A mesh independence study is a process used to determine the optimal mesh size for a specific region of a geometry. This is important because a mesh that is too coarse will result in inaccurate results, while a mesh that is too fine will result in a waste of computational resources. Here are the steps involved in a mesh independence study:
1. Identify the region of interest on the geometry. If the geometry is small, the whole model can be used. However, for larger models, it is more practical to focus on a specific area.
2.Determine a way to quantify the nodes on the region of interest. This can be done by counting nodes on a radius, chamfer, body, or a named selection.
3.Refine the mesh in the region of interest. It’s important to only change the mesh sizing and not other mesh settings to ensure a fair comparison between runs. A minimum increase of 10% in mesh count is recommended between runs. Obtaining at least three different meshes is typical for establishing mesh independence.
4.Choose a result to study, such as a stress value, equivalent stress, stress component, or one of the principal stresses. Other results such as plastic strain, accumulated damage, or fatigue life may also be chosen depending on the scenario.
5.Compare the results for the various runs by plotting the number of nodes vs the stress. Look for a point where a significant increase in node count does not result in a significant increase in stress. The analyst may plot raw data, percentages, or both.
Check out the picture down below – it’s a mesh independence plot! Basically, it tells us how many nodes we need to get an accurate stress reading. In this example, we need at least 5,000 nodes to get a stress value of 690 MPa.
Mesh Independence Study – Raw results, Number of nodes vs Stress
FYI, when we use a rough mesh, it often leads to lower stress values than what’s actually happening. But sometimes, the opposite happens too (weird, right?). We’ll cover how meshes affect stress in another post – stay tuned!