Commit 5b00648
committed
Revert HLL/LF adv gsrc change; update comment to explain PR #800 intent
The previous commit incorrectly set flux_gsrc(adv) = 0 for HLL and LF,
which would have reverted PR #800. Per the analysis in issue #1402 comment
by ChrisZYJ: flux_rs(adv) in HLL/LF is the numerical diffusion component
(not transport), and any face-flux in cylindrical coordinates requires the
full geometric divergence (1/r)*d(rF)/dr = dF/dr + F/r. The flux_gsrc term
provides the F/r correction for the diffusion. The transport half is handled
separately via flux_src_rs (non-conservative source). PR #800 was correct.
The comment is updated to document this so future readers do not repeat the
mistake. The convergence test is unaffected: for num_fluids=1, alpha=1
everywhere so flux_rs(adv) = 0 at all faces regardless.1 parent 743016a commit 5b00648
1 file changed
Lines changed: 4 additions & 4 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
772 | 772 | | |
773 | 773 | | |
774 | 774 | | |
775 | | - | |
| 775 | + | |
776 | 776 | | |
777 | 777 | | |
778 | | - | |
| 778 | + | |
779 | 779 | | |
780 | 780 | | |
781 | 781 | | |
| |||
1382 | 1382 | | |
1383 | 1383 | | |
1384 | 1384 | | |
1385 | | - | |
| 1385 | + | |
1386 | 1386 | | |
1387 | 1387 | | |
1388 | | - | |
| 1388 | + | |
1389 | 1389 | | |
1390 | 1390 | | |
1391 | 1391 | | |
| |||
0 commit comments