Skip to content

Commit 067e200

Browse files
JonathanMaceCopilot
andcommitted
[infra] Rewrite README in academic voice
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
1 parent 84f5c85 commit 067e200

1 file changed

Lines changed: 62 additions & 91 deletions

File tree

README.md

Lines changed: 62 additions & 91 deletions
Original file line numberDiff line numberDiff line change
@@ -1,116 +1,91 @@
1-
# The Brown Note
1+
# Browntone: Vibroacoustics of Fluid-Filled Soft Shells
22

3-
Can airborne sound make you soil yourself? No — and the mechanical-vs-airborne asymmetry is the interesting part.
3+
This repository contains the analytical framework, source code, and manuscript drafts for a research programme investigating the low-frequency vibroacoustics of fluid-filled viscoelastic shells — with applications ranging from abdominal infrasound coupling to non-destructive produce testing and spectral identifiability theory.
44

5-
The abdomen does have a real low-frequency flexural resonance near 4 Hz. The problem for the "brown note" is not resonance; it is coupling. The airborne path couples so weakly to that mode that the real story is the asymmetry between airborne forcing and mechanical vibration — and that turns out to be a useful problem in shell mechanics, vibration, and inverse theory.
5+
The programme is **purely analytical at present**. A phantom validation protocol has been designed but not yet executed; see [docs/](docs/).
66

7-
**Purely analytical at present.** A phantom validation protocol has been designed but not yet executed; see [docs/](docs/).
7+
## Motivation
88

9-
The same shell model that explains why sound cannot shake your insides also explains when you can, and cannot, invert a resonance spectrum to recover the parameters that generated it.
9+
The human abdomen possesses a real flexural resonance near 4 Hz. The long-standing conjecture — sometimes called the *brown note* — is that sufficiently intense airborne infrasound could exploit this resonance to induce involuntary gastrointestinal effects. We show that the resonance is genuine but the airborne coupling to it is not: at 120 dB SPL the energy-consistent wall displacement is only 0.028 μm, well below the 0.5–2.0 μm range associated with PIEZO-channel mechanotransduction. Whole-body mechanical vibration, by contrast, couples roughly 3.3 × 10⁴ times more effectively (SDOF upper bound; ~1.6 × 10⁴ with modal participation Γ₂ ≈ 0.48).
1010

11-
## The Short Version
11+
The interesting result is not the debunking itself but the asymmetry it reveals — and the observation that the same oblate-spheroidal shell model that explains *why* sound cannot shake your insides also determines *when* a resonance spectrum can or cannot identify the parameters that produced it.
1212

13-
At 120 dB SPL, the energy-consistent airborne displacement of the whole abdominal wall is only 0.028 μm. That sits below the 0.5-2.0 μm cellular activation range usually associated with PIEZO mechanotransduction. The resonance is real; the whole-abdomen airborne coupling to it is not.
13+
## Principal Papers
1414

15-
A second way to say the same thing is that H_mech/H_air ≈ 6.5 × 10^6. Paper 1 resolves the mechanical-vs-airborne asymmetry for whole-abdomen coupling, while Paper 2 shows that localised gas pockets are the notable exception.
15+
### [Paper 1 — The Brown Note](papers/paper1-brown-note/)
16+
*Targeting Journal of Sound and Vibration*
1617

17-
Paper 7 is a restricted scalar inversion that estimates effective rind stiffness from a watermelon tap tone. That forward problem is exactly what motivates the broader identifiability question taken up in Papers 8-10.
18+
Paper 1 resolves the mechanical-vs-airborne coupling asymmetry from first principles. At the n = 2 flexural resonance, ka ≈ 0.01: the acoustic wavelength dwarfs the body, the impedance mismatch is severe, and the incident field barely drives the mode. Mechanical excitation through the body wall suffers none of these penalties. The conclusion is not that abdominal resonance is fictional; it is that the resonance is real while the airborne path to it is negligible.
1819

19-
## If You Read Two Papers
20+
### [Paper 7 — Watermelon Rind Stiffness](papers/paper7-watermelon/)
21+
*Targeting Postharvest Biology and Technology*
2022

21-
If you read only one paper, read Paper 1. If you want to see why the framework is useful beyond the original question, read Paper 7.
23+
Paper 7 applies the same shell framework to a problem with clearer practical value: mapping a watermelon's tap-tone frequency to its effective rind stiffness. The claim is deliberately narrow — inference to eating ripeness requires cultivar-specific calibration not yet demonstrated — but the forward model works well enough to make the inverse question unavoidable: under what conditions does a resonance spectrum actually identify the parameters of interest? That question motivates Papers 8–10.
2224

23-
### [Paper 1: The Brown Note](papers/paper1-brown-note/) — targeting *Journal of Sound and Vibration*
25+
## Supporting Analyses (Papers 2–6)
2426

25-
This is the paper that answers the question. It starts from a simple empirical asymmetry: whole-body vibration in the 4-8 Hz range can produce gastrointestinal effects, while airborne sound at the same frequencies does not. The model explains that asymmetry from first principles.
27+
Papers 2–6 are not independent contributions; they probe the same shell framework across complementary organs, forcing conditions, and physiological scenarios.
2628

27-
At the n = 2 flexural resonance, ka is only about 0.01. The wavelength is enormous compared with the body, the air-tissue mismatch is severe, and the incident acoustic field barely drives the mode at all. Mechanical excitation transmitted through the body wall does not suffer those penalties. The result is not that abdominal resonance is fictional; it is that the resonance is real while the airborne coupling is not.
29+
| Paper | Focus | Contribution |
30+
|-------|-------|-------------|
31+
| [P2](papers/paper2-gas-pockets/) | Gas pocket transduction | Tissue-constrained bubbles couple far more strongly than the whole cavity |
32+
| [P3](papers/paper3-scaling-laws/) | Cross-species scaling | The low-frequency shell mechanics are not uniquely human |
33+
| [P4](papers/paper4-bladder/) | Bladder resonance | Organ-specific geometry yields a distinct resonance landscape |
34+
| [P5](papers/paper5-borborygmi/) | Borborygmi | Bubble-shell acoustics reproduce clinical stomach-growl frequencies |
35+
| [P6](papers/paper6-sub-bass/) | Sub-bass perception | Structural transmission dominates the airborne path at concert levels |
2836

29-
### [Paper 7: The Watermelon Thump Test](papers/paper7-watermelon/) — targeting *Postharvest Biology and Technology*
37+
## Identifiability Theory (Papers 8–10)
3038

31-
This is the unexpected bridge paper. People really do tap watermelons and listen for a deep, resonant thump, but the practice has mostly lived as anecdote and market lore. Paper 7 turns that intuition into a quantitative shell model that maps tap-tone frequency to effective rind stiffness.
39+
The theoretical contribution of the programme rests on four formal results:
3240

33-
That is a narrower claim than "ripeness prediction". Inference to eating ripeness requires cultivar-specific calibration not yet demonstrated. What Paper 7 contributes is a forward model that works well enough to make the inverse question unavoidable: when does a resonance spectrum actually identify the parameters you care about?
41+
1. **Rank deficiency under scalar geometric reduction.** Equivalent-radius formulations reduce the Jacobian's column space and can render spectral inversion structurally ill-posed ([P8](papers/paper8-kac/), targeting a suitable inverse-problems venue).
42+
2. **Identifiability lifting by oblate asphericity.** Breaking spherical symmetry in the oblate direction restores local rank and improves conditioning by ten orders of magnitude ([P8](papers/paper8-kac/)[P10](papers/paper10-capstone/)).
43+
3. **Near-spherical conditioning asymptotics.** As eccentricity ε → 0, the smallest singular value σ₃ vanishes as λ₁ε² + O(ε⁴); the finite floor reported in earlier Ritz calculations was a discretisation artefact ([P9](papers/paper9-lifting-theorem/)[P10](papers/paper10-capstone/)).
44+
4. **Forward adequacy does not imply inverse adequacy.** A model can predict resonance frequencies to within 10% yet remain catastrophically ill-conditioned for parameter recovery ([P10](papers/paper10-capstone/)).
3445

35-
## Supporting Analyses
46+
[Paper 9](papers/paper9-lifting-theorem/) establishes that comparable prolate perturbations do not produce the same lifting — the mechanism is geometry-selective. [Paper 10](papers/paper10-capstone/) contains the formal proofs for the axisymmetric case; extensions to non-axisymmetric geometries and experimental validation are the principal next steps.
3647

37-
Papers 2-6 are supporting analyses, not five unrelated grand claims. They test the same framework across related organs, forcing conditions, and physiological scenarios.
48+
## Selected Quantitative Results
3849

39-
| Paper | Focus | What it adds |
40-
|---|---|---|
41-
| [P2 Gas Pockets](papers/paper2-gas-pockets/) | Local gas inclusions | Tissue-constrained bubbles can couple much more strongly than the whole abdominal cavity. |
42-
| [P3 Scaling Laws](papers/paper3-scaling-laws/) | Cross-species scaling | The low-frequency shell mechanics are not uniquely human. |
43-
| [P4 Bladder Resonance](papers/paper4-bladder/) | Organ-specific variation | The same framework predicts a different resonance landscape for the bladder. |
44-
| [P5 Borborygmi](papers/paper5-borborygmi/) | Gut sounds | Bubble-shell acoustics can account for clinically realistic stomach-growl frequencies. |
45-
| [P6 Sub-Bass Coupling](papers/paper6-sub-bass/) | Airborne vs structural bass | Reinforces that the floor and seat matter far more than the air. |
46-
47-
## Formal Results
48-
49-
The theoretical spine of the programme is compact enough to state plainly:
50-
51-
1. **Rank deficiency of equivalent-radius models.** Sphere-like equivalent-radius formulations are badly ill-conditioned for spectral inversion (P8, targeting a suitable inverse-problems venue).
52-
2. **Identifiability lifting by oblate asphericity.** Breaking spherical symmetry in the oblate direction lifts identifiability and improves conditioning (P8-P10).
53-
3. **Near-spherical conditioning asymptotics.** As spherical symmetry is restored, conditioning worsens sharply; the apparent finite floor in earlier low-order Ritz calculations was a discretisation artefact, and the correct limit is κ → ∞ as ε → 0 (P9-P10).
54-
4. **Forward adequacy ≠ inverse adequacy.** A model can predict resonance frequencies well and still fail as a parameter-identification tool; P7 is the motivating example, P8-P10 make the point formal.
55-
56-
P9 adds the cross-check that comparable prolate perturbations do not produce the same lifting. P10 currently contains the formal results for the axisymmetric case; extension to non-axisymmetric geometries and experimental validation are the next steps.
57-
58-
## Key Results
59-
60-
| Quantity | Value | Paper |
61-
|----------|-------|-------|
50+
| Quantity | Value | Source |
51+
|----------|-------|--------|
6252
| Flexural frequency f₂ (n = 2) | 3.95 Hz | P1 |
63-
| Breathing mode (n = 0) | 2490 Hz | P1 |
64-
| Airborne displacement ξ_air at 120 dB | 0.028 μm | P1 |
65-
| PIEZO cellular activation range | 0.5-2.0 μm | P1 |
66-
| Transfer-function ratio H_mech/H_air | ≈ 6.5 × 10^6 | P1 |
67-
| Mechanical displacement ξ_mech at 0.1 m/s² RMS (SDOF upper bound) | 917 μm | P1 |
68-
| Coupling ratio R (120 dB vs 0.1 m/s², SDOF upper bound) | ≈ 3.3 × 10^4 (≈ 1.6 × 10^4 with Γ₂ ≈ 0.48) | P1 |
69-
| Abdominal Sobol total-order sensitivity S_T(E) | 0.86 | P1 |
70-
| Dimensionless frequency Π₀ (cross-species) | 0.07 | P3 |
71-
| Watermelon Sobol total-order sensitivity S_T(E_rind) | 0.54 ± 0.05 | P7 |
72-
| Condition number κ_sphere | ≈ 1.37 × 10^10 | P8 |
73-
| Condition number κ_oblate | 69.4 | P8 |
74-
| Scattering parameter ka | 0.0114 | P1 |
53+
| Breathing mode (n = 0) | ~2490 Hz | P1 |
54+
| Airborne displacement at 120 dB | 0.028 μm | P1 |
55+
| Coupling ratio R (120 dB vs 0.1 m/s², SDOF) | 3.3 × 10⁴ | P1 |
56+
| Sobol sensitivity S_T(E), abdomen | 0.86 | P1 |
57+
| Sobol sensitivity S_T(E_rind), watermelon | 0.54 ± 0.05 | P7 |
58+
| Condition number κ (equivalent sphere) | ~1.37 × 10¹⁰ | P8 |
59+
| Condition number κ (canonical oblate) | 69.4 | P8 |
7560

76-
## What This Repo Is and Isn't
61+
## Scope and Limitations
7762

78-
- **An analytical programme, not an experimental one.** The papers and code are modelling work; an experimental validation protocol exists, but it has not yet been executed.
79-
- **A rind-stiffness framework, not a ripeness oracle.** Paper 7 maps tap-tone frequency to effective rind stiffness; inference to eating ripeness requires cultivar-specific calibration not yet demonstrated.
80-
- **A mostly finished arc, not a closed book.** Papers 1-9 contain the current core results. Paper 10 is the synthesis still being refined.
81-
- **Latest Paper 10 draft.** [papers/paper10-capstone/drafts/draft_2026-04-04_1400.pdf](papers/paper10-capstone/drafts/draft_2026-04-04_1400.pdf)
63+
- **Analytical, not experimental.** All results derive from shell-theoretic models. A phantom validation protocol exists ([docs/](docs/)) but has not been executed.
64+
- **Rind stiffness, not ripeness.** Paper 7 recovers effective stiffness from tap-tone data; the link to consumer-relevant ripeness requires cultivar calibration not yet performed.
65+
- **Papers 1–9 are submission-ready drafts.** Paper 10 is under active revision.
66+
- **Latest P10 draft:** [draft_2026-04-04_1400.pdf](papers/paper10-capstone/drafts/draft_2026-04-04_1400.pdf)
8267

83-
## For Researchers
84-
85-
The core model treats the abdomen as a fluid-filled viscoelastic oblate spheroidal shell. In the forward problem, an equivalent-radius formulation captures the dominant low-frequency behaviour. In the inverse problem, the full oblate geometry matters because asphericity is exactly what breaks the spectral degeneracy. Flexural modes (n ≥ 2) govern the infrasonic response; the breathing mode (n = 0) sits around 2490 Hz and is not the low-frequency story.
86-
87-
### Quick Start
68+
## Getting Started
8869

8970
```bash
9071
pip install -e .[dev]
9172
python -m pytest tests/ -v
9273
```
9374

9475
```python
95-
from src.analytical.natural_frequency_v2 import (
96-
AbdominalModelV2,
97-
flexural_mode_frequencies_v2,
98-
)
76+
from src.analytical.natural_frequency_v2 import AbdominalModelV2, flexural_mode_frequencies_v2
9977
from src.analytical.energy_budget import self_consistent_displacement
100-
from src.analytical.mechanical_coupling import compare_airborne_vs_mechanical
10178

10279
model = AbdominalModelV2(
10380
E=0.1e6, a=0.18, c=0.12, h=0.01, nu=0.45,
10481
rho_wall=1100, rho_fluid=1020, K_fluid=2.2e9,
10582
P_iap=1000, loss_tangent=0.25,
10683
)
107-
10884
freqs = flexural_mode_frequencies_v2(model, n_max=5)
10985
disp = self_consistent_displacement(model, mode_n=2, spl_db=120)
110-
ratio = compare_airborne_vs_mechanical(model)
11186
```
11287

113-
### Citation
88+
## Citation
11489

11590
```bibtex
11691
@misc{browntone2026,
@@ -121,27 +96,23 @@ ratio = compare_airborne_vs_mechanical(model)
12196
}
12297
```
12398

124-
## Repository Structure
125-
126-
```text
127-
browntone/
128-
├── papers/
129-
│ ├── paper1-brown-note/ # Resolves the mechanical-vs-airborne asymmetry
130-
│ ├── paper2-gas-pockets/ # Supporting analysis
131-
│ ├── paper3-scaling-laws/ # Supporting analysis
132-
│ ├── paper4-bladder/ # Supporting analysis
133-
│ ├── paper5-borborygmi/ # Supporting analysis
134-
│ ├── paper6-sub-bass/ # Supporting analysis
135-
│ ├── paper7-watermelon/ # The thump-test application
136-
│ ├── paper8-kac/ # Identifiability theory
137-
│ ├── paper9-lifting-theorem/ # Oblate-prolate asymmetry
138-
│ └── paper10-capstone/ # Axisymmetric synthesis, extensions next
139-
├── src/analytical/ # Core shell models and coupling calculations
140-
├── tests/ # Regression suite
141-
└── docs/ # Research logs, validation plans, and notes
99+
## Repository Layout
100+
101+
```
102+
papers/ src/analytical/
103+
├── paper1-brown-note/ ├── natural_frequency_v2.py
104+
├── paper2-gas-pockets/ ├── energy_budget.py
105+
├── paper3-scaling-laws/ ├── oblate_spheroid_ritz.py
106+
├── paper4-bladder/ ├── kac_identifiability.py
107+
├── paper5-borborygmi/ ├── universality.py
108+
├── paper6-sub-bass/ └── ...
109+
├── paper7-watermelon/
110+
├── paper8-kac/ tests/ 487 regression tests
111+
├── paper9-lifting-theorem/ docs/ Research logs, protocols
112+
└── paper10-capstone/
142113
```
143114

144-
This work was produced with AI assistance; for methodology and disclosure details, see [docs/ai-assisted-research.md](docs/ai-assisted-research.md).
115+
This work was produced with AI assistance; see [docs/ai-assisted-research.md](docs/ai-assisted-research.md) for methodology and disclosure.
145116

146117
## Licence
147118

0 commit comments

Comments
 (0)