[Nek5000-users] Variable density flows
nek5000-users at lists.mcs.anl.gov
nek5000-users at lists.mcs.anl.gov
Sat Oct 12 23:41:31 CDT 2013
The c=0 bc Is not consistent with the initial condition. You are generating a big error at t=0 that persists throughout your simulation. My understanding is you want to simulate the dispersion of a scalar that is already injected in a flow at t=0 right? I would spatially localize the concentration so that it decays to 0 at the inflow boundary ..a Gaussian exp(- x-a)^2) is a good candidate for your initial conditions
Ammar
Sent from my iPhone
> On Oct 12, 2013, at 11:07 PM, nek5000-users at lists.mcs.anl.gov wrote:
>
> Thanks Ammar for the help. My channel is quite long so I can use the 'ON' boundary condition at outlet for the passive scalar. I have a scalar equation
>
> c_t + u.grad(c) = div(D*grad(c))
>
> with
>
> c = 1 as initial condition
>
> I want to use inlet bc
>
> c = 0 at x=0 (inlet)
>
> but this leads to a discontinuous initial condition. Is it ok to use such bc ?
>
> To avoid specifying such dirichlet bc, I tried to use zero flux bc for c at inlet but then c did not change with time at all.
>
> Thanks
> praveen
>
>> On Sun, Oct 13, 2013 at 12:34 AM, <nek5000-users at lists.mcs.anl.gov> wrote:
>> If you are to use outflow BCs' at the channel exit plane it should be located sufficiently far from the inlet (i.e. the channel length long enough) to allow the flow and scalars to be fully developed d phi /dx =0
>> Insulation is usually used at solid walls where there is no gradient in the wall normal direction (no heat or mass flow across the boundary) so it works for temperature and scalars.
>>
>> try to at least check on mass conservation in an integral sense..
>>
>> Ammar
>
>
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/nek5000-users/attachments/20131012/401fe6d3/attachment.html>
More information about the Nek5000-users
mailing list