“…However, only a single winding trajectory was simulated, rather than a path showing the width of the yarn. Several works of the literature (Hongya et al, 2007;Xianfeng et al, 2010;Zu et al, 2012) used a similar approach to simulate the winding process for other mandrel shapes, such as toroidal tanks, S-elbows, aircraft inlets, and vanes. Nonetheless, they were not real filament winding simulations because no consideration was given to the cross-section shape of yarn in the model.…”