Research On Multi-level Optimization Design Of Railway Vehicle And Its Typical Applications

China's railway system is being challenged by higher speed and heavier load which brings great opportunities to the locomotive and rolling stock industry. The structural design of the railway vehicles has become one of the key technology development areas. Railway vehicles are products of multi-disciplinary technologies with lots of coupling, it is particular more complex for those EMUs at speed 200 km/h and higher. The development of China's high-speed railway vehicles not only benefited from the experience gained from China's railway speed increases over the years but also the technology advancements throughout the world, however, lots of improvement opportunities remain. The traditional trial and error approach is still practiced widely in design processes, and the concept of structural optimization design is only in its early adoption. This thesis studied in detail the railway vehicle's structural optimization design and its applications. Structural optimization design is applied to various stages of the design process such as the topological optimization in conceptual design, the shape optimization in detailed design, and the multidisciplinary optimization in the integration design. Multi-level optimization is applied in complex engineering design and the overall performance of the complex engineering products is optimized. With its product quality enhancement, short development cycle, and design efficiency improvement, multi-level, multi-disciplinary structural optimization design is the future of railway vehicle design.This thesis reviewed railway vehicle's structural optimization design and discussed its recent developments throughout the world. The research work took a multi-level approach, targeted railway vehicle's structural optimization design, covered topological optimization of conceptual design, shape (size) optimization of detailed design and multidisciplinary optimization of integration design. The thesis also discussed in detail the optimization methodology and strategy, and provided examples and typical applications. The main research areas of the thesis are(1) Detailed discussion of the structural optimization theory, summary of the main structural design algorithms. Detailed discussion of the methodologies and principles for those commonly used engineering optimization algorithms and multidisciplinary optimization.(2) Discussion of the optimization strategy using approximate model for efficiency improvement of complex structural designs. Introduction of Design of Experiment (DOE), response surface approximate model, Kriging approximate model, RBF approximate model, and the Taylor sequence model. Analysis of evaluation criteria for approximate model accuracy and comparison of RSM, Kriging and RBF models through examples.(3) Topological optimization design of the bogie pivoted arm using variable densities. Topological optimization results were ensured by strength test using detailed finite element model with contact relationships accounted, taking into consideration of stress constraint issues and convenience of operation for a real complex structure. The optimization of the structural conceptual design for railway vehicle is made possible.(4) Discussion of the principles of displacement sensitivity and stress sensitivity. Finite element model of a high-speed aluminum alloy body. Finite element analysis using specifications; Using plate thickness as design variables, calculation of vehicle body structure's displacement sensitivity and stress sensitivity to the design variables in an finite element analysis. The structure weight lightening is achieved using the calculated sensitivities. This chapter provides a good example of fast and optimized design of complex structure using sensitivity information.(5) Multidisciplinary optimization study of the blades of the locomotive diesel engine turbocharger compressor. Three-dimensional parametric model was built for the blades. The most optimized results were obtained by multidisciplinary integrated design on structure and vibration frequency, using blade thickness at different cross-section as design variables, without sacrificing aerodynamic performance of the blades. This result also has typical model role。(6) Multidisciplinary optimization study of welded structure's reliability upon fatigue. Discussion of Goodman fatigue's safety coefficient, linear cumulative damage theory, and the international welded joint fatigue evaluation criteria (IIW, BS). Virtual Fatigue Test(VFT) technique is outlined, fatigue test on a Virtual Prototype, predicting designed fatigue life for the products; Multidisciplinary Feasible(MDF) method based on approximation model is proposed, with which optimization efficiency is greatly improved. Optimization model of welded bogie frame is created with considering welded joints fatigue damage and stress constrains. Developed program to calculate Goodman fatigue safety coefficient and welded joints fatigue damage. Integration of multiple programs and automatic optimization are achieved for welded frame fatigue damage and structural analysis. The weight of the welded frame was reduced by 11.6%, using Multidisciplinary Feasible (MDF) method based on approximation model, while satisfying the requirements of stress, fatigue safety coefficient and welded joint cumulative damage. This method provides a useful reference for the reliability design of the lightweight welded vehicle structures.This research is funded by the State "863" high-tech research and development project: "Coordinated design of complex product, simulation and optimization----integrated platform research, development and application" (Project Number: 2006 AA04Z160).