Fibre orientation distribution function mapping for short fibre polymer composite components from low resolution/large volume X-ray computed tomography
Robert M. Auenhammer,Anuj Prajapati,Kaldon Kalasho,Lars P. Mikkelsen,Philip J. Withers,Leif E. Asp,Renaud Gutkin
DOI: https://doi.org/10.1016/j.compositesb.2024.111313
IF: 11.322
2024-02-23
Composites Part B Engineering
Abstract:Short glass fibre injection moulded composites, used in interior and exterior automotive parts, are exposed to complex stress states, for example during a crash. As the fibre scale dominates the composite's material properties, numerical models need to account for the local fibre orientation. In recent years, mould flow simulation results have been exploited to predict the fibre orientations for finite element models, albeit with limited accuracy. Alternatively, X-ray computed tomography can be used to directly image and analyse fibre orientations. Traditionally, achieving the necessary resolution to image individual fibres restricts the imaging to small regions of the component. However, this study takes advantage of recent advancements in imaging and image analysis to overcome this limitation. As a result, it introduces, for the first time, a reliable, fast, and automated fibre orientation mapping for a full component based on image analysis on a single fibre level; even for cases where the pixel size is significantly larger than the fibre diameter. By scanning at lower resolutions, a drastically larger volume of interest can be achieve. The resulting fibre orientation analysis and mapping algorithm, based on X-ray computed tomography, is well matched to the level of information required for automotive crash modelling with a standard element-size of a few millimetres. The entire process, encompassing image acquisition, image analysis and fibre orientation mapping, can be directly integrated in an industrial full component application in a matter of hours.
engineering, multidisciplinary,materials science, composites