Welcome to the forums. Please post in English or French.

You are not logged in.

#1 2021-01-20 20:36:15

Johannes_ACKVA
Member
From: Ingenieurbüro für Mechanik, DE
Registered: 2009-11-04
Posts: 716
Website

problem with Homard and cao boundary

hello

I try with the Homard Module in SalomeMeca to make a uniform refinement considering the curved cao boundary.

The very coarse original mesh is based on a cylinder geometry. The mesh is so coarse that it is a polygon with 7 edges.
I exported the  cylinder geometry in .cao format and used it as prescribed boundary in the Homard case.

I was expecting a 14-edges-polygon refined mesh, but it is a 7-edges-polygon as well

I attach some pictures.

Please which is the right manner to do that (if there is . . ) ?  Thank you,

Regards

Johannes_ACKVA


Attachments:
pictures_homard.zip, Size: 132.02 KiB, Downloads: 25

Ingenieurbüro für Mechanik
D 91717 Wassertrüdingen / Germany

www.code-aster.de                                                Training & Support for NASTRAN and CODE-ASTER

Offline

#2 2021-01-22 14:31:51

Nicolas
Member
From: EDF Lab Saclay
Registered: 2007-11-29
Posts: 98

Re: problem with Homard and cao boundary

Hello Johannes,
glad to have some news from you!

Perhaps you made mistake when you wrote: I see 9 edges and then 18 edges, and not 7 / 14.

I think that the problem appears because of the link between the CAD and the mesh. This link is made through a group. You have to define a group for the vertical surface of the cylinder, the non plane surface you want to follow. This group must exist in the CAD that is stored in the xao file as a boundary. And this group must exist in the initial mesh with the same name (most of the times it has been automatically built from the group defined in the CAD). The link is made by the names of that group.

If it is not OK, I suggest that you send to me your hdf or python file.

Regards.


Gérald NICOLAS
EDF R&D
Responsable du logiciel HOMARD

Offline

#3 2021-01-26 11:03:47

Johannes_ACKVA
Member
From: Ingenieurbüro für Mechanik, DE
Registered: 2009-11-04
Posts: 716
Website

Re: problem with Homard and cao boundary

hello  Gerald Nicolas,

thank you for your engangement in Homard!  I did not try your hint yet but hope to take the time during this week.

What I understand so far: it is now via xao (and only with this means) possible to refine a mesh and ensuring that the additional nodes will be on an arbitrary shaped 2-dimensional bounday. I looked before to the examples in the SalomeMeca installation:
....modules/HOMARD_V9_3_0/share/salome/homardsamples/test_1.*
....modules/HOMARD_V9_3_0/share/salome/homardsamples/test_2.*
....modules/HOMARD_V9_3_0/share/salome/homardsamples/test_3.*
The 3rd considers also a 2-dimensional boundary, but by means of defining a cylindrical surface. This is what since many years is included Homard but what is not really very general (outer boundary is seldom a primitive surface).

one more question:

it seems that the "xao-method" is not included in the basic command MACR_ADAP_MAIL for refinement during a Code-Aster run.
Does it mean that the Homard Module in Salome will deviate from the Homard in MACR_ADAP_MAIL? And is the latter in a certain sense obsolete?

many thanks! and regards
Johannes_ACKVA


Ingenieurbüro für Mechanik
D 91717 Wassertrüdingen / Germany

www.code-aster.de                                                Training & Support for NASTRAN and CODE-ASTER

Offline

#4 2021-01-26 11:31:08

Nicolas
Member
From: EDF Lab Saclay
Registered: 2007-11-29
Posts: 98

Re: problem with Homard and cao boundary

Hello Johannes,

Using the xao files allows you to deal with non conventionnal shapes: a skew roof for example. But if the boundary can be described by analytical functions such sphere, cylinder,... the classical way is effective. In your case, the simplest way is that you use the option FRONTIERE_ANALYTIQUE of MACR_ADAP_MAIL and that you describe the cylinder by its basic characteristic.

You're right: the xao solution is not implemented in MACR_ADAP_MAIL. There are some differences between the HOMARD module in Salome and the use of HOMARD inside the commands of code_aster. But that difference is low!

Gérald


Gérald NICOLAS
EDF R&D
Responsable du logiciel HOMARD

Offline