Develop and demonstrate at commercial production rates an intelligent extruder monitoring and control system that will reduce production variability, energy use, and offgrade waste stream generation. This project will develop and implement three major aspects of machine intelligence in extruder systems for polymer processing:
(1) model-based process and equipment diagnostic software to detect problems in compounding line components before unacceptable material is produced,
(2) an inferential estimation system using a few vital variables from readily available extruder data to estimate material properties for use in continuous quality audits and process control, and
(3) closed-loop multivariable process control to keep process parameters within acceptable limits, despite incoming material variations and correctable process upsets.
Technologies for new product or retrofit applications of diagnostics, inferential estimation, and closed-loop control will be developed and demonstrated at interim milestones over the three-year course of development, culminating in commercial product and service offerings from a GE Industrial Systems / Werner-Pfleiderer partnership. System prototypes will be tested, going from a 30-mm (50 lb/hr) laboratory scale to 58-mm (500 lb/hr) pilot scale to 133-mm (>2,000 lb/hr) commercial production scale compounding line.