The accuracy of measuring acoustic intensity using two closely spaced microphones is examined. The phase and gain mismatch errors are corrected by measuring the transfer function between the two microphone systems exposing them to the same sound (phase and pressure levels) over a wide range of frequencies. The accuracy of the measurement method was verified by creating a sound field in an anechoic room and by generating plane-wave propagation inside a long length of pipe with an anechoic termination. The measurement accuracies were very satisfactory. This method has the advantage of eliminating the recording and processing of two sets of data required in the circuit switching technique.

Noise intensity measurements per octave band have been carried out in the far field of a small circular nozzle, at angles 0-150° from the jet axis, with and without vortex generators. The results, interpreted in the light of recent theoretical conclusions, confirm Lighthill's original suggestion that the dominant noise radiator is the pressure-shearx-rquadrupole at all but the lowest frequencies. At very low frequencies Reynolds stress-shearx-xand x-r quadrupoles contribute comparable amounts of radiation. When the product of Strouhal number (based on nozzle diameter and jet velocity) and jet Mach number exceeds unity the geometrical acoustics approximation becomes valid, but even at such high frequencies the shear noise contribution dominates. Vortex generators, while reducing noise intensity, do not appear to modify drastically the above composition of noise radiators.

In this paper, a new bearing defect diagnostic and classification method is proposed based on pattern recognition of statistical parameters. Such a pattern recognition problem can be described as transformation from the pattern space to the feature space and then to the classification space. Based on trend analysis of six commonly used statistical parameters, four parameters, namely, RMS, Kurtosis, Crest Factor, and Impulse Factor, are selected to form a pattern space. A 2-D feature space is formulated by a nonlinear transformation. An intraclass transformation is used to cluster the data of different bearing defects into different regions in the feature space. The classification space is constructed by piecewise linear discriminant functions. Training the classification space is performed, in this paper, by using data of bearings with seeded defects. Diagnosis of the defected bearings in the classification space then becomes straightforward. Numerical experiments show that the proposed method is effective in indicating both the location and the severity of bearing defects.

The discharge distribution decreases as the Froude number increases.