BC in fluent

Posted 2021-02-16 code-saturne

tags:

篇首语：本文由小常识网(cha138.com)小编为大家整理，主要介绍了BC in fluent相关的知识，希望对你有一定的参考价值。

Boundary conditions in Fluent

1. Boundary Conditions (BC)

1 Boundary Conditions (BC)

1.1 Turbulence Parameters

1000358 setting Turbulence Parameters at inlet for RANS two categories, those which are very sensitive to the inlet values and those which are less sensitive.

Parameters

Specified Values. Spatially varying profiles can be specified using UDFs or boundary profiles
Intensity and Length Scale
Intensity and Viscosity Ratio
- free stream boundaries in external flow, intensity <=1%, turbulent viscosity ratio =2
Intensity and Hydraulic Diameter

Perferred Parameters: intensity and length scale

k, ε, ω, or overline{u_i‘ u_j‘} must be specified for turbulence model > https://www.sharcnet.ca/Software/Fluent6/html/ug/node217.htm

Turbulence intensity: < 1%, low, >10%, high

All CFD simulations performed using a RANS turbulence model require boundary values for turbulence variables at all flow inlets.

It is convenient to divide applications into two groups:

Applications requiring very accurate description of the boundary profiles

at the inlet in order to obtain an accurate solution. For instance, when simulating flow over a backwards facing step, if the inlet is very close to the step, the predicted reattachment length downstream of the step depends strongly on the turbulence profiles.

Applications where single, uniform values can be used for the turbulence variables at the inlet.

These can be further divided into two categories, those which are very sensitive to the inlet values and those which are less sensitive.

For instance, some external aerodynamics applications may be sensitive to turbulence levels at free stream inlet boundaries, whereas in flows driven by mechanical input, such as pumps or fans, the solution may be relatively insensitive to the turbulence values so long as they are reasonably realistic.

Backflow values for turbulence values are also required at pressure outlet boundaries.

Outlets should always be located such that they are far enough downstream of any areas of interest in the simulation that backflow values will not have any effect on the converged results.

LES and DES have very different requirements for inlet boundary conditions than those of RANS models.

Model Comparison

Applies to all RANS models
Any variation of the k-ε family of models
Any variation of the k-ω family of models
The Spalart-Allmaras model
Reynolds stress models
v2fmodel

Four options to specify boundary conditions for the turbulence models:

Explicitly input Values, k, ε, ω Spatially varying profiles

can be specified using UDFs or boundary profiles

Intensity,I, and Length Scale, l
- Length scale is related to size of large eddies that contain most of energy.
  - For boundary layer flows: l ≈ 0.4 δ99
  - For flows downstream of grid: l ≈ opening size
Intensity and Viscosity Ratio
- External flow, 1<μ_t/μ < 10
- Turbulence intensity, I=u_rms ‘ /U <20%
Intensity and Hydraulic Diameter
- Ideally suited for internal flows (duct/pipe flow)

Options 2-4 allow the user the opportunity to provide inflow values in terms of physically intuitive quantities, which Fluent then transforms internally into equivalent values of the turbulence model variables. Interested readers can find the details of these transformations in Section 7.3.2 of the Fluent 12.0 User’s Guide.

The checklist below provides strategies that can be used to determine which option to use in your case and how to determine the values.

1.1.1 Checklist

Do not ever use the default settings (in Fluent v12.0 or Fluent 6.x or lower) for boundary values for whatever turbulence model was selected.

a. There is one possible exception. The default for the Spalart Allmaras model is modified turbulent viscosity = .001 m2/s. That is approximately equivalent to air at standard temperature and pressure with a turbulent viscosity ratio of 70, which could be a reasonable value in some cases, although it would probably be too high for many external aerodynamics calculations.

If your application does not require specification of detailed, spatially varying turbulence profiles at the inlet,

choose one of the following options a. Intensity and Hydraulic Diameter – this is the preferred option for duct flows. It can be used for any flow but hydraulic diameter is only meaningful only when the inlet is the cross section of a pipe or duct. Calculate the turbulence intensity as described in the FLUENT 12.0 User’s Guide, Section 7.3.2, Equation 7.3-1.

b. Intensity and Length Scale – this is the preferred approach when it is reasonably easy to estimate the turbulence length scale. Examples include :

flow downstream of obstructions such as turning vanes or perforated plates (length scale ≈ size of flow openings)
boundary layer flows (length scale ≈ 40% of boundary layer thickness).
Downstream of a flow obstruction, turbulence intensity is generally high.

Correlations or data may be available, but in the absence of additional information such as these, a value of 10% will provide a good starting point.

c. Intensity and Viscosity Ratio – in indeterminate cases the turbulent viscosity ratio (TVR) is a useful option. Guidelines suggested in the FLUENT 12.0 User’s Guide, Section 7.3.2, can be summarized as follows:

i. In flows such as high Reynolds number boundary layers, shear layers, fully developed duct flows the viscosity ratio could be in the range of 100 to 1000.
ii. For free stream boundaries in external flows, recommended values are intensity <= 1% and TVR = 2.

In a completely indeterminate case, 10 is the recommended value for viscosity ratio.

Turbulence intensity may be known from measurements or values that have been reported for similar flows, but in the absence of such additional information a moderate value for turbulence intensity (4-5%) is recommended.

Whenever there is any uncertainty about one or more of the inputs for turbulence values, a sensitivity study can be performed by recalculating the flow using one or more different values of the unknown variables to determine if they have a significant effect on the simulation results.

In applications requiring accurate representation of a boundary layer at the inlet,

a spatially varying profile of the turbulence boundary values will be needed. Profiles can be created in a number of ways. If experimental data is available, it can be put into an appropriately formatted boundary profile file; however it is unusual to have measurements for every single variable used by the turbulence model. For instance in the k- model, data might be available for k but would probably not be available for . In certain cases, correlations may be available for profiles of the turbulence model variables. These can be used either alone or in conjunction with a UDF (see Example 2 here for an example of implementing profiles from a correlation through a UDF) for instance a profile of experimental data might be used for turbulence kinetic energy together with a UDF with a correlation for dissipation. For internal flows with a fixed flow rate, a fully developed profile can be generated easily and quickly by performing an auxiliary 3D periodic calculation to generate the inlet profile.

If you are using the Reynolds stress model and select “K or Turbulent Intensity” to specify the Reynolds stresses, uniform values for each of the individual Reynolds stress components will be assigned based on the value that you have entered for the turbulence kinetic energy or intensity as described in Section 7.3.2 of the FLUENT 6.3 User’s Guide.
Backflow values at pressure outlets

a. If reversed flow does not take place, either during the iterations or as part of the final solution, then it does not matter what values are entered in the panel.

b. If reversed flow does take place then it is still not so important what values are used unless they are dramatically different from the values in regions of the boundary without reversed flow. That can cause convergence problems in some cases. Intensity and Viscosity Ratio is generally a good choice, starting with values of 10 for intensity and 100 for viscosity ratio. c. After the first few iterations, the solution near the outlet can be inspected and backflow values can be adjusted if necessary

1.1.2 Reference

1000358-turbulent-parameters-inlet-RANS.pdf

1.2 inlet

1.2.1 static pressure

supersonic/initial gauge pressure The Supersonic/Initial Gauge Pressure is ignored by FLUENT whenever the flow is subsonic,

1.3 interior

interior: faces have fluid cells on both sides and do not require any boundary conditions to be set.

1.4 pressure outlet

input: relative static pressure ( gauge pressure)
the specified static pressure is used only when the

flow is subsonic

1.4.1 operating pressure

use 0 for air >> easy for post use large value for water to reduce roundoff error

1.4.2 Supersonic/Initial Gauge Pressure

Whenever the flow is subsonic, the Supersonic/Initial Gauge Pressure is ignored by ANSYS Fluent, in which case it is calculated from the specified stagnation quantities.

The static pressure (termed the Supersonic/Initial Gauge Pressure) must be specified if the inlet flow is supersonic or if you plan to initialize the solution based on the pressure inlet boundary conditions.

1.5 Periodic BC

periodic BC includes:

conformal periodic BC: the nodes of the two zones have to match one-for-one.
Non-conformal Periodic BC : the nodes of the two zones do not have to match one-for-one.
- non-conformal periodic BC can‘t be used in 3D sliding mesh

1.5.1 Non-conformal interface with the periodic boundary condition option enabled

Non-conformal interfaces can be used to implement a periodic boundary condition (see Periodic Boundary Conditions).

The advantage of using a mesh interface is that, unlike the standard periodic boundary condition, the nodes of the two zones do not have to match one-for-one.

The BC type is setup as interface rather than periodic

> 5.4.1.1. The Periodic Boundary Condition Option > https://www.sharcnet.ca/Software/Ansys/16.2.3/en-us/help/flu_ug/x1-2670007.4.1.html

1.5.2 Creating Conformal Periodic Zones

The conformal periodic boundaries can be created in the meshing mode of Fluent or GAMBIT when you are generating the volume mesh. See the Fluent Meshing User’s Guide or the GAMBIT Modeling Guide for more information. Alternatively, you can create the conformal periodic boundaries in the solution mode of Fluent using the mesh/modify-zones/make-periodic text command

For conformal periodic boundaries, the periodic zones must have identical meshes.
mesh nodes should match

> mesh/modify-zones/list-zones >/mesh/modify-zones/make-periodic Periodic zone [()] 10 Shadow zone [()] 9 Rotational periodic? (if no, translational) [yes] yes Create periodic zones? [yes] yes

Note: When you create a conformal periodic boundary, Fluent will check to see if the faces on the selected zones “match” (that is, whether or not the nodes on corresponding faces are coincident). The matching tolerance for a face is a fraction of the minimum edge length of the face. If the periodic boundary creation fails, you can change the matching tolerance using the mesh/modify-zones/matching-tolerance text command, but it should not exceed 0.5 or you may match up the periodic zones incorrectly and corrupt the mesh.

mesh → modify-zones → matching-tolerance

1.6 outflow

Outflow boundary conditions in ANSYS Fluent are used to model flow exits where the details of the flow velocity and pressure are not known prior to solution of the flow problem.