Research On The Morphology Reconstruction Of The Overall Puncture Needle Of Woven Carbon Cloth And The Uniformity Of Fiber Distributio

Carbon/carbon(C/C)composites are featured by the excellent properties such as high specific strength,high specific modulus and ablation resistance,which are important thermal structural materials in the aerospace field.Integrated piercing technique is one of the key techniques for preparing C/C composite reinforcement.The impact of the needle tip and the instability of the pressure bar are the main mechanical behaviors that affect the integrated piercing,and the fiber uniformity in the preform directly affects the properties of C/C composites.Therefore,it is of great significance to improve the reliability of the piercing steel needle and the uniformity of fiber in the preform.Taking the integrated piercing process as the research object,this paper studied the mechanical behavior of the integrated piercing steel needle and the fiber uniformity in the preform,including five parts: the mechanism of the integrated piercing process of carbon cloth,the eccentric displacement of the piercing steel needle,the curve reconstruction of the piercing tip,the interlayer uniformity of the preform,and the integrated piercing experiment of carbon cloth.(1)The mechanism of the integrated piercing process of carbon cloth was studied.Firstly,the representative cells of carbon cloth and puncture equipment were analyzed and geometrically modeled respectively.Secondly,the integral stiffness matrix of the digital fiber was derived,and the mechanism of the interaction between the steel needle and the fiber during the piercing process was modeled.The analysis results showed that the impact of the needle tip and instability of the pressure bar were the main reasons affecting the integrated piercing.(2)In order to reduce the collision between the needle tip and the hole wall of the template,a shape reconstruction method of steel needle was proposed with the aim of minimizing the eccentric displacement of the needle tip.Firstly,based on the superposition method,the displacement function of the needle tip with variable crosssection was constructed,and the correlation between the shape parameters of the steel needle and the displacement was calculated using the feature selection method.Secondly,the length and diameter of the bar were selected for model optimization based on the sorting results.The analysis results showed that the deviation of the piercing needle tip was significantly reduced after morphological reconstruction.(3)Aiming at the instability of the pressure bar of the needle tip,a reconstruction method of the needle tip curve based on the mechanical characteristics of piercing was proposed.Fourteen needle tip curves were constructed using three methods,and their puncture mechanical characteristics were evaluated.On this basis,particle swarm optimization(PSO)was introduced and a general optimization method for needle tip curve was proposed.The analysis results showed that this method could effectively improve the mechanical characteristics of needle tip piercing.(4)Aiming at the problem of fluctuation of density between layers of preforms,a prediction model of preform compaction parameters was established by introducing machine learning method.Taking the performance metric evaluation score as the standard,the k-nearest neighbor operator was selected as the base learner of ensemble learning,and the model optimized by grid search was used to predict the compaction parameters.(5)Finite element simulation and experiment.The integrated piercing simulation results showed that the morphological parameters were the main factors affecting the mechanical behavior of the piercing steel needle.Four groups of steel needles with different shapes were used for piercing experiments.The results showed that,compared with the fourth group(non-optimized),the average eccentric displacement of the other three groups of steel needles after optimization could be reduced by 18.16%～31.42%,indicating that the optimized reconstruction method could effectively reduce the impact damage of steel needles.Compared with Proportional-Integral-Derivative(PID)control,the deviation of the average layer height of the preform prepared by the prediction model was reduced from 12% to 6.8%,which proved that the method improved the interlayer density uniformity of the preform.