Meeting increasingly stringent targets for vehicle performance, economy and emissions requires a deep understanding of the overall IC engine system behavior and the ability to optimize it considering all control and noise factors and their variations. The tradeoffs in exhaust gas temperature, exhaust valve temperature, engine performance, economy and emissions demand a combination of capable CAE analytical tools and a methodology capable of leading the design to a reliable and robust solution. This paper presents a newly developed methodology that uses a Ford in-house quasi-dimensional combustion model called GESIM (General Engine Simulation Program) and a 3D conjugate heat transfer (CHT) model to predict crank angle resolved exhaust gas temperatures and cycle average valve temperatures in a 6-Sigma context, which considers a wide range of engine factors and their variations, to determine a feasible robust design solution. Engine design factors used in this study are: air temperature, compression ratio, air-to-fuel ratio, spark timing, exhaust valve opening (EVO), exhaust valve closing (EVC), intake cam timing, camshaft dynamics and valve seat area. The engine responses considered in this study are: exhaust gas temperature (EGT), engine power, BSFC, BSCO, BSNOx, BSHC, maximum exhaust valve temperature (EVT), and exhaust flange temperature (EFT). The models are capable of predicting the effect of engine design and operating condition changes and were calibrated to actual engine data. Main effects on the exhaust valve temperature, engine performance, economy and emissions were determined and transfer functions were created considering all the factors and their interactions. These transfer functions were found to be extremely useful to the engine development team in providing quick data, resulting in fast and reliable decision making process.
The CAE predictions are in good agreement with the actual measured data. The methodology was also validated for different engine architectures and operating conditions. In this paper, we present a full analytical robustness and reliability technique in which the variations in the system factors and responses are calculated and the main contributing factors to these variations are identified. The final design was determined through a shrink and shift process (Reliability and Robustness analysis) and optimized with real life system constraints and limitations.