Deconsolidation of thermoplastic prepreg tapes during the heating phase of LAFP

Abstract

Laser-assisted automated fiber placement (LAFP) is a promising additive manufacturing technique for the production of large aerospace components. Thermoplastic prepreg tapes will be placed ply-by-ply on top of a mould and in-situ consolidated. This is beneficial for higher production rates. The mechanical performance of thermoplastic composite laminates is highly dependent on the consolidation quality. One of the quality indicators is the maximum allowable void content after LAFP-manufacturing, which still exceeds the 1% limit for aerospace standards. Challenges remain before LAFP can be completely industrialized with the desired throughput and still obtaining the minimum required quality.

During the heating phase the material is heated to a processing temperature of around 400 ±C within a very short heating time (0.2-0.8s) and no pressure application. As a result of rapid heating, interconnected mechanisms can occur which affect the final consolidation quality. These mechanisms and their relation with processing parameters need to be understood to achieve a high quality laminate manufactured by LAFP. Previous research at Delft University of Technology has shown that deconsolidation phenomena, such as decompaction of the fiber reinforcement network, waviness formation and void thermal growth occur during the heating phase. This will give rise to an increase in void content, surface roughness, out-of-plane deformation and dimensional changes. However, this research did not include tape pre-tension in the experimental setup which is the main component of a LAFP tape placement head. Also, thermoplastic prepreg tapes with a resin-rich surface have been suggested in literature to contribute to a higher consolidation quality of the final laminate. The effect on deconsolidation of thermoplastic prepreg tapes with resin-rich surface during the rapid heating phase is not known yet.

Therefore, the focus of this study is on the effect of a resin-rich surface and tape pre-tension on the deconsolidation response during the heating phase of LAFP. Deconsolidation was quantified through the following response variables (output): surface roughness, maximum out-of-plane deformation, void content, thickness increase and arc-length increase. The results showed that the processing parameters (heating time, heated spot length, resin-richness and tape pre-tension) affect the deconsolidation response through similar interlinked mechanisms as were observed before. However, it has been demonstrated that the mechanisms affecting the deconsolidation response are different due to the presence of a resin-rich surface and as a result of tape pre-tension.

Suprem resin-rich thermoplastic prepreg tapes have a great potential to be used together with the LAFP- process. It has been shown that a resin-rich surface contributes to significantly less decompaction of the fiber reinforcement network. This resulted in less surface roughness and less out-of-plane deformation after the heating phase. Because the fibers are surrounded by resin, no dry fibers are popping-out of the heated surface. The smoother and more resin-rich surface are beneficial for intimate contact development during the consolidation phase of LAFP. It is therefore expected that a higher degree of effective intimate contact can be reached with Suprem resin-rich tapes which is favourable for a higher final quality of a LAFP-manufactured laminate.

Laser heating experiments were performed with three levels of tape pre-tension: 5N, 10N and 15N. Increasing the level of tape pre-tension improved the contact between the tape and the surface of the tool. Since the tool worked as a heat sink in this case, it was shown that local heat absorption occurred only at locations where fiber clusters were present. Increasing the tape pre-tension level towards 10N and 15N seem to be disadvantageous for the LAFP-process. Large out-of-plane deformation was observed and the temperature data showed a highly non-uniform temperature across the simulated nip-point (for 10N and 15N) which is unfavourable for intimate contact development during the consolidation phase of the LAFP-process. It is therefore expected that the optimum pre-tension level lies around 5N. It is assumed that global out-of- plane decompaction and global surface roughness increase will remain lower if a low (5N) pre-tension force is applied.