A contribution to the number of measurement positions necessary to determine the sound power within a certain span of uncertainty

摘要

The main quantity to describe the noise emission of machines and equipment is its radiated sound power. In practice the so called enveloping surface methods are the most common procedures to measure this quantity where the time averaged sound pressure squared or intensity components are to be determined on a limited number N of discrete positions distributed over the measurement surface. From these data the spatial mean value of the pressure squared respectively the spatial mean of the intensity component are to be calculated. The "quality" of these mean values respectively of the yielded sound power described by its uncertainty depends both on the irregularity of the sound field quantity existing across the chosen measurement surface and the number N of measurement positions. For the measurement effort it is an important question how many positions N are necessary to fulfill a certain given accuracy requirement. The ISO Standards 3745, 3744, 3746 using sound pressure and the 9614 Standards using sound intensity measurements offer for this several rules by formulas for the N-determination in function of the expected upper-limit of the uncertainty Δ and a parameter describing the sound field irregularity. For the same goal the formulas of these standards are different and are based on statistical laws requiring especially non-correlated signals of the measured quantities. The paper presents results from experimental and theoretical correlation investigations, where the objects are sources radiating airborne noise caused by vibrating surfaces and by aerodynamic generations. This investigation is a first step for establishing an assured rule for the determination of the number N of measurement positions guaranteeing a sound power value within a certain limit of the uncertainty for practical situations.