[Nek5000-users] Small random perturbations

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

Hi,
to circumvent the problem of specifying initial conditions (which should 
even be divergence free) we have made good experience with a 
physics-inspired tripping (i.e. a small vertical force which then 
amplifies due to lift-up). This has really solved our problems with 
getting a flow turbulent.

The fact that a flow relaminarises when going from low to high 
resolution I have seen before, in particular for marginally turbulent 
cases. The recommendation is to increase the resolution in small steps, 
and making sure that you let the flow adapt at the intermediate resolutions.

Best regards,
Philipp

On 2018-03-16 09:22, nek5000-users at lists.mcs.anl.gov wrote:
> Hello everyone
> 
> I have been troubled by the exactly same problem with Ali for the last 
> two months with the turbulent channel case. When I increase elements, 
> the flow is laminarized even though the Re is quite big (10000). I tried 
> many different ways. Even added a constant random forcing, but the 
> disturbance just disappear. (because the disturbance shape is not the 
> same as that in linear stability theory, it is possible that the 
> disturbance just decay. )
> 
> How I finally solve this problem is running a coarse mesh calculation 
> first, and then interpolating the result to fine mesh. Then I got a 
> sustaining turbulence and the result is quite satisfying. I hope this 
> could solve Ali's problem.
> 
> However, I still cannot understand why increasing resolution would 
> laminarize flow. hoping someone else could explain the problem.
> 
> Regards
> 
> Bien cordialement
> 
> Zhenrong JING
> 
> Doctorant (Doctor Candidate)
> 
> LHEEA
> 
> Ecole Centrale de Nantes
> 
> 1 Rue de la Noë, 44321 Nantes, France
> 
> 
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>