6.4.3. Impedance Boundary Conditions

The impedance boundary condition (IBC) lies in between a traditional reflective boundary condition and a fully non-reflective boundary condition. It provides the ability to specify a partial reflection in the range from full-reflection to no-reflection. Impedance is a complex value; it is the reflection that changes the amplitude and the phase of the incoming wave. The use of impedance boundary conditions comes in cases where the flow in the simulation is highly influenced by reflected waves from objects outside the computational domain. In such cases the acoustic wave interaction from the larger domain can be modeled in the smaller domain through the use of impedance boundary conditions.

6.4.3.1. Restrictions and Limitations

The impedance boundary condition is available only in the pressure-based solver. It is incompatible with steady-state flow, multiphase, or compressible liquid models (compressible-liquid method for density).

6.4.3.2. Theory

Impedance specifies an acoustic resistance in the frequency domain. It is a characteristic of the properties of the media and specific geometry described by a ratio of the pressure perturbation to the normal velocity perturbation at the boundary Blackstock [[11]].

(6–192)

where apostrophe denotes acoustic perturbation and hat denotes quantity in the frequency domain.

Fluent is a time domain solver. It cannot use the impedance from the frequency domain directly. The above expression and all its variables have to be converted to the time domain. After the conversion the relation between pressure and normal velocity perturbations is expressed through a convolution integral.

(6–193)

If impedance, , is unbounded in the time domain, then admittance is used (admittance is the inverse of impedance). Fluent uses the reflection coefficient instead of impedance/admittance, to uniformly treat unbounded cases Fung [[33]]. The reflection coefficient is a ratio between reflected and incoming wave amplitudes at the boundary. It is expressed through the impedance as:

(6–194)

Using a reflection coefficient, the relation between pressure and normal velocity perturbation is:

(6–195)

The discretized form of this expression is used in Fluent to connect acoustic pressure and normal velocity. The computed acoustic perturbations are superimposed onto the pressure and velocity from non-reflecting boundary condition equations. The non-reflecting boundary condition equations provide mean flow values at the boundary, which drive the flow in the domain.

The data for the reflection coefficient are available in the frequency domain. As such they usually do not satisfy the causality and reality conditions. Fluent asks you to provide the reflection coefficient data in the form of a special approximation. This approximation is based on the system theory, which ensures that the reflection coefficient in the time domain will satisfy the above conditions Fung [[32]]. The reflection coefficient is represented as a sum of zero, first and second order systems. The zero system is described with a real value, the first order system is described with a real pole, and the second order system is described with a pair of complex conjugate poles. Introducing a system variable, the complete approximation for the reflection coefficient is:

(6–196)

where is a real term, is a number of real poles, and are real pole and its amplitude, is a number of complex conjugate pole pairs, , are real and imaginary part of the complex conjugate pole, and are real and imaginary part of the amplitude of the complex pole.

To obey the causality and reality conditions, real pole , real and imaginary part of the complex conjugate pole should be positive. The passivity condition requires that the absolute value of zero order term be less than 1. The above restrictions are enforced in the user interface. In addition you should ensure that the absolute value of the reflection coefficient computed by this formula is less than 1.

The impedance data can be obtained from measurements or from an acoustic solver. Running these data through a mathematics package will provide an approximation in terms of first and second order poles.

6.4.3.3. Using the Impedance Boundary Condition

The example below shows you how to enable the impedance boundary condition (IBC) in the Pressure Outlet dialog box. Similarly, you can enable IBC in the Pressure Inlet, Velocity Inlet, and Mass-Flow Inlet dialog boxes for compressible flows with the pressure-based solver.

  1. Select pressure-outlet from the Boundary Condition task page and click the Edit... button.

  2. In the Pressure Outlet dialog box, choose Impedance under Acoustic Wave Model.

    The dialog box will expand to reveal Impedance Parameters.

  3. In Impedance Parameters enter data for the reflection coefficient according to the approximation formula (Equation 6–196).

    1. Specify the Zero Order Term, .

    2. Under Real Pole Series, set the Number of Real Poles and for each real pole specify the Pole, , and the Amplitude, .

    3. Under Complex Pole Series, set the Number of Complex Poles and for each complex pole pair, specify Pole Real, , Pole Imaginary, , Amplitude Real, , and Amplitude Imaginary, .


    Important:  If flow is tangential to the boundary, then specify either From Neighboring Cell or Direction Vector for Back flow Direction Specification Method. Do not select Normal to the boundary because this computes the face velocity components from flux to zero values during initialization, which impairs solver convergence.


    The IBC is implemented on top of the non-reflective boundary condition. Choose a time step that will not make the CFL number exceed a value of 1 in the cells adjacent to the impedance boundary.


    Note:  The mean flow in the domain should be well established before enabling IBC.



Release 18.2 - © ANSYS, Inc. All rights reserved.