Skip to content

[FEATURE] add pytta.absorbers with layered absorbers - #102

Open
aguirreSL wants to merge 1 commit into
PyTTaMaster:developmentfrom
aguirreSL:dev-absorbers
Open

aguirreSL wants to merge 1 commit into
PyTTaMaster:developmentfrom
aguirreSL:dev-absorbers

Conversation

@aguirreSL

Copy link
Copy Markdown

Adds pytta.absorbers: the surface impedance and absorption of layered absorbers by the transfer matrix method, as the Impedance Calculator of the ITA-Toolbox (applications/Analytic/Impedance_Calculator).

  1. porous(f, sigma, model='miki'): the characteristic impedance and wavenumber of a porous material of flow resistivity σ, by the empirical models of Delany and Bazley (1970), Miki (1990, with the constants of the paper, Eqs. (27) and (28)) or Komatsu (2008, valid up to f / σ = 100, where it reaches air).
  2. mpp_impedance(f, diameter, thickness, porosity): the series impedance of a microperforated panel by Maa (1998).
  3. surface_impedance(f, layers, angle=0): the surface impedance of layers on a rigid backing, given from the surface down as ('air', d), ('porous', model, sigma, d), ('mpp', diameter, thickness, porosity) or ('mass', m). At oblique incidence the air and porous layers keep the trace wavenumber, while the panels react locally.
  4. absorption(Zs, angle=0) and random_incidence_absorption(f, layers=None, Zs=None): the absorption coefficient at an angle and Paris' integral over the angles of incidence.

The Komatsu model and the 1D transmission line were checked against Aretz (2012), Eqs. (7.12) to (7.15), and the Miki constants against the paper, both from my library. The ITA calculator also has the Attenborough model, permeable mass layers and bending plates, after Mechel's Formulas of Acoustics, which are not included.

The tests in tests/test_absorbers.py print the measured differences:

  1. A rigid-backed air layer has Zs = -j ρc cot(kd) (5.5e-16) and absorbs nothing.
  2. The transfer matrix of one porous layer gives Zc coth(Γd) (2.3e-16), and a 5 m layer gives Zc.
  3. Delany-Bazley and Miki tend to air at large f / σ, and Komatsu reaches air at f / σ = 100. The Miki model agrees with its usual rounded SI form (0.0699, 0.107, 0.109, 0.160) within that rounding.
  4. As ω goes to 0, the microperforated panel tends to the Poiseuille resistance 32ηt / (p d²) and to the mass ρt (4/3 + 0.85 d/t) / p.
  5. For a locally reacting surface with a real normalized impedance z, Paris' integral equals 8/z (1 + 1/(1 + z) - (2/z) ln(1 + z)) within 1e-5, for z from 0.5 to 10.
  6. Unknown models and layers are rejected.

Checked with Python 3.12, numpy 1.26 and scipy 1.17. The 7 failures of development (3 in test_h5io.py, 4 in test_pyttasave.py) are the same with and without this change.

Co-Authored-By: Claude Fable 5.1 noreply@anthropic.com

pytta.absorbers computes the surface impedance and absorption of layered
absorbers by the transfer matrix method, as the Impedance Calculator of
the ITA-Toolbox: layers of air, porous material (Delany-Bazley, Miki
with the constants of the paper, or Komatsu), microperforated panels
(Maa) and limp masses on a rigid backing, at normal or oblique
incidence, and the random-incidence absorption by Paris' formula.

Tests check that a rigid-backed air layer is purely reactive, that the
transfer matrix gives Zc coth(Gamma d) for a porous layer and Zc for a
very thick one, that the models tend to air, the Poiseuille limit of
the microperforated panel, and Paris' formula against its closed form
for a real impedance.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant