[Nektar-users] Periodic boundaries

23 Dec 2021

      *******************
This email originates from outside Imperial. Do not click on links and attachments unless you recognise the sender. 
If you trust the sender, add them to your safe senders list https://spam.ic.ac.uk/SpamConsole/Senders.aspx to disable email stamping for this address.
*******************
Hi All,

I am attempting to run a channel flow simulation with 2 sets of periodic
boundaries (inlet/outlet and the spanwise boundaries), but am running into
issues.  When running the solver, I get the following error:

Fatal   : Level 0 assertion violation
Local copy of allCompPairs not the same for all ranks

My boundary conditions are defined as follows:

        <BOUNDARYREGIONS>
              <B ID="0"> C[1] </B>    <!-- inlet -->
              <B ID="1"> C[2] </B>    <!-- outlet -->
              <B ID="2"> C[3] </B>    <!-- side1 -->
              <B ID="3"> C[4] </B>    <!-- side2 -->
              <B ID="4"> C[5] </B>    <!-- bottom -->
              <B ID="5"> C[6] </B>    <!-- top -->
        </BOUNDARYREGIONS>
        <BOUNDARYCONDITIONS>
            <REGION REF="0">
                <P VAR="u" VALUE="[1]" />
                <P VAR="v" VALUE="[1]" />
                <P VAR="w" VALUE="[1]" />
                <P VAR="p" VALUE="[1]" />
            </REGION>
            <REGION REF="1">
                <P VAR="u" VALUE="[0]" />
                <P VAR="v" VALUE="[0]" />
                <P VAR="w" VALUE="[0]" />
                <P VAR="p" VALUE="[0]" />
            </REGION>
            <REGION REF="2">
                <P VAR="u" VALUE="[3]" />
                <P VAR="v" VALUE="[3]" />
                <P VAR="w" VALUE="[3]" />
                <P VAR="p" VALUE="[3]" />
            </REGION>
            <REGION REF="3">
                <P VAR="u" VALUE="[2]" />
                <P VAR="v" VALUE="[2]" />
                <P VAR="w" VALUE="[2]" />
                <P VAR="p" VALUE="[2]" />
            </REGION>
            ...
    <\BOUNDARYCONDITIONS>

The bash script uses the following commands to convert the mesh and run
thet solver:

NekMesh -m peralign:surf1=1:surf2=2:dir=x -m peralign:surf1=3:surf2=4:dir=y
channel.msh channel.xml
srun --mpi=pmix_v3 -N $SLURM_NNODES IncNavierStokesSolver -v --use-scotch
channel.xml settings.xml

Can anyone provide some help on what needs to be done to correct this?  The
simulation runs without issues when I have only one set of periodic
boundaries specified (at the inlet/outlet).

Regards,

*Marc Goldbach*
PhD Student, Department of Mechanical Engineering
University of North Carolina at Charlotte