Wood chip sound absorbers: Measurements and models

Abstract

Normal incidence absorption coefficient spectra of samples made from glued wood chips have been measured for various mesh sizes, bulk densities, thicknesses, and air gaps. Increasing thickness introduces additional layer resonance peaks and shifts the initial peak towards lower frequencies. The wood chip samples composed of the smallest mesh sizes were found to offer the highest sound absorption, comparable with that of the same thickness of materials made from synthetic fibers. Measured absorption spectra are compared with predictions of four models for the acoustical properties of rigid porous media. These include a model for slanted parallel identical uniform slits (SS), the Johnson-Champoux-Allard (JCA) and Johnson-Champoux-Allard-Lafarge (JCAL) models for arbitrary pore structures, and model for a non-uniform pore size distribution (NUPSD). Porosity and flow resistivity values have been determined non-acoustically. However, the tortuosity and characteristic lengths required for the JCA model have been obtained by fitting the measured absorption spectra. The thermal permeability required for the JCAL model has been deduced indirectly from the fitted tortuosity through a relationship with standard deviation of the pore size distribution due to the NUPSD model. JCAL and JCA models give the best agreement overall, but predictions of the SS and NUPSD models that use only the fitted tortuosity in addition to measured porosity and flow resistivity are found to give comparable agreement with data for many samples. SS and NUPSD predictions are improved by increasing the tortuosity values compared with those obtained by fitting the JCA model. The study should encourage the creation of sustainable sound-absorbing materials from wood chip wastes.