[Paper Review] Chaotic Hypothesis and Universal Large Deviations Properties

Giovanni Gallavotti|ArXiv.org|Aug 5, 1998

Computational Physics and Python Applications4 citations

TL;DR

This paper proposes the Chaotic Hypothesis as a unifying framework for understanding nonequilibrium statistical mechanics in chaotic systems, particularly smooth hyperbolic systems. By assuming chaoticity and hyperbolicity, it derives universal large deviations properties—specifically, a fluctuation theorem for entropy production—without relying on specific models, establishing a general principle applicable across diverse physical systems.

ABSTRACT

Chaotic systems arise naturally in Statistical Mechanics and in Fluid Dynamics. A paradigm for their modelization are smooth hyperbolic systems. Are there consequences that can be drawn simply by assuming that a system is hyperbolic? here we present a few model independent general consequences which may have some relevance for the Physics of chaotic systems. Expanded version of a talk at ICM98, Berlin.

Motivation & Objective

• To establish general, model-independent consequences of assuming chaoticity in physical systems, particularly hyperbolic dynamics.
• To investigate whether universal large deviations properties can be derived from the chaotic hypothesis alone.
• To provide a theoretical foundation for nonequilibrium statistical mechanics applicable to systems like fluids and statistical mechanical models.
• To extend the fluctuation theorem to a broader class of chaotic systems beyond specific solvable models.
• To unify the description of irreversible processes in chaotic systems through a single, general principle based on hyperbolicity.

Proposed method

• Adopt the Chaotic Hypothesis as a foundational assumption: that a system behaves as if it were transitive and uniformly hyperbolic.
• Use the formalism of smooth hyperbolic dynamical systems to analyze time-averaged observables, particularly entropy production.
• Apply large deviations theory to derive the probability distribution of time-averaged entropy production in such systems.
• Leverage the existence of a Sinai-Ruelle-Bowen (SRB) measure to define statistical ensembles and compute probabilities of rare events.
• Derive a symmetry in the large deviation function of entropy production, leading to the fluctuation theorem.
• Use the invariance of the SRB measure under time reversal and the structure of the phase space to prove the symmetry without assuming time-reversal invariance of the dynamics.

Experimental results

Research questions

• RQ1 Can universal large deviations properties be derived from the chaotic hypothesis without reference to specific models?
• RQ2 What is the role of hyperbolicity in determining the statistics of time-averaged observables in nonequilibrium systems?
• RQ3 Does the fluctuation theorem for entropy production emerge universally from the chaotic hypothesis in smooth hyperbolic systems?
• RQ4 How does the SRB measure facilitate the derivation of large deviation functions in chaotic systems?
• RQ5 Can the symmetry of the large deviation function for entropy production be proven using only the assumptions of hyperbolicity and ergodicity?

Key findings

• The paper establishes that under the Chaotic Hypothesis, the large deviation function for entropy production in smooth hyperbolic systems exhibits a symmetry that implies the fluctuation theorem.
• The fluctuation theorem is derived as a general consequence of hyperbolicity and the existence of an SRB measure, not from specific dynamical assumptions.
• The symmetry of the large deviation function implies that the probability of observing a negative entropy production is exponentially suppressed, but not zero, in agreement with the second law on average.
• The result is universal: it applies to all smooth hyperbolic systems, including those modeling fluids and statistical mechanical systems.
• The derivation does not require time-reversal invariance of the dynamics, only the existence of an SRB measure and the chaotic hypothesis.
• The paper confirms that the fluctuation theorem is not a feature of specific models but a general property of chaotic systems with hyperbolic structure.