Feature/full analytical plume model

tests/plume/data/digitized/README.md

@@ -0,0 +1,29 @@
+# Digitized reference data for the plume verification figures
+
+Drop digitized curves from Cai & Wang 2012 (JSR 49(1), DOI
+10.2514/1.A32046) here and the manual-run figure scripts
+(`tests/plume/plume_verification_test_04` ... `_27`) will overlay them
+automatically on their next run. No code changes are needed.
+
+## File convention
+
+- Name: `<figstem>_<series>.csv`, where `<figstem>` is the paper figure
+  (e.g. `fig19`) and `<series>` names the dataset. Examples:
+  - `fig03_dsmc.csv` — DSMC centerline density (S0 = 2), Fig. 3
+  - `fig19_dsmc.csv` — DSMC centerline density, Kn = 100, Fig. 19
+  - `fig25_dsmc_kn100.csv`, `fig25_dsmc_kn0p1.csv`,
+    `fig25_dsmc_kn0p01.csv` — the three DSMC mass-flux curves, Fig. 25
+  - contour figures: one CSV per digitized contour polyline, e.g.
+    `fig06_dsmc_0p001.csv` (the DSMC n/n0 = 0.001 line of Fig. 6)
+- Content: two comma-separated columns `x,y` with one header row.
+  x/y are in the figure's plotted units (X/D, Z/D, theta in degrees,
+  normalized quantity values).
+- Series suffixes are prettified for legends: `dsmc` -> `DSMC`,
+  `kn0p1` -> `Kn=0.1`.
+- Lower-half-plane series in the split contour figures (Figs. 6-8,
+  11-18) should be digitized with positive Z/D values taken from the
+  paper's lower half; enter them as printed (negative y) — overlays are
+  drawn exactly as given.
+
+Each figure script's docstring lists the exact filenames it looks for.
+Files that are absent are silently skipped.

tests/plume/output/.gitignore

@@ -0,0 +1,2 @@
+*
+!.gitignore