Suppression of tonal noise in a centrifugal fan using guide vanes

Journal of Sound and Vibration 357 (2015) 95–106

Contents lists available at ScienceDirect

Journal of Sound and Vibration journal homepage: www.elsevier.com/locate/jsvi

Suppression of tonal noise in a centrifugal fan using guide vanes Kishokanna Paramasivam a, Srithar Rajoo b, Alessandro Romagnoli c,n a b c

Malaysia–Japan International Institute of Technology, Universiti Teknologi Malaysia, Kuala Lumpur, Malaysia UTM Centre for Low Carbon Transport in Cooperation with Imperial College London Universiti Teknologi Malaysia, Johor, Malaysia Nanyang Technological University, School of Mechanical & Aerospace Engineering, 50 Nanyang Drive, Singapore, Singapore

a r t i c l e i n f o

abstract

Article history: Received 22 January 2015 Received in revised form 29 June 2015 Accepted 4 July 2015 Handling Editor: L.G. Tham Available online 30 July 2015

This paper presents the work aiming for tonal noise reduction in a centrifugal fan. In previous studies, it is well documented that tonal noise is the dominant noise source generated in centrifugal fans. Tonal noise is generated due to the aerodynamic interaction between the rotating impeller and stationary diffuser vanes. The generation of tonal noise is related to the pressure fluctuation at the leading edge of the stationary vane. The tonal noise is periodic in time which occurs at the blade passing frequency (BPF) and its harmonics. Much of previous studies, have shown that the stationary vane causes the tonal noise and generation of non-rotational turbulent noise. However, omitting stationary vanes will lead to the increase of non-rotational turbulent noise resulted from the high velocity of the flow leaving the impeller. Hence in order to reduce the tonal noise and the non-rotational noise, guide vanes were designed as part of this study to replace the diffuser vanes, which were originally used in the chosen centrifugal fan. The leading edge of the guide vane is tapered. This modification reduces the strength of pressure fluctuation resulting from the interaction between the impeller outflow and stationary vane. The sound pressure level at blade passing frequency (BPF) is reduced by 6.8 dB, the 2nd BPF is reduced by 4.1 dB and the 3rd BPF reduced by about 17.5 dB. The overall reduction was 0.9 dB. The centrifugal fan with tapered guide vanes radiates lower tonal noise compared to the existing diffuser vanes. These reductions are achieved without compromising the performance of the centrifugal fan. The behavior of the fluid flow was studied using computational fluid dynamics (CFD) tools and the acoustics characteristics were determined through experiments in an anechoic chamber. & 2015 Elsevier Ltd. All rights reserved.

1. Introduction Nowadays centrifugal turbomachines such as fans, compressors and turbine has wide range of applications from industrial to home appliance. The aerodynamic noise generation from these machines has been a major concern to many designers aiming for quieter solutions. The aerodynamic noise from turbomachines can generally be characterized by tonal noise and broadband noise [1]. Tonal noise arises primarily due to the strong flow interaction between the rotating impeller and the stationary vane/volute cut-off where else broadband noises are due to separated flows in the impeller shroud,

n

Corresponding author. Tel.: þ65 6790 5941. E-mail address: [email protected] (A. Romagnoli).

http://dx.doi.org/10.1016/j.jsv.2015.07.003 0022-460X/& 2015 Elsevier Ltd. All rights reserved.

96

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

Nomenclature b2 c0 C d2 Fi H Lpi Lp Mi Mr ni n Nm p P Q r s t T ij u2 vn w xi

blade width at discharge (m) speed of sound ðm s  1 Þ blade length blade diameter (m) force acting on fluid (N) Heaviside function background noise corrected sound pressure level at ith position overall surface-averaged sound pressure level Mach number in i direction the component of convection Mach number in the direction r of the observer normal surface impeller rotational speed (rev/min) number of microphone sound pressure (Pa) pressure (Pa) volume flow rate ðm3 s  1 Þ radius (m) blade spacing time (s) Lighthill's tensor impeller peripheral speed ðm s  1 Þ velocity normal to surface ðm s  1 Þ average eddy-induced velocity cartesian coordinates of the observer (m)

xj Z Πt;t ϵR β δðf Þ ρ σ φ ψ ϕ

cartesian coordinates of the source (m) number of blade total to total pressure ratio relative error blade exit angle Dirac-delta distribution air density slip factor flow coefficient total pressure coefficient generic variables

Subscripts D.L.E G.V.L.E I.T.E t

diffuser leading edge guide vane leading edge impeller trailing edge total

Abbreviations BEP BPF CFD dB SPL

best efficiency point blade passing frequency computational fluid dynamics decibel sound pressure level

turbulence/secondary flows in the blade channel and blade trailing edge interactions [2–4]. A centrifugal fan used in domestic vacuum cleaner is adopted in this study. The centrifugal fan consists of an impeller, a diffuser, a circular casing and a universal motor as shown in Fig. 1. Several theoretical studies and experimental works on the centrifugal fan have been conducted over the years. One of the comprehensive reviews is given by Neise which summarized all the efforts on improving the geometry of the impeller and its

Fig. 1. Exploded image of the parts from centrifugal fan.

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

97

volute cut-off to reduce tonal noise in centrifugal fan [5]. Neise and Koopmann experimentally investigated a method by which an acoustic resonator can be used to reduce the aerodynamic noise generated by turbomachinery [6]. Choi focused on identifying the aspects of fluid dynamics that are associated with noise generations in centrifugal turbomachinery [7].An experimental study by Cudina and Prezelj presents the aerodynamic, mechanical, and electromagnetic noise emitted by vacuum cleaner in different operational conditions. They state that the minimum sound pressure level is at the best efficiency point (BEP) of the centrifugal fan [8]. Jeon used two-dimensional vortex method to analyze the unsteady flow field in centrifugal fan of a vacuum cleaner. This is used to predict the sound pressure level so that it can be implemented during the design stage of the centrifugal fan [9]. In addition to that, Jeon also proposed trailing edge of impeller to be tapered where the inclination of the trailing edge will cause some phase shift to the pressure fluctuation around the impeller and diffuser vanes. These phase shifts results in local cancellation which reduces the strength of the fluctuating pressure consequently suppress the strong peak sound at BPF [10]. Khelladi et al. presented work on numerical prediction of dipole and monopole tonal noises in frequency domain using Ffowcs Williams and Hawkings (FW–H) equation. It was shown that the interaction between the impeller–diffuser is the origin of tonal noise [11]. Beside studies on centrifugal fan, some investigations have been conducted on the geometry of diffuser vane leading edge in compressor. For instance, Raitor and Neise studied on dominant sound generation mechanisms governing the overall noise level in centrifugal compressors. They concluded that blade tone noise and buzz-saw noise at design speed are the main contributors [12]. Ohta et al. showed that by tapering the leading edge from suction side to pressure side, noise reduction and improvement in term of performance can be achieved [13]. All the previous studies have contributed to the understanding the role of stationary vane in tonal noise generation. In the present work, an investigation was carried out to study the effect of tapered leading edge stationary vane on tonal noise generation. The existing diffuser vane was replaced with the tapered guide vane. Acoustic experiments according to the standardized procedure were conducted in an anechoic chamber to measure the sound pressure level of the centrifugal fan. This paper also includes computational fluid dynamics (CFD) modeling which was validated through experiment for detailed visual of the flow in the centrifugal fan. 2. Experimental methodology 2.1. Investigated prototypes Stationary vanes in turbomachines can serve as nozzles to accelerate the flow as diffusers to diffuse the flow. The diffuser vanes are designed in a manner where the flow from the impeller encounters increasing in flow area as it passes through it. The increase in flow area causes reduction in the flow velocity which leads to increase in the flow pressure. For centrifugal fan in the current study, the impingement of impeller outflow and wake/jet flow structure on the leading edge of diffuser vanes causes periodical pressure fluctuation which leads to the tonal noise generation at blade passing frequency. In addition, the impeller trailing edges and diffuser vane leading edges in the conventional design are parallel to each other causing the pressure fluctuations on the leading edges which are excited by the outflow from the impeller to be in phase. Therefore in order to reduce the tonal noise and improve sound quality, the leading edge of stationary vane was modified. Since the discharge from the impeller forms a wake/jet flow, the leading edge was tapered with inclining angle from the shroud to hub. The ratio between radius of guide vane leading edge at hub and radius of impeller trailing edge was r G:V:L:E =r I:T:E ¼ 1:06 as shown in Fig. 2. The ratio 1.06 was chosen based on previous studies where the optimum distance between the impeller trailing edge and volute cut-off in centrifugal pump for noise reduction [14]. Furthermore an investigation by Shum et al. in determining the effect of impeller–diffuser interaction in centrifugal compressor, suggested that the optimal value between the impeller trailing edge and the diffuser leading edge should be between 1.05 and 1.09 [15].

Fig. 2. Geometry of the tested prototypes: (a) diffuser vane and (b) tapered guide vane.

98

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

Fig. 3. Actual and theoretical velocity diagram at impeller exit.

Table 1 Aerodynamics characteristics at operating conditions. Descriptions

Value

Rotational speed Flow rate

34,560 (rev/min) 0.0479 (m3 =s)

Table 2 Geometrical characteristics of the centrifugal fan. Description Radius of blade inlet (mm)

Shroud Hub

Inlet blade angle (deg) Radius of blade exit (mm) Angle of blade exit (deg) Blade number

Impeller

Diffuser

Tapered guide vane

18 18 9.2 51.5 32.18 11

53 53 2.8 62.5 19.8 15

53 54.6 14.14 62.5 13.8 15

In order to reduce the outflow impingement, the inlet for the guide vane was determined using velocity triangle at impeller exit as shown in Fig. 3. It should be noted that, slip factor must be taken into consideration when designing the guide vane. The slip factor is defined as actual tangential velocity at the impeller exit to the theoretical tangential velocity. According to Dixon, even under ideal conditions the relative flow leaving the impeller will not receive perfect guidance from the vane and the flow is said to slip [16]. Von Backström unifies all the centrifugal impeller slip factor prediction techniques based on a single relative eddy method (SRE) as shown in Eq. (1) [17]. After taking slip factor into account, the inlet angle for the guide vane were set to 14.14°. The aerodynamics and geometrical characteristics of the centrifugal fan are presented in Table 1 and 2. σ slip ¼ 1 

Δwslip 1 ¼ 1
 u2 1 þ 5 C=se ð cos βÞ0:5

(1)

where the blade length, C ¼ r e  r i = cos β and the blade spacing, se ¼ 2πr e =Z blade . 2.2. Aerodynamic performance measurement In order to obtain aerodynamic performance information of the centrifugal fan, a test facility was built as shown in Fig. 4. The centrifugal fan was fitted with long straight PVC pipe at the inlet in order to have the simplest geometry for CFD modeling. In this experiment the flow rate is obtained based on the static pressure difference across the orifice plate which was designed in accordance to ISO 5762-2 [18]. The inlet pressure of the centrifugal fan, the exit pressure from the impeller and the diffuser are measured with strain gauge pressure transducer. These pressure transducers are flushed mounted and the time averaged values are used to validate CFD model. The impeller rotational speed which corresponds to the universal motor rotational speed was measured by using inductive proximity sensor. Only one operating speed was considered in this work since the centrifugal fan is from a fixed-speed vacuum cleaner which uses cyclone system. Unlike conventional vacuum cleaner, vacuum cleaner with cyclone system does not have dust bag. The dusty air flows into dust tank, where the dirt particles and air particles are separated due to the centrifugal and gravitational force. By using the cyclone system the centrifugal fan constantly operates at its best efficiency point (BEP).The measured rotational speed was 34,560 rev/min and

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

99

Fig. 4. Schematic diagram of centrifugal fan test bench.

the number of blades was 11, so the blade passing frequency (BPF) would be 6336 Hz as calculated in Eq. (2) BPF ¼

nrpm  Z blade 60

(2)

The following uncertainties were established for the measured and calculated magnitudes: (I) (II) (III) (IV)

U-Tube Manometer: 719.6 Pa; Pressure Transducer: 71.7 kPa; Inductive Proximity Sensor: 75 percent; K-type Thermocouple: 72.2 1C.

2.3. Centrifugal fan acoustics measurement The acoustic measurement was conducted in an anechoic chamber based on the methodology stated in the ISO 3745 standard to determine the sound pressure level (SPL) as shown in Fig. 5 (a) [19]. The dimension of the chamber is 4.1 m  5.2 m  3.1 m and the cut-off frequency is 230 Hz. ISO Standards recommend that the radius of test hemi sphere should be either 1 m or twice the largest source dimension. For the current case, the centrifugal fan is the noise source and the hemi sphere is taken as 1 m [19]. The Gras Free Field 12 ″ Preamplifier microphone was used to measure the SPL. Before the measurements, the microphone was calibrated using the B&K 4231 sound level calibrator. The calibrator emits continuous sound pressure level of 94 dB at frequency 1 KHz, and the sound measuring equipment was calibrated against it to ensure the sensitivity and accuracy are within the acceptable uncertainties. These are the uncertainties for the acoustics devices: (I) Gras 12 ″ Free Field Preamplier microphone: 71 dB (5 Hz to 10 kHz) and 72 dB (3.15 Hz to 20 kHz) (II) B&K sound level calibrator: 70.2 dB The measured background noise of the anechoic chamber is 21.2 dB (A). The sound pressure level (SPL) is measured at each of the 20 positions as shown in Fig. 5 (b) and the overall surface-averaged SPL is computed using Eq. (3). " # Nm 1 X Lp ¼ 10 log 100:1 Lpi (3) Nm 1 where Nm is the number of microphone positions and Lpi is the background noise corrected sound pressure level at ith microphone positions in decibels

100

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

Fig. 5. (a) Centrifugal fan measured in semi anechoic chamber and (b) the microphone position on the measurement surface (ISO 3745:2012).

Fig. 6. Fluid domain used in numerical modeling (CFD).

3. Numerical modeling 3.1. Aerodynamic modeling This section presents the overview of the computational fluid dynamic (CFD) modeling using ANSYS FLUENT. The computational domain consists of 5 fluid domains: the upstream inlet, gap between the casing, the impeller, the volute, and the downstream outlet. These entire domains are attached using 5 conformal interfaces as presented in Fig. 6. Hybrid meshing of 4.62 million elements with tetrahedral and hexahedral type was used to cater for the component complexity. Tetrahedral meshing was used for the impeller and the volute–return channel volumes, while hexahedral used for the rest of the fluid volumes. This is similar to the meshing technique used by Khelladi et al. where study on impeller– diffuser interface of a centrifugal fan was conducted [20]. The unsteady term of implicit second order conservation equations was applied. Turbulence model k–ω SST variant of Detached Eddy Simulation (DES) was used, because it is capable in resolving eddies for noise prediction. In addition, DES models have been specifically designed to address high Reynolds number wall bounded flows [21]. According to Xu et al. Detached Eddy Simulation (DES) approach combines the advantages of RANS and LES [22]. DES turbulence formulation is also not as demanding as pure LES formulation and the resulting

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

101

Table 3 Comparisons between experiment and simulation for aerodynamic validation.

EXP CFD ϵR ð%Þ

φ

ψ

Πt;t

0.136 0.129 4.86

0.604 0.594 1.67

1.129 1.125 0.35

computational time saving is a significant factor to take into account [23]. SIMPLE algorithm is used for pressure–velocity coupling and second order upwind for the turbulence [24,25] For the boundary conditions, the inlet is defined as “pressure inlet” using the value from experimental measurement and the downstream outlet is specified as “pressure outlet”. Since the air flow from the centrifugal fan is discharged to ambient, the downstream outlet geometry is extended and assigned as atmospheric pressure. The surfaces that rotate relatively are defined as “moving wall”. Fluid zone in the inner volume is defined as “moving mesh” and the rotational speed defined as 34,560 rev/min. In order to define sliding mesh property, the same surfaces in the inner and outer volume families are defined as “interface”. The interfaces type boundary conditions are applied in between the inlet and impeller surface and also between the impeller outlet and the volute inlet. This configuration is known as non-conformal interfaces [11,26].The time was set to 4:8225  10  6 s so that the impeller mesh turns 11 per time step, where it takes 360 time steps to complete one revolution. The resolution of CFD time data is 32.73 time steps per blade passing frequency (BPF). In order to validate the CFD model, the results were compared with experiments. Performance of the centrifugal fan was obtained by measuring the required parameters at specific locations during the aerodynamic experiment. The performance parameters for validations are the flow coefficient φ, total pressure coefficient ψ, and the total-to-total pressure ratio, Πt;t as shown in Eqs. (4)–(6). The total pressure can be calculated based on the flow rate and static pressure from the aerodynamic experiment since the total pressure is sum of static pressure and dynamic pressure. φ¼

Q πb2 d2 u2

(4)

ψ¼

ΔP t 0:5 ρ u2 2

(5)

Πt;t ¼

P t;out P t;in

(6)

Comparisons between the measured and predicted results are shown in Table 3, where the relative error, ϵR for the generic variables, ϕ is defined as in Eq. (7):   ϕCFD  ϕexp  ϵR ð %Þ ¼  100 (7) ϕexp As seen in Table 3, the CFD model under-predicts the unsteady performance of the centrifugal fan. The under-predictions are probably due to the simplified geometry used to represent the universal motor in the modeling. In experiment the universal motor's commutator rotates to provide momentum to the air flow. However, in the simulation it is defined as stationary to reduce the total calculation time. Nevertheless the small errors between the prediction and experimental values indicate the unsteady flow simulation can be accepted.

3.2. Aeroacoustics modeling The Ffowcs, Williams, and Hawkings (FW–H) equation was used to predict the noise from the unsteady flow in the centrifugal fan. The FW–H equation is extension of Lighthill's acoustic analogy. The use of the FW–H equation allows noise sources to be separated into monopole, dipole, and quadrupole sources. [27] The FW–H equation is written as in Eq. (8):      1 ∂2 ∂ ∂ ∂2  2 ρvn δðf Þ∇f  ∇ n Pδðf Þ∇f þ T H ðf Þ (8) p¼ ∂t ∂xi i ∂xi ∂xj ij c0 2 ∂t 2 The first term is monopole sound source which is associated with the fluid volume displacement due to the surface motions. The second is dipole, which represents the sound generated due to unsteady forces exerted by the moving surfaces. The third term is quadrupole source, which represents generation of sound by volume distribution outside of surfaces [11,27,28]. Khelladi et al. showed that the dipole which results from the fluctuating pressure is the source of tonal noise generation in the centrifugal fan. [11] In addition, Liu et al. stated that the most prominent source of dipole in centrifugal fan in the rotating impeller and the stationary vane/cut-off. [28] The solution for FW–H equation in predicting the noise radiated by the dipole source is obtained by using Green functions in free field. The acoustics pressure in far field for a point force in

102

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

Fig. 7. (a) Comparison between measured and predicted tonal noise SPL of diffuser vane and (b) comparison of predicted tonal noise SPL between diffuser vane and tapered guide vane.

Table 4 Comparison of tonal noise sound pressure level between diffuser vane and tapered guide vane.

Diffuser vane (original) Tapered guide vane Difference

1st BPF (dB)

2nd BPF (dB)

3rd BPF (dB)

Overall (dB)

75.4 68.6 6.8

68.4 64.3 4.1

65.6 48.1 17.5

91.2 90.3 0.9

arbitrary motion is written as in Eq. (9): pðxi ; t Þ ¼



∂F i Fi ∂Mr þ 4πr 2 c0 ð1  Mr Þ2 ∂t ð1 M r Þ ∂t τ ðxi  yi Þ

Mi ¼

vi ; c0

Mr ¼

ðxi  yi Þ Mi r

(9)

Eq. (9) shows that the acoustic pressure of a moving point force is related to the time variation of force and velocity. The sound pressure level (SPL) is obtained by Fast Fourier transform algorithm. The SPL is represented as in Eq. (10): SPLðLp Þ ¼ 20log10

p ; pref

P ref ¼ 2  10  5 ½Pa

(10)

In aeroacoustics formulation, noise is assumed to be radiated into free field. Hence the effects scattering, diffractions, as well as the reflections of casing and blades are not considered [11,28]. The predicted SPL for the diffuser vane is compared with the experimental results in Fig. 7(a). The predicted values show similar pattern with the experiments, but there is consistent deviation in their magnitude. The deviation is contributed by the simplifications and assumptions in the CFD model. Nevertheless, focusing on the prediction patent, it can be said the CFD model is suitable for further analysis. Using the validated CFD model and aeroacoustics modeling, tonal noise for the proposed tapered guide vane is predicted and compared with the predicted SPL of diffuser vane as shown in Fig. 7(b). Reduction in tonal noise at the fundamental BPF and its harmonic is expected with the application of tapered guide vane. 4. Results and discussion The SPL frequency spectrum from the acoustics experiments comparing the diffuser vane and tapered guide vane is presented in Fig. 8. From the figure, the tapered guide vane shows reduction in the tonal noise at the BPF and its harmonics. Tonal noise at the fundamental BPF was reduced by 6.8 dB, where else for its harmonics in 2nd and 3rd harmonics was 4.1 dB and 17.5 dB respectively. The overall SPL was reduced by about 0.9 dB as presented in Table 4. As mentioned earlier, the tonal noise is generated due to the strong interaction between outflow from the impeller and the stationary vane. By referring to Eq. (9), tonal noise can be reduced by reducing the acoustic pressure at the leading edge. In addition, strength of the dipole source depends on the pressure fluctuation. Fig. 9 shows the pressure fluctuation at the leading edge of the diffuser vane and one will find there are 11 peaks for one period of rotation, 1:7361  10  3 s. As the impeller blade approaches the leading edge of a stationary vane, amplitude of the pressure increases and then decrease as it

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

103

Fig. 8. SPL frequency spectrum of diffuser vane and tapered guide vane.

Fig. 9. Comparison of pressure fluctuation at leading edge between the diffuser and guide vanes.

moves away. These 11 peaks represent the number of blade impeller and each of the pressure fluctuation periods correspond to the BPF. For the tapered leading edge, amplitude of pressure fluctuations is smaller compared to the leading edge of diffuser vanes. Fig. 10 clearly shows the smaller pressure distribution at the guide vane's leading edge compared to the diffuser vanes. The acoustics pressure generated by the guide vane is lower compared to the diffuser vane in Fig. 11. The jet/wake structure at the impeller exit exhibits non-uniform velocity profile causing the leading edge of the diffuser vane to experience strong pressure pulsation. Through the inclination of leading edge, the radial distance between the impeller and leading edge from shroud to hub is increased which makes the impingement on the leading edge more uniform. Hence this suppresses the strong peak sound at BPF. Besides that, leading edge inclination creates phase shift in the pressure fluctuation between the impeller and the leading edge, thus reducing the fluctuation strength [10]. Even though the overall noise reduction is quite low, but sound quality in term of human feeling may be different [29]. Besides the reduction in tonal noise, application of the tapered guide vane also reduces the broadband noise at the higher frequency from 14 kHz till 20 kHz. This improvement can be explained as the secondary effects of the tapered leading edge. The pressure fluctuation leads to the formation of turbulence flow which contributes to the increase in broadband noise. By reducing the fluctuation strength, the flow becomes less turbulent leads to the lowering of broadband noise as shown in Fig. 12. In term of performance, the tapered guide vane shows similar characteristics to the original diffuser vane. Table 5 presents the comparison between tapered guide vane and diffuser vane.

104

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

Fig. 10. Contour of static pressure on leading edge of stationary vane during interaction: (a) original diffuser vane and (b) tapered guide vane.

Fig. 11. Comparison of acoustic pressure fluctuation at leading edge between diffuser vane and guide vane.

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

105

Fig. 12. Contour of turbulent intensity (%) on leading edge of stationary vane during interaction: (a) original diffuser vane and (b) tapered guide vane.

Table 5 Comparison of aerodynamic performance between guide vane and diffuser vane.

Diffuser vane Tapered guide vane Difference (%)

Mass flow rate (kg/s)

Total-to total pressure ratio, Π t;t

0.0544 0.0543

1.129 1.131

0.2

0.2

5. Conclusion The effects of tapered guide vane on the tonal noise of centrifugal fan are presented in this paper. An acoustics measurement based on standard procedure was conducted to capture the spectral component of noise radiated by the centrifugal fan. Furthermore, CFD simulation of the aerodynamic in the centrifugal fan was performed which provides the information on the pressure fluctuation due to the interaction between the impeller and stationary vane. This interaction

106

K. Paramasivam et al. / Journal of Sound and Vibration 357 (2015) 95–106

causes the tonal noise radiation in the centrifugal fan. The application of tapered guide vane shows that the tonal noise can be reduced. The inclining angle with ratio of r G:V:L:E =r I:T:E ¼ 1:06, can reduce the tonal noise at fundamental BPF up to 6.8 dB with the overall noise reduction of about 0.9 dB. The results presented in this paper shows that pressure fluctuation's amplitude reduces the tonal noise. When the leading edge is tapered, the circumferential velocity profile from the impeller outflow become smoother hence reduces the pressure fluctuation amplitude. Moreover, this paper shows the broadband noise at higher frequency can also be reduced by using guide vane. As a final conclusion, it can be stated that the proposed tapered guide vane can reduces the tonal noise and broadband noise without affecting the performance of the centrifugal fan. Further studies on the geometric design parameters are required in order to determine the optimum design of tapered leading edge. Acknowledgments The authors would like to thank Universiti Teknologi Malaysia for the Flagship Grant VOT 01G49 and Energy Research Institute (ERI@N), Nanyang Technological University, Singapore for their support in allowing acoustics experiment to be conducted in their anechoic chamber facility. References [1] R.K. Turton, Principles of Turbomachinery, 2nd ed. Chapman & Hall, 1995. [2] Claus Wagner, Thomas Hüttl, Pierre Sagaut (Eds.), Large-Eddy Simulation for Acoustics, 1st ed. Cambridge University Press, Cambridge, 2007. [3] Daniel Wolfram, Thomas H. Carolus, Experimental and numerical investigation of the unsteady flow field and tone generation in an isolated centrifugal fan impeller, Journal of Sound and Vibration 329 (21) (2010). [4] Q. Liu, D. Qi, H. Tang, Computation of aerodynamic noise of centrifugal fan using large eddy simulation approach, acoustic analogy, and vortex sound theory, Proceedings of the IMechE, Part C: Journal of Mechanical Engineering Science 221 (2007) 1321–1332. [5] W. Neise, Noise reduction in centrifugal fan: a literature survey, Journal of Sound and Vibration 45 (3) (1976) 375–403. [6] W. Neise, G.H. Koopmann, Reduction of centrifugal fan by use of resonators, Journal of Sound and Vibration 73 (2) (1980) 297–308. [7] Soo Choi Jong, Aerodynamic noise generation in centrifugal turbomachinery, KSME Journal 8 (2) (1994) 161–174. [8] M. Cudina, J. Prezelj, Noise generation by vacuum cleaner suction units. Part I: noise generating mechanisms – an overview, Applied Acoustic 68 (5) (2007) 491–502. [9] W.H. Jeon, S. Baek, C. Kim, Analysis of the aeroacoustic characteristics of the centrifugal fan in a vacuum cleaner, Journal of Sound and Vibration 268 (2003) 1025–1035. [10] W.H. Jeon, S. Baek, C. Kim, Aeroacoustics characteristics and noise reduction of centrifugal fan for a vacuum cleaner, KSME International Journal 18 (2) (2004) 185–192. [11] S. Khelladi, S. Kouidri, F. Bakir, R. Rey, Predicting tonal noise from a high rotational speed centrifugal fan, Journal of Sound and Vibration 313 (2) (2008) 113–133. [12] Till Raitor, Wolfgang Neise, Sound generation in centrifugal compressors, Journal of Sound and Vibration 314 (3–5) (2008) 738–756. [13] Y. Ohta, T. Goto, E. Outa, Effects of tapered diffuser vane on the flow field and noise of a centrifugal compressor, Journal of Thermal Science 16 (4) (2007) 301–308. [14] W. Neise, Review of noise reduction methods for centrifugal fans, Journal of Manufacturing Science and Engineering 104 (2) (1982) 151–161. [15] Y.K.P. Shum, C.S. Tan, N.A. Cumpsty, Impeller–diffuser interaction in a centrifugal compressor, Journal of Turbomachinery 122 (4) (2000) 777–786. [16] S.L. Dixon, C.A. Hall, Fluid Mechanics and Thermodynamics of Turbomachinery, 6th edition, Butterworth-Heinemann, Boston, 2010. [17] W. Von Backström Theodor, A unified correlation for slip factor in centrifugal impellers, Journal of Turbomachinery 128 (1) (2005) 1–10. [18] ISO 5167-2:2003 – Measurement of Fluid Flow by Means of Pressure Differential Devices Inserted in Circular Cross-section Conduits Running Full – Part 2: Orifice Plates. [19] ISO 3745:2012 – Acoustics: Determination of Sound Power Levels and Sound Energy Levels of Noise Sources Using Sound Pressure – Precision Methods for Anechoic Rooms and Hemi-anechoic Rooms. [20] S. Khelladi, S. Kouidri, F. Bakir, Flow study in the impeller–diffuser interface of a vaned centrifugal fan, Journal of Fluids Engineering 127 (3) (2005) 495–502. [21] ANSYS FLUENT User Guide, 2011. [22] Chang-Yue Xu, Li-Wei Chen, Xi-Yun Lu, Large-Eddy and detached-Eddy simulations of the separated flow around a circular cylinder, Journal of Hydrodynamics, Series B 19 (5) (2007). [23] S. Fontanesi, S. Paltrinieri, G. Cantore, CFD analysis of the acoustic behavior of a centrifugal compressor for high performance engine application, Energy Procedia 45 (2014) 759–768. [24] P.R. Spalart, W.H. Jou, M. Strelets, S.R. Allmaras, Comments on the feasibility of LES for wings, and on a hybrid RANS/LES approach, Advances in DNS/LES 1st AFOSR International Conference on DNS/LES,Greyden Press, Columbus, 1997. [25] M. Tournour, Z.E. Hachemi, A. Read, F. Mendonca, F. Barone, P. Durello, Investigation of the tonal noise radiated by subsonic fans using the aeroacoustic analogy, Proceedings of the Fan Noise International Symposium, 2003. [26] S. Rama Krishna, A. Rama Krishna, K. Ramji, Reduction of motor fan noise using CFD and CAA simulations, Applied Acoustics 72 (2) (2011) 982–992. [27] F. Farassat, K.S. Brentner, The acoustic analogy and the prediction of the noise of rotating blades, Theoretical and Computational Fluid Dynamics 10 (1–4) (1998) 155–170. [28] Q. Liu, D. Qi, Y. Mao, Numerical simulation of centrifugal fan noise, Proceedings of the IMechE, Part C: Journal of Mechanical Engineering Science 220 (2006) 1167–1178. [29] W.H. Jeon, A numerical study on the acoustic characteristics of a centrifugal impeller with a splitter, GESTS International Transactions of Computer Science and Engineering 20 (1) (2005).