Does slip conditions work for quadratic elements?

[jgonzalezusua]

Does slip conditions work for quadratic elements?

[loumalouomega]

It could also be the utility we are using for the calculation of the normal (NormalCalculationUtils). We should confirm that we are passing through the correct function (and not the simplified ones for the simplex).

The error shown is from geometry.h itself, not the normal utilities. The normal compute the simplex only for triangles a priori.

[rubenzorrilla]

Frankly speaking, I've never tried it...

[loumalouomega]

A priori the algorithm is general, but maybe it depends on things not defined, let me check.

[loumalouomega]

I just chekced and a priori there is a test for the normal of the Quad3D9, so it should work

[loumalouomega]

Can you provide the geometry failing?

[loumalouomega]

Can you provide the geometry failing?

In fact can you provide the whole thing, because unitNormal is computed ina set of local coordinates, maybe your input is on global coordinates.

[jgonzalezusua]

Can you provide the geometry failing?

Do you mean the .mdpa? The problem is that quadratic conditions are not defined in master branch for the FluidDynamicsApplication, so I do not think you are not able to run it.

[loumalouomega]

Can you provide the geometry failing?

Do you mean the .mdpa? The problem is that quadratic conditions are not defined in master branch for the FluidDynamicsApplication, so I do not think you are not able to run it.

The error is in geometry.h, meaning that fails in a specific geometry. Can you provide the geometry points and input to check that it is a bug or something?

[jcotela]

As @loumalouomega is saying, from the code it looks like UnitNormal should be working for Quadrilateral3D9. I would suggest trying to find out which is the condition that is generating this error in your problem to see if there is something strange going on (is the geometry degenerate? Is it aligned in a particular way?)

[jgonzalezusua]

As @loumalouomega is saying, from the code it looks like UnitNormal should be working for Quadrilateral3D9. I would suggest trying to find out which is the condition that is generating this error in your problem to see if there is something strange going on (is the geometry degenerate? Is it aligned in a particular way?)

Ok, I will provide the output. Referring to the geometry, it is not degenerate as I am using a structured mesh in the wall.

[loumalouomega]

As @loumalouomega is saying, from the code it looks like UnitNormal should be working for Quadrilateral3D9. I would suggest trying to find out which is the condition that is generating this error in your problem to see if there is something strange going on (is the geometry degenerate? Is it aligned in a particular way?)

Ok, I will provide the output. Referring to the geometry, it is not degenerate as I am using a structured mesh in the wall.

Maybe not well defined (order of nodes in Kratos may be the issue)

[rubenzorrilla]

It could also be the utility we are using for the calculation of the normal (NormalCalculationUtils). We should confirm that we are passing through the correct function (and not the simplified ones for the simplex).

[jgonzalezusua]

Ok, I have found the error and there is nothing to do with the geometry. I registered the condition as it was 2 dimensional quadrilateral instead of 3d. Sorry for the inconvenience.

[rubenzorrilla]

OK. Closing then.