A High Resolution Continuous Scan Acoustic Measurement Method for Turbofan Engine Applications

Abstract

Detailed mapping of the sound field produced by a modern turbofan engine, with its multitude of overlapping noise sources, often requires a large number of microphones to properly resolve the directivity patterns of the constituent tonal and broadband components. This is especially true at high frequencies where the acoustic wavelength is short, or when shielding, scattering, and reflection of the sound field may be present due to installation effects. This paper presents a novel method for measuring the harmonic and broadband content of complex noncompact noise sources using continuously moving (referred to here as continuousscan (CS)) microphones in conjunction with a stateoftheart phasereferencing technique. Because the microphones are moving through the sound field produced by the noise sources, they effectively provide infinite spatial resolution of the sound directivity over the scan path. In this method, harmonic (i.e., shaftcoherent) content at the integer multiples of the instantaneous shaft rotational frequency is first extracted from the time signal using a tachometer signal and the VoldKalman (VK) filter. The residual broadband signal is then filtered in the time domain in fractional octave bands. The broadband spectra of the signals from the moving microphones are then computed at arbitrary positions along their scan paths using weighted averages (based on Chebyshev polynomial zerocrossings) and the assumption of a complex envelope that varies slowly over a spatial scale whose lower bound is set by the acoustic wavenumber. A benefit of this method is that the decomposition of the total measured sound field into a stochastic superposition of components preserves a meaningful phase definition for each “partial fieldâ€‌ associated with a given shaft order (SO). This preservation of phase data enables the forward or backward projection of each of these partial fields using acoustical holography (AH). The benefits of the CS method are demonstrated using acoustic data acquired for a 22in. scalemodel fan stage run at the NASA Glenn Research Center's 9foot by 15foot wind tunnel. Two key outcomes of the work include (1) significant improvement in the spatial resolution of the measured sound field and (2) reduction in the overall data acquisition time. Additionally, the methods described here lead to new opportunities for noise source diagnostics and visualization.