Control of the flow of thermal energy in an air-cooled engine is important to the overall performance of the engine because of potential effects on engine performance, durability, design, and emissions. A methodology is being developed for the assessment of thermal flows in air-cooled engines, which includes the use of cycle simulation and in-cylinder heat flux measurements. The mechanism for the combination of cycle simulation, the measurement of in-cylinder heat flux and wall temperatures, and comparison of predicted and measured heat flux in the methodology is presented.The methodology consists of both simulation and experimental phases. To begin, a one-dimensional gas dynamics code (WAVE) has been used in conjunction with a detailed in-cylinder flow and combustion model (IRIS) in order to simulate engine operation in a variety of operating conditions. The methods used to apply the model to the air-cooled engine case are described in detail. Sensitivity studies have been performed demonstrating the effects of unknown parameters of importance to the model predictions, but not available from experiments. After the engine model has been tuned to produce the measured pressure history, it is then used to predict the in-cylinder heat flux.Experimentally, a thermopile-based heat flux sensor has been used to measure in-cylinder surface heat flux and surface temperatures. Cycle-resolved heat flux and surface temperature measurements have been obtained over a range of air/fuel ratios. It is observed that the individual cycle heat flux varies substantially from one cycle to the next. The amount of variability in the instantaneous flux was a strong function of air/fuel ratio, with the least variability occurring near the peak power point. Peak wall temperatures were observed at the peak power point (298.5°C), with some reduction further on the rich side of stoichiometric, and substantial reductions on the lean side (>279.8°C at A/F = 15.75.)Comparison of the model predictions of the instantaneous in-cylinder heat flux with that measured shows that the measured heat flux tends to peak earlier in the burn, and the predicted value for the integrated heat flux over the cycle is about 15% higher than that measured.