Digital Twin Validation for the In-Situ Automated Fiber Placement Process of Thermoplastic Composites

The digital twin is a methodology accelerating process knowledge for the automated fiber placement (AFP) of unidirectional thermoplastic tape with in-situ consolidation. It implies a continuous assessment and evaluation of data allowing a faster process optimization and reducing ramp-up times. In a previous study, a digital twin is developed allowing the interactive visualization of process data within a digital shadow and the interpretation of quality measures based on analytical methods. In this study, the developed digital twin for predicting the degree of bonding is validated by experimental results. By using destructive testing methods, such as interlaminar shear strength tests and microsections, the degree of bonding is quantified. The results are feedbacked to the initial digital twin forecast and a comparison is made. The results disclose the need to calibrate the analytical models as the underlying assumptions do not hold in a realistic environment. Based on these findings, a methodology is presented enabling a fast calibration of the analytical models. Aim is to reduce extensive experimental work for process optimization and, thus, to accelerate the ramp-up for a flexible series production.