Search Results

Noise problems are often system issues rather than component issues. Component manufacturers have been putting continued efforts into constantly improving the quality of their products. There are numerous tests and standards to assess the vibro-acoustic performance of individual components. But once all components are put together, the system response might be entirely different from those of individual components. Typical system level testing has primarily been used to identify bad assembled products from good ones. These tests are usually done as part of a quality control process and slow down production. Such tests usually provide little information about the root causes of noise and vibration problems and no insight into improving engineering designs for noise abatement. This paper presents a new way of conducting system level noise diagnoses by using the Helmholtz Equation Least Squares (HELS) based Nearfield Acoustical Holography (NAH) technology [1].

A semi-empirical formula [1] for predicting noise spectra of an engine cooling fan assembly is developed. In deriving this formulation it is assumed that sound radiation from an axial flow fan is primarily due to fluctuating forces exerted on the fan blade surface. These fluctuating forces are correlated to the total lift force exerted on the fan blade, and is approximated by pressure pulses that decay both in space and time. The radiated acoustic pressure is then expressed in terms of superposition of contributions from these pressure pulses, and the corresponding line spectrum is obtained by taking a Fourier series expansion. To simulate the broad band sounds, a normal distribution-like shape function is designed which divides the frequency into consecutive bands centered at the blade passage frequency and its harmonics. The amplitude of this shape function at the center frequency is unity but decays exponentially. The decay rate decreases with an increase in the number of bands.

An extensive experimental study of noise generating mechanisms of two production models of automotive alternators is presented. It was established that aerodynamic noise (generated by cooling fans) is dominating at high speeds (above 3,000 rpm), while electromagnetic noise is the most intensive at low rpm. Two directions of noise reduction are proposed and validated: reduction of noise levels generated by alternators to be achieved by using axial flow fans for cooling instead of presently used bladed discs, and radical reduction of operating speed of alternators by using variable transmission ratio accessory drives.