like 7+ not 2-3) :)
@@Enterfea OK, that would be great.. Wish you good luck!
@@ambasinha4993 Thank you so much! I'm actually sitting in a cafe right now... writing a chapter/example about nonlinear geometry :)
I don't know about you man but I'm modeling entire aircraft... 3D meshing is not popular with me. Shell elements ftw!!
Totally 2D mesh to rule them all!