getfem-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Error of projection


From: Yves Renard
Subject: Re: Error of projection
Date: Tue, 16 Feb 2021 09:51:28 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.10.0


Dear Tetsuo,

It seems that on some distorded quadrilateral elements, the inversion of the geometric transformation converges well, but outside the element (you can trace the inversion in the function geotrans_inv_convex::invert_nonlin in bgeot_geotrans_inv.cc to see it) which means that there are possibly several solutions to the inversion. For the moment, I do not see how to solve that. Appart to shift a bit the tranformation such as

md.add_interpolate_transformation_from_expression("Proj1", mesh1, mesh2, "[X(1);-X(2)-1e-2]")

Best regards,

Yves

On 13/02/2021 11:04, Tetsuo Koyama wrote:
Dear GetFEM community

I want to solve the contact problem between two cylinders. The mesh is like the following.
This is a 1/4 model with symmetry conditions on the left and having Dirichlet condition [0.-0.1] on the top and [0.0, 0.0] on the bottom.
I used this tutorial `Example of wheel in contact (Assembly between two meshes, transformations, use of fixed size variables)` as a guide to create a program like the attached script "hertz.py", but an error occurred:

RuntimeError: (Getfem::InterfaceError) -- Error in getfem_generic_assembly_compile_and_exec.cc, line 1433 virtual int getfem::ga_instruction_interpolate_base::exec():
No valid element for the transformation. Probably transformation failed

I think my projection is correct.
md.add_interpolate_transformation_from_expression("Proj1", mesh1, mesh2, "[X(1);-X(2)]") 

Do you have any idea how to solve this problem?
Thanks a lot for reading.

Best regards Tetsuo

mesh.png


-- 

  Yves Renard (Yves.Renard@insa-lyon.fr)       tel : (33) 04.72.43.87.08
  INSA-Lyon 
  20, rue Albert Einstein
  69621 Villeurbanne Cedex, FRANCE
  http://math.univ-lyon1.fr/~renard

---------

reply via email to

[Prev in Thread] Current Thread [Next in Thread]