What is Meshing in ANSYS workbench or in FEA and CFD?

Meshing is the distribution or discretization of the body into individual small chunks called elements. In EFA or CFD, it is impossible to solve the object without dividing it into smaller elements. Each element has its own stiffness while loading. Each element of mesh has an individual solution of partial Differential Equations. Adding all those elements creates the global stiffness matrix which gives the global solution of results (Stress, Strain, Velocity, Pressure, Temperature, etc.) developed in the body.

Why is good meshing in the ANSYS workbench necessary?

As meshing is the discretization of the body into individual small elements, it can be said that mesh is an approximation of the actual geometry. The mesh quality has a great influence on the accuracy of results and stability of a simulation. Poor or irregular shape mesh generation will create poor mesh discretization of the body which is not capable to capture the actual physics of loads and results in false results.

A CFD solver or FEA solver is comfortable with regular standard-shaped mesh. Meshing with irregular shapes causes difficulties to converge the solution, and takes more time to solve the problem. That’s why good quality mesh with a standard shape is the first step to solving the problem numerically. The mesh should be said good quality if the solution converged in the expected time, mesh improves at least one important property of simulation.

How to verify mesh quality in the ANSYS workbench?

There are a few methods by which mesh quality can be determined and verified whether a mesh is acceptable or not.

  1. Mesh quality metrics
  2. Result convergence study (Graph of Residuals)
  3. Grid or Mesh independence test

Mesh quality metrics

Mesh metrics are one of the easiest and most widely useful features in the ANSYS meshing workbench that identify the correct shape and size of the elements. There are many mesh quality metrics in ANSYS meshing workbench but the most popular metrics are Aspect ratio, Skewness, and Jacobian ratio. Even, I personally use only these three metrics to check the mesh quality.

Aspect ratio of Mesh element

Aspect ratio- Ansys meshing
Aspect ratio- Ansys meshing

The Aspect ratio is the ratio of the longest length of the mesh element to the shortest length of the mesh element. The Aspect ratio determines the quality of the shape of the element. Ideally, for a perfect shape, the Aspect ratio is equal to 1. A higher aspect ratio results in the worst mesh elements.

Skewness 

Skewness is the other most widely used mesh metric used to check mesh quality. Skewness is the angular measurement of mesh quality with respect to the angular measurement of the ideal mesh element. Skewness represents the deviation between the ideal mesh elements to the generated or existing mesh elements. Generally, the acceptable range of skewness is 0 to 0.5.

Skewness and orthogonal quality ansys meshing
Skewness and orthogonal quality in ANSYS meshing | Image credit – Ansys.com

Higher skewness or highly skewed cell should be avoided because this can lead to convergence difficulties and inaccuracies in the numerical solution.

skewness in ansys meshing

Jacobian ratio

The jacobian ratio represents the deviation of generated elements from ideal shape elements. The jacobian ratio ranges from -1.0 to 1.0. For perfect shape, the jacobian ratio is 1.0

Result convergence study (Study of Residuals Graphs)

We all know that FEA or CFD software solves the problem by Numerical methods. The Numerical methods are sets of the iterative solution. The residuals are the measurements of stability or imbalances of solutions. For an accurate result, residuals should be as small as possible. Generally, residuals below 1e-3 are a good and acceptable start.

Residuals Graphs Ansys meshing
Residuals Graphs Ansys meshing | Image credit- simscale.com

The Residuals Graphs represent the imbalance or stability of the numerical solution. By observing residual graphs, the sudden jump shows that the solution faced the problem of being converged. The results obtained from such simulations are not reliable.

Grid or Mesh independence test

A mesh independence test is a study in which a Design engineer or analyst or researcher performs some simulation to determine the dependence of results on mesh density. A mesh independence test is an effective method for identifying the mesh quality by refining meshing in ANSYS or any simulation software until the satisfactory result in the interested location converges.

The major drawback of the mesh independence test is, it is a time-consuming method as it requires multiple re-meshing and re-solving of problems. The mesh independence test is suitable for simple problems but it is very time-consuming for complex problems. However, In the ANSYS workbench, you can perform the mesh independence test automatically by using the Convergence tool option.

Graph and result photo

One thing remembers is that the Mesh independence test is performed on the basis of mesh density only, not on mesh type or quality.

Example: You performed a simulation with mesh size = 5mm with tetrahedrons mesh and obtained results. Now, during the mesh independence test, you should just change the size of the mesh from 5 mm. you should not change the types of mesh which is tetrahedrons mesh. Then, compare the results with previous results.

How to improve Mesh quality?

There are a few tips that improve the mesh quality.

Clean CAD Geometry

Before importing the model in the simulation wizard, keep the CAD model as simple as possible. But remember that you cannot compromise with its design. Make sure that the CAD model includes all shapes on which physical effects need to check.

For example – The name imprinted on an object can be deleted or suppressed this feature because you may not need to study the physical effect of that name. Remove unnecessary holes made on an object that is modeled for bolting or weight reduction. But remember that, the hole size should be small compared to the object It will not affect it too much with results. Simplify the object near the simple shape as much as possible.

By doing this, software avoids creating a mesh that will save meshing time, and the possibility of un-stability during solution may reduce.

Remove the sharp corners to avoid high skewness. If there is a high skewness in the model then just chop off or just apply a little fillet.

method to reduce skewness

During the simulation of assembly, two or more parts are shared their face with each other. Different faces share their nodes with adjacent faces. In that case, apply “Share topology”. This option is available in ANSYS SpaceClaim (Backlink) in the ANSYS workbench. Share topology helps to create a conformal mesh where body parts share their faces with adjacent faces. Share topology is the one way to ensure that mesh is created perfectly at the intersections.

Leave a Reply

Your email address will not be published. Required fields are marked *