Kratos icon indicating copy to clipboard operation
Kratos copied to clipboard

[StructuralMechanicsApplication] Adding patch test for axisymmetry

Open loumalouomega opened this issue 2 years ago • 13 comments

📝 Description

Adding patch test for axisymmetry

🆕 Changelog

  • Adding patch test for axisymmetry

@AlejandroCornejo the test fails, by a lot, any idea why?

loumalouomega avatar Aug 22 '22 07:08 loumalouomega

Can it be because the geometry has non-positive x components?

AlejandroCornejo avatar Aug 22 '22 08:08 AlejandroCornejo

actually not, I've checked

AlejandroCornejo avatar Aug 22 '22 08:08 AlejandroCornejo

Can it be because the geometry has non-positive x components?

Huuuuuuuuum, let me see

actually not, I've checked

Faster than me

loumalouomega avatar Aug 22 '22 08:08 loumalouomega

The element I'm pretty sure that is fine cause I've compared the result with another solver (Ramseries) and works fine. Maybe I'm wrong hehe

AlejandroCornejo avatar Aug 22 '22 08:08 AlejandroCornejo

The element I'm pretty sure that is fine cause I've compared the result with another solver (Ramseries) and works fine. Maybe I'm wrong hehe

Could be..., what I don't understand is why your cpp test works

loumalouomega avatar Aug 22 '22 08:08 loumalouomega

The element I'm pretty sure that is fine cause I've compared the result with another solver (Ramseries) and works fine. Maybe I'm wrong hehe

Could be..., what I don't understand is why your cpp test works

well my cpp is simpler, I'm only comparing stresses comparing with the current status, not against any analytical value. (I assumed that was ok)

AlejandroCornejo avatar Aug 22 '22 08:08 AlejandroCornejo

I suppose that the ordering of the nodes within the elements is consistent and not generating negative jacobians

AlejandroCornejo avatar Aug 22 '22 08:08 AlejandroCornejo

I suppose that the ordering of the nodes within the elements is consistent and not generating negative jacobians

A priori is the same geometry as the regular 2D tests

loumalouomega avatar Aug 22 '22 08:08 loumalouomega

And what about the 2*PI factor? does it have to be taken into account somehow in the test?

AlejandroCornejo avatar Aug 22 '22 09:08 AlejandroCornejo

And what about the 2*PI factor? does it have to be taken into account somehow in the test?

I would say that the displacements are the same. The 2Pi is to integrate

loumalouomega avatar Aug 22 '22 09:08 loumalouomega

Any news @AlejandroCornejo ?

loumalouomega avatar Aug 25 '22 16:08 loumalouomega

@AlejandroCornejo could you think about this?

loumalouomega avatar Sep 01 '22 07:09 loumalouomega

@AlejandroCornejo could you think about this?

👋

loumalouomega avatar Sep 13 '22 09:09 loumalouomega