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

Arthur Conlas Dela Peña

doi:10.2478/tar-2026-0006

.blurhash-client-img { display: none !important; }

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

Transactions on Aerospace Research

Volume 2026 (2026): Issue 1 (March 2026)

By: Arthur Conlas Dela Peña

Open Access

|Mar 2026

Figures & Tables

Certification-style conceptual framework for blended wing body (BWB) evacuation compliance under uncertainty.

Conceptual BWB cabin configuration used for evacuation simulations (baseline occupancy ≈ 225 passengers; illustrative, not manufacturer-specific).

Global sensitivity of evacuation robustness metrics under uncertainty (total-effect indices). (a) Total-effect sensitivity for compliance probability PoC; (b) total-effect sensitivity for T95; (c) total-effect sensitivity for T99.

BWB cabin layout with geometric exit identification for the deterministic baseline (N = 225; Exits: FWD-L, FWD-R, AFT-L, AFT-C, AFT-R).

Density heatmap snapshots over time for the deterministic baseline evacuation scenario (t = 10 s, 30 s, and 45 s; N = 225; k = 5).

Exit utilization over time for the deterministic baseline evacuation scenario (cumulative evacuated per exit; N = 225; k = 5).

Probability-of-compliance and evacuation-time distributions under uncertainty: (a) PoC by scenario with 95% CI, (b) CDF of evacuation time by scenario with 90-s threshold, and (c) boxplots of evacuation time by scenario with 90-s threshold (R = 5000).

Stress scenario ranking and worst-case exit-block patterns for the probability-of-compliance analysis (R = 5000; threshold = 90 s).

Tornado-style Morris sensitivity plot for PoC and T95.

Pareto trade-off between compliance probability and structural penalty proxy, with representative recommended layout families (R = 5000; PoC = P(T ≤ 90 s)).

Robustness scorecard: PoC by scenario with 95% CI and target PoC threshold (R = 5000; threshold = 90 s; target PoC = 0.95).

Scenario outcomes summary and ranking by compliance risk (PoC, median, T99; R = 5000; threshold = 90 s)_

Risk rank	scenario	label	R	T limit sec	PoC hat	PoC CI95 low	PoC CI95 high	T Median sec	T95 sec	T99 sec
1	S7	Compound stress	5000	90	0.000	0.000	0.000768	142.127	182.972	201.981
2	S1	Random 1- exit blocked	5000	90	0.000	0.000	0.000768	121.599	149.168	163.554
3	S2	AFT-C blocked	5000	90	0.000	0.000	0.000768	117.821	140.425	151.509
4	S3	Reduced visibility	5000	90	0.007	0.005	0.010640	97.272	102.578	104.439
5	S4	Center-heavy loading	5000	90	0.039	0.034	0.045369	112.033	142.963	157.591
6	S6	Reduced crew	5000	90	0.088	0.081	0.096796	94.549	100.186	102.269
7	S5	Edge-heavy loading	5000	90	0.134	0.125	0.144341	98.617	117.778	128.536
8	S0	Baseline	5000	90	1.000	0.999	1.000000	72.254	74.4360	74.9619

Scenario matrix for stress-testing (repeatable condition sets)_

Scenario ID	Scenario label	Exit availability	Visibility	Passenger spatial distribution	Crew effectiveness	Purpose/interpretation
S0	Baseline certification scenario	All required exits available (baseline set)	Normal (baseline)	Nominal (regulatory/standard seating distribution)	Nominal guidance/compliance	Reference case for PoC and tails; anchors comparisons [49]
S1	Random 1- exit blocked	Exactly one exit is unavailable per run; blocked exit is sampled uniformly from the baseline exit set; seeded & repeatable	Normal	Nominal	Nominal	Captures expected degradation from single-exit unavailability without assuming a specific failure location
S2	AFT-C blocked (targeted worst-case	AFT-C unavailable; remaining baseline exits available	Normal	Normal	Nominal	Targeted topology stressor aligned to worst-case singleexit loss results
S3	Reduced visibility	Baseline exits	Low-visibility tier (penalty active)	Nominal	Nominal	Test mobility/decision degradation under smoke/visibility impairment [41]
S4	Centerheavy loading	Baseline exits	Normal	Centerheavy (higher initial density in central cabin zones)	Nominal	Probes the wide-cabin redistribution burden and the sensitivity to merge formation.
S5	Edge-heavy loading	Baseline exits	Normal	Edge-heavy (higher initial density near edges/outer aisles)	Nominal	Contrasts with S4 to detect geometry-driven advantages/disadva ntages.
S6	Reduced crew effectiveness	Baseline exits	Normal	Nominal	Reduced (lower assertiveness /compliance probability)	Tests sensitivity to guidance quality and compliance; aligns with evidence that guidance materially shifts outcomes [2,5]
S7	Compound stress	AFT-C blocked + low-visibility tier + centerheavy loading + reduced crew effectiveness	Low-visibility tier	Center-heavy	Reduced	Harsh-but-credible compound case for tail amplification and failure modes

Robust design candidate outcomes (PoC, median, T95, T99; R = 5000; threshold = 90 s) with a structural penalty proxy_

n	Layout family	Key features	Structural penalty proxy	PoC (95% CI)	T95 (s)	T99 (s)
D1	Baseline 5-exit	k=5	5.0	0.247 [0.235, 0.259]	111.25	119.37
D2	5-exit + zoning	k=5, zoned exits	5.0	0.067 [0.061, 0.075]	131.63	144.94
D3	5-exit + cross-aisle	k=5, cross-aisle	5.5	0.541 [0.528, 0.555]	101.59	108.02
D4	5-exit + wide aisle	k=5, wide aisles	5.3	0.520 [0.507, 0.534]	105.46	113.97
D5	4-exit (reduced)	k=4	4.0	0.000 [0.000, 0.001]	136.05	146.34
D6	4-exit + cross-aisle + wide	k=4, cross-aisle, wide aisles	4.8	0.010 [0.008, 0.013]	114.65	121.27
D7	6-exit (added)	k=6	6.0	0.879 [0.869, 0.887]	94.60	102.34
D8	3-exit (minimal)	k=3	3.0	0.000 [0.000, 0.001]	175.05	187.43
D9	5-exit + cross-aisle + wide	k=5, cross-aisle, wide aisles	5.8	0.834 [0.823, 0.844]	95.34	100.58

Probability of compliance and tail evacuation times by scenario (R = 5000; threshold = 90 s)_

Scenario	Label	Run	T limit sec	PoC hat	PoC CI95 low	PoC CI95 high	T median sec	T95 sec	T99 sec
S0	Baseline	5000	90	1.0000	0.999232	1.00000	80.67222	85.51355	86.77391
S1	Random blocked exits	5000	90	0.0000	0.000000	0.000768	132.6695	167.1254	184.6882
S2	Worst-case blocked	5000	90	0.0000	0.000000	0.000768	132.4394	166.9959	188.9797
S3	Reduced visibility	5000	90	0.0070	0.000000	0.000768	106.9491	114.9699	118.4541
S4	Center-heavy loading	5000	90	0.0134	0.010566	0.016981	113.5104	149.3857	168.2943
S5	Edge-heavy loading	5000	90	0.0178	0.014488	0.021853	111.9598	142.6022	158.8951
S6	Reduced crew	5000	90	0.0006	0.000204	0.001763	104.1563	112.7216	116.3755
S7	Compound stress	5000	90	0.0000	0.000000	0.000768	159.7831	209.5973	240.6371

Per-exit evacuation breakdown for the deterministic baseline (S0; N = 225, k = 5)_

Exit ID	Final evacuated n_j	Share e
FWD-L	45	0.2
FWD-R	45	0.2
AFT-L	45	0.2
AFT-C	45	0.2
AFT-R	45	0.2

Robustness scorecard by scenario (PoC, CI, and Tail Risk) with required modifications and residual assumptions (R = 5000; threshold = 90 s; target PoC = 0_95)_

Scen ario	Stress condition/variant	PoC (95% CI)	Median T (s)	T95 (s)	T99 (s)	Meets PoC target?	Required modification to reach PoC = 0.95	Residual risk (under assumptions)
S7	Compound stress	0.000 [0.000, 0.001]	142.13	182.97	201.98	No	Combine: add redundancy + widen aisles + reduce premovement; mitigate compounding stressors	Non-compliance is structural under current topology/capacity assumptions
S1	Random 1-exit blocked	0.000 [0.000, 0.001]	121.6	149.17	163.55	No	Add exit redundancy and reroute flow (crossaisle); avoid single-exit dependency	Non-compliance is structural under current topology/capacity assumptions
S2	AFT-C blocked	0.000 [0.000, 0.001]	117.82	140.43	151.51	No	Add exit redundancy and reroute flow (crossaisle); avoid single-exit dependency	Non-compliance is structural under current topology/capacity assumptions
S3	Reduced visibility	0.008 [0.006, 0.011]	97.27	102.58	104.44	No	Reduce premovement + improve wayfinding (lighting/markings/crew guidance)	Non-compliance driven by tail events (delays + bottlenecking) under adverse draws
S4	Centerheavy loading	0.040 [0.035, 0.045]	112.03	142.96	157.59	No	Rebalance zoning/exit-choice to equalize queues; de-bias AFT-C loading	Non-compliance driven by tail events (delays + bottlenecking) under adverse draws
S6	Reduced crew	0.089 [0.081, 0.097]	94.55	100.19	102.27	No	Capacity-preserving geometry (wider aisles/door area) and cross-aisle redistribution	Non-compliance driven by tail events (delays + bottlenecking) under adverse draws
S5	Edge-heavy loading	0.135 [0.125, 0.144]	98.62	117.78	128.54	No	Rebalance zoning/exit-choice to equalize queues; de-bias AFT-C loading	Non-compliance driven by tail events (delays + bottlenecking) under adverse draws
S0	Baseline	1.000 [0.999, 1.000]	72.25	74.44	74.96	Yes	No modification required under modeled assumptions	Residual risk dominated by rare routing imbalance and parameter tail draws

Morris screening sensitivity indices (μ*, σ) for PoC and T95_

factor	range	μ Star PoC	σ PoC	μ Star T95	σ T95	Rank PoC	Rank T95	Rank sum
μ scale	[0.75, 1.05]	0.035000	0.071644	44.16160	13.07616	1	1	2
μ AFT-C	[1.00, 1.30]	0.007083	0.017803	23.32390	20.92023	4	2	6
t₀ mode	[8.00, 20.0]	0.013333	0.029169	12.92360	11.49410	2	6	8
κ	[10.0, 60.0]	0.010833	0.030178	16.18077	20.21300	3	5	8
bAFT-C	[1.00, 1.80]	0.005833	0.017327	21.02509	24.23203	5	3	8
p block	[0.00, 0.50]	0.002083	0.006765	19.15942	13.02821	6	4	10

Deterministic baseline performance summary (S0: N = 225,k = 5)_

Metric	Symbol	Baseline value
Total occupants	N	225
Available exits	k	5
Evacuated through each exit (final)	n_j	45 for each of the 5 exits
Exit shares (final)	Si=njn{S_i} = {{{n_j}} \over n}	0.20 for each exit
Most-loaded exit share	max_{S_j}	0.2
Exit balance index (CV across exits)	Cv_exit	0.00 (perfectly balanced under equal split)
Total evacuation time (last occupant exits)	T_det	t0+45μ seconds{t_0} + {{45} \over \mu }{\rm{ seconds}}
Mean overall discharge rate	q¯=NTdet\bar q = {N \over {{T_{det}}}}	225t0+45μ person /s( if t0=0=5μ){{225} \over {{t_0} + {{45} \over \mu }}}{\rm{ person }}/s\left( {{\rm{ if }}{t_0} = 0 = 5\mu } \right)

Worst-case exit-block patterns (fixed single-exit block; PoC, median, T95, T99; ranked)_

Block risk rank	Blocked exit	R	T limit sec	PoC CI95 high	T Median sec	T95 sec	T99 sec
1	AFT-C	5000	90	0.000768	117.920	139.774	152.698
2	FWD-R	5000	90	0.000768	114.155	136.921	148.523
3	FWD-L	5000	90	0.000768	114.477	136.740	148.426
4	AFT-L	5000	90	0.000768	114.412	136.117	147.098
5	AFT-R	5000	90	0.000768	114.212	135.732	146.786

Uncertain parameters used for Monte Carlo evacuation simulations_

Parameter name	Physical meaning	Distribution type	Range/mean / SD	Justification/source category
U₀	Free walking speed (uncongested)	Lognormal	Median 1.20 m/s; GSD 1.20 (approx. SD ≈ 0.20 m/s)	Literature-calibrated (evacuation walking variability; distribution fitting under reduced visibility) [44]
k_P	Congestion/friction sensitivity in speed≈density relation	Uniform	(kP in [0.8, 1.2]) (multiplier on baseline congestion curve)	Conservative bound capturing uncertain crowd-friction effects in wide geometry
τ	Pre-movement / reaction time before entering flow	Lognormal (scenario-shifted)	Baseline: Median 8 s; 95th ≈ 35 s. Stress: Median 12 s; 95th ≈ 60 s	Literature-calibrated; stress shift represents conservative inflation [43]
P_crew	Compliance with crew guidance (follow commands, directed exits)	Beta (scenario-shifted)	Baseline: Beta(9,3) mean 0.75. Stress: Beta(6,4) mean 0.60	Literature-supported sensitivity to guidance; heterogeneity + conservative stress reduction [45]
b_exit	Exit choice bias (nearest/known exit vs distributed choice)	Truncated Normal	Mean 0; SD 0.5; truncated to ([-1, +1])	Engineering assumption; stress can shift mean toward “nearest-exit” behavior [9]
k_vis	Visibility penalty factor applied to speed / movement potential	Uniform (scenario-shifted)	Baseline: ([0.85, 1.0]); Stress: ([0.50, 0.85])	Literature directionality (visibility reduces speed) + conservative bounds for cabin analogs [44]
x_exit state (replaces x_block)	Exit availability state (scenario–defined): which exit(s) are unavailable	Scenario-defined categorical (deterministic by scenario; with optional within-scenario randomizatio n)	S0: none blocked. S1: exactly one exit blocked per run; blocked exit sampled uniformly from baseline exit set (seeded). S2: fixed targeted block (e.g., AFT-C blocked). S7: same targeted block as S2 + additional stressors	Aligns regulatory “unavailable exit” interpretation to reported scenarios; prevents mismatch between Methods and Results (scenario matrix)
x_aisle	Partial aisle obstruction (e.g., debris, spillback)	Bernoulli (scenario-shifted)	Baseline (p = 0.05);Stress (p = 0.20)	Conservative bound (rare baseline; elevated under stress)
C_e	Exit flow capacity/service rate (persons/s) for available exits	Truncated Normal (>0)	Mean 1.4; SD 0.3; truncated to ([0.6, 2.2])	Engineering assumption with conservative bounds; explicitly models throughput variability while avoiding unrealistic discharge
k_c	Capacity degradation multiplier under stress (applied to available exits)	Uniform	([0.6, 0.9])	Conservative bound representing door/slide interface inefficiency; note: blocked exits are modeled as zero capacity (not multiplied)
P_irr	Probability of disruptive passenger behavior affecting flow (hesitation, counterflow, nonideal actions)	Beta (scenario-shifted)	Baseline Beta(2,18) mean 0.10; Stress Beta(5,15) mean 0.25	Literature-informed direction + conservative stress inflation [9]

References

Authors

Metrics

Articles in this issue

DOI: https://doi.org/10.2478/tar-2026-0006 | Journal eISSN: 2545-2835

Journal RSS Feed

Language: English

Page range: 121 - 157

Submitted on: Feb 4, 2025

Accepted on: Mar 5, 2026

Published on: Mar 18, 2026

Published by: ŁUKASIEWICZ RESEARCH NETWORK – INSTITUTE OF AVIATION

In partnership with: Paradigm Publishing Services

Publication frequency: 4 issues per year

Keywords:

blended wing body,

emergency evacuation,

probability of compliance,

uncertainty quantification,

robust design

Related subjects:

Engineering,

Introductions and overviews,

Materials sciences, other,

Physics,

Physics, other

© 2026 Arthur Conlas Dela Peña, published by ŁUKASIEWICZ RESEARCH NETWORK – INSTITUTE OF AVIATION
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.

Volume 2026 (2026): Issue 1 (March 2026)

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

Figures & Tables

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

Fig. 11.

Scenario outcomes summary and ranking by compliance risk (PoC, median, T99; R = 5000; threshold = 90 s)_

Scenario matrix for stress-testing (repeatable condition sets)_

Robust design candidate outcomes (PoC, median, T95, T99; R = 5000; threshold = 90 s) with a structural penalty proxy_

Probability of compliance and tail evacuation times by scenario (R = 5000; threshold = 90 s)_

Per-exit evacuation breakdown for the deterministic baseline (S0; N = 225, k = 5)_

Robustness scorecard by scenario (PoC, CI, and Tail Risk) with required modifications and residual assumptions (R = 5000; threshold = 90 s; target PoC = 0_95)_

Morris screening sensitivity indices (μ*, σ) for PoC and T95_

Deterministic baseline performance summary (S0: N = 225,k = 5)_

Worst-case exit-block patterns (fixed single-exit block; PoC, median, T95, T99; ranked)_

Uncertain parameters used for Monte Carlo evacuation simulations_

Paradigm

My account