In order to simulate the working process, an accurate description of heat transfer occurring between in-cylinder gases and combustion chamber walls is required, especially regarding thermal efficiency, combustion and emissions, or cooling strategies.
Combustion chamber wall heat transfer models are dominated by zero-dimensional semi-empirical models due to their good compromise between accuracy, complexity and computational efficiency. Classic models such as those from Woschni, Annand or Hohenberg are still widely used, despite having been developed on rather ancient engines. While numerous authors have worked on this topic in the past decades, little information can be found concerning the systematic calibration process of heat transfer models.
In this paper, a systematic calibration method based on experimental data processing is tested on the complete operating map of a turbocharged GDI engine. This method mainly relies on heat release rate calculations using in-cylinder pressure during the compression stroke or/and after the end of the combustion. Practical sub-models have been chosen for wall temperatures, trapped mass and residual gas fractions estimations. The described method is tested by calibrating two simple wall heat losses models, the Hohenberg model and the reduced Woschni model by defining calibration objectives on compression and after the end of combustion. The influence of calibration objectives over simulated results is then analyzed. While tested models behave properly when applied to single operating points on either of the calibration objectives, they do not provide accurate results for both objectives at the same time. Moreover, such simple models cannot cover the whole engine operating map with a unique set of calibration parameters for the compression objective.