Same physical picture as the 3D simulator: ARED replaces free weights (useless in microgravity) with vacuum cylinders. As the piston extends by x, the sealed volume grows and — by Boyle's Law at fixed temperature — the trapped pressure drops, so the resisting force actually rises through the stroke:
P(x) = P_v0 · V0 / (V0 + A·x)
F(x) = n · (P_atm − P(x)) · A
Bug found in the 3D engine, fixed only here: the 3D sim's own piston area (A = 0.0045 m², ⌀ ≈ 7.6 cm) can never produce more than ≈890 N (200 lbf) even at max vacuum and both cylinders — yet it normalizes its bone-load fraction against a 2,700 N (600 lbf) constant taken directly from the real ARED spec, so that fraction is capped at ≈0.33 no matter how hard you crank the vacuum slider. The bone-density trend can therefore never show the near-full loading a real astronaut running maxed-out ARED settings actually gets. This 2D companion corrects only the piston area to A = 0.01365 m² (⌀ ≈ 13.2 cm, in line with ARED's real documented cylinder bore) so the force curve can actually reach the 2,700 N ceiling it's compared against — every other equation, slider and the dose-response model itself are unchanged from the 3D sim.
Verified numerically (standalone Node/Python scratch script, never touching the 3D source): because V0/A is a fixed geometric ratio (half the cylinder length) independent of the piston area, P(x) at any given vacuum level and stroke position is exactly the same in both engines — only F = ΔP·A scales with A. Solving F(STROKE, vac=95%, n=2) = 2700 N for A gives A = 2700 / (2·98,925) ≈ 0.013647 m², confirmed against a direct re-evaluation of the Boyle's-law integral used for the mechanical-work readout.
- Evacuation level — how much air is pumped out of the cylinder at rest; higher vacuum means a bigger pressure differential and a stronger, more residual-gas-molecule-sparse chamber (fewer blue dots).
- 1 / 2 cylinders — engaging both roughly doubles resisting force for the same vacuum level, same as the real ARED's dual-canister design.
- Session duration / frequency — set the weekly training dose fed into the muscle and bone dose-response curves.
The dose-response model is identical to the 3D sim's: muscle responds mainly to time-under-tension (improves with dose alone), while bone remodeling needs mechanical load too — so with the corrected area, a low vacuum setting still leaves bone loss closer to the unexercised baseline even at full duration, and only a near-maximal vacuum with both cylinders now genuinely approaches the real 600 lbf ceiling the model was always meant to be judged against.