Appendix Z: Quasicrystalline Lattice QCD (QLQCD) Algorithm

Z.1 Transition to Non-Perturbative Simulation

The results presented in Volumes 1-3 rely on Geometric Harmonic Analysis (GHA)—an analytic approach validating the low-energy limit of the $E_8 \to H_3$ projection. However, to extract exact, high-energy hadronic observables (such as dynamic parton distribution functions or deep-inelastic scattering cross-sections), the field must transition from analytic approximation to full non-perturbative simulation.

This appendix formalizes the mathematical framework and algorithmic blueprints for Quasicrystalline Lattice QCD (QLQCD), replacing the standard $SU(3)$ hypercubic LQCD with the native geometry of Geometric Consciousness Theory. It is designed for future exascale implementation.

---

Z.2 The Discrete Geometric Action

In standard LQCD, spacetime is discretized as a hypercubic lattice, with fermion fields $\psi$ existing at the nodes and gauge fields $U_\mu \in SU(3)$ existing as parallel transporters (links) along the strictly orthogonal edges.

In QLQCD, the physical vacuum is an Ammann-Kramer-Neri (AKN) 3D tiling generated by the projection of $\mathbb{Z}^6$. The fundamental unit cells are Rhombohedra, not cubes (specifically, prolate and oblate rhombohedra formed by the 6 canonical basis vectors $\mathbf{e}_i$ of the icosahedral star).

Z.2.1 Gauge Links on Rhombohedral Edges Gauge links $U_i(x)$ are defined exclusively on the edges of the rhombohedra. Because the AKN tiling has 6 independent generator directions, the index $i \in \{1, \dots, 6\}$.

The standard Wilson action is replaced by the Icosahedral Plaquette Action: $S_{gauge} = \frac{\beta}{3} \sum_{\text{plaquettes } p} \mathrm{Re}\left[ \mathrm{Tr} (1 - U_p) \right]$ where a plaquette $p$ is a rhombus face rather than a square of the 3D unit cells. The path-ordered product $U_p = U_1 U_2 U_3^\dagger U_4^\dagger$ traces the perimeter of the rhombus.

---

Z.3 The Matching Rule Constraint (Monte Carlo Updates)

The most profound algorithmic departure from standard LQCD lies in the Monte Carlo (MC) update constraints. In hypercubic LQCD, local link variables can fluctuate independently, provided the global action is minimized.

In QLQCD, the vacuum is not a generic graph; it is a rigid projection from 6D defined by specific Matching Rules. The canonical AKN tiling permits a small finite set of distinct local neighbourhoods around any node, classified by how the cut-and-project window intersects the 12 nearest-neighbour displacements of $\mathbb{Z}^6$ projected into $E_\perp$. Direct enumeration on the I_h-closed cage geometry (see GCT_Physics_Engine/src/protocol_akn_vertex_stars.py) finds 9 distinct edge-incidence vertex-star types under $I_h$ equivalence, with coordination histogram $\{4{:}1,\ 5{:}1,\ 6{:}2,\ 7{:}1,\ 8{:}1,\ 9{:}1,\ 10{:}1,\ 12{:}1\}$. With matching-rule decorations enforcing aperiodicity, the decorated-star count is larger. Under the cut-and-project identity that a matching-rule decoration is the radius-2 local window patch (the next-shell data that fixes a vertex's admissible nearest-neighbour updates: $v+d_k$ present iff $\pi_\perp(v)+\pi_\perp(d_k)\in W$, and $v+d_k+d_j$ present iff $\pi_\perp(v)+\pi_\perp(d_k)+\pi_\perp(d_j)\in W$), the radius-2 patch is a function of the perp-coordinate alone, so the decorated stars are the cells of the partition of $W$ by the radius-2 ball of window translates. Window sampling of $W$ (engine: GCT_Physics_Engine/src/protocol_akn_decorated_stars.py) gives the realised decorated-star set $\mathcal{V}_{\text{AKN}}$ with $|\mathcal{V}_{\text{AKN}}| = 101 \quad \text{(radius-2 patches, full } I_h \text{ equivalence),}$ the radius-1 restriction recovering the 9 edge-incidence stars. The count is stable across $90^3$, $130^3$, $170^3$ sampling grids ($\sim 0.42$–$2.9 \times 10^6$ window points); the exactness certificate is a formal cell-arrangement enumeration of the window partition, which a fixed grid does not by itself supply. The often-cited Penrose-style $\sim 30$–$40$ is a lower-dimensional ($2$D) heuristic rather than a $6$D-AKN radius-2 prediction, so the larger value is a change of decoration convention, not a correction of that figure. $\mathcal{V}_{\text{AKN}}$ is the nearest-neighbour viability set used by the §Z.3.1 Metropolis acceptance check; larger-radius decorations refine it further.

Z.3.1 The Algorithmic Bottleneck A Monte Carlo trajectory updating the gauge field $U_i$ or the phason displacement field $\delta \mathbf{w}$ must be rejected if it creates a vertex star outside of the allowed set $\mathcal{V}_{\text{AKN}}$.

• Hypercubic MC: Unconstrained local updates.
• QLQCD MC: Constrained graph-rewriting. An update represents a localized phason flip (a De Bruijn dual surface passing through a node).

The algorithm requires a Metropolis-Hastings acceptance step where the proposal distribution $Q(x \to x')$ includes a binary viability check: $\mathcal{P}_{accept} = \min \left( 1, e^{-\Delta S} \right) \times \prod_{v \in \text{update zone}} \Theta(v \in \mathcal{V}_{\text{AKN}})$ where $\Theta$ is an indicator function returning 1 if the updated vertex $v$ belongs to the set of legal AKN stars $\mathcal{V}_{\text{AKN}}$, and 0 otherwise. This enforces the $K_\perp$ restoring force at the algorithmic level, structurally forbidding "unphysical" vacuum states.

---

Z.4 Resolution of the Fermion Doubling Problem

A historic plague of Lattice QCD is the Fermion Doubling Problem. The Nielsen-Ninomiya theorem proves that any local, translationally invariant, hermitian discretization of fermions on a periodic lattice inevitably produces spurious "ghost" particles (2$^D$ fermions for every 1 intended fermion). Standard LQCD patches this by introducing "Wilson fermions," explicitly breaking chiral symmetry by adding an ad-hoc mass term that diverges in the continuum limit, forcing the ghosts infinitely massive.

QLQCD naturally bypasses the Nielsen-Ninomiya theorem.

Z.4.1 The Chiral Cut The underlying cause of fermion doubling is the discrete translation symmetry of the hypercubic lattice, which causes the momentum-space Dirac operator to have multiple zeros at the edges of the Brillouin zone ($p = \pi/a$).

By definition, the AKN tiling is generated by an irrational cut ($\lambda = \phi$). It is limit-quasiperiodic. It possesses no discrete translational symmetry. Therefore, there is no generic Brillouin zone, and the global periodicity required by the Nielsen-Ninomiya theorem is fundamentally broken at the lattice scale.

Z.4.2 Proof of Extinction Because the icosahedral projection is inherently chiral, the fermion propagator $S_F(x,y)$ lacks the periodic poles that generate doublers. The GCT Dirac operator $D_{GCT}$ is defined on the aperiodic grid: $\sum_{\mu=1}^6 \gamma^\mu \nabla_\mu^{(AKN)} \psi(x) = 0$ Because the discrete derivative $\nabla_\mu^{(AKN)}$ operates along the 6 aperiodic axes, the Fourier transform $\tilde{D}(p)$ only vanishes at the true physical origin $p = 0$. The 15 other parity-inverted momentum modes that plague hypercubic lattices ($\mathbf{p} \propto \pi$) do not map to continuous spectrum states in the quasicrystal.

Conclusion: GCT requires no Wilson mass term. Chiral symmetry is exactly preserved for massless fermions on the QLQCD lattice, solving the doubling problem via pure geometric necessity.

---

§Z.5 — Atiyah-Patodi-Singer Boundary Pathway for $\alpha$ [Tier 3 — Open Program]

Pathway, not closure. The canonical fine-structure mechanism in GCT is the bare baseline $\alpha^{-1}_{bare} = 360\phi^{-2}$ (V1 Ch13 §13.2), with the specific 360 multiplier and anti-screening magnitude retained as O.19/O.5 closure targets, refined by the $1/288$ bilayer correction (V1 Ch13 §13.2.4, Tier 2 Motivated Derivation), which gives $\alpha^{-1}_{GCT} = 137.0303$ with a 41.6 ppm residual against CODATA. The 41.6 ppm residual is the present QLQCD-1L target.

The present section sketches an alternative topological pathway: identifying the residual as the value of an Atiyah–Patodi–Singer (APS) $\eta$-invariant of the Dirac operator restricted to the boundary $\partial\mathcal{M}_{RT}$ of the Rhombic Triacontahedron. If the bosonic phason loop modes on the 30 rhombic faces of the RT cage carry the requisite spectral asymmetry, the $\eta$-invariant supplies the residual exactly and the $\alpha$ derivation becomes a topological boundary computation rather than a perturbative loop calculation. This is an open program: the manifold $\mathcal{M}_{RT}$ has not yet been equipped with the smooth-manifold-with-boundary structure required by the APS theorem, and the boundary integral has not been evaluated.

Z.6 The APS Boundary Pathway [Tier 3 — Open Program]

Z.6.1 Conjectural Identification. Under the hypothesis that the icosahedral spectral triple $(\mathcal{A}, \mathcal{H}, D_F)$ admits a smooth-manifold-with-boundary completion compatible with APS, the identification $\alpha^{-1}_{obs} \;\stackrel{?}{=}\; \alpha^{-1}_{bare} + \tfrac{1}{2}\eta\!\left(D_F|_{\partial\mathcal{M}_{RT}}\right)$ would close the electromagnetic sector without further perturbative input. The conjecture is consistent with the Tier 2 bilayer pathway only if the $\eta$-invariant evaluates to the bilayer-equivalent $\sim 41.6$ ppm shift after restriction to the bilayer cage; whether the two pathways coincide is itself an open question.

Z.6.2 The Chern–Simons Boundary Term. The secondary characteristic class associated with the conjectural $\eta$-invariant contribution is the Chern–Simons 3-form $\mathcal{CS}_3(\omega) = \operatorname{Tr}\!\left(\omega \wedge d\omega + \tfrac{2}{3}\omega \wedge \omega \wedge \omega\right)$ evaluated on $\partial\mathcal{M}_{RT}$ with the $H_3$-symmetric connection $\omega$. Were the integral $\delta\alpha^{-1} \;\stackrel{?}{=}\; \tfrac{1}{2}\int_{\partial\mathcal{M}_{RT}} \mathcal{CS}_3(\omega)$ to be both well-defined and equal to the observed residual, the program would close. As of this writing the integral has not been computed and the boundary structure has not been formalized; the section therefore stands as a research target, not as a derivation.

Z.7 QLQCD-2: Strong Coupling Bare Prediction

Status: TREE-LEVEL COMPUTED. Full QLQCD closure pending.

The Prediction: $\alpha_s^{-1}(\text{bare}) = 10\phi^2 \approx 26.18$, derived from the area-law of the 10 three-fold RT axes (V3 Ch04 §4.5). The geometric RGE has been extended (protocol_geometric_rge.py) to run this from the GUT boundary condition to $M_Z$, yielding a tree-level prediction of $\alpha_s(M_Z) \approx 0.0382$. This leaves a 67.6% gap relative to the PDG value ($0.1179 \pm 0.0010$).

Implementation Path: The tree-level geometric track has been implemented. The remaining task is to compute the full non-perturbative QLQCD-2 corrections to close the 67.6% gap.

Priority: LOW. Does not block current release certification but constitutes a natural and important extension of the geometric RGE program.