TMM is a Python toolbox for building and evaluating multilayer acoustic treatments with transfer-matrix methods. It supports porous layers, air cavities, membranes, circular perforated panels, long-slot panels, porous facings, and material-model helpers based on the included database.

The package can compute normal, angle, and diffuse-incidence absorption and impedance, plot common treatment quantities, save/load HDF5 checkpoints, and export XLSX or CSV result files.

Install the package directly from GitHub:

pip install git+https://github.com/rinaldipp/tmm.git

For local development, clone the repository and install it in editable mode:

pip install -e .

Runtime dependencies are declared by the package and include numpy, scipy, matplotlib, pandas, mpmath, xlsxwriter, and h5py.

from tmm.tmm import TMM

# Define the frequency range, resolution, and sound incidence.
treatment = TMM(
    fmin=20,
    fmax=5000,
    df=1,
    incidence="diffuse",
    diffuse_method="field",
    incidence_angle=[0, 78, 1],
    filename="example_perforated_resonator",
    display_name="Perforated resonator",
)

# Add layers from the incident face toward the rear termination.
treatment.perforated_panel_layer(t=19, d=8, s=24, method="bessel_ingard", rho=2700)
treatment.porous_layer(model="db", t=50, sigma=27)
treatment.air_layer(t=50)

# Compute, plot, export, and save the treatment.
treatment.compute(backing="rigid", show_layers=True)
treatment.plot(plots=["alpha"], save_fig=True, show_fig=True)
treatment.save2sheet(n_oct=3)
treatment.save()
bands, filtered_alpha = treatment.filter_alpha(view=True, n_oct=3)

incidence="normal" evaluates normal incidence, incidence="angle" evaluates a single configured angle, and incidence="diffuse" evaluates a diffuse-field approximation. For diffuse incidence, diffuse_method="field" computes a field-incidence impedance and absorption result. diffuse_method="paris" computes only a scalar diffuse absorption coefficient from angular absorption integration, so it does not produce a diffuse complex impedance.

Standalone examples are available in the examples folder.

[1] R. Petrolli, A. Zorzo and P. D'Antonio, "Comparison of measurement and prediction for acoustical treatments designed with Transfer Matrix Models", in Euronoise, October 2021.

For questions about usage, bugs, licensing and/or contributions contact rinaldipp@gmail.com.