|
1488 |
|
Wednesday, June 04, 2008, Room 08 2:00 PM Diagnostics and Sensors 2 Modern process diagnostics offer a lot of possibilities for gaining data for both quality control and process knowledge. The recorded data is the base for process models and enhances the process knowledge |
|
Sensor issues for real time control of coating thickness in plasma spray |
|
Donald Wroblewski / Boston University, USA Brian Vattiat/ Cyber Materials, USA Onomitra Ghosh/ Boston University, USA Michael VanHout/ Boston University, USA Soumendra Basu/ Boston University, USA Michael Gevelber*/ Cyber Materials, USA |
|
Currently, the plasma spray process is run open-loop with respect to the critical deposition physics and is characterized by large variations in coating thickness and structure which affect yield and quality and pose challenges for new coating designs. Although closed loop control offers the potential for dramatic improvement in the quality and reliability of coatings, development of such advanced control will require sensors that can rapidly detect plume characteristics that most closely map onto the critical coating qualities. This paper will discuss the requirements of sensors for advanced control and the critical issues associated with design and implementation of such sensors. The overarching requirement is that the sensor must capture the subset of particles that most contribute to coating properties. This critical sub-distribution must be filtered and aggregated in a manner that reflects the volume-weighted contribution to the coating buildup. This approach demands a sensor capable of imaging the entire plume at a rate suitable for real-time control, with a dynamic range commensurate with the full range of intensities from particles that constitute the critical sub-distributions. This paper will focus on sensing for coating thickness control, for which molten particles make up the critical sub-distribution. Our experiments have shown that molten mass flux correlates better with coating thickness than bulk-average temperatures or intensity, since it captures the volume weighted characteristic of coating buildup, and includes just the molten particles, which represent the critical sub-distribution. Since the volume-weighted nature of the molten volume flux requires accurate diameter measurements, calibration procedures have been developed that match measured distributions with powder distributions, using a volume-weighted approach. |