Empirical Prediction of Workshop Fitting Densities for Noise Prediction by Ray Tracing

Abstract

Empirical models were developed for predicting frequency-varying fitting densities in industrial workshops for use in the prediction of noise levels by a ray-tracing model. Eleven typical workshops with varying dimensions, types, quantities and distributions of fittings, in which octave-band sound-propagation curves and the fitting dimensions had been measured, were involved. The workshops were modeled and sound-propagation curves were predicted for a range of fitting densities. The predicted curves were compared with the measured curves to determine the ‘best-fit’ fitting density. Linear-regression analysis was then used to find empirical models for predicting the best-fit fitting densities from physical parameters calculated from the fitting and workshop dimensions. The average fitting-to-workshop height ratio, the fitting-to-workshop volume ratio and the number of fittings were the parameters that predicted the fitting density best. Preliminary validation work, involving the comparison of sound-propagation curves predicted with the empirically-predicted fitting densities by ray tracing and the curves measured in four other workshops, suggests that the empirical models are inherently valid.