When the Universe Reaches Zero but Does Not End

Details: Written by: admin; Category: UNNS Research; Published: 10 June 2026

UNNS Substrate Research Program · Cosmological Boundary Routing · 2026

Admissible Boundary Routing:
What Structurally Replaces the Big Bang Singularity?

Structural Classification of Terminal Approach, Finite Bounce, and Cosmological Recession — how the UNNS Substrate adds a third level of description to singularity removal, going beyond equations and observable boundedness to classify the topology of continuation.

3/3 Full Percolation Turning-Locus Routing κ Ratio 7.5× 98.83% α-Persistence 0 Unstable at Bounce Crossing Bridge v1.1 Pilot Corpus

Instruments: STRUC-PERC-I v2.5.0 · STRUC-I v1.0.4 · Bridge v1.1 · α-Deformation v2.2 Source: Planck 2018 · Synthetic Friedmann · Synthetic LQC (effective, v2.2) Status: Pilot manuscript · 2026

Abstract

Standard descriptions of cosmological singularity removal operate at two levels: the equation level (whether the dynamics diverge) and the observable level (whether density and curvature remain bounded). Neither fully characterizes what structurally happens at the critical regime.

This manuscript introduces a third level — the route-topology level — and applies it to three matched synthetic cosmological trajectories: classical Friedmann contraction (Dataset A), an effective Loop Quantum Cosmology bounce (Dataset B), and Planck 2018-anchored ΛCDM expansion (Dataset C). All three canonical direct encodings return Full percolation, but the orientation-sensitive Bridge v1.1 identifies three fundamentally different route classes: represented terminal approach, finite turning-locus routing, and boundary recession.

The central finding: singularity avoidance is not fully characterized by bounded curvature alone. Structurally, it is the replacement of terminal route topology by admissible turning topology. This result is supported by direct connectivity, structural-pressure, α-deformation, and orientation-sensitive bridge analyses across all three datasets.

▶ Interactive Dashboard — Cosmological Boundary Routing Open fullscreen ↗

▶ Cosmological Boundary Routing — Visual Overview Open in new tab ↗

Visual Overview. An animated introduction to the Cosmological Boundary Routing corpus — covering the three-trajectory design (classical Friedmann contraction, effective LQC bounce, Planck-anchored ΛCDM expansion), the route-topology classification, connectivity and structural-pressure results, the turning-shell localization pattern, and the cross-model boundary-routing taxonomy.

🔭 The Question That Started This

Physics has long known how to remove the classical Big Bang singularity at the equation level: quantum gravity corrections, loop quantization, or effective field theory modify the Friedmann dynamics so that density remains bounded and the scale factor never reaches zero. Standard Loop Quantum Cosmology, for example, replaces the classical divergence with a finite bounce at a critical density ρ_c.

But a fundamental question remained unanswered:

The Structural Question

If the classical singularity is removed, what actually replaces it structurally? Bounded curvature is not the same as a characterization of the topology of continuation. A trajectory may remain finite while still approaching a terminal structural limit. Or it may reach a finite critical locus and reverse. These are different outcomes — and they require a different analytic language.

The UNNS Substrate program introduces that language. Rather than asking only whether a singularity is removed, it asks what kind of structural route replaces the classical terminal path. This is the route-topology level of description.

Route classes identified

terminal · turning · recession

Datasets processed

A · B · C

Structural analysis layers

STRUC-PERC-I · STRUC-I · α · Bridge

Full percolation verdicts

3/3

zero Hard / Giant / Tail

α-admissibility persistence (B)

98.83%

982/84,000 deformation rows unstable

Unstable rows at exact bounce

2 crossing intervals — all admissible

Two-panel infographic: left shows three levels of description (equation, observable, route-topology); right shows the matched pilot corpus with Datasets A, B, C and their route-coordinate classifications.

Figure 1. Three Levels of Description and Route-Topology Classification. The left panel distinguishes the equation level (whether dynamics diverge), the observable level (whether density and curvature remain bounded), and the route-topology level introduced here. The right panel presents the matched pilot corpus: Dataset A (classical Friedmann contraction ending at the terminal cutoff a = 10⁻⁸), Dataset B (effective LQC bounce at H = 0, ρ = ρ_c), and Dataset C (Planck-anchored ΛCDM expansion). On the shared route coordinate s ∈ [−1, +1], the three cases are classified as represented terminal approach, finite turning-locus routing, and boundary recession.

🌌 Three Matched Trajectories, Three Route Classes

The pilot corpus is built around three matched cosmological model constructions, each representing a distinct way a cosmological trajectory can behave at the critical regime. Their route classes are not assumed — they are measured.

Dataset A — Classical Friedmann

Classical contracting universe approaching the terminal numerical cutoff. Scale factor decreases, density increases, curvature increases, provisional margin decreases. Route class: −1 → 0. Represented terminal approach in the selected classical chart. Not a simulation of crossing the mathematical singularity.

Dataset B — Effective LQC Bounce

Full contraction–bounce–expansion trajectory from the effective LQC equation H² = (8πG/3)ρ(1 − ρ/ρ_c). At the bounce: H = 0, ρ = ρ_c, a_min > 0. Route class: −1 → 0 → +1. Finite turning-locus route. Direct analysis covers the post-bounce expansion ladder (n = 2000); full path is analyzed via α-deformation and Bridge v1.1.

Dataset C — Planck ΛCDM

Expansion away from an early critical reference regime, calibrated to Planck 2018 best-fit parameters (H₀ = 67.32 km/s/Mpc, Ω_m = 0.3158). Route class: 0 → +1. Boundary recession. Provides the observationally anchored reference background against which the other trajectories are compared.

The Key Design Principle

The three datasets are not merely different equations. They represent different continuation topologies at the route-topology level. This distinction is invisible to equation-level or observable-level analysis — it requires orientation-sensitive structural encoding.

Route-Coordinate Topology — Shared coordinate s ∈ [−1, +1]

Arrows indicate route-coordinate progression (not physical direction). The bounce circle at s=0 for Dataset B marks the tested-admissible turning locus.

⚡ Discovery 1: Full Percolation Does Not Determine Route Topology

When all three direct ladder encodings returned Full percolation, the initial reading might be that the systems are structurally equivalent. They are not. Full percolation is a necessary but not sufficient condition for route equivalence.

Core Discovery — Three Distinct Axes

Admissibility ≠ structural pressure ≠ route topology.
Final connectivity class ≠ connectivity depth.
Full(A) ∧ Full(B) ∧ Full(C) ⇏ A ≡_route B ≡_route C.

Direct Connectivity Threshold κ_conn

Dataset B requires a connectivity scale approximately 7.5× larger than A and C. All three eventually connect, but they do not connect with equal ease or equal structural depth.

Mean Structural Pressure ρ̄ (STRUC-I)

Dataset B’s direct post-bounce encoding maintains Aκ=1 throughout the tested STRUC-I range, while the full path remains 98.83% admissible under α-deformation.

The Deeper Theoretical Statement

A system may remain realizable and connected while moving into a shallower, more deformation-sensitive region of admissibility space. This is not failure — it is high-pressure admissibility. The distinction separates: whether a state is allowed; how robustly it is allowed; how close it is to structural reorganization; and what route it follows through that region.

Four-panel figure: bar charts of κ_conn and mean structural pressure for A, B, C; schematic giant-ratio curves; percolation verdict icons.

Figure 2. Direct Connectivity and Structural Pressure (Direct Encodings). All three canonical direct encodings reach Full percolation with A_κ = 1 across the tested STRUC-I range, but they occupy different structural regimes. The post-bounce LQC expansion encoding reaches full connectivity at κ_conn = 0.10000, approximately 7.5 times the A/C threshold of 0.01334. Its mean structural pressure, ρ̄ = 0.30605, is approximately 7.36 times the A/C values. A and C are classified as Stable Structure; B occupies the higher-pressure Weak Persistence regime. Dataset B's direct values refer specifically to its 2000-point post-bounce expansion encoding.

🔍 Discovery 2: Magnitude Geometry Erases Route Topology

One of the most conceptually important findings concerns Datasets A and C. Under sorted-magnitude ladder encoding, they are nearly degenerate — their gap structures are nearly identical because sorting removes the direction of evolution.

Sorted Magnitude Encoding

A ≈ C

κ_conn, GR, A_κ, α-vectors all identical. Route orientation completely erased. Classical contraction and Planck expansion appear as the same structural object. This is a false equivalence.

Orientation-Sensitive Encoding

A ≠ C

Phase-sign vectors are exactly opposite. Dataset A: (−, +, +, −). Dataset C: (+, −, −, +). Every channel reverses. This is the full physical distinction: a contracting universe versus an expanding one.

This finding directly strengthens the Canonical Ladder problem within the UNNS Substrate:

Canonicality Is Purpose-Sensitive

A magnitude-only ladder may be sufficient for connectivity analysis. It is not sufficient for direction-sensitive route classification. The correct lesson is not that one representation is wrong — it is that a canonical chart must preserve the structure relevant to the question being asked. Route topology requires orientation to be retained.

Dataset B, by contrast, remains magnitude-distinct from A and C even in sorted-magnitude geometry, through its elevated connectivity threshold, higher structural pressure, and different α-vector. B's turning topology, however, is visible only through signed-flow encoding.

🔄 Discovery 3: The Effective LQC Bounce is a True Turning Topology

The full Dataset B path is not merely an expanding branch with different numbers. It is a continuous trajectory joining contraction and expansion through a finite turning locus — and this reversal is visible in the structural data.

The pre-bounce and post-bounce empirical phase-sign vectors are:

    
          Phase
          Empirical Phase-Sign Vector Σ̄P
          Δln a mean
          Margin flow
        
          A — Terminal approach
          (−, +, +, −)
          −0.500
          −0.093
        
          B pre-bounce
          (−, +, ≈0, −)
          −0.305
          −0.100
        
          B turning surface
          (≈0, ≈0, ≈0, ≈0) · H̄ = 1.76×10⁻⁴
          0
          0
        
          B post-bounce
          (+, −, ≈0, +) ≈ −Σ̄pre
          +0.305
          +0.100
        
          C — Boundary recession
          (+, −, −, +)
          +0.500
          +0.093

Phase	Empirical Phase-Sign Vector Σ̄_P	Δln a mean	Margin flow
A — Terminal approach	(−, +, +, −)	−0.500	−0.093
B pre-bounce	(−, +, ≈0, −)	−0.305	−0.100
B turning surface	(≈0, ≈0, ≈0, ≈0) · H̄ = 1.76×10⁻⁴	0	0
B post-bounce	(+, −, ≈0, +) ≈ −Σ̄_pre	+0.305	+0.100
C — Boundary recession	(+, −, −, +)	+0.500	+0.093

The Route-Reversal Signature

Σ̄_B,post ≈ −Σ̄_B,pre

This is the operational structural signature of route reversal. The reversal is not inferred only from H changing sign. It is simultaneously visible in the reversal of Δln a, density flow, margin flow, and branch orientation. Dataset B is not simply "non-singular." It is structurally a finite turning-locus route.

Signed-flow panel comparing A, B pre, B turning, B post, and C phases with phase-sign vectors and magnitude degeneracy inset.

Figure 4. Signed Flows, Phase-Sign Vectors, and Magnitude Degeneracy. Schematic phase-mean profiles summarize four orientation-sensitive channels across the A, B, and C route phases. Datasets A and C are nearly degenerate under sorted magnitude encoding but possess opposite signed-flow orientations and are therefore route-distinct. Dataset B exhibits an approximate phase-sign reversal across its finite turning locus (Σ̄_B,post ≈ −Σ̄_B,pre), while the turning-surface vector remains approximately zero. The inset demonstrates that a representation may preserve scale and gap geometry while erasing the topology and direction of continuation.

🌀 Discovery 4: The Turning Shell

The α-deformation analysis of the full 4001-row LQC path produced one of the most striking findings of the project. The naive expectation would be:

The Naive Expectation (Not What Was Found)

Maximum criticality ⇒ maximum structural instability at the exact bounce. If ρ/ρ_c = 1 and H = 0 mark the most extreme point, the α-deformation analysis should reveal maximum sensitivity there.

The data show something different. The instability pattern across the full 4001-row LQC path is:

Total: 982 unstable rows of 84,000 tested · α_persist = 98.83%

Shell region: 0.03 ≤ ρ/ρ_c ≤ 0.82 on each side

The Turning-Shell Discovery

Maximum route reversal ≠ maximum detected deformation sensitivity.

The instability pattern is 0 | 491 | 0 | 491 | 0: stable outer regions, sensitive near-bounce shells, and a stable exact crossing. Instability does not peak at the turning locus — it is localized symmetrically around it.

The manuscript distinguishes two things carefully: the measured result (symmetric turning-shell localization) and the proposed interpretation (Turning-Locus Shielding Hypothesis). The hypothesis — that the exact route-reversal locus may be structurally protected relative to its surrounding transition neighborhood — is formally labelled Hypothesis 9.1 and is proposed for replication, not presented as a theorem.

Horizontal route-coordinate diagram showing the five instability regions of Dataset B, with 0|491|0|491|0 pattern and bounce crossing marked tested-admissible.

Figure 3. Turning-Shell Localization in Dataset B (Alpha-Deformation Results). Alpha-deformation sensitivity is absent from the outer contraction and expansion regions and from the exact bounce-crossing intervals. All 982 detected unstable rows are confined symmetrically to the two near-bounce shells, with 491 rows on each side. The exact crossing at H = 0, ρ/ρ_c = 1, route coordinate s = 0 remains tested-admissible under every applied deformation. The measured result is symmetric turning-shell localization; its interpretation as local protection of the exact turning locus is the Turning-Locus Shielding Hypothesis 9.1.

🔮 What Zero Actually Means at the Critical Point

At the effective LQC bounce, H = 0 and ρ/ρ_c = 1. A density-based provisional margin therefore reaches zero. This might be interpreted as structural failure. The analysis shows otherwise.

Critical Coordinate Zero ≠ Universal Failure

The meaning of zero in a critical coordinate depends entirely on route topology. The same numerical value — zero in a margin component — can mark three different structural events depending on which route class the trajectory belongs to:

Route Class	What Zero Marks	Structural Outcome
Represented terminal approach (A)	End of the represented route	No continuation in selected chart
Finite turning-locus routing (B)	Route-orientation reversal coordinate	Tested-admissible; the route continues
Boundary recession (C)	Early boundary reference starting point	Recession begins from this reference

This insight is one of the deepest conceptual refinements of the project. The classification of zero in a structural coordinate as either obstruction or reversal requires the route-topology level of description — it cannot be determined from equation-level or observable-level analysis alone.

🔬 Relation to Standard Loop Quantum Cosmology

The UNNS contribution is strongly aligned with standard effective LQC but operates at a different level. Standard LQC already provides the modified Friedmann equation, finite critical density, a nonzero minimum scale factor, and bounded background quantities. The UNNS work is not the invention of the bounce — it is the structural classification of the bounce.

Standard Effective LQC Asks

What quantum correction removes the classical divergence? Are density, expansion, and curvature bounded? Does the effective spacetime continue through the bounce? What perturbations and observables follow?

UNNS Route-Topology Adds

What structural route replaces the classical terminal path? Is the critical locus terminal, turning, or receding? What signed structural signature characterizes that continuation? Which admissibility, pressure, and connectivity class does the model occupy?

Complementary, Not Competing

The equation-level result is: the singularity is replaced by a bounce. The UNNS route-level result is: the terminal route is replaced by finite admissible reversal. These are different statements at different levels of description. The UNNS analysis cannot independently establish geodesic completeness — and the manuscript says so explicitly. Its contribution is a structural classification of a trajectory supplied by the effective theory.

⭐ Two Routing Morphologies: A Broader Theory Emerges

The Cosmological Boundary Routing result is best understood alongside the earlier Stellar Boundary Dynamics corpus, which established the first routing morphology. Together, they reveal that boundary routing is a family of topological transition mechanisms.

Stellar — Branching Routing

Core-collapse supernovae. Pre-collapse, light curves, spectral series — all 10/10 Full percolating, but ABC bridge: d_AB=0.401 weak, d_BC=1.287 strong. C branches into a distinct admissible regime. Identity-preserving path broken; multiple admissible downstream regimes.

Cosmological — Turning-Locus Routing

Classical → LQC bounce → Planck expansion. 3/3 Full percolating. Σ̄_post ≈ −Σ̄_pre. Route reversal through a finite tested-admissible locus. Identity-preserving path continuous; reversed orientation.

The Broader Theory

Boundary events reorganize continuation rather than merely destroying or preserving states.
Some boundaries branch structure. Some reverse structure. Some terminate structure. Some permit recovery.
This is a general language for comparing different cosmological models without claiming they share the same physics.

🗺️ A General Taxonomy of Boundary Routing

The combined UNNS research supports a provisional taxonomy of how physical systems interact with structural boundaries. The unifying question is not which microscopic mechanism generates the transition, but which structural event occurs at the critical regime.

Morphology	Structural Behavior	Example	Status
Represented terminal approach	Route ends at a limiting locus; no continuation in selected chart	Classical Friedmann contraction (A)	Realized
Finite turning-locus routing	Route reaches finite critical locus, reverses, continues	Effective LQC bounce (B)	Realized
Boundary recession	Route moves away from early critical reference	Planck ΛCDM expansion (C)	Realized
Branching routing	One catastrophic event creates multiple non-equivalent downstream admissible regimes	Stellar Boundary Dynamics I	Realized
Recovery routing	Fragmented representation lifted into connected admissible form	Δ-lifting / ACG cases	Theoretically motivated
Repeated turning routing	Multiple turning loci in cyclic evolution	Cyclic bounce cosmologies	Prediction
Cross-chart continuation	Continuation through a representation change	Conformal Cyclic Cosmology	Open

📡 What Comes Next: Perturbation Implications

The current manuscript concerns background trajectories. It does not yet calculate scalar or tensor perturbations. But route topology creates a natural framework for the next stage of the program.

Terminal Routes

Supply no background-defined transfer map through the endpoint. Pre-terminal perturbations do not connect to post-terminal observables without additional physics.

Turning-Locus Routes

Permit a transfer map: (v_k, v′_k)_pre → (v_k, v′_k)_post. Three structural hypotheses: perturbation continuation, route-localized spectral modification, and background-symmetry transfer test.

Recession Routes

High-k modes may approach standard evolution; low-k modes may retain pre-bounce information. Whether this separation occurs is a model-dependent question for perturbation calculation.

The Critical Non-Implication

Γ₁ ≡_route Γ₂ ⇏ P_ℛ,1(k) = P_ℛ,2(k)

Two models can share route topology and still produce different observable spectra. Route-topology equivalence is a necessary but not sufficient condition for perturbation equivalence. This distinction is essential for comparing LQC, matter-bounce, ekpyrotic, and modified-gravity models.

🔓 What Remains Open

The Canonical Boundary Metric

The strongest open problem is the canonical metric on the admissibility manifold ℳ_adm. The present route coordinate is path-intrinsic — not yet a universal geometric distance. A valid metric must be nonnegative on all represented routes; vanish only at a verified terminal boundary; remain strictly positive at an admissible turning locus; and distinguish terminal obstruction from finite turning.

The Strongest Justified Conclusion

The project has not proved that the Big Bang "never happened." It has shown something more precise and structurally meaningful: within the processed pilot corpus, the classical Friedmann trajectory and the effective LQC trajectory belong to different continuation classes. The classical construction approaches a represented terminal route-zero locus; the effective LQC construction reaches a finite tested-admissible turning locus, reverses its empirical phase-sign structure, and continues onto an expanding branch.

Singularity avoidance is the replacement of represented terminal topology by tested-admissible turning topology.

Open extensions include: replication in alternative bounce models; robustness under grid refinement and normalization changes; perturbation-level transfer and observational likelihood analysis; topological classification of conformal continuation; and extension to black-hole interiors.

📦 Resources and Downloads

📄 Primary Manuscript

Admissible Boundary Routing

34 pages · 7 figures · formal definitions · pilot-corpus propositions · cross-model taxonomy · perturbation hypotheses

📊 Corpus Analytics

Full Numerical Analytics Dashboard

Complete STRUC-PERC-I, STRUC-I, α-deformation, and Bridge v1.1 results · all tables · route diagrams · signed-flow profiles

↓ Corpus Archive

Cosmological Boundary Routing Corpus ZIP

Datasets A / B / C · canonical ladders · α-deformation grids · Bridge v1.1 route profiles · phase summary CSVs · synthesis report

▶ Interactive Dashboard

Full Mission Dashboard

Route topology diagram · κ/pressure bars · α instability · signed flows · cross-model taxonomy · stellar comparison · real-time display

References and Companion Manuscripts

UNNS Substrate Research Program · unns.tech · 2026 · Instruments: STRUC-PERC-I v2.5.0 · STRUC-I v1.0.4 · Bridge v1.1 · α-Deformation v2.0/v2.2 · All formal results scoped to the pilot corpus · route coordinate path-intrinsic · canonical boundary metric open · inv(P_ε; L) ≤ ν(V_ε(L))