Improving engine cooling performance requires sophisticated and intelligent engine cooling design especially when interactions of all engine parts are to be considered. The cooling system would highly influence engine thermal efficiency, durability and engine design criteria. Several attempts have been made by engine designers to improve the cooling design during the past decades, each with particular purpose considerations.
In this paper, based on the cylinder head flame face of an existing heavy duty medium speed diesel engine, three other flame face cooling systems are designed, modeled and changes are implemented using a three dimensional computer aided design modeling software. Modeled cylinder head flame face cooling concepts are experiencing the effect of cooling passages geometry changes on performance of thermal efficiency, effective subcooled regions and other resultant factors.
A detailed coupled computational fluid dynamic and thermal finite element analysis for one cylinder bank assembly is performed several times; paying special attention to the risky areas to get comparative results to assess the flame face cooling designs. Engine specifications and loading conditions together with the engine performance data are available from test rig. Initial and boundary conditions have been determined through a global model simulation and analysis.
The subcooled nucleate boiling heat transfer computation is carried out using the boiling departure lift-off model. In order to obtain the temperature for components under consideration, a comprehensive thermal analysis has been performed coupling with the detailed CFD analysis to reach an accepted value through transferring data between the CFD and FEA software. This method leads to an accurate prediction of the wall temperature and heat flux.
It is observed that proper cooling design could improve wall temperature and thermal stress related phenomena significantly. The advantages and disadvantages of each concept are discussed and preferred flame face design is demonstrated. Calculated results of original design are validated with test cell records.
