#1 2008-08-28 10:54:42

todd_alan_martin

Member

Registered: 2008-03-06

Posts: 131

Arlequin method

Hi.

I've been looking at test case SSNP135B and I'm a little confused.

Am I correct in assuming the mesh representing the circular patch (mesh 2) is added to the main plate (mesh 1) in a cumulative manner? In other words, there is no actual hole in the assembled mesh?

If there is a hole in the assembled mesh, why are stress results found in elements where the hole should be?

Todd.

sébastien meunier

Member

From: EDF Lab Les Renardières

Registered: 2007-12-02

Posts: 287

Re: Arlequin method

Hi Todd,

In the Arlequin method, you have to define so-called weight parameters F1 and F2. These parameters stand for the importance you give to each model 1 and 2. You have the relation F1+F2=1.

The more F1 (resp. F2) is near to one, the more the model 1 (resp. 2) is predominant and the model 2  (resp. 1) is weak.

In ssnp135b, the model with a hole is predominant (see poids_fin). Therefore, stresses you obtain in zone GM9-GM111 are like there was a hole in the geometry.

I don't remember but I hope the stresses in the hole are near to 0. Anyway, do not pay attention to stresses in this zone.

Good luck !

---

Sébastien Meunier - EDF Lab Les Renardières

todd_alan_martin

Member

Registered: 2008-03-06

Posts: 131

Re: Arlequin method

Hi Sebastien

Thanks for the insight.
Unfortunately the stresses in the hole appear to be similar to those in the "fin" mesh region, unless I'm misinterpreting the output. That concerns me, when I want to display a colour contour map of the Von Mises stresses for example.

I will play with the test case a bit more.

Please, if anyone else would like to add to this discussion (in French or English), that would be great.

Todd.