[Nek5000-users] Checking convergence to steady solution

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

Hi praveen, 

Yes you can ..

         call opsub2(uo,vo,wo,vx,vy,vz)
         call normvc(H1,SEMI,L2,LINF,uo,vo,wo)
         residu = L2/dt
         call opcopy(uo,vo,wo,vx,vy,vz)

The subroutine 

1) opsub2 computes the difference between the previous time-step and the current for velocity 
uo = uo - vx ..
2) Normvc computes the L2-norm.
3) opcopy copies the [vx,vy,vz] to [uo,vo,wo] to use it for next time step 

you can write the the value of residu to a separate file or to the logfile to see it.

Best regards,
Kamal 


On 31 May 2014, at 21:03, nek5000-users at lists.mcs.anl.gov wrote:

> Hello neks
> I am computing flow past 2d cylinder at Re=100. To get steady solution I run the unsteady solver with selective frequency damping. Looking at the forces on the cylinder and visualizing the solution I can see that I am getting a steady solution. But I would like to also quantitatively measure the convergence to the steady state. Is there a way that I can compute some residual to measure convergence to steady state ?
> Thanks
> praveen
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users