Reapply GK passive advection work (PR #1084) by Antoinehoff · Pull Request #1102 · gkeyllorg/gkeyll

Add a new GK equation object to do passive advection, with a velocity given in the input file. Add sample regression test (gkylcas 16cfe76398a327fa8617ed162509585f0c6c6f47).

manauref

authored and

Antoinehoff

committed

c7a6260

Change field_id in rt_gk_passive so that the omega_H dt doesn't show up. We should make it so that omega_H is only computed if the field evolves. Passive eqn object is clearly not ready as the resu…

manauref

authored and

Antoinehoff

committed

93e5a40

Sync advection speeds so that parallel runs come out correctly. Add 2x2v passive advection reg test. Tried using TS BCs in passive advection but I think there's a memory bug, need to switch to clus…

manauref

authored and

Antoinehoff

committed

21b5951

I think the changes to rect_grid_find_cell in a recent PR might've broken its use in bc_twistshift. Also, I think there was a bug in find_cell all along, though I don't know if it was ever triggere…

manauref

authored and

Antoinehoff

committed

a730b0d

Add TS BC unit test corresponding to figure 14 of the paper. Seems to show that the shift at the lower boundary is OK, but the shift at the upper boundary (which -1 times the lower shift) produces …

manauref

authored and

Antoinehoff

committed

59812b2

Protect against integrating over zero-volume regions in some sub-cell integrals of bc_twistshift, it was causing NaNs in the fig14 scenario. Fix boundary kernel selection in gk_collisionless_passiv…

manauref

authored and

Antoinehoff

committed

fe7b669

Add passive advection CBC test. Works fine with constant q, but the quadratic q has issues near the upper radial boundary. We'll examine this in a unit test setting, and try it with TCV geo.

manauref

authored and

Antoinehoff

committed

00e5450

Add a TS BC unit test with the same setup as the CBC reg test. We may delete or simplify this in the future, but it already reproduces the checkerboard pattern seen in the simulation.

manauref

authored and

Antoinehoff

committed

cc92dbc

Add an assert in twistshift BC updater that errors out of the change in the shift across a single cell is equal or greater to the length of the domain along y. I tried to add support for this sort …

manauref

authored and

Antoinehoff

committed

057b35f

Change the way we look for shifted cells in periodic copies along y in TS BC updater. The previous approach looked in periodic copies of the domain independently for each point, but this could give…

manauref

authored and

Antoinehoff

committed

1854d08

Fix y extents in CBC input files so the wedge number is indeed an integer. This explains a mysterious shift we were seeing when structures crossed the IMP.

manauref

authored and

Antoinehoff

committed

34859d5

Fix kernel selection in device code, and add some missing casts.

manauref

authored and

Antoinehoff

committed

d262b5e

Add device CBC unit test. Add bc_twistshift_new method, instead of inew, so we can call it from python.

manauref

authored and

Antoinehoff

committed

b5afcd9

Change signature of bc_twistshift_new so that structs are passed by address, this makes it easier to call it from python. Rename num_cell_ variables in tcv reg tests so that they use Nx like all ot…

manauref

authored and

Antoinehoff

committed

718cd63

Detected through compute sanitizer that delta_ts_x_dev was never released in its local scope.

Antoinehoff

committed

f9ac3ff

Increase resolution in d3d_iwl and tcv_iwl tests so they run (they can't run at too low a res because of constraints on TS BCs. Increase tolerance in unit test to 1e-8 so it passes on Perlmutter.

manauref

authored and

Antoinehoff

committed

881ec87

Increase another tolerance to 1e-9 in TS BC unit test so it passes on Perlmutter CPU.

manauref

authored and

Antoinehoff

committed

d7d773c

Fix mem leak in GK app where an array was not released at program's end.

manauref

authored and

Antoinehoff

committed

12e1f57

Update perlmutter CPU machine files to remove all possibility of GPU compilers/libraries being used/linked.

manauref

authored and

Antoinehoff

committed

948ce2b

Reduce Nmu in TCV iwl test to reduce runtime.

manauref

authored and

Antoinehoff

committed

3933c44

Navigation Menu

Uh oh!

Reapply GK passive advection work (PR #1084) - #1102

Reapply GK passive advection work (PR #1084)#1102
Antoinehoff merged 20 commits into
mainfrom
restore_gk_passive_1084

Commits on Sep 1, 2026

Add a new GK equation object to do passive advection, with a velocity given in the input file. Add sample regression test (gkylcas 16cfe76398a327fa8617ed162509585f0c6c6f47).

Change field_id in rt_gk_passive so that the omega_H dt doesn't show up. We should make it so that omega_H is only computed if the field evolves. Passive eqn object is clearly not ready as the resu…

Sync advection speeds so that parallel runs come out correctly. Add 2x2v passive advection reg test. Tried using TS BCs in passive advection but I think there's a memory bug, need to switch to clus…

I think the changes to rect_grid_find_cell in a recent PR might've broken its use in bc_twistshift. Also, I think there was a bug in find_cell all along, though I don't know if it was ever triggere…

Add TS BC unit test corresponding to figure 14 of the paper. Seems to show that the shift at the lower boundary is OK, but the shift at the upper boundary (which -1 times the lower shift) produces …

Protect against integrating over zero-volume regions in some sub-cell integrals of bc_twistshift, it was causing NaNs in the fig14 scenario. Fix boundary kernel selection in gk_collisionless_passiv…

Add passive advection CBC test. Works fine with constant q, but the quadratic q has issues near the upper radial boundary. We'll examine this in a unit test setting, and try it with TCV geo.

Add a TS BC unit test with the same setup as the CBC reg test. We may delete or simplify this in the future, but it already reproduces the checkerboard pattern seen in the simulation.

Add an assert in twistshift BC updater that errors out of the change in the shift across a single cell is equal or greater to the length of the domain along y. I tried to add support for this sort …

Change the way we look for shifted cells in periodic copies along y in TS BC updater. The previous approach looked in periodic copies of the domain independently for each point, but this could give…

Fix y extents in CBC input files so the wedge number is indeed an integer. This explains a mysterious shift we were seeing when structures crossed the IMP.

Fix kernel selection in device code, and add some missing casts.

Add device CBC unit test. Add bc_twistshift_new method, instead of inew, so we can call it from python.

Change signature of bc_twistshift_new so that structs are passed by address, this makes it easier to call it from python. Rename num_cell_ variables in tcv reg tests so that they use Nx like all ot…

Detected through compute sanitizer that delta_ts_x_dev was never released in its local scope.

Increase resolution in d3d_iwl and tcv_iwl tests so they run (they can't run at too low a res because of constraints on TS BCs. Increase tolerance in unit test to 1e-8 so it passes on Perlmutter.

Increase another tolerance to 1e-9 in TS BC unit test so it passes on Perlmutter CPU.

Fix mem leak in GK app where an array was not released at program's end.

Update perlmutter CPU machine files to remove all possibility of GPU compilers/libraries being used/linked.

Reduce Nmu in TCV iwl test to reduce runtime.