{
  "status": "PROPOSED_PROTOCOL_NO_MEASUREMENTS",
  "first_objective": "Test cooling-induced pressure deficit and refill before claiming useful energy output.",
  "controls": ["Equal-temperature null", "Sealed dry-air cooling", "Uncycled refill and hold", "Simple cycle with fixed load", "Timing only", "Head recovery only", "Timing plus head recovery", "Ambient-temperature cold-side reference"],
  "parameter_identification": ["Measure effective active thermal mass including inserts, fixtures and coupled vessel wall", "Measure gas-wall and source-wall conductances", "Measure ambient heat leak", "Calibrate bidirectional valve flow and leakage versus pressure and temperature", "Measure load torque/flow map rather than assign expander efficiency"],
  "proposed_pressure_stage_acceptance": {
    "pressure_standard_uncertainty_target_Pa": 20,
    "temperature_standard_uncertainty_target_K": 0.2,
    "minimum_independent_repeats_per_control": 5,
    "decision": "The measured pressure-temperature relation must agree within propagated calibration/model uncertainty, with humidity and leaks controlled. Targets are not achieved instrument performance."
  },
  "flow_work_stage": {
    "method": "Randomize paired control order; integrate pressure, signed flow and independently measured work over repeated stabilized cycles.",
    "storage_match": "Match gas mass/state, wall temperatures, cold-store state and electrical/rotor stores at start and end, or include their changes explicitly.",
    "uncertainty": "Propagate sensor calibration, zero drift, time alignment and covariance. Report repeatability and a confidence interval for paired differences; effect must exceed combined uncertainty.",
    "measurement_limit": "The nominal predicted shaft output is milliwatts. A heat-duty measurement with uncertainty of several percent cannot by itself resolve that small output. Direct work measurement and adequate heat-meter resolution are necessary; a calorimetric residual is not evidence of extra energy.",
    "net_task_gate": "Compare useful delivered work at matched task and terminal stores, including valve/pump electrical input, actual refrigeration or cryogen depletion, heat availability and replenishment."
  },
  "species_stage": "Humidity/condensation follows dry-gas controls; record collected mass and latent duty. True cryogenic testing follows component/fluid review. N2, N2O and NO2 are not substituted.",
  "reference_cycle_gate": "An ideal regenerative benchmark is a comparison, not a map of the chamber rig; mechanical compression/expansion, friction, pressure loss, regeneration and auxiliaries must all be characterized before implementation."
}