An improved method for direct impedance deduction of a locally reacting ground

Abstract

An accurate deduction of the acoustic impedance of a locally reacting ground depends on a precise measurement of sound fields at short-ranges. However, measurement uncertainties exists in both the magnitude and the phase of the acoustic transfer function. By using the standard method, accurate determination of the acoustic impedance can be difficult when the measured phases become unreliable in many outdoor conditions. An improved technique, which relies only on the magnitude information, has been developed. A minimum of two measurements at two source/receiver configurations are needed to determine the acoustic impedance. Even in the absence of measurement uncertainties, a more careful analysis suggests that a third independent measurement is often needed to give an accurate solution. When experimental errors are inevitably introduced, a selection of optimal geometry becomes necessary to reduce the sensitivity of the deduced impedance to small variations in the data. A graphical method is provided which offers greater insight into the deduction of impedance and a downhill simplex algorithm has been developed to automate the procedure. Physical constraints are applied to limit the search region and to eliminate the rogue solutions. Several case studies using indoor and outdoor data are presented to validate the proposed technique.