柔性防护系统环形拦截网分区等代模型

环形网是被动柔性防护系统中的重要结构单元，其冲击变形具有明显的多柔体非线性动力学特征，其计算模型的构建是柔性防护结构理论研究的难点与关键点. 为了建立环形网的计算模型，首先通过柔性被动网整体结构足尺冲击试验，获取了环形拦截网的区域化变形特征以及网环的典型变形状态；其次通过网环拉伸试验与数值仿真计算的方式得到了各典型变形状态下，网环的荷载位移关系,建立了分区等代计算模型；最后通过与网片顶破试验、动力冲击试验展开对比计算的方式校验了模型的准确性. 研究结果表明:环形拦截网具有明显的分区域变形特征，网环（环形网的基本单元）具有3种典型变形状态，各变形状态的网环均具有两阶段受力特征；据此采用连接于圆心、初始长度为网环半径的“X”形4桁架单元模拟单个网环，将典型变形状态下网环的荷载位移关系等效为桁架单元应力应变关系，连接对应节点形成整片网的单元模型，根据网环所处的变形区域赋予桁架单元相应的应力应变关系，建立了整片环形网的高效率分区域等代模型；通过与静力试验及动力试验的对比，模型计算误差小于10%. 

Ring-Net Subdivision Equivalent Model of Flexible Protection System

The ring-net is an important interception component in passive flexible protection systems, and it has obvious nonlinear dynamic characteristics. Constructing a mechanical model of the ring network is the key problem in the theoretical study of the flexible protection structure. To propose an equivalent mechanical model of the ring net, an full-scale impact test was conducted with an actual structure, and the regional deformation characteristics of the ring-net and typical deformation state of the ring-net were obtained. Then, the load-displacement relationships of the net ring under typical deformation states were obtained using a tension test and numerical simulation calculation; on these bases, a subdivision equivalent calculation model was established.. Finally, the accuracy of the model was verified by comparison with the breaking and dynamic impact tests. Results show that the deformation of the ring net has obvious regional characteristics, and the ring-net (basic unit of the net) displays three typical deformation states (each with two-stage characteristics). Based on this, an ‘X’ type 4-truss element, connected to the centre of the circle with the initial length of the ring radius, is used to simulate a single ring. The load-displacement relationship of the ring in a typical deformation state is treated as equivalent to the stress-strain relationship of the truss element, and the corresponding nodes are connected to form the element model of the entire ring-net. A comparison with the static and dynamic tests indicates that the model calculation error is less than 10%.