Decompression
4/4 ✓Decompression, decoded. Inert gas (nitrogen, helium) loads and unloads each tissue compartment exponentially — Haldane’s 1908 model (five compartments, a ~2:1 supersaturation ratio) generalised by Bühlmann’s ZHL-16 (sixteen compartments, a = 2/∛t½, b = 1.005 − 1/√t½, the tolerated ceiling P_amb,tol = (P − a)·b), with the dual-phase bubble models (VPM, RGBM) adding microbubble control. Henry’s law drives the loading; the oxygen window drives the elimination; decompression sickness is bubble formation on ascent, treated definitively by recompression / hyperbaric oxygen, which has ~15 evidence-based indications. Honest bound: DCS risk is probabilistic — the models lower it but never to zero — and this is physics, not a replacement for training, certified tables/computers, or a doctor.
- ✓inert-gas loading is exponential per tissue compartment — Haldane (1908): five compartments (half-times 5/10/20/40/75 min) with a ~2:1 supersaturation ratio; P(t) = P0 + (P_insp − P0)(1 − 2^(−t/t½)) (one half-time → halfway, here 1.58 bar)
- ✓Bühlmann ZHL-16 sets the ascent ceiling — 16 compartments (N₂ half-times ~4–635 min); a = 2/∛t½ (1.17 bar) and b = 1.005 − 1/√t½ (0.558) give the tolerated ambient pressure P_amb,tol = (P − a)·b; variants A/B/C; the basis of dive tables and computers. The dual-phase bubble models (VPM, RGBM) add free-phase microbubble control
- ✓the physics is real — Henry’s law (dissolved gas ∝ partial pressure), the oxygen window (inherent unsaturation driving off-gassing), DCS as bubble formation on ASCENT (not pressure at depth); hyperbaric oxygen therapy has ~14–15 UHMS-evidenced indications (DCS, gas embolism, CO poisoning, gas gangrene, non-healing wounds…)
- ✓HONEST BOUNDS — DCS risk is PROBABILISTIC (one dataset ~5.7%; the US Navy targets ~2% mild, ~0.1–0.2% serious): the models REDUCE risk, they never make it zero, and eliminating it entirely is the genuinely impossible task. This is physics knowledge, NOT a substitute for proper dive training, certified tables/computers, or medical supervision
граница: HONEST (the saved README-first wave, Wikipedia/PMC/UHMS/DAN-cited): DOCUMENTED — Haldane’s compartment model (1908), Bühlmann ZHL-16 (the a/b coefficients computed here in src/0), M-values, the VPM/RGBM dual-phase models, Henry’s law, the oxygen window, and HBOT’s ~14–15 UHMS-approved indications. The CARDINAL honest bound: decompression is PROBABILISTIC — DCS occurs in a small but nonzero fraction of divers on an identical profile (one analysis ~5.7%; the USN designs to ~2% mild / ~0.1–0.2% serious), so "no-decompression" and dive computers REDUCE risk, they do not guarantee safety; driving DCS risk to exactly zero, or off-gassing arbitrarily fast (the slowest compartment limits it), is genuinely impossible. This fold is physics knowledge and decision-support context — NOT dive instruction or medical advice; real dive planning uses certified training, tables and computers, and DCS is a medical emergency for professionals.