[cmccomb]

Scope

Investigate issue #599 with a derived heave-only mass shift on disposable copies of the pinned Nonlinear_Hydro/ode45 inputs. The WEC-Sim core, Applications repositories, and the existing published-input 0.05 s numerical solver-envelope gate remain unchanged. This PR does not claim published-input parity; issue #599 remains open.

The core shifts D = adjMassFactor × trace(A_xyz) into rigid translational mass and sends the residual added-mass acceleration force through a 100 ns Transport Delay. Each copied input derives adjMassFactor = A_zz / trace(A_xyz) from the pinned source HDF5 and selected wave mode. For these single active heave-coordinate models, the delayed heave coefficient becomes zero while effective heave mass remains m + A_zz. The shift also changes inactive surge and sway masses, so this is not a general multidimensional core correction.

Measured evidence

The full 150 s R2025b source run at head 963f97442da8258578c1fde94c9f6d3bcb78050e completed Regular and RegularCIC at the published 0.05 s clock and a refined 0.005 s solver/nonlinear-force/CIC clock. Both retain a 0.05 s output grid. The actual applied delayed heave force stays below 6e-22 N; inactive accelerations stay below 1.3e-21.

Maximum heave gap over 150 s	Regular	RegularCIC
Corrected 0.05 s vs independent Python	23.525 mm	23.770 mm
Corrected 0.05 s vs corrected 0.005 s	21.268 mm	21.414 mm
Corrected 0.005 s vs independent Python	2.265 mm	2.378 mm

The corrected 0.005 s comparison passes the established pointwise physical limits for heave, speed, PTO force, and absorbed power. The corrected 0.05 s comparison fails those 3/3.5 mm physical heave limits and even the unchanged-input test's broader 22/23 mm numerical solver envelope. Its source self-change also exceeds the established 3/3.75 mm heave stability screens. Removing active heave added-mass feedback alone does not close the published-clock gap.

Reproducibility

Measured results and provenance include the exact source commits, run, generated HDF5, and compact traces. Ordinary Python CI independently advances motion and PTO against these traces. It has a positive refined physical gate and two strict expected failures for corrected published-clock physical closure and source step stability; an unexpected pass fails. The source workflow regenerates and repacks both cases and compares them with the archived traces. No status or badge changes are included.