Fix boundary conditions for implicit radiation; add the option of pure radial rays and function for user-defined frequency grid #639
Conversation
…es are not working for implicit radiation module
… direction in spherical polar coordinate
|
Can one of the admins verify this patch? |
|
add to whitelist |
felker
changed the title
|
it's much better but also does not fix to machine precision -- for the equilibrium test I raised in the issue, the old code would not invoke the reflecting bcs and as a result velocities at the level of 1.0e-2 would develop after t=1.0. With reflecting bcs put in by hand through a user-defined rad bc, the velocity remained 0 to machine precision (1.0e-16). This fix produces v~1.0e-10 at t=1, obviously better but not equivalent to calling the rad bcs by enrolling them through the user-def framework. Took a brief look at the commit but did not see obvious cause of this discrepancy |
|
pgen and input files are in original post #603, it includes the user-def bc so just change between x1 bcs=user or bcs=reflecting to test the two implementations, no configure options other than -implicit_radiation |
|
Please check again. I just committed a fix. It is because the new radiation
physical boundary was not added during initialization.
…On Wed, Dec 11, 2024 at 12:02 AM Avery Bailey ***@***.***> wrote:
pgen and input files are in original post #603
<#603>, it includes
the user-def bc so just change between x1 bcs=user or bcs=reflecting to
test the two implementations, no configure options other than
-implicit_radiation
—
Reply to this email directly, view it on GitHub
<#639 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACCWEJZ2M363OG4D5ZOR74T2E7BKNAVCNFSM6AAAAABTL2WUSGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDKMZTGY2DCMJRG4>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
--
Best,
Yan-Fei Jiang
Research Scientist,
CCA, Flatiron Institute,
162 Fifth Avenue
New York, NY 10010
|
|
yes now the equilibrium is correct to machine precision, thanks for the fix @yanfeij |
| //---------------------------------------------------------------------------------------- | ||
| //! \fn void RadBoundaryVariable::ApplyRadPhysicalBoundaries() | ||
| // \brief Apply physical boundaries for specific intensities only | ||
| void RadBoundaryVariable::ApplyRadPhysicalBoundaries(const Real time, const Real dt) { |
There was a problem hiding this comment.
I am a bit confused by the introduction of this new function--- it seems to also be used for explicit radiation? Does that mean that the original issue #603 is also true for explicit radiation transfer?
|
No. Explicit radiation boundary is updated in the same task list as in
hydro (also the main integrator). But implicit radiation is not. I also do
not want to add implicit radiation to the main integrator, that is because
during iterations, implicit radiation needs to update the boundaries while
all other hydro variables are fixed. So I need these boundary functions for
radiation only.
…On Tue, Dec 17, 2024 at 1:18 PM Kyle Gerard Felker ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In src/bvals/cc/nr_radiation/bvals_rad.cpp
<#639 (comment)>
:
> @@ -398,3 +401,244 @@ void RadBoundaryVariable::PolarBoundarySingleAzimuthalBlock() {
}
return;
}
+
+
+//----------------------------------------------------------------------------------------
+//! \fn void RadBoundaryVariable::ApplyRadPhysicalBoundaries()
+// \brief Apply physical boundaries for specific intensities only
+void RadBoundaryVariable::ApplyRadPhysicalBoundaries(const Real time, const Real dt) {
I am a bit confused by the introduction of this new function--- it seems
to also be used for explicit radiation? Does that mean that the original
issue #603 <#603> is
also true for explicit radiation transfer?
—
Reply to this email directly, view it on GitHub
<#639 (review)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ACCWEJ7K6RFSR4J34LDWFMT2GBTGVAVCNFSM6AAAAABTL2WUSGVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDKMBZG4YTIOBVHA>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
--
Best,
Yan-Fei Jiang
Research Scientist,
CCA, Flatiron Institute,
162 Fifth Avenue
New York, NY 10010
|
Closes #603