When trying to simulate the sound generated by a drone propeller, a jet engine, or even human speech, engineers face a daunting computational bottleneck known as scale disparity.
The physical fluctuations that generate sound—such as turbulent eddies, vortices, and shearing flows—are highly non-linear, localized, and energetically dense. Conversely, the resulting sound waves are tiny, linear pressure perturbations that carry a fraction of the flow's energy but travel across vast distances. If you attempt to simulate both the chaotic near-field fluid dynamics and the far-field acoustic propagation using a single, monolithic computational grid, the simulation becomes impossibly expensive. You would need a mesh fine enough to capture sub-millimeter turbulent structures extended across kilometers of open air.
To bypass this limitation, modern aeroacoustics relies on a hybrid, two-step modeling strategy. Instead of tracking sound waves as they move through a complex fluid, researchers mathematically decouple the problem: they solve the non-linear fluid dynamics only in the immediate vicinity of the source, and then use an acoustic analogy to project those near-field fluctuations into the far-field.
The Core Mechanism: Flow-Sound Separation
The foundational insight behind hybrid aeroacoustics is that near-field fluid mechanics can be computed independently of the far-field sound propagation. As Meng et al. demonstrate, separating these fields allows researchers to use the Incompressible Navier-Stokes (INS) equations for low-speed flows where compressibility doesn't affect the fluid dynamics, and then pass the resulting data to equations like the Linearized Compressible Perturbation Equation (LPCE) to simulate the acoustic waves.
By isolating the acoustic perturbations from the background flow, the complex fluid dynamics are simplified into small perturbations around an ambient medium. This two-step process ensures that the intricate interactions between aerodynamic turbulence and acoustic wave generation are captured accurately without wasting computational power on meshing the entire acoustic domain. For example, Yang et al. utilize high-fidelity Large Eddy Simulations (LES) to map the chaotic feedback loops and vortex impacts inside open cavities, using the separate acoustic step to determine how those localized oscillations radiate outward as low-frequency noise.
Mathematical Bridges: Lighthill, Curle, and FW-H
The mathematical engine that makes this decoupling possible is the acoustic analogy, originally formulated by James Lighthill. The core idea is to rearrange the exact Navier-Stokes equations into the form of a classical wave equation. In this rearrangement, all the complex, non-linear fluid terms that are difficult to propagate are grouped onto the right-hand side of the equation, acting as equivalent acoustic "sources" embedded in an idealized, perfectly still medium.
Depending on the physical system, these sources are categorized by how they radiate energy:
[Fluid Turbulence] --> Quadrupole Sources (Volume-based)
[Static Obstacles] --> Dipole Sources (Surface forces)
[Moving Surfaces] --> Thickness/Loading (Permeable surfaces)
Volume Sources: Lighthill’s Analogy
In pure fluid flows without solid boundaries, turbulence acts as a collection of quadrupole sources (opposing pairs of forces). Shin applies this concept to synthesize physically meaningful vortex sounds directly from 3D fluid data, integrating the contributions of these discrete quadrupole sources to approximate continuous wave propagation. Similarly, Yoshinaga et al. use Lighthill's analogy in the frequency domain to convert simulated turbulent airflow in a vocal tract into the characteristic sound of a sibilant /s/, proving that high-fidelity near-field grids are essential for capturing the precise turbulence kinetic energy that drives noise production (Zhao et al.).
Surface Sources: Curle’s Formulation
When fluid interacts with a rigid, stationary object, the surface exerts a force back on the fluid. Curle extended Lighthill's analogy by showing that these surface pressure fluctuations act as highly efficient dipole sources. This mechanism explains why wind howling past an obstacle is often louder than pure turbulence. Cisonni et al. leverage Curle's formulation to model how airflow striking a tooth-shaped constriction generates acoustic energy, mapping the precise distribution of surface pressure fluctuations to the emitted sound.
Moving Boundaries: The Ffowcs Williams-Hawkings (FW-H) Equation
For rotating machinery—such as drone rotors, cooling fans, or aircraft propellers—the boundaries themselves are moving through space. The Ffowcs Williams-Hawkings (FW-H) equation generalizes the acoustic analogy to account for these moving surfaces by introducing both aerodynamic loading forces (dipoles) and the physical displacement of fluid by the moving blade volume (monopoles).
The true elegance of the FW-H formulation lies in its use of a permeable integration surface. Instead of integrating over the complex, moving geometry of the blade itself, engineers can place an imaginary, stationary surface around the entire rotating component. As long as all the non-linear flow features are contained inside this surface, the far-field noise can be calculated purely by integrating the pressure and velocity fluctuations recorded on that surface.
This strategy is widely adopted across modern aeroacoustics:
- Aviation and Multirotors: Yin et al. and Zanotti et al. apply FW-H to surface pressure data from aerodynamic simulations to predict the complex acoustic directivity and wake-interaction noise of next-generation eVTOL tandem propellers (Zanotti et al.).
- Jet Propulsion: Lindblad et al. project near-field flow data from permeable surfaces to calculate jet noise interactions near aircraft wings without meshing the far-field domain.
- Industrial Design: Yang et al. couple LES with FW-H to optimize non-smooth surface features on heavy machinery fans, disrupting noise-generating vortices at the source.
- Biomimetics: Seo et al. represent swimming fish as compact dipole sources within an FW-H framework to discover how destructive wave interference grants acoustic stealth to schooling fish.
Accounting for the Propagation Medium
While acoustic analogies excel at translating near-field flow into far-field sound, they assume that the medium between the source and the observer is uniform and quiescent. In reality, sound waves must often travel through turbulent shear layers, atmospheric wind gradients, or complex urban terrain.
To handle these environmental effects, researchers couple the acoustic analogy output with specialized propagation solvers. For instance, when sound passes through a violent jet engine shear layer, the interaction causes spectral broadening. Gay et al. analyze this using a multiscale perturbative approach based on the linearized Euler equations to isolate the acoustic fluctuations from the background flow.
On a larger scale, atmospheric conditions can completely alter noise footprints. Botteldooren et al. demonstrate that wind velocity profiles create screen-induced refraction, causing traditional noise barriers to lose effectiveness in downwind conditions. To simulate this efficiently, the authors integrate CFD wind models with acoustic wave equations, using a hybrid model that couples Finite-Difference Time-Domain (FDTD) simulations for complex near-field geometries with the Parabolic Equation (PE) method for long-range propagation through the atmosphere (Botteldooren et al.).
Scaling Up to Large Scenarios
The ultimate utility of the hybrid approach is its scalability. By mastering the physics at the component level, engineers can construct comprehensive noise models for entire systems.
As outlined by Delfs et al., a multi-level, multi-fidelity framework allows developers to take detailed computational aeroacoustics (CAA) data from individual aircraft components—like slats, flaps, or landing gears—and integrate them into full-aircraft radiation patterns. These patterns can then be evaluated within macroscopic ground noise exposure models, seamlessly bridging the gap between micro-level fluid mechanics and community noise assessment.
Key Foundational Work
Auralization I: Vortex Sound Synthesis (2004) — contributed 2 concept elements to this direction.
Parameter study of sound propagation between city canyons with coupled FDTD-PE model (2005) — contributed 1 concept elements to this direction.