Adaptmesh chaging boundary labels

General Discussion
1

Hello everyone,

I have a mesh resulting from adaptmesh function, now, I’m trying to use this mesh to solve a new problem, however, my problem has neumann and non homogeneous dirichlet boundary conditions and the solution is coming out wrong, is there a way to use adaptmesh but keep the boundary labels?

For exemple, this is my initial mesh

real x0 = 0.;
real x1 = 1200;
real y0 = 0;
real y1 = 2200;

int globalH = 2^3;
mesh calP = square(globalH,globalH);

since I’m using square, the boundary labels are 1, 2, 3 and 4, I would like to keep this labels in the adapted mesh.

I’m using this function:

ReMeshIndicator(calP, Etah, V0, etaTot[], 1, nDoFu0);

as specified in the documentation section 5.1.9 Adaptation using residual error indicator.

2

I think you might be having a different problem… adaptmesh shouldn’t change your boundary labels.

3

Interesting…
because I have another example and the only thing that changes is the boundary condition, I though it might have something to do with it.

Thanks though!

4

I’m imposing a global boundary condition like this:

varf muhgbndDir(unused,muh) = int1d(calP, 1)((1.*muh)) + int1d(calP, 3)((0.*muh));

and then, when I look at the solution, it seems that for these specific corner elements, it doesn’t work and I have no idea why, beacause the other side is basically the same

Screenshot from 2024-05-03 13-34-43
Screenshot from 2024-05-03 13-34-43691×111 23.4 KB

5

its unclear to me what you are trying to do. Your BC on boundary 3 is not doing anything since it is multiplied by 0. Did you mean to use the on() function?

6

either way, is the same problem
the boundary condition on boundary 3 is suposed to do nothing, is there just in case I wanted to change examples
but my problem is for the condition on boundary 1, it seems that it’s not working on that last element

7

What are you imposing on boundary 2?