Structural Design And Mechanical Properties Prediction Of Graphene With Negative Poisson’s Ratio Based On Machine Learning And Kirigami Technology

The kirigami technique has become a prominent approach for tailoring of material properties and show great promise for the fabrication of mechanical metamaterial with customizable properties by introducing rationally designed geometric structures.Special kirigami patterns have been shown to induce auxetic behavior in graphene,for example,negative Poisson’s ratio,which can potentially further expand the application of graphene in flexible and elastic electronic devices.However,how to design graphene kirigami structures is still an open topic,and the influence of the kirigami geometric parameters on the mechanical properties of graphene has not been fully understood yet.In this work,two types of auxetic graphene kirigami structures are fabricated,which corresponds to two typical mechanisms inducing negative Poisson’s ratio: rigid unit rotation mechanism and rib bending deformation mechanism.Machine learning approach is adopted to analyze and predict the influence of kirigami structure parameters on the negative Poisson’s ratio of graphene,which shows great precision with reduced computational costs and may speed up of the design process.The main content of this thesis is as follows:1.The influence of geometrical perforation parameters of four different perforation types(rectangle,diamond,ellipse,and peanut-shaped)and the associated patterns on the Poisson’s ratio of graphene is investigated.The results show that both rectangular and diamond-perforated kirigami graphene exhibits auxetic behavior,which is caused by the presence of perforation voids that leads to the rotation of rigid units and is also accompanied by out-of-plane deformation of graphene.As the aspect ratio(AR)of the perforation increases,the auxetic behavior of the graphene kirigami is enhanced,while the auxetic behavior decreases as the spacing ratio(IS)of the perforations increases.The predictive mechanical model built using multilayer perceptron algorithm shows that the spacing ratio(IS)of the perforations has a greater impact on the simulation results.The mechanical properties of the kirigami can be effectively controlled by adjusting the modeling parameters.By comparing the effects of different perforation shapes on the auxetic behavior,it was found that the rectangular-perforated graphene kirigami exhibits the strongest auxetic behavior,while peanut-shaped perforated graphene kirigami exhibits the weakest auxetic behavior.2.The auxetic behavior of double-arrowhead and concave hexagonal graphene kirigami,as well as the influence of kirigami geometric parameters on the auxetic behavior is further studied.The results show that the auxetic behavior of both types of graphene kirigami is originated to the fact that the tensile force acting on the diagonal ribs during stretching is not along the same action line and then results in a moment that causes the diagonal ribs to undergo bending deformation.For the concave graphene kirigami structure,the major factors that affect its auxetic behavior are the length and wall thickness of the diagonal ribs.When the length of the diagonal ribs increases,the intensity of local bending during the stretching processes also increases,which leads to stronger auxetic behavior.When the wall thickness increases,it will enlarge the individual diagonal rib resistance to bending during the stretching process,which results in weakened auxetic behavior.At the same time,the comparison results between the actual values and predicted values also indicate that the neural network algorithm can accurately predict the target values,which can be quickly obtained through the provided parameters.