Material processing and intelligent manufacturing will be studied in an interdisciplinary format involving process modeling, in-situ product quality monitoring, and real-time process control. The program will initially focus on molded and advanced composite products. Two fundamental manufacturing science problems will be addressed, which are (1) the majority of manufacturing processes are not completely understood, and (2) product quality is not monitored and optimally adjusted during fabrication. Improved process models for molding and advanced composites will then be employed in real-time environments to direct manufacturing processes toward optimal product quality in a feedback control manner. Doing so will require the development of sensor subsystems for in-situ product quality assessment and knowledge based tools for real-time manufacturing precess adaption. Methodologies for operations will be developed. Several sensory transducer mechanisms will be investigated, with an initial focus on embedded optical and electronic fibers. Real-time process adaptation will be accomplished through the development of knowledge based tools incorporating neural networks. A novel component of the effort will be the procurement of neural network training information from process simulation models rather than from actual manufacturing data. Overall, the synergistic manufacturing methodology development will decrease both process development time and cost while improving product quality and process flexibility.
