Resonant frequencies in web lines and filtering of web tension

Scope and Method of Study. The thesis proposes a dynamic model for an idler system and investigates the relationship between the idler system's minimum resonant frequency and the system parameters. A Kalman filter is designed and implemented in the web tension feedback loop to improve the tension control performance. A large experimental platform, which mimics most of the features of an industrial web process line, is used to verify the proposed idler system model experimentally and to demonstrate the effects of the tension Kalman filter.;Findings and Conclusions. A state space system model is proposed to predict the resonant frequencies of an idler system consisting of idler rollers and web spans between two driven rollers. A close match between the experimental results and the model simulations validate the correctness of the dynamic model. The relationship between system parameters and the system's minimum resonant frequency is investigated through analytical and numerical analyses. An analytical approximation of the minimum resonant frequency is developed for a special class of idler systems consisting of equal span lengths and any number of idle rollers. A Kalman filter is designed based on the web span tension dynamic model in terms of state errors. Extensive comparative experiments with and without the Kalman filter show improvement in the web tension signal when the Kalman filter is used.