Faculty Early Career Development: Research in Combinatorial Structure of Mechanical Parts

机械系统快速原型设计的一个主要障碍是缺乏组合模型来描述机械形式、功能和制造过程中离散的、简单的和相互作用的原语。本研究的目标是通过(1)研究、试验和修改一类工业零件的现有生产实践，来调查机械形式、功能和单件机械部件制造之间的组合关系；(2)发展一种新的理论，能够支持基于物理和制造特性的这些部件的组合模型的系统创建和修改；(3)开发可与现有商业CAD软件结合使用的计算机算法、表示和系统。该教育计划认为实体建模是工程语言发展的中间阶段。通过研究产生的新知识，工程分析、设计和制造中的学科可以重新组织和系统化。这将有助于沿着从实体建模到更一般的机械和物理建模学科的迁移路径取得进展。从这项研究中产生的见解和新知识应该会导致更好的工程设计方法，在许多工程活动之间更有效的沟通，以及更有效的生产过程。将这些结果推广到涉及多个相互作用的部件和现象的机械功能、过程和系统中，可以提高我们在结合重要的人类工程知识和实验的同时有效地自动化某些设计方面的能力。

9502728 Shapiro A major roadblock for rapid prototyping of mechanical systems is the absence of combinatorial models to describe mechanical forms, functions, and fabrication processes in terms of discrete, simple, and interacting primitives. The goal of this research is to investigate a combinatorial relationship among mechanical form, function, and fabrication for single-part mechanical components by (1) studying, experimenting with, and modifying existing production practices for a class of industrial parts; (2) developing a new theory capable of supporting systematic creation and modification of combinatorial models of such parts based on physical and manufacturing characteristics; and (3) developing computer algorithms, representations, and systems that can be used in conjunction with existing commercial CAD software. The education plan considers solid modeling as an intermediate stage in the evolution of engineering languages. With new knowledge resulting from the research, subjects in engineering analysis, design, and manufacturing can be reorganized and systematized. This will contribute to progress along a migration path from solid modeling toward the more general disciplines of mechanical and physical modeling. Insights and new knowledge resulting from this research should lead to better approaches toward engineering design, more effective communication among many engineering activities, and more efficient production processes. Generalizing these results to mechanical functions, processes and systems that involve multiple interacting parts and phenomena can improve our ability to effectively automate some aspects of design while still incorporating important human engineering knowledge and experimentation.