Search

Thermal deformations that occur during formation of long-fiber-reinforced composites have been a continued challenge for manufacturers as the final shape of a given part can be different from the original mold shape. The ensuing dimensional distortions can be difficult to predict due to complex thermo-mechanical behaviour of composite laminates during different forming cycles. This study intends to model the fundamental mechanisms that lead to thermal deformations during forming of a thermoplastic matrix composite comprised of comingled polypropylene and E-glass fibers. While the discussion is framed around a custom-design multi-stage roll-forming process, it is also relevant to a wider range of thermoplastic composites manufacturing processes. A methodology is developed to characterize the thermal mechanical behavior of the material, optimize the manufacturing process, and predict the magnitude of resulting spring-in angle due to thermal deformations. It is found that the process control parameters can be optimized first such that the crystallization of the matrix occurs at an ideal position along the forming line. Once the process is optimized, the developed numerical model, with a thermoelastic material behaviour, can give an adequate prediction of spring-in at the end of the process. Finally, through a comparative study, it is discussed how for other manufacturing processes, such as compression molding, including a thermoviscoelastic liquid/solid material behaviour may be required to yield accurate spring-in predictions. [ABSTRACT FROM PUBLISHER]

Experimental results in the literature indicate notable non-repeatabilities during mechanical testing of woven fabrics, even under the same loading conditions, which may be linked to the presence of defects and uncertainties at meso or micro levels in yarns. Sources of such uncertainties can include both yarn dimensional tolerances and variations in the fiber/matrix material properties. The aim of this article is to conduct a systematic sensitivity analysis on the meso-level uncertainty factors in a typical woven fabric and identify the most significant factors and their interactions under a trellising mode. A finite element model capable of capturing behavior of dry yarns, along with a two-level full-factorial design approach has been employed. Factorial designs are split into two categories of the geometrical and material factors. It is shown how the obtained range of variations from the above statistical designs may be used to capture non-repeatability in some trellising tests. [ABSTRACT FROM AUTHOR]