An experimental investigation of the spreading behaviour of carbon-fiber rovings

This work investigated the production of UD tapes based on a continuous extrusion process: Multiple fiber rovings are unwound from a creel and spread mechanically to a certain width to form a homogeneous fiber carpet. The fiber carpet is then impregnated with thermoplastic polymer melt while being pulled through an extrusion die. Impregnation between the fibers must be complete to ensure good fiber-matrix adhesion in the tape, which is essential for transferring mechanical loads in the final application. In the downstream processing steps, the tape is calibrated to the final thickness, cooled, and rolled up. An integral step of the process is fiber spreading. A perfectly spread and homogeneous fiber carpet is a prerequisite for complete fiber impregnation and thus a sufficiently high tape quality. Various techniques may be used to spread rovings from their initial to the required width. From these, we systematically studied the factors that affect deflection-based fiber spreading, which is a technique that applies higher stress to the outer roving filaments, they evade to smaller radii, which widens the roving. A test rig was developed to investigate fiber spreading under near-process conditions. Using a carbon fiber roving, numerous spreading tests were carried out in which the roving was guided over deflection rods and the evolving roving width was recorded with cameras before and after spreading. In these experiments, (i) the number of rods, (ii) the rod diameter, (iii) the rod immersion depth, and (iv) the roving take-off speed were varied. It could be shown that the spreading behavior is mainly affected by the number of rods and to a lesser extent by the immersion depth. The results of the test-rig experiments shall be used to optimize the spreading configuration in the industrial production of UD tapes later on.