Front. Energy Power Eng. China 2008, 2(4): 433–437 DOI 10.1007/s11708-008-0082-6
RESEARCH ARTICLE
Jianfeng MA, Datong QI, Yijun MAO
Noise reduction for centrifugal fan with non-isometric forwardswept blade impeller
E
Higher Education Press and Springer-Verlag 2008
Abstract To reduce the noise of the T9-19No.4A centrifugal fan, whose impeller has equidistant forward-swept blades, two new impellers with different blade spacing were designed and an experimental study was conducted. Both the fan’s aerodynamic performance and noise were measured when the two redesigned impellers were compared with the original ones. The test results are discussed in detail and the effect of the noise reduction method for a centrifugal fan using impellers with non-isometric forward-swept blades was analyzed, which can serve as a reference for researches on reduction of fan noise. Keywords centrifugal fan, non-isometric blade impeller, noise reduction
1
Intr In trod oduc ucti tion on
One of the effective methods to reduce the noise of a fan is to control the noise sources. There are several ways to control the aerodynamic noise sources of centrifugal fans [1], but applying the same noise-reduction method to different kinds of fans may lead to different results. One method met hod may obt obtain ain not notice iceabl ablee noi noise se red reduct uction ion,, whi while le another may show no effect at all. Therefore, it is necessary that further investigations be conducted in this field. A fan’s aerodynamic noise is often dominated by tones at the blade passing frequency (BPF) and its higher harmonics caused by the strong flow interaction between the Translated from Fluid Machinery, Machinery, 2007, 35(9): 1–4, 37 [译自: 流体机 械] Jianfeng MA School of Energy and Power Engineering, Xi’an Jiaotong University, Xi’an 710049, China; Shenyang Blower Works Group Co., Ltd, Shenyang 110142, China E-mail:
[email protected] Datong QI, Yijun MAO Datong School of Energy and Power Engineering, Xi’an Jiaotong University, Xi’an 710049, China 1)
rotating impeller and stationary volute casing. To reduce the BPF sound pressure level (SPL), the method of using non-isometric blades has been proposed as this can disperse the SPL at BPF and reduce the total SPL. In 196 1967, 7, Lo Lowso wson n pro propose posed d tha thatt by dis distri tribut buting ing bla blade de spacing unequally, the BPF sound level of axial flow compressors can be reduced [2]. In 1970, Melin and Sovran also suggested that the method of using non-isometric blades can reduce reduce the discrete discrete noi noise se lev level el of the im impel peller ler.. The They y deduced an equation to compute the discrete sound level of non-isometric blades [3]. Later, Ewald [4] adopted some similar noise-reducing methods by which the BPF sound level of a 22-blade axial flow fan was decreased by about 8 dB. Other researches conducted by Duncan also corroborated this method of reducing the sound level by spacing blades unequally on the impeller of an axial flow fan [5]. In 1980, using the blade spacing model, Krishnappa 1) theoretically ica lly ded deduce uced d gov govern erning ing equ equati ations ons of sou sound nd rad radiat iation ion,, con con-side si deri ring ng no nonn-is isom omet etri ricc bl blad ades es fo forr th thee wh whol olee fa fan, n, an and d conducted cond ucted several exper experimen iments. ts. In Chin China, a, Sun Xiao Xiaofeng feng discus dis cussed sed the inf influe luence nce of a non non-is -isom ometri etricc bla blade de on the BPF sound level at the observation point of an axial flow fan [6 [6]. ]. Du Duee to de devel velop opme ment ntss in ae aero rona naut utic ics, s, ea earl rlie ierr resear res earche chess mos mostly tly focused focused on the axial flow fan using a non-isomet non-i sometric ric blad blade, e, but few researc researches hes were conducted on a centrifugal fan with forward-swept blades. In this paper, based on previous studies which introduced non-i non-isomet sometric ric blades blades for centri centrifugal fugal fans, fans, the effect effectss of two types of different blade spacing impellers on aerodynamic performance and aerodynamic noise in a centrifugal fan with forward-swept blades were compared and analyzed. This centrifugal fan used had 12 forward curved blades driven by a 3.0 kW AC motor, rotating at 2900 r/min. r/m in. It con consis sisted ted of an imp impell eller, er, a van vanele eless ss dif diffus fuser, er, and a volute. The blade angles at the inlet and outlet were 38u and 126u, respectively, and the inlet and outlet diameters of the impeller were 156 and 400 mm, respectively. The vaneless diffuser, whose outlet diameter is 460 mm, rotated with the impeller.
Krishnappa G. Effect of modulated blade spacing on centrifugal fan noise. Proceedings of Inter-Noise’80, 1980, 215–218
434
Jianfeng MA, et al.
2 Design method of non-isometric blade impeller
Hence, the span of phase-modulated quantity A is obtained by using the above constraint condition, as shown in Table 1.
Two constraints should be considered when designing a non-isometric blade impeller. One is that after modifying the blade spacing, the dynamic balance condition should be guaranteed when the impeller is running. The other is that the variation of the angle between blades should not be too large; otherwise it may cause a large deterioration of aerodynamic performance. The circumferential angles of blades with respect to the original and redesigned fans are denoted by h and h9. Assuming that the first circumferential angle of 0 the blades is h1 ~h1 ~0 , the other circumferential angle of 0 0 0 the blades are [h2,h3,…,hN ] and h2 ,h3 , . . . ,hN . Since the tested fan is running at low speed, the influence of aerodynamic load distribution on the dynamic balance of the impeller is ignored. Thus, the mathematical representation of the dynamic balance constraint is 0
h
N
P cos
1z
i ~2 N
P sin
i ~2
0
i
9> 0, > = >; 0: >
hi ~
0
hi ~
0
0
hi z1 {hi f
2p za, N
ð2Þ
0
Dh~hi z1 {hi ~ðhi z1 {hi ÞzA½sinðshi z1 Þ{ sinðshi Þ
s|30 sð2hi z30 Þ cos : 2 2 0
~30 z2A sin 0
0
ð3Þ
The above formula should satisfy
s A sin
|30
0
2
cos
Experimental installation and results
A test installation for the aerodynamic and acoustic characteristics of the investigated fan was designed according to China standards GB/T2888-91: Methods of Noise Measurement for Fans, Blowers Compressors and Roots Blowers, and GB/T1236-2000: Industrial FansPerformance Testing Using Standardized Airways [7,8]. Figure 1 is the illustration of this test installation with its main elements. In order to compare the original impeller with the redesigned ones, the same volute, air inlet, inlet duct, electric motor and measurement system are used. Hence, the only difference between the original and redesigned fans is the use of different impellers. The performance curves and the SPL spectra are obtained in a half anechoic chamber. In Figs. 2 and 3, the A-weighted SPL and specific A-weighted SPL are represented. The total pressure and the efficiency are represented as a function of the volume flow, as shown in Figs. 4 and 5. The SPL spectra for the investigated fans operating at three volume flow rates are shown in Figs. 6– 8. The test results show that the aerodynamic performances of the two redesigned fans are similar, but the noise of redesigned-fan II is weaker than that of redesigned-fan I. Thus, only the comparison between the redesigned-fan II and the original fan is presented. The redesigned-fan II attains its maximum efficiency when the volume flow is about 17 m3/min. In the Table 1
30 Þ 2
sð2hi z
According to the design method described in the previous section, two different types of blade design scheme are employed. For convenience, the unmodified fan is called the original fan and the two redesigned fans are called redesigned-fan I and redesigned-fan II, respectively. Table 2 shows the specific parameters of these two design schemes, wherein a is the circumferential angle, and Da is the circumferential angle interval. Due to the limitation of machining accuracy and processing conditions in the fan factory, it is inevitable that there exists some deviation in the impeller dimension from that in the drawing paper. Both the design values and the actual dimensions of the redesigned impellers are shown in Table 2.
4
where N is the blade number. The selection of angle a depends on different conditions, and in this paper, a 5 5u. Many different schemes of combining blades can be derived from meeting the constraint conditions, i.e., Eqs. (1) and (2). In this paper, the blade grouping scheme detailed in Ref. [6] is adopted. In this scheme, the circumferential angle of blade No. i of the redesigned impeller is 0 defined as hi ~hi zA sinðshi Þ, where s is the number of groups, A is the phase-modulated quantity. The compound mode of a blade on the T9-19No.4A centrifugal fan was redesigned according to the above formula. To keep the aerodynamic performance of the fan almost unchanged, the circumferential angle interval of the blades was designed to be between 25 u and 35u. The angle interval between two adjacent blades is Dh 0
Selection of design scheme
ð1Þ
The geometric constraint condition on the circumferential angle interval between two adjacent blades is 2p {af N
3
0
2:5 :
v
0
ð4Þ
Span of phase-modulated quantity s
1
2
3
4
span of A/(u)
0–9.6
0–5
0–3.5
0–2.8
Noise reduction for centrifugal fan Table 2 No.
Parameters of two types of design scheme redesigned-fan I (s 5 2,A 5 5)
original fan a/(u)
1 2 3 4 5 6 7 8 9 10 11 12
435
30 60 90 120 150 180 210 240 270 300 330 360
design value
actual value
design value
actual value
a/(u)
Da/(u)
a/(u)
Da/(u)
a/(u)
Da/(u)
a/(u)
Da/(u)
34.3 64.3 90 115.7 145.7 180 214.3 244.3 270 295.7 325.7 360
34.3 30 25.7 25.7 30 34.3 34.3 30 25.7 25.7 30 34.3
35 63.5 88 116 144.5 178 213.5 243.5 270.5 295 324.5 358.8
35 28.5 24.5 28 28.5 33.5 35.5 30 27 24.5 29.5 35.5
33.5 60 86.5 120 153.5 180 206.5 240 273.5 300 326.5 360
33.5 26.5 26.5 33.5 33.5 26.5 26.5 33.5 33.5 26.5 26.5 33.5
32.5 58.5 84.5 118 151.5 179 206 238.5 272 298.5 325 359
32.5 26 26 33.5 33.5 27.5 27 32.5 33.5 27 27 34
Fig. 1
Test installation
A-weighted SPL spectra, there is a conspicuous peak at BPF (i.e., at 595.66 Hz), which is slightly lower than that of the original fan. At 749.89 Hz, a new peak emerges, while there is no such evident one in the SPL spectra of
Fig. 2
redesigned-fan II (s 5 3,A 5 3.5)
A-weighted SPL vs volume flow
the original fan. The deviation of the total SPL between redesigned-fan II and the original fan is small. In small volume flow (about 7 m3/min), the SPLs of the original fan at BPF and 668.34 Hz are almost the same. The SPLs
Fig. 3
Specific A-weighted SPL vs volume flow
436
Jianfeng MA, et al.
Fig. 4
Total pressure vs volume flow
Fig. 7 Comparison of measured A-weighted SPL at three volume flow (redesigned-fan II)
mode. At an even larger flow condition (34 m3/min), in the entire SPL spectra of the original fan, the main contributions are noise whose frequencies are BPF at 578.76 Hz, whose second harmonic is at 1154.78 Hz and whose third harmonic is at 1727.83 Hz, respectively, of which the BPF SPL is the maximum. Compared with the original fan, the total A-weighted SPL of redesigned-fan II increases by 1 to 4 dB at this operating mode. In addition, as far as aerodynamic performance is concerned, the total pressure of redesigned-fan II drops a little over the entire operating mode, as does its efficiency.
5 Fig. 5
Efficiency vs volume flow
of the redesigned-fan II at BPF and 749.89 Hz are obviously lower than that of the original fan. Compared with the original fan, the total A-weighted SPL of redesigned-fan II is reduced by 1 to 2.5 dB at this operating
Fig. 6 Comparison of measured A-weighted SPL at three volume flow (original fan)
Conclusions
The method of noise reduction for a centrifugal fan with a non-isometric forward-swept blade impeller is studied. Some key conclusions of this investigation are summarized as follows. 1) In this paper, the attempt to adopt a non-isometric blade impeller for a centrifugal fan is not successful.
Fig. 8 Comparison of measured A-weighted SPL at three volume flow (redesigned-fan I)
Noise reduction for centrifugal fan
Compared with the original fan, at the domain of larger flow, the sound level of the redesigned fan is increased by about 1 to 4 dB, with the maximum efficiency operating mode and the sound level of the redesigned fan almost remaining unchanged or slightly reduced. However, at small flow, the sound level of the redesigned fan can be reduced by about 1 to 2.5 dB, which indicates that the possibility of fan noise reduction does exist. Further investigations and studies should be conducted on the adoption of a non-isometric blade impeller. 2) The test indicates that by using a non-isometric blade impeller, the sound levels of certain frequencies of the centrifugal fan with a forward-swept blade can be changed, which provides the possibility of noise reduction and sound quality improvement for a centrifugal fan with a forward-swept blade. 3) There exists some deviation in impeller dimension from that in the drawing paper; hence, it is difficult to analyze the regularity of some experimental phenomena. More attention should be paid to working on accuracy in future research. 4) There are only two improvement schemes in this test. More improvement schemes should be attempted to explore more appropriate distribution regularities of a non-isometric blade to obtain more believable analyses and conclusions. 5) The test indicates that changing the isometric blade into a non-isometric blade may lead to deterioration in the aerodynamic performance of the fan. It is suggested that the blade profile be modified, with changes in the circumferential spacing of blades, so as to explore the best design method for blade profiles while using a non-isometric blade. This can both keep better flow in the blade channel
437
of the impeller and maintain the aerodynamic performance of the fan. Moreover, it may be of benefit for displaying the noise reduction effect of non-isometric blades. 6) The fan investigated in this test is a T9-19No.4A centrifugal fan. The method described in this paper is mainly for small types of forward-swept centrifugal fans which have a higher pressure coefficient and a smaller flow coefficient. It is hoped that this paper can serve as a reference for research in similar products.
References 1. Zhong Fangyuan. Translation of Blade Machinery on Aeroacoustic. Beijing: China Machine Press, 1987 (in Chinese) 2. Lowson M V. Reduction of compressor noise radiation. Journal of the Acoustic Society of America, 1968, 43(1): 37– 50 3. Mellin R C, Sovran G. Controlling the tonal characteristics of aerodynamic noise generated by fan rotors. Journal of Basic Engineering (Series D), 1978, 72: 143–154 4. Ewald D, Pavlovic A, Bollinger J G. Noise reduction by applying modulation principles. Journal of the Acoustical Society of America, 1971, 49: 1385 5. Duncan P E, Dawson B. Reduction of interaction tones from axial flow fans by suitable design of rotor configuration. Journal of Sound and Vibration, 1974, 33(2): 143–154 6. Sun Xiaofeng. The aeroacoustic nature of unequally spaced fan. Journal of Beijing Institute of Aeronautics and Astronautics, 1986, (4): 137–145 (in Chinese) 7. Xu K C. Fan Manual. Beijing: China Machine Press, 1999 (in Chinese) 8. Standardization Administration of People’s Republic of China. GB/T2888-91 Classification Compilation of Standards of China, Machinery Volume (20). Beijing: Standards Press of China, 1993 (in Chinese)
