Sensor-Based Prediction of Thermally Induced Lateral Dynamics in Roll-to-Roll Web Processing

In roll-to-roll (R2R) manufacturing processes, lateral displacement of moving webs in the cross-machine direction (CMD) is a critical issue that directly affects coating uniformity, registration accuracy, and overall product quality.In drying sections, spatially nonuniform thermal conditions induce asymmetric thermal expansion of the web, resulting in time-varying geometric imperfections and complex lateral dynamics that are difficult to predict using conventional analytical approaches.This study proposes a sensor-based thermomechanical modeling framework for predicting thermally induced lateral web displacement in R2R drying processes.CMD temperature distributions, acquired using distributed in-situ thermocouples, are directly incorporated into an extended lateral dynamics model to represent temperature-dependent geometric variations and their coupling with web tension.By explicitly linking sensormeasured thermal fields with lateral displacement prediction, the proposed approach enables physics-based interpretation of thermally driven lateral behavior under asymmetric heating conditions.Experimental validation was performed using an industrial-scale R2R coating system equipped with temperature and edge-position sensors.The results demonstrate that the proposed sensor-informed model provides significantly improved prediction accuracy compared with conventional uniform-temperature assumptions, particularly in cases with pronounced CMD temperature gradients.Statistical analysis further confirms that web speed is a dominant factor governing lateral displacement behavior.The proposed framework provides a practical basis for sensor-based process monitoring and predictive assessment of lateral web dynamics in R2R manufacturing systems and supports systematic evaluation of thermally induced process variability.