Author:
Osterberger Joachim,Maier Franz,Keller Sophia,Hinterhölzl Roland Markus
Abstract
In order to demonstrate the accuracy of macroscopic finite element draping simulations it is necessary to compare the results with experiments. In this work, a compact overview of evaluation methods for draping simulations based on experiments, in the recent literature, is provided. Then, a method using state of the art 3D laser scans (RS6, Hexagon) and robot supported fiber angle scans (FScan, Profactor) is described. The presented comparison of draping results with the tool geometry in 3D allows for an evaluation of wrinkles and bridging areas. For the evaluation of the edge contour, the commonly used method of projecting the edge contour on a 2D plane is extended to a comparison in 3D space. To determine fiber orientations and compare them with the predictions from simulations, a robot supported fiber angle sensor and a script-based mapping and comparison algorithm are used. The results are further analyzed statistically, to derive comparative figures to other results found in the literature. The location and dimensions of macroscopic manufacturing effects such as fiber bridging and wrinkles could be predicted accurately. The final component contour could be on average predicted within 5.2 mm. The fiber orientation could be predicted with a deviation of less than 2° for approx. 65% and within 6° for 95% of the part surface for UD laminas. Cross-ply laminas showed larger deviations, only 40% of the part surface was within 2° and 80% within 6°, compared to the experiment. Overall, the results for the presented methods show good agreement between multi-ply draping experiments and macroscopic simulations conducted with the Abaqus Fabric material model.
Subject
Materials Science (miscellaneous)