【作者】 杨晓东；

【导师】 李成； 时建纬；

【作者基本信息】 郑州大学 ， 机械设计及理论， 2023， 硕士

【摘要】 碳纤维增强复合材料(Carbon Fiber Reinforced Polymer,CFRP)是由有机碳纤维与环氧树脂等基体组合而成的复合材料,因为其具有许多的优点,所以在许多领域都有广泛的应用,同时复合材料制作的连接结构也得到了广泛的关注,连接结构作为复合材料部件之间的传力枢纽,决定了整体结构的承载能力和可靠性,因此对CFRP连接结构的分析和设计具有十分重要的意义。本文结合实验分析、有限元仿真以及强化学习算法,对CFRP制作的胶螺混合连接结构(Hybrid Bonded-bolted Joint,HBBJ)的力学性能分析和设计进行了系统的研究。本文的主要研究工作包括:首先建立了HBBJ的有限元模型,并基于实验数据验证了模型的有效性。根据HBBJ的失效形式,证明了三个螺栓的载荷分布是不均匀的,发现载荷分布大的螺栓孔附近的层合板会首先出现损伤导致HBBJ失去最佳连接性能,且间隙配合关系对HBBJ的载荷分布有显著的影响。其次基于有限元分析,揭示了搭接长度、螺栓孔间距和间隙配合关系的改变对HBBJ载荷分布不均匀度与极限失效力的影响规律,发现载荷分布不均匀度与极限失效力之间呈负相关,且搭接长度、螺栓孔间距能大幅提高HBBJ的极限失效力,但无法使载荷分布均匀;间隙配合关系虽能使载荷分布均匀,但对极限失效力影响幅度不大。因此不能通过改变单一结构参数使HBBJ达到最佳的连接性能。最后开发了基于强化学习的HBBJ力学性能设计方法,实现了HBBJ的极限失效力预测、载荷分布预测、载荷均匀分布设计、最大化和指定的极限失效力设计。同时研究了该方法的设计效率和设计过程中的分布转移,并通过有限元仿真验证了设计结果的准确性。该方法能够通过很小的计算代价设计出符合期望的结构参数,验证了其在设计问题中的可行性。本研究结合实验、有限元分析和强化学习算法,研究了HBBJ结构参数与力学性能间的影响关系,并开发了可实现更多分析和设计功能的方法,研究成果为HBBJ更广泛的应用和更复杂的设计奠定一定的基础。更多还原

【Abstract】 Carbon fiber reinforced polymer(CFRP)is a composite made of organic carbon fiber and epoxy resin matrix.Because of its many advantages,it is widely used in many fields.At the same time,the connection structure made of composite materials has also been widely paid attention to.As the transmission hub between composite components,the connection structure determines the bearing capacity and reliability of the whole structure,so the analysis and design of CFRP connection structure has very important significance.The mechanical property analysis and design of the Hybrid Bonded-bolted Joint(HBBJ)made of CFRP were systematically studied by combining experimental analysis,finite element simulation and reinforcement learning algorithm.The main research work include:Firstly,the finite element model of HBBJ was established,and the validity of the model was validation based on the experimental data.According to the failure mode of HBBJ,it is proved that the load distribution of the three bolts of HBBJ is uneven,and the laminate plate near the bolt hole with large load distribution will be damaged first,leading to the loss of the best connection performance of HBBJ.It is found that the clearance fit relationship has a significant effect on the load distribution of HBBJ.Then,based on the finite element analysis,the influence law of the change of overlap length,bolt hole spacing and clearance fit relationship on the load distribution unevenness and ultimate failure force of HBBJ was revealed.The negative correlation between load distribution unevenness and ultimate failure is revealed.It is found that overlap length and bolt hole spacing can greatly improve ultimate failure force of HBBJ,but the load distribution can not be uniform.Although the clearance fit relationship can make the load distribution uniform,it does not affect the ultimate failure force of HBBJ greatly.Therefore,it is not possible to achieve the best connection performance by changing a single parameter.Finally,the analysis and design method of mechanical properties of HBBJ based on reinforcement learning is developed,and the ultimate failure force prediction,load distribution prediction,uniform load distribution design,maximum and specifiedAbstractultimate failure force design of HBBJ are realized.The design efficiency of the method and the distribution transfer in the design process are studied,and the accuracy of the design results is validation by finite element simulation.The structure parameters can be designed with a small computational cost,which proves the feasibility of the method in the design problem.In this study,experiments,finite element analysis and reinforcement learning algorithm were combined to study the influence relationship between the structural parameters and mechanical properties of HBBJ,and to develop a method to realize more analysis and design functions.The research results lay a foundation for more extensive application and more complex design of HBBJ.更多还原