Geometric Formulation of Power-Efficiency Bounds in Carnot-like Engines
R. X. Zhai (翟若迅) · 2026
Read Geometric Formulation of Power-Efficiency Bounds in Carnot-like Engines by R. X. Zhai (翟若迅) (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-02732.txt.
Introduction
R. X. Zhai recasts a classic trade-off in heat engine design as a problem in plane geometry. Carnot-like engines face a linked limit on power and efficiency: pushing one tends to cut the other. Zhai plots normalized losses from the engine's two branches, and in that plane the curves of constant efficiency become straight lines.
With efficiency contours straight, the best efficiency at fixed power reduces to bounding the slope of an allowed line. When branch losses scale as a power of the process time, the fixed-power condition carves out a two-dimensional region, and the slope problem becomes a linear program. The geometric frame then yields the exact power-efficiency limit for engines with power-law losses. The paper is short and formal, aimed at readers in thermodynamics theory.
Catalog record
| CREATOR | R. X. Zhai (翟若迅) |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 18 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.02732
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Geometric Formulation of Power-Efficiency Bounds in Carnot-like Engines R. X. Zhai (翟若迅) Abstract We formulate the power-efficiency constraint of Carnot-like heat engines as a geometric optimization problem in the plane of normalized branch dissipations. Efficiency contours are straight lines in this plane, so maximizing efficiency at fixed power reduces to bounding the slope of an admissible line. We apply this framework to branch-resolved power-law dissipation, where the irreversible loss on each isothermal branch decays with the branch duration with a common exponent rather than following the standard inverse-time law. After optimizing over the dissipation-asymmetry parameter, the fixed-power attainable set becomes a two-dimensional region, and the resulting slope-bound problem reduces to linear programming. The framework yields the exact power-efficiency frontier within this model and gives closed-form constraints for representative dissipation exponents, including the maximum-power limit. I Introduction A Carnot heat engine operates between two reservoirs and converts part of the absorbed heat into work Huang 1987 ; Ma 2004 . Its efficiency is bounded by the Carnot efficiency, which is attained only in the quasistatic limit. In this limit the cycle time diverges, so the output power vanishes. Finite-time thermodynamics addresses this compromise between reversibility and useful power Salamon and Berry 1983 ; Andresen et al. 1984 . A central question is therefore the power-efficiency constraint: at a given power, what efficiency is attainable, and at a given efficiency, what power is attainable? The power-efficiency constraint is the direct extension of the familiar single-point benchmark set by the efficiency at maximum power. Early endoreversible models and later finite-time engines showed that maximum-power operation generally occurs below the Carnot efficiency Yvon 1955 ; Chambadal 1957 ; Novikov 1958 ; Curzon and Ahlborn 1975 ; Tu 2012 . This question has since been studied in linear-response and stochastic models, as well as in quantum, thermoelectric, and Rankine-type heat-engine settings Van den Broeck 2005 ; Schmiedl and Seifert 2008 ; Tu 2008 ; Izumida and Okuda 2008 ; Allahverdyan et al. 2008 ; Rutten et al. 2009 ; Sheng and Tu 2015 ; Brandner and Seifert 2015 ; Tu 2014 ; Martinez et al. 2015 ; Rossnagel et al. 2016 ; Josefsson et al. 2018 ; Josefsson et al. 2019 ; Liang et al. 2025 ; Cui and Dong 2025 . In the low-dissipation regime, bounds on the efficiency at maximum power were derived in Ref. Esposito et al. 2010 . For low-dissipation heat engines where the irreversible entropy production is proportional to 1/\tau , fixed-power bounds and universal constraints were developed in Refs. Holubec and Ryabov 2016 ; Ryabov and Holubec 2016 ; Ma et al. 2018 ; Zhai et al. 2026 , and related results have been obtained for steady-state, stochastic, minimally nonlinear, shortcut-assisted, chemical, and fluctuation-constrained engines Long and Liu 2016 ; Shiraishi et al. 2016 ; Holubec and Ryabov 2017 ; Pietzonka and Seifert 2018 ; Chen 2022 ; Zhao et al. 2022 ; Zhai and Dong 2025 . Experiments have tested both the low-dissipation scaling and the resulting power-efficiency trade-off Ma et al. 2020 ; Zhai et al. 2023 . Usually, the inverse-time low-dissipation law assumes that the irreversible entropy production of a finite-time isothermal branch is proportional to the inverse branch duration. When a driven system crosses a critical region or a first-order transition, irreversible entropy production, excess work, or hysteresis indicators can obey nonstandard finite-time scaling Deffner 2017 ; Wu et al. 2025 ; Chen et al. 2026 . When the system-bath coupling is controlled during an isothermal transformation, the total dissipation can decay with tunable powers of the protocol time Pancotti et al. 2020 . When memory kernels or relaxation laws are algebraic, finite-time corrections need not be governed by a single relaxation time Lutz 2001 ; Burov and Barkai 2008 . These examples motivate, without deriving microscopically, a branch-resolved power-law dissipation model. Finite-time engines beyond the inverse-time law have been studied through generalized dissipative Carnot-like models Wang and Tu 2012 ; Wang and Tu 2013 . A recent unified approach derived power-efficiency trade-off relations for generic thermal machines from a cycle-level power-law irreversibility Ma and Fu 2025 ; related finite-time refrigerators, heat pumps, and heat devices have also been optimized Wang et al. 2012 ; de Tomas et al. 2013 ; Holubec and Ye 2020 ; Ye and Holubec 2022 ; Ye and Holubec 2025 . The present work is complementary to these cycle-level approaches: it focuses on the branch-resolved attainable region and the geometry of its power-efficiency frontier. We do not introduce a new heat-engine mechanism. Instead, we give a geometric method for Carnot-like engines with branch-resolved power-law dissipation. The normalized irreversible entropy productions on the hot and cold isothermal branches are used as coordinates (\sigma_{h},\sigma_{c}) . At fixed normalized power, the admissible protocols form a set in this plane. Efficiency level curves are straight lines, so maximizing efficiency at fixed power becomes a problem of bounding the slope of an admissible line. This formulation also clarifies the role of the time-allocation asymmetry parameter \xi . If \xi is fixed, a prescribed normalized power selects a one-dimensional closed curve in the (\sigma_{h},\sigma_{c}) plane, and the efficiency bound follows from tangency with an efficiency line. For the low-dissipation case, this fixed- \xi construction corresponds to the exact bound of Ref. Zhai et al. 2026 . Here \xi is optimized as well, so the family of fixed- \xi curves projects onto a two-dimensional attainable region. For power-law dissipation, the resulting slope-bound problem reduces to linear programming. This yields the exact power-efficiency frontier within the model and gives closed-form constraints for representative dissipation exponents, including the maximum-power limit. II Geometry of Power-Efficiency Constraints We consider a Carnot-like cycle operating between two heat reservoirs at temperatures T_{h}>T_{c} . The cycle consists of two adiabatic branches and two finite-time isothermal branches. During the finite-time contacts of durations \tau_{h} and \tau_{c} , the heats exchanged with the hot and cold reservoirs are Q_{h,c}=\pm T_{h,c}\Delta S^{\text{(re)}}-Q_{h,c}^{(\text{ir})}(\tau_{h,c}) . Here, \Delta S^{(\text{re})}\geq 0 is the isothermal entropy change in the corresponding reversible cycle. We assume that the irreversible heat transfer scales with \tau as a power law, Q_{h,c}^{\text{(ir)}}=M_{h,c}\tau_{h,c}^{-\alpha} . For \alpha=1 , this recovers the well-studied low-dissipation case. The power and efficiency are defined as P=\frac{(T_{h}-T_{c})\Delta S^{\text{(re)}}-M_{h}\tau_{h}^{-\alpha}-M_{c}\tau_{c}^{-\alpha}}{\tau_{h}+\tau_{c}}, (1) and \eta=\frac{(T_{h}-T_{c})\Delta S^{\text{(re)}}-M_{h}\tau_{h}^{-\alpha}-M_{c}\tau_{c}^{-\alpha}}{T_{h}\Delta S^{\text{(re)}}-M_{h}\tau_{h}^{-\alpha}}. (2) We now rewrite these quantities in terms of three dimensionless parameters, \xi , \sigma_{h} , and \sigma_{c} . The variables \sigma_{h} and \sigma_{c} are the dimensionless irreversible heats for the finite-time isothermal processes, defined as \sigma_{h,c}\equiv\quantity( M_{h,c} / T_h \Delta S^{\text{(re)}} )\tau_{h,c}^{-\alpha}\geq 0 . The variable 0\leq\xi\leq 1 , defined by \xi\equiv M_{h}^{1/(1+\alpha)}/\quantity(M_h^{1/(1+\alpha)} + M_c^{1/(1+\alpha)}) , characterizes the dissipation asymmetry between the two isothermal branches. For later convenience, we set \epsilon\equiv 1/\alpha . In these variables, the efficiency becomes \eta=\frac{\eta_{C}-\sigma_{h}-\sigma_{c}}{1-\sigma_{h}}. (3) By normalizing the power by its maximum with respect to \sigma_{h} and \sigma_{c} , i.e., \tilde{P}\equiv P/P_{\max} , we obtain \tilde{P}(\sigma_{h},\sigma_{c},\xi)=\frac{(1+\epsilon)^{1+\epsilon}}{\epsilon^{\epsilon}\eta_{C}^{1+\epsilon}}\frac{\quantity(\eta_C - \sigma_h - \sigma_c)\sigma_{h}^{\epsilon}\sigma_{c}^{\epsilon}}{(1-\xi)^{1+\epsilon}\sigma_{h}^{\epsilon}+\xi^{1+\epsilon}\sigma_{c}^{\epsilon}}. (4) For heat-engine operation, the positive-work condition requires \eta_{C}-\sigma_{h}-\sigma_{c}>0 . With the above notation, the mathematical problem of determining the power-efficiency constraint becomes the following: under the constraint that the function in Eq. ( 4 ) takes the value \tilde{P}=\tilde{P}_{0} , find the maximum and minimum values of the efficiency \eta in Eq. (