What Is a Mesh Analysis?

Mesh division is to divide the model into many small units. As the most important pre-processing of finite element analysis, the degree of matching between the mesh division and the calculation target and the quality of the grid determine the quality of the later finite element calculation.

1. Conversion expansion method. If it is a curved surface and the shape is relatively regular, you can use the transformation expansion method. The transformation expansion method starts with nodes, expands into line units, then expands into planar two-dimensional units, and then expands into three-dimensional units. The meshes generated in this way are relatively high-quality and fast, and can not only generate three-dimensional meshes, but also one-dimensional meshes and two-dimensional meshes. You can also move,

1. Observe the model. Models created by ourselves or CAD software and imported into the analysis software must be observed and processed first. For example, the model is symmetric, and we'd better keep only part of the model, so that the number of meshes will be small and the calculation time will be reduced. When setting the boundary conditions, it can be set to symmetrical.

2. Model boundary line processing. For example, if there is a welding line, we must cut a boundary line artificially at the junction when dividing. After the grid is divided in this way, a grid line is naturally left here, which is in line with the actual situation. Sometimes the same is done for welding lines, and then rigid elements are connected according to the nodes.

3. Select the element type and order. The best case is to divide into hexahedral meshes, with a small number and fast solution. But sometimes the model is complicated and cannot be divided into hexahedrons, so it has to be divided into tetrahedrons. Tetrahedral high-order elements also have high accuracy, but compared with low-order elements, the solution time is also long. This requires you to weigh your choices yourself.

4. Grid density. The density of the grid is large, the grid is dense, and the solution accuracy is high, but the solution cost is also increased. This is how we deal with it. Where the parameter gradient is required, we need to use a dense grid, for example, where the stress is concentrated. If there is no such thing as a lost net, the result is often incorrect.

5. Number of grids. The higher the density, the larger the number, and the longer the solution time. We need to control this number. Within the range of accuracy and under comparable circumstances, we can reduce the number appropriately to reduce the calculation time. For structural analysis, the density of key points should be large. For modal analysis, try to unify the size and consider the number. ^[2]

The unit type mainly refers to the shape of the unit. When analyzing, you should consider the requirements for the characteristics of the unit, the calculation cost, and the calculation accuracy to select the appropriate unit type.