Crewed spaceflight architecture has been studied in detail for low-Earth-orbit and cislunar operations (Mission Class A, days to months), for multi-year missions inside the inner Solar System with return (Mission Class B, exemplified by NASA Design Reference Architecture 5.0 for Mars), and for multi-generational interstellar concepts (Mission Class D, the generation-ship literature). The intermediate regime — decadal, one-way, deep-space missions of 15–35 years duration at 30–50 AU heliocentric distance, conducted without resupply and without reproduction at the target — is treated only fragmentarily. We name this regime Mission Class C and argue that it constitutes a distinct architectural category requiring its own crew-composition logic. We propose the Stratified Crew Model v2 (SCM v2): a crew of N = 22 (defensible corridor 18–24) stratified along two orthogonal axes — career stage (Junior 25–30 / Mid-career 35–45 / Senior 50–60, target quota 25/50/25 per cent) and functional profile (depth-specialist / resilience-generalist, target balance 50/50) — composed under a T-shaped competency model. The stratification is derived not from energy or mass economics but from three compounding failure modes over thirty years: growing praxis asymmetry between the crew and Earth-side research under narrow-bandwidth contact (the crew can read Earth science but cannot import the apparatus needed to extend it), compounded by cumulative galactic-cosmic-ray-induced cognitive substrate degradation; habitat maintenance without resupply under common-mode aging; and psychosocial stability of small collectives over decades. The model is extended with a selection pipeline favoring best-available individuals over best initial team fit, a three-layer constitutional welfare-and-governance architecture with seven non-amendable core protections, and a logistics architecture pairing mandatory in-situ manufacturing with optional adaptive lunar-mass-driver cargo flow. The maximum empirical isolated-confined-extreme (ICE) analogue dataset (520 days, Mars-500) requires a roughly 20-fold extrapolation to Class C duration; we identify the corresponding research gaps as a proposed precursor research program for the 2030s and 2040s, rather than as a limitation of the present work.
Riswan Hassen (Mon,) studied this question.