Torsion problem with ASTER

Do you know a book or internet page with a simple torsion problem having an analytical solution in order to compare the results of FEM with it ? It seems pretty difficult to find examples where the cross-section of the bar isn't circular.

Francisc

Well, not on the internet.
I have an old book on mechanics I used many years ago (you are not allowed to make jokes about my age): Technische Mechanik - Teil 3 - Festigkeitslehre by Holzmann, Meyer und Schumpich, Teubner Stuttgart. Nice book, still valid.

It says, in German:
fuer Rechteck: n = h/b>=1 --> I* = ''Itorsion'' = eta2*h*b^3
h: height
b: width or vice versa, n>=1.

n 1 2 3 00 (infinite)
eta2 0.141 0.229 0.263 0.333
(if you need numbers inbetween, I can give them to you).

So, in your case, for very large n eta2 tends to 1/3 and
phie = Mtorsion*Length/(GI*).
I donot have all the dimension, so you have to determine the result for yourself

** kind regards - kees

Boloni wrote:

Thanks Kees,

now jokes about your age, promise

I got this formula too, now in order to meet the same conditions as the equations demands, it would be better to apply a torque and not a force in ASTER.

How is possible to do that ? Knowing that my model is 3D, do I apply the torque on a group of nodes of the surface parallel to the clamped surface ?

I don't (for some reason) have a handy .comm file I can upload, but you are right about the approach:

Create a node group on the face at the end of the structure you want to apply moment to, then create a single node a bit offset (and normal to) from that surface - link the two node groups with AFFE_CHAR_MECA --> LIASON_GROUP/SOLIDE

*
* * <-- apply moment here to get torsion around X
*

y
|
|____x

I think you need to create a new node outside the mesh, only for the purpose of applying torsion - give that a try