Design and Control of Capstan-Pulley Draw System for Optical Fiber Manufacturing

Document Type

Article

Language

eng

Publication Date

7-7-2008

Publisher

American Society of Mechanical Engineers

Source Publication

ASME 2008 9th Biennial Conference on Engineering Systems Design and Analysis Volume 2: Automotive Systems; Bioengineering and Biomedical Technology; Computational Mechanics; Controls; Dynamical Systems

Source ISSN

9780791848364

Abstract

Optical fiber for data communication is manufactured by the draw process, which involves heating and pulling high purity glass cylinders to diameters of 125 μm. The diameter of the glass fiber and its light-guide core must remain constant to create a product capable of transmitting high-bandwidth optical data. The draw capstan design has a significant impact on the optical fiber quality. As the draw speed is used to control the fiber diameter, the ability of the draw capstan to follow velocity commands directly affects the resulting fiber diameter. To improve the control of the optical fiber diameter, the design of the overall system was revisited. A lumped-parameter model of the capstan drive was developed. It accounts for disturbances in the draw process that arise from sources such as the variation in the diameter of the input glass cylinder and the draw tension control, affecting the glass temperature and viscosity. The selection of the motor and the design of the speed controller in the optical fiber draw capstan pulley system were studied. Simulation studies over a range of parameters demonstrate that speed regulation, necessary to manufacture optical fiber within allowable diameter tolerances, can be achieved in the presence of estimated process disturbances without motor current saturation. The model predictions suggest that an effective draw capstan system can be synthesized and controlled. This case study has broad applications for the design of practical engineering systems that include control, electrical, and mechanical subsystems, typical in modern manufacturing. The paper highlights the use of mechanical and electrical modeling, system identification, and control design as necessary parts of product and process improvement.

Comments

Published as a part of ASME 2008 9th Biennial Conference on Engineering Systems Design and Analysis, Volume 2: Automotive Systems; Bioengineering and Biomedical Technology; Computational Mechanics; Controls; Dynamical Systems, Haifa, Israel, July 7–9, 2008. Paper No. ESDA2008-59290, pp. 471-478. DOI.

Share

COinS