<div dir="ltr"><div><div><div>Hi Paul,<br><br></div>I have simply used the ext_cyl.rea and .map files so I guess it is dirichlet boundary condition at the inlet (which I have set to zero in usrbc since I am looking at perturbations), periodic boundary conditions on the sides, no-slip condition on the wall of the cylinder and classical outflow condition at the outlet.<br>
<br></div>I initially thought this could come from the fact I had been using a call to random_number to initialise vxp and vyp. I made sure to use dsavg to ensure that the velocity on each side of the interface between elements has the same value. It did not change anything. I also tried with a hand-made divergence-free perturbation. Did not changed anything either. I am kind of clueless, particularly because it works fine with all of my other cases. Attached are the files I am using if you want to peak a glance at them.<br>
<br></div>Cheers,<br>JC<br></div><div class="gmail_extra"><br><br><div class="gmail_quote">2014-02-06 5:11 GMT+01:00 <span dir="ltr"><<a href="mailto:nek5000-users@lists.mcs.anl.gov" target="_blank">nek5000-users@lists.mcs.anl.gov</a>></span>:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><br>
Hi JC,<br>
<br>
Are you trying to look at the perturbation from the<br>
steady cylinder base flow ?<br>
<br>
Am curious about your boundary conditions (though I<br>
recognize that your bc set for the external cylinder<br>
flow shouldn't be too different from the boundary<br>
layer case)...<br>
<br>
Just inquiring to get a better idea of what's going on.<br>
<br>
Paul<div><div class="h5"><br>
<br>
<br>
On Wed, 5 Feb 2014, <a href="mailto:nek5000-users@lists.mcs.anl.gov" target="_blank">nek5000-users@lists.mcs.anl.<u></u>gov</a> wrote:<br>
<br>
</div></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div class="h5">
Hi Nek's,<br>
<br>
I am trying to run a linearised DNS of the flow over a two-dimensional<br>
cylinder at Re=60 for illustration purposes using the ext_cyl example. I am<br>
however encountering some troubles... I have been able to compute the<br>
unstable steady equilibrium using a selective frequency damping approach<br>
without any trouble. However, it is quite a different story when it comes<br>
to the linearised DNS. Indeed, the perturbation velocity field blows up<br>
quite rapidly and I am not sure why. I have tried other flows (lid-driven<br>
cavity, boundary layer, ...) just to make sure I had no problem with my<br>
nek install and everything was running fine. I have also tried on three<br>
different computers to make sure it was not platform-dependent and it<br>
actually blew up every time. I have been using the linearised version of<br>
nek quite extensively for the past three years and but never had such<br>
problem.<br>
<br>
Is there something I have missed there?<br>
<br>
Thanks a lot,<br>
JC<br>
<br>
-- <br>
Jean-Christophe Loiseau<br></div></div>
Homepage <<a href="https://sites.google.com/site/loiseaujc/" target="_blank">https://sites.google.com/<u></u>site/loiseaujc/</a>><br>
<br>
</blockquote>
______________________________<u></u>_________________<br>
Nek5000-users mailing list<br>
<a href="mailto:Nek5000-users@lists.mcs.anl.gov" target="_blank">Nek5000-users@lists.mcs.anl.<u></u>gov</a><br>
<a href="https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users" target="_blank">https://lists.mcs.anl.gov/<u></u>mailman/listinfo/nek5000-users</a><br>
</blockquote></div><br><br clear="all"><br>-- <br><div dir="ltr">Jean-Christophe Loiseau<br><a href="https://sites.google.com/site/loiseaujc/" target="_blank">Homepage</a><br></div>
</div>