Combining Physical and Statistical Methods for Modeling Novel Manufacturing Processes

点击翻译按钮获取中文摘要

9415001 Finger The award is for creating a methodology for acquiring process models of novel manufacturing processes using statistical methods, physical models, and experimentation. The process that will be studied is the shape deposition melting process, which is a solid freeform fabrication process. The part is built up by successively depositing molten materials, such as steel and copper, in thin layers using a microcasting process. Other subprocesses used are thermal spraying, shot peening, and machining. Only the microcasting and machining subprocesses will be studied using the proposed methodology. Statistical design of experiments will be used to create coarse process models that identify important process variables and relationships among them. For subprocesses not well understood at present, models will be developed based on the underlying physics of the subprocesses. Surrogate models of different levels of complexity and detail will also be abstracted from the detailed models. The methodology to be developed is particularly useful for studying the shape deposition process which is a rapid prototyping process. More conventional processes would also benefit from the methodology.