SCIENTIFIC ACCEPTANCE ROADMAP — INDEPENDENT REVIEW, 2026-10-06

These are proposed engineering acceptance gates, not passed physical tests,
standard-mandated thresholds, hardware instructions, or a claim of patentability.
Thresholds must be selected before new outcome data are examined and should
reflect measurement uncertainty and the practical improvement being sought.

1. ACCEPT THE NUMERICAL EXPERIMENT WITHIN ITS DECLARED THEORY

For iteration06, verify cantilever roots and boundary conditions, mass-weighted
modal orthogonality/norms, three-mode equivalence to the unchanged prior engine,
passive contact storage/dissipation, source energy balances, common timesteps
resolved for the highest retained mode, and unchanged controller commands.
Report target-energy and protocol-ratio sensitivity to BOTH timestep and mode
count. Keep incomplete/failed rows and changes in outcome sign. A 5-to-7-mode
comparison is observed truncation sensitivity, not proof of infinite-mode
convergence or correctness of Euler-Bernoulli theory for a physical specimen.
Uniform high-mode damping, line loading, shear/rotary inertia, fixture behavior,
and possible plate/three-dimensional modes remain modeling questions.

2. CALIBRATE ONE REAL SPECIMEN AND ITS COUPLING BEFORE TRANSFERRING GEOMETRY

Record geometry, mass/density, support conditions, crystal/material orientation,
contacts, and sensor/drive calibration. Use low-amplitude observations to fit
mode frequencies, damping and measured input/output transfer functions. Keep
some excitation/sensor positions and records withheld from fitting. Accept a
model only when it predicts those records within declared uncertainty/tolerance.
Raw piezo volts or a single frequency peak do not identify force, effective
modal mass, joules, complete geometry or a fracture threshold.

Quartz requires direction-sensitive elasticity; NIST's resonance-ultrasound
work estimates six independent elastic constants by matching measured and
predicted specimen frequencies. It also reports specimen-dependent internal
friction. These support a material-specific calibration route, not assigning
one universal quartz resonance or using the aluminum beam as quartz data. [Q]

Measure contact force and relative motion, including separation and preload,
to identify the contact law. Mechanical work is the time integral of force
times velocity at the stated interface. Include released preload/fixture
energy. Measure electrical input separately: neither ledger substitutes for
the other, and the probe/driver/launcher/reset contributions must be counted.

3. ADD A CALIBRATED FRACTURE MODEL AND VALIDATE ITS PREDICTIONS

Declare the specific output: crack initiation, completion of a split, path
deviation, off-path damage, or fragment-size distribution. Start with a defined
specimen/loading/notch problem whose force-displacement and crack-extension
data can identify the chosen model. Fit strength/fracture parameters and their
orientation/rate dependence on calibration specimens, then predict different
specimens without retuning. Compare crack-initiation load, response curve,
energy, crack path and damage extent with uncertainty and repeatability.

Published granite/gneiss wedge-splitting experiments show that direction and
notch geometry affect cracking, and obtain fracture information from measured
force/displacement/crack evolution. Their discussion distinguishes a finite
fracture process zone from an ideal LEFM crack. These are useful calibration
examples, not an acoustic cutter demonstration. [R]

An anisotropic phase-field/cohesive/bonded-particle solver is an implementation
choice. Its fracture energy, strength/softening law, spatial regularization and
mesh dependence require calibration/verification. MOOSE documents anisotropic
phase-field fracture capability, but does not provide verified parameters for
this project's quartz, sandstone, diamond or wood specimens. [M]

4. TEST THE VALUE OF VIBRATION/TIMING AGAINST A FAIR BASELINE

Freeze policies and evaluation cases before comparing results. Compare an
ordinary baseline, fixed-timing excitation and measured-frequency excitation
with total energy, preparation/preload, permitted processing time and delivered
outcome explicitly accounted for. Report joules per completed accepted split,
processing time, path error and collateral damage separately. An improvement
in reversible peak modal energy is not itself a fracture-efficiency gain.
If a new decision policy is tuned on the present synthetic cases, use a fresh
frozen evaluation set; retain negative cases and do not call repeated plant
conditions independent evidence of a probability of success.

5. TREAT CONTACTLESS AIR AND LIQUID AS DISTINCT COUPLING BRANCHES

For air, measure the actual source boundary/pressure field and received solid
vibration. Validate an aperture/propagation/interface/finite-solid model on
withheld locations before using it to infer internal tensile stress. Record
distance, aperture, incidence, atmosphere, geometry and orientation. A scalar
plane-wave transmission coefficient is a baseline, not a resonant-device
efficiency or arbitrary-geometry upper bound. A pressure focus does not itself
establish a narrow internal fracture plane.

Air-coupled resonant pulse-echo research demonstrates frequency-based NDT on
regular plates/pipes, with geometry and angle sensitivity. It supports an
initial measurement test, not remote acoustic fracture of a boulder. [A]
Liquid coupling introduces its own propagation, interface and, where relevant,
bubble/nonlinear effects; it cannot be treated as measured dry-air performance.

6. ACCEPT SIZE/MATERIAL TRANSFER ONLY THROUGH NEW HELD-OUT VALIDATION

Each new material/geometry/support/contact regime needs its own measured
response and fracture validation. Increase size in documented stages and
compare predicted versus observed results. Record total system footprint,
energy supply, drivers, fixtures and coupling, not only the contact head.
The current reference-beam results provide no accepted size limit for crystal
fracture and no demonstrated diamond-to-boulder operating envelope. The next
credible milestone is accurate prediction on one instrumented specimen family.

PRIMARY SOURCES CHECKED ON 2026-10-06

[Q] NIST, Elastic Constants of Natural Quartz (2003):
https://www.nist.gov/publications/elastic-constants-natural-quartz

[R] Jacobsson et al., Tensile fracture initiation and propagation of granite
and gneiss at wedge splitting tests, Part 1 (2025):
https://link.springer.com/article/10.1007/s10704-025-00857-z

[M] MOOSE, Compute Cracked Stress, official documentation:
https://mooseframework.inl.gov/source/materials/ComputeCrackedStress.html

[A] Alvarez-Arenas and Camacho, Air-Coupled and Resonant Pulse-Echo Ultrasonic
Technique, Sensors 19(10), 2221 (2019):
https://www.mdpi.com/1424-8220/19/10/2221

The ordered gates, acceptance recommendations and experimental priorities are
this reviewer's engineering judgments. The cited works support their stated
material, measurement and solver facts; they do not validate the proposed
device or prescribe these exact project gates.
