Time-step check of the JOREK growth rates; the grid check did not run
Under the owner’s convergence rule (OQ-39, decision 7), a JOREK growth rate γ is quoted only if halving the time step Δt and doubling the radial and poloidal resolution each leave it within a fixed tolerance, with the mode staying where it was. This entry applies the rule to the case (a) modes of entry 68: the pedestal modes n = 12 and 16, and the axis modes n = 8 and 20, to test whether the latter are numerical.
What ran
The third and last job allowed under OQ-39 had three parts: - Δt = 0.5 on the 80 × 128 grid for n = 8, 12, 16, 20, restarting from the same axisymmetric state as entry 68, so that only Δt changes; - the doubled grid, 160 × 256: a new axisymmetric start, then n = 12 and 16 at Δt = 1; - n = 8 and 20 on the doubled grid, after the first part, until the time limit.
The comparison uses the same energy-band fit as entry 68 and also requires the same location class. It reads both runs’ saved energies, so it doesn’t depend on how the runs were launched. It is still our own fit applied to our own runs.
Result
| n | where the mode peaks | Δt = 1 → 0.5 | 160 × 256 | verdict |
|---|---|---|---|---|
| 12, 16 | pedestal | n = 12: 0.04397 → 0.04397 (−0.01%); n = 16: 0.03744 → 0.03633 (−2.94%); both within the 5% toleranceC-227 | not run | converged in Δt only |
| 8, 20 | magnetic axis | n = 8: 0.02810 → 0.02967 (+5.56%); n = 20: 0.03300 → 0.03473 (+5.27%); both outside the 5% toleranceC-228 | not run | fails the Δt check |
All four Δt = 0.5 runs have clean linear phases, keep their location, and end on the same solver abort after it. The pedestal modes pass the time-step check. The axis modes fail it, which is consistent with a numerical origin.
The doubled grid never got past its axisymmetric start (about 126 s per step against 9.7 s on 80 x 128 (13x); timed out after 43 of 50 steps, so no doubled-grid linear run startedC-229). Scaled from that cost, each linear run on the doubled grid would take longer than one job allows. So no growth rate is quotable under decision 7: n = 12 and 16 are converged in time step only.
The job also wasted compute (about 290 of 1046 core-h: after the last run ended the script waited on its watchdog (an empty wait list became a bare wait); fixed in both job scriptsC-232). After the last run ended, the batch script kept waiting on its watchdog process: an empty list of process IDs turned the wait command into a wait for every background process. We cancelled the job once it was idle, and both J5 job scripts are fixed.
Decision
The owner chose the recommended options of OQ-41: - Case (a) is closed on the time-step check. n = 12 and 16 are reported as “converged in Δt, grid not checked”, not quotable. The other six n are “not measured: core or axis mode faster”. - The full check, time step and grid, moves to case (b), the TORAX state. - Case (b) gets up to five one-node jobs of at most six hours, each announced before submission.
Where this stands: case (b) needs its own equilibrium first (entry 71).
Technical details → JOREK. Decision → OQ-41.
JOREK acknowledgement. JOREK is developed by the JOREK team and is open source under the GNU LGPL v3. Results here use JOREK at commit 01a52be3. Citations: M. Hoelzl et al., Nuclear Fusion 61, 065001 (2021); M. Hoelzl et al., Nuclear Fusion 66, 116006 (2026). For a list of JOREK team members, refer to the author list of M. Hoelzl, G.T.A. Huijsmans, F.J. Artola, E. Nardon, M. Becoulet, N. Schwarz, A. Cathey, S.J.P. Pamela et al., Nuclear Fusion 64, 112016 (2024).
Technical details → JOREK › convergence check
Decisions → OQ-39 · OQ-41