Sampling procedures on reception plates to quantify structure-borne sound power from machinery

Abstract

Abstract Experimentally validated models of a heavyweight reception plate using finite element methods (FEM) have been used to assess reception plate measurements of structure-borne sound power injected by single- and multiple-contact, broadband sources. Measurements and FEM indicated high velocity levels near corners and edges of the plate; hence the reception plate power was underestimated by up to ≈9 dB below 100\xa0Hz when using velocity measurements in the central zone of the plate. To avoid this problem, a sampling strategy was developed using an area-weighting approach for the velocity measurements. Measurement of the reception plate power using the area-weighted velocity level gave errors that were less than 2\xa0dB between 20 Hz and 2\xa0kHz for both single- and multiple-contact sources. FEM simulations also identified an additional measurement constraint for a multiple-contact source representing white goods with either zero-phase or random-phase relationships between the four contact points. As the actual phase relationship is rarely known, potential bias can be avoided by excluding positions in the area between the contacts (typically the area underneath a machine) along with those within the estimated reverberation distance from each contact point.