Discrete Topological Invariants and the Algebraic Frontier of Gauge Structure: Closure Saturation, Absence of Local Gauge Fields, and the 1/4 Area Coefficient
The conclusions of this paper are of two kinds, and the hard ones come first: on the horizon triangulation of a discrete graph the microscopic entropy is (3/2)F ln3, and once the gravitational coupling is derived independently the Bekenstein–Hawking coefficient comes out exactly 1/4 — the two expressions are two ways of writing the same number; the claim that can be shot down experimentally is the upper bound of 1.648 nats on the entropy density per cell. The remaining conclusions are labelled one by one on a four-level scale (proved, conditional, phenomenological, prediction). Only one geometric setting is used throughout: hoppings connect only sites a fixed step k apart, with k taken as 3 (the smallest step at which the underlying topology can carry a Z₃ centre phase). The first group is purely algebraic: from the three-dimensional representation of the third-order Heisenberg group a nine-dimensional tensor space is built, five initial path operators are iterated under commutators, and the closure grows round by round as 5, 10, 29, 80, ending at the 80-dimensional traceless algebra sl(9,C), which with the central element becomes the 81-dimensional u(9); the subalgebra generated by only two single-factor operators is sl(3,C) of complex dimension 8, which with the central element becomes gl(3,C) (complex dimension 9); this conclusion has been rechecked round by round in exact arithmetic over the rational field Q(ω). The second group is identities: for fixed-step hopping in the low-momentum expansion the ratio of the quartic to the quadratic term is (kℓ)²/12, that is k²/12; it is not a new β-function correction but the equivalence between "k-step hopping" and "coarse-graining the lattice spacing by a factor k", both giving the same effective spacing kℓ. The third group is phenomenological output: with two frozen inputs (mode count N=16 and matching factor gSW=1/2) the effective clover coefficient is 32/23≈1.391, with a test window of ±0.029 and a sensitivity band of [1.34,1.46], explicitly labelled phenomenological and not used as a first-principles prediction. The fourth group is conditional: given the Z₃ linear form of the shear operator and the definition of the shear field, the spatial cross-section eigenvalues are 0 and ±3, longitudinal exclusion forbids magnetic monopoles, and transverse closure gives Faraday induction and a light speed c=kℓ/Δt; given Z₃ phase closure and the (1,3) signature as two premises, the spatial dimension is locked to 3; given that the homology group is Z₃ and that there is no global anomaly, only three fermion generations are allowed. The fifth group is graph counting: the horizon triangulation satisfies 3F=2E and V−E+F=2, the number of cut edges across the boundary is 3F/2, and Z₃ edge labels give the microscopic entropy (3/2)F ln3. This paper states explicitly that 1/4 is an algebraic identity within this derivation, not an independent prediction, and that its finite-size correction is O(1/F); the genuinely falsifiable hard conclusion is the upper bound of 1.648 nats on the entropy density per cell. Whenever a conclusion is conditional, the premises it relies on are written out; whenever it is phenomenological, all inputs are listed.
Authors
- Wenjun Luo (ORCID: https://orcid.org/0009-0002-2082-5651)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-21
- DOI
- https://doi.org/10.5281/zenodo.24420845
- Primary Topic
- Noncommutative and Quantum Gravity Theories
- Type
- preprint