Investigation of One-Dimensional Modeling Technique for Engine Thermal Systems

Zielinski, J. V. Investigation of One-Dimensional Modeling Technique for Engine Thermal Systems. University of Wisconsin-Madison, 2020.

Thermal management is a critical engine function that tends to be disregarded during the early design phase when there is limited information available. Auxiliary components often are added later to achieve the targeted steady-state uid and component temperatures. To meet the demand for higher power density an analytical tool that is capable of predicting thermal loads and steady-state temperatures is needed.

An integrated engine thermal model was developed with the intent of investigating energy transfer throughout the engine, and ultimately its dependence on various engine variables. The Mercury Marine 40 HP, 3-cylinder outboard engine was modeled using the GT-Suite one-dimensional modeling domain. The integrated GT model consisted of detailed submodels for each of the engine’s main subsystems: air-handling, combustion, cooling, and lubrication. Multi-objective optimization techniques and empirical correlations were used to calibrate the model’s numerous input parameters.

When compared to conjugate heat transfer (CHT) analysis data provided by Mercury for qualitative evaluation, the spatial temperature distributions output by the integrated GT model compared quite well; the locations of the maximum and minimum temperature on the engine components showed adequate agreement. At 6000 rpm, WOT, the integrated GT model’s prediction of steady-state water, oil, and exhaust gas temperatures were compared to Mercury data. A 0.1 C dierence between predicted and measured stead-state water and oil temperatures were observed, while the exhaust gas temperatures also exhibited acceptable agreement. Part-load tests were performed on the engine at UW-Madison for additional model validation at 1800, 2400, and 3000 rpm. The integrated GT model’s prediction of steady-state water temperature was between 0.2 and 5.2 C of the measured values, while the model’s prediction of steady-state oil temperature diered by between 0.2 and 5.0 C. A 12 C dierence was realized between the simulated and measured block metal temperatures, while the head metal temperature showed the largest discrepancy of all the components, at 33 C.

A sensitivity analysis was performed using the integrated GT model and showed that the heat transfer coecient associated with the underside of the piston had the largest inuence on the steady-state oil temperature. The same analysis was performed for the steady-state water temperature and indicated that the block water jacket heat transfer coecient had the largest inuence. Utilizing the integrated GT model, steady-state uid temperatures were successfully predicted, and the dependence of `hard-to-measure’ engine parameters was also examined.