Integrated robust design using computer experiments and optimization of a diesel HPCR injector

Robust design has been gaining plenty of attention in both academia and industry over the past two and half decades. However, there is still plenty of room for improvement. In this dissertation research, several drawbacks of the existing robust design methods were identified and analyzed. These shortcomings mainly include incompatible robust parameter design and tolerance design optima, inadequate attention to internal noise factors, limitation of Steepest Descent search in parameter design process, limitation of robust design using computer experiments, limitation with the Loss Function, and insufficient connection between the academia and industry on robust design.An integrated robust design framework was proposed to help achieve robustness against both external and internal noises. A Steepest Descent Driven Parameter Design method was developed to enhance parameter design. Also developed was an improved loss function for robust tolerance design.The proposed methodology was applied using computer simulation experiments to optimize the design of a state-of-the-art injector in a High Pressure Common Rail (HPCR) injection system in a diesel engine. The injector simulation model was built using Advanced Continuous Simulation Language (ACSL) and well calibrated through extensive hardware experiments. iSIGHT, an integration and optimization software package, was utilized to assist the computer experiments. Design Experts and Minitab were used to conduct statistical designs and analysis.It was demonstrated that the injector performance was significantly improved with the proposed method. The proposed method was also compared with Taguchi's method and Response Surface Methodology based robust design method.