1、Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-c EFFECTS OF SPIRAL LONGITUDINAL VORTICES ON TURBUUNT BOUNDARY LAYER SKIN FRICTION By Jack G. Spangler and C. Sinclair Wells, Jr. SUMMARY As a part of an experimental investigation of the effects of ord
2、ered mixing on turbulent skin friction, spiral longitudinal vortices were generated in a turbulent boundary layer. The boundary layer was formed on the test sec- tion wall of a facility designed especially for low-speed boundary layer studies. Screening tests which resulted in the choice and optimiz
3、ation of standard rec- tangular planform wall-mounted elements for producing this type of vortex are described. Skin friction and velocity profiles were measured as a function of longitudinal and transverse distance for several element configurations, and for the same flow condition. Element heights
4、 of 2, 10, and x) percent of the boundary layer thickness were used. Effects on skin friction of the vortex- producing elements were measured up to 87 element heights downstream. Direct measurement of the element form drag was also made. I“I!RODUCTION The characteristics of the turbulent boundary la
5、yer have long posed a problem to those concerned with the efficiency of viscous flow. high values of skin friction drag and heat transfer rate bulent boundary layer, compared to those for laminar flow, are well known. There has been considerable study into the feasibility of extending the range of l
6、aminar flow; that is, delaying the onset of turbulence. A certain d-egree of success has resulted from these studies, but in most practical cases the eventual transition to turbulent flow is inevitable, It would seem, therefore, that, in addition to developing techniques to retard the onset of turbu
7、lence, it is worthwhile to study the possibility of modifying the turbulent boundary layer in such a way as to reduce the undesirable effects. The relatively associated with a tur- I The interest in using vortices to modify the turbulent boundary layer stems from somewhat fragmentary evidence (refer
8、ences I, 2, 3) of their effects on a boundary layer. reduction in skin friction. In at least one case (reference 3) a marked decrease in I heat transfer resulted from the presence of the vortices implying a similar The choice of spiral longitudinal vortices in the subject investigation was due to th
9、eir persistence, the desirability of extending any beneficial effects being obvious. In addition, a great deal of information on the pro- duction of this type of vortex in a shear layer is available. Finally, the Provided by IHSNot for ResaleNo reproduction or networking permitted without license fr
10、om IHS-,-,-results; i.e., the effects of this common type of vortex-boundary layer inter- 4 action, would be of interest regardless of whether the effects are favorable. This report presents the results of a set of experiments with spiral longitudinal vortices in a turbulent boundary layer. The effo
11、rts to opti- mize the experimental configuration are also described. The testing en- vironment was created in the boundary layer facility of the LTV Research Center. Skin friction was measured directly,as well as mean velocity pro- files,to determine the effects of the vortices. ments were negative,
12、 in that no significant reduction in skin friction was measured; however, the experimental data, which are presented in sufficient detail to be useful, and the conclusions regarding the effects of the vortices should be of general interest. The results of these experi- SYMBOLS AR cf C P D K P 9 r 0
13、U X X Y CY 6 2 aspect ratio local skin friction coefficient, T0/q pressure coefficient, p - p/qlo (where subscript 10 refers to 0 X = 163.3 inches) element drag element height static pressure reservoir pressure 2 dynamic pressure, 3 p U, boundary layer channel radius momentum thickness Reynolds numb
14、er, U,O/v local velocity free stream velocity longitudinal distance measured from origin of laminar boundary layer longitudinal distance downstream of element location distance normal to wall angle of attack boundary layer thickness Provided by IHSNot for ResaleNo reproduction or networking permitte
15、d without license from IHS-,-,-bo boundary layer momentum thickness for axisymmetric flow, defined by: d) azimiithnl pnsf tinn a, I rotational speed in a vortex I V kinematic viscosity I P density 7 wall shear stress 0 I friction velocity, Note: subscript sw refers to smooth wall conditions I INITIA
16、L STUDIES A. LITERA!IWE SEARCH The spiral longitudinal vortex in a turbulent boundary layer has been dealt with by other experimenters on several occasions. work has been devoted to the use of vortices as a mixing mechanism. tent of the experiments has been to delay separation of boundary layer flow
17、 from its bounding surface in the presence of an adverse pressure gradient,such as exists in diffusers or on the aft portion of airfoil surfaces. The standard technique is the production of arrays of vortices near the edge of the boundary layer that mix the high energy fluid from the inviscid stream
18、 with the shear layer near the surface. This can be done with numerous types of vortex genera- tors. The resut is a net increase in energy near the surface which allows the flow to advance farther into an adverse pressure gradient before separation than is possible when the vortices are absent. Exte
19、nsive experiments of this nature have been conducted at United Aircraft Corporation and the results can be found in references 4, 5, 6, and 7. done by the NACA and the National Bureau of Standards as reported in references 8, 9, and 10. The majority of this The in- The same type of work has also bee
20、n In most of these previous experiments the main objective has been to determine the controlling parameters and to optimize the mixing effects. of the important parameters have been found to be generator shape (i.e., lifting flat plates of various planforms, wedges, ramps, hemispheres, etc. ) aspect
21、 ratio, generator height in the boundary layer, generator spacing, angle of attack to the flow (for lifting plates), and rotational sense of the vortices. In addition to studies of these parameters Schubauer and Spangenberg (ref- erence 10) have investigated the effects on forced boundary layer mixi
22、ng of Some 3 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-elements that were designed to mix the flow by deflection and displacement rather than through vortex action. designed to have weak or non-existent trailing vortices were tested. general th
23、ese elements were of a wedge or plow-type shape. thought that intentional suppression of trailing vortices would reduce the momentum loss associated with the induced drag of the vortices. However, it was found that the body shapes necessary for vortex suppression had such an excessive surface area t
24、hat the skin friction drag on the element itself nullified any gain realized by eliminating the induced drag of the vortices. A number of different mixing elements In It was at first B. CHOICE OF PARAMETERS In order to conduct a meaningful and systematic study of the vortex- boundary layer interacti
25、on phenomenon it was concluded that only a few of the many parameters should be varied at first and the effects of these varia- tions analyzed. Then, based on these findings, the need for and the type of further experiments could be evaluated. The following paragraphs explain the process by which ce
26、rtain parmeters were fixed. 1. Element TvDe Considering all of the results found in references 1 through 10 it can be concluded that for forced boundary layer mixing the simple lifting flatplate element (plate in a plane normal to the surface) is an efficient device and is by far the simplest type t
27、o fabricate and use. Based on this conclusion the experiments described in this report were restricted to flat plate elements. 2. Planform It would be desirable to have all of the vorticity created by an element concentrated into a single vortex which is released at the tip of the element. To accomp
28、lish this the circulation over the element span must be uniform and,for a non-uniform velocity field such as presented by a boundary layer, this can be achieved only by an appropriate choice of planform. In particular, the proper criterion is thought to be that the product of element chord and veloc
29、ity, as a function of distance normal to the wall, should be held constant. Comparative tests (references 4 and 5) of elements tapered to match the velocity profile, rectangular elements, and trapezoidal elements (an approximation of the tapered case) have shown that the tip vortex strength, as impl
30、ied by mixing efficiency, is indeed increased when the planform is tapered but the increase is small and from practical considerations the rec- tangular elements are the more desirable. Accordingly a rectangular planform was chosen for these experiments. 3. Element Height The most important paramete
31、r would seem to be the height of the vortex-generating element as this determines the location of the vortex with respect to the bounding surface of the flow. lies in the fact that a shear flow will tend to dissipate any concentrated vortex motion. Thus, due to the velocity gradient in the boundary
32、layer, the The importance of this parameter 4 Provided by IHSNot for ResaleNo reproduction or networking permitted without license from IHS-,-,-* distance of the vortex from the surface will, to some extent, affect its persistence in the flow. Also, little information is available in the litera- tur
33、e concerning variations in this parameter, and it was obvious from the purpose of the reported experiments on the delay of separation that the opti- mum location for the vortices was at the edge of the boundary layer. For these reasons the element height was set as the primary variable. 4. Sense of
34、Rotation Elements may be inclined to the flow in such a way as to produce Both situations vortices that have the same rotational sense (eo-rotating) or that will al- ternately have opposite rotational sense ( counter-rotating) . have been examined experimentally and analytically and compared for eff
35、ective- ness (references 4, 5, and 7). shown counter-rotating vortex arrays to be suprior mixers to co-rotating arrays. Potential flow analysis also shows that vortices in a counter-rotating array will tend to be attracted to or repelled from adjacent vortices in such a way as to group themselves in
36、 pairs due to interference effects caused by the induced velocity fields. However, experimental results show that this effect is actually very weak or non-existent. This is probably explainable by the fact that the true vortex motion damps out quickly downstream of the elements. The effects of the v
37、ortex, such as local three-dimensional distor- tion of the boundary layer, may persist quite a distance downstream, but the regions of distortion will not be moved laterally because of the quick dissipa- tion of the induced velocity fields. Based on the above information the ex- periments reported h
38、erein have been restricted to counter-rotating vortex arrays. Experiments and potential flow analyses have 5. Element Spacing The choice of a spacing arrangement for the elements was somewhat arbitrary. The data in reference 7 show that mixing effectiveness is increased as the element spacing is dec
39、reased for equal spacing arrangements. Likewise the momentum deficit in the flow resulting from the form drag of the elements is increased due to the larger number of elements per transverse unit length. The velocity profiles show this effect in their change relative to the smooth wall profiles. The
40、 transversely averaged velocity near the wall is increased indicating stronger mixing action and implying higher skin friction for the closer spacings. as great as four to one (ratio of distance between elements lifting toward each other to distance between elements lifting away from each other) the
41、 velocity near the wall is significantly decreased in the decelerated regions and only slightly increased in the accelerated regions. the condition of lowest transversely averagdvelocity and hence lowest skin friction for all of the experiments described in reference 7. the experiments described her
42、ein, the element spacing parameter was fixed such that the distance between elements in a pair lifting away from each other was two element heights and the distance between elements of adjacent pairs was eight element heights, giving the four to one spacing ratio. Limited data for unequal spacings s
43、how that at spacing ratios This situation implies Therefore, for 6. AsDect Ratio and hle of Attack The remaining parameters are element aspect ratio and angle of attack. he to the strong influence of these parameters for an element of a 5 Provided by IHSNot for ResaleNo reproduction or networking pe
44、rmitted without license from IHS-,-,-J given height it was decided to conduct a series of screening tests to determine their optimum values. are described in the following section. L These tests along with the final detailed experiments EXPERIMENTAL PROGRAM A. EXPERIMENTAL FACILITY All of the experi
45、ments in this study were performed in the LTV Boundary Layer Channel. acteristics may be found in reference ll. By way of brief explanation the boundary layer channel is a straight circular tube having a test section 25 feet in length and eight inches internal diameter made up of six clear plexiglas
46、s sections. Operation is continuous and flow conditions are extremely steady. A vacuum blower, isolated from the test section by a sonic throat, pulls air at atmos- pheric pressure through a series of damping screens, a contraction section, and the test section. The flow may be investigated in detai
47、l at any location in the 25 foot test section. A sketch of the facility is shown in Figure 1. A complete description of this facility and its operating char- The test environment for these experiments was a fully turbulent boundary layer created by artificial tripping of the laminar boundary layer n
48、ear the entrance of the test section of the boundary layer channel. A con- stant mass flow rate was maintained during all testing. Figure 2 presents the rate of growth of the boundary layer, the free-stream velocity, and wall shear stress with distance down the channel for this mass flow rate. The a
49、bscissa in this figure is the distance in the downstream direction measured from the fictitious origin of the initially laminar boundary layer (see reference 11). The location of all elements tested in these experiments is also shown in this figure. Figure 3 is a photograph of a portion of the boundary layer channel showing the set-up for these experiments. It has been pointed out in reference 11 that the boundary layer ch
