1、432 2009 ASHRAEThis paper is based on findings resulting from ASHRAE Research Project RP-1223.ABSTRACTFan performance data measured as installed may show lower performance than manufacturer ratings, primarily because of improper inlet or outlet connections. An apparatus and test procedure to experim
2、entally measure air and sound performance of propeller fans with systematic variation of inlet flow components is described. The test program is intended to simulate typical “in the field” installations of the fans. INTRODUCTIONFan performance data measured as installed may show lower performance th
3、an manufacturer ratings, primarily because of improper inlet or outlet connections. In this study, experimentally measured air and sound performance of propeller fans with systematic variation of inlet flow compo-nents, intended to simulate typical “in the field” installations of the fans, was consi
4、dered. The fan inlet appurtenances consisted of mitered elbows mounted at various angles at the fan inlet plane, inlet duct contractions of various area ratios, and walls perpendicular to the fan axis which were located at various distances from the inlet. The resulting aerodynamic performance data
5、are presented in Young et al. (2009a), and acoustic data are presented in Young et al. (2009b). The pres-ent paper documents the test apparatus and procedure that were developed in order to experimentally measure air and sound performance of propeller fans with systematic variation of inlet flow com
6、ponents, so as to simulate typical fan instal-lations. For complete details, refer to Darvennes et al. (2008).Many researchers have noted that fan performance can be affected by fan inlet or outlet connections, for example see Traver (1970). It was observed that optimum fan performance requires that
7、 a uniform velocity without swirl be present at the inlet to the fan, and that a duct of sufficient length be placed at the fan outlet in order to permit the disturbed fan flow to achieve a uniform velocity without significant swirl. Christie (1971) performed tests with various inlet elbows affixed
8、to fans. It was concluded that a fan which is selected to operate near its best efficiency will more nearly approach its open inlet performance than a fan which is expected to operate at a rating point which is near free delivery, when inlet boxes or elbows are present. It was also observed that the
9、 use of a larger fan operating at a lower speed should help to reduce performance problems caused by poor inlet conditions.Seminal work regarding system effects was first reported by Farquhar (1973), Meyer (1973), and Brown (1973). Farqu-har (1973) noted that fan performance measured in the labo-rat
10、ory often displayed significant differences from that measured in the field. These differences were attributed to flow conditions near the fan inlet or outlet; for the first time they were referred to as SEFs. These factors were quantified in terms of the familiar system effect curves, and procedure
11、s for their use were described. It was claimed that considerable judgment must be employed in using these curves, since every field installation is different. Meyer (1973) discussed the determination of duct system characteristics essential to the selection of a fan from a fan performance curve in t
12、erms of duct system characteristics. It was noted that the system itself is capable of adversely affecting the performance of a fan, and this system effect must be included in the determination of the characteristics of the system to permit the proper selection of the fan. Some common causes of syst
13、em effect were cited, Test Apparatus and Procedure to Measure Inlet Installation Effects of Propeller FansM.N. Young, PhD C. Darvennes, PhD S. Idem, PhDMember ASHRAEM.N. Young is an engineer with the Tennessee Valley Authority, Knoxville, TN. C. Darvennes and S. Idem are professors in the Depart-men
14、t of Mechanical Engineering, Tennessee Tech University, Cookeville, TN.LO-09-039 (RP-1223) 2009, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (www.ashrae.org). Published in ASHRAE Transactions 2009, vol. 115, part 2. For personal use only. Additional reproduction,
15、distribution, or transmission in either print or digital form is not permitted without ASHRAEs prior written permission.ASHRAE Transactions 433namely nonuniform fan inlet flow, swirl at the fan inlet, and improper fan outlet conditions. All of these conditions adversely affect the fan, such that it
16、will not achieve its rated performance.Brown (1973) characterized the test apparatus and proce-dures required to measure system effects. Briefly, the system effect of a fitting can be determined through a series of two tests: (1) with the fan set up as it would be for a performance test without the
17、appurtenance installed and (2) with the fan set up with the fitting installed. Therein the effect of the fitting can be determined by comparing the results of the two tests. These methods were employed by Clarke et al. (1978) to measure system effects associated with centrifugal fans. It was propose
18、d to calculate a dimensionless factor (the ratio of the average fan inlet or outlet velocity to the tip speed) to correlate the data. Therein SEF expressed as a percentage of the inlet or outlet velocity pressure was presented as a function of this dimensionless velocity. This approach was found to
19、success-fully collapse SEF data onto a single curve, for each fan appur-tenance. Likewise Zaleski (1988) measured system effects related to axial fans. The experimental results of Clarke et al. (1978) and Zaleski (1988) constitute the majority of the SEF data presented by AMCA in their Publication 2
20、01 (2002). Likewise these system effect factors, converted to local loss coefficients, are in the ASHRAE Duct Fitting Database (2006).Other attempts to experimentally characterize system effects were reported in a series of papers presented by Cory (1982, 1984a, 1984b), wherein data for the influenc
21、e of inlet and outlet elbows on fan performance were presented. Like-wise the discrepancies in measured system effects obtained using different test protocols were discussed. The inclusion of an elbow directly at the fan inlet significantly affected fan performance, whereas the addition of only 1D o
22、f straight duct between the elbow and the fan recovered some of this loss. Elbows with 2.4D of straight duct between the elbow and the inlet to the fan resulted in a loss which could be accounted for by a straight duct pressure drop calculation. As the radius of the bend in the elbow became larger,
23、the system effects were less apparent. It was observed that system effects were more pronounced towards free delivery, but not in proportion to the square of the velocity. Therein it was claimed that the data correlation approach advocated by Clarke et al. (1978) was inappropriate. Coward (1990) lik
24、ewise attributed system effects to uneven velocity profiles or swirling flow entering or leaving the fan. In practice the unrecovered energy lost due to system effects appears as an additional pressure loss (perhaps unan-ticipated by the design engineer), such that there may be a corresponding reduc
25、tion in the air flow below the design value. It was noted that fan performance tests are conducted with precisely defined inlet and outlet conditions which provide carefully controlled inlet and outlet velocity profiles. This ensures optimum fan performance and repeatability from one laboratory to a
26、nother. However, these conditions are rarely achieved in field installations where such issues as space limi-tations or the presence of fittings, dampers, or belt guards, etc., close to the fan inlet or discharge interfere with the desired uniform flow. Bolton (1990) reviewed previous work related t
27、o changes in fan aerodynamic, acoustic, and vibration perfor-mance caused by fan-system interactions. Bevirt (1992) stated that the system effect increases proportionally to the square of the fan inlet or discharge velocities. Therein employing larger fans operating at slower speeds with lower inlet
28、 and outlet velocities could also potentially alleviate system effects.Vanderburgh and Paulauskis (1994) qualitatively discussed system effects on fan aerodynamic and acoustic performance. They commented that for optimum fan effi-ciency air flows along the fan blade for a substantial portion of its
29、length before it experiences separation from the blade. Flow separation results in shed vortices, which contribute to the noise generated by the fan. If system effects force the fan to operate at other than the design point, the separation point tends to move forward along the surface of the blade.
30、Under those circumstances the fan operates with decreased airflow, and generates additional noise while requiring greater power consumption. Therein it may be possible to restore the system to the design airflow by increasing the fan speed. However, this requires additional fan horsepower, and it wa
31、s claimed that fan noise may increase by up to 5 dB. However, it was noted that fan system interactions may even cause the fan to stall, thereby causing 15-20 dB additional noise accompanied by a drastic air flow deficiency. This requires either that the fan be replaced, or the entire system be rede
32、signed. It was further noted that a classical description of fan noise distinguishes between two primary noise components, namely vortex flow noise and rotational flow noise. Vortex noise comes from the vortex shedding at the fan blade and impingement of vortices on other fixed surfaces. This noise
33、output is termed “broad-band” regarding the frequency spectrum. The rotational noise derives from pulses of air, which are created each time a fan blade passes a fixed surface. Therein the noise output occurs at discreet frequencies of sound (pure tones), which is often referred to as the blade pass
34、ing frequency. However, they described another frequently ignored noise source attributed to the vibration of fan enclosures or motor housings enclosures that resonate with flow turbulence and vorticity, as well as fan or motor frequencies. This phenomenon often leads to a low frequency noise termed
35、 duct rumble.Swim (2005) described a series of tests wherein the distance of a wall from the inlet plane of various axial fans was varied. It was concluded that at distances greater than 1/2D, the pressure losses were too small to be effectively measured. Murphy (2005a) employed airfoil theory in an
36、 attempt to predict flow performance degradation as measured by Swim. Sound data were also presented for various inlet configura-tions in an effort to explain the flow performance data. It was concluded that the change in efficiency due to inlet distortion was essentially zero, until the wall was pl
37、aced unreasonably 434 ASHRAE Transactionsclose to the inlet. Therein Murphy (2005b) related the vortex shedding frequency to the sound data measured by Swim.This research program consisted of measuring air and sound performance of propeller fans with a systematic varia-tion of inlet flow components.
38、 The objective of this test program was to measure inlet system effects for both air and sound for three sizes of small propeller fans of the same series, at two fan speeds. The aerodynamic and acoustic investiga-tions were performed in accordance with the ANSI/AMCA 210/ASHRAE 51 (1999) air performa
39、nce testing standard and the AMCA 300 (1996) reverberant room sound testing stan-dard. The apparatus was set up per Figure 1 (which is excerpted from ANSI/AMCA 210/ASHRAE 51 (1999). The fan was situated in a reverberant chamber, thereby permitting measurement of fan sound power level. Discharge duct
40、s having a length of two to three times the fan impeller diameter were used to connect the fan to a nozzle chamber; this was employed to measure fan static pressure and flow rate. These short ducts had a square cross section closely matched to the height and width of the fan. Baseline air performanc
41、e and sound data were measured using an “ideal” installation with an open inlet. Subsequently appurtenances were mounted adja-cent to the fan inlet plane in order to determine the magnitude of the system effect. A unique aspect of the present test program was that aerodynamic and acoustic performanc
42、e data were obtained simultaneously on the test apparatus. The test procedure and apparatus are herein described in detail.TEST APPARATUSThe three fans chosen for study in this test program were six-blade propeller fans having nominal impeller diameters of 610-mm (24-in), 914-mm (36-in), and 1219-mm
43、 (48-in). In every instance the fans were belt-driven, with the motor offset from the fan axis. Each 3-, 240-V fan motor was nominally rated at 1.0-, 2.0-, and 5.0-hp, respectively. The fans were intended to be of a high efficiency, low noise design, consist-ing of the same series from a single manu
44、facturer.The inlet appurtenances considered in this study included a 90 round inlet duct with a mitered elbow; a schematic diagram is shown in Figure 2. The dimensions of the duct that was parallel to the fan axis, and between the fan inlet and the elbow, were equal to the diameter of the fan impell
45、er. Similarly the distance along the side of the inlet duct upstream of the elbow, which was normal to the fan axis and between the open end of the inlet duct and the elbow, was equal to the diameter of the fan impeller. Where possible, the opening of the inlet section of inlet duct was positioned f
46、ive angular positions rang-ing from 0 to 270 (relative to the vertical orientation) in 45increments. In addition, three round inlet ducts with contrac-tion area ratios of 1.0, 1.25, and 1.5 were mounted at the fan inlet plane; a schematic diagram is depicted in Figure 3. The length between the fan i
47、nlet and the downstream end of the transition section was one fan impeller diameter. Similarly the length of ductwork between the open end of the inlet duct and the upstream end of the transition section was one fan impeller diameter. In every instance the transition incorporated a 90 included angle
48、. The inlet fittings were constructed from combi-nation of 18 and 20-ga galvanized steel.Likewise a wall perpendicular to the fan axis was posi-tioned at various distances away from the fan inlet, at locations including 0.25, 0.5, 0.75, 1.0, 1.25, 1.5, and 2.0 times the fan impeller diameter; a sche
49、matic diagram is included in Figure 4. In every case the wall height and width were three times the fan impeller diameter. The wall was centered horizontally and vertically to the fan axis. The plywood wall was mounted to a collapsible gate fence to give a sturdy foundation. The gate fence was sandwiched between two sheets of 19-mm (3/4-in) thick plywood. Rubber strips were installed between the mounting surfaces of the gate and plywood wall in order to Figure 1 Outlet chamber setupmultiple nozzles in chamber.Figure 2 Schematic of 90 round inlet duct with mitered elbow.ASHRAE Tran
