A numerical methodology for the thermal assessment of high performance engines addressing the boiling phenomenon

Internal combustion engines (ICEs) are machines that convert the thermal energy produced by the combustion of the fuel in mechanical energy. This process is inherently inefficient and a significant portion of energy is lost as heat, thus an effective cooling system is needed in order to dissipate it. In high-performance engines, thermal and mechanical loads act together and lead to thermomechanical fatigue phenomena, making thermal management a critical factor to consider. Hence, designing cooling systems for such engines is a complex task, since the fluid temperature often approaches saturation, creating conditions that favour the formation of vapor films in the cooling jacket, hindering the heat transfer. Currently, the most advanced method for designing water jackets in ICEs is Computational Fluid Dynamics (CFD), which can simulate both fluid dynamics and the potential formation of vapor bubbles or films. Despite offering high accuracy, CFD comes with the significant drawback of requiring substantial computational resources and extended simulation times. This thesis presents a methodology that uses empirical correlations to consider the fluid-wall heat transfer coefficient (HTC) variation due to nucleate boiling phenomena within thermal Finite Element Analysis (FEA). This approach enables early detection of boiling phenomena in the engine head design, where complex biphase CFD simulations are too computationally expensive and time-consuming, allowing improved thermal predictions. A review of the relevant literature identifies the most suitable empirical boiling correlations, balancing accuracy and simplicity. These correlations are then implemented within the FEA software through the use of user subroutines, specifically utilizing Dassault Systems Abaqus. Finally, this approach is used to optimise the thermal simulation of the FEA model of a real high-performance ICE, with the objective of reaching a correlation between the simulated data and the real data obtained at the test bench.