TDE Example Inputs

This page mirrors inputs/TDE_examples/README.md in the repository, which is the authoritative version.

TDE example inputs

These five input files reproduce the fiducial (FID) tidal disruption calculation of Jiang et al., An End-to-End Numerical Framework for Tidal Disruption Events with AthenaK (ApJS). A 1 Msun, 1 Rsun star built from the tabulated H/He equation of state is disrupted by a 1000 Msun black hole on a parabolic orbit with beta = 1, and the debris is followed to five times the fallback time of the most bound debris, P_mb = 88.04 t0 = 1.295 d. The physical and numerical parameters are those of Table 1 of the paper.

Step	File	Time (t0)	Box (Rsun)	AMR levels	Frame	What it shows
1	tde_01_disruption.athinput	0 to 21	64^3	9	translating	EOS-balanced star, first pericenter passage
2	tde_02_remap_box256.athinput	21 to 24	256^3	10	translating	remap to a larger box
3	tde_03_remap_box512.athinput	24 to 40	512 x 512 x 256	9	translating	second remap, BH-centered refinement shells
4	tde_04_bh_frame.athinput	40 to 70	512 x 512 x 256	8	BH rest frame	conversion to the BH rest frame
5	tde_05_fallback_lat.athinput	70 to 440	1024 x 1280 x 256	8	BH rest frame	fallback with localized adaptive time stepping

Pericenter is at t = 21 t0 (0.24 P_mb). Every step starts from the last restart file of the previous step through the <remap> block, which resamples the conserved variables onto the new mesh. Code units are M0 = 1 Msun, L0 = 2 Rsun (so the stellar radius is 0.5), t0 = 1271 s, and 4 pi G = 1.

The black hole is live from step 2 on (bh_live = true, with external_bh_gravity_source = true). In step 1 it starts outside the 64 Rsun box, where the gas potential cannot be sampled, so it follows the two-body orbit until the first remap (bh_live = false). A live BH is a softened point mass with an excised inner boundary whose position and velocity are integrated with the gas pull on it. The "BH rest frame" of steps 4 and 5 is the inertial frame in which the BH was at rest while it followed the two-body orbit in step 1; the live BH keeps its small reflex velocity and moves in it. Step 4 is the conversion itself, a Galilean boost of the state by the frame velocity done by the remap. Every remap carries the BH position, velocity, and acceleration from the source restart file. The reciprocal force (the gas pull on the BH applied with the same stencil as the BH force on the gas) needs LAT, so it is off in steps 1 to 4 and on in step 5.

Build

cmake -S . -B build_tde -D PROBLEM=tde_external -D Athena_ENABLE_MPI=ON \
      -D Kokkos_ENABLE_CUDA=ON -D Kokkos_ARCH_VOLTA70=ON \
      -D CMAKE_CXX_COMPILER=$PWD/kokkos/bin/nvcc_wrapper
cmake --build build_tde -j 16

Set Kokkos_ARCH_* for your GPU. The paper runs used NVIDIA V100 cards.

Equation of state table

The runs use eos_tables/chabrier2021_t13_helm_union_prad_640.table, the Chabrier et al. (2021) H/He table joined with the Tomida et al. (2013) and HELM tables, including radiation pressure (Sec. 4 of the paper). Build it with ./get_eos_table.sh in the top-level directory of the repository. The input files look for it one directory above the directory each step runs in.

Running the chain

Each step runs in its own directory, and the directories sit next to each other so that the relative restart paths in the <remap> blocks resolve.

REPO=/path/to/AthenaK-HXJ
ATHENA=$REPO/build_tde/src/athena
DECKS=$REPO/inputs/TDE_examples

mkdir -p tde_run && cd tde_run
ln -s $REPO/eos_tables/chabrier2021_t13_helm_union_prad_640.table .

for step in 01_disruption 02_remap_box256 03_remap_box512 04_bh_frame 05_fallback_lat; do
  mkdir -p $step
  (cd $step && mpirun -np 10 $ATHENA -i $DECKS/tde_$step.athinput)
done

A remap never runs on a restart (athena -r). Step 5 remaps and runs with LAT on, together with the reciprocal force of the gas on the BH, to 5 P_mb (t = 440.2). The settle steps of the remap run without LAT, and LAT starts after the last remap pass. A step that stops early is continued the usual way, with athena -r rst/<last file>.rst inside its directory.

The restart files used by the next step are

Written by	File	Time (t0)
step 1	01_disruption/rst/TDEExternalLTEPrad.00007.rst	21
step 2	02_remap_box256/rst/TDEExternalLTEPrad.00008.rst	24
step 3	03_remap_box512/rst/TDEExternalLTEPrad.00010.rst	40
step 4	04_bh_frame/rst/TDEExternalLTEPrad.00016.rst	70

Output numbers continue from the source restart file. A different source can be given on the command line, for example remap/source=../01_disruption/rst/<file>.rst.

Resources

Steps 1 to 4 run on 10 GPUs with 16 GB each, and step 2 needs all 10 cards because the AMR hierarchy grows quickly after pericenter. The memory of step 5 grows as the debris spreads. If a step runs out of memory, lower mesh_refinement/max_nmb_per_rank and use more ranks.

Variations used in the paper

• Self-gravity cadence (Sec. 8.1). Steps 1 to 4 with gravity/solve_dt=0.0025 added on the command line, compared with the value solve_dt = 0.03 set in the input files.
• LAT benchmark (Sec. 7.4). Run step 5 with time/lat=false problem/bh_reciprocal_force=false time/tlim=71.0 to get a checkpoint at t = 71, then restart it once with time/lat=true problem/bh_reciprocal_force=true and once without both (the reciprocal force needs LAT), keeping every other setting identical.
• HR run. The high-resolution run of the paper continues from the step-4 restart with 10 AMR levels and a finer vertical grid. It is not included here.

Synthetic observables

athenak_rt computes line-of-sight luminosities, spectra, and images from the bin/*.bin snapshots with the multifrequency H/He continuum transfer of Sec. 9 and the grey reference modes. See athenak_rt/README.md.