Parked spec · dates and the radiation-pressure arithmetic checked 24 October 2026 · the link budget is still blank
The radiator is half the engine
A closed vehicle spends power just to stay alive. Call that standing draw k. The useful question is how to offset k with energy that arrives from outside. Sunlight is one arrival. A laser beamed from a grid on Earth is another. Vibration, crew motion, and a real thermal gradient are smaller ones. The machine that uses those arrivals is a harvester. Its output stops when the input stops.
The audit that comes first
Any sketch that loops the working fluid back so the heat can be reused is rejected on the whiteboard, before a simulation. A heat engine makes work from heat flowing from hot to cold. Carnot efficiency is η = 1 − Tcold/Thot. Remove the cold side and the efficiency goes to zero. The radiator is half the engine. A simulation built on the closed loop would only draw an impossibility in the shape of a chart.
Primary chain
| Stage | Choice | Why it is the first path |
|---|---|---|
| Source | Sunlight, or a laser beamed from Earth | JAXA’s IKAROS deployed its sail in June 2010 and was reported to accelerate on sunlight. LightSail 2 launched on 25 June 2019 and deployed its sail on 23 July 2019. Breakthrough Starshot is a research program, not a ship in flight. Sending grid power as photons is a real power-beaming idea, and this page does not size a beam. |
| Capture | A photovoltaic cell matched to the beam wavelength | One wavelength skips the turbine, the fluid, and the moving parts. This spec does not lock a cell efficiency. The number that matters is the cell at that wavelength and at the temperature it will actually run. |
| Conditioning | Maximum-power tracking, then a buffer | A capacitor in front of the battery smooths beam jitter and eclipse, and it separates capture from the load. |
The thermal path is the spare
Where a panel is the wrong tool, the spare path is a laser, a boiler, a turbine, a radiator, a condenser, and a pump. Mercury vapor is one historical working fluid for that spare. A TIME report of 8 July 1929 says Hartford Electric Light had used a mercury-vapor turbine since 1923, with the exhaust used to raise steam. The Smithsonian holds the Emmet mercury-boiler records for that work. This page did not read the plant logs. The radiator stays in the drawing. Mercury vapor is poisonous, which is why this path is the spare and not the baseline.
Two ideas that do not survive the audit
A photon windmill does not power the vehicle. Radiation pressure on a perfect mirror is 2P/c. For a 1 kW beam that is 2 × 1000 / 299792458, which is 6.67 micronewtons. A surface that absorbs the beam sees half of that. This is the arithmetic, not a force measured on a vehicle here. Photon pressure is useful as propulsion over months. It is a poor way to harvest watts. A Crookes radiometer spins on residual-gas heating, and it stops in a hard vacuum.
An orthogonal mercury loop that returns the heat to the start fails for the Carnot reason above. A cycle already returns to its start state. The heat still has to leave at the condenser. Pistons winding back spend work on friction.
The small vehicles are a parts list
The notebook calls the loads “mice”: small platforms the harvester would power. The list is a wheeled base, dry or electrostatic adhesion for a wall in low gravity, a tethered crane head, and onboard harvesting from sun, radio or laser, motion, and a thermal gradient. There is no mass budget in the note, and there is no split between Earth test, habitat, and vacuum. The crane head points at a separate Mars tether and fiber-relay spec in the public repository. That file is a written architecture. It is not a vehicle that has flown. mars_tethered_pushline_relay.md
The same notebook asks for hidden checks on amplitude decay, a lost beam, and thermal runaway before any larger machine is unlocked. Those checks are not built, and they are not one of the downloads on this site.
Blanks still open
- Delivered beam intensity at the range that matters.
- Matched-cell efficiency at the chosen wavelength and temperature.
- Radiator area against the duty cycle.
- The standing draw k this harvester is actually supposed to offset.
Until those four numbers exist, this page is a premise audit and a parts list. It is not a power system for sale.