Theoretical Frameworks for Estimating Internal Combustion Engine Power

Abstract

This review examines the principal theoretical routes used to estimate engine power without relying exclusively on full-scale dynamometer testing. The analysis spans ideal cycle formulations, zero-dimensional and quasi-dimensional combustion models, one-dimensional gas-exchange simulations, high-fidelity CFD, and emerging data-driven surrogates. Particular attention is given to how heat release, in-cylinder pressure evolution, volumetric efficiency, friction, pumping work, and fuel properties influence the prediction of indicated and brake output. Representative figures are used to compare modeling hierarchies, visualize fuel-energy partitioning, track pressure and heat-release phasing, map parameter sensitivity, and illustrate the dependence of brake power on engine speed and mean effective pressure. The review shows that low-order thermodynamic methods remain effective for screening and preliminary design, whereas combustion-resolved and CFD-based approaches are better suited to calibration, abnormal-combustion assessment, and fuel-flexibility studies. Machine-learning models can accelerate prediction, but their reliability improves when they are anchored to physics-based constraints. Overall, the most appropriate framework depends on the intended application, the level of fidelity required, the availability of input data, and the acceptable computational cost.