Fiber tension prediction and control methods in automated fiber placement

Abstract

In automated fiber placement (AFP), the fiber tension is an important process parameter. Sharp fluctuations in fiber tension are unacceptable in AFP process. The focuses of this article are on fiber tension modeling and control methods for AFP. Firstly, the mechanical model of fiber transport based on frictional loss is established, and the control objective, which should really be focused on in the tension system, is pointed out. Then, identification experiments are carried out to obtain the unknown parameters in the mechanical model. In addition, a dynamic model of fiber tension based on a passive dancer roller is established and linearized by curve fitting. Finally, a variable‐parameter Proportion‐Integration‐Differentiation (PID) controller is established to achieve stable control of the fiber tension. The experimental results show that the proposed control method can accurately stabilize the actual tension at the target value and has a lower control error in the face of strong velocity fluctuations than the traditional tension control system.Highlights The fiber tension at the end of the tow transfer path cannot be measured directly by the sensor in automated fiber placement. The tension model predicts the fiber tension at various points in the tow path. The control method using the tension prediction model is much more accurate than the traditional method.