[Nek5000-users] Variable density flows

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

Hi Paul and Praveen

div U=0    In the low Mach limit,  in the absence of external heat sources q' , time dependent variation of the background pressure dPo/dt, and reactions source terms 

and Mass conservation reduces to  conservation of density along particle path so D rho /Dt=  drho/dt + u. grad rho= 0


Ammar





On Oct 11, 2013, at 1:27 PM, nek5000-users at lists.mcs.anl.gov wrote:

> 
> Hi Praveen,
> 
> If you want to follow the model of the Guermond paper that you
> pointed out then I shouldn't think low Mach is necessary given
> that, in their model, div U = 0.
> 
> I've not looked into this in detail, however, so I suggest
> undertaking a bit of exploration and test.
> 
> To specify variable density in Nek, you set ifuservp=.true.
> and ifstrs=.true. in usrdat, and then in uservp() you set:
> 
>      if (ifield.eq.1) then
> 
>         utrans = mydensity
>         udiff  = my viscosity
> 
>      endif
> 
> You also need to set lx2=lx1-2, (same for ly2 and lz2 (if3d));
> and set lx1m=lx1,ly1m=ly1,lz1m=lz1 in the SIZE file.
> 
> Paul
> 
> 
> ----- Original Message -----
> From: nek5000-users at lists.mcs.anl.gov
> To: nek5000-users at lists.mcs.anl.gov
> Sent: Friday, October 11, 2013 11:54:38 AM
> Subject: Re: [Nek5000-users] Fwd: Variable density flows
> 
> 
> 
> 
> Hi Paul 
> 
> For variable density case do I need to use the low mach number version or some other way ? 
> 
> 
> Thanks 
> praveen 
> 
> 
> 
> On Fri, Oct 11, 2013 at 8:24 PM, < nek5000-users at lists.mcs.anl.gov > wrote: 
> 
> 
> 
> In principle, this should work with the Pn-Pn-2 STRESS formulation, but 
> I don't think it has been thoroughly tested. 
> 
> Regarding large viscosity variations, work needs to be done on the 
> preconditioners for the viscous terms. This hasn't been a traditional 
> focus for Nek5000, which has been aimed at high Reynolds number flows 
> where the viscous solves are essentially trivial (compared to the 
> pressure, say). 
> 
> Certainly, there are things that could be done to be effective in 
> this space. I would start by running some test problems with the 
> STRESS formulation to see if you get the accuracy you're after. If 
> so, then we can look into performance enhancement via better preconditioning 
> strategies for the case at hand. 
> 
> HTH, 
> 
> Paul 
> 
> 
> 
> ----- Original Message ----- 
> From: nek5000-users at lists.mcs.anl.gov 
> To: "nek5000" < nek5000-users at lists.mcs.anl.gov > 
> Sent: Friday, October 11, 2013 9:16:24 AM 
> Subject: [Nek5000-users] Fwd: Variable density flows 
> 
> 
> 
> Hello 
> 
> 
> 
> 
> 
> Can we solve variable density incompressible flows with nek5000 ? For e.g. see the model given by eq. 2.1 here 
> 
> 
> http://www.math.tamu.edu/~guermond/PUBLICATIONS/guermond_quartapelle_JCP_2000.pdf 
> 
> 
> 
> Also, can nek handle variable viscosity with large variations ? 
> 
> 
> Thanks 
> praveen 
> 
> _______________________________________________ 
> 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 
> 
> 
> _______________________________________________
> 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