Vignesh Gopakumar
  • Home
  • Research
  • Talks
  • Blog
  • Tokamak Toolkit

The first EC reference run stops at its plotting device

negative
Published

October 8, 2026

negative result Entry 33 · milestone M3

We tried to produce native electron-cyclotron (EC) heating and current profiles from the original ITER reference case. These will eventually supply the transport sources described in the Primer; terminology is in the Glossary.

The owner approved continuation with the proposed original ITER numerical case and its one-run resource ceiling. The first run started and stopped before producing scientific output. Its plotting dependency rejected the requested portrait colour PostScript device. GENRAY’s error branch then returned zero, which would look successful if we checked only the process exit code.

The pilot runner also requires the expected output file. It correctly retained a failure, with unchanged inputs, immutable binary/input snapshots, stdout, stderr and partial files. No heating or driven-current profiles were produced.

We traced the failure to a missing device alias in the pinned replacement plotting library. We added a small explicit patch that maps the alias to the library’s existing portrait colour PostScript driver. GENRAY source and physics inputs remain unchanged. The correction builds separately, retaining the patch and executed build helper.

A plotting-only Fortran probe now checks the actual interface against both libraries: all eight plotting-device checks match their expected outcomesC-155. Successful output has the expected page orientation and colour; an unknown device still fails. This establishes the device correction, while a completed GENRAY run and full scientific/plotting comparison remain pending.

At this checkpoint, the original one-run approval is exhausted. A concrete retry request keeps the same allocation and conservative charging ceiling and awaits fresh approval. No retry, labels or training occurred at this stage. The subsequently authorized retry completes the numerical pilot. The current-source and ITER pedestal boundary contracts remain open, and historical full-suite test statistics are unchanged. M3 remains in progress.

Technical details → first EC execution and plotting correction. Decision → OQ-21.

Technical details → Surrogates › genray first run
Decisions → OQ-21

© Copyright 2026 Vignesh Gopakumar

 
 
Code and first draft by Claude (Anthropic) and Codex, working to a brief by Vignesh Gopakumar, who reviewed and approved this page. How this is built · Tokamak Toolkit home