You have access to many parameter settings that can be used to control or improve the performance and convergence behavior of the CHEMKIN-CFD solver within Ansys Fluent. These parameters are accessed through the Ansys Fluent Text User Interface (TUI), through typed text commands. This appendix describes the available parameters, their default values, and their recommended use.
In general, default settings for the CHEMKIN-CFD parameters should be sufficient for most problems. However, it may be necessary to change these settings to improve performance for specific types of problems. In particular, the most appropriate solver parameters may vary depending on the nature of the chemistry set.
For example, for chemistries with many trace species that have a large impact on solution results (that is, the chemistry is “stiff”), solution tolerances should be relatively small (for example, 1.E-9 to 1.E-12 for absolute tolerances and 1.E-4 to 1.E-5 for relative tolerances). For chemistry descriptions where all of the species included are within a few orders of magnitude in terms of mass fraction, however, much looser tolerances are more appropriate. Also, depending on the characteristic time scales of the dominant chemical reactions, you may need to adjust the maximum time-step used by the CHEMKIN-CFD solver.
/define/models/species/CHEMKIN-CFD-parameters> basic-options Print Level [0] Absolute Tolerance For Gas Species [1e-08] Relative Tolerance For Gas Species [0.0001] Absolute Tolerance For Surface Species [1e-08] Relative Tolerance For Surface Species [0.0001]
/define/models/species/CHEMKIN-CFD-parameters> advanced-options Pseudo Time Stepping Solver Option (0:Basic 1:Robust) [1] Pseudo Time Stepping Size [0.0001] Maximum Pseudo Time Step Attempts [1] Maximum Time Step Allowed [0.01] Minimum Time Step Allowed [1e-10] Initial Time Step For Solver [0] Upper Bound For Gas Species Mass Fractions [1.01] Lower Bound For Gas Species Mass Fractions [-1e-12] Upper Bound For Surface Species Mass Fractions [1.1] Lower Bound For Surface Species Mass Fractions [-0.0001] Small Positive Value To Replace Negative Mass Fractions [1e-12] Enter Auxiliary Input Filename [""] Enter Arbitrary String [""] Enter Arbitrary Value [0]
Table 1: Summary of Basic CHEMKIN-CFD Parameters and Table 2: Summary of Advanced CHEMKIN-CFD Parameters summarize basic and advanced parameters, respectively. These tables include descriptions of the default values and physical units, where applicable, for each of the CHEMKIN-CFD parameters, as well as a discussion of the recommended parameter usage.
Table 1: Summary of Basic CHEMKIN-CFD Parameters
Parameter | Usage | Default Value |
|---|---|---|
| Controls the level of diagnostic printing to the CHEMKIN-CFD diagnostic output file, where 0 indicates no printing, 1 indicates standard printing, and 2 enables a verbose level of printing that provides details about the solver operations. Note that diagnostic output can significantly contribute to compute time such that high levels of printing should be avoided unless problems are encountered that require diagnosis. | 0 |
| Controls the smallest value of a species mass fraction that will be tracked by the CHEMKIN-CFD solver convergence algorithm. This should represent either the smallest mass fraction that is of importance in the problem or, for stiff chemistry, the smallest fraction pertaining to a species that can have a significant effect through chemical reaction on other species of interest. Typical values range from 1.E-6 to 1.E-12. | 1.E-8 |
| Controls the accuracy to which gas species mass fractions are resolved. Roughly this parameter controls the number of significant digits in the solution. However, this parameter can also be important for convergence of problems with stiff chemistry, as it is important to sufficiently resolve species that have a major impact on trace species of interest. Typical values range from 1.E-3 to 1.E-5. | 1.E-4 |
| Controls the smallest value of a surface species site fraction that will be tracked by the CHEMKIN-CFD solver convergence algorithm. This should represent either the smallest mass fraction that is of importance in the problem or, for stiff chemistry, the smallest fraction pertaining to a species that can have a significant effect through chemical reaction on other species of interest. Typical values range from 1.E-6 to 1.E-15. | 1.E-8 |
| Controls the accuracy to which surface species site fractions are resolved. Roughly this parameter controls the number of significant digits in the solution. However, this parameter can also be important for convergence of problems with stiff chemistry, as it is important to sufficiently resolve species that have a major impact on trace species of interest. Typical values range from 1.E-3 to 1.E-5. | 1.E-4 |
Table 2: Summary of Advanced CHEMKIN-CFD Parameters
Parameter | Usage | Default Value |
|---|---|---|
| Determines which solver to use during pseudo time-stepping used to improve the initial guess in a cell for steady-state simulations. By default, it is set to 1 and a robust solver with strict error control will be used when the steady-state algorithm requires time-stepping assistance. You may optionally revert to a less accurate time-stepping technique, which may be computationally faster but more prone to failure, by setting it to zero. See Advanced Parameters Used in the Steady-State Solution Algorithm for more information. This parameter is ignored for transient simulations. | 1 |
| This option is only relevant if the Pseudo Time-stepping Solver Option is set to 1 (the default setting). This controls the time interval for time-stepping when time-stepping is used to assist a steady-state calculation. See Advanced Parameters Used in the Steady-State Solution Algorithm for more information. A value of 1.E-4 seconds is recommended for combustion problems. | 1.E-6 [seconds] |
| This parameter controls how many iterations between time-stepping and steady-state solution attempts can be tried before the solver gives up and registers a failure for the cell. Usually only one attempt is needed, but the maximum can be increased from this default value in difficult cases. Note that in some cases it is most efficient to ignore a small number of cell failures early on in the simulation as long as these cells recover on subsequent iterations. See Advanced Parameters Used in the Steady-State Solution Algorithm for more information. A value of 10 to 100 is recommended to minimize error messages. | 1 |
| Maximum pseudo time-step value for time integration during steady-state solution. This value is relevant to steady-state simulations only. | 1.E-2 [seconds] |
| Minimum pseudo time-step value for time integration during steady-state solution. This value is relevant to steady-state simulations only. | 1.E-10 [seconds] |
| Initial time-step value for use in time integration. By default, the solver will automatically choose an appropriate initial time step, which is the recommended approach. In rare cases, however, control of initial time step value may improve convergence behavior. | 0 (that is, not specified) [seconds] |
| Sets the upper bounds for the gas-phase species mass fractions during steady-state solution. In some cases, it may help convergence to allow species to go slightly over the physical bounds during interim steady-state iterations rather than forcing the bounds to be one. | 1.01 |
| Sets the lower bounds for the gas-phase species mass fractions during steady-state solution. In some cases, it may help convergence to allow species to go slightly negative during interim steady-state iterations rather than forcing the bounds to be zero. A value as negative as about -1.E-4 may be used. | -1.E-12 |
| Sets the upper bounds for the surface species site fractions during steady-state solution. In some cases, it may help convergence to allow species to go slightly over the physical bounds during interim steady-state iterations rather than forcing the bounds to be one. | 1.1 |
| Sets the lower bounds for the surface species site fractions during steady-state solution. In some cases, it may help convergence to allow species to go slightly negative during interim steady-state iterations rather than forcing the bounds to be zero. | -1.E-4 |
| This value will be used to substitute any non-positive species fractions during interim steady-state iterations. In some cases, setting this to a very small positive number may aid convergence for the steady-state solver. | 1.E-12 |
| Currently not used. Provided for future use. | ““ |
| Currently not used. Provided for future use. | ““ |
| Currently not used. Provided for future use. | 0 |