Digital Fiber Length Determination for Injection Molded Glass Fiber Reinforced Composite Materials

Andreas Grießer presents a novel, non-destructive method for determining fiber length distributions in injection-molded composite materials using CT-scans. This approach combines advanced image processing with AI-driven segmentation techniques, enabling precise analysis of individual fiber properties for more detailed insights into fibrous structures and facilitated modeling of similar structures. Andreas Grießer will detail the neural network training process, its application to CT data, and the comparison of results with traditional experimental methods such as the FASEP apparatus. The presentation will illustrate the utility of this methodology in assessing material aging and improving composite performance through detailed statistical analyses of fiber properties.