1、451 2 An Experimental Investigation of Turbulent Wall Jets Zou Yue, Ph.D., P.E. ABSTRACT In this papel; wall jets discharged from round nozzles (D = 43 mm 1.7 in., 76 mm 13 in., and 152 mm 6 in.) were tested in a room (4.2 m x 3.6 m x 2.7 m 13.8ft x 11.8 ft x 8.9 ft) and measurement results show the
2、 follow- ing. In the case of big nozzles (D = 76 mm 3 in. and I52 mm 6 in.), the maximum velocity decay coefficient K-values of a wall jet can always be estimated by multiplying 2/2 with those of a free jet with the same exit area. Howevel: for the small size nozzles (D = 43 mm 1.7 in.), K-values be
3、come lower: The end maximum velocity of zone 3 could be propor- tional to outlet velocity, and gradient can be related to room size. A smaller test room gives a higher value. INTRODUCTION As stated in ASHRAE handbooks, the objective of heat- ing, ventilation, and air-conditioning systems is to creat
4、e a proper combination of temperature, humidity, and air motion in the occupied zone. If enough heating or cooling capacity is available to maintain the desired average temperature and humidity within the space, the ability of the system to satisfy the comfort requirements of the occupants will then
5、 almost certainly depend on the air system. The three-dimensional wall jet is one of the most impor- tant flow patterns for mixing air distribution systems. The jet follows the ceiling, entrains air from the occupied area, and generates a recirculating flow in the conditioned room. The wall jet flow
6、 is strongly influenced by the condition of the diffuser and surrounding details, such as the distance to the ceiling and the ceiling structure. For instance, if there is a certain distance between the supply outlet and ceiling, the Coanda effect will give rise to a net force on the jet curved towar
7、d the ceiling (see Figure 1). However, three-dimensional wall jets still have many features in common with their free jet counterparts. Both flow fields may be conveniently divided into four distinct regions based on the streamwise decay of the maximum mean velocity: (1) potential core zone in which
8、 the maximum velocity of the airstream remains practically unchanged; (2) transition zone in which the orifice condition has an important effect; (3) fully developed turbulent zone where the velocity profile is similar and flow behaves as if it were generated by a point source; and (4) termination z
9、one where the maximum velocity decreases rapidly and the jets start to degenerate into room air movement. The flow charac- teristics in this last zone are not well understood. In this paper, some measurement results of three-dimen- sional wall jets from nozzle-type outlets under isothermal condition
10、s are presented. Investigation is concentrated on the ,-Y ,Attachment point I core I transition I profile similarity i termination Zone ? Zone 3 %one 4 1 X I Figure 1 A non-jlushed-mounted three-dimensional wall jet. Zou Yue is a researcher in the Building Science Department, Royal Institute of Tech
11、nology, Stockholm, Sweden. ASHRAE Transactions: Research 203 flow characteristics at centerline in the ceiling region since it is the background for “throw length“ calculation. Experimen- tal data could also be used to evaluate and improve numerical simulation models of room air motion. LITERATURE R
12、EVIEW The earliest three-dimensional wall jet work in HVAC applications to which we have had access is Farquharson (1952) in which the effect of the proximity of a wall on the behavior of the jet from square orifices was studied. Measure- ment results showed that the maximum velocity decay coef- fic
13、ient K, which is 6.5 in a free jet, should be replaced by 9.0 when the edge of the discharge opening is adjacent to a wall. This work was followed by those of Tuve (1953), Koestel (1957), and Sforza and Herbst (1970). Sforza and Herbst (1 970) presented an extensive experimental investigation of the
14、 mean properties of turbulent wail jets from various rect- angular orifices. From the results obtained, the wail jet flow field was found to be characterized by axial velocity decay. An analytical approach to estimate the shear stress distribution at the plate was also presented. Nielsen and Mller (
15、1988) measured a nozzle-like diffuser (D = 140 mm 5.5 in.) located close to the ceiling under both isothermal and thermal conditions and found that the flow was independent of the Reynolds number for all the tests (U, 2.9 mis 571 ft/m). The temperature distribution was independent of the Archimedes
16、number for all supply velocities and could also be described by an equation similar to the velocity distribution equations. A jet produced from a rectangular outlet was studied by Kirkpatrick and Kenyon (1998). Using a simple jet model, the general nature of jet flow characteristics, such as velocit
17、y decay coefficients, virtual origins, and spread angles, was deduced and compared with previous studies by Sforza and Herbst. In Sandberg (1998), experimental studies of both three- and two-dimensional jets in rooms were presented, and it was found that the traditional wall jet theory, based on the
18、 expan- sion of a jet in an infinite ambient, was also useful in room air flow. These studies also showed that there may be confine- ment phenomena that have strong influence on jet develop- ment. Finally, a good number of publications on the character- istics of turbulent wall jets can also be foun
19、d in the field of fluid mechanics, such as Bakke (1957), Newman et al. (1972), Koso and Ohashi (1982), and Padmanabham and Gowda (1991). An excellent review of wall jet flows is given in Laun- der and Rodi (1981, 1983), Abrahamsson et al. (1996), and Raharatnam (1976). EXPERIMENTAL SETUP All the mea
20、surements were carried out under isothermal conditions in a full-scale test room located within the labora- tory hall of a Swedish university. The dimensions of the test room were 4.2 m x 3.6 m x 2.7 m (13.8 ft x 11.8 ft x 8.9 ft). IL .- 5 lo g -10 5 -20 U al 50 E U In .- O -30 -40 # f I I 2 Mean Ve
21、locity (m/s) Figure 2 Velocity profile in the exit of the nozzle (D =43mm1.7in.J, U, = 8.3ds1634ft/m). All four walls of the test room were insulated, and the ceiling was smooth. The supply and exhaust were located in the same wall. To study the influence of outlet size, three ASME-stan- dard long r
22、adius nozzles (D = 43 mm 1.7 in., D = 76 mm 3 in., and D = 152 mm 6 in.) were tested. The nozzles were located as close as possible to the ceiling (about 3 cm from the ceiling to the nozzle edge) so that they could generate three- dimensional wall jets along the ceiling. Air was supplied by a freque
23、ncy controlled centrifugal fan, and the flow was led through a cooling coil before it entered the settling chamber, which was 1.2 m (3.9 ft) long and 0.8 m (2.6 ft) in diameter. Five internal fine mesh screens were used to produce auniform velocity profile and reduce the turbulence level (see Figure
24、 2). Velocity measurements were made with a constant temperature hot-wire probe with a single, unplated tungsten sensor, 1.5 mm long and 5 pm in diameter. This probe was operated at an overheat ratio of 1.5 and placed on a stand with wheels. Data were acquired and converted by a computerized anemome
25、ter system. Typical sampling time was 180 seconds with sampling frequency at 160 Hz. The calibration of the anemometer probe was made with two apparatus: For air speeds less than 2.5 ds (490 ft/m), a laminar pipe-flow method was used (see Zou and Malmstrm 1998a). For air speeds larger than 2.5 ds (4
26、90 ft/m), the nozzle for the experiments was used. It is well known that the constant temperature anemom- eter is highly sensitive to the variation of surrounding air temperature. One of our studies shows that when the room air temperature changes by IT (1.8“F), the output of the anemometer can devi
27、ate by 2%. To avoid measurement errors caused by air temperature, two sets of calibration curves were applied for different temperature ranges, where one evaluated at 22T (71.6“F) and the other at 24T (75.2“F). A calibrated 360 type thermometer (accuracy = 10.3“C OSF) was also 204 ASHRAE Transaction
28、s: Research applied to check the temperature in the nozzle exit and the measurement area. Of primary concern in studies of the present type is the decay of the maximum mean velocity with x, the streamwise coordinate. The maximum mean velocity in a three-dimen- sional wall jet did not occur necessari
29、ly on the symmetry axis, since the jets had a tendency to slightly move their centerline from time to time (“buckling” and “meandering”); see Figure 3. This characteristic made the jet centerline velocity measurement always questionable, especially for measure- ment points far away from the outlet o
30、r for low-velocity air jets. It was therefore necessary to check during the measure- ments that the chosen probe position really represented the centerline. In our experiments, all measurements with veloc- ities less than 2 ms (390 ftm) were at first measured at the geometrical centerline at least f
31、our times. Then the probe was moved (with 0.5 cm 0.2 in. spacing) up and down, left and right, from the geometrical centerline point. Every checking position was measured at least twice. If the velocity of one checking position was found to be larger than that in the geometrical centerline position,
32、 then this checking position was assumed as a new “geometrical centerline” position, and the measurement procedure was repeated until the maximum velocity was found. The measurements were also repeated if results seemed to fali away from the trend associated with jet zone 3 in an evaluation graph. R
33、ESULTS AND DISCUSSION Maximum Velocity Decay in Zone 3 A matter of primary concern for HVAC applications in studies of wall jets is the distance in which supply air needs to mix with room air in order to reduce air velocities and temper- ature differences to acceptable levels before it enters the oc
34、cu- pied zone. The throw conception has been developed for this application. The throw of the jet is the distance from the diffuser to the point where the maximum velocity has decreased to a special level called the terminal velocity (0.25 ms or 0.20 ms 40 ftm or 50 ftm in most ventilation appli- ca
35、tions). According to the 1997ASHRAE Handbook-Funda- mentals (ASHRAE 1997), throw can be divided into four zones, and zone 3 is normally the longest one. Therefore, zone 3 is the one most often studied by engineers because of its ventilation potential. The maximum velocities of zone 3 for wall jets f
36、rom round nozzles can be determined with accept- able accuracy using the following formula: Equation 1 shows that for a wall jet with known outlet size and outlet velocity, the K-value is a key parameter to deter- mine the jets performance. A higher K-value means a higher velocity at a given distanc
37、e and often less surrounding air entrained into air jets. In this paper, all K-values were evalu- $ 5 1.5- I I I O I )i I 2 1.4 - I - I al I I - d 1.1 O 20 40 60 80 100 120 140 160 180 200 Seconds Figure3 One example of jet meandering: velocity in geometrical centerline, average values of five secon
38、ds (D = 43 mm t1.7 in., x/D = 45, U, = 1 O ds 1968ft/m). I Mirror image of wall jet Figure 4 Method of image. ated by measuring mean velocities in different centerline posi- tions of the air jet. Considering there is no entrainment on the ceiling side and neglecting the wall friction, by treating th
39、e wall jet with the method of image (see Figure 4), the flow in a wall jet from a “flush-mounted” nozzle with a supply area of A, is in prac- tice identical to the flow in a free jet with a supply area of 2 A, (the opening and its mirror image). Therefore, the K-value of a wall jet might be estimate
40、d by multiplying n/z with that of a free jet with the same exit area (see also ASHRAE 1997). This fact indicates that the K-values of a free jet can be used as reference cases for those of a wall jet. In Malmstrm et al. (1992), free jets from nozzles similar to those used in the present study were e
41、xamined to determine the effect of nozzle diameter and supply velocity on the maxi- mum mean velocity decay of the jet (Figure 5). Two sets of data are displayed: measurements by Nottage (1951) with a nozzle of 6 in. diameter and measurements by Malmstrm et al. with three nozzles with diameters of 1
42、.5 in., 3 in., and 6 in., ASHRAE Transactions: Research 205 7 - . . -. . . . . . . . o 0: 6.- d 6-i XXA o 00 5 -: o 5.- A* o A 4 * : I( Olt K: 4 - K: 3 0- 3.: o o M15m 2 - x Mlm 1 - 2 A m.m 1 o- “i“:“ “i “; “ Figure 5 U-value for free jets from round nozzles with different diameters (based on Malmst
43、rm et al. 1992): (a) us a jnction of outlet Reynolds number und (b) as a function of the outlet velocity. .,. o A* A AO 0 B A utx -. o 0- OM ADS x L1.S 0 - -. .- O-“i“i“i“ respectively. Obviously the K-value is not dependent on the outlet Reynolds number (see Figure Sa), which is expressed by the fo
44、rmula 6- 5- 4- 3 2- 1- Y U,. D Re = - V + + D= 43mm - mD= 76mm - + D=i52mm - - In fact, a better correlation was obtained when efflux velocity was used as the independent variable (see Figure 5b.) The data almost collapse on the same curve. For free air jets from round nozzles, the maximum velocity
45、decay coefficient K is around 6 for higher outlet velocities, and it starts to decrease at outlet velocity 6 ds. 9, 1 O 5 10 Uo (m/s) In this paper, the velocity profiles were measured at different distances x from the outlet, and maximum velocity U, was obtained from the profile measurements. A K-v
46、alue was then evaluated from a depiction of U, vs. distance x. The K-values for all the present test nozzles (D = 43 mm 1.7 in., 76 mm 3 in., and 1.52 mm 6 in.) are shown in Figure 6a as functions of outlet velocities. However, considering that the diffusers were not flush-mounted (3 cm from ceiling
47、 to nozzle edge), all the K-values need to be rectified. Farguharson (19.52) proposed to multiply a correction factor with the K- value to take into account the surface proximity for an axisym- metric wall jet with a short distance from supply to ceiling (Figure 7). The modified data of Figure 6a ar
48、e shown as Figure 9, 4 ta D= 76mm O 5 10 15 Uo (m/s) (b) Figure 6 K-values based on outlet velocities: (a) original data and (b) rectification data considering surfuce proximi0 (see Figure 7). 206 ASHRAE Transactions: Research TABLE 1 K-Values for Wall Jets from “Flush-Mounted” Nozzles with Small Di
49、ameters Investigators Davis and Winarto Wu and Raiaratnam Date Media u, (ds) D (cm) K 1980 air 1 O0 2.54 6.67 1990 water 3.35 2.54 5 Wu and Rajaratnam 1990 water 4.27 1.27 5.3 Pani Swamy and Bandvopadhvav * Approximate value calculated from the data available. 1972 air 60 6.25 1975 air 20 5.1 5.7* 0.0 4 1 I I f 01234567 Padmanabham and Gowda Abrahamsson et ai. Distance from surface - Grimitlin 1970; Skret 1973; Schneider 1985; and Hussein et al. 1994). In Zou (2000), the effect of outlet veloci
