Modeling Nucleate Boiling in Engine Coolant Jacket: Analysis & Development of Semi-Mechanistic Boiling Methods in Heavy-Duty Diesel Engine Applications

Abstract

This thesis investigates subcooled nucleate boiling in a coolant jacket of a heavyduty diesel engine, using a Conjugate Heat Transfer - Computational Fluid Dynamics (CHT-CFD) framework. As modern engines operate at increasingly high thermal loads under strict efficiency and emissions requirements, an accurate prediction of local cooling performance is essential. In cast-iron heavy-duty diesel engines, regions exposed to high thermal loads may experience nucleate boiling, which can enhance heat transfer if controlled, but may also lead to deterioration of cooling performance if excessive boiling occurs.

The work focuses on the implementation, calibration and validation of the Blended Boiling Model (BBM) and an extended Dry-spot Blended Boiling Model (DBBM) in STAR-CCM+ for two heavy-duty diesel engine variants. The models are evaluated in both a simple academic framework and a complete engine simulation framework, both validated with relevant experimental data. Their performance is also compared with that of other boiling models available in STAR-CCM+.

The results show that both BBM and DBBM predict boiling heat flux with reasonably good accuracy. The BBM performs well in the partial boiling regime and provides a useful indication of boiling intensity with a parameter called probability of bubble interactions. However, under low pressure and low coolant mass flow conditions, where the occurrence of a transition from nucleate boiling to film boiling is evident in the experiments, the BBM underestimates the wall temperatures. By introducing a dry-spot probability, the DBBM extends the modeling capability into more severe boiling regimes and improves the prediction in regimes where excessive boiling is observed in the experiments.