Guided wave mechanics modeling for sensor development and process control in powder injection molding.

摘要

This dissertation reports the application of an elastodynamics model, incorporating viscoelastic effects, to comprehend quality of powder injected molded parts in situ. Work reported here goes beyond quasi-point measurements involving normal incident bulk waves. Ultrasonic guided wave sensors are mounted in the mold of a powder injection molding machine to capture changes in the part microstructure on-line. Comprehensive information of the part quality is obtained since improved sensitivity is possible due to variation in ultrasonic guided wave structure that occurs across the thickness of the part being manufactured. As a result, properties along the center of the product as well as on either surface can be evaluated.; This dissertation reports computations of oblique incidence reflection and transmission factors for the powder injection molded feedstock embedded between two steel inserts. The computations of the reflection and transmission factors aid in determining the optimum angles and frequencies for the waves to strike the steel-feedstock interface. Thus, knowing the material properties of the feedstock, it is possible to predict optimum angles and frequencies of incidence, for a part of any composition and shape.; In situ experiments involve acquiring the guided wave propagating through the feedstock, and eventually leaking off into the mold insert. To interpret and quantify the guided wave mode sensitivity to part microstructure, a numerical method is reported to extract the complex roots of the dispersion equation. The dispersion curve calculations finally lead to wave displacement computations. From the shape and energy carried by the displacements, it is feasible to analyze the real-time filling signals during powder injection molding.; Experiments are carried out on parts with different geometries and compositions. State-of-the-art hardware and software developments are carried out to implement in situ monitoring. Theoretical wave mechanics model predictions are applied in an industrial environment successfully. Extraction of features, from the in situ signals, for various modes are considered for their ability to evaluate product quality and subsequent feedback for improvement in process control. This dissertation covers theoretical wave mechanics modeling efforts of this problem as well as experimental results on a variety of proper and improper material processing situations.