Acoustics
Acoustic analysis predicts pressure modes and pressure response in a fluid domain. The physical model needs density and sound speed, a pressure/velocity source, and boundary behavior such as a rigid wall, radiation condition, or impedance. RunMat does not infer those conditions from CAD face names or from an enclosed-looking volume.
The .fea language checks static, modal, transient, harmonic, and nonlinear acoustic procedure
contracts. The Rust numerical library has a canonical Tet4/Tet10 harmonic
entry, physics::acoustic::solve_canonical_harmonic, with exact pressure, normal-velocity,
rigid-wall and positive real impedance boundaries. The public typed .fea executor does not
connect that entry end to end. The older reduced acoustic Helmholtz graph remains separately
identified as a proxy; it is not the canonical FE formulation.
The canonical acoustic qualification record states the independent references, supported library envelope and outstanding gates.
What acoustic equation is represented?
For small-amplitude harmonic pressure in a homogeneous medium,
with pressure, normal-velocity, impedance, rigid-wall, radiation, or periodic boundary conditions. An impedance boundary has the frequency-domain form
Air with loss is described per material by damping_ratio in the acoustic block: the
bulk modulus becomes complex, with loss factor , so the sound speed is
and a resonance's half-power bandwidth is times its frequency. Harmonic
studies apply it region by region.
The implemented reduced component constructs a bounded structured graph from the analysis model,
uses nearest-neighbor Laplacian weights , adds and damping/boundary diagonal terms,
maps pressure sources onto graph nodes, and solves the complex Helmholtz system. Its admitted
realization is named ReducedAcousticHelmholtz; source evidence distinguishes a hashed-target graph
projection from a synthetic fallback. This component is useful for pipeline integration and
diagnostics; it is not a general solver for arbitrary CAD acoustics.
Which acoustic procedures are defined?
| Procedure | Required domain/state | Current numerical status |
|---|---|---|
acoustic static | Static pressure state | Typed contract; reduced component is harmonic, not static |
acoustic modal | Positive mode count; acoustic modes state | Canonical Tet4/Tet10 pressure eigen-solve; publishes modal.frequency_hz, modal.eigenvalue and one acoustic.mode_shape.<mode> field per mode |
acoustic transient | Time samples; acoustic pressure state | Canonical Tet4/Tet10 Newmark pressure solve; standalone studies publish retained nodal pressure snapshots, time and energy series, and authored native nodal pressure histories sampled over time. Without an authored history, only the final pressure snapshot is retained. |
acoustic harmonic | Frequency samples and complex pressure | Canonical Tet4/Tet10 complex pressure; one certified sparse LU per frequency solves and proves the operator nonsingular; publishes acoustic.pressure, acoustic.particle_velocity and acoustic.intensity at every frequency, and authored outputs (pressure, level in dB re 20 µPa) at the sampled frequencies |
acoustic nonlinear | Load-factor domain | Typed contract only |
No unsupported procedure is approximated by silently selecting the reduced harmonic component.
Standalone cavity-mode studies do not admit authored observations. Standalone pressure transients
admit history outputs of scalar acoustic_pressure at native nodes with every-accepted-step,
explicit-time, or time-interval sampling. Final-only history and other observation kinds,
locations, recoveries, and sampling domains are rejected. Explicit times must uniquely resolve
to accepted transient grid points; the original authored time labels remain in the history.
How do I describe an acoustic cavity modal study?
model cavity_model {
assign material.air to geometry.cavity.region.fluid
constraint pressure_reference {
acoustic.pressure {
target = geometry.cavity.region.reference_face
value = 0 Pa
}
}
}
study cavity_modes {
geometry = geometry.cavity
mesh = mesh.cavity
model = model.cavity_model
analysis modes { acoustic modal { modes = 20 } }
}The discrete eigenproblem is
. runmat run solves it on the
canonical mesh and publishes each natural frequency with its pressure mode shape and residual.
How do I apply an acoustic pressure or speaker velocity?
signal drive : velocity over frequency {
samples = [
{ at = 100 Hz, value = 0.005 m/s },
{ at = 1000 Hz, value = 0.010 m/s },
]
interpolation = linear
extrapolation = reject
}
excitation speaker {
normal_velocity {
target = geometry.cavity.region.driver
signal = signal.drive
}
}Use acoustic_pressure for a pressure drive and normal_velocity for a boundary velocity. The
signal quantity, frequency domain, interpolation, and complex/real shape must match the excitation.
Static direct forms are available when one constant value is sufficient.
How do I model rigid, radiating, or impedance boundaries?
load enclosure {
acoustic.rigid_wall { target = geometry.duct.region.walls }
}
load termination {
acoustic.impedance {
target = geometry.duct.region.outlet
impedance = 415 Pa*s/m
}
}acoustic.rigid_wallimposes zero normal velocity.acoustic.radiationrepresents an outgoing-wave truncation.acoustic.impedancerelates pressure and normal velocity through the declared specific impedance.acoustic.pressureprescribes the primary pressure unknown.- periodic and phase-shifted periodic constraints pair repeating acoustic boundaries.
These are different physical models. A radiation boundary is not a generic replacement for a measured termination impedance, and a pressure reference is not an absorbing boundary.
Why are my acoustic resonances or sound levels wrong?
Investigate:
- wrong geometry units or sound speed;
- a missing pressure reference for a formulation that needs one;
- rigid-wall assumptions on openings that should radiate;
- source area/direction or impedance units that do not match the experiment;
- insufficient spatial resolution relative to wavelength ;
- frequency samples that skip narrow resonances;
- damping or boundary loss added without physical basis;
- reduced/synthetic realization evidence where full CAD-backed FE accuracy is required.
Pressure amplitude, RMS pressure, and sound-pressure level are not interchangeable. When SPL is reported, its reference pressure and amplitude convention must be part of the output definition:
See acoustic constraints, acoustic loads, fluid–structure interaction, outputs, and current status.