Theoretical Foundations of Mechanical Tolerancing

9500140 Bisgaard这项拨款为开发机械和机电产品公差的理论基础提供资金。这个基础将以现代统计理论和决策理论为基础。它将提供一个统一问题的技术和经济方面的框架。具体来说，该方法将考虑来自原材料、制造过程和测量误差的成品的固有变异性。此外，它将纳入有关偏离目标规格、制造和检验的功能损失成本的信息。它将整合之前几次提供基础的尝试，包括日本工程师田口博士的尝试。研究计划是从简单的问题开始，建立一个统一的理论，并在与工业合作的实际应用中测试这些想法，从而增加复杂性。如果成功，这项研究将为公差信息的工程通信带来更简洁的语言，从而避免当前的歧义；以安全、可靠和可预测的方式执行所需功能的产品的经济设计、制造和使用取决于清晰和明确的公差信息。此外，这项研究的成功完成将为工程师提供目前无法用于设计和评估公差性能的科学和数学工具。当前工程信息计算机化的趋势，特别是计算机辅助设计（CAD）和计算机辅助制造（CAM）工具的使用，需要一个理论和数学基础。如果本研究成功，将会缓解当前CAD/CAM工具存在的许多问题。

9500140 Bisgaard This grant provides funding to develop a theoretical foundation for the tolerancing of mechanical and electro-mechanical products. This foundation will be based on modern statistical theory and decision theory. It will provide a framework that unifies technical as well as economic aspects of the problem. Specifically the approach will take into account the inherent variability in maunfactured products coming from raw materials, maufacturing processes and measurement errors. In addition it will incorporate information about the cost of functional losses coming from deviating from target specifications, manufacturing, and inspection. It will integrate several previous attempts to provide a foundation, including that of the Japanese engineer Dr. Taguchi. The research plan is to build up a unified theory starting with simple problems and increase the complexity as the ideas are tested on practical applications in collaboration with industry. If successful, this research will lead to a more concise language for engineering communication of tolerance information that will avoid current ambiguities; the economic design, manufacturing, and us of products that perform the desired function in a safe, reliable, and predictable manner depend on clear and unambiguous tolerance information. Moreover, the successful completion of this research will provide engineers with scientific and mathematical tools not currently available for the design and evaluation of the performance of tolerances. The current trend toward computerization of engineering information and in particular the use of computer aided design (CAD) and computer aided manufacturing (CAM) tools necessitate a theoretical and mathematical foundation. If this research is successful many problems with current CAD/CAM tools will be alleviated.