Real-time imaging for thermal spray process development and control

Abstract

Thermal spray and other high-temperature industrial processes are quite difficult to monitor with the human eye, because the luminous volume of the plasma or flame obscures the behavior of the solid or molten material in the heat-affected area. When a photographic or video camera is used, viewing is further degraded by the extreme contrast variation across the image area, making it impossible to achieve proper exposure throughout the image—except possibly for small areas of comparable brightness. Optical filtering with neutral density filters, such as those used in a welder’s helmet, are of no practical benefit. With thermal spray processes, the injection and flow of particles within the plasma flame is almost totally concealed by the extreme brightness of the plasma, flame, or arc. In addition, the particles quickly accelerate to very high speeds, making their detection even more difficult. This article discusses the development of integrated thermal spray process monitoring and analysis techniques based on two principles. The first is a unique vision sensing system that suppresses the flame, plasma, arc, or other high-luminosity phenomena in the video image. A further improvement is the use of dedicated image and analysis processing to enhance the sensor images and extract features of interest or dimensional measurements. These experimental techniques can be used as feedback for automated process monitoring and control.