Real-time web tension prediction using web moving speed and natural vibration frequency

Abstract

Precise web tension measurement is critical for web handling and process control in roll-to-roll (R2R) manufacturing systems. Calculating web tension by using out-of-plane vibration measurement offers a contactless solution to tension measurement during the R2R process, which is suitable for situations where contacting web tension measurement sensors cannot be applied. In this paper, a real-time prediction of web tension based on contactless sensing of out-of-plane web vibration is presented for the first time. We design and test a fast frequency estimation algorithm, based on the Fourier series, which solves the frequency resolution problem currently existing in real-time applications when conventional FFT is applied. Our simulation results show that the proposed method has an accurate estimation of frequency under harmonics and white Gaussian noise based on few observed cycles. A R2R testbed was built for testing the real-time performance of the proposed algorithm including accuracy, robustness, and computation speed. The experimental results on the testbed show that the proposed method can predict the web tension with an error less than 10% for various tension values and track the tension changes within one analysis window of 50 ms.