Modelling and control of dieless wire drawing.

A novel technique of drawing wire, i.e., dieless wire drawing, is described in which no conventional reduction dies are used. The wire is passed through a unit having stepped parallel bores filled with polymer melts, the smallest bore diameter being greater than the initial nominal wire diameter. The technique eliminates the need for a leader wire, prevents breakage during start-up, and eliminates the use of conventional reduction dies and, hence, the problem of die wear.

An analysis has been developed for predicting the produced wire sizes for given drawing speeds, the pressure distribution within the unit and the drawing load during the drawing process. The performance of the drawing process is effected by means of the plasto-hydrodynamic action of the polymer melt within the unit. Heat is generated by mechanical dissipation during the drawing process. The effect of this upon the performance of the process is considered by coupling the energy equation into the analysis and allowing the polymer viscosity to be dependent on temperature. Agreement between the theory and experiment is found to be close. To further investigate the performance of the drawing process, an extensive experimental study has been undertaken, in which parameters such as drawing speeds, the temperature of the polymer and type of polymers are varied.

Based on the theoretical and experimental results, a microcomputer-based control system for the dieless wire drawing process has been designed and built. A method with which the percentage reduction in area of the wire can be measured on line is proposed. Wires of desired qualities in terms of consistency in diameter over long lengths, or uniform change in area of the wire, are capable of being produced from the new system.

A self-tuning controller for determinisic systems is proposed, which has the same structure as a conventional PID controller. The new self-tuning PID controller is based on a generalised self-tuning controller with pole assignment for deterministic systems. Simulation and experimental studies for several examples indicate that the new controller performs well and can be further improved to provide a fundamental method of tuning a PID controller.