[Nek5000-users] Passive scalar equation

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

Hi Andreas,

Unless you have a good reason (highly coupled and/or stiff scalar
equation system) you don't need to use CVODE to solve a passive scalar
equation with Nek. Yes, the discrete weak diffusion term (multiplied
by -1) is computed within axhelm (helm1). The implementation does not
assume that lambda is constant. Just work out the weak formulation of
the diffusion term and you'll see what I mean.

Cheers,
Stefan

On 10/3/11, nek5000-users at lists.mcs.anl.gov
<nek5000-users at lists.mcs.anl.gov> wrote:
> Dear all,
>
> I was having a look into the code to see how the terms of the passive scalar
> equation are computed.
>
> As far I understand, for iftran=true and ifcvode=true, the term
> "divergence*(lambda gradient(T))" is computed into the subroutine "makeq" by
> calling the subroutine "wlaplacian". The subroutine "wlaplacian" calls the
> subroutine "axhelm". It seems to me that the subroutine "axhelm" computes
> "lambda(laplacian T)" instead of "divergence*(lambda gradient(T))". Of
> course the two expressions are the same for a constant lambda, which is not
> my case. Why is it like this?
> The subroutine "axhelm" is quite complex, therefore I am not sure that I
> understood it correctly.
>
> Thanks in advance for your answer.
>
> Regards,
> Andreas.
> _______________________________________________
> Nek5000-users mailing list
> Nek5000-users at lists.mcs.anl.gov
> https://lists.mcs.anl.gov/mailman/listinfo/nek5000-users
>