Re: [sparta-users] EXT :Re: Re: Re: Re: modelling an emitting surface with spatially varying properties

[Steve P. <sj...@gm...>]

Hi Sekar - it has taken longer than I expected, but the PR with
enhancements to SPARTA that I mentioned
(ability to define and read in user-defined custom properties for each
surface triangle) is now merged into
main SPARTA (in the master branch on GitHub).  I think the only thing that
might be lacking to enable the
model you described are some enhancements to fix emit/surf to have it use
the precise custom properties
you define (output by the CFD code).

If you want to work with me on this, I can add the new options for fix
emit/surf.

Steve

On Wed, Sep 13, 2023 at 12:54 PM Ramakrishnan, Sekaripuram [US] (SP)
(Contr) <Sek...@ng...> wrote:

> Yes!  It is a one-way coupled simulation in which CFD++ generates input
> for DSMC simulation.
>
>
>
> Rama
>
>
>
> *From:* Steve Plimpton <sj...@gm...>
> *Sent:* Wednesday, September 13, 2023 7:24 AM
> *To:* Ramakrishnan, Sekaripuram [US] (SP) (Contr) <
> Sek...@ng...>
> *Cc:* spa...@li...
> *Subject:* EXT :Re: Re: [sparta-users] Re: Re: modelling an emitting
> surface with spatially varying properties
>
>
>
> Is this a one-time transfer of info to SPARTA?  I.e. one set of triangle
> boundary data for
>
> one DSMC simulation.  I.e. the boundary data does not dynamically change
> during
>
> a single DSMC simulation.
>
>
>
> Steve
>
>
>
> On Tue, Sep 12, 2023 at 7:32 PM Ramakrishnan, Sekaripuram [US] (SP)
> (Contr) <Sek...@ng...> wrote:
>
> I forgot to ‘reply all’ in my last email.  Hence, I am sending it again.
>
>
>
> My coupling procedure starts with generating a CFD solution with
> prescribed chamber conditions for the nozzle flow.  The geometry employed
> includes the body that is housing the nozzle.
>
>
>
> From the converged CFD solution, I extract Kn=0.01 surface.  The surface
> is smoothed and triangulated.  An interpolation tool included in the CFD
> software used, computes the required quantities, namely, number density
> (Nd), temperature(T) and species concentration at the grid points.  The
> grid is usually in the ‘stl’ format and the interpolated values are listed
> in a text file.  I use a MATLAB script to combine the 2files and create
> data in the format required by DAC97.  Therefore, I can create the data in
> any format required by the DSMC solver.
>
>
>
> I looked at ‘emit/face/file’ command and I believe it won’t work for me
> since it is limited to constant x or y or z surface while my Kn=0.01
> surface is quite arbitrary.
>
>
>
> Rama
>
>
>
>
>
>
>
> *From:* Steve Plimpton <sj...@gm...>
> *Sent:* Tuesday, September 12, 2023 10:30 AM
> *To:* Ramakrishnan, Sekaripuram [US] (SP) (Contr) <
> Sek...@ng...>
> *Cc:* spa...@li...
> *Subject:* EXT :Re: [sparta-users] Re: Re: modelling an emitting surface
> with spatially varying properties
>
>
>
> Hi again - Michel reminded me the that fix emit/face/file command is meant
> to do
>
> something similar to what you are asking for emission from a box face.  it
> has the
>
> added bonus that the mesh of grid points it defines can be coarser (or
> finer) than
>
> the grid cells on that face.  I.e. it does interpolation.
>
>
>
> So the first Q is can that command work for what you want ?
>
>
>
> If you really need this to work for surface elements doing the emission
>
> (presumably surface elements are the effective boundary of the DSMC model),
>
> then probably a new fix emit/surf/file command could do something similar.
>
> Probably without the interpolation.  I.e. it would be simplest if the CFD
> code that produces
>
> the file (or middleware code) knew about the collection of SPARTA
> triangles.
>
>
>
> Steve
>
>
>
> On Tue, Sep 12, 2023 at 8:16 AM Steve Plimpton <sj...@gm...> wrote:
>
> ok - thanks - that's helpful details.
>
>
>
> Can you also explain the mechanics of running the coupled calculation
>
> and the format of the exchanged file(s) ?
>
>
>
> Steve
>
>
>
> On Mon, Sep 11, 2023 at 3:43 PM Ramakrishnan, Sekaripuram [US] (SP)
> (Contr) <Sek...@ng...> wrote:
>
> Thank you, Steve, for your response.
>
>
>
> In the case of DAC97, I specify number density, temperature, species
> concentration and velocity vector(u,v,w) for each triangle on the surface
> of interest.
>
>
>
> Rama
>
>
>
> *From:* Steve Plimpton <sj...@gm...>
> *Sent:* Monday, September 11, 2023 8:05 AM
> *To:* Ramakrishnan, Sekaripuram [US] (SP) (Contr) <
> Sek...@ng...>
> *Cc:* spa...@li...; Steve Plimpton <
> sj...@gm...>
> *Subject:* EXT :Re: Re: [sparta-users] modelling an emitting surface with
> spatially varying properties
>
>
>
> The same examples/custom folder you asked about in your other message is
> where
>
> scripts that use the custom surf attributes are.  The folder is in the
> pending PR on the
>
> SPARTA GitHub site.
>
>
>
> If you can give more details about what per-surf info needs to pass from a
> CFD code to
>
> SPARTA and how you expect it to be communicated, we can likely add the
> necessary
>
> modifications to enable it to work.  E.g. is there a file with snapshots
> of per-surf values
>
> that the CFD code outputs prior to running SPARTA?  How frequently does
> SPARTA need
>
> to read a new snapshot?
>
>
>
> Steve
>
>
>
> On Fri, Sep 8, 2023 at 5:49 PM Ramakrishnan, Sekaripuram [US] (SP) (Contr)
> <Sek...@ng...> wrote:
>
> Thank you, Steve, for confirming that the read-custom-surf will enable me
> to simulate a one-way coupled CFD-DSMC simulation that I have been able to
> do with DAC97.
>
>
>
> Is there a sample case to illustrate the use of this new command?
>
>
>
> Rama
>
>
>
> *From:* Steve Plimpton <sj...@gm...>
> *Sent:* Friday, September 8, 2023 4:43 PM
> *To:* Ramakrishnan, Sekaripuram [US] (SP) (Contr) <
> Sek...@ng...>
> *Cc:* spa...@li...
> *Subject:* EXT :Re: [sparta-users] modelling an emitting surface with
> spatially varying properties
>
>
>
> With the new pull request that I mentioned it would be possible to assign
> per-triangle values
>
> for quantities like temperature and number density, possibly the other
> ones you mention.
>
> As custom per-surf attributes.
>
>
>
> What would also be needed is modifications to the fix emit/surf class to
> allow the mixture
>
> properties (which is what the current version use, to be overridden by the
> custom per-surf
>
> attributes.
>
>
>
> Steve
>
>
>
>
>
> On Thu, Sep 7, 2023 at 6:31 PM Ramakrishnan, Sekaripuram [US] (SP) (Contr)
> <Sek...@ng...> wrote:
>
> Is it possible to model an emitting surface with unique values for number
> density, temperature, species concentration and velocity assigned to each
> triangular element on the surface?
>
>
>
> Rama
>
> _______________________________________________
> sparta-users mailing list
> spa...@li...
> https://lists.sourceforge.net/lists/listinfo/sparta-users
>
> _______________________________________________
> sparta-users mailing list
> spa...@li...
> https://lists.sourceforge.net/lists/listinfo/sparta-users
>
>