# Extended 2D Dataset — NeurIPS Closure Challenge additions Three new 2D cases added to the closure challenge from the **NASA TMR Collaborative Testing Challenge 2022** ([turb-prs2022](https://tmbwg.github.io/turbmodels/turb-prs2022.html)): | Folder | Case | Required metrics | |--------|------|------------------| | `NASA_2DZP/` | 2D Zero-Pressure-Gradient flat plate | Cf(x); u⁺(log y⁺) at x=0.97 — vs theory | | `NASA_ASJ/` | Axisymmetric Subsonic Jet (M_jet ≈ 0.5) | u/Uj along axis; u/Uj and u'v'/Uj² at 5 stations — vs Bridges & Wernet PIV | | `NASA_2DN00/` | NACA 0012 airfoil, α = 10/15/17/18° | CL(α), CD vs CL, Cp(x/c), Cf(x/c upper) — vs Ladson + Gregory | ## Per-case layout ``` NASA_/ ├── README.md Case description, refs, submission format ├── baseline_komegasst/ k-ω SST RANS baseline (.dat NASA Tecplot + .csv flat) │ └── README.md ├── highfidelity/ Theory / experiment / CFD-cross-check references │ └── README.md └── plots/ ├── plot_profiles.py Adapt: set MODEL_DIR env var to overlay your model └── *.pdf Pre-rendered baseline-vs-experiment plots ``` ## Tools at this level - `dat_to_csv.py` — converts NASA Tecplot zone-format `.dat` files to flat CSV (single `zone` column to preserve zone identity). Re-run after editing data: ``` python dat_to_csv.py NASA_2DZP/baseline_komegasst NASA_2DZP/highfidelity \ NASA_ASJ/baseline_komegasst NASA_ASJ/highfidelity \ NASA_2DN00/baseline_komegasst NASA_2DN00/highfidelity ``` ## Provenance All `.dat` files copied from `OLD_NASA_CHALLENGE_SUBMISSION/finalModel/` (a baseline RANS submission to the 2022 NASA TMR Collaborative Testing Challenge), folders `baselineResults/` (k-ω SST baseline) and `NASAResults/` (theory / experiment) — see [plotFinal.py:10-14](OLD_NASA_CHALLENGE_SUBMISSION/finalModel/plotFinal.py#L10-L14) for the original mapping. The corresponding OpenFOAM cases live under `OLD_SETUPS/{00_2DZP, 02_ASJ, 04_2DN00}/00Baseline/`. ## TODO — full OpenFOAM `00Baseline_Data/` per case The integral profiles + plots are complete. To match the structure of the existing 4 challenge cases (`ML4Fluids_2D_Dataset/{CBFS, NASA_2DWMH, PH_Breuer, Parm_PH_29}/00Baseline_Data/`), each new case still needs the converged k-ω SST OpenFOAM run packaged. Candidate sources: | Case | OpenFOAM source | Notes | |------|-----------------|-------| | 2DZP | `OLD_SETUPS/00_2DZP/00Baseline/simpleFoam/` | `kOmegaSST`, converged at t=10000 | | ASJ | `OLD_SETUPS/02_ASJ/BaselineRenzhi/Case/` or `00Baseline/CFDDomain/compressibleTransient/` | both `kOmegaSST`; pick one | | 2DN00 | `OLD_SETUPS/04_2DN00/00Baseline/rhoSimpleFoam_Final/Case2/{10,15,17,18}_deg/` | only initial state — needs to be run, OR use the per-AoA cases under `OLD_NASA_CHALLENGE_SUBMISSION/finalModel/04_2DN00/` (latter were run with `propagationClassifierkOmegaSST` augmented model, not pure baseline) |