Shepherded Gyroscopic Mass Streams: A Modular Dynamic-Support Architecture for Cislunar Orbital Infrastructure

Dynamic-support structures use high-velocity mass streams to generate forces far exceeding the static strength of their materials, potentially enabling orbital infrastructure from conventional steels and composites. The Launch Loop (Lofstrom, 1985) is the most developed such concept, but its continuous segmented rotor requires microsecond centralized control and presents a cascading-failure mode capable of releasing terajoule-scale energy. This paper proposes an alternative architecture: the Shepherded Gyroscopic Mass Stream (SGMS).In SGMS, discrete 2 kg steel-composite balls, each spinning at up to 50,000 rpm and carrying integrated Halbach permanent-magnet arrays, travel at 10–15 km/s along trajectories maintained by co-moving and stationary shepherd stations in cislunar or interplanetary space. Because the stream is discretized, failures are inherently localized: a single-ball loss releases energy only at the unit scale, with no direct propagation mechanism between neighboring masses. Gyroscopic angular momentum provides passive attitude stability on timescales of months to years, allowing shepherd corrections to replace the sub-microsecond global control demands associated with continuous-rotor architectures, while local station interactions remain millisecond-scale. The SGMS framework also permits incremental deployment. A 100-ball proof-of-concept segment could be established within months of initial lunar electromagnetic launch, using lunar-sourced mass, gravitational slingshot velocity gain, and shepherd-mediated trajectory correction. More broadly, the architecture shifts dynamic support from a monolithic continuously coupled rotor to a packetized, failure-tolerant stream that is more compatible with staged deployment and bounded-risk testing. Within the regime analyzed here, no immediate first-principles physical inconsistency has been identified. However, several key subsystem claims—especially payload coupling at operational velocity and shepherd deflection geometry—remain unvalidated and require experimental demonstration.