A step towards predicting the mechanical properties of weld lines in injection-molded short fiber-reinforced thermoplastics

2019 
Weld Lines frequently appear by injection molding when separate polymer melt fronts meet. They induce a significant reduction in the failure strength and strain, especially for composites. It is therefore essential to predict reliably their mechanical properties during the product design phase, but current simulation tools are still not able to do it. Literature points to two main reasons of WL weakness: an incomplete polymer matrix healing and a change in the fiber orientation distribution. The objective of this work is to characterize and quantify the contribution of these factors and to contribute to improving the prediction of the mechanical properties of injection molded short-fiber reinforced polymers.Samples of 30 % wt. glass fiber-reinforced PBT were injection molded with frontal and flowing weld lines. The deformations in the mechanical tests were measured by of digital image correlation to quantify the location of deformation in the weld lines. The microstructure was quantified using X-ray computed tomography scans. In particular, we were able to show that the flowing weld lines were fading very slowly and that the material fronts behave towards each other like walls. A physical model based on reptation theory was implemented to determine a criterion for interface healing. Using the measured orientation distribution and an appropriate homogenization scheme for each element of a finite elements simulation, the mechanical properties up to the failure of the composite could be calculated and explain the reduction in mechanical properties at the weld lines, knowing that the interfaces were fully healed for the study material. Finally, for this semi-crystalline polymer, we have shown the need to use an elastoplastic constitutive law with a damage threshold depending on the orientation of the fibers.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    0
    References
    1
    Citations
    NaN
    KQI
    []