Correlation between Aerodynamic Noise and Velocity Fluctuation of Tip Leakage Flow of Axial Flow Fan ()
1. Introduction
Low noise level is an important sales point of the various kinds of machines as well as high performance and miniaturization. This situation is also applied to axial flow fans used, for example, in air conditioners.
The controlling noise source generated from an axial flow fan is turbulent noise due to vortex shedding when the fan is operated near the design point [1,2]. Fukano et al. have investigated the discrete frequency noise generated by Karman vortex shedding from a flat plate blade immersed in a uniform two-dimensional flow field, and theoretically introduced a formula to predict its sound pressure level [3].
On the other hand, aerodynamic noise increase by enlarging a tip clearance is studied by Longhouse [4], Fukano et al. [5], and Kameier and Neise [6]. Their studies showed that the spectral peaks occurred in sound spectra although a tip clearance noise is broadband naturally. Fukano and Jang [7] also reported that the noise increase due to tip clearance flow at low flow rate condition was analyzed with relation to the distribution of velocity fluctuation due to the interference between the tip leakage vortex and the adjacent pressure surface of the blade. However, the relation between the aerodynamic sound and the tip leakage vortex near the rotor tip in axial flow fan is unclear in detail.
The purpose of the present study is to clarify the relation between fan noise and velocity fluctuations of tip leakage flow around rotor blade of a low pressure axial flow fan in the cases of the design and off design point. The correlation measurements between the velocity fluctuation of tip leakage flow and the aerodynamic noise are made using a rotating hot-wire sensor near the rotor tip in the rotating frame.
2. Experimental Apparatus and Procedure
The schematic view of the experimental apparatus is shown in Figure 1. It was an open-loop facility having the duct inner diameter of 579 mm. The facility consisted of a bellmouth inlet, a fan driving motor connected by the belt, a damper and a booster fan. The aerodynamically designed damper was used to adjust the flow rates.
The present study was performed on low speed axial flow fan of the tip clearance of 2 mm, the outer radius of 287.5 mm and the number of blades of 9. A flow coefficient (F) and a total pressure rise coefficient (Yt) are defined as
(1)
(2)
where Q is the volume flow rate, Dt the rotor tip diameter, Dh the rotor hub diameter, Ut the rotor tip speed, ΔPt the total pressure rise, and ρ the air density. The fan has a design flow coefficient F of 0.41. Figure 2 shows the total pressure rise Yt plotted against flow coefficient F of the test fan. The total pressure rise was measured by using 5-hole pitot tube. The rotor blade has NACA 65 series profile sections designed by free vortex operation. The blade stagger angle at the rotor tip is 63.9 deg. The experimental measurements were carried out at the design operating condition of F = 0.41 and off design conditions of F = 0.31, 0.47 while rotational speed of the fan rotor was kept constant, 1000 rpm. The blade tip section of the rotor has the solidity of 0.65 and the chord length of 131 mm. Reynolds number based on the rotor tip speed and the rotor tip chord length is 2.5 × 105. The trailing edges of the rotor blades were semicircle shape which thickness δt were 2.0 mm over the whole span of the blades.
The velocity fluctuation near blade tip was measured from a relative frame of reference fixed to the rotating blade by using an I-type hot-wire sensor which rotated with the same speed of the blade. The hot-wire was a tungsten filament wire of 5-μm diameter. The schematic view of the experimental apparatus used in the present experiment is shown in Figure 3. The hot-wire sensor was traversed and fixed at pre-determined location by a computer controlled traversing system even when it was turning. The traversing probe was controlled by the three-dimensional traversing system, i.e., radial, axial and rotational directions, installed inside of the hub with traverse resolution of 0.3 mm.
The wire of the probe sensor was set parallel to the radial direction of the rotor blade. The output from this sensor was automatically sampled by a computer and the statistical values were calculated. The spectrum analysis
Figure 2. Total pressure rise of test fan.
Figure 3. Schematic of the test section.
of the velocity fluctuation was performed using FFT analyzer. The output from the hot-wire probe was carried from a rotating frame to a stationary frame through slip ring unit installed inside of the hub as shown in Figure 1. Figure 4 shows a measuring area around rotor blade in the blade passage. The measuring area on the L-Z plane is 96 per cent span of blade and R-Z plane is from 0.90 to 0.97.
The aerodynamic sound generated from the rotor blade was measured at the position of 1 m upstream from the fan rotor and on the rotational axis. In the measurement of the sound, the background noise kept 5 dB below the sound pressure level of all frequencies. The frequency
Figure 4. Measuring area around rotor blade.
resolution was estimated to be 5.0 Hz.
The resonance curve was measured by the supply of sinusoidal tone from the speaker at the outlet of test duct as shown in Figure 1, and the damping ratio ζ was obtained by the half power method. This measurement of acoustic damping ratio was carried out in the condition without the flow.
3. Results and Discussion
3.1. Acoustic Resonance Frequency in the Duct
The resonant frequency in the Z-direction of the duct is given by
(1)
where a is the sound velocity, L the length of the duct, and m the number of standing waves in the duct. Natural resonance frequencies in the Z-direction of the duct are about 86.9Hz (m=3), 144.8 Hz (m = 5), 173.8 Hz (m = 6), 231.7 Hz (m = 8), 405.5 Hz (m = 14), 550.3 Hz (m = 19), etc., when the temperature of the flow t is 20˚C.
Figure 5 shows the typical result of measured frequency response characteristic in the test duct. The multiple peaks were formed at the frequency response spectrum. The peak frequencies of f5 = 140 Hz, f6 = 175 Hz, f8 = 235 Hz, f14 = 405 Hz and f19 = 550 Hz agree with the resonance frequencies calculated by Equation (1) although there are the bellmouth, hub, honeycomb and chamber in the test duct. It is considered that the other peaks of 185 Hz, 210 Hz and 355 Hz etc. were formed by the acoustic resonance of the Z-direction coupled with the other direction.
Figure 6 shows the variation of acoustic damping ratio ζ plotted against the resonance frequencies. The acoustic damping ratio became local minimum at about 175 Hz. It is considered that it is easy to generate the acoustic resonance of this mode in the test duct.