T1: Phase-Guided Graphene Thermal Interface — Technical Whitepaper and Numerical Evidence Ledger

This structured technical whitepaper presents the T1 numerical evidence chain for a phase-guided graphene thermal interface within the Phase-Locked Quantum-Plasma Processor research program. The study develops a claim-bounded reduced-order model that progresses from thermal-balance validation and a virtual RG-21-like reference calibration to passive graphene-network optimization, fixed-budget phase-guided redistribution, mechanism-isolation controls, directional conductivity-tensor steering, geometry tests, state-budget recovery, action normalization, and final joint action-state topology control. The public release contains 88 curated figures covering experiments T1.1–T1.10. Corrective datasets internally labeled T1_7_dual_recovery and T1_8_recovery are incorporated into T1.7 and T1.8 as superseding evidence rather than counted as separate publication experiments. The evidence supports several bounded conclusions within the declared phenomenological model: passive network morphology and through-plane organization strongly affect effective thermal transport; fixed-budget spatial redistribution can reduce hotspot thermal impedance; thermal-demand localization provides the dominant active contribution, with a smaller phase-alignment component; directional tensor orientation affects heat removal; and the tested persistent-amplitude and broad action-normalized topology-control claims are not supported. The final T1.10 confirmatory experiment reproduces a small topology-dependent thermal-response candidate at full production resolution. Relative to the numerical reference, the source-aligned topology improves thermal impedance by approximately 0.148%, while the anti-hotspot topology degrades it by approximately 0.255%, producing a source-to-control separation of about 0.403%. The corresponding hotspot-state localization separation is approximately 0.671. The result is accepted only under a corrected, read-only action-schedule audit and remains explicitly limited to the declared model. This work does not report a manufactured thermal paste, formulation recipe, atomistic graphene simulation, physical phase-locking mechanism, actuator-energy model, certified commercial comparison, or laboratory thermal-conductivity measurement. The apparent conductivity values and thermal-impedance improvements are numerical model outputs. Raw arrays, calibrated parameters, solver inputs, execution schedules, seeds, source code, machine-readable metric tables, and reconstruction-enabling records are retained privately under NDA/proprietary restrictions and are not included in the public release.