Uncertainty-Aware Robustness Analysis of Blended-Wing-Body Cabin Evacuation Under the Faa 90-Second Requirement (14 CFR § 25.803)

Blended-wing-body (BWB) aircraft are currently being explored for potential efficiency gains, but emergency evacuation remains a key certification feasibility requirement for wide, non-cylindrical cabins under the FAA 90 s rule (14 CFR §25.803). This study reframes evacuation compliance as a probabilistic safety assessment problem and develops an uncertainty-aware pipeline for certification-style robustness evidence. Evacuation is posed as a limit-state problem, g ( θ ) = 90 - T ( θ ), where T is evacuation time under uncertainty in behavioral and flow parameters. Monte Carlo simulation propagates bounded aleatory variability and evidence-tagged epistemic assumptions across a predefined stressor matrix (S0–S7) to estimate the distribution of T . Compliance is reported as PoC = P ( T ≤ 90 s) with 95% confidence intervals and tail-risk metrics ( T 95 , T 99 ); Morris screening identifies dominant contributors to failure probability P f = 1 - PoC and extreme delays. Baseline conditions meet the 90 s requirement, whereas topology/capacity disruptions – especially exit loss and compound stress – drive PoC toward zero and inflate upper-tail percentiles, indicating discharge-capacity limits. Reduced visibility and degraded crew effectiveness primarily thicken the right tail, yielding partial compliance but elevated P f at 90 s. Robust layout screening (D1–D9) shows that exit-demand redistribution and capacity-preserving geometry recover safety margin with limited structural penalty growth‥