Author: Benjamin Eckenfels — Co-Founder, SQUIRL OS Technologies LLC, U.S. Navy Veterunteer (Nuclear Submariner), Project SOLACE Sponsor. Engineering and receipt machinery: SQUIRL OS Technologies LLC.
Project SOLACE is a concept for a seafloor HALEU-fueled micro-reactor quietly powering an unmanned-underwater-vehicle dock for decades: Stirling conversion, lead-bismuth primary coolant, seawater immersion as the ultimate heat sink, 5 kW continuous load on a 48 V DC bus.
This repository is the open engineering record of what happened when the concept was run through a deterministic gate machine: six conservation-law gates with no LLM in the decision path.
- The original dossier failed 0/6. Numbers gathered from third-party AI research
tools were tested against conservation laws. One figure was 1,849x above what
physics will pay a direct converter.
sim/spec_original.jsonis the failed spec, preserved exactly as received. - The honest rebuild passed 6/6. 60 kWth, 12 kWe at 20% conversion, Stirling
converters, 420 mm core, LBE coolant.
sim/spec_stirling.json. - 122-day frozen-schedule endurance trial (
endurance_sim.py, results sha 43c040a0607fc4fa, schedule sha cd0f13aa1785eb83): 14,640 kWh delivered, 99.89% uptime, +36.5 C worst-case thermal margin, station-blackout drill recovered in 193 minutes, aged plant re-passing all six gates. - The Ten-Thousand-Year Test (
sim_solace_3.py, sha 6ba234c8): 10,000 seeded one-year missions, zero margin violations, worst unit +31.0 C, time-compression validated at 1x/5x/20x. - BREAKER (
sim_solace_4_breaker.py, sha 6f6675c3): the failure-mode battery. Gear breach in 20-43 minutes unprotected; all modes survivable with scram; the derate ladder and tank-and-steam loop demote the pump from killer to derate. Published on purpose: a receipt system that hides bad news is a marketing system.
| Path | What it is |
|---|---|
sim/ |
Gate machine + the four frozen specs (failed original and honest rebuilds) |
endurance_sim*.py, endurance_results*.json |
SIM SOLACE 1-4 endurance trial |
sim_solace_3.py, solace3_results.json |
Ten-Thousand-Year Monte Carlo ensemble |
sim_solace_4_breaker.py, solace4_results.json |
BREAKER failure-mode battery |
papers/ |
The sponsor paper (published) and the per-simulation reports |
THESIS.md |
Simulation-before-build as engineering due diligence |
RECEIPTS.md |
Every hash and verdict in one table |
PROOF_OF_PRACTICE.md |
The evidence bundle for SIM SOLACE 1-4 |
- Sponsor paper (published under the author's name): https://doi.org/10.5281/zenodo.23463567
- The paper's 32-byte hash is anchored on Bitcoin mainnet, block 972104, tx 54764b6d...8dded (hash-only OP_RETURN, no monetary output), and walked a custody round trip on the XRP ledger: hop 1, hop 2.
- Company ledger: https://github.com/LLong2026/four-horsemen
Dual Mesh Benchmark status: MEASURED (scoped) — no neural-mesh inference in the decision path of any result in this repository. Program benchmark: https://doi.org/10.5281/zenodo.23321637
This software is a prototype and is provided for educational and research purposes only. It is not intended for production use, commercial deployment, or safety-critical environments. All systems are experimental and may contain defects on them.
Nothing in this repository is a nuclear design, a licensed reactor design, or engineering advice. The simulations are deterministic arithmetic models of a concept, published as receipts of method, not as hardware claims.