Modeling and control of thrust force in the drilling of composite materials

This dissertation deals with the characterization and modeling of drilling processes of composite materials, and the development and test of an intelligent supervisory controller for their delamination-free drilling.;Characterization of the thrust force is achieved through drilling experiments with constant feedrate. The drilling process consists of different stages such as entrance, middle, exit and withdrawal; the control strategy at each stage is explained. The major contribution of this dissertation is the development of various force controllers that regulate the actual force at the required reference force.;Global modeling of thrust force including entrance and exit stages is investigated. On-line process gain estimation is performed so that the linear model may be applied to the whole drilling process under the situation of large variation of operation range. It can also compensate for the gain variation due to tool wear and nonlinearity of the drilling process.;When the on-line gain estimation is available, adaptive PI control is investigated in which the PI control signal is scaled by the estimated gain to make the overall gain unchanged. This is helpful in achieving same transient performance for different operation regions and compensating for gain variations.;A new hybrid control scheme is proposed in which the on-line gain adaptation and fuzzy learning control are coordinated together to achieve a better overall control performance.;Simulation and drilling experiments are conducted to show the advancement of the concepts introduced in this dissertation. Experimental results show that this control strategy is able to shorten the rise time and reduce the overshoot simultaneously and achieves efficient delamination free drilling. (Abstract shortened by UMI.).