[Nek5000-users] G.E. Global Research: Multiphase flow

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

Greg,

It's really not designed for multiphase -- sorry.

Documentation is up and coming - google nek5000 for project webpage.

There is an svn repo (documented on webpage).

Principal strength of the code is incompressible and low-Mach
number single phase at very large processor counts (>100,000).

Paul

On Fri, 9 Apr 2010, Laskowski, Gregory M (GE, Research) wrote:

> Dear Ramesh and Paul-
>
> I would be interested to learn more about nek5000.  I worked with it briefly
> while I was a postdoc at Sandia National Labs in Livermore several years
> ago.  I'm sure it's advanced since then. Can you please either point me to
> documentation of code capabilities or send me a report directly?
>
> I'm interested in DNS/LES of single phase and multiphase flows (2-phase and
> 3-phase) for some fundamental (canonical) type flows.  Does nek5000 have
> (scalable) VOF or Levelset capabilities?
>
> Regards,
>
> Greg
>
> Gregory M. Laskowski, PhD
> Computational Heat Transfer Lab
> GE Global Research Center
> Niskayuna NY  12309  USA
>
> (o) +1 518 387-6126
>
> -----Original Message-----
> From: Gupta, Anurag (GE, Research)
> Sent: Tuesday, April 06, 2010 1:53 PM
> To: bramesh at anl.gov
> Cc: Paul Fischer; Laskowski, Gregory M (GE, Research)
> Subject: RE: G.E. Global Research: Multiphase flow
>
> Hi Ramesh, Paul
>
> I've asked Greg Laskowski, who's leading one of the relevant projects here,
> to get in touch with Paul directly. As I mentioned before, he was keenly
> interested both in the experiment and the LES validation work that Paul
> showed me during my visit
>
> regards
> Anurag
>
>
> -----Original Message-----
> From: Ramesh Balakrishnan [mailto:bramesh at anl.gov]
> Sent: Tuesday, April 06, 2010 11:50 AM
> To: Gupta, Anurag (GE, Research)
> Cc: Paul Fischer
> Subject: Re: G.E. Global Research: Multiphase flow
>
> Hello Anurag,
> I understand from Paul (cc'ed) that the nek5000 spectral element code does
> handle complex geometries and does have the necessary models to handle
> multiphase flows. That is, it does have Eulerian-Lagrangian coupling (it
> does two way coupling). However, the Langrangian (particle) model and two
> way coupling have not been tested extensively. If your team could tolerate a
> bit of a learning curve, it would be worth the effort to try this out. On
> the other hand, the Navier-Stokes solver in nek5000 scales extremely well.
> It has been run on all 40 racks of the Blue Gene/P and has been run on
> almost all the 70 racks of the Blue Gene/P at Juelich. Do let us know how
> you'd wish to proceed.
>
> Sincerely,
> Ramesh
>
> Ramesh Balakrishnan wrote:
>> Hello Paul,
>> I understand, from Anurag Gupta (cc'ed), that  G.E. Global Research is
>> going to be working on projects that involve multiphase incompressible
>> flows in complex geometries. Could you please let us know if your
>> nek5000 code handles multiphase flow with Lagrangian particles?
>>
>> Thanks,
>> Ramesh
>