Poor quality meshes are meshes containing highly skewed cells, highly non-orthogonal cells, non-convex cells, or cells with left-handed faces. Such mesh elements tend to decrease the numerical stability of traditional CFD discretization algorithms. These mesh elements require special treatment, namely, a numerical correction of the transport equation discretization, which is intended to improve the numerical properties of the solution algorithms at mesh cells of poor quality.
Note: If your mesh exhibits any type of poor quality except
left-handed faces, you can maintain second order accuracy by using
the warped-face gradient correction and disabling poor mesh numerics.
To disable poor mesh numerics, set solve/set/poor-mesh-numerics/enable? to no. For additional information on this
method, see Warped-Face Gradient Correction.
In order to facilitate solution convergence on meshes of poor quality, the ANSYS Fluent solver can apply a local solution correction, limited spatially to distorted cells of the mesh. If you read in a poor quality mesh containing cells and faces with corrupt metrics, you will see a warning in the TUI of the form:
Info: The mesh contains elements that are invalid or of poor quality.
A different numerical scheme will be applied to these elements,
which may affect the quality of the solution. It is recommended
that you consider removing the invalid and poor quality elements
in the mesh.
For more information on the invalid and poor quality elements,
please use the following TUI commands:
/mesh/check
/mesh/repair-improve/report-poor-elements.
Additionally, in the Solution Methods task page, the Report Poor Quality Elements button will appear, which can be used to report more statistics on the number and type of cells that ANSYS Fluent has identified as having poor quality (see Figure 28.52: Reporting Poor Quality Elements).
In order to facilitate solution convergence on meshes of poor quality, the ANSYS Fluent solver can apply a local solution correction, limited spatially to distorted cells of the mesh. The corrected solution can be of 0th, 1st, or 2nd order:
The 0th order scheme applies an algorithm that computes the solution variable for the transport equation in the bad cells by assembling the solution directly from the surrounding solution in the better quality cells.
The 1st order scheme applies locally low order discretization methods and neglects some non-orthogonal contributions to the gradients when computing the diffusive fluxes.
The 2nd order scheme only modifies the numerics in the bad cells by assembling the gradient vector for the given solution variable from the gradients in the surrounding better quality cells.
In other words, the discretization error will be independent of the mesh size in this region if 0th order is used. It will decrease linearly with subsequent grid refinement when 1st order is used, and quadratically when the 2nd order option is selected.
By default, Fluent will apply the 1st order correction scheme to any invalid cells. Using text commands, you can further control the application of the poor mesh numerics by using one of the other schemes, including highly skewed and highly non-orthogonal cells, and including cells from a user-defined register. The text commands that are available are as follows:
/solve/set/poor-mesh-numerics/enable?
Choose whether or not to apply the poor mesh numerics corrections and specify which order scheme to use.
In general, it is recommended that you use the 1st order solution correction, which provides a reasonable compromise between accuracy and stability gain. For meshes of better (and yet still low) quality it is advisable to try the 2nd order option, which will preserve the mesh convergence behavior provided by the convection term discretization schemes. In case no convergence is obtained with any of the aforementioned schemes, you should use the 0th order option, which will provide the highest stability and at the same time the lowest accuracy. In regions with highly nonlinear flow physics this scheme can yield highly nonphysical results.
The computational time (or cost or run time requirements) for all three schemes will increase linearly as the number of cells identified as poor quality cells increases. Provided that there is reasonable convergence behavior with all three schemes, the 1st order option will be the fastest; 0th and 2nd order schemes will have similar run times to each other.
/solve/set/poor-mesh-numerics/cell-quality-based?
Choose whether or not the poor mesh numerics should also be applied to cells that are highly non-orthogonal or highly skewed.
/solve/set/poor-mesh-numerics/set-quality-threshold
Set the threshold for including cells based on cell
quality metrics when cell-quality-based? is
enabled. The value you enter is applied as a threshold for the cell
orthogonality and the complement of cell skewness below which the
poor mesh numerics will be applied. The default value is 0.05. Some cases may require tuning this parameter to
improve stability.
/solve/set/poor-mesh-numerics/user-defined-on-register
Answer yes at the prompt to
Include cells from a register in the poor mesh numerics treatment.
You will be prompted for the name or ID of the register to include.
Note that if you answer no at the prompt any
cells previously included from registers are removed from the list
of cells for poor mesh numerics.
If you include a user-defined
register in the poor mesh numerics, reporting of poor mesh element
statistics will also include the cells added by the user-defined-on-register command. For example:
Poor Mesh Element Statistics: Identified 0 faces with too small area. Identified 0 faces adjacent to negative volume cells. Identified 0 faces adjacent to bad quality cells. Identified 43 cells from user-defined register.
Note: The statistics not only report the cells included in the register, but also their neighboring cells, resulting in an overall larger number of cells being identified.
Important: It is your responsibility to maintain the register-based cells.
If you change the mesh, (for example, using mesh manipulation, adaptation,
and so on), then you must use the user-defined-on-register command again. Enter yes to the command
prompt in order to update the cells based on a new register, otherwise,
enter no to remove all cells from the poor
mesh numerics list. For more information about registers, see Registers.
/solve/set/poor-mesh-numerics/print-poor-elements-count
Print out a listing of the poor cells for each criterion (default, cell quality, and user-defined).
> /solve/set/poor-mesh-numerics/print-poor-elements-count - Number of marked cells based on default criteria = 5 - Number of marked cells based on cell quality criteria = 11 - Number of marked cells based on user-defined criteria = 0
/solve/set/poor-mesh-numerics/reset-poor-elements?
Reset the poor cells included from user-defined registers and cell-quality metrics. The poor cells identified from the internal mesh check will be retained.