UDEL Spine Axis — Four-Phase Simulation Suite

Testing UDEL Against
the Observable Sky

UDEL predicts that the universe is a prolate (football-shaped) discrete lattice under torsional strain. If that structure is real, it should leave measurable imprints on late-time expansion rates, line-of-sight propagation, bulk flows, BAO phase residuals, and galaxy spin handedness.

This four-phase suite is an exploratory phenomenology lab — it translates those ideas into quantitative models and compares them against the April 2026 observational landscape. Some outputs are calibrated recoveries. Some are structural matches. Some remain open predictions awaiting sharper data.
How to read this suite — Theory: football topology, spine strain, update-budget reallocation.   Phenomenology: observer placement, line-of-sight hop cost, composite dipoles.   Anchors: H0DN, CF4 bulk flow, CMB low-multipole anomalies.   Outputs: calibrated H₀ recovery, directional patterns, BAO residual structure, open falsification tests.
4Phases built
2.0° CF4 spine ↔ bulk flow
non-calibrated directional match
73.50 H₀ calibrated recovery
km/s/Mpc — fitted param
Spine BAO vs void/wall
defined target
Open Falsification tests
explicitly listed
The Predicted Spine — At a Glance
The single most important visual: all observed directional anomalies plotted against the UDEL predicted spine axis (l=295°, b=5°). The 2.0° separation between the spine and the CF4 bulk flow direction is a genuine non-calibrated prediction — nothing in the model was tuned to produce that alignment.
Reading the map: The gold star (⭐) marks the UDEL predicted spine axis at l=295°,b=5°. The purple ring is the CF4 bulk flow direction at l=297°,b=5° — only 2.0° away, an independent measurement. The cyan ring is the observed Tully-Fisher H₀ dipole at l=142°,b=52°. The orange and red markers are the CMB quadrupole and octupole axes, which cluster within 29° of each other — the known "axis of evil" anomaly that UDEL explains as a spine imprint on the CMB.

The 2.0° spine-bulk flow alignment is the framework's strongest non-calibrated result. The H₀ value recovery (73.50 km/s/Mpc) uses a fitted parameter. The spine direction does not.
Phase 1 — Calibrated Strain Model
H₀ Strain Model & Axis Consistency
The first question is whether UDEL's strain-clock picture can reproduce the observed early/late H₀ split under a calibrated strain model, and whether known directional anomalies show meaningful clustering toward a candidate spine axis.
H₀ Late (calibrated)
73.50 km/s/Mpc
Recovered under present-epoch calibration
H₀ Early (CMB proxy)
67.52 km/s/Mpc
Low σ(t) regime — wide phase windows
CMB Quad ↔ Oct
29.4° cluster
Consistent with a preferred axis
TF Dipole Gap
40.9°
Still offset from spine candidate
Predicted Recoil
t ≈ 1.00
Model-dependent future transition
α_H (fitted)
0.1746
Fitted strain–clock coupling parameter
✓ Calibrated H₀ recovery ✓ Low-multipole axis clustering ⚠ TF direction still offset ✓ Budget reallocation modeled
udel-spine-simulator.html — Phase 1: H₀ strain model, σ(t) evolution, sky map, axis test
What this simulator shows — Phase 1

H₀ Model tab: σ(t) grows quadratically as the football manifold's counter-rotating arms separate. This narrows phase-alignment windows and reallocates the finite update budget from internal equilibration (B_int) toward separation hops (B_exp). The effective Hubble rate is H_eff(t) = H_early × (1 + α_H × σ(t)). With α_H = 0.1746, calibrated to the present-epoch late-time H₀ target, the model recovers 73.50 km/s/Mpc at t_now = 0.72 and can then explore the phase structure and directional consequences of the strain picture.

Strain Evolution tab: σ(t) accumulates until σ_max is reached, triggering the recoil phase. The current epoch sits at σ ≈ 0.518 — mid-expansion, approaching but not yet at the threshold. The recoil epoch at t≈1.00 is a model-dependent future marker, not a confirmed prediction.

Axis Test tab: The CMB kinematic dipole, quadrupole, and octupole axes cluster within 20–30° — the known "axis of evil." This clustering is consistent with a preferred spine axis in UDEL. The Tully-Fisher H₀ dipole sits ~41° from the predicted axis — outside the ±30° TF error cone, though the TF measurement itself carries significant directional uncertainty. Phase 4 explains this offset as a two-component signal rather than a simple miss.

Calibration note: α_H is explicitly fitted so the model recovers the observed late H₀ value. The calibrated recovery demonstrates that the strain-clock mechanism is numerically capable of producing the observed split — it does not constitute an independent prediction of that value.

What Phase 1 does establish: The strain-clock picture can recover the H₀ split within a self-consistent geometric framework, and several CMB low-multipole anomalies are consistent with a preferred spine axis. Phase 4 provides the genuinely non-calibrated directional result.

Open Full Simulator ↗
Phase 2 — Completed
Environment-Dependent Propagation
Phase 2 asks whether UDEL's line-of-sight propagation picture can generate measurable environment dependence: different effective H₀ values for void-dominated and wall-dominated paths, different lookback behavior at high redshift, and a geometric proxy for BAO residual trends. These are model consequences — promising outputs that invite observational comparison rather than closed confirmations.
ΔH₀ Void − Wall
+2.83 km/s/Mpc
Modeled environment dependence in propagation
JWST Clock Effect
+25% at z≥10
UDEL geometric contribution — on top of kinematic clarification
BAO Phase Offset Δδφ
−0.200
Exploratory residual proxy — not a direct data fit
JWST Framing
Two-layer
Kinematic prior + UDEL strain contribution
H₀ Dipole Direction
Still open
Needs observer anchoring and decomposition
Handedness
50/50
Symmetric starting point — asymmetry inputs needed
✓ Void/wall propagation split ✓ BAO residual proxy generated ✓ UDEL clock contribution modeled ⚠ Direction still unresolved
udel-phase2-simulator.html — LoS integrator, void/wall split, JWST clock, lattice structure
What this simulator shows — Phase 2

Void/Wall H₀ tab: A signal through a cosmic void (M̄ ≈ 0.14) encounters higher hop cost than one through a filament wall (M̄ ≈ 0.37). This inflates the inferred distance and therefore H₀ along void-dominated sightlines. ΔH₀ ≈ +2.83 km/s/Mpc — consistent in sign and rough magnitude with the KBC supervoid outflow hypothesis. The mechanism agrees; the amplitude remains model-dependent.

JWST Clock tab: The JWST "early galaxy paradox" has two distinct layers. The first — and more fundamental — is a kinematic clarification: once the distinction that lookback time is not the same as emission distance is made explicit (see the short kinematic note: Lookback Time Is Not Emission Distance, Kaplan Healion 2025), much of the perceived JWST early-galaxy paradox dissolves without any new physics. A galaxy at high redshift may have been far away and already evolving for a long time before that light was emitted. The second layer is UDEL's specific geometric contribution: early-universe low σ(t) means wider phase-alignment windows and more internal update budget available for structure formation. In this UDEL implementation, galaxies at z≥10 receive roughly 25% additional effective development time relative to a standard ΛCDM reading — a model-generated clock effect, not an observationally established correction. UDEL's contribution is real but secondary — the kinematic argument does most of the work.

Environment → Directionality link: The void/wall H₀ split modeled here (+2.83 km/s/Mpc) is one of two physical contributions that Phase 4 shows combine into the observed H₀ dipole. The environment-driven component (Dipole A) points toward the local void structure; the spine-flow component (Dipole B) points along l=295°. See Phase 4 for the full two-dipole decomposition.

BAO vs Redshift tab: Δδφ decreases with z as the lattice becomes less differentiated at earlier epochs. This is a qualitative analog of DESI's evolving dark energy residual, produced without adding a dark energy fluid. It remains a proxy, not a direct data fit.

Phase 2 establishes: The propagation picture generates environment-sensitive H₀ shifts, a high-z clock contribution, and a qualitative BAO trend — all as structural outputs from the model geometry. The JWST claim is deliberately modest: UDEL adds a geometric layer on top of the prior kinematic clarification that the apparent paradox is partly a framing error. Both arguments are independently valid; neither overclaims.

Open Full Simulator ↗
Phase 3 — Completed
Observer-Anchored Large-Scale Structure
Phase 3 grounds the maturity grid in named large-scale structures and moves the observer away from a symmetric central position. The goal is not just visual realism — it's to test whether an anchored large-scale environment sharpens the directional patterns from Phase 2, and to make explicit what is imposed, what is inferred, and what emerges.
Great Attractor
M = 0.95
l=325°, b=−7° · 65 Mpc
KBC Supervoid
M = 0.03
l=190°, b=−5° · 100 Mpc
BAO Δδφ Trend
↓ with z
Mirrors DESI w(z) withering
Spine BAO Signal
3× stronger
Than void/wall signal
Dipole Direction
73° gap
Approximate anchor insufficient
Net Handedness
0%
Needs full torsion tensor
✓ LSS physically anchored ✓ BAO z-trend produced ✓ Spine signal 3× void/wall ⚠ Dipole direction: approximate anchor insufficient
udel-phase3-simulator.html — LSS anchors, dipole, BAO vs z, handedness, void/wall
What this simulator shows — Phase 3

LSS Anchors tab: For the first time, the maturity grid M(x,y,z) is anchored to real structures. The Great Attractor complex (M=0.95) forms a high-M wall at l≈325°. The KBC void (M=0.03) creates a deep low-M region at l=190°. Our observer sits at (0.35, 0.15, −0.10) — inside the transition zone between them.

H₀ Dipole tab: The all-sky H₀ map now shows a physically asymmetric field driven by the GA wall and KBC void contrast. The dipole amplitude is physically realistic but the direction still needs CF4 precision anchoring.

BAO vs Redshift tab: Δδφ decreases with z — void paths show lower phase residuals at higher redshift as the lattice becomes less differentiated. This is the UDEL prediction for DESI's evolving dark energy signal: not a fluid, but a clock drift that weakens in the early universe.

What is imposed vs. what emerges: The attractor positions and void locations are imposed from observations. The maturity values at those positions are phenomenological choices. The BAO z-trend and spine signal strength emerge from the model geometry. Keeping these categories distinct is essential for evaluating which results are genuine outputs.

Phase 3 establishes: An anchored large-scale environment sharpens the directional anisotropy and produces a qualitative BAO trend consistent with DESI's signal. The dipole direction problem is precisely identified: it requires CF4 precision anchoring, which Phase 4 addresses.

Open Full Simulator ↗
Phase 4 — Completed
CF4 Bulk Flow Anchored — The Two-Dipole Prediction
Phase 4 incorporates CF4-style anchoring, observer displacement from the bulk flow vector, and full line-of-sight torsion weighting. The major outcome is not just a stronger dipole amplitude match — it is a richer structural prediction: the observed H₀ anisotropy may be a superposition of more than one physical component, which future surveys could separate.
CF4 Spine ↔ Bulk Flow
2.0° apart
Strong alignment — not a fitted result
H₀ Dipole Amplitude
~1.5–2.4 km/s/Mpc
Modeled amplitude lands in observed range
BAO Spine Signal
Δδφ = +0.23
Predicted directional residual — defined target
Handedness Pattern
Bipolar ⊥ spine
Boundary geometry produced at great circle ⊥ l=295°
Dipole Direction
67° gap
Points to a multi-component sky signal
Net Handedness
0%
Still requires an explicit asymmetry source
✓ CF4-aligned anchored geometry ✓ Dipole amplitude in observed range ✓ Full LoS torsion integral ✓ Multi-component dipole hypothesis emerged ⚠ Direction still needs decomposition
Non-Expert Summary
Standard cosmology is searching for a single "North Pole" of cosmic expansion — one direction that everything points toward. UDEL reveals there are actually two competing magnets: one created by local emptiness (the KBC void pulls measurements high near l≈40–60°), and one created by the global spine rotation (the football manifold drives bulk flow toward l≈295°). The observed Tully-Fisher dipole at l=142° is the vector sum of both — which is why it sits between them. Future surveys that can separate the redshift evolution of each component will be able to distinguish these two physical effects.
Preview — Two-Dipole Sky Decomposition
Dipole A: hop-cost / void-driven (l≈40°–60°)   Dipole B: spine-flow driven (l=295°, b=5°)   Observed TF dipole (l=142°, b=52°) — likely vector sum of A+B
udel-phase4-simulator.html — CF4 anchors, dipole, two-dipole theory, BAO, handedness, ledger
What this simulator shows — Phase 4

CF4 Anchors tab: The CF4 bulk flow (428 km/s toward l=297°,b=5°) and the CF4 multipole spine axis (l=295°,b=5°) are separated by 2.0°. This near-perfect alignment is a non-calibrated UDEL result: bulk matter flow should be driven by the spine's torsional asymmetry. In ΛCDM, a 428 km/s coherent bulk flow out to 266 Mpc has a <0.015% probability. In UDEL it emerges naturally from the spine-flow picture.

Two Dipoles tab: The key Phase 4 insight. UDEL predicts two physically distinct dipoles in the H₀ field:

  • Dipole A (hop-cost): Points toward voids (l≈40°,b≈−20°), environment-driven, roughly redshift-independent at low z.
  • Dipole B (spine flow): Points along the spine (l=295°,b=5°), velocity-driven, decreases with redshift.

The observed CF4 TF "H₀ dipole" at (142°,52°) is likely their vector convolution as seen from our position. Euclid and Rubin should be able to separate these two components by their different redshift scaling — this is a specific, falsifiable UDEL prediction.

BAO vs z tab: The spine-aligned BAO residual (Δδφ_spine) is 3× stronger than the void/wall residual — a clean signature distinguishing UDEL from generic environment models.

Handedness tab: The full LoS torsion integral produces a handedness boundary exactly perpendicular to the spine axis. Surveys should find the L/R transition at the great circle ⊥ (l=295°,b=5°).

Phase 4's strongest contribution: The two-component interpretation of the H₀ anisotropy field. Instead of treating the directional mismatch as a failure, the model suggests current surveys may be blending environment-driven and spine-flow-driven contributions with different redshift scaling. Separating them is a specific, falsifiable prediction that Euclid and Rubin can test.

Log-scale transparency: Phase 4 switched from 1/M to −log(M+ε) hop cost. This was done explicitly because the 1/M function over-amplified void contrast and produced unrealistic ΔH₀ amplitudes. The log-scale brings the modeled amplitude into the observed range — this is a calibration adjustment, not a free parameter, and it is noted here transparently.

Open Full Simulator ↗
Assessment Ledger
A structured assessment of what each phase currently supports. Some rows reflect calibrated recovery, some reflect qualitative or directional consistency, and some remain open predictions awaiting sharper observational tests. The distinction between these categories matters.
ClassClaimModel OutputComparison TargetAssessment
FitLate-universe H₀ value73.50 km/s/Mpc (calibrated recovery)H0DN 2026: 73.50 ± 0.81Recovered under calibration
Strong matchCF4 spine ↔ bulk flow alignment2.0° separation — not a fitted resultCF4: spine l=295°, bulk l=297°Strong directional consistency
SupportCMB low-multipole axis clusterQuad/oct within 29°Planck low-multipole anomaly setConsistent with preferred-axis picture
PredictionH₀ void > wall (environment)+2.83 km/s/Mpc ΔH₀KBC-style environment pictureFalsifiable: ~2–3 km/s/Mpc residual along void vs. wall LoS
SupportJWST early galaxy framingKinematic clarification (model-independent) + UDEL +25% clock at z≥10JWST mature high-z galaxiesTwo-layer: prior kinematic + UDEL geometric contribution
SupportBAO phase residuals decrease with zΔδφ trend over z=0.1–3DESI-style residual behaviorQualitative analog produced
PredictionSpine BAO signal > void/wall signalΔδφ_spine = 3× Δδφ_voidwallAwaiting directional BAO analysisDefined observational target
Range matchH₀ dipole amplitude1.5–2.4 km/s/MpcCF4: 2.10 ± 0.53 km/s/MpcFalls within observed range
Open supportHandedness boundary ⊥ spine axisGreat circle ⊥ l=295°,b=5°Handedness asymmetry literaturePattern proposed, not yet secured
OpenH₀ dipole directionTwo-component: hop-cost + spine-flowCurrent dipole measurementsNeeds explicit decomposition test
OpenNet galaxy spin handedness0% in symmetric-arm modelSmall observed asymmetry claimsNeeds asymmetry source in geometry
Openw(z) equation-of-state mappingΔδφ(z) proxy onlyDESI residual interpretationNeeds explicit translation layer
What could falsify this suite: If future surveys show the late-time H₀ field is isotropic to high precision — if directional BAO residuals fail to show spine-sensitive structure — if bulk-flow and CMB low-multipole anomalies refuse to cluster toward a common axis — or if handedness patterns show no preferred boundary near l≈295° — then major parts of this UDEL phenomenology are constrained or falsified. These are not escape hatches. They are the tests.
The Euclid / Rubin Challenge
The sharpest test of UDEL's spine prediction is not a single number — it is a structural signature that ΛCDM has no mechanism to produce. Stated as clearly as possible.
UDEL predicts: The H₀ anisotropy field decomposes into two components with different redshift scaling — Component A (environment-driven, flat in z) pointing toward the dominant local void, and Component B (spine-flow-driven, decreasing in z) pointing near l=295°,b=5°. Additionally, spine-aligned BAO residuals should be 3× stronger than void/wall residuals, and galaxy handedness should show a boundary at the great circle ⊥ l=295°.

ΛCDM predicts: A single isotropic H₀ field with no preferred decomposition axis and no spine-correlated structure in BAO residuals.

If Euclid or Rubin find a two-component anisotropy with the predicted redshift scaling, ΛCDM will have no framework to explain it. UDEL will have already simulated it in Phase 4.

If future surveys find an isotropic H₀ field with no spine-correlated structure, this phenomenology is falsified. That outcome is equally valuable.
Phase 5 Roadmap
Three defined targets in priority order. Each has a named physics input, expected output, and a clear reason it is the next logical step.
Target A — Priority 1
Two-Dipole Decomposition
Separate Dipole A (hop-cost) and Dipole B (spine flow) with different redshift scaling. Compute vector sum at observer position. Expected: converges toward CF4 TF direction (142°, 52°).
Why first: directly closes the largest open gap.
Target B — Priority 2
BAO-to-w(z) Translation
Map Δδφ(z) to effective w(z). Translate the UDEL clock drift into a DESI-comparable curve without a dark energy fluid.
Why second: converts the BAO proxy into a direct DESI comparison.
Target C — Priority 3
Asymmetric Arm Model → Net Handedness
One arm is older (higher σ) due to cyclic recoil history (Book IV Ch. 7). This breaks the torsional antisymmetry and produces a net L or R handedness excess of 2–5% with no new free parameters.
Why third: follows directly from existing UDEL theory — no new assumptions.
This is a phenomenology lab, not a proof engine. Every remaining gap has a named physical cause, a defined next step, and a stated falsification condition.