Firstly, each component is modeled separately, the finite element model of each subcomponent is assembled, and then the corresponding degrees of freedom of the joint are coupled at the corresponding parts of the bolts of the respective components, thus establishing a rough model of the structure; o rough The model applies constraints and loads to perform the initial calculation of the rough model; a sub-model containing detailed geometric features is generated for the local regions of the bolted joints of interest. The position of the submodel should be the same as the corresponding part of the rough model; cut boundary interpolation is provided. This step is the key to the submodel approach. The user defines the nodes of the cutting boundary. The program uses the rough model results to interpolate the degrees of freedom (such as displacement) at these points, and uses this value as the boundary condition of the submodel; calculates the submodel; verifies the distance between the cutting boundary and the stress concentration region. It should be far enough. Verification by comparing whether the results (stress, etc.) on the cutting boundary are consistent with the results of the corresponding positions of the rough model. If not, re-define and calculate the sub-model by redefining the cutting boundary farther from the area of ​​interest. The advantages of this method are mainly: when the model has a large number of bolted joints, it is not necessary to establish detailed features such as screw holes, which greatly reduces the workload of modeling; and when meshing, there is no need to finely draw near the screw holes. The area thus greatly reduces the number of units. Avoid the problems caused by the finite element model of handling a large number of bolts separately. For the area of ​​interest, a model that is closer to the actual one can be established, and a finer grid can be divided for calculation, regardless of the influence of the number of units on the calculation. Not only can the constraints of hardware and software conditions in complex structural analysis be solved, but also the analysis results with higher precision can be obtained. The calculation example combines the requirements of the above coupled node sub-model method to calculate the quasi-static stress distribution of a cabinet, and investigate the strength reserve of the electronic equipment rail fixing parts in the vicinity of the bolt connection. There are four sides on the left and right sides of the rail rail fixing parts of the cabinet. The joints of the left and right front pillars and the left and right front fixing members of the cabinet are connected by 21 bolts. The rail rail fixing parts of the electronic equipment and the front fixing parts of the cabinet pass through two. The bolts are coupled together, and the rail fixing members are required to be connected to the front pillars for different sizes and different positions, and three pairs of mounting holes are reserved on each side, that is, four pairs of holes on each side. There are five pull members on the left and right sides of the cabinet, which are connected with the front and rear columns through bolts. At work, the cabinet will withstand up to 9g of acceleration. The calculation procedure using the coupled node submodel method is as follows: the cabinet rough model 11 establishes and calculates the rough model. When the mesh is divided, the bolt joints should be taken into account, so that when the node degrees of freedom are coupled later, the position of the nodes can be compared with the actual The bolt positions correspond. The degrees of freedom of the corresponding nodes are coupled at the bolt joints. In this example, the six corresponding degrees of freedom of the corresponding nodes are all coupled. This establishes a rough model of the entire cabinet, as shown. The model uses a shell-and-shell unit. The rough model has a total of 44,515 nodes, 32,831 units, and 2,481 coupling points. After the rough model of the cabinet is established, first calculate the first few modes of the model, verify the accuracy of the model, whether the bolt connection is missing, and whether there is overlap when coupling the nodes. It is calculated that the first-order natural frequency of the whole main cabinet is 15.9 Hz, which is the lateral bending mode of the whole cabinet. This is in good agreement with the first-order natural frequency of about 15.0 Hz obtained by the experimental test, indicating the finite element model of the whole cabinet. Very high precision. This way the model is considered to be basically accurate. Apply a static calculation of 9g to the rough model to see the strain and stress distribution of the entire cabinet. When the sub-model is established to establish the sub-model, the position where the joint position of the rail fixing member and the front fixing member is far away and the stress variation gradient is not large is selected as the cutting boundary. The submodel is directly taken from the rough model by cutting. A comparison graph of the finite element model of the position of the sub-model and the rough model and the sub-model is shown. These two bolts are simulated using solid elements. The sub-model after meshing has a total of 13486 nodes and 3118 units. Verification Submodel In this study, it is considered whether the selection of the cutting boundary is appropriate by comparing the VONMISSES stress of the node on the cutting boundary with the node on the rough model. The following table lists the stresses of the corresponding nodes of the rough model and the sub-model on the cutting boundary. The VONMISSES stress of the corresponding node on the cutting boundary of the two models compares the stress value of the cutting boundary node. Finally, the maximum equivalent stress value of the rail fixing member in the vicinity of the screw hole of the joint is 280.7 MPa. The entire calculation is completed.
Application of integrated frame point simulation method in the analysis of bolts and sequence equipment