Model-based integration of control and supervision for one kind of curing process

The optimization of the cure schedule for one kind of adhesive die-attach curing process in the electronics industry is very difficult to achieve due to the lack of tools for the online measurement of the extent of reaction adhesives during curing. In practice, the cure schedule is typically determined in a trial-and-error process, even though this is costly and may not guarantee the reliability of the adhesive die attach. A novel model-based integration of cure schedule optimization, supervision, and decoupling control is introduced to maintain both reliability and throughput. First, a novel hybrid spectral/ neural method is used to model the curing process. The model developed can accurately estimate the temperature field inside the chamber. Then, an approximate decoupling linearization controller is developed to suppress the coupling effects from different heating sources for a better temperature tracking. Finally, the optimal cure time and temperature setpoints are accurately calculated from the characteristics of the cure oven and the cure kinetics of the adhesives used. The method is straightforward and effective, and can be easily applied to the curing supervision. Such a system-wide integration of control and supervision can be utilized to replace the traditionally used, unreliable trial-and-error process, and will provide an optimal production that is able to adapt to varying operating conditions. © 2004 IEEE.