calculating Mass accretion rate for SN post-processing #211

mari2895 · 2024-03-20T00:38:24Z

PR Summary

PR Checklist

Adds a test for any bugs fixed. Adds tests for new features.
Format your changes by calling scripts/bash/format.sh.
Explain what you did.

Yurlungur · 2024-03-20T15:36:51Z

src/progenitor/progenitordata.cpp

+           ReduceInGain<class internal_variables::GcovCool>(md, 0, 0);
+  };
+  auto Mdot400 = [](MeshData<Real> *md) {
+    Real rc = 0.04;


Maybe make this a runtime set-able parameter. Might even be good to support an arbitrary number of these at different radii.

I am trying to imagine how is the second part going to work

ParameterInput can read a std:::vector so you could do something like this:

auto mdot_radii = pin->GetOrAddVector("analysis", "mdot_radii", std::vector<Real>{0.04}); for (auto rc : mdot_radii) { auto mdot = [=](MeshData<Real> *md) { return History::CalculateMdot(md, rc, 0); }; hst_vars.emplace_back(HistoryOutputVar(HstSum, mdot, "Mdot at r = " + std::to_string(rc))); }

Yurlungur · 2024-03-20T15:39:20Z

src/progenitor/progenitordata.cpp

-    return ReduceInGain<class internal_variables::GcovHeat>(md, 0) -
-           ReduceInGain<class internal_variables::GcovCool>(md, 0);
+    return ReduceInGain<class internal_variables::GcovHeat>(md, 0, 0) -
+           ReduceInGain<class internal_variables::GcovCool>(md, 0, 0);


Is the word class here actually required?

deleted and still compiles and runs so is not required.

Yurlungur · 2024-03-20T15:40:10Z

src/analysis/history.hpp

+        bool is_netheat = (v(b, iv::GcovHeat(), k, j, i) - v(b, iv::GcovCool(), k, j, i) >
+                           1.e-8); // checks that in the gain region
+        auto analysis = pmb->packages.Get("analysis").get();
+        const Real outside_pns_threshold = analysis->Param<Real>("outside_pns_threshold");


This needs to be moved outside of the kernel. It can't be called inside the par for.

src/analysis/history.hpp

src/analysis/history.cpp

Yurlungur · 2024-03-20T15:42:46Z

src/analysis/history.cpp

+                                                const F &f) {
+  Real div_mass_flux_integral;
+  div_mass_flux_integral =
+      -(f(b, 1, k, j, i + 1) - f(b, 1, k, j, i)) * coords.FaceArea<X1DIR>(k, j, i);


Remove the minus signs here, and add them to the value in mdot as the divergence doesn't have a minus sign in its definition but mdot does, due to the orientation of surface normals.

src/analysis/history.cpp

Yurlungur · 2024-03-20T15:45:04Z

src/analysis/history.cpp

+            auto rad = pmb->packages.Get("radiation").get();
+            const parthenon::AllReduce<bool> *pdo_gain_reducer =
+                rad->MutableParam<parthenon::AllReduce<bool>>("do_gain_reducer");
+            const bool do_gain = pdo_gain_reducer->val;


This can't be here---you can't access packages from within a par for on GPU. Move this above.

Yurlungur · 2024-03-20T15:45:25Z

src/analysis/history.cpp

+            auto analysis = pmb->packages.Get("analysis").get();
+            const Real inside_pns_threshold =
+                analysis->Param<Real>("inside_pns_threshold");


Can't be here. Move this out of the kernel.

Yurlungur · 2024-03-20T19:43:28Z

src/analysis/analysis.cpp

+  Real outside_pns_threshold =
+      pin->GetOrAddReal("analysis", "outside_pns_threshold",
+                        2.42e-5); // corresponds to entropy > 3 kb/baryon
+  Real inside_pns_threshold = pin->GetOrAddReal("analysis", "inside_pns_threshold",
+                                                0.008); // corresponds to r < 80 km
+  Real radius_cutoff_mdot =
+      pin->GetOrAddReal("analysis", "radius_cutoff_mdot", 0.04); // default 400km


This stuff might better belong in the progenitor package.

moved to progenitor pkg

Yurlungur · 2024-03-20T19:47:21Z

src/progenitor/progenitordata.cpp

+           ReduceInGain<class internal_variables::GcovCool>(md, 0, 0);
+  };
+  auto Mdot400 = [](MeshData<Real> *md) {
+    Real rc = 0.04;


ParameterInput can read a std:::vector so you could do something like this:

auto mdot_radii = pin->GetOrAddVector("analysis", "mdot_radii", std::vector<Real>{0.04}); for (auto rc : mdot_radii) { auto mdot = [=](MeshData<Real> *md) { return History::CalculateMdot(md, rc, 0); }; hst_vars.emplace_back(HistoryOutputVar(HstSum, mdot, "Mdot at r = " + std::to_string(rc))); }

Yurlungur

This is looking good! A few more minor changes...

src/analysis/history.cpp

src/analysis/history.hpp

Yurlungur · 2024-03-20T23:07:39Z

src/progenitor/progenitordata.cpp

+  Real LengthCGSToCode = units.GetLengthCGSToCode();
+  Real rc = 400e5 * LengthCGSToCode;
+  auto mdot_radii = pin->GetOrAddVector("progenitor", "mdot_radii",
+                                        std::vector<Real>{rc}); // default 400km


I don't think you want to conversion here.

Suggested change

Real LengthCGSToCode = units.GetLengthCGSToCode();

Real rc = 400e5 * LengthCGSToCode;

auto mdot_radii = pin->GetOrAddVector("progenitor", "mdot_radii",

std::vector<Real>{rc}); // default 400km

Real LengthCGSToCode = units.GetLengthCGSToCode();

auto mdot_radii = pin->GetOrAddVector("progenitor", "mdot_radii",

std::vector<Real>{400}); // default 400km

Yurlungur · 2024-03-20T23:08:47Z

src/progenitor/progenitordata.cpp

+    auto Mdot = [rc, LengthCGSToCode](MeshData<Real> *md) {
+      return History::CalculateMdot(md, rc, 0);


Suggested change

auto Mdot = [rc, LengthCGSToCode](MeshData<Real> *md) {

return History::CalculateMdot(md, rc, 0);

auto rc_code = rc*1e5*LengthCGSToCode;

auto Mdot = [rc_code](MeshData<Real> *md) {

return History::CalculateMdot(md, rc_code, 0);

src/progenitor/progenitordata.cpp

Yurlungur

As soon as tests pass here (and you confirm it works for you locally) I will merge.

mari2895 · 2024-03-21T02:41:22Z

As soon as tests pass here (and you confirm it works for you locally) I will merge.

All tests have passed, compiles, runs, produces reasonable output.

calculating Mdot for SN post-processing

4f761b2

mari2895 requested review from Yurlungur and jonahm-LANL March 20, 2024 00:38

Yurlungur requested a review from AstroBarker March 20, 2024 15:35

mari2895 self-assigned this Mar 20, 2024

Yurlungur requested changes Mar 20, 2024

View reviewed changes

comments applied

2b604be

Yurlungur reviewed Mar 20, 2024

View reviewed changes

mari2895 added 2 commits March 20, 2024 16:20

moved thresholds from analysis pkg to progenitor pkg

d4073c5

Mdot can now take a set or thresholds

994e2cc

Yurlungur requested changes Mar 20, 2024

View reviewed changes

few more details chainged to address suggestions

7f3bc17

Yurlungur approved these changes Mar 20, 2024

View reviewed changes

mari2895 added 2 commits March 20, 2024 20:56

typo

4658fa0

print statements deleted

c6f72c5

Yurlungur merged commit 799e02e into main Mar 21, 2024
2 checks passed

Yurlungur deleted the mg/mdots branch March 21, 2024 16:23

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

calculating Mass accretion rate for SN post-processing #211

calculating Mass accretion rate for SN post-processing #211

mari2895 commented Mar 20, 2024 •

edited

Loading

Yurlungur Mar 20, 2024

mari2895 Mar 20, 2024

Yurlungur Mar 20, 2024

Yurlungur Mar 20, 2024

mari2895 Mar 20, 2024

Yurlungur Mar 20, 2024

mari2895 Mar 20, 2024

Yurlungur Mar 20, 2024

mari2895 Mar 20, 2024

Yurlungur Mar 20, 2024

mari2895 Mar 20, 2024

Yurlungur Mar 20, 2024

mari2895 Mar 20, 2024

Yurlungur Mar 20, 2024

mari2895 Mar 20, 2024

Yurlungur Mar 20, 2024

Yurlungur left a comment

Yurlungur Mar 20, 2024

Yurlungur Mar 20, 2024

Yurlungur left a comment

mari2895 commented Mar 21, 2024

		auto Mdot = [rc, LengthCGSToCode](MeshData<Real> *md) {
		return History::CalculateMdot(md, rc, 0);

-    auto Mdot = [rc, LengthCGSToCode](MeshData<Real> *md) {
-      return History::CalculateMdot(md, rc, 0);
+    auto rc_code = rc*1e5*LengthCGSToCode;
+    auto Mdot = [rc_code](MeshData<Real> *md) {
+      return History::CalculateMdot(md, rc_code, 0);

calculating Mass accretion rate for SN post-processing #211

calculating Mass accretion rate for SN post-processing #211

Conversation

mari2895 commented Mar 20, 2024 • edited Loading

PR Summary

PR Checklist

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Yurlungur left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Yurlungur left a comment

Choose a reason for hiding this comment

mari2895 commented Mar 21, 2024

mari2895 commented Mar 20, 2024 •

edited

Loading