CAD-based virtual prototyping and behavioral modeling of blade assembly process in vehicle torque converters

This dissertation presents a methodology aimed at quantifying the blade assembly process for automating the assembly process. A virtual prototype is created for the blade assembly in a torque converter of an automobile to reduce the physical prototyping needs. In this work a comprehensive method is proposed for analyzing, optimizing, and quantifying the assembly design for turbine and pump blades in vehicle torque converters. This model development is based on the behavioral modeling technique and implements ProEngineer ®. Based on the behavioral modeling requirements, the design intents, pertinent parameters, and their relationships, the entire blade assembly process must be captured and integrated into the model. The proposed method can be used to obtain the blade trajectory during the assembly process and to optimize the loci of blade motion for automated assembly purposes. The novelty of this work lies in that, through comprehensive kinematics analysis, it computes and offers animated visual prototyping to show the interference of the blade tabs with the turbine and the pump shells during the assembly process. The developed method enables the determination of the feasibility of blade design and assembly in the early phases of design and is superior to all existing blade design methods. As opposed to existing methods, the proposed method can eliminate the expensive and time-consuming physical prototyping needs. Aiming at blade assembly automation, the research effort focused on quantifying the blade assembly process by expertly defining certain benchmarks and indices based on the kinematics behavior of the assemblages. These benchmarks can be used for evaluating and guiding the development of the blade assembly design in torque converters or more generically, in any kind of turbo machine. Moreover, the finalized benchmarks can be used for documenting, evaluating, and comparing the blade assembly designs in various assemblages of torque converters to improve and transfer the experiences of the designers embedded in the model. In the experimental phase of this dissertation, the proposed methodology has been tested and verified to be effective with a realistic industrial torque converter. The results of the experiments and comparisons of real and virtual prototypes are demonstrated and discussed in detail.