[Nek5000-users] Periodic BC on non-parallel boundaries

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

Hi David,

What you need is a cyclic not periodic BC. They are similar but not
the same. There is one important assumption in the current
implementation: The normal vector of a cyclic boundary needs to be
plane and orthogonal to the Z-axis. Then, just set IFCYCLIC to true in
the logical switches of your rea-file and give it a whirl.

Cheers,
Stefan

On 10/10/11, nek5000-users at lists.mcs.anl.gov
<nek5000-users at lists.mcs.anl.gov> wrote:
> Hi neks,
>
> I would like to employ periodic boundary conditions on the sides of a slice
> of an annulus.  I have constructed this domain by deforming the domain
> [0,1]x[0,1] in userdat2.  Setting the initial domain up with periodic sides
> does not work.  I`m guessing this is because nek is trying to mathc x- and
> y-velocities at the boundaries, when really I need to match velocities in
> the normal and parallel directions.
>
> Is the P boundary condition unable to handle matching non-parallel
> boundaries, and if so is there a work-around?
>
> If this makes it easier, I can settle for having only 1/4, 1/2 and 3/4
> slices, so the BC`s amount to mappings of +/-u,v to +/- u,v.
>
> Thanks!
>
> David
>