Gambling Carnot Engine
arXiv:2409.17212 · doi:10.1103/w8cx-xx1z
Abstract
We propose a theoretical model for a colloidal heat engine driven by a feedback protocol that is able to fully convert the net heat absorbed by the hot bath into extracted work. The feedback protocol, inspired by gambling strategies, executes a sudden quench at zero work cost when the particle position satisfies a specific first-passage condition. As a result, the engine enhances both power and efficiency with respect to a standard Carnot cycle, surpassing Carnot's efficiency at maximum power. Using first-passage and martingale theory, we derive analytical expressions for the power and efficiency far beyond the quasistatic limit and provide scaling arguments for their dependency with the cycle duration. Numerical simulations are in perfect agreement with our theoretical findings, and illustrate the impact of the data acquisition rate on the engine's performance.
20 pages (including supplemental material), 8 figures (including supplemental figures)
References in corpus (13)
- Efficiency at maximum power: An analytically solvable model for stochastic heat engines
- Nonequilibrium Detailed Fluctuation Theorem for Repeated Discrete Feedback
- Second-law-like inequalities with information and their interpretations
- Large work extraction and the Landauer limit in a continuous Maxwell demon
- Adiabatic processes realized with a trapped Brownian particle
- Maximizing power and velocity of an information engine
- Information engine in a nonequilibrium bath
- Martingales for physicists: A treatise on stochastic thermodynamics and beyond
- Thermodynamics of computations with absolute irreversibility, unidirectional transitions, and stochastic computation times
- Efficiency at maximum power of a Carnot quantum information engine
- Dissipation reduction and information-to-measurement conversion in DNA pulling experiments with feedback protocols
- Survival and extreme statistics of work, heat, and entropy production in steady-state heat engines
- Performance limits of information engines