Measurement of air temperatures and velocities in the fiber glass forming environment

Abstract

The manufacture of continuous strand fiber glass is accomplished by draining molten glass through an array of small orifices in a heated bushing plate and then drawing the glass onto a winder. Upon exiting the orifices, the fibers pass through a cloud of atomized water to assist in the cooling of the fibers and then over a surfactant applicator before they are wound. The rate at which the cooling of the fibers occurs relies heavily on the operating conditions used in the process. This report deals with measurements made in a scaled-down process using a variety of different operating conditions. Experimental data on air temperatures and velocities were recorded within the forming environment without water sprays. Trials were run using an apparatus similar to one used in production, but of a much smaller scale. In production, a granular batch is continuously fed into a furnace, melted, mixed, and sent down a slightly downward sloping channel, via gravity. A series of electrically heated bushings, each with hundreds of tips, lies below the channel, allowing the molten glass to drain. The apparatus used in this study, however, consisted of one isolated nine-tip bushing with a small well above it. Glass marbles were placed in the well, and the system was electrically heated, melting the glass and allowing it to drain from the well through the bushing tips. The speed at which the fiber was drawn was varied, as was the bushing temperature. It was found that fiber diameter, which is information essential for predicting cooling rates, varies with both of these parameters, but is more heavily influenced by bushing temperature. By running the process at a constant bushing temperature, while using a series of fiber speeds, it was found that varying the speed of the fiber noticeably changes the temperature field in the forming environment. By increasing the fiber speed from 12.2 m/s to 21.3 m/s there are more noticeable temperature peaks surrounding the individual fibers, indicating significant differences in convective cooling rates as fiber speed is varied. Similar trends are seen for a constant fiber speed as the bushing temperature is increased from 1505 K to 1533 K, though the differences are not as dramatic. In addition it was necessary to correct temperature measurements using a theoretical heat transfer model, to account for error due to radiative heating of the thermocouple from the bushing plate and conductive heat loss through the thermocouple leads. Air velocity measurements were also taken using the same series of operating conditions. Increasing the fiber speed resulted in noticeably higher air velocities; increasing the bushing temperature also raised air velocity, although the change was less significant.