[Nek5000-users] Warning : DIV(V)-QTL too large

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

Dear Nek experts,

I'm trying to simulate a differentially heated cavity with a Rayleigh 
number of 10e9 (turbulent convection).
I modified the tutorial of the Rayleigh Benard cell to build my case.
I increased progessively the Rayleigh number. As my Ra increase, I get 
this warning "DIV(V)-QTL is too large" so I refined the mesh and the 
timestep as proposed in this thread 
https://lists.mcs.anl.gov/pipermail/nek5000-users/2016-January/003553.html.

I have now still this warning even if my courant number is very low 
(around 0.05).
So I would like to know if I could use a bigger time step keeping C 
below 0.5.

My other question is the meaning of QTL, the thermal divergence. In my 
simulation it is equal to 0, this is partly why DIV(V)-QTL is too large. 
Should I specify a QTL ?


Thank you in advance for your answers,
Best,
Samuel