NASA NASA_I_p_238819850004006

_'_Techn_,N.Paper2388

November 1984

Vort ex GeneratingCoolant,Flow,PassageDesign for IncreasedFilm Cooling Effectivenessand Surface Coverage

S. Stephen Papell

AUg2 5 1988LANGLEY RESEARCHCENTER

LIBRARY NASA_.. HAMP[ON. VIRGINIA

N/A

https://ntrs.nasa.gov/search.jsp?R=19850004006 2020-03-31T19:27:42+00:00Z

3 1176 01326 5898NASATechnicalPaper2388

1984

Vortex-GeneratingCoolant-Flow-PassageDesign for IncreasedFilm-Cooling Effectivenessand Surface Coverage

S. Stephen PapellLewis Research CenterCleveland, Ohio

NI NANational Aeronauticsand Space Administration

Scientific and TechnicalInformation Branch

Summary In actual use the coolant enters the base of the bladingand is ejected through holes drilled at some angle through

The study used an inclined jet in crossflow as a model the blading surfaces. Although the coolant passages areto aid in the development of the hypothesis that a generally straight with a circular cross section, some at-coolant-channel contour could increase film-cooling effi- tempts have been made to increase the effectiveness ofciency. The coolant channel, by virtue of its shape, the film-cooling process by design modifications. Enginegenerated a counterrotating vortex pair secondary flow manufacturing companies are working to develop inletwithin the coolant flow. The interaction of the vortex and exit passage contouring techniques with encouragingstructure generated by the flow passage and the vortex results. References 1 and 2, in which the influence ofstructure generated by the mainstream and boundary coolant-passage curvature on film cooling was examined,flow was examined in terms of film-cooling effectiveness report that curvature increases the effectiveness of film-and surface coverage, cooling by about 35 percent at the low blowing rates but

Reported herein are the thermal film-cooling foot- that at higher blowing rates curvature decreases theprints observed by infrared imagery for three coolant- effectiveness.passage configurations embedded in adiabatic test plates. The present study uses an inclined jet in crossflow as aA round-hole cross section was used as a reference with model to hypothesize a coolant-channel contour that in-which to compare the performance of two orientations of creases the effectiveness of film-cooling. The design con-the new, vortex-generating flow passage. For both orien- cept requires that the coolant channel generate a counter-tations the vortex-generating coolant passages showed up rotating vortex pair secondary flow within the coolant jetto factors of four increases in both film-cooling effec- by virtue of the shape of the flow channel. The interac-tiveness and surface coverage over that obtained with the tion of the vortex structuregenerated by the flow passageround coolant passage. The crossflow data covered a and that generated by the crossflow was examined inrange of tunnel velocities from 15.5 to 45 m/sec with terms of film-cooling effectiveness and surface areablowing rates from 0.20 to 2.05. coverage. Comparative data obtained with a round

In addition, it was determined that flow separation of coolant passage were used to evaluate the worth of thisan inclined jet from the surface containing the hole con- new design concept.figuration could be inhibited by a coolant-passage Reported herein are the thermal film-cooling foot-geometry that generates a counterrotating vortex pair prints obtained by infrared imagery for the vortex-structure in the coolant flow that rotates in opposition to generating and the round coolant passages embedded atthe vortex pair rotation generated by the crossflow, an angle of 30* in adiabatic test plates. The crossflow

A photographic streakline flow visualization technique covered a range of velocities from 15.5 to 45.0 m/sec withwas used that supported the concept of the counter- blowing rates from 0.20 to 2.05. Temperature differencesrotating capability of the flow passage design and gave between the coolant and crossflow was about 22* C.visual credence to the explanation offered for the dif-ferences found in the data.

Symbols

Introduction c centrifugalforce

Film cooling is well established as a viable method for D dragprotecting turbine blades and vanes in high-performance d coolant-tube diameter

engines for advanced aircraft. Since the coolant itself is M blowing rate, (p V)c/6oV)**drawn from compressor pressurized air that is part of theaerodynamic cycle of the engine, it becomes necessary to m momentum changeuse it as efficiently as possible. R Tadius

T temperature and up through the center. This direction of rotation, in-

t adiabatic plate thickness duced by the crossflow is quite significant and will be fur-u crossflow velocity ther discussed later.

From this simple model it is evident that the discrete-V velocity hole film-cooling process is controlled by the energyx distance from downstream edge of coolant port to available in the crossflow. It is suggested that this energy

downstream edge of isotherm trace is distributed in a manner to influence the jet trajectory,x/d dimensionless distance the three turbulent mixing processes described above, and

the spreading of the jet on the surface to be protected.adiabatic film-cooling effectiveness, The design concept was to modify the coolant-jet

(T_-Taw)/(T_-Tc) stream by inducing a vortex structure within the coolant

P density flow, similar to that shown in figure 2, by virtue of the

Subscripts: shape of the coolant passage. It was postulated that thisgeometry-induced vorticity could replace the crossflow

aw adiabatic wall energy required to generate the vortex flow structurec coolant within the jet and, as a result, make available more

j jet crossflow energy to turn the jet closer to the wall and/oroo free stream or tunnel air increase the vortex strength to spread the jet. Both of

these flow mechanisms should increase the effectivenessof the film-cooling process.

It was also suggested that the direction of rotation ofFluid Dynamic Modeling the counterrotating vortex structure generated by the

Inclined Jet in Crossflow passage geometry could be reversed with respect to thecrossfiow generated vortex structure. In this manner both

The film-cooling mechanism is the injection of a cool vortex structures would tend to cancel each other at thefilm of air into a boundary layer to provide an insulating coolant-injection port. If this were possible, the mixinglayer between a hot gas and metal surfaces. For the pro- that occursdue to the tunnel air being pulled sidewaysandtection of turbine blading the coolant is ejected through under the jet stream could be preventedor minimized.discrete holes drilled at some angle in the blade wall sothat the basic fluid dynamic process is essentially an in-

Coolant Passage Design Conceptclined jet in crossflow. The literature contains manystudies relating to the mixing process associated with a jet The vortex-generating coolant flow passage designin crossflow (e.g., refs. 3 and 4). concept is illustrated in figure 3 along with a round

Figure 1 is a schematic of an inclined jet in crossflow coolant flow passage for comparison. The cross sectionshowing the forces on a control element of the jet. The of the vortex-generating passage in figure 3(a) isdrag D acts opposite to the direction of the relative somewhat elliptical, with a flat surface on one of the longvelocity caused by the interaction of the crossflow veloci- ends of the ellipse and a cusp surface on the opposite end.ty Vooand the jet velocity Vc. The centrifugal force C is Figure 3(b) shows the edge of cusp extending along thenormal to the jet path, and the momentum change m acts length of the flow passage that provides the vortex-along the jet path. generating capability of the coolant passage.

There arethree well-defined turbulent mixing processes The double curved surfaces of the cusp should split theassociated with this momentum change that contribute to flow into two streams that rotate in opposite directions,the breakup of the jet. The first is the mixing that occurs producing a vortex pair structure within the coolantalong the interface between the jet and the crossflow. The stream. Within the passage the direction of this vortexsecond is just downstream of the jet exit port where a pair rotation is from the center of the flat surface (fig.pressure rarefaction pulls the tunnel air sideways under 3(a)) toward the cusp edge and out along the curved sur-and into the jet stream. The third mixing process occurs faces in opposite directions. The direction of thiswithin the jet stream itself with the formation of a vortex geometry-induced rotation with respect to the crossflow,structure (schematically illustrated in fig. 2) as a counter- therefore, depends on the orientation of the cusp edge inrotating vortex pair. The vortex strength and the distance the flow passage. Figure 4 illustrates the two cusp orien-between vortex centers, which influence the width of the tations used in this study. With the cusp surface on thejet, are also functions of the crossflow momentum. The top of the jet stream (fig. 4(a)), the direction of rotationcross section of the jet (fig. 2) shows that the direction of is upward toward the cusp edge and down on the sides.rotation for this vortex pair structure is down on the sides With the cusp surface at the bottom of the jet stream (fig.

4(b)), the rotation is downwardtoward the cusp edgeand coolant-injection port. Under the flow conditions of thisupward along the sides, study the tunnel flowboundary layer thicknesswas about

It, therefore, follows from figure 4 that the vortex 1.65 cm, and the turbulence intensity of the free streamstructure generated by the top-cusp geometry enhances was about 3 percent.the vortex structure generated by the crossflow by The coolant was supplied to the plenum by use of arotating in the same direction,whilethat generatedby the Hilsch tube connected to a 8.3-kPa (120-psi) dry airbottom-cusp geometry tends to cancel the crossflow source. This source, which incorporates a vortex-generated vortex structure. The interactions of these sets generator element, separates the inlet air into hot andof vortex pairs and their influence on film-cooling,effi- coldstreams. This results from the forcedvortex or wheelciency is the subject of this study, type of angular velocityimparted to the air entering the

device.Conservation of the total energyof the inner por-tion of the contained vortex causesheat to be transferredto the outer region of the vortex. Consequently, a

Apparatus relativelycold inner core of air and a warm outer ring ofair are available. In the Hilsch tube design the warm and

The actual coolant passage designused is illustrated in cold air dischargeports are on opposite ends of the tube.figure 5, which also shows the dimensionsof the coolant Cold side temperatures of 0* C are available with thispassages. Both the round and vortex-generatingpassages deviceso that temperature differencesbetweenthe tunnelwere cast in epoxy as inserts to be installed in the flat and coolant airflow of about 22* C could be readily ob-plate shown in figure 6. The round passage had a flow tained.diameter of 1.27 cm; the contoured vortex-generating The tunnel air temperature was measured with a ther-passage was sized to have the same flow area. Thus, for mocouple mounted in the contoured inlet. The coolant-the same mass flow, the dischargevelocityof the coolant air temperature was measured with a thermocouplewasthe same. Both passagesdischargedthe coolant at an mounted in the plenum between the screens and theangle of 30* to the surface in line with the tunnel flow. mahogany test plate. The coolant airflow rate was

The test plate was made of mahoganywith a thickness measuredwith a turbine type flowmeterinstalledbetweenof 6.35 cm that minimizedheat transfer through its wall. the Hilsch tube and the coolant plenum.In addition, the thick plate also provided a ratio of The infrared camera and detector unit used to measurethickness to coolant-passage effective diameter t/d of 5, the surface temperature downstream of the coolant-whichis typical of holes drilled in the wallsof cooled tur- injection port was capable of operating within thebine blading. The coolant-passage inserts were hand fit- temperature range of -30* to 200° C and couldted to an opening in the test plate designed to accept discriminate temperature differences as small as 0.2* C.them. The surfaces of the plate and the passage inserts The detector elementwasindium antinomide cooled by awere finished smooth and painted a uniform dullblack to liquid-nitrogenbath. The detector displayedthe isothermincreasetheir radiation emissivity.The top of the plenum imageon the cathode-ray screen, from whichit was then(fig. 6) was attached to the bottom of the test plate. The photographed for a permanent record. A referencecoolant air supplied to the plenum entered the coolant isotherm at a known temperature was obtained for eachpassage at very low velocity. The plenum assemblycon- data run by focusing the camera on the test plate at atained a system of baffles and screens and was covered position not influenced by the cooling air jet. Thewith fiberglass insulation, temperature at this position was measured with ther-

The test plate and plenum assembly were installed as mocouplesembedded in the plate.part of the tunnel floor shown in figure 7. The tunnel Examples of photographic data are shown in figure 8,itself, designed for flow visualization, was made of clear which presents equal temperature isotherm traces of theplastic sectionswith a flow area cross sectionof 15by 38 round and cusped coolant channels for data obtained atcm. A contoured inlet and a transition piececonnectedto identical test conditions. The temperatures of thethe altitude exhaust system of the laboratory completed isotherm tracesare indicated in thermal scaleunits by thethe tunnel assembly, tick marks along the abscissa on the photographs. The

The tunnel section containing the test plate was posi- comparison of these scale readings previously describedtioned so that there was 1.3 m of tunnel length (not in- permit the evaluation of the isotherm temperatures. Theeluding the contoured inlet) in front of the coolant- two verticallineson the photographs are the field-of-viewinjection port. In a previousinvestigation(ref. 2)velocity limits of the infrared detector. These linesare centered soprofiles were obtained in this tunnel by both hot-wire and that the left line abuts the downstream edge of thePitot tube probes. These measurements assured the coolant-injection port, which is not shown in theestablishment of fully developed turbulent flow at the photograph.

Experimental Procedure and Data The results of this investigation, presented graphically,Reduction covered data obtained at specific test conditions of tunnel

velocity 15.5, 22.8, 30.5, and 45.0 m/s and blowing rates

The experimental data obtained for this study enabled from 0.20 to 2.05 for each of the three coolant-passagethe evaluation of film-cooling effectiveness and surface geometries.area coverage for each isotherm trace. The controlledparameters for these tests were tunnel air velocity(crossflow), coolant velocity, and coolant temperature.The coolant temperature, measured in the plenum, was Presentation of Data and Flowset near 0° C by controlling the Hilsch tube valves. The Phenomenatunnel air temperature, depending on ambient condi-tions, was about 25° C. Variations in the blowing rate Centerline Film Cooling Effectiveness (Basic Data)were obtained by control of tunnel and Hilsch tube

Distributions of film-cooling effectiveness downstreamairflow valve settings, of the coolant-injection port are presented in figures 9 toThe testing procedure required an initial cool down of

11 for the round, top-cusp, and bottom-cusp coolantthe plenum before a desired blowing rate was set. Thepassages. The data are plotted as film-cooling effec-

temperatures were then allowed to come to equilibrium tiveness as a function of dimensionless distance (x/d)before data collection commenced. Eight infraredphotographs were taken for each test setting that covered along the centerline from the edge of the coolant-

injection port. The round coolant passage data (fig. 9)the temperature range of the film-cooling footprint. The show an expected trend that is influenced by tunneltesting procedure was repeated over a range of tunnelvelocity settings and blowing rates for all coolant velocity and blowing rate. An increase in tunnel velocitypassages studied, generally increases the level of each curve for a constant

blowing rate. For a constant tunnel velocity an increase inThe film-cooling effectiveness of each isotherm traceblowing rate first increases the level of the curves up to

was calculated as about M= 0.50, and then decreases the level of the curvesas the blowing rate increases.

Too-Taw The distance downstream of the injection port for_1- Too-Tc (1) maximum effectiveness depends on the blowing rate.

Figure 9(a) shows that at M= 0.20 the maximization ofeach curve occurs at the coolant-injection port, sug-

where Too, Tc, and Taware temperatures of the tunnel air, gesting an immediate attachment of the jet stream to thecoolant, and plate, respectively. Information gained plate. With increasing blowing rate the maximization offrom these traces also included interpretations of each curve shifts downstream from about 2 holeisotherm shapes with regard to the mixing processes of diameters at M= 0.34 to 6 hole diameters at M= 2.05.both air streams. For example, the photographs in figure This position shift in maximum effectiveness is caused8 illustrate the differences in isotherm traces obtained by flow separation of the jet due to the angularity of theunder the same test conditions for the round and top- coolant passage with respect to the crossflow. It is sug-cusp coolant-passage geometries. The differences in size gested that maximization occurs where the energy in theand shape of the traces are readily observed. The round crossflow forces the jet stream to attach to the plate sur-coolant passage trace (fig. 8(a)) closes on itself at the end face.nearer to the coolant-injection port. Since the plate sur- The individual curves plotted in figure 9 were obtainedface temperature outside of the enclosed trace is warmer from lines drawn through data points as shown in figurethan the trace itself, the warmer plate temperature near 9(c), which is representative of the extent of the scatter inthe coolant-injection port suggests that the jet did not at- the data. At the higher blowing rates (figs. 9(f) to (h)), rigtach to the surface until some distance downstream. The limitations in the availability of the coolant flow ratespressure rarefaction that occurs just downstream of the through the Hilsch tube did not permit covering the fullinjection port due to the passage angularity with respect range of tunnel velocity data.to the tunnel flow is responsible for this flow separation The top-cusp film-cooling effectiveness data obtainedphenomena, under the same test conditions as the round coolant

Two types of measurements were obtained from the in- passage data are shown in figure 10. The tunnel velocityfrared photographs. The first was the centerline length of and blowing rate trends are similar to those obtained witheach isotherm trace measured from the downstream edge the round coolant passage (fig. 9). In addition, the pro-of the coolant-injection port to the downstream edge of files of the curves are also similar, including the shift inthe trace. The second measurement, performed with a maximum effectiveness due to the delayed jet stream at-planometer, was the surface area enclosed by the trace, tachment phenomena described above.

The bottom-cusp data obtained under the same test bottom-cusp data fall close to the round coolant passageconditions are shown in figure 11. The tunnel velocity data but, as the blowing rate increased, shift closer to theand blowing rate trends are similar to those obtained with top-cusp data.the other coolant-passage geometries, but the position of For dimensionless distances (x/d) less than 4, which ismaximum effectiveness and the shapes of the data curves close to the coolant-injection port, a reversal occurs inare quite different, the relationship between the cusp orientations and the

An examination of the eight sets of data in figure 11 vortex-generating passage, with the bottom-cusp passageshows that for all blowing rates the position of maximum becoming more effective. The effectiveness differenceseffectiveness occurs at the coolant-injection port. Ap- increases considerably at the higher blowing rates.parently, the vortex-generated flow due to the bottom- The general results drawn from the data of figures 12cusp geometry did not permit jet separation under any to 15 is that the top-cusp coolant-passage is superior inblowing rate. performance to both the bottom-cusp and the round

The shapes of the effectiveness profiles shown in figure coolant passage configurations except for close to the11 are functions of the blowing rates. At the lowest blow- coolant-injection port, where, at M>_ 1.00, the bottom-ing rate (M= 0.20) the shapes of the curves are similar to cusp passage becomes superior. Differences in the datathose of figures 9 and 10. At higher blowing rates up to a factor of 4 in film-cooling effectiveness can be(M=0.34 and 0.50; figs. 9(b) and (c)) the horizontal readily obtained by choice of coolant-passage geometry.dashed lines indicate a flattening of the profiles within 4diameter downstream of the port, depending on the tun- Film Cooling Coverage (Basic Data)

nel air velocity. At the higher blowing rates (M= 1.00 to The surface areas enclosed by the isotherm traces were2.05; figs. 9(e) to (h)) the shapes of the curves presented obtained from the infrared photographs and plotted inchange significantly. An inflection appears in the data terms of film-cooling effectiveness based on theprofiles at a position downstream of the coolant-injection measured adiabatic temperatures of the isotherms. Theport between x/d values of 4 to 5. data curvesare presented in figures 16 to 18 for the three

An explanation for this unique inflection that appears coolant passage geometries. The individual curves werein the effectiveness profiles for the bottom-cusp data lies obtained from lines drawn through data points as shownin the direction of vortex pair rotation induced by the in figure 16(b), which is representative of the scatter incoolant-passage geometry, which in this orientation is op- the data.posite to the rotation induced by the crossflow. Within 4 An examination of the three sets of data shows similarto 5 diameters of the coolant exit, the geometry-induced trends, which are also influenced by tunnel velocities andvortex flow structure predominates but is finally over- blowing rates. Increasing tunnel velocity generally in-come by the crossflow induced vortex flow further creases the level of each curve for a constant blowingdownstream, rate, while increasing blowing rates first increases then

decreases the level of the curves. In general, the curves ofData Comparisons figures 16 to 18"show a maximization of film-cooling ef-

Differences in centerline film-cooling effectiveness per- fectiveness occurring at the lowest values of isothermformance of the round, top- and bottom-cusp coolant area coverage with a gradual reduction in effectivenesspassages are illustrated in figures 12 to 15. Each figure with increasing area coverage.

presents data for a constant value of tunnel velocity over Data Comparisona range of blowing rates. The individual curves, plottedas film-cooling effectiveness against hole diameters Differences in film-cooling effectiveness as a functiondownstream from the edge of the coolant port, were oh- of isotherm area coverage for the three coolant-passagetained from the basic data presented in figures 9 to 11. geometries are illustrated in figures 19 to 22. The curvesThe trends that appear in the comparative data plots of were obtained from the basic data in figures 16 to 18.the figures are quite similar for all four sets of data. Each figure presents data for a constant value of tunnel

In all cases the film-cooling performance of the round velocity over a range of blowing rates.cross section passage was inferior to that of the new, The trends illustrated in figure 19 for a tunnel velocityvortex-generating passage in either cusp orientation. For of 15.5 m/s show the superior performance of the top-dimensionless distances (x/d) greater than 4, the data cusp passage over the round passage for the entire rangetrends are quite regular. The top-cusp coolant-passage of blowing rates from (M=0.20 to 2.05); the bottom-data show the highest film-cooling effectiveness, and the cusp curves lie between these two sets of curves in a posi-round coolant passage data shows the lowest. The tion that depends on the blowing rate. At the lowestbottom-cusp data fall in between, with its position blowing rate (M=0.20; fig. 19(a))the bottom-cusp datarelative to the top-cusl_and round coolant passage data a fall on or close to the round coolant passage data. Withfunction of blowing rate. At low blowing rates the increasing blowing rate, the bottom-cusp curve shifts in

the direction of the top-cusp curve, and at M= 2.05 the When the cusp surface is oriented on the bottom of thecurves are quite close to each other. These same data jet stream, both vortex structures rotate in opposite direc-trends are repeated in figures 20 to 22 for tunnel velocities tions and influence the jet separation phenomenaof Voo=22.5, 30.5, and 45.0 m/s. previously described. Both the round and top-cusp

In general it appears that, for a constant value of film- coolant passages allow flow separation to occur due to acooling effectiveness, the coolant flow from the top-cusp pressure rarefaction that occurs just downstream of thepassage covers a larger surface area than is available from injection port. The flow separation results in a side flowthe round coolant passage. In addition, the area covered pumping of the tunnel air under and into the jet stream.with coolant flow from the bottom-cusp passage falls be- The direction of vortex pair rotation with the passage intween the other two passages in a position that depends the bottom-cusp orientation must therefore be responsi-on the blowing rates. Differences in surface area coverage ble for preventing or minimizing this pumping from oc-under the same flow conditions of up to a factor of 4 can curring.be readily obtained by choice of coolant-passage Although significant isotherm data differencegeometry, measurements obtained with the round, top- and bottom-

cusp coolant passages could be attributed to geometricdifferences, the counterrotating pair vortex-generating

Discussion capability of the cusp shaped passage could only bespeculated. Therefore, a flow visualization study was

The improved film-cooling performance of the vortex- conducted to attempt to visually establish the expectedgenerating coolant passage over the round-cross-section vorticity in the flow. A streakline flow visualizationpassage has been demonstrated herein for the range of technique (ref. 5) was used to photograph very small,blowing rates relevant to turbine blade cooling (M= 0.20 neutrally buoyant, helium-filled soap bubbles, about 1to 2.05). The evidence is graphically presented in plots mm in diameter, that followed the flow field. With aobtained under identical test conditions. It was deter- camera shutter speed of 12frames per second, the bubblemined that the coolant passage, by virtue of its geometry, appears as a streak line on the film. The bubble generatorgenerates a counterrotating vortex pair structure within outlet probe was set up to discharge into the coolantthe coolant flow that is similar to the vortex structure plenum thereby seeding the coolant flow. The indicationgenerated by the crossflow. The cusp surface coolant- of vorticity in the flow is then captured on film and canpassage design evolved along with the following be identified by the twisting motions of the streaklines.arguments. Because the cusped surface in the coolant The photographs in figure 23 were obtained using apassage controls the vortex pair direction of rotation, it telephoto lens on a camera facing downstream from thewas believed that a change in rotation direction could be tunnel inlet toward the front edge of the test plate. Theachieved by orienting the cusp surface on the top or at the horizontal white lines in these photographs are the sur-bottom of the jet stream. It was also believed that this face of the test plate illuminated by the lamp that definesdifference in induced vortex rotation would influence the the exit of the coolant passage. In the absence of tunneldata. The vortex structure generated by the top-cusp flow, a streakline rises from the the top-cusp passagegeometry enhanced the vortex structure generated by the toward the top of the tunnel (fig. 23(a)). The cameracross flow because they rotate in the same directions. The caught the bubble streak after it had emerged from thevortex structure generated by the bottom-cusp geometry passage. Since there is no crossflow, vorticity must haverotates in opposite the directions and tends to cancel the been generated by the geometry of the coolant passage.vortex structure of the crossflow. The fluid dynamic in- Figures 23(b) and (c) were obtained under conditionsteractions of these sets of vortex pairs were used to ex- of crossflow with both the top-and bottom-cusp passagesplain the differences in the data. subjected to a blowing rate of 0.34 but at different tunnel

There are separate arguments that apply to the dif- flow rates. The crossflow momentum appears to haveferent orientations of the cusp-shaped coolant passage, turned the coolant jet toward the surface away from theWhen the cusp surface in the passage is on top of the jet camera. The streak lines that appear to be moving intostream, both geometry and crossflow induced vortex the paper reveal the twisting motion of the jet as itrotations are in the same direction. For this orientation it emerges from the edge of the coolant passage indicatingwas hypothesized that, when the flow is discharged into vorticity. The counterrotating vorticity in the flow is alsothe tunnel mainstream, the geometry-induced vortices are supported by these photographs in which the streak linesalready present and do not require initiation by the main appear to be rotating in opposite directions. The divisionstream. Consequently, the mainstream has more momen- of flow appears quite clear; unfortunately the sense ofturn available to turn the jet closer to the wall, resulting in rotation direction for the individual streams can only beenhanced film-cooling effectiveness, speculated.

One of the speculations that could be proven by presented graphically as film-cooling effectiveness inphotography is the change in the trajectory of the jet terms of the centerline length of the isotherm tracesstream caused by differences in coolant-passage measured from the edge of the coolant-injection port andgeometry. This was accomplished by photographing also the surface area enclosed by the isotherm.from the side edge of the test plate. Figure 24 shows a Interactions of the vortex pair structure in the coolantstreak line emerging from top-cusp passage with no flow generated by passage geometry with the vortex paircrossflow. The jet flow ejects from the passage at an generated by the crossflow was influenced the film-angle of 30° with respect to the plate surface. The pointer cooling process. Data obtained under the same test condi-on the transparent wall of the tunnel is located at the tions showed up to factors of 4 increases in bothdownstream edge of the ejection port and establishes a centerline film-cooling effectiveness and surface areareference measurement position. A bolt through the test coverage over that obtained with the round cross sectionplate and the edge of the tunnel wall serves as a reference passage. In addition, flow separation of an inclined jetposition for measuring the height of the stream line at a from the surface containing the passage could befixed position downstream from the coolant-injection depressed by a coolant-passage geometry that generates aport. As did figure 23(a), figure 24 shows that the vortici- counterrotating vortex pair structure within the coolantty appearing in the streak line could be attributed to the flow which rotates in opposition to the crossflow-cusp passage geometry, generated vortex pair rotation.

The change in trajectory of the jet stream caused by A streak line flow visualization technique was used todifferences in passage geometry, under identical flow support the concept of the counterrotating capability ofconditions, is illustrated in figure 25. The jet trajectory the cusp-faced flow passage design. Photographic datafrom the round passage (fig. 25(a)) is significantly higher showed the coolant flow split into two streams that rotate(farther from the reference bolt) than the trajectories in opposite directions. In addition, the streak line trajec-from the top- and bottom-cusped passages (figs. 25(b) tory from the round passage was considerably higherand (c)). The reason for the inferior film-cooling perfor- than from either the top- or bottom-cusp passages, at-mance of the round jet passage (see figs. 12(f) and 19(0) testing to differences obtained in the film-cooling data.is revealed by these differences in trajectory shown infigure 25.

Lewis Research CenterNational Aeronautics and Space Administration

Summary of Results Cleveland, Ohio, July 13, 1984

The film-cooling isotherm data reported herein wereobtained by infrared photography along with an ap-

paratus designed to determine the film-cooling perfor- Referencesmance of three descrete-hole coolant-flow passages. Thepassages were embedded in adiabatic test plates and 1. Papell,s.s.; Graham,R.W.; and Cageao, R.P.: Influence ofdischarged the flow at an angle of 30° in line with the tun- Coolant Tube Curvature on Film Cooling Effectiveness asnel flow. The standard round hole cross section coolant Detected by Infrared Imagery. NASA TP-1546, 1979.

2. Papell, S.S.; Wang, C.R.; and Graham, R.W.: Film-Coolingpassages was used as a reference with which to comparethe performance of two orientations of a counterrotating Effectiveness With Developing Coolant Flow ThroughStraightand Curved Tubular Passages. NASA TP-2062, 1982.vortex-pair generating coolant passage that evolved from 3. Abramovich, G.N.: The Theory of Turbulent Jets. M.I.T. Press,an examination of the basic fluid dynamics of an inclined 1963.jet in crossflow. 4. Keffer, J.F.; and Baines, W.D.: The Round Turbulent Jet in a

Test conditions included blowing rates from 0.20 to Cross Wind, J. Fluid Mech., vol. 15, pt. 4, Apr. 1963, pp.2.05 and tunnel velocities from 15.5 to 45.0 m/s, with 481-496.

5. Colladay, R.S.; and Russell, L.M.: Streakline Flow Visualizationtemperature differences of about 22° C between the tun- of Discrete-Hole Film Cooling With Normal, Slanted, andnel and coolant-air streams. Data comparisons are CompoundAngleInjection.NASATND-8248,1976.

C_ _ Vc__'_ JET CROSSFLOW

CROSSFLOW .._ _V_ PATH

/// 7-"-.-./ / LVORTEX/ D /

/ , / PATH//

Figure!.--Inclinedjetin crossflowshowingforceson a control Figure2.--Counterrotatingvortexpairstructurewithinthejetinducedelement, by the crossflow.

I-CUSPINDUCEDVORTEXI ROTATIONDIRECTIONI

CROSSFLOW II

CUSP

_ SURFACE_ FLOWINDUCED

VORTEXROTATIONDIRECTION

/-- CUSP-INDUCEDVORTEXII ROTATIONDIRECTION

ROUND VORTEX jSTANDARD GENERATING CROSSFLOW iHOLE HOLE

(a)

c o, ,owI"AIId VORTEXROTATION

DIRECTION

JETFLOW

(b)

(a) Top cusp orientation.(a) Coolant flow passage cross sections. (b) Bottom cusp orientation.

(b) Counterrotating vortex-generating coolant passage. Figure 4.--Geometry-induced vortex pairs with respect to crossflow-Figure 3.--Design concept, induced vortex rotation.

8 /

CROSSFLOW TUNNEL•=_ FLOW

COOLANTFLOWPASSAGEGEOMETRY

I r MAHOGANY16.0 ' CASTEPOXY :

_,-B I_-1"27 I ,NSERT-,,, I TESTPLA_;-// 3. 8 I SCREENS-

_6 35 BAFFLES-,,

"" '"- , I ]tl///_ // I INSULATED /

,..I _/ L___ PLENUMWALL-'0 J..- coo, ,8 SURFACES_" SUPPLY

Figure 6.--Test plate and coolant-supply plenum.

°0?_AD--

_-0.02_BB- i BB- 2ROUND CUSP

Figure 5.--Design details of passage cross sections. (All dimensions incentimeters.)

INFRAREDOSCILLOSOCOPECAMERA DISPLAYUNIT

THINFILM

WINDOW_!

ADIABATIC '\ ALTITUDEWALL-, \ /'-

ROOMr---J-----.._-._ _ ,/i,::_=-.,._,-i,....11EXHAUST

AIR "_'_L-------'_L._ L"]""_I° _ ["J _ HOT/ _Z_" ' / "°"_l_ _ AIR

TUNNEL-" PLENUM-" ,/ COLDAIR]][ HILSCHTURBINETYPE / IITUBEOFFLOWMETER-/ (a) Round. (b) Top cusp.

AIRSUPPLY,83kPa (120psi) Figure 8.--Equal temperature isotherm traces. Infrared photographs

obtained under same test conditions. Tunnel velocity, V, 15.5 m/s;

Figure 7.--Film-cooling rig. blowing rate, M, 1.00; film cooling effectiveness, 7, 0.16.

9

.6 -- Tunnel --velocity.

Voopm/s

o4

/_45.0 r45.0

o

_"15.5 'x-22.8_

Ial I I I I (e) I I I I0

,6 --

.4 _ _ p45.0

/-30.5

.2 -- / /

.__-_ o Ibl I I I I (f_ I I I II

E'_-..6 --

_ ,-45.0

.4 _,._ ____30.5

•2 -- -- //-_22.8

15.5J

ic_ I I I I Igl I0

.6_-

.4 /-45.0 _

.2 __ -- /-22.8//rlS. 5//

I(d! I I I I (h)_T I_ I0 4 8 12 16 0 4 8 12 16

Dimensionlessdistancefromport, x/d

(a) M=0.20. (e) M=I.00.(b) M=0.34. (f) M=1.25.(c) M=0.50. (g) M=l.50.(d) M= 0.75. (h) M= 2.05.

Figure 9.--Basic round coolant passage data.

10

.6 -- Tunnel

velocity, z- 45.OVoo.

m/s /r30. 5

L-15.5

(a) I I I I (e) I I I I0

°6 --

/-30. 511-22.8

_._- _ _",_L15.5

._ o (hi I I I I (fl I I I !.6 --

.- /-45.0

-\\ L-22.8k-15.5

o (c) I I I I cgl I I I I,6 --

/-42.0

.,__S/-_°.'

(a) I I I I (h) I I I I0 4 8 12 16 4 8 12 16

Dimensionlessdistancefromport,xld

(a)M=0.20. (e) M= 1.00.(b) M=0.34. (f) M= 1.25.(c) M=0.50. (g) M=l.50.(d) M=0.75. (h) M=2.05.

Figure lO.--Basi¢ top-cusp coolant passage data.

11

,-45.5 _ /'-30.5

_:_ .4 ___F30.5

.2

15.5J

o (b) I I I I (f) I I I IE

.6 --

15"5J

(c) I I I I !(g) I I I I0

,6 --

/-45.0 . /-22.8

.4 -- .IF30.5 ._i_ 15.5

(d) I I I I (h) I I I IO 4 8 12 16 0 4 8 12 16

Dimensionlessdistancefromport,xld

(a) M=0.20. (e) M= 1.00.(b) 2=0.34. (f) M= 1.25.(c) 2=0.50. (g) M= 1.50.(d) M= 0.75. (h) M= 2.05.

Figure 11.--Basic bottom-cusp coolant passage data.

12

o6 -- M

iI I { I 1 Lhl _ I0 4 8 12 16 4 8 12 16

Dimensionlessdistancefromport,x/d

(a) M=0.20. (e) M=I.O0.(19)M=0.34. (f) M=1.25.(c) M=0.50. (g) M=l.50.(d) M=0.75. (h) M=2.05.

Figure 12.--Comparison of centerline film-cooling effectiveness distributions for free stream velocity of 15.5 m/s.

13

Coolant

•6 -- passage _configuration

-- Round.... Topcusp

.4 ,=,,.-....._ _m Bottomcusp s._...--_-....__

(a) I I I I ,I I I I I0

7_.=_=o ,hi I I I i '" I I I I_' .6 --E

.4 -- _ __ ._,,_.

I "____' "_..

0 (C)i I I I I gl_ [

0 4 8 12 16 0 4 8 12 16Dimensionlessdistancefromport,xld

(a) M=0.20. (e) M= 1.00.(lo) M=0.34. (f) M=1.25.(c) M=0.50. (g) M=l.50.(d) M-0.75. (h) M=2.05.

Figure 13.iComparison of centerline film-cooling effectiveness distributions for free-stream velocity of 22.8 m/s.

14

Coolantpassage

--_'_, configuration

.4 _:. --Round.... Topcusp I_,_ _ Bottomcusp

°2 _

(a) I I I I0

.4 __ _"_,_Z" -"-....

-(bl I I I I (e)_o i I I I

._ .6 -

.4J _'-.,, __x.,==z._,.

o (_ I I I I (n I I I l

• _'_ -_><.T_._..

(d_ I I I I (gr I I " I -" I0 4 8 12 16 0 4 8 12 16

Dimensionlessdistancefromport,xld

(a) M=0.20. (e) M= l.O0.(b) M=0.34. (f) M=1.25.(c) M=0.50. (g) M= 1.50.(d) M= 0.75.

Figure 14.--Comparison of centerline film-cooling effectiveness distributions for free-stream velocity of 30.5 m/s.

15

.6 --

Coolant

•4 -- passageconfiguration

Round_--- Topcusp

•2 -- _ Bottomcusp

= o <a_ 1 I I I_.6 --

.4

.__

_o Ibl I I I I !dl I I I I

• 6 €__f_.._

[c) I I I I (e) I I I I0 4 8 12 16 4 8 12 16

Dimensionlessdistancefromport,x/d

(a) M=0.20. (d) M=0.75.(b) M=0.34. (e) M= 1.00.(c) M= 0.50.

Figure 15.--Comparison of centerline film-cooling effectiveness distributions for free-stream velocity of 45.0 m/s.

16

.2 .30.5 m {15.5 .2 ;22.8

L22. 8 f _30.5 L15.5

0 20 40 60 0 20 40 0 20 40 0 20 40

Isothermarea,cm2 Isothermarea,cm2

(a) M=0.20. (e)M= 1.00. (a) M=0.20. (e) M=I.00.(b) M=0.34. (f) M=1.25. (b) M=0.34. (f) M=1.25.(c) M=0.50. (g) M= 1.50. (c) M=0.50. (g) M= 1.50.(d) M=0.75. 0a) M=2.05. (d) M=0.75. (h) M=2.05.

Figure 16.--Basic round hole film-cooling coverage data. Figure 17.--Basic top-cusp film-cooling coverage data.

17

.6 --

Tunnelvelocity,

.4 _- Voo,_k m/s

._45.0 _45.02 ._N_,_.5 30.5 CoolantI passage

•4 -- configuration _

L-22,8 _ 15,5j" \\ _ Round0 (a)15.5-" I _ I L --- Topcusp \

_. \ 1 Bottomcusp _ ,

.2 -- \_

,4. _ L f./_.

_45.0 0 (a) I I I _e) J I30.5

'- o ml ""'1 .I m I I _='2E

_ ICb_ I I I I=+o.4 - .+

,--45.0 _ .4 _-- --

]--2_ 8

.2 5_.2 1

o I I __.6 - ocl I I I _91 _"" I

.4 \-- I,4 -- \\\

-_,+P

.2 _,. : _L_8Id_ I I I

l(h) l I 0 20 40 60 0 200 20 40 20 40 Isothermarea,cm2

Isothermarea,cm2 (a) M=0.20. (e) M= ].00.(a) M=0.20. (e) M= LO0. Co)M=0.34. (0 M= 1.25.(b) M=0.34. (_ M=].25. (c) M=0.50. (g) M=1.50.(c)M=0.50. (g)M=l.50. (d)M=0.75. (h)M=2.05.

(d)M=0.75. (h)M=2.05. Figure19.--Film-coolingeffectivenessasfunctionofisothermareafor

Figure18.--Basicbottom-cuspfilm-coolingcoveragedata• free-streamvelocityof15.5m/sec.

18

6F Coolantpassage

configuration

.4 _ Round_---- Topcusp_ Bottomcusp

.2-\\\-. _ o I"6F -o (al I (el I I

.6 -- _ c- .4Coolantpassage 8,

configuration.4 _-_ - Round -- _ .2

"\_ x - Topcusp '_ '_- _x._\x - Bottomcusp "_ "'_"

_ ,. 8 (b} I I I

__.6 --

'--= _.4 --.6 -- -- -

E

E

-- \

.2- "__... o I I I I

o (c_ I I I __]

.4 r'- -- .4 _-_\ --

(al I I I (%_"'_ _- I (dl I I I (gl I_""0 20 40 60 20 40 0 20 40 60 0 20 40

Isothermarea,cm2 Isothermarea,cm2

(a) M=0.20. (e) M= 1.00. ta) M=U.ZO. (e) M= 1.00.Co)M=0.34. (f) M= 1.25. (b) M=0.34. (f) M=1.25.(c) M=0.50. (g) M= 1.50. (c) M=0.50. (g) M=l.50.(d) M=0.75. (h) M=2.05. (d) M=0.75.

Figure 20.--Comparison of film-cooling coverage data for free-stream Figure 21.--Comparison of fillm-cooling coverage data for free-streamvelocity of 22.8 m/sec, velocity of 30.5 m/s.

19

• 6 -- Coolantpassage

\ configuration.4 _-_ -- Round

.... Topcusp

---- Bottomcusp

.2

lal I I0

c-

.__

o COOLANTV . 2 HOLE'--_"E

EI I I I_ o

.6 --

.4

.2

__1 I I0 20 40 20 40

Isothermarea,cm2 (a) Top cusp; tunnel velocity, 0; coolant velocity, !.6 m/s.(a) M=0.20. (d) M=0.75. (b) Top cusp; tunnel velocity, 8.7 m/s; coolant velocity, 2.9 m/s;(b) M=0.34. (e) M= 1.00. blowing rate, 0.34.(c) M=0.50. (c) Bottom cusp; tunnel velocity, 15.2 m/s; coolant velocity, 5.0 m/s;

Figure 22.--Comparison of film-cooling coverage data for free-stream blowing rate, 0.34.velocity of 45.0 m/s. Figure 23.--Tunnel view looking downstream.

20

POINTERSHOWINGTUNNELDOWNSTREAM FLOW"'_ STREAKLINEEDGEOFHOLE

POINTERSHOWlNG "STREAKLINEDOWNSTREAMEDGEOFHOLE

Figure 24.--Side view of top-cusp configuration. Tunnel velocity, 0;coolant velocity, 7.1 m/s.

(a) Round.(b) Top cusp.

(c) Bottom cusp.

Figure 25.--Side view. Free-stream velocity, 15.2; coolant velocity,18.7; blowing rate, 1.25.

21

1. Report No. 2. Government Accession No. 3. Reclplent's Catalog No.

NASATP-23884. Title and Subtitle 5. Report Date

Vortex-Generating Coolant-Flow-Passage Design for November 1984Increased Film-Cool i ng Effectiveness and Surface 6.Performing Organization Code

Coverage 505-41-32

7. Author(s) 8. Performing Organization Report No.

E-2147

S. Stephen Papel I 1o.WorkUnitNo.

9. Performing Organization Name and Address11. Contract or Grant No.

National Aeronautics and Space AdministrationLewis Research CenterC1eve1and, Ohi o 44135 13. Type of Report and Period Covered

12. Sponsoring Agency Name and Address Techni ca1 PaperNational Aeronautics and Space Administration 14. Sponsoring Agency CodeWashington, D.C. 20546

15. Supplementary Notes

16. Abstract

Reported herein are the thermal film-cooling footprints observed by infraredimagery for three coolant-passage configurations embedded in adiabatic-testplates: A standard round-hole cross section and two orientations of a vortex-generating flow passage. Both orientations showed up to factors of fourincreases in both film-cooling effectiveness and surface coverage over thatobtained with the round coolant passage. The crossflow data covered a rangeof tunnel velocities from 15.5 to 45 m/sec with blowing rates from 0.20 to 2.05.A photographic streakline flow visualization technique supported the conceptof the counterrotating capability of the flow passage design and gave visualcredence to its role in inhibiting flow separation.

17. Key Words (Suggested by Author(s)) 18. Distribution Statement

Film cooling Unclassified - unlimitedVortex STARcategory 34Heat transferFluid mechanicsTurbine cooling

19. Security Classlf. (of this report) 20. Security Classif. (of this page) 21. No. of pages 22. Price*

Unclassified Unclassified 22 A02I

*For sale by the National Technical Information Service, Springfield, Virginia 22161NASA-Langley, 1984

National Aeronautics and THIRD-CLASS BULK RATE Postageand Fees Paid _=""=_="=_

SpaceAdministration National Aeronautics andSpace AdministrationNASA-451

Washington, D.C. _ _s_c_ O_TE_20546

Official Business 3 1176 01326 5898Penalty for Private Use, $300 __....

DO NOT REMOVE SLIP FROM MATERIAL

Postal Manual) ]Do Not Return

to the library.

NAME DATE MS

NASA Langley (Rev. Dec. 1991) R AD N-75