Multi-objective optimization of fiber orientation and grammage of recycled carbon fiber felt

It is of great significance to realize the recycling and reuse of waste composite materials. Traditional recycling methods consume high energy and are easy to damage the material structure of the fibers. The recycled fibers are oriented in a disorderly manner and cannot be utilized for high value. In this research, the new recyclable epoxy resin developed by our team was used to manufacture type IV hydrogen storage vessel, and the efficient recycling method of carbon fiber from waste composite hydrogen cylinder was studied. The fiber wet orientation device was designed independently to prepare oriented fiber felt. The effects of various process in the wet orientation process on the quality of regenerated fiber felt samples, including orientation degree and weight, were explored. The response surface method (RSM) was used to establish the objective models related to the orientation and grammage of the fiber felt, and the reliability of the models were analyzed. Based on the models, the non-dominated sorting genetic algorithm Ⅱ (NSGA-Ⅱ) was used to perform multi-objective optimization of the parameters of fiber wet orientation process. The optimal solution was selected using technique for order preference by similarity to ideal solution (TOPSIS), and experiments were designed to verify the results of multi-objective optimization. The results show that the impact of various process conditions on the orientation of fiber felt is ranked as fiber length > fiber content > dispersant content > filter mesh hole size. And the order of the impact of various process conditions on the grammage of fiber felt is fiber content > dispersant content > filter mesh hole size > fiber length. The objective function models are very reliable for analyzing the orientation and grammage of the fiber felt. Through the multi-objective optimization algorithm, the optimal process parameters for the preparation of regenerated fiber felt are fiber length 3 mm, fiber content 6.37 g/L, dispersant content 13.37 g/L, and filter mesh hole size 0.75 mm. The orientation degree of the fiber felt prepared by verification experiment is 81.08%, with an error of 0.94% from the orientation degree(81.84%) predicted by genetic algorithm. The grammage of the fiber felt experimentally prepared is 42.86 g/m², with an error of 0.68% from the grammage(42.57 g/m²) predicted by genetic algorithm.