This is a real numerical integration, not a scripted animation. Every physics step computes drag from the actual drag equation and lets it fight gravity:
F_drag = ½·ρ·v²·Cd·A (opposes motion)
a = g − F_drag / m
v += a·dt ; altitude −= v·dt
Body-only (belly-to-earth arch), a jumper's drag area Cd·A ≈ 0.45 m² gives a freefall terminal velocity v_t = √(2mg / ρ·Cd·A) ≈ 54 m/s (194 km/h) — the speed where drag exactly cancels weight and acceleration reaches zero. That's why the velocity curve below flattens out instead of climbing forever: it's an asymptote the integration approaches, never a capped value.
Pulling the chute doesn't teleport you to a slow speed — it ramps Cd·A up to ≈ 42 m² over about 1.2 s of canopy inflation. Because drag scales with v², that huge area increase against a still-high velocity produces a genuinely enormous deceleration (tens of g for a fraction of a second — the real "opening shock"), which the integrator resolves in small sub-steps so it stays numerically stable. The new terminal velocity under canopy is only ≈ 5.5 m/s, so if there's enough altitude left to coast down that new curve, the landing is soft; if the ground arrives before v gets close to it, the landing is still fast.
- Exit altitude — how much runway the flight has for freefall + full canopy deceleration. BASE objects range from ~60 m spans to 600 m+ cliffs.
- Body position — a stable arch (Cd·A≈0.45) is the safe default; a head-down track (Cd·A≈0.20) roughly doubles fall speed to gain horizontal separation from the object, at the cost of far less time to react.
- Pull chute — fires whenever you click it, at whatever altitude/time you choose. Pull low on a fast track and the ground can win the race against the deceleration curve — that's the real risk this simulator lets you feel.
Landing is graded on the actual integrated impact velocity: ≤3 m/s stand-up soft, ≤6 m/s a standard parachute-landing-fall is survivable, ≤9 m/s a hard landing with real injury risk, above that (or no canopy at all) is not survivable.