Acoustic Modal Analysis

The Acoustic Modal Analysis model computes the acoustic frequencies, normalized mode shapes and their linear growth rates for a given geometry and CFD solution.

Acoustic Wave Equation

The acoustic wave equation (the inhomogeneous Helmholtz equation) describes sound waves in a fluid:
Figure 1. EQUATION_DISPLAY
2 p t 2 ∇⋅ ( c ¯ 2 p ) = ( γ 1 ) q ˙ t
(4770)
where p is the acoustic pressure fluctuation, t is the time, c ¯ is the mean speed of sound, γ is the specific heat ratio, and q ˙ is the heat release fluctuation.

Harmonic Fluctuations

Assuming harmonic fluctuations at frequency f = ω 2 π (where ω is a complex number), the pressure fluctuation p and the heat release fluctuation q ˙ are expressed as:
Figure 2. EQUATION_DISPLAY
p = p ^ e ( i ω t )
(4771)
and
Figure 3. EQUATION_DISPLAY
q ˙ = q ^ e ( i ω t )
(4772)
respectively.
where:
  • p ^ is the pressure fluctuation magnitude,
  • q ^ is the heat release fluctuation magnitude.
Substituting Eqn. (4771) and Eqn. (4772) into the acoustic wave equation Eqn. (4770) gives:
Figure 4. EQUATION_DISPLAY
ω 2 p ^ + ∇⋅ ( c ¯ 2 p ^ ) = i ω ( γ 1 ) q ^
(4773)
Assuming that p and q ˙ denote the cell center values, integrating Eqn. (4773) over the cell volume gives:
Figure 5. EQUATION_DISPLAY
ω 2 p ^ d V + ∇⋅ ( c ¯ 2 p ^ ) d V = i ω ( γ 1 ) q ^ d V
(4774)
Discretizing Eqn. (4774) gives the following matrix eigenvalue equation—whose eigenvalues are the acoustic frequencies ω , and the corresponding eigenvectors are the acoustic mode shapes P ^ :
Figure 8. EQUATION_DISPLAY
[ A + ω B ( ω ) + ω 2 C ] P ^ = D ( ω ) P ^
(4777)
where matrix A accounts for the wave propagation term, B ( ω ) accounts for the impedance boundary conditions, C is equal to the identity matrix I (except when using the quadratic impedance boundary), and D ( ω ) denotes the combustion heat release source term in the Flame Transfer Function (FTF)—which is omitted from Eqn. (4777) for non-reactive flows. (See Flame Transfer Function). P ^ is a size N c e l l s vector for the acoustic mode (eigenvector) and ω is the acoustic frequency (eigenvalue).

A , B , C , and D are sparse matrices of size N c e l l s × N c e l l s in which N c e l l s is the number of cells that are in the mesh. When discretizing the diffusion term in Eqn. (4774), Simcenter STAR-CCM+ provides the option of including the secondary gradient term, or omitting it.

The phase of the acoustic pressure p ^ is calculated as:
Figure 14. EQUATION_DISPLAY
φ = arctan ( I m ( p ^ ) Re ( p ^ ) )
(4783)

Unstable Modes

The acoustic frequency is composed of real ( ω r ) and imaginary ( ω i ) parts:
Figure 15. EQUATION_DISPLAY
ω = ω r + i ω i
(4784)
Using the above expression, the pressure fluctuation can be expressed as:
p = p ^ e i ω t = [ p ^ e ω i t ] e i ω r t
(4785)
When ω i < 0 , the pressure fluctuations decay with time—and the mode is stable. If ω i > 0 , the pressure fluctuations grow with time—and the mode is unstable. In situations where ω i = 0 , the amplitude of the pressure fluctuations remains constant.

Boundary Conditions

Simcenter STAR-CCM+ allows you to define the acoustic boundary condition in the following ways:
Perfectly Reflecting (Hard Wall) Boundary
Figure 16. EQUATION_DISPLAY
p ^ n B C = 0
(4786)
Zero Acoustic Pressure Boundary
Figure 17. EQUATION_DISPLAY
p ^ = 0
(4787)
Specified (Constant) Impedance Boundary
Figure 18. EQUATION_DISPLAY
c ¯ Z p ^ n B C i ω p ^ = 0
(4788)
where the acoustic impedance specifies the magnitude and phase of the reflected wave, and is the same for every frequency.
Quadratic Impedance Profile Boundary
The acoustic impedance is a function of acoustic frequency Z Z ( ω ) and the impedance is specified using the quadratic form:
Figure 19. EQUATION_DISPLAY
1 Z = 1 Z 0 + ω Z 1 + Z 2 ω
(4789)
in which Z 0 , Z 1 , and Z 2 are complex valued constants.