Hi, Wolfgang
>>I found one problem with inp file - in line "1 0 quad1 2 3 4". I
>> guess the second number should be nonzero here. I have try to set 1 and
>> get better results. Now I can see velocity changes around the
>> input-output bounds.
> Is this an input file you created yours
Evgeny,
> I found one problem with inp file - in line "1 0 quad 1 2 3 4". I
> guess the second number should be nonzero here. I have try to set 1 and
> get better results. Now I can see velocity changes around the
> input-output bounds.
Is this an input file you created yourself, or ar
Hi Daniel,
I found one problem with inp file - in line "1 0 quad1 2 3 4". I
guess the second number should be nonzero here. I have try to set 1 and get
better results. Now I can see velocity changes around the input-output
bounds.
I think with cell face iterator I have to use cell quad
Eugeniy,
my first guess would be that you are setting boundary conditions
inadequately and that this gives you a singular matrix.
What happens if you apply homogeneous Dirichlet boundary conditions for the
velocity everywhere?
Best,
Daniel
--
The deal.II project is located at http://www.deali
Dear Jean-Paul,
I have the same error with increased max_iterations value:
n_of_refines = 2
pressure_fe_degree = 2
dt = 5e-4
max_iterations = 1
Iterative method reported convergence failure in step 1. The residual
in the last step was 44.4628.
max_iterations = 5
Iterative method
Dear Eugeny,
Have you tried to increase the “max_iterations” in the parameter file?
Basically, with the way the tutorial program is defined the solver is currently
limited to use the number of iterations defined by the variable “vel_max_it”.
This seems to be prescribed using a parameter file, a
Hello,
i have tried step-35 with other ".inp" files that I have created. It
works fine with original inp but fails with any my inp with the same error
below.
I have tried simple rect without hole by hyper_rectangle and rect with
hole channel_with_cylinder. I use rects with about the same siz
