Compute thermodynamic derivatives and implement dens+pres mode in primordial chem EOS #1302
Conversation
… to first derivative of sqrt (sound speed) w/r/t dens
psharda
changed the title
| @@ -27,6 +27,7 @@ struct eos_t:eos_base_t { | |||
| amrex::Real dsdr{}; | |||
| amrex::Real dpde{}; | |||
| amrex::Real dpdr_e{}; | |||
| amrex::Real G{}; // fundamental derivative (Thompson 1971), we use equation 2.24 of Menikoff & Plohr 1989 | |||
There was a problem hiding this comment.
Is it necessary for this to go in the eos_base_t?
There was a problem hiding this comment.
But this is in eos_t and not eos_base_t. We use eos_t for primordial chem and chem_eos_t for other Quokka problems.
There was a problem hiding this comment.
@maxpkatz Is there a better place we can put it?
There was a problem hiding this comment.
Sorry my comment above should have been less ambiguous: my concern was about adding a new quantity to eos_t which only some of the equations of state fill in. Part of the motivation for creating chem_eos_t was so that Quokka could modify that without affecting any of the other equations of state. So in my ideal world, we would only add this G field to chem_eos_t, and then add a has_G function analogous to has_pressure that wraps around the filling of this field in the EOS, so that the other applications could continue to use the gamma_law EOS without picking up a new quantity. But I may have misunderstood and didn't realize that Quokka was using eos_t in addition to chem_eos_t.
There was a problem hiding this comment.
We still use eos_t for the primordial_chem network, since we need the species fractions to compute the EOS quantities. I suppose we could add the species fraction array to chem_eos_t, but my recollection of the rationale for this was that we didn't want to include that when just using the gamma_law EOS.
There was a problem hiding this comment.
Yes, @maxpkatz created chem_eos_t so that we can use Microphysics EOS for Quokka problems that have no species.
|
catching up on this -- we should add G to all of the other EOSes if it is going to be in the |
|
That would be useful for us, so we can run the test problems in your EOS paper. Is it straightforward to do for the tabulated EOSes? |
|
Not at present, we don't have the second derivatives of pressure in Helmholtz. However, this is something we would likely have to add if we were to re-implement the electron-positron portion of Helmholtz from the Timmes EOS, so we could revisit it then. |
AMReX-Astro/Microphysics#1302 introduced roundoff level changes in this EOS.
…mordial chem EOS (AMReX-Astro#1302) Add the eos_input_rp mode for primordial chem EOS. Also, compute the sound speed and stores it in state.cs if pressure is defined.
This PR adds the
eos_input_rpmode for primordial chem EOS. It also computes the sound speed and stores it instate.csif pressure is defined.