The aluminum industry is in need of high-speed in-situ elemental identification technology for molten metals. wTe Corporation of Bedford, MA was granted funds to research this technology for aluminum through the Advanced Technology Program (ATP) of the National Institute of Standards and Technology (NIST). As a secondary metal scrap processor, wTe Corporation is interested in increasing the value of scrap and better identification technology is one way of doing this. With better identification technology, foundries and smelters will be more confident in the melt chemistry and more apt to use secondary aluminum which decreases the energy required for processing by approximately 85%. wTe Corporation is exploring two avenues for in-situ molten metal identification: Laser Induced Breakdown Spectroscopy (LIBS) and X-Ray Fluorescence (XRF). The objectives of this work were to contribute to developing XRF technology for in-situ identification of molten metal by establishing a method of data instrumentation and analysis for XRF to determine aluminum melt composition, investigate the major alloying elements in the Al380 alloy series and determine the relationship between intensity and concentration, and to determine the effect of temperature on XRF Spectra. The XRF instrumentation development and the technical challenges associated with high temperature measurements are presented. The relationship between intensity and concentration is presented for three alloys within the 380 alloy series, and lastly it is observed that there are significant differences between liquid and solid spectra and that a calibration curve for liquid data is necessary. Several hypotheses are given as to why this difference between liquid and solid state spectra may occur.

Manufacturers of cast metal parts are interested in the development of a feedback control system for use with the Precision Sand-Casting (PSC) process. As industry demands the ability to cast more complex geometries, there are a variety of challenges that engineers have to address. Certain characteristics of the mold, such as thick-to-thin transitions, extensive horizontal or flat surfaces, and sharp corners increase the likelihood of generating defective casts due to the turbulent metal-flow during fills. Consequently, it is critical that turbulent flow behavior within the mold be minimized as much as possible. One way to enhance the quality of the fill process is to adjust the flow rate of the molten metal as it fills these critical regions of the mold. Existing systems attempt to predict the position of the metal level based on elapsed time from the beginning of the fill stage. Unfortunately, variability in several aspects of the fill process makes it very difficult to consistently predict the position of the metal front. A better approach would be to embed a sensor that can detect the melt through a lift-off distance and determine the position of the metal-front. The information from this sensor can then be used to adjust the flow rate of the aluminum as the mold is filled. This thesis presents the design of a novel non-invasive sensor monitoring system. When deployed on the factory floor, the sensing system will provide all necessary information to allow process engineers to adjust the metal flow-rate within the mold and thereby reduce the amount of scrap being produced. Moreover, the system will exhibit additional value in the research and development of future mold designs.