This simulator installs N+1 packaged cooling modules to serve N units of cooling demand. You pick N, fail modules one by one and see whether the remaining capacity still meets the demand, alongside the probability that enough modules survive the mission.
• A 3D bank of packaged cooling modules, condenser fans, a supply and return manifold and a cooling-demand exchanger. • Sliders for mission duration (100-3000 h), component failure rate (0.05-2 per 1000 h), required units N (2-4) and failed modules (0-3). • Five live readouts: installed modules, healthy modules, capacity margin in units, mission capacity probability and whether the demand is met. • Two presets: One failed module (three of four units remain, so demand is met) and Two failed modules (two remaining units cannot meet three units of demand).
With n = N + 1 installed units and unit survival p = exp(-lambda T), the system meets demand if at least N of the n units are working, R = P(X >= N) = p^n + n (1 - p) p^(n-1). The capacity margin is the healthy units minus N, so one spare gives a margin of 1 when nothing has failed, 0 after one failure and a shortfall after two.
The model is an independent active k-out-of-n mission calculation, not cold standby, and all installed units are assumed to be running. It has no switching delay, no load-dependent failure rates, no partial derating and no common failures, and the modules share demand ideally. N+1 covers a single failure; it does not cover a second failure at full demand.
Not at full demand. Only one unit of spare capacity is installed, so after two failures the healthy units fall below N and the demand is not met.
No. In this model all N+1 units are installed and active, and the probability is the independent active k-out-of-n calculation.
R = P(at least N of N+1 units survive) = p^n + n (1 - p) p^(n-1), where p = exp(-lambda T) and n = N + 1.
It leaves out switching delay, load-dependent failure rates, partial derating, repair and common failures, and assumes ideal demand sharing.