Modeling and Optimal Design of a Glass RTP System

The problem of modeling and generating an optimal design of a rapid thermal processing (RTP) system for flat panel display glass was addressed. An RTP system using bulb-type tungsten-halogen lamps was considered, and a dynamic model to describe the glass temperature distribution was established. The power dispersion function comprising the model was established in an experimental RTP setup for 2G display glass. The model was used to determine an optimal lamp arrangement for a 4G glass RTP system such that the maximum temperature error meets the specifications while both heating power and lamp occupation area are moderate. An iterative learning control method was applied and found to be able to closely match the predicted ultimate temperature error in dynamic situations with model uncertainty.