We introduce a geometric framework for dissipative dynamical systems whose limit cycles carry structured phase, multi-scale coherence, and stochastic stability. The central object is the dm³ system — a smooth Riemannian manifold equipped with a hyperbolic limit cycle, a Lyapunov function, and a stochastic extension — formalized through eight axioms. On this foundation we construct a complete operator algebra (the g-, L-, R-, and U-operators) governing expansion, coherence, resonance, and unification, and embed the entire structure in contact geometry.
Four main results are established. Theorem A: any dm³ system admits a unique exponentially stable hyperbolic limit cycle with a topological invariant (the winding integral) and stochastic stability below an embodiment threshold \(\tau\). Theorem B: the category dm³ is closed under a unification operator, with \(\tau_{12} \leq \min(\tau_1, \tau_2)\). Theorem C: every dm³ system near its limit cycle is locally contact-diffeomorphic to a universal three-parameter normal form \((\mu_{\max}, \omega, \beta)\). Theorem D: the full operator algebra is C¹-structurally stable with explicit stability radius \(\varepsilon_0 = |\mu_{\max}| / [2(1 + \sup\|\text{Hess}\,V\|)]\). Complete verification on an explicit toy model appears in the companion paper [2].
A wide class of natural systems — neural circuits, circadian clocks, cardiac oscillators, allostatic stress responses, plasma reconnection events — are governed by dissipative dynamics whose long-run behavior is not mere convergence to a fixed point but structured convergence to a periodic orbit with phase. The orbit encodes an organizing principle: it persists under noise, survives perturbation, and reestablishes itself after disruption. Classical dynamical systems theory provides Poincaré–Bendixson, normal hyperbolicity, and Floquet theory; classical stochastic analysis provides Fokker–Planck and Itô calculus. What has been lacking is a unified axiomatic framework that simultaneously accounts for the geometric structure of the orbit, the operator algebra governing the approach to it, and the stochastic stability threshold.
The Generative Contact Mechanics (GCM) framework proposed here fills this gap. The central observation is that contact geometry — specifically, the odd-dimensional counterpart of symplectic geometry introduced by Gibbs, developed by Eliashberg and Etnyre — is the natural home for dissipative limit-cycle systems. Liouville's theorem forbids attractors in compact symplectic systems; contact geometry overcomes this by adding a dissipation direction \(z\) whose evolution records accumulated phase. The contact form \(\alpha = dz - \lambda\) (with \(d\lambda = \omega\) the symplectic form of the reduced system) provides the invariant structure.
The paper is organized as follows. §2 defines the dm³ system axiomatically. §3 constructs the operator algebra. §4 embeds everything in contact geometry. §5 states and proves the four main theorems. §6 derives the explicit stability radius formula. The companion paper [2] provides complete numerical verification on an explicit three-dimensional toy model.
A dm³ system is a quintuple \((M, g, \Gamma, V, \xi)\) satisfying the following axioms.
On a dm³ system \((M, g, \Gamma, V, \xi)\) we define four canonical families of operators acting on smooth functions on \(M\) (or on trajectories).
| Operator | Symbol | Action | Geometric meaning |
|---|---|---|---|
| Compression | C: X → X_C | Reduces degrees of freedom; projects onto slow manifold | Singularity theory: rank drop before the fold |
| Curvature drive | K: κ ↑ κ* | Accumulates curvature \(|\kappa|\) to threshold \(\kappa^*\) | Geodesic focusing; Whitney A₁ precursor |
| Fold | F: rank(J)↓1 | Whitney A₁ fold — Jacobian loses one rank | Singularity of first kind; contact Hopf bifurcation |
| Unfolding | U: ∇Φ-flow | New topology emerges; post-fold stable branch | Thom–Mather unfolding; \(\Gamma\) is the stable post-fold orbit |
The operators satisfy an algebra over composition: \(G = U \circ F \circ K \circ C\) is one full generative cycle. The g-operators \(\{g_t\}_{t \geq 0}\) form a one-parameter semigroup (expansion); the L-operators are Lie-bracket derived from \([F, K]\); the R-operators are resonance selectors; and the U-operator is the unification map between two dm³ systems (Theorem B).
Given two dm³ systems \(X_1 = (M_1, g_1, \Gamma_1, V_1, \xi_1)\) and \(X_2 = (M_2, g_2, \Gamma_2, V_2, \xi_2)\), their unification is the dm³ system \(X_{12} = U(X_1, X_2)\) on \(M_1 \times M_2\) with contact form \(\alpha_{12} = \alpha_1 + \alpha_2\) and threshold \(\tau_{12} = \sqrt{\mu_1^2 + \mu_2^2} / \max(\kappa_1, \kappa_2)\).
The passage from symplectic to contact geometry is forced by a classical obstruction:
On a compact symplectic manifold \((N, \omega)\), every Hamiltonian flow preserves the Liouville measure. Hence no compact attractor (limit cycle or otherwise) can exist. Dissipative systems with attractors require a non-exact structure.
Contact geometry resolves this by passing from \((M, \omega)\) to the contact manifold \((\hat M, \alpha)\) where \(\hat M = M \times \mathbb{R}\) and \(\alpha = dz - \lambda\) (with \(d\lambda = \omega\)). The variable \(z\) records accumulated dissipation. The Reeb vector field of \(\alpha\) is \(\partial_z\), and orbits of the dm³ system spiral in the \(z\)-direction as they approach \(\Gamma\).
The contact normal form of Theorem C expresses every dm³ system near \(\Gamma\) in canonical coordinates as:
Let \((M, g, \Gamma, V, \xi)\) be a dm³ system. Then:
(i) \(\Gamma\) is the unique exponentially stable hyperbolic limit cycle in the Gronwall basin \(\mathcal{B}(\Gamma)\). The exponential rate is \(|\mu_{\max}|\).
(ii) The winding integral \(W(\Gamma) = \oint_\Gamma \lambda \in \mathbb{Z}\) is a topological invariant of the contact class of \(\alpha\).
(iii) For the stochastic extension with noise amplitude \(\sigma\), the stationary measure \(\rho_\sigma\) concentrates on \(\Gamma\) as \(\sigma \to 0\), and the system is stochastically stable for \(\sigma < \tau\), where \(\tau = \sqrt{c/\kappa_{\text{noise}}}\) is the embodiment threshold.
The category dm³ (with objects dm³ systems and morphisms contact maps) is closed under the unification operator \(U\). Moreover, the embodiment threshold satisfies:
Unification weakens noise tolerance: the threshold of the unified system does not exceed that of either component. Equality \(\tau_{12} = \min(\tau_1, \tau_2)\) holds iff the two Lyapunov functions are \(g\)-orthogonal.
Let \((M, g, \Gamma, V, \xi)\) be a dm³ system with parameters \((\mu_{\max}, \omega, \beta)\). Then there exists a neighborhood \(\mathcal{U}(\Gamma)\) and a local contact diffeomorphism \(\varphi: \mathcal{U}(\Gamma) \to \mathbb{R}^{2n+1}\) such that \(\varphi^*\alpha_{\text{normal}} = \alpha\), where the normal form contact field is
The three parameters \((\mu_{\max}, \omega, \beta)\) are contact invariants of \(\Gamma\): they classify the local contact geometry uniquely up to contact diffeomorphism.
The dm³ operator algebra \((G, U, F, K, C)\) is C¹-structurally stable: for any perturbation \(\tilde f\) with \(\|f - \tilde f\|_{C^1} < \varepsilon_0\), the perturbed system \(\tilde X\) is contact-diffeomorphic to \(X\). The explicit stability radius is
In the dm³ toy model: \(\mu_{\max} = -2\), \(\sup\|\text{Hess}\,V\| = 2\) on \(\mathcal{U}(\Gamma)\), giving \(\varepsilon_0 = 2/(2 \cdot 3) = 1/3\).
The stability radius \(\varepsilon_0\) has a direct geometric interpretation: it is the largest perturbation of the vector field that the dm³ system can absorb while maintaining its contact structure. The derivation proceeds via Gronwall's inequality applied to the transverse eigenvalue equation.
For initial conditions \(x_0\) with \(V(x_0) \leq V_0\), the orbit satisfies
for all \(t \geq 0\). The Gronwall basin is \(\mathcal{B}(\Gamma) = \{x : V(x) < \varepsilon_0^2 (1+\sup\|\text{Hess}\,V\|)^2\}\).
Note: the stability radius \(\varepsilon_0 = 1/3\) applies to the outer Gronwall basin. The inner basin (\(r < r_{\text{inner}}\) in the toy model) is governed by a separate asymmetry — the Gronwall asymmetry open problem documented in AXLE Issue #13 and listed as thm_gronwall_asymmetry (sorry ★★★, AXLE VolumeTwo.lean).
The GCM framework is the abstract backbone of the Principia Orthogona series. Its relation to the companion papers is as follows:
| Paper / Volume | DOI | GCM role |
|---|---|---|
| Vol I · GOMC — Mathematics of Generative Transitions | 19117400 | Establishes the operator chain G = U∘F∘K∘C on Riemannian manifolds; GCM axioms 1–4 are proved here |
| Vol II · TOGT — Contact Realization | 20159456 | Lifts to contact geometry; GCM Theorems A–C instantiated; Lean 4 skeleton |
| Toy Model · SIAM — Global Dynamical Analysis | 20230624 | Explicit verification of GCM Theorems A–D on exact ODE; ε₀ = 1/3 confirmed numerically |
| GOMC Opus — CatGT + Plasma + 30 Problems | 19117399 | Applications: zeolite catalysis (CatGT), MHD reconnection (plasma), 18-domain coherence bridge |
The toy model paper provides complete numerical verification. The dashboard gives live interactive exploration.
Toy Model → Dashboard → Vol II →