Assessment of fiber orientation models predictability by comparison with X-ray µCT data in injection-molded short glass fiber-reinforced polyamide

The distribution of fibers in injection molded thermoplastic-reinforced parts is known to significantly affect both the final mechanical properties and the appearance of defects related to undesired shrinkages and warpages. Even if the numerical modelling of the process is in the state-of-the-art, the selection criteria of the model for the fiber orientation predictability and the influence of the models’ parameters are not yet clearly understood. The complexity of the matter increases further considering that, inside the same injection molded part, the fiber orientation distribution could differ from one region to another depending on the local melt shear-flow type. In this context, the aim of this study is to investigate the orientation of fibers in regions of non-simple shear flow in an injection molded short glass fiber reinforced part. X-ray micro-computed tomography is used to experimentally investigate fiber distribution and orientation. Furthermore, a number of numerical simulations of the injection molding process are performed in Moldflow® by varying the mesh type (2D/3D), the selected predictive model and the models’ coefficients. The main findings indicates that, in regions with non-simple shear flow, a 2D mesh fails to capture accurate fiber orientation. A 3D mesh is essential for reliable predictions. MRD-specific parameters from the literature reduce prediction error by 37.5% compared to Moldflow's default MRD parameters. Regarding the RSC model, both Wang’s optimal retarding rate parameter and AMI's default retarding rate parameter result in similar 5% prediction errors. However, employing a literature-suggested interaction coefficient raises the prediction error to 26%.