D01: ANALYTICAL DELIVERY AND GEOMETRIC SCALING
Independent supplement. Does not change the frozen resonance studies or R07.

WHAT THIS PROGRAM DOES

Evaluates six illustrative acoustic interfaces, 241 dimensionless impedance
ratios, nine outgoing rod-wave scenarios, five geometric scales, and 36
wavelength/aperture rows. These are algebraic scenario evaluations, NOT time
integrations, laboratory trials, fracture predictions, or device rankings.

All material values are explicit illustrative assumptions. No actual specimen,
steel grade, rock structure, quartz orientation, temperature, flaw distribution
or contact condition has been measured here. No guessed actuator efficiency or
high-energy electrical circuit is included.

EQUATIONS AND SI UNITS

For equal-area, lossless normal-incidence longitudinal plane waves with perfect
interface continuity, impedance Z=rho*c (Pa*s/m) gives energy fractions:
  R = ((Z2-Z1)/(Z2+Z1))^2
  T = 4*Z1*Z2/(Z1+Z2)^2
  R+T=1.
Transmitted stress amplitude is 2*Z2/(Z1+Z2) times incident stress. It can exceed
one even though energy transmission never exceeds one. Do not confuse pressure
or stress amplitude ratios with energy efficiency.

For a linear, low-frequency slender rod, c=sqrt(Young_modulus/rho), Z=rho*c,
A=pi*diameter^2/4. Outgoing RMS stress, force and particle velocity obey:
  power = A*stress_rms^2/Z = force_rms*particle_velocity_rms
  packet_energy = power*duration
  force_rms = A*stress_rms; particle_velocity_rms=stress_rms/Z.
RMS is over the stated packet interval; it is not a peak-amplitude claim.
The representative frequency supplies a wavelength diagnostic, not a waveform.

Bulk longitudinal steel speed and slender-rod extensional speed are different
models and are explicitly kept separate. The outgoing component's stress does
not decay with guide length in this ideal lossless model; that is an assumption,
not measured evidence that a long real beam has lossless delivery.

RECEIVER AND VALIDITY LIMITS

Rod receiver T, power and energy use a SEPARATE equal-area one-dimensional
junction abstraction. They are not measured coupling from a finite rough rod
end into a boulder or mountain. Finite contact compliance, preload, gaps,
receiver radiation impedance, joints, diffraction, mode conversion, attenuation,
and support motion require an actual dynamic model and calibration.

Every guide row carries length/diameter, diameter/wavelength, and reflected-wave
return-time indicators. Aspect ratio below 10 and diameter/wavelength above .1
are explicitly declared descriptive screens; satisfying them is not a proof
of validity. Short, wide guide rows are retained as model-limit examples. Every
guide has unknown real contact impedance. A reflected wave can return during
the declared 10 ms packet in these finite-guide cases, so the outgoing-component
calculation is not their complete transient or resonant response.

The plotted guide uses length 10 m, keeping all three plotted diameters above
the declared aspect-ratio screen. It still represents an outgoing component,
not the complete finite-guide dynamics.

GEOMETRIC SCALING, NOT A FRACTURE THRESHOLD

For geometrically similar shapes with linear scale s, projected crack area grows
as s^2 and volume/mass as s^3 at fixed density. If a calibrated critical energy
per projected crack area Gc were unchanged, the corresponding area-energy term
would scale as s^2. No Gc value or fracture-initiation threshold is supplied.
Crack-face counting must match the chosen measured Gc convention.

At fixed rod stress and pulse duration, required packet energy scales as area,
s^2. If pulse duration also scales with traversal time, it scales as s^3. At
fixed energy and duration, geometrically scaled rod stress falls as 1/s. These
relations do not show that a small device can split an arbitrary large object.

WAVELENGTH/APERTURE TABLE

lambda=c/f and D/lambda are reported for declared air/water/rock wave speeds,
20, 200, 2,000 and 20,000 Hz, and .1, 1 and 10 m apertures. A lower frequency
has a larger wavelength for the same medium. This table does not predict focal
spot size, near-field behavior, pressure, directivity or an acoustic laser.

PORTABLE REPRODUCTION FROM THE EXTRACTED PACKAGE DIRECTORY

Python 3 with matplotlib is needed for plots. The equations/tests use only the
standard library. Install plot dependencies if they are not already available:
  python -m pip install matplotlib

Use an EXPLICIT FRESH output directory:
  python scaling.py --output ./reproduced-d01
  python -m unittest discover -s . -p test_scaling.py -v

Existing output directories are refused, including nonempty ones, to preserve
evidence. Choose a new output name for a second run. The default, if omitted,
is ./delivery-output relative to the current directory, never a parent path.
The optional --plot-deps argument adds an already installed local package
directory; it is not necessary for a normal matplotlib installation.

The committed evidence reports 13 passing tests covering conservation,
reciprocity, amplitude-versus-energy, units/flux, geometry, finite/invalid inputs,
zero input energy and output-directory preservation. TEST-report.json binds the
tested source hashes. manifest.json binds source and evidence files. Recreated
timestamps and plot metadata may change; compare formulas/results/tolerances.

EQUATION REFERENCES (NOT MATERIAL-PROPERTY VALIDATION)

University of Texas Austin, longitudinal rod speed:
https://farside.ph.utexas.edu/teaching/315/Waves/node34.html
University of Texas Austin, longitudinal-wave energy flux:
https://farside.ph.utexas.edu/teaching/315/Waveshtml/node43.html
MIT OpenCourseWare 6.013, acoustic plane waves/interfaces, section 13:
https://ocw.mit.edu/courses/6-013-electromagnetics-and-applications-spring-2009/d3be4ea78b036a6362230fb41780cf54_MIT6_013S09_notes.pdf
Abaqus documentation, traction-separation fracture-energy convention:
https://docs.software.vt.edu/abaqusv2025/English/SIMACAEITNRefMap/simaitn-c-cohesivebehavior.htm

No hardware control, account access or publication is performed by this code.
