NSF Young Investigator: Formal Tolerance and Set-Based Methods in Engineering Design and Manufacturing

点击翻译按钮获取中文摘要

This project focuses on a balance between an academic research program, engineering education, and industrial applications. In this program, the general problems of engineering design and design for manufacturing will be addressed. In particular, one part of the research program considers the development of analytical and experimental models for determining the physical surface-structures produced from both material removal processes and solid-freeform fabrication. The basis for these models is fractal-based representations, or more generally the Karhunen-Loeve method, which will provide indices for linking manufacturing process variables to design parameters. While a fractal-based approach is known to be applicable at surface finish scales, surface structures that will be identified in this research are at tolerance scales (e.g., 40-300 microns for journal bearings). Tolerances usually carry the function of the device. For example, in a rotating mechanism, the tolerances associated between a shaft and a bearing assembly may be critical, i.e., if the shaft-bearing assembly is misaligned or out-of-tolerance, catastrophic failure of the system will result due to excessive wear and possible seizure. Providing design methods and measures that link manufacturing process variables to the design parameters, performance parameters, and functional requirements of a design is crucial, especially if manufacturing information is to be combined with functional considerations in the early stages of design. This research will seek to verify the fractal-based models with experimentation, represent the results within a formal set-based theory for design, and apply the method to rotating machinery, structures, and microautomation. Beyond the fundamental theories of fractal tolerances and set-based design representations, computational methods will be developed for linking global geometric information to the design for manufacturing techniques. The general scope for these computational methods will be a feature-based design system, including the formal representation of spatial relationships.