Dear All,

I'm trying to run a constricted pipe flow problem at Re=1000 in Nektar++ with the following boundary conditions:

 <BOUNDARYREGIONS>
            <B ID="0"> C[6] </B>  <!-- Inlet -->
            <B ID="1"> C[8] </B>  <!-- Outlet -->
            <B ID="2"> C[7] </B>  <!-- Wall -->
        </BOUNDARYREGIONS>

        <BOUNDARYCONDITIONS>
            <REGION REF="0">
                <D VAR="u" VALUE="2*(1-(4*(y^2+z^2)))" />
                <D VAR="v" VALUE="0" />
                <D VAR="w" VALUE="0" />
                <N VAR="p" USERDEFINEDTYPE="H" VALUE="0" />
            </REGION>
            <REGION REF="1">
                <N VAR="u" VALUE="0" />
                <N VAR="v" VALUE="0" />
                <N VAR="w" VALUE="0" />
                <D VAR="p" VALUE="0" />
            </REGION>
            <REGION REF="2">
                <D VAR="u" VALUE="0" />
                <D VAR="v" VALUE="0" />
                <D VAR="w" VALUE="0" />
                <N VAR="p" USERDEFINEDTYPE="H" VALUE="0" />
            </REGION>
        </BOUNDARYCONDITIONS>

The simulation seems converging but after a while starts to diverge and ended by Nektar due to exceedence of maximum iteration number. Time history plot of a node at the center of the pipe is attached to the mail. Does anybody have idea about the source of this problem? Is it directly related with the structure of the mesh I've used or do I set something wrong in my *.xml file?

Best Regards,
Kamil

Bu e-posta Avast tarafından korunan virüssüz bir bilgisayardan gönderilmiştir. www.avast.com