| Three-dimensional structured preforms containing Z-directional fibers are the reinforcement of advanced composite materials,and the molding process of the material directly affects the final performance of the material.The fibers in X,Y and Z directions inside the 3D woven preforms are interwoven to form a threedimensional mesh structure.The Z-directional fibers are subjected to different degrees of frictional wear during the implantation process leading to fracture,which affects the forming quality and efficiency of the preforms.With the rapid development of the aerospace industry,the Z-directional properties of fiberreinforced composites have put forward higher requirements,and a new process method for enhancing the Z-directional properties of composites is urgently needed.This paper focus on high quality and high efficiency implantation of Zdirectional fibers into preforms to improve the mechanical properties of 3D woven composites,and investigate the friction distribution law at different locations of preforms,the mechanism and method of low damage implantation of fibers,the mechanism of enhancing the tensile and shear properties of composites,and the digital implantation process,method and equipment of Z-directional fibers.Firstly,two BP neural network models were developed to enhance the continuity of Z-directional fiber implantation into preforms,one is a prediction model for the distribution of Z-directional fiber implantation force by inputting the number of preform layers,the number of guide sleeve rows and the number of guide sleeve columns,the other is a model for the prediction of Z-directional fiber tensile strength by inputting the number of Z-directional fiber implantations and the implantation force.The relationship between the tensile strength of the fibers and the implantation position and number of implantations is revealed,and the critical strength value for fracture of the fibers is determined,effectively avoiding the occurrence of fracture during Z-directional fiber implantation.Secondly,to address the problem that the Z-directional fibers rub against the preform during implantation,resulting in serious wear and affecting the mechanical properties of the composite,this paper proposes a method of anti-wear by applying dimethyl silicone oil on the surface of Z-directional fibers,and builds a platform for tension control during Z-directional fiber implantation,which effectively reduces the fiber wear.Based on Hertz’s theorem and Howell’s theory,a model of the relationship between normal force and fiber width and thickness of preform during Z-directional fiber implantation and a model of the relationship between friction coefficient and implantation force and normal force were created to reveal the mechanism of the improvement of the wear degree of the fiber after surface treatment and tension application.In order to further improve the Z-directional tensile properties of the composites,this paper reveals the effect of twisting of different K-beam carbon fiber bundles on fiber properties,establishes the relationship models of axial strain,radial strain and Poisson’s ratio of fiber bundles,realizes the prediction of tensile strength,bundle length and elongation of twisted fibers,and analyses the internal stress distribution during fiber twisting.A model for the filling factor of fiber bundles and a model for the relationship between fiber bundle length and twist was developed.A fiber volume fraction model was established for both twisted and untwisted Z-directional fiber preforms and the elastic modulus of the two composites was predicted by theoretical calculations and finite element method respectively.In combination with the flexible guided 3D woven preform forming process,the Z-directional tensile strength of the composite was increased by 60.9%by changing the Z-directional fiber type.In order to further enhance the Z-directional shear performance of the composite material,the carbon nanotube grafting method and device developed independently were used to complete the carbon nanotube grafting of large quantities of fibers,and the carbon nanotube fibers were implanted as Zdirectional fibers into the prefabricated body,and the interlaminar shear strength of the composite was increased by 90%..Finally,this paper proposes a digital fiber implantation process based on machine vision localization,designs and builds a Z-directional fiber implantation device.A deep learning-based multi-scale visual detection and positioning algorithm,including a modified YOLOv3-m model and a modified Hough-m transform model,is proposed to achieve the identification of guide sleeves in the context of large batches,small sizes and complex working conditions,to meet the demand for automatic and accurate positioning of Z-directional fibers to the axis of the guide sleeve under complex working conditions,and the degree of automation of fiber insertion is increased by 6.3 times compared to manual fiber insertion. |
