1、EFFECTS OF SEVERAL ARRANGEMENTS OF RECTANGULAR VORTEX GENERATORS ON THE3 STATIC-PRESSURE A FUSE THROUGH A SHORT 2:l DIFFUSER By E. Floyd Valentine and Raymond B. %axroll Langley Aeronautical Laboratory Langley Field, Va. 1 Provided by IHSNot for Resale-,-,-1 I - ” NACA IiM L50L04 * 0 UNCUSSIFIEO NAT
2、IONAL ADVISORY %WITTEE FOR AERONAUTICS RESEIRCH MEMORANDUM EFFECTS OF SEVERAL ARRANGEMENTS OF RECTANGULAR VORTEX (XIOBATORS ON THE SWIC-PIEESSW FUSE THROUGH A SHORT 2: 1 DIFFUSER By E. Floyd Valentine and Raymond B. Carroll SUMMARY An investigation was made of a 2:l area-ratio diffuser of length equ
3、al to the inlet diameter with several arrangements of simple rectan- gular vortex generators over a speed range up to an inlet Mach number of 0.5. The investigation was-for an inlet boundary layer of 7 percent of the inlet diameter,. a condition for which this diffuser had substan- tial separated ar
4、eas with no vortex generators. The effects varied considerably between different vortex-generator arrangements. Some arrangements actually reduced the diffuser static-pressure rise. The effect of one.of the better vortex-generator arrangements was to increase the diffuser effectiveness by 30 percent
5、; this arrangement made it equal to that of a diffuser of twice the length with no vortex generators. INTRODUCTION The ineffectiveness of wide-angle conical diffusers has been sham by previous investigations to be associated with separation of the flow from the diffuser boundary. The separation resu
6、lts from the inability of the flow to negotiate the high static-pressure =adient required by the rate of expansion of the diffuser area. A substantial part of the cross section at and beyond the first separation is then occupied by low or negative velocity air. The main mass flow of air is then taki
7、ng place through a reduced area and, consequently, at considerably higher speed. Effectively, the area ratio has been reduced below the geometric value. The skin-friction losses over the unseparated surfaces will be greater and the over-all static-pressure rise less than if no separation had taken p
8、lace. Some further insight into the mechanism.of the diffu- sion process and its relation to the characteristics.of the boundary layer may be gained by a study of references 1 and 2. Reference 2 brings UNCLASStFIED Provided by IHSNot for Resale-,-,-2 NACA RM L5OL04 n out that the flow in a conventio
9、nal, smooth, short diffuser is.defi“ . . . -. nitely a three-dimensiond phenomena lacking axiaLsymmetry and that- * it is dependent on the initial boundary layer and Mach number as well as on the area-ratio change accomplished in a givenlength-. .- The obvious method of-avoiding boundary-layer separ
10、ation in a diffuser would, of course, be to increase its length and thereby decrease the adverse.pressure gradient. Since, in many practical cases, space limitations-do not permit this solution, my other method of preven5ing or ube readings were depended upon to give the static pressure representati
11、ve of stations 1 and 6. These points we from the readings for the tubes plotted at Oo in. figure 4 and would lead to the conclusion that. the dififuser has o operating conditions differing considerably in their resulting static- pressure rises. In contpast to this, the points in figure 6, which were
12、 obtained by using an arithmetic average of, the values for station 1 and also an average for station 6, give results which do not differ much for the two apparentoperating conditions. Averages were therefore used throughout the program for. _the diffuser outlet as already stated under the section e
13、ntitled “Appar.Btus and Methods. I Since the static orif.ices at the diffuser inlet would be in the local pressure field of.the vortei generators, a single orifice upstream from this influence wss used to obtain the inlet static pressure. The use of an upstream orifice gives a conservative result si
14、nce some of the pressure drop along the-pipe is, being subtracted from the pressure rise attributed to the diffusion process. , . I - - “ c. “ . Figure 6 indicates that considerable pressure recovery takes place in the tail pipe. A. s+il.ar pressure ret-overy in the tail pipe was also Provided by IH
15、SNot for ResaleNo reproduction or networking permitted without license from IHSNACA RM L50L04 , - 7 d h obtained in the results from the same diffuser in reference 2. In fact figure 6 represents.a duplicate condition,to the thicker boundary-layer investigation of reference 2. However, the results ar
16、e presented on a different basis and slightly different measuring-tube positions were used. When changed to the same basis and corrected for the measurement- technique differences, the results from reference 2 are then in-what is considered good agreement with the present results considering the pos
17、si- bility of differences due to elapsed time, reassembly of the duct system, and the different method of dealing with the fluctuation of the flow I conditions. All diffuser static-pressure recoveries of this investigation were from measurements of static pressure at the diffuser inlet and outlet an
18、d were made with the tail pipe in place. This condition was also true for the investigation of reference 2. However, no data indicate that the vortex-generator effects measured xith a tail pipe in place are not equally applicable to vortex-generator installations in diffusers intended to operate wit
19、h no tail pipe in place. Counterrotating vortex generators.- The static-pressure rise in the diffuser for 22 counterrotating vortex generators at three different angles of attack is shown in figure 7 in terms of the indicated inlet dynamic pressure. This rise is the pressure recovery up to the end o
20、f the diffuser and includes no gains obtained in the tail pipe. The eurve from figure 6 for the diffuser without vortex generators is included for ready comparison. As in the case of the bare diffuser, two fairly definite operating conditions were found at which the flow would stabilize for each of
21、the three angles of attack. The vortex generators improvedthe pressure recovery over this speed range for all three angles of attack with the 15 setting having the best over-all effect. The circumferential static-pressure distribution at the diffuser exit is given for two flow rates in figure 8 for
22、22 counterrotating vortex generators set at 1.50 and for 22 vortex generators set at 20. For the angle of attack giving the greatest pressure rise, 150r the exit static-pressure variation at the lower flow rate is quite small and is considerably less than for the corresponding curve of figure 4 for
23、the bare diffuser. For the less effective 20 setting, the varia- tion at the lower flow rate was greater. For both the 15O and the 20 setting, there was considerable nonuniformity at the higher flow rate. Cross plots of the pressure rise in the diffuser for angles of attack of Oo, 15O, 17. .5O, and
24、20 are given in figure 9 for two flow rates. The 0 values are from a straight-line interpolation between 0 values measured with 14 vortex generators and with 28 vortex generators. An angle of attack of around 17O appears from ,this result to be favorable for 22 counterrotating vortex generators at t
25、he lower flow rate and an angle of attack of about 1.5 for the higher flow rate. Provided by IHSNot for Resale-,-,-8 The pressure rise in.tbe diffuser with the 22 counterrotating vortex generators never reached-.the value. of over 0.6 obtained .at- the . . . end of the.tai1 pipe with no vortex gener
26、ators. (See fig. 7.) The values obtained at the end of.the tail pipe with 22 counterrotating vortex generators set at 15 are shown in figure 10. The curve for no vortex generators is repeated from figure 6. In this case, the pressure recovery at the end of the tail pipe with the vortex generators ex
27、ceeds that with no vortex generators except at the highest speed. The effect-of varying the number of these vortex generators was investigated by also running tests with 14 and with 28 counterrotating vortex generators. With 14 vortex generators the results were not .- particularly favorable. Howeve
28、r, with 28 counterrotating vortex generators set at 15O (see fig. I“( a) ) the pressure. rise through the diffuser was higher over.most of the speed range than that for the diffuser with tail pipe when no vortex generators were used. In Figure ll(b) for 28 vortex generators set at 20 shows pressure
29、recoveries smaller than those for the 150 setting. Reference to figure 12 shows that-the circumferentie3.static-pres.sure distribu- tions at the end of the diffuser for 28 counterrotating vortex generators at 150 was quite uniform as contrasted with the distri- . . . butions with the diffuser by its
30、elf (fig. 4) and with the higher speed run with the diffuser and 22 counterrotating vortex generators (fig. 8). addition, there was still some pressure rise in the tail pipe. The effect of number of vortex generators set at 15 is sham in figure I3.which gives cross plots at two. flow:rates from info
31、rmation .- obtained with 0, 14, 22, and 28 vortex generators. These data indicate that no advantage is gained by less than 14 vortex generators. Above 14, however, the diffuser pressure recoveries for both flow rates increase with the number. of vortex generators up to 28, the largest nuaber investi
32、gated. Other arrangements investigated.- A small part of this investigation was on the effects of lowitudinal location of the.vartex generators and on the effect of corotation as differentiated from counterrotation of the vortex generators. If the effects are considered to.be mainly the mixing actio
33、n frm the -tip vortices of the airfoils, the longirtudinal location of the airfoils relative to the first separation areas of the bare diffuser couid be expected to have a definite relation to the static-pressure rises.?bserye. Whether the cmotating or the counter- rotating arrangement were used wou
34、ld,also be an important factor. At. the low flow rate, moving the vortex generators upstream one- fourth of the Wet- diameter improved the pressure recovery sliglltly at- 15 and considerably at 20. (See fig. 14( a) . ) Moving the vortex . a- - Provided by IHSNot for Resale-,-,-2 . generator one-half
35、 the inlet diameter u-pstream gave a further increase over the position one-fourth the inlet diameter upstream. Figure 14(b) for the higher speeds shows a similar but evecmore favorable result. This result confirms the idea that the effect is, at least in part, a mixing effect. Points at 150 with th
36、e vortex generators 0.36d down- stream of the inlet vortex-generator station are also included in figure 14. These vortex generators were in the conical part of the diffuser and in a region in which the flow for the bare diffuser gave separated areas. The static-pressure recoveries, 1ower.than with
37、no vortex generators, support the conclusions from reference 3 that the vortex generators must be upstream of the original separation area. The theory and experiment of reference 3 indicated that more favorable results could be expected from counterrotat-ing than from corotating vortex generators. S
38、ome check on this result was made for this diffuser by running 22 and 28 corotating vortex generators at angles of attack of 15O and 20. The more favorable results of the 22 and 28 corotating vortex-generator arrangements are shown in figure 15; Curves for no vortex generators and for 28 counterrota
39、ting vortex generators are repeated to facilitate comparison The two corotating arrangements -me better than the arrangement with no vortex generators but not as favorable as the counterrotating vortex-generator arrangement. In fact, neither corotating arrangement (fig. 15) has a6 great a pressure r
40、ise in the diffuser as can be obtained with 22 counter- rotating vortex generators (fig. 7). The circumferential static-pressure distribution at the diffuser exit, figure 16, was, however, quite good for both corotating arrange- ments. The exit static-pressure distribution for 22 corotating vortex g
41、enerators set at 20 was very good in contrast to the poor distribution (fig. 8( b) ) for 22 counterrotating vortex generators set at ZOO. As a check on whether a simple obstruction would have a.favorable effect similar to the effects from the so-called vortex generators, a few additional arrangement
42、s were tested. These tests,included running the diffuser with 28 vortex generators at Oo angle of attack, 28 rods 1/8 inch by 1 inch projecting in from the surface, and 28 rods 3/8 inch by 1 inch projecting in from the surface. In every case, the pressure recovery in the diffuser and in the diffuser
43、 plus tail pipe was substan- tially less than with the unobstructed diffuser. Comparison with results from previous diffuser research.- If the values of static-pressure rise as a fraction of indicated inlet dynamic pressure given in this report are divided by 0.75, they become vaiues of diffuser eff
44、ectiveness Apactual/Apideal on the basis of assumption of incompressible-flow relations and uniform-velocity profiles. These same assumptions were used in obtaining the diffuser effectiveness in Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS10 . referen
45、ces 1 and 2 Values of diffuser effectiveness for the best vortex-generatar arrangement investigated, 28 counterrotating vortex generators set 150, have been plotted in: figure 17 for the diffuser and for the diffuser plus tail pipe. Curves for no vortex generators from figure 7 of reference 2 are in
46、cluded The diffuser effectiveness of the diffuser with this vortex-generator arrangement is about equal to that of the bare diffuser plus tail pipe. There is also some pressure rise in the tail pipe ding the effectiveness of the diffuser with vortex generators and tail pipe greater thaathat of.the b
47、are diffuser plus tail.pipe. -.The. diffuser effectiveness obtainable with the 28 counterrotating vortex generators at 15O is, for the speed range investigated, practically the same as that- of a similar diffuser of twice the length (reference 1). L. . -. “ . . . - “ .“ .- “ CONCLUSIONS “ The follow
48、ing conclusions relate to the static-pressure rise measured in a 2:l short conical diffuser oelength roughly equal.to the inlet diameter and do not consider the- total-pressure losses. All comparisons are from measurements made with a tail pipe in place. The inletboundary layer had a thickness of 5
49、percent-of theinlet diameter. The conclusions are for a single vortex-generator configuration which was rectangular and noncambered and had a spanwise length of one-half the chord. .“ . . . . 1. The effectof one of the better vortex-generator arrangements was to make the diffuser effkc-tlveness ApIAp, equal to that of a aiffuser of twice the length with no vortex generators. This result was obtained by using 28 counterrotating vortex generators setat
