[FLASH-USERS] Temperature error while using PPM solver

[Max Katz]

Jeevan,

Oscillations in temperature like that around shocks seem to be fairly
common in my experience. Trying to understand how to get a better
temperature field, especially for real astrophysical equations of state,
was one of the motivations for Zingale and Katz (2015)
<http://adsabs.harvard.edu/abs/2015ApJS..216...31Z>. We found that the
temperature field is sensitive to both the method you use for predicting
interface states to the Riemann solver, and the Riemann solver itself.
While our technique was mainly implemented in the code CASTRO, we do show
comparisons to FLASH as well. One of the things we were able to concretely
say is that turning off the characteristic variable limiting in FLASH
improved our results on these particular test problems. However it does not
make the oscillations near the interface go away, it just makes them
smaller.

However, while we were able to make the temperature field look better for
these 1D test problems we used, I don't think we came up with a general
understanding of how to improve the temperature field in all cases;
possibly there is none. Even with these new interface prediction methods
and an accurate Riemann solver, I still see temperature fields just like
that in multi-D in some cases (though note that CASTRO is using an unsplit
solver and maybe you are not). Still, if you are interested, it might be
worth looking into trying these other methods we investigated -- maybe they
will work for you.


Max Katz
Ph.D. Candidate, Department of Physics and Astronomy
Stony Brook University
http://astro.sunysb.edu/mkatz/

On Wed, Mar 9, 2016 at 2:45 PM, jeevan dahal <jeevan.dahal1 at gmail.com>
wrote:

> Dear all,
>
> I am trying to run simulation of  a sine wave interface by shock strength
> of M 1.5. When I run my simulation using the PPM in FLASH code, I see an
> alternating temperature profile on the interface of sine wave. I have
> attached a picture of temperature profile to make it clear. In addition, I
> see a lot of diffusion in my simulation which is around 5 times the width
> of the zone even though I have turned the diffusion model to be off.
>
> I would appreciated any insights on solving the temperature problem. In
> addition, how can diffusion be reduced in a simulation?
>
> Thanks
> Regards,
> Jeevan
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20160311/fc697c2d/attachment.htm>