[Nek5000-users] what exactly is thermal flux and how to set it ondifferent sides?

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

Paul, 
Thank you very much for the quick answer, now it works fine

Van



> 6/17/2014 2:04:30 AM пользователь  (nek5000-users at lists.mcs.anl.gov) написал:
> 
> 
> Van,
> 
> It is a real flux:        q" = k dT/dn
> 
> If it is blowing up it's probably because of the nonlinear response of the Boussinesq
> equations to the I.C./B.C. combination that you are using --- you might be able to
> cure this with sufficiently small timestep.      For inhomogeneous Neumann conditions
> there is no stability issue of the type that's encountered with Robin (i.e., Newton 
> law of cooling) BC's, where a sign change can indeed lead to an instability.
> 
> 2.    I like things like:
> 
> if (y.lt.0.5) flux = 1
> if (y.gt.0.5) flux=2
> 
> as discriminators for different surfaces.
> 
> 3. The BC types are assigned in the .rea file.      The functionality for each type
> is prescribed in the .usr file.      Having extra definitions in the .usr file doesn't matter
> if there is no reference to them.
> 
> Paul
> 
> ________________________________________
> From: nek5000-users-bounces at lists.mcs.anl.gov [nek5000-users-bounces at lists.mcs.anl.gov] on behalf of nek5000-users at lists.mcs.anl.gov [nek5000-users at lists.mcs.anl.gov]
> Sent: Monday, June 16, 2014 4:55 PM
> To: Nek5000 Users
> Subject: [Nek5000-users] what exactly is thermal flux and how to set it on            different sides?
> 
> 1.
> if n is a normal direction,
> is it that
> flux = dTEMP/dn (what I supposed)
> or is it a    real thermal flux
> flux = k dTEMP/dn with k as conductivity?
> 
> A am asking because I dont understand the source of instability in solution when I simulate a stratified flow
> between two boundaries with
> TEMP=1-z as initial condition
> When I set flux on upper and lower boundaries to 0    the program works,
> but when I set it to -1 the program blows up in few steps.
> 
> 2. The next question is how to set differen user defined BC on different boundaries
> 
> if you set in .rea files lowcase "f"
> for two boundaries it means that the BC is determined by "flux" in userbc subroutine,
> but I have different expressions for fluxes for different sides.
> I have not found this question described in docs.
> 
> 3. And sorry for another qiestion:
> In benard example there are files like ray_nn.rea ray_dn.rea ray_dd.rea for different combinations of BCs.
> But only one .usr file - ray_cr.usr.
> I suppose that these BCs must be defined in userbc subroutine of .usr file.
> So how does it work?
> 
> Van
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
> 
>