In recent years, continuous fiber-reinforced thermoplastic composites have attracted attention for their high mechanical properties, secondary workability, and recyclability. However, it is not easy to impregnate thermoplastic resin into reinforcing fiber bundles because the melt viscosity of thermoplastic resin before curing is higher than that of thermosetting resin. One of the methods to solve this problem is to use intermediate materials, and this study focused on a commingled yarn by using the fiber tow spreading-commingling technology (hereinafter referred to as spread commingled yarn). In this study, spread commingled yarns were produced under different fabrication speed conditions, and the impregnation properties of molded products using the spread commingled yarns were investigated to obtain guidelines for high-speed production of highly impregnated spread commingled yarn. The effects of un-impregnation ratio of the molded products on the mechanical properties of the molded products were also investigated. The results of this study revealed that increasing the dispersion ratio of carbon and resin fibers was most important for the use of spread commingled yarn as an intermediate material for molding. In order to increase the dispersion ratio, it was important to increase the f/v value, which was the vibration frequency of the transducer divided by the spread commingled yarn production speed under conditions where the transducer and carbon fibers were in contact. The quantitative demonstration of the boundary condition that switch between contact and non-contact suggests a design guideline for improving the production rate of highly dispersed spread commingled yarn.

A conventional shuttle loom was equipped with an industrial camera to capture textile images while weaving. Both defect-free textiles and four types of defect textiles were weaved for five kinds of patterns. The defect textiles were intentionally added the defects at the weaving stage. A unique convolutional autoencoder was individually trained on the normal image data of these textiles, and differences were taken between defective input and output images using a trained model. Model accuracies of defect segmentation were verified by calculating the AUC values taken from residual maps and ground truths prepared by visual inspection. The time until the model achieved satisfactory accuracy was also examined. The resulting model achieved extremely high accuracy for specific defects, regardless of patterns. Longer one repeating pattern required longer training time, although training time was short enough to train while weaving. Thus, this study showed the feasibility of an automatic learning defect segmentation system to train for each pattern individually.