Mixed Relaxation Phase Dynamics

• Mixed relaxation phase is a regime where multiple relaxation behaviors coexist, influencing the system's evolution from early oscillatory modes to eventual equilibrium.
• It is modeled across physical, chaotic, quantum, plasma, and optimization systems, each with distinct signatures such as power-law decay, two-step exponential relaxation, and hybrid algorithms.
• Diagnostic methods leverage velocity correlations, survival probabilities, and order parameters to capture the complex, non-exponential dynamics of mixed relaxation.

A mixed relaxation phase refers to an intermediate, nontrivial regime of relaxation dynamics in which more than one relaxation behavior, pattern, or mechanism coexists or competes, either in time, in space, or in the configuration/phase space of the system. Across physical, mathematical, and computational contexts, the mixed relaxation phase typically arises when the system’s evolution is governed by the interplay of multiple metastable modes, competing patterns, or intertwined dynamical channels. Mixed relaxation regimes underpin complex approaches to equilibrium, anomalous decay laws, and rich spatiotemporal organizations that defy simple exponential or uniform dynamical descriptions.

1. Physical Models and Prototypical Emergence

In finite, dissipative chains of interacting particles embedded in periodic potentials, the mixed relaxation phase manifests as a transient coexistence of in-phase and π-out-of-phase oscillation domains. The model is formalized by the equations

$$m\ddot{x}_i + \gamma \dot{x}_i + V'(x_i) + K[2x_i - x_{i+1} - x_{i-1}] = 0,$$

with viscous dissipation γ, spring constant K, and on-site periodic potential $V(x)$. Throughout relaxation:

• Early times feature in-phase ("PC") collective oscillations, dominated by the center-of-mass (CM) mode.
• An intermediate regime exhibits domains of in-phase and antiphase ("AC") oscillations, both with significant amplitudes, and cross-correlations near zero: the definitive mixed relaxation phase.
• Ultimately, the system settles into an antiphase oscillation dominated by the π-mode.

Parameter windows for the persistence of this mixed phase require moderate damping, strong mode nonlinearity, and sufficient initial energy. The mixed regime persists over a scaling window $\tau_{\text{mix}}\propto\gamma^{-1}$, with the amplitude of π-patterns growing as $$A_\pi(t)\sim [1-e^{-\alpha E_{\text{init}} t}]^{1/2}$$ and spatial domains characterized by coexisting synchronous and checkerboard-like regions (Liebchen et al., 2014).

2. Mixed Relaxation in Chaotic and Quantum Systems

In deterministic or Hamiltonian systems with mixed ("divided") phase space—regions of chaos and regularity—a mixed relaxation phase is the algebraic-decay regime of survival (return) probability,

$F(t)\sim t^{-\gamma_0}h(\ln t),$

rather than the exponential decay typical of fully chaotic systems. The universal exponent $\gamma_0$ emerges from the hierarchical island structure. Numerical studies (e.g., via the Meiss–Ott tree model) reveal that the apparent power-law exponent extracted from finite-duration simulations varies widely, with extremely slow (oscillatory, log-periodic) convergence to the asymptotic value, governed by ensemble sample-to-sample fluctuations that decay as $t^{-\epsilon}$, $\epsilon\ll1$ (Ceder et al., 2012).

In quantum many-body systems (driven chains, Floquet systems), mixed (two-step) relaxation describes scenarios where, for a local observable such as bipartite purity $I(t)$, the relaxation proceeds by two exponentials: $$I(t)-I_\infty \simeq \begin{cases} Ae^{-\Gamma_\text{init} t}, & t\ll t_c \ Be^{-\Gamma_\text{asymp} t}, & t\gg t_c \end{cases}$$ with $t_c$ set by the system size or entanglement saturation time. The initial decay is kinematically constrained (by gate connectivity), while the late regime is set by the transfer-matrix spectral gap. Crossovers can involve slower-than-expected ("phantom") relaxation at early times, with clear implications for information scrambling and ergodicity (Znidaric, 2023). In variationally projected quantum dynamics, mixed phase space structure (regular islands, chaotic sea) yields robust long-time revivals (quantum scars) or strongly state-dependent thermalization rates—a quantum analog of Kolmogorov–Arnold–Moser theory (Michailidis et al., 2019).

3. Mixed Relaxation Phases in Plasma and Kinetic Models

In plasma physics, the relaxation of broad beams of energetic particles to equilibrium under weak turbulence is traditionally modeled by quasilinear diffusion. The mixed diffusive–convective relaxation phase is identified when beam–wave nonlinearities enable a convective ("ballistic-like") regime on intermediate time scales: $$\frac{\partial f}{\partial t} + v\frac{\partial f}{\partial x} = \frac{\partial}{\partial v}\left[ D_{QL}\frac{\partial f}{\partial v} + V_{drag}f \right]$$ Here, diffusion dominates for $t\gg\tau_{\text{diff}}$, convection (drag) emerges for $\tau_{\text{conv}}\ll t\ll\tau_{\text{diff}}$. The coexistence or competition of these phases (enabled by strong resonance overlap, nonlinearity, excitation of stable modes) determines the transport characteristics and plateau formation in velocity space (Carlevaro et al., 2015).

4. Mixed Relaxation Algorithms in Optimization

In mathematical programming, the mixed relaxation phase constitutes the initial phase of a hybrid framework for large-scale binary MIPs or MIQPs. The mixed relaxation phase combines three convexification strategies: linear relaxation (variables $b\in[0,1]$), Lagrangian dual relaxation (introducing multipliers $\lambda$), and augmented Lagrangian relaxation (with quadratic penalties). Randomized sampling of relaxation parameters generates a diverse pool of near-feasible binary patterns that then serve as high-quality seeds for downstream metaheuristics (genetic algorithms, VNS). The coverage and quality of this pre-solution pool directly influence the convergence and final solution quality, outperforming any single relaxation pipeline (Wang, 31 Jan 2026).

5. Mixed Relaxation in Multiphase and Phase-Transition Models

In multiphase fluid models, the mixed relaxation phase typically refers to a computational strategy in which each phase is coupled to an auxiliary relaxation dynamics—usually a Suliciu-type pseudo-pressure evolution—designed to linearize the fluxes and guarantee entropy stability and positive volume fractions. The resulting system is updated using explicit fixed-point solvers for contact speeds and wave structures, yielding robust, oscillation-free solutions even when phase fractions vanish. The relaxation ensures that nonlinear equations of state are approached dynamically, while the interfacial coupling encodes the mixed-phase nature of the system (Saleh, 2018).

In phase-transition models (liquid-vapor), relaxation dynamics enforces the free-energy minimization towards stable, metastable, or coexistence equilibria. Here, the mixed relaxation phase can refer both to the coexistence region in the (p, ρ)-plane and the numerically/time-resolved mass transfer driven by the gradient of the (nonconvex) Helmholtz free energy. The system’s dynamical basin structure determines whether a metastable state persists or whether phase transfer proceeds to a stable mixture (James et al., 2015).

6. Diagnostics, Methods, and Theoretical Insights

Characterization of mixed relaxation phases relies on tailored order parameters, such as velocity cross-correlations, Fourier-based phase coherence measures, survival probabilities, entanglement spectra, or functionals over populations in phase space. Analytic rate equations, scaling laws, and fixed-point theorems underpin predictions of phase boundaries, lifetimes, and amplitude growth.

A recurring theme is the presence of strong sample-to-sample or region-to-region fluctuations, slow algebraic convergence to asymptotia, and log-periodic modulations or kinks at phase transitions, signaling the underlying hierarchical or spatially heterogeneous organization.

The theoretical significance of mixed relaxation phases is substantial: they challenge the universality of simple exponential decay, force a re-examination of thermalization and ergodic hypotheses, and demand robust computational frameworks capable of capturing both local and global relaxation modes.

7. Context, Universality, and Applications

The mixed relaxation phase is a unifying concept across a spectrum of physical, mathematical, and algorithmic settings:

Context	Mixed Relaxation Mechanism	Diagnostic/Signature
Particle chains in cosines	In-phase/antiphase domain coexistence	$K_v(t)$, $|Z_0|, |Z_\pi|$
Hamiltonian chaos	Power-law decay, hierarchical trapping	Survival $F(t)\sim t^{-\gamma_0}$
Quantum Floquet circuits	Two-step decay, rank constraints	Purity/OTOC exponent crossover
Plasma kinetic relaxation	Diffusion-convection competition	$\langle \delta v^2\rangle, f(v)$
Mixed-integer programming	Multiple relaxations, sampling diversity	Hamming gaps, solution pool metrics
Multiphase flow	Pseudo-pressure relaxation, phase interface	Positivity, discrete energy law

This concept provides a rigorous language for describing intermediate, metastable, or nontrivial dynamical regimes characterized by coexistence, competition, or strong fluctuation-induced effects. Mixed relaxation phases often delineate boundaries of method validity, indicate where universal scaling breaks down, and inform both numerical and analytical strategies for probing complex systems.

References:

• Liebchen & Schmelcher, "Spatiotemporal Oscillation Patterns in the Collective Relaxation Dynamics of Interacting Particles in Periodic Potentials" (Liebchen et al., 2014)
• Ceder & Agam, "Fluctuations in the relaxation dynamics of mixed chaotic systems" (Ceder et al., 2012)
• Carlevaro et al., "Mixed diffusive-convective relaxation of a broad beam of energetic particles in cold plasma" (Carlevaro et al., 2015)
• Xu et al., "A Hybrid Relaxation-Heuristic Framework for Solving MIP with Binary Variables" (Wang, 31 Jan 2026)
• Ljubotina et al., "Two-step relaxation in local many-body Floquet systems" (Znidaric, 2023)
• Michailidis et al., "Slow quantum thermalization and many-body revivals from mixed phase space" (Michailidis et al., 2019)
• James & Mathis, "A relaxation model for liquid-vapor phase change with metastability" (James et al., 2015)
• Coquel et al., "A relaxation scheme for a hyperbolic multiphase flow model. Part I: barotropic eos" (Saleh, 2018)