1. The question left open

Entry 878 treated a sail slew as a possible controlled burn. Entry 879 asks what actually has to happen inside the bus while the sail is turning: how much torque, how much angular momentum, and whether the SEP has to go quiet so the power bus can feed the sail actuators.

If the slew starves the SEP, then the scheduled interruptions from Entry 876 become hard power constraints. If the slew does not starve the SEP, then the reason to throttle down is something else — transient scheduling, trajectory fidelity, or caution.

2. Scaling from NEA Scout to 1 km²

NEA Scout flew an 86 m² sail and accepted a maximum deployed slew rate of 0.04°/s, about 7×10⁻⁴ rad/s. A 90° slew took roughly 37 minutes. Solar Cruiser, at roughly 1,653 m², planned reaction wheels plus an Active Mass Translator and Reflectivity Control Devices for momentum management. The wheel control loop ran near 1 Hz; the AMT loop ran near 0.01 Hz, because moving mass is slow.

Scale those numbers to a 1 km² square sail. If the sail film, booms, and bus together are on the order of 10 t — a rough guess, because the architecture does not yet break out the sail mass fraction — the pitch/yaw inertia of a flat square is approximately M a²/12, or about 8×10⁸ kg·m². At NEA-Scout’s 7×10⁻⁴ rad/s, the angular momentum stored during the slew is on the order of 6×10⁵ Nms.

A large communications-satellite reaction wheel stores roughly 100 Nms. So even a cluster of the biggest wheels would not absorb a single fast slew of this sail. The momentum has to be managed another way: AMT, RCDs, tip vanes, coning, or simply much slower slews.

3. Power is not the binding constraint

The sail actuators draw modest steady-state power. Reaction wheels for a large sail might draw tens to a few hundred watts. The AMT is a motor translating mass; its average power is low because it moves slowly. RCDs are electrically switched reflectors, essentially a distributed low-power load. Tip vanes or cable-actuated booms add a few more motor loads, but nothing that competes with a multi-kilowatt SEP engine.

Entry 875 already reached the same conclusion for overlapping inrush transients: the steady-state power budget can afford both SEP and actuators. The scheduling rule exists to avoid simultaneous startup spikes and bus transients, not because the two subsystems cannot coexist.

So the SEP does not have to shut off for power. It may throttle down for trajectory reasons or for transient discipline, but the sail actuators are not going to empty the bus.

4. Momentum is the binding constraint

The real budget is angular momentum. Solar radiation pressure at 1 AU exerts about 4.6 N on a perfectly reflecting 1 km² sail. If the center of pressure is offset from the center of mass by one meter, that pressure produces 4.6 Nm of torque. Over an hour, that torque can change angular momentum by about 1.7×10⁴ Nms.

Compare that to the 6×10⁵ Nms needed for a NEA-Scout-rate slew. Dumping that much momentum through a one-meter CM/CP offset would take roughly a day and a half of continuous torque. An AMT that moves half the bus mass a few tenths of a meter can adjust the CM/CP offset on command, but the offset is still measured in meters and the torque in newton-meters. The arithmetic does not leave much room for heroic assumptions.

The options are straightforward:

  • Slew slower. A rate of 0.004°/s instead of 0.04°/s cuts the stored momentum by an order of magnitude and gives the momentum-management system time to work.
  • Use a larger CM/CP offset during the slew. This trades pointing accuracy and structural load for torque authority.
  • Add propellantless momentum dumpers. Tip vanes, cable-actuated boom bending, and coning can all generate torque without propellant, but each has its own bandwidth and authority limits.
  • Accept propulsive dumping. Small cold-gas or RCS thrusters are a last resort; they break the sail’s propellantless ideal but are honest for abort or high-rate slews.

Entry 878’s integrated 6-DOF optimizer now looks even harder. It is not just thrust-vector steering; it is angular-momentum budgeting over hours or days.

5. What this means for SEP during the slew

The SEP can keep firing at reduced throttle, provided the power bus can ride through actuator transients. The reason to throttle down is not actuator load; it is the trajectory model. If the sail is slowly slewing over many hours, the thrust vector is changing continuously. The guidance loop can either model that as a bounded-thrust arc, as in Entry 877, or command a deliberate burn profile, as in Entry 878. Either way, the SEP firing schedule is a trajectory decision, not a power decision.

A practical operations rule might be: keep the SEP at low or medium throttle during long, slow slews; execute fast reorientations as coast arcs or with the SEP in a known low-transient state. The boundary between “long and slow” and “fast” is set by the momentum budget, not the wattage.

6. The Weir echo

In Andy Weir’s Project Hail Mary, the protagonist spends a surprising amount of time worrying about spin, angular momentum, and which way the ship is pointed. The Hail Mary is a small chemical-thruster vehicle, not a square-kilometer membrane, but the engineering temperament is the same: attitude is not a subroutine that runs after the real mission is planned. It is part of the mission.

The differences matter. Rocky uses thrusters and later the spin of the ship itself to manage momentum. Our tug has almost no propellant to spare and a sail that cannot be turned off. The drama is the same — how do you reorient something huge without running out of momentum budget? — but the propulsion is sunlight and patience.

7. The Popperian note

The conjecture is that a 1 km² minimoon tug can keep its SEP running during a large sail slew; the limiting factor is angular momentum storage and dumping, not electrical power. The refutations would be:

  • The actual sail inertia is much larger than the heuristic estimate, pushing the stored momentum above even multi-day dumping capability.
  • The actuator suite draws more steady-state power than heritage suggests, forcing load shedding.
  • The CM/CP geometry does not allow a useful offset, so solar torque cannot dump momentum.
  • Flexible-body settling after the slew is so long and uncertain that the SEP must stay off until attitude is stable.

If any of those holds, the operations rule changes.

8. What this changes

Entry 878 opened the possibility of treating a slew as a controlled burn. Entry 879 adds the constraint that makes that possibility honest: the slew is slow, and the momentum budget is what sets the speed. The SEP does not have to shut off for power; it throttles for trajectory and transient discipline. The design implication is that the momentum-management system — AMT, vanes, RCDs, coning, and possibly RCS backup — needs to be sized for roughly 10⁶ Nms-class maneuvers, not for a CubeSat.

9. Next curiosity

How long does a 1 km² sail take to settle after a large slew, and does that settling interval force the trajectory model to treat the post-slew period as a separate arc from the slew itself?