Accurate simulation of bolted joints is not always consistent with industrial requirements, since numerous nonlinearities in the vicinity of the bolt can lead to overly expensive calculations. For this reason, commercial finite element (FE) codes preferentially use substitutes for bolts, such as simplified models or connectors. In this paper, a nonlinear FE connector with its identification methodology is proposed to model the behavior of a bolted assembly. The connector model is based on practical design parameters, such as bolt preload, friction coefficients, or plastic material parameters. The connector is based on a separation of the different phenomena governing the macroscopic behavior of a bolted assembly. The axial behavior of the connector reflects the preload effect and the axial stiffness of the assembly, while the tangential behavior of the connector takes into account the friction phenomena that occur in the vicinity of the bolt. At the same time, a plastic behavior law with isotropic hardening for the bolt is proposed. The identification of the connector parameters is performed on a generic elementary one-bolt assembly. The connector is implemented in ABAQUS through a user-element subroutine. Comparisons of quasi-static responses between large-scale full 3D calculations and simulations with connectors on different bolted assemblies are provided. The results obtained are very similar, while observing a significant reduction in CPU time.

螺栓连接的精确模拟并不总是与工业要求一致，因为螺栓附近的许多非线性会导致计算成本过高。因此，商业有限元（FE）代码优先使用螺栓的替代品，例如简化模型或连接器。在本文中，提出了一种非线性有限元连接器及其识别方法，以对螺栓装配的行为进行建模。连接器模型基于实际设计参数，例如螺栓预紧力，摩擦系数或塑料材料参数。连接器基于控制螺栓组件宏观行为的不同现象的分离。连接器的轴向性能反映了预紧效应和组件的轴向刚度，而连接器的切线行为考虑了在螺栓附近发生的摩擦现象。同时，提出了螺栓各向同性硬化的塑性行为定律。连接器参数的识别是在通用基本单螺栓组件上执行的。该连接器通过用户元素子例程在ABAQUS中实现。提供了在使用不同螺栓连接的连接器上进行的大规模完整3D计算和仿真之间的准静态响应的比较。获得的结果非常相似，同时观察到CPU时间的显着减少。连接器参数的识别是在通用基本单螺栓组件上执行的。该连接器通过用户元素子例程在ABAQUS中实现。提供了在使用不同螺栓连接的连接器上进行的大规模完整3D计算和仿真之间的准静态响应的比较。获得的结果非常相似，同时观察到CPU时间的显着减少。连接器参数的识别是在通用基本单螺栓组件上执行的。该连接器通过用户元素子例程在ABAQUS中实现。提供了在使用不同螺栓连接的连接器上进行的大规模完整3D计算和仿真之间的准静态响应的比较。获得的结果非常相似，同时观察到CPU时间的显着减少。