[Nek5000-users] mesh problems

[nek5000-users at lists.mcs.anl.gov]

Hi Paul,

thanks a lot.

1/ we used the AMG preconditioner since it should scale better for bigger processor numbers.
I remember of having long waiting times at the beginning with the XXt. I will check it out.

2/ I got about 80 disconnected elements while running genmap. How many did you get? I will repeat 
the run with tol=0.1.

3/ did you use my file y5.rea and my file zpts_compare?

4/ did you have to enforce the outer elements to radius r=0.5 in the usrdat?

Janet and Joerg.

 

Am 14.01.2013 um 17:52 schrieb nek5000-users at lists.mcs.anl.gov:

>> 
>> I faced the same trouble with 2.17 million elements. Let me follow your
>> suggestion and set up a mesh in cell with z between 0 and 3 while keeping the
>> x-y mesh exactly the same, just to check if the same errors occur.
>> 
> 
> Hi Joerg,
> 
> I had no problems with the 2.17 million case.
> 
> I ran it with P=32768 and the XXt solver.
> 
> When I ran genmap, I used a tolerance of .01 --- this can
> be important for certain classes of meshes.  If the genmap tolerance
> is too loose then it's conceivable that you can have spurious
> connectivities in the graph that represents the sparse matrices,
> which might be the source of the problem you encountered.
> 
> Paul
> 
> 
> 
> 
> 
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users