Gravity as Boundary-Layer Ebbing

Every orbital system in this framework is wrapped in counter-rotating boundary layers — co- and counter-rotating flows that nearly cancel, forming the barriers that make stable matter possible. These boundaries do three things, depending on how they are disturbed:

• Push back against internal flow → the quantum potential that gives particles their wave nature
• Leak a tiny current of substrate particles → gravity
• Eject a counter-rotating vortex dipole → a photon

All three use the same boundary layer, the same dc1 substrate, the same counter-rotating mechanics. The difference is the mode of interaction: reaction force, leak current, or ejection. The rates are set by the two mutual friction channels, and they are set very differently — modes 1 and 2 run continuously on the dissipative and reactive faces of one response function, while mode 3 has a threshold and does not run at all below 0.776\,c (The HVBK Bridge § Three Modes, One Layer). This chapter is about the second mode — the leak current that we experience as gravity, and why it is so extraordinarily weak.

The Mechanism

Gravity is a physical flow, not a force at a distance. It arises from the net dc1 current that leaks through the counter-rotating boundary layers of orbital system complexes.

Within a boundary layer, the co- and counter-rotating systems nearly cancel, creating an approximately neutral zone. But a tiny fraction f_\text{cross} of dc1 particles transit the boundary per unit time, carrying momentum towards the gravitational source. This transit fraction is tied to the framework’s mutual friction machinery: a single counter-rotating boundary passes the reactive (spectral-flow) share \mathcal C of the same Kopnin vortex-friction response whose dissipative coefficient is \alpha_{mf} (The HVBK Bridge § Gravity, Outer Rim Onset § Route A). At the outer scale \mathcal C\approx6\times10^{-6} per boundary, far larger than gravity’s per-tick f_\text{cross}\approx4\times10^{-31} (below); how the one compounds into the other is open. This “ebbing current” applies force to each boundary as a whole — and since mass is the total rotational energy enclosed by those boundaries (see Mass as Leaking Rotational Kinetic Energy), the force is proportional to the enclosed mass.

Interactive: open the boundary-layer simulation — the close-up between the gravity and galaxy scenes. Three of the boundaries the river crosses, one lattice cell across, each drawn as a breathing pair of vortex sheets (+ω₀ facing the river, −ω₀ behind it) with the vortex lines threading every cell. The river piles up against each face and bounces many times before a tracer gets through, and the chart beside it is the reason: the transit probability is the reactive face of the same response function whose dissipative face is \alpha_{mf}. The step each boundary holds is drawn above them; it is the pull. Slide the baryons’ pull and watch the paired breath hold (a taller step than Newton’s: MOND) or get pried open (Newton’s step), and a guided tour (T) walks through the regimes, the external-field knob, z = 11.4, and the tear past v_L.

Mathematical Form

The gravitational ebbing current density is:

j_\text{grav} = f_\text{cross} \cdot n_1 \cdot m_1 \cdot v_\text{drift}

where v_\text{drift} is the net drift velocity of dc1 particles between massive systems, driven by the asymmetry created by a mass M at distance r. For this to reproduce Newtonian gravity (F = GMm/r^2), the drift velocity must scale as M/r^2 — which follows if the dc1 current is sourced by the total orbital system energy of the source and falls off as 1/r^2 due to geometric dilution in 3D.

From the substrate’s three dimensional inputs — \hbar, c, and the dc1 mass m_1 (that also determine the cell size \xi = \hbar/m_1 c) there is one form the dimensional analysis allows for G:

\boxed{\;G = \frac{\hbar c}{m_1^2}\;\mathcal F\;},\qquad \mathcal F \equiv \frac{G m_1^2}{\hbar c} = \left(\frac{\ell_\text{Pl}}{\xi}\right)^2 \approx 2.78\times10^{-62}.

The other substrate parameters are pure numbers (\nu, \alpha_{mf}, the packing f, 4\pi, v_L/c) that can feed into the derivation of \mathcal F.

\hbar c/m_1^2 is the gravitational constant the substrate would have if its own quanta coupled at full strength; \mathcal F is the unit-free fraction by which real gravity falls short of that. Deriving G means deriving the pure number \mathcal F, which sets the leak rate through a counter-rotating boundary.

A rate \Gamma and a density span mass, length and time exactly, and \Gamma^2/\rho has the same units as G with no leftover factor:

4\pi G\,\rho_\text{DM} = \Gamma^2,\qquad \Gamma = f_\text{cross}\,\omega_1,\qquad \omega_1 = c/\xi .

\Gamma \approx 1.37\times10^{-18}\ \text{s}^{-1} is the substrate’s Jeans rate; \omega_1 \approx 3.1\times10^{12}\ \text{s}^{-1} is the cell’s Compton clock. Their ratio is the transit probability per breath of the cell,

f_\text{cross} = \frac{\Gamma}{\omega_1} = \sqrt{4\pi f}\;\frac{\ell_\text{Pl}}{\xi} \approx 4.4\times10^{-31}\ \text{per tick},

that finds f_\text{cross} as a probability of the crossing, and \mathcal F is close to its square - within the packing factor 4\pi f.

The geometric form holds at any epoch, because it uses the substrate’s own fixed cell density f\hbar/c\xi^4. The Jeans-rate reading equates that cell density with the cosmic-mean \rho_\text{DM}, which is true only today: the cell density is fixed, while the cosmic mean dilutes as (1+z)^3 (WIP-15 item 2a asks why the two coincide now; Galactic Dynamics § Which density sets a_0).

WarningRetired: G = f_\text{cross}\,v_\text{rot,outer}/4\pi

Earlier drafts wrote G = f_\text{cross}\,v_\text{rot,outer}/4\pi with v_\text{rot,outer}\approx750 km/s, giving “f_\text{cross}\approx1.1\times10^{-15}.” That recipe is a coincidence of SI units, not an incomplete derivation, and it has been withdrawn. Two ways to see it:

• A velocity and a pure number carry no mass dimension, so no product of them, with any coefficient, can reach G’s \text{kg}^{-1}.
• The recipe, together with the outer-rim law v_L = c\,(4\pi/\nu)^{1/3}, reduces to the bare numerical statement G \approx 0.155\,c/\nu^2. In CGS the same ratio is 1.55. A physical relation cannot change by a factor of ten when the units change.

The “missing projection factor” P = G\nu^2/c that later drafts chased was this same coincidence, read backwards. The outer-rim speed itself is untouched: v_L = c\,(4\pi/\nu)^{1/3}\approx740 km/s is a clean, G-free result (Outer Rim Onset). It simply does not produce G. Verification: scripts/g_dimensionless_audit.py.

One number, three hats. Combined with relations the framework already uses, \mathcal F turns out to be the same number as two other open quantities. From the MOND scale a_0 = c\sqrt{G\rho_\text{DM}} (Galactic Dynamics) and close-packing \rho_\text{DM} = f\hbar/c\xi^4 (both evaluated today, when the cell density and the cosmic mean coincide),

\mathcal F = \frac{1}{f}\left(\frac{a_0\,\xi}{c^2}\right)^2 ,

and from the Friedmann equation, with N_* = \ln(c/H_0\xi) \approx 69.4 the inflationary e-fold count of WIP-27,

\mathcal F = \frac{3\,\Omega_\text{DM}}{8\pi f}\;e^{-2N_*} \qquad\Longleftrightarrow\qquad \xi^2 = \sqrt{\frac{8\pi f}{3\,\Omega_\text{DM}}}\;\ell_\text{Pl}\,\frac{c}{H_0} \approx 4.2\,\ell_\text{Pl}\,\frac{c}{H_0}.

The lattice cell is the geometric mean of the Planck length and the Hubble radius. That is the long-known “dark-energy scale” coincidence, and the substrate inherits it rather than explains it. Weak gravity, the MOND scale in cell units, and the e-fold count are one number; deriving any of them derives the other two.

The statement is shortest in the Jeans rate. Substituting 4\pi G\rho_\text{DM} = \Gamma^2 into a_0 = c\sqrt{G\rho_\text{DM}} gives

a_0 = \frac{c\,\Gamma}{\sqrt{4\pi}} = \frac{c\,f_\text{cross}\,\omega_1}{\sqrt{4\pi}} \approx 1.16\times10^{-10}\ \text{m/s}^2 ,

so the MOND acceleration is the speed of light times the boundary’s leak rate, up to the Gauss factor \sqrt{4\pi}. One rate sets the strength of G, drives the ebb v_\text{ebb}, and fixes the acceleration at which the ebb’s pull turns MONDian — the keystone of the headliner figure. It carries the same caveat as the Jeans-rate reading, and the caveat matters here. \Gamma is fixed microphysics (it sets G, which cannot drift), while a_0 tracks the cosmic-mean \rho_\text{DM}, which dilutes. So a_0 = c\Gamma/\sqrt{4\pi} is a present-epoch identity, and at redshift z the framework predicts a_0(z) = (c\Gamma/\sqrt{4\pi})(1+z)^{3/2} (Galactic Dynamics § Which density sets a_0; why the two densities coincide today is WIP-15 item 2a).

Two guardrails on where the number can come from.

• Not from cosmic expansion. The Friedmann form invites the Dirac reading, that gravity is weak because the universe is old. That route is closed. If G tracked H^2 at fixed substrate density, it would drift at |\dot G/G|\sim 6\times10^{-11} per year; lunar laser ranging bounds the drift near 10^{-13} per year. Volovik reaches the same verdict from induced gravity (The Universe in a Helium Droplet, §10.5.4). So \Gamma must be fixed microphysics. The a_0 and N_* forms measure \mathcal F today; they are not routes to it.
• Not from the substrate’s own fluctuations. Sakharov–Volovik induced gravity, G^{-1} = N_F\Delta_0^2/9\pi\hbar c^5, with the dc1 scale as cutoff gives a G some 5\times10^{62} times too large (with m_\text{eff}c^2 as cutoff, 7\times10^{44}). The Einstein–Hilbert form and its 4\pi (bridge equation Step A) and the magnitude of G are therefore separate problems. A strong suppression — the leak — is structurally required, not an optional embellishment.

A caution on the search itself. Simple monomials \nu^a\alpha_{mf}^b2^c\pi^d are cheap: 19 of 9,800 reproduce M_\text{Pl}/m_\text{eff} within the present uncertainty on \nu. A power-law fit to \mathcal F therefore carries no evidential weight on its own. A derivation must supply the exponents from a mechanism — for instance, a per-boundary transmission compounded over a counted number of boundaries between the effective quantum and the cell. There is not yet a derivation of \mathcal F. That is the gravity sector’s one owed number (WIP-15 item 2).

Gravity’s extraordinary weakness is this statement read physically. At the outer scale the substrate is deep in its clean limit, where spectral flow across a counter-rotating boundary is almost completely blocked (\mathcal C \to (\omega_0\tau)^{-2}; Outer Rim Onset § Route A). The counter-rotating layers are nearly perfect barriers, and roughly one breath in 10^{31} lets a dc1 quantum through.

Recovering General Relativity

The substrate does not merely approximate Newtonian gravity — it reproduces general relativity exactly. The self-consistent steady-state dc1 inflow velocity is:

v_\text{ebb}(r) = \sqrt{\frac{2GM}{r}}

Substituting this into the Unruh-Visser-Volovik acoustic metric gives

ds^2 = -c^2\,dt^2 + \bigl(dr + v_\text{ebb}(r)\,dt\bigr)^2 + r^2\,d\Omega^2

— the exact Painlevé-Gullstrand form of the Schwarzschild solution, not approximate, not linearized (the + sign encodes the inward direction of the ebb; expanding the square reproduces g_{tt} = -(c^2 - v_\text{ebb}^2) = -(c^2 - 2GM/r), the Schwarzschild lapse). The substrate’s Euler equation produces this flow self-consistently (v_\text{ebb} \cdot dv_\text{ebb}/dr = -GM/r^2 exactly), closing the loop through a fixed-point argument; continuity is where the account is still incomplete, and the next section says exactly where.

All classical static GR tests — gravitational redshift, light deflection, Shapiro delay, perihelion precession, GPS corrections — are exact consequences of this acoustic metric. The density that accompanies this flow is uniform, \rho(r) = \rho_0 — not the hydrostatic profile \rho_0\exp(-\Phi/c^2) a substrate at rest would adopt. The two are the two branches of a single Bernoulli integral, and the ebb selects the free-fall branch. It matters, because a constant \rho is exactly what makes the acoustic metric’s conformal prefactor constant, and so makes the Schwarzschild match exact rather than conformal.

One Integral, Two Branches: Why Gravity Must Be a Flow

This section supports the substitution of the ebb into the acoustic metric. It helps show why this makes sense for the substrate’s mechanical state, why it makes sense for a gravitational variable to be a velocity. A density perturbation in a stiff medium produces a pressure response, that response varies with position, and a spatially varying pressure is a perfectly good candidate for the gravitational field — with the flow, if any, a consequence rather than a cause.

The substrate’s equation of state shows the flow is the only description that works.

The two pictures are two branches of one integral. Take the steady, spherically symmetric Euler equation with any barotrope P = P(\rho), and define the specific enthalpy by dh = dP/\rho:

v\,\frac{dv}{dr} = -\frac{1}{\rho}\frac{dP}{dr} - \frac{d\Phi}{dr} \qquad\Longrightarrow\qquad \boxed{\;\tfrac12 v^2 + h(\rho) + \Phi = 0\;}

with the constant fixed by v\to0, \rho\to\rho_0, \Phi\to0 at infinity. One equation, two unknowns. The mass sets the total, -\Phi = GM/r; how that total divides between kinetic energy and enthalpy is not fixed by Euler at all. The two limits are exactly the two pictures:

Branch	Condition	Result
Hydrostatic	v = 0	h(\rho) = -\Phi; for the acoustic-metric response h = c^2\ln(\rho/\rho_0), giving \rho(r) = \rho_0\,e^{-\Phi/c^2}
Free-fall	dh/dr = 0	\rho(r) = \rho_0 and \tfrac12 v^2 = GM/r, i.e. v = \sqrt{2GM/r} = v_\text{ebb}

So the ebb sets the pressure response to the free-fall branch of the same invariant that produces the hydrostatic profile.

The two rows are endpoints, and the framework uses both — for different configurations. The field of an isolated mass with steady throughput is the free-fall end: that is the ebb, and it is what carries the metric. A bound, relaxed substrate distribution carrying no net throughput — the virialized halo of a cluster or a galaxy — sits at the hydrostatic end, and there \rho_0 e^{-\Phi/c^2} is the right profile (Bullet Cluster, Tidal Dwarfs, constraint SC3). The distinction that must not blur: in the bound case the density profile says where dark-matter mass sits, and that mass then sources gravity through the ordinary channel. It never is the gravitational field — as the no-go below shows, a substrate at rest has no field to be. (The enthalpy in the top row is worth noticing on its own: for the row-2 response \delta P/\delta\rho = c^2, dh = c^2\,d\rho/\rho integrates to h = c^2\ln(\rho/\rho_0) — the substrate’s Bernoulli function is its logarithm.)

Two things follow:

The free-fall branch is immune to the three-pressures ambiguity. Which pressure response governs a quasi-static gravitational gradient — the luminal Bogoliubov slope \delta P/\delta\rho = c^2 or the gentle logotrope c_{s,\text{bg}}^2 = A/\rho (Three pressures, not one) — is a real question, and the selected branch does not care: it is the branch with dh/dr = 0, which holds for any barotrope. Only the hydrostatic branch depends on which slope is inserted. The ebb profile \sqrt{2GM/r} is an equation-of-state-independent consequence of Euler alone, which is part of why it can be exact rather than leading-order.

The ebb’s density is uniform — and that is what makes the Schwarzschild match exact. Read the free-fall row again: \tfrac12 v_\text{ebb}^2 = GM/r and \Phi = -GM/r cancel identically, so h = 0 and \rho \equiv \rho_0. The substrate around a mass is not hydrostatically stratified; it is uniform and falling. This is the same statement as the Euler result dP/dr = 0 already derived in Substrate Dynamics — no pressure gradient in a freely-falling frame, the equivalence principle in fluid form — and it is exactly what the exactness claim needs, because the acoustic metric’s conformal prefactor is \rho/c_s. Constant \rho means a constant prefactor, which can be absorbed into the units; the Painlevé–Gullstrand form is then Schwarzschild on the nose, not Schwarzschild up to a conformal factor. A stratified \rho(r) would leave a residual conformal distortion and spoil the timelike geodesics. The framework asserted uniformity where it needed it; Bernoulli supplies it.

The no-go: a static substrate bends no light

Suppose the gravitational field were the hydrostatic branch — a structured region of substrate at rest, carrying all of -\Phi in its density and pressure. What geometry does a modon see? Set \mathbf v = 0 in the Unruh–Visser–Volovik metric:

g_{\mu\nu} = \frac{\rho}{c_s}\,\mathrm{diag}\!\left(-c_s^2,\;1,\;1,\;1\right).

Everything now turns on whether c_s can vary with position. It cannot. The substrate’s sound speed is density-independent by construction — that is the defining property of the logarithmic equation of state, the one that upgrades exact Lorentz invariance from enforced to structural: dc^2/d\rho \equiv 0 everywhere, against the cubic’s residual g/m_1 \neq 0. With c_s = c constant the bracket is exactly Minkowski, and the metric collapses to

g_{\mu\nu} = \Omega^2(r)\,\eta_{\mu\nu}, \qquad \Omega^2 = \rho(r)/c

— conformally flat, whatever \rho(r) does. A conformally flat metric has vanishing Weyl tensor. Schwarzschild is Ricci-flat, so all of its curvature is Weyl (R_{\mu\nu\rho\sigma}R^{\mu\nu\rho\sigma} = 48G^2M^2/c^4r^6, and none of it in the Ricci part). And null geodesics are conformally invariant, so light in a static logarithmic substrate travels in straight lines no matter how steep the density profile.

A static substrate therefore delivers gravitational redshift — static clocks tick at the rate \Omega(r), so the frequency ratio survives — and zero light deflection. That pairing is the signature of scalar gravity: it is the Nordström branch, which Eddington’s 1919 measurement closed. The effective refractive index the framework uses for lensing, n(r) = 1 + 2GM/rc_0^2, is read off the flow — it is the modon’s coordinate speed in a moving medium, not a density-induced change in c_s — and there is no version of it a resting substrate can supply.

The logarithm forces the flow picture. The irony runs opposite to the objection. A cubic (Gross–Pitaevskii) substrate has c_s^2 = g\rho/m_1 and could do exactly what the objection asks: a density profile would carry a genuine refractive-index gradient and bend light with no flow at all. The framework rejected the cubic on three independent grounds (three routes to one logarithm), and the price of that rejection is that the static route to gravity closes with it. The same property that buys exact Lorentz invariance forbids static-gradient gravity. The substrate must move, because a medium whose sound speed cannot notice its own density has nothing else to offer a photon.

The one escape is an anisotropic effective index — a radial propagation speed differing from the tangential one, which is what a static metric needs in order to have g_{tt}/g_{rr} \neq -c_s^2 and therefore nonzero Weyl. A barotropic scalar medium at rest cannot produce it. Supplying it would require the vortex lattice to align radially around every mass, i.e. gravitationally induced vacuum birefringence at order GM/rc^2 — many orders above the laboratory and astrophysical bounds, and in flat contradiction with the framework’s own Michelson–Morley position.

Cause, consequence, or neither

The question “is the ebb the cause of the geometry or its consequence?” dissolves once the shift vector is named. In general relativity the velocity appearing in the Painlevé–Gullstrand line element is the shift N^r of a time slicing — pure gauge, removable by rewriting the same geometry in static Schwarzschild coordinates. In the substrate it is a medium velocity with a preferred rest frame. These are not the same object, and the distinction is the whole content of the claim that the substrate says more than GR does:

The metric can always be written statically. The medium cannot be static.

The no-go above is what makes that second sentence a theorem rather than a preference. GR’s freedom to slice the geometry is not a freedom to bring the substrate to rest, because the resting substrate is a different geometry — conformally flat, Weyl-free, lensing-free. What GR treats as gauge, the substrate treats as a physical degeneracy that its equation of state has already broken.

What this leaves open

The free-fall branch solves Euler, not continuity. With \rho = \rho_0 and v \propto r^{-1/2},

\nabla\!\cdot\!(\rho_0\,\mathbf v) = -\tfrac32\,\rho_0\sqrt{2GM}\;r^{-3/2} \neq 0,

so a steady ebb requires a distributed sink of dc1 at that rate at every radius — equivalently a throughput \dot M(r) = 4\pi\rho_0\sqrt{2GM}\,r^{3/2} that grows without bound with distance rather than being fixed by the central mass. A point sink alone forces \rho_0 v r^2 = \text{const}, hence v \propto 1/r^2, which is Visser’s canonical acoustic hole and not Schwarzschild. This is the one place the objection’s instinct — that something is being supplied rather than derived — has real teeth, and it deserves a name rather than the word “entrainment.” It is tracked as WIP-36. The leading candidate there reads the ebb as the superfluid half of a two-fluid counterflow with zero net mass flux. The co-rotating flow falls in, the counter-rotating eddies rise out at the same speed, and the two convert locally at exactly the r^{-3/2} rate, so continuity holds with no sink at all. It survives only if gravitational waves see the superfluid velocity; if they do not, GW170817’s Shapiro delay rules it out.

Rotating bodies: frame-dragging from the azimuthal flow

Real masses spin, and their gravitational fields carry angular momentum. In the substrate this must show up as an azimuthal entrainment v_\phi added to the radial ebb — the spinning boundary layers dragging the surrounding dc1 into helical flow. The acoustic metric makes the connection exact: the cross term -2\,\vec v\cdot d\vec x\,dt that already carries the gravitomagnetic sector has an azimuthal piece -2\,v_\phi\,(r\sin\theta)\,d\phi\,dt, so g_{t\phi}=-v_\phi\,r\sin\theta and g_{\phi\phi}=(r\sin\theta)^2 (the conformal factor cancels). The frame-dragging (zero-angular-momentum-observer) rate is then

\omega_\text{drag} = -\frac{g_{t\phi}}{g_{\phi\phi}} = \frac{v_\phi}{r\sin\theta}.

Frame-dragging is the local angular velocity of the substrate flow. The radial ebb v_\text{ebb} lives in g_{tt} and gives Schwarzschild; the azimuthal v_\phi is the only new field needed for the rotating (Kerr) sector.

Why the falloff is 1/r^3, not the bathtub’s 1/r^2. The tempting picture — a draining vortex — is wrong. A free vortex conserves circulation \Gamma = 2\pi r\,v_\phi=\text{const}, giving v_\phi\propto 1/r and \omega_\text{drag}\propto 1/r^2, which is not the Lense–Thirring law. The substrate around a spinning body is not freely draining; it is entrained by the rotating boundary. A purely azimuthal, force-free flow v_\phi=f(r)\sin\theta obeys the \phi-component of the vector Laplacian,

f'' + \frac{2}{r}f' - \frac{2}{r^2}f = 0,

an Euler equation with indicial roots n=+1 (the rigid co-rotating interior) and n=-2 (the decaying exterior). The unique decaying entrainment is therefore

v_\phi(r,\theta) = \frac{2GJ}{c^2}\,\frac{\sin\theta}{r^2} \quad\Longrightarrow\quad \omega_\text{drag}(r) = \frac{2GJ}{c^2 r^3},

which is exactly Kerr / Lense–Thirring to linear order. This is the spin-dipole (\ell=1) analog of the mass-monopole’s Newtonian 1/r: a mass sources a monopole (the radial ebb \sqrt{2GM/r}), an angular momentum sources a dipole (the azimuthal rotlet). The radial law is fixed by geometry alone — zero parameters, the same status as the Schwarzschild profile; only the amplitude carries a coupling, and it is the same G as SC1, applied to the boundary’s angular-momentum current J rather than its mass M.

Gravity Probe B. Feeding Earth’s parameters into this picture, the geodetic precession (from the radial SC1 sector and orbital motion) and the frame-dragging precession (from the new v_\phi) are, for the polar GP-B orbit (a=7027.4 km):

Effect	Substrate sector	Predicted	GP-B measured
Geodetic (de Sitter)	radial ebb (SC1)	6604 mas/yr	6601.8\pm18.3
Frame-dragging (Lense–Thirring)	azimuthal v_\phi	41 mas/yr	37.2\pm7.2

The geodetic value matches GR’s 6606.1 mas/yr to \sim0.03\%; the frame-dragging value (a simple circular orbit-average; the full GP-B model’s 39.2 mas/yr adds eccentricity and Schwarzschild-coupling corrections) sits well within the measured 37.2\pm7.2. The calculation is in scripts/kerr_frame_dragging.py.

Where v_\phi goes supersonic (|v_\phi|=c) the substrate has an acoustic ergosphere — the rotating analog of Unruh’s “dumb hole,” inside which no static substrate parcel can exist. What remains for a full, non-linearized Kerr match (the standard the Schwarzschild sector already meets) is the combined radial-plus-azimuthal acoustic metric and the explicit boundary-layer entrainment fixing the amplitude from first principles; see open problems WIP-24.

The Cosmological Constant Problem — and Its Resolution

The cosmological constant \Lambda is the term Einstein added to his field equations for the energy of empty space itself. Today it is the simplest account of dark energy: the roughly 70% of the universe’s energy budget that does not clump, does not dilute as space expands, and drives the accelerating expansion discovered in 1998. Anything that counts as “vacuum energy” should show up in \Lambda, because gravity couples to every form of energy.

In the substrate, the obvious candidate is the energy held in the counter-rotating boundary layers between all orbital systems. That energy is gravitationally invisible in equilibrium because co- and counter-rotating contributions nearly cancel, it doesn’t couple to electromagnetic probes (dark particles only), and it shows up only through its gravitational effects.

Quantum field theory treats every field as a collection of oscillators, one per mode, and quantum mechanics forbids any oscillator from being fully at rest: each keeps a zero-point energy \tfrac12\hbar\omega. Summing that energy over all modes up to the Planck scale, where our theories stop being trustworthy, gives a vacuum energy density of order the Planck density, \rho_\text{Pl} \approx 5\times10^{96} kg/m³. The observed dark-energy density is \rho_\Lambda \approx 6\times10^{-27} kg/m³. The ratio is about 10^{123}, usually quoted as “10^{120},” and it is widely called the worst prediction in physics. A vacuum that heavy would have blown the universe apart long before atoms could form. The standard escape is that some unknown contribution cancels it to about 120 decimal places. That fine-tuning is the cosmological constant problem.

Volovik’s answer from condensed matter. Grigory Volovik pointed out that physicists inside a superfluid would face the same problem and could see why it isn’t real. Picture an observer living in superfluid helium-4 who knows only its sound waves (phonons). Summing the phonons’ zero-point energies up to the interatomic cutoff, they would predict an enormous vacuum energy for their liquid. But the liquid’s actual energy is set by its atoms, and it obeys an exact thermodynamic identity that never refers to the mode sum. The identity involves two quantities:

• \varepsilon, the vacuum energy density: energy per unit volume of the ground state. Strictly, it is the energy measured relative to the chemical potential of the medium’s conserved constituents, \varepsilon \equiv \varepsilon_\text{total} - \mu n, where n is their number density and \mu is the energy cost of adding one more. This combination is the one that gravitates, as the q-theory paragraph below makes precise.
• P, the pressure of that same ground state: the force per unit area the vacuum exerts on a boundary.

For any medium at zero temperature in complete thermodynamic equilibrium, the Gibbs–Duhem relation (the bookkeeping identity linking energy, pressure and particle number) gives:

\varepsilon + P = 0

Why this is dark energy’s equation of state. Cosmologists label each component of the universe by its equation-of-state parameter w = P/\varepsilon, the ratio of pressure to energy density. Ordinary matter has w = 0 (no pressure worth mentioning), radiation w = \tfrac13, and a cosmological constant w = -1: negative pressure equal and opposite to its energy density. It is also the only equation of state a vacuum can have if it looks the same to every observer, since a Lorentz-invariant stress-energy must be proportional to the metric. Negative pressure matters because general relativity sources gravity with \varepsilon + 3P, not \varepsilon alone. With P = -\varepsilon that source is -2\varepsilon, so a positive vacuum energy repels and the expansion accelerates. The superfluid identity hands us the dark-energy equation of state, w = -1, for free.

Why it is zero, not huge. The identity says more than w = -1. In equilibrium, \varepsilon and P are both exactly zero, not just their sum. Think of a droplet of liquid floating in empty space. Nothing pushes on it from outside, so its pressure at the surface must be zero, or it would expand or contract. The liquid adjusts its density until P = 0, and the identity then forces \varepsilon = 0. This holds no matter how large the microscopic energies inside are, because the atoms rearrange until they cancel. The universe has no outside to push on it, so the vacuum is that droplet. The superfluid self-tunes: any attempt to add vacuum energy changes the density, which changes the chemical potential, which drives flows that relax the energy back to zero. No fine-tuning is needed. The enormous zero-point sum is not wrong as arithmetic. It is simply not what gravitates, because the constituents’ \mu n term cancels it automatically.

The identity is the backbone of q-theory (Klinkhamer & Volovik [R159]). There, a conserved vacuum variable q, whose chemical potential \mu supplies the counterterm, plays the role of n above, so that the Gibbs–Duhem relation \varepsilon_\text{total} - \mu q = -P nulls the gravitating vacuum energy in equilibrium at zero external pressure. Volovik’s working example for q is an abstract 4-form field; the substrate gives q-theory a face: \boldsymbol{q = n_1}, the countable, conserved dc1 number density, with \mu its ordinary chemical potential, driven to zero at the marginal point. His recent generalization sharpens the container further ([R155]): rewriting the Einstein–Hilbert term as a matter Lagrangian KR, the total energy of matter-plus-gravity obeys \varepsilon^\text{gen} = \varepsilon_\text{Matter} + KR - \sum_a \mu^{(a)} q^{(a)} = 0 identically across all homogeneous universes — a cleaner formal statement of SC2’s “the substrate is its own gravitational source,” and a consistency constraint (the conjugate-pair form, not the free-energy form, is the general law) that any substrate thermodynamics the framework writes must satisfy.

Perfect equilibrium therefore predicts \Lambda = 0 exactly. The dark energy we actually observe is small but not zero, so it has to come from the substrate being out of equilibrium. That residual is the subject of the next section.

The residual: an order-unity disequilibrium

Interactive: the vacuum’s ledger draws the paired lattice with its gravitating vacuum energy above it — a kick relaxes away (self-tuning), and the relaxation tracker below runs live against Friedmann, so the history, the attractor and the drive-sets-the-value negative can be seen directly.

The observed \Lambda comes from the universe not being in perfect equilibrium. Cosmic expansion prevents the substrate from fully relaxing, and the residual vacuum energy is quadratic in the departure from equilibrium:

\rho_\Lambda = \rho_\text{ref} \cdot \left(\frac{\delta T}{T_c}\right)^2

The disequilibrium fraction \delta T/T_c this implies depends entirely on which density \rho_\text{ref} the residual is measured against — and that choice, usually left implicit, is the whole story.

Against the substrate’s own density, the disequilibrium is order unity. The substrate’s natural energy density is its ground-state value \rho_\text{ref} \approx n_1 m_1 \approx \rho_\text{DM} (close-packing). Measured against it,

\frac{\delta T}{T_c}\bigg|_\text{substrate} = \sqrt{\frac{\rho_\Lambda}{\rho_\text{DM}}} \approx \sqrt{\frac{5.8\times10^{-27}}{2.25\times10^{-27}}} \approx 1.6 = \mathcal{O}(1).

There is no fine-tuning of the substrate’s state at all — it sits an order-unity fraction away from full relaxation. This is what the dark-energy data measure directly: the local density today is f(0) = 1.25, and the moraine’s harmonic edge z_\text{harm} = -0.25 lies in our future (see Dark Energy and the Crust). We are still inside the downstream wake of the previous bubble; the substrate has not yet re-equilibrated. The nonzero \Lambda today is that un-drained disequilibrium. (This is constraint C7.)

The famous 10^{-61.5} appears only against the gravitational Planck density. The “worst prediction in physics” measures \rho_\Lambda against \rho_\text{Pl} = c^5/\hbar G^2 \approx 5\times10^{96} kg/m³. That reference gives

\frac{\delta T}{T_c}\bigg|_\text{Planck} = \sqrt{\frac{\rho_\Lambda}{\rho_\text{Pl}}} = 3.4\times10^{-62} \approx 10^{-61.5}.

The two readings differ by exactly the ratio of the two Planck scales, \sqrt{\rho_\text{DM}/\rho_\text{Pl}} = (m_1/M_\text{Pl})^2, so the entire “10^{-61.5} fine-tuning” is the order-unity disequilibrium times the substrate-to-gravitational hierarchy factor:

\frac{\delta T}{T_c}\bigg|_\text{Planck} = \mathcal{O}(1)\times\left(\frac{m_1}{M_\text{Pl}}\right)^2 = 1.6 \times 2.1\times10^{-62} = 3.4\times10^{-62}.

Small \Lambda is the same number as weak gravity. (m_1/M_\text{Pl})^2 is not an independent small number. With M_\text{Pl}^2 = \hbar c/G and the Volovik speed m_1 = \hbar/c\xi (C1) alone, it is simply the squared ratio of the Planck length \ell_\text{Pl} = \sqrt{\hbar G/c^3} to the substrate cell \xi; the leak-rate relation 4\pi G\rho_\text{DM} = (f_\text{cross}\,\omega_1)^2 of the Mathematical Form, with close-packing \rho_\text{DM} = f\hbar/c\xi^4, then rewrites that same ratio through the per-tick boundary-transit probability:

\left(\frac{m_1}{M_\text{Pl}}\right)^2 = \left(\frac{\ell_\text{Pl}}{\xi}\right)^2 = \frac{\hbar G}{c^3\xi^2} = \frac{f_\text{cross}^{\,2}}{4\pi f} \approx 2.1\times10^{-62}.

(This section uses the f\to1 cosmology cell, \xi = 112\;\mum and m_1c^2 = 1.77 meV; the adopted 97\;\mum cell of the Mathematical Form gives 2.8\times10^{-62}. The difference is the packing factor f^{1/2}, and every conclusion below holds for either cell.)

The two middle forms make plain that the number is geometric — the cell sits some 10^{31} Planck lengths across (\ell_\text{Pl}/\xi \approx 1.45\times10^{-31}, just m_1/M_\text{Pl} read as lengths, since \xi and \ell_\text{Pl} are the reduced Compton wavelengths of m_1 and M_\text{Pl}) — while the last ties it to the same boundary-transit physics that makes gravity weak. (Every form is genuinely unit-free: f_\text{cross}\approx4\times10^{-31} is a probability per breath of the cell, not the retired SI-only 10^{-15}.) The cosmological constant and Newton’s G are the same problem: deriving the pure number f_\text{cross} (see Open Problems WIP-15 item 2, WIP-16) would predict both at once — the same collapse the framework found between the bridge 4\pi and the Higgs 8\pi.

This same hierarchy is the one hand-tuned input of the closest published log-EOS dark fluid. Chavanis’s logotropic model carries a single dimensionless constant B = 1/\ln(\rho_P/\rho_\Lambda) = 1/283, which he notes is quantum in origin (B\to0 as \hbar\to0); its argument is identically the substrate’s weak-gravity number, \ln(\rho_P/\rho_\Lambda) = 2\,|\ln(m_1/M_\text{Pl})^2| = 283, so B_\text{Chavanis} = 1/(2\,|\ln(m_1/M_\text{Pl})^2|). The number Chavanis inputs to make his logarithm cosmological is the one the substrate routes through the marginal point — see the logotropic comparison for the full three-way map.

The dark-energy scale is the dc1 scale — and the two lengths differ by a known factor. The close-packing relation \rho_\text{DM}c^2 = (m_1c^2)^4/(\hbar c)^3 fixes the substrate’s infrared quantum from the matter density alone, m_1c^2 = \rho_\text{DM}^{1/4} = 1.77\ \text{meV}, whose Compton length is the lattice cell \xi_\text{DM} = \hbar c/\rho_\text{DM}^{1/4} = 111.8\ \mu\text{m} — Route 1 of the bridge equation. The dark-energy density defines its own length by the same fourth root,

\xi_\text{DE} = \frac{\hbar c}{\rho_\Lambda^{1/4}}, \qquad \rho_\Lambda^{1/4} = 2.24\ \text{meV}, \qquad \xi_\text{DE} = 88.1\ \mu\text{m}.

This length is not our invention: \xi_\text{DE} = (\hbar c/\rho_\Lambda)^{1/4} \approx 85\ \mu\text{m} is the canonical dark-energy length (Beane 1997; Kapner & Adelberger 2007 title their sub-millimetre-gravity paper exactly “Tests of the Gravitational Inverse-Square Law Below the Dark-Energy Length Scale”), and it is why Eöt-Wash torsion-balance experiments probe \sim100\ \mu\text{m} at all. The two lengths are not equal, and their ratio is not free:

\frac{\xi_\text{DM}}{\xi_\text{DE}} = \frac{\rho_\Lambda^{1/4}}{\rho_\text{DM}^{1/4}} = \left(\frac{\Omega_\Lambda}{\Omega_\text{DM}}\right)^{1/4} = 1.269.

So the long-noted “\rho_\Lambda^{1/4}\sim meV \sim m_1c^2” coincidence, sharpened, is an identity carrying one dark-sector number: the lattice cell exceeds the dark-energy length by exactly (\Omega_\Lambda/\Omega_\text{DM})^{1/4}. The substrate has a single energy-density scale, and dark matter and dark energy are two fourth roots of it — the coincidence problem (\rho_\Lambda\sim\rho_\text{DM} today) dissolves into one density, and the residual 27\% between “the medium” and “its vacuum energy” is the dark-sector ratio. This is the framework’s honest replacement for the broken electroweak cube root (C-08/C-09): not a second, independent determination of a length, but a single dimensionless identity whose measured side, \Omega_\Lambda/\Omega_\text{DM}, comes from cosmology.

One number, three hats — and a caution against counting it thrice. That same ratio \Omega_\Lambda/\Omega_\text{DM} = 2.59 is the only dark-sector datum in what follows, and it reappears at three powers. The moraine crest we sit on (Dark Energy and the Crust) is its fourth root, f(0) = (\Omega_\Lambda/\Omega_\text{DM})^{1/4} = 1.27; the substrate-referenced disequilibrium above is its square root, \delta T/T_c = \sqrt{\rho_\Lambda/\rho_\text{DM}} = 1.61; and the flatness-normalized Volovik base density is \Omega_{\Lambda,\text{base}} = \Omega_\Lambda^{3/4}\Omega_\text{DM}^{1/4} = 0.540 (the DESI self-consistent fit quotes 0.548, using its fitted f(0)=1.25 rather than the predicted 1.27; WIP-18). These obey f(0)^2 = \delta T/T_c and \Omega_{\Lambda,\text{base}} = \Omega_\Lambda/f(0) identically — one quantity wearing three hats, not three independent confirmations, and the framework must not present them as such. Two honesty notes follow. First, the DESI moraine fit constrains only the shape f(z)/f(0) of the crust — the flatness-normalized expansion law E^2(z) = \Omega_m(1+z)^3 + \Omega_\text{tot}\,f(z)/f(0) depends on the ratio, never on the absolute crest f(0) — so the crest’s agreement with 1.27 is a consistency check, not an independent second read of the ratio. Second, this base density (0.548) is not the close-packing equilibrium (\rho_\Lambda = \rho_\text{DM}, \Omega = 0.26): the base is a weighted geometric mean tilted toward today’s \Omega_\Lambda, and the two must not be conflated.

The Planck-referenced number is a frozen photon mass. The split above is bookkeeping; underneath it is a mechanism, and the substrate’s logarithmic equation of state supplies it. The photon is the substrate’s sound mode (see Emergent Speed of Light); the logarithm makes that mode exactly massless only when the vacuum sits at its marginal, perfectly Lorentz-invariant density — the \mu\to0 critical point that, on the log’s Bogoliubov spectrum, is the literal edge of stability the substrate self-tunes toward (massless light \Leftrightarrow marginal stability; see Substrate Particles § The Marginal Point, grounded in § The Logarithmic Equation of State). The approach slows critically: the rate that relaxes the residual super-criticality (the vacuum sitting just above marginality) vanishes at the critical point, so cosmic expansion outruns it and freezes a small residual in place. Read as a photon mass, that frozen residual is the Hubble energy itself,

m_\gamma c^2 \;\sim\; \hbar H_0 \;\approx\; 1.4\times10^{-33}\ \text{eV}, \qquad \lambda_\text{Compton} \;\sim\; \frac{c}{H_0} = R_\text{Hubble}

— a Compton wavelength of order the horizon, the smallest photon mass the observable universe can resolve, and some 10^{15} below the observational bound m_\gamma\lesssim10^{-18}\,\text{eV}. Expressing this rate-scale residual as an energy density is exactly what the Friedmann equation H_0^2 = (8\pi G/3)\,\rho_\text{tot} does, and carrying it through returns the Planck-referenced disequilibrium (m_1/M_\text{Pl})^2, the rate-to-density conversion factor being the dark-energy fraction 8\pi/3\Omega_\Lambda — Friedmann itself. So the famous 10^{-61.5} is not merely a choice of reference frame: it is the frozen super-criticality of a relaxing vacuum, and the celebrated hierarchy is cosmology’s own dictionary between a rate and a density.

The mechanism selects the history, not the number. The relaxation ODE above is a tracker: written out, \dot{(\delta T)} = -\delta T/\tau + \alpha H with \rho_\Lambda = \rho_\ast(\delta T)^2 and critical slowing \tau\to\infty as \delta T\to0, coupled to Friedmann, has been integrated (Open Problems WIP-16). Two results are robust and one negative is load-bearing. It reproduces the correct dark-energy history — \rho_\Lambda\to0 at high redshift, rising through \Omega_\Lambda\sim0.7 today and freezing toward de Sitter — which is exactly the behaviour a naive \rho_\Lambda\propto H^2 tracker gets wrong (that one holds \Omega_\Lambda constant and fails at high z); critical slowing is what freezes the residual late instead. And it is a genuine attractor in initial conditions: the starting \delta T is forgotten, so “now” is not a fine-tuned instant but a generic point on a universal freeze-out. But the frozen value \rho_\Lambda/\rho_m today is set one-to-one by the drive strength (the inherited relaxation depth) — no drive law tested, self-limiting or not, selects it. So the mechanism explains why \rho_\Lambda\sim\rho_\text{DM} is natural and late without predicting the ratio 2.59. This is a sharper competitive position than the closest published twin, Chavanis’s logotropic dark fluid, which fixes the analogous number (\Omega_\text{de}/\Omega_\text{dm}=e) only through an admitted “cosmic coincidence”: the substrate trades that coincidence for a mechanism (attractor + correct history) at the cost of the precise value. Chavanis himself suggests his e-relation “may correspond to a fixed point in a more sophisticated theory” — the substrate’s tracker attractor is a candidate for exactly that (full comparison).

What remains open. Consistent with the tracker, the value of \Lambda — equivalently the de Sitter horizon entropy S_\text{dS} = 1/(\delta T/T_c)^2 \approx 2\times10^{122}, or the bubble’s size at nucleation — is an inherited initial condition: the residual relaxation depth of the previous cycle, the same status as the crust amplitude B. Volovik self-tuning plus the horizon-entropy fluctuation law \delta T/T_c = 1/\sqrt{S_\text{dS}} make \rho_\Lambda = \rho_\text{Pl}/S_\text{dS} a marginal, scale-free self-consistency — it fixes the form of the relation but not the value, exactly what one expects at the substrate’s critical (\mu\to0) point. So the framework predicts the dark-energy scale (m_1) and reduces the apparent tuning to the gravitational hierarchy; the value is set by the nucleation dynamics of \mathcal{B}^{-1} (see A Universe That Boils breadcrumb 1, and Open Problems WIP-17).

Gravity’s Place in the Framework

Gravity, the quantum potential, and photon propagation all arise from the same boundary physics operating at different scales. The quantum potential (Two Fluids → Quantum Potential) is the reaction force of counter-rotating layers on co-rotating flow. Gravity is the macroscopic leak current through those same layers. And photons are modons — counter-rotating vortex dipoles ejected when boundaries reorganize — which is the subject of the next chapter.