NASA Technical Memorandum 104428 AIAA-91- 3354 :i Evaluation of Panel Code Predictions With Experimental Results of Inlet Performance for a 17-Inch Ducted Prop/Fan Simulator Operating at Mach 0.2 D.R. Boldman, C. Iek, D.P. Hwang, and R.J. Jeracki Lewis Research Center Cleveland, Ohio and M. Larkin and G. Sorin Pratt and Whitney Aircraft East Hartford, Connecticut Prepared for the 27th Joint Propulsion Conference cosponsored by the AIAA, SAE, ASME, and ASEE Sacramento, California, June 24- 27, 1991 N/ A ( ',) : _ _ : t t :; _/ cJ;7 EVALUATION OF PANEL CODE PREDICTIONS WITH EXPERIMENTAL RESULTS OF INLET PERFORMANCE FOR A 17-INCH DUCTED PROP/FAN SIMULATOR OPERATING AT MACH 0.2 D. R. Boldman 1,C. lek, D. P. Hwang, and R. J. Jeracki National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 M. Larkin and G. Sorin Pratt and Whitney Aircraft East Hartford, Connecticut 06108 Abst_ct Subscripts An axisymmetric panel code was used to evaluate c corrected to standard day conditions the performance of a series of ducted propeller inlets i incompressible value which were designed by P&W Aircraft and tested in the LOC local condition NASA Lewis 9- by 15-Foot Low Speed Wind Tunnel as MAX maximum value part of ajoint program with P&W Aircraft. Three basic 0 total condition inlets having ratios of shroud length to propeller diame- PROP propeller face ter of 0.2, 0.4, and 0.5 were tested with the P&W ducted S static condition prop/fan simulator. A fourth "hybrid" inlet consisting of cor corrected for compressibility the shroud from the shortest basic inlet coupled with the spinner from the longest basic inlet was also tested. This Superscript latter configuration represented the shortest overall inlet. The simulator duct diameter at the propeller face was - average value 17.25 inches. The short and long spinners provided hub- to-tip ratios of 0.44 at the propeller face. The four inlets Backaround were tested at a nominal free stream Mach number of 0.2 and at angles-of-attack from 0° to 35° (upper limit of Panel codes have been demonstrated to be powerful the model support system). The panel code method tools for the design of a variety of subsonic inlets includ- incorporated a simple two-part separation model which ing short inlets for VTOL and STOL aircraft operating yielded conservative estimates of inlet separation. at high angles-of-attack 02) (often approaching 90°). Panel methods have been extended to include complex 3- Nomenclature D geometries with and without slats ¢3"8)and have yielded good predictions of the surface static pressures at A area conditions in which the inlet was operating free of Cf friction coefficient separated flow. In many of the applications panel codes D diameter were compared to data from inlets where the pumping of L shroud length the flow was achieved by some external means rather M Mach number than by an integral propeller or fan. In other related p static pressure experiments in which an integral fan was used to pump P total pressure the flow, the inlets were relatively long (X/DF_ > 0.5). V velocity In either case the pumping mechanism probably did not W mass flow rate exert a strong influence on the inlet flow. Consequently, X axial distance from propeller face the panel codes, which do not account for the presence angle-of-attack of a fan or propeller (except through the mass flow rate), p density would be expected to yield good predictions of the LAerospace Research Engineer Assoc. Fellow AIAA surface pressures inside the inlet, particularly when the unity velocity vectors parallel to each of the two cartt;sian inlet was free of separated flow. In the present paper, coordinate axes. The X-axis represents the axial direc- the Douglas-Neumann panel code EOD c9"11)in conjunc- tion for zero angle-of-attack. The basic solutiom are tion with SCIRCL 02) and COMBYN °2) were used to obtained by replacing the surface by a number of p;mels predict the inlet static pressures and separation observed having a surface source (or sink) distribution of unknown in recent tests with the P&W ducted prop/fan simulator. strength. In the present applications, a piecewise-l_ara- The code COMBYN incorporated a compressibility bolic source strength distribution was assumed in con- correction. A boundary layer calculation, which will be junction with a higher order calculation which uses described later, was added to COMBYN in order to curved surface elements. A distribution of unit vortices predict diffuser separation. This unique data base on the shroud surface is used to induce a static mass flow provided an excellent opportunity to re-examine the through the inlet (in addition to the distribution of panel code for subsonic inlet design since the inlet local sources that represent the inlet profile). Arbitrary static Mach number levels were often perceived to be too high mass flows are obtained by using a multiplication factor. for this class of design codes. The basic solutions obtained from EOD are com- In comparing the panel code results with inlet data bined linearly in a program called COMBYN to provide from the simulator, emphasis has been placed on (1) the solutions of interest for the prescribed flow conditions, comparison of inlet surface static pressures on the free stream Mach number, angle-of-attack, and mass windward side as the angle-of-attack approached the flow. The linearly combined incompressible solution is separation value, and (2) the comparison of predicted corrected for compressibility by the Lieblein-Stockm;m c13) and observed angle-of-attack corresponding to the "onset" compressibility correction. This correction, simply stated, of separation. is ,1, Panel Code Potential Flow where the terms on the right side are obtained fron, the Since the theory for the panel code is well docu- input conditions and the incompressible flow solution. mented °2), only a brief description of the method will be The other desired properties such as Mach number and given. The basic problem consists of calculating the pressure ratio are obtained from the compressible potential flow (with a correction for compressibility velocity, V_o_. effects) about an axisymmetric ducted propeller inlet at any combination of inlet mass flow, We, and inlet angle- The solutions from COMBYN require the specifi- of-attack, a. A series of programs developed at NASA cation of a control station which, in the present study, Lewis Research Center in the early seventies are used to was chosen to be the throat of the inlet. Since the solve this problem. The first program, SCIRCL, repre- compressibility correction does not exactly satisfy conti- sents the axisymmetrie inlet geometry by its meridional nuity, the COMBYN solutions are most accurate near profile with the shroud and spinner extended far down- the control station (or near the throat region in the stream in order to obtain an accurate potential flow solu- present application). tion in the region between the highlight and propeller face. SCIRCL breaks the profile into segments with a Boundary Laver and Separation Modgi control point on each segment used for the potential flow calculations. The program also generates off-body points The code capability as described above allow:; for such as flow measuring rakes at prescribed axial stations. the calculation of the inlet static pressure distribtttion One such rake station represents a "control" station for upon accounting for compressibility. It now remains to subsequent use in the COMBYN program. determine the maximum angle-of-attack that car_ be obtained prior to boundary layer separation in the inlet. The Douglas-Neumann program, EOD ¢9"_1)p,rovid- Several empirically based separation models have been ed the fundamental incompressible potential flow field used in the past with varying degrees of success °4"_. The for the geometry specified by SCIRCL. In EOD, three separation model used in this paper is based, in par_, on basic flow conditions are computed; namely, a static experimental observations from past studies of inlets condition (M 0-- 0), and two stream flow conditions with 2 operatingat highangles-of-attaackndatflow rates Euler code, and data were made for the conventional sufficiently high to produce locally supersonic flow in the inlet. internal lip region of the windward side of the inlet. Based on these past experiments, it was concluded that Experimental Rig a shock-induced lip separation of the internal flow in the shroud could be expected when the Mach number Installation reached a value of about 1.504). However, if the Mach number remains below 1.5, separation of the boundary The P&W 17-in. diameter ducted propeller simula- layer can still occur in the diffuser starting near the exit tor was installed in the 9- by 15- Foot Low Speed Wind or propeller face and moving upstream with an increasing Tunnel as shown in Fig. 2. The centerline of the simula- adverse pressure gradient resulting from an increasing tor was 51 in. from the tunnel floor and was offset 21 in. angle-of-attack. These two observations were combined in a lateral direction from the tunnel centerline. This to provide the simple separation model used in the offset resulted in a model centerline which was 65 in. present study. This model is depicted in Fig. 1. The first from the near wall of the tunnel and 113 in. from the far limiting condition in climb angle-of-attack arises when wall. The simulator was rotated about the pivot axis in the Mach number limit of 1.5 is exceeded as shown in a counterclockwise direction to increase the angle-of- Fig. la. The second condition which can limit the climb attack, i.e. an increasing angle-of-attack was obtained as angle-of-attack is when the local Mach number does not the model was rotated laterally toward the wind tunnel reach a value of 1.5 but separation occurs in the diffuser centerline. The maximum angle-of-attack for the support (Cf -- 0) as suggested in Fig. lb. Since the panel code system was 35°. The propeller was driven by a 1,000 HP model of the inlet is based on a flow-through duct air turbine drive system at rotational speeds up to 12,000 extending far downstream, the calculated boundary layer RPM. can separate anywhere inside the long duct. For purpos- es of identifying the diffusion-limited angle-of-attack, it is Inlets assumed that this limit is reached when the calculated separation occurs upstream of the diffuser exit. The The inlets which were tested on the ducted propel- diffuser exit is defined as the plane representing the ler simulator are shown in Fig. 3. These inlets were location of the propeller face. assembled from three different shrouds and two spinners. The shorter of the two spinners had a length approxi- The analysis of the boundary layer is based on the mately equal to the length of the conventional shroud 2-dimensional compressible boundary layer program of (0.5 DpRop). The longer spinner had a length of about Herring ¢t6). This numerical method calculates the usual 0.7 Dpaop and was used only with the shortest shroud to boundary layer parameters including the skin friction form the "plug" inlet shown in Fig. 3d. The short spinner coefficient, Ct . Flow separation occurs when Ct = 0. was also combined with the three shrouds to form the The program contains several options for controlling the "conventional", "hybrid", and "midlength" inlets shown in boundary layer development including the initialization Figs. 3a to 3c, respectively. The inlet lengths based on and transition criterion. A fixed set of assumptions were the ratio of shroud length to propeller diameter ranged made concerning the boundary layer options and no from 0.2 to 0.5. effort was made to alter these options for the results presented herein. Instrumentation 3-D Euler Code The simulator was extensively instrumented through- out the flow path in order to provide data for this multi- In addition to the panel code predictions, a few purpose program in which the present inlet results comparisons of the inlet static pressure distributions were comprised only a small but important part. Static made with a 3-D version of an Euler flow solver devel- pressure taps, thermocouples for health monitoring and oped by NiC17).The Euler flow solver uses a fast explicit rakes, rake total pressures, and blade clearance proximity numerical scheme for solving the unsteady Euler flow probes accounted for nearly 550 channels of steady state equations to obtain steady solutions. The scheme is data. Several additional types of instrumentation for constructed by combining the multiple grid technique dynamic measurements and acoustics were included to with a second order accurate finite volume integration support other aspects of the program. The inlet results method. Comparisons between the panel code, 3-D presented in this paper were based primarily on measure- 3 ments of surface static pressures on the windward side of tions of the inlet static pressure distributions for the the inlets, total pressure contours for assessing the conventional inlet are shown in Fig. 4. These results are degree of distortion ahead of the propeller, angle-of- for three angles-of-attack ranging from 25.0° (Fig. 4a) to attack, captured mass flow rate, and free stream condi- 27.3° (Fig 4c). At a = 25.0°, the observed minimum tions. value of P/P0 reaches a value of 0.27 which corresp,_nds to a local Mach number of 1.5 (Fig. 4a). At the higher Emphasis has been placed on the data obtained at values of a, the experimental minimum pressure ratio angles-of-attack above about 25° where inlet separation starts to increase resulting in lower peak values of iocal starts to become a concern. The number of static Mach number (Figs. 4b and 4c). Both the axisymmetric pressure measurements in the inlet ranged from 9for the panel and 3-D Euler codes yield excellent agreement with short shroud to 14 for the midlength shroud. The each other and with the data at a = 25.0° where the spacing between pressure taps was designed to provide experimental peak Mach number was observed. At good resolution in the highlight region where shock- higher angles-of-attack where the peak Mach nuraber induced separation can occur. The concentration of starts to drop (minimum value of P/P0 increases), the pressure taps in this region will become apparent upon two codes continue to predict higher values of peak examining the pressure distributions presented in the Mach number since viscous effects are not included in section entitled Results. the codes. The inlet total pressure distortion profiles were The results of these comparisons indicate thai the obtained from four rakes, each containing 12 tubes. potential flow code (with a compressibility correclion) Distortion profiles will be presented for the midlength provides excellent agreement with the 3-D Euler code at inlet with the rakes located 3.0 in. upstream of the conditions where separation has not occurred. Also, _oth propeller face. The tube spacing ranged from 0.5 in. codes yield excellent agreement with the data for the near the spinner to 0.23 in. near the shroud. The tube conventional inlet up to the angle-of-attack where the nearest the shroud was displaced 0.06 in. from the flow separates. surface. The rakes were positioned circumferentially at 0° (leeward side), 20°, 170°, and 185° (representing the Panel Code Predictions vs. Exl_riment, Shorter Inlets windward side). Symmetry was assumed about the 0° - 180° axis. This symmetry assumption coupled with an The panel code calculations were extended to interpolation program provided an additional nine pseudo include the other three inlets operating at the highest rakes which were used to estimate the total pressure mass flow rate. The predicted inlet static pressure contours. distributions for the hybrid, midlength, and plug inlets are compared with the experimental results in Figs. 5 to Test Conditions 7. The experimental peak values of local Mach number increased relative to the value obtained in the conven- Results are presented for a nominal free stream tional inlet and permitted separation-free operation at Mach number of 0.2. The tests were performed over an much higher angles-of-attack as shown in Figs. 5 and 7. angle-of-attack range of 0° to 35°and at corrected speeds This can be noted by comparing the experimental of 7,500 to 12,000 RPM. Corrected flows ranged from minimum values of P/P0 to the value of 0.27 which nominal values of 30.5 to 45.5 lb/sec. In the comparisons corresponds to a Mach number of 1.5. The local Mach between experiment and theory, the experimental values numbers for the short shroud hybrid and plug configura- of corrected mass flow were used in the theory rather tions reached levels of 1.65 at angles-of-attack of 34.3°. than the nominal values. These experimental values may At slightly higher values of a, full separation ca_ be vary by about ± 1lb/sec from the nominal values given noted by an abrupt increase in the minimum value of above. static pressure followed by a flattening of the internal pressure distribution (Figs. 5c and 7c). The midlength Results inlet peak Mach number reached 1.5; however this Mach number was attained at a = 29.0° rather than at 25.0° Panel Code vs. 3-D Euler Predictions, Conventional Inlet which was observed for the conventional inlet. Again, the panel code predictions of static pressure were generally in good agreement with experimental results up The results of panel and 3-D Euler code predic- to the point of separation. Close examination of Figs. 5 to 7 reveals a slightly lower predicted minimum static As indicated above, the value of a corresponding to pressure than observed experimentally, however, in view the maximum value of Mt_C.MAX(a) was selected as the of the difficulty in obtaining higher experimental resolu- limiting angle-of-attack for a separation-free inlet. In tion of the pressures in the highlight region, it cannot be Fig. 8b, the distributions of Mt_c_o.x(a ) at low and high determined whether these differences are real or whether corrected flows show a distinct peak value of Mt.oc_aAx. the true minimum in static pressure was not measured However, the distribution at the low corrected flow does because it may not have occurred precisely at the loca- not exhibit the pronounced drop in Mach number that tion of a static tap. was characteristic of the results at high flow rates. Although the decrease in Mt_C.MAx after the peak value Experimental Seuaration is more gradual with lower flows, separation is still present in the inlet. This separation occurs in the In order to perform comparisons between predicted diffuser and, as shown below, the extent of this separated and experimental separation, consideration must be given region tends to increase with increasing angle-of-attack to the method of determining, from the measurements, until the inlet becomes fully separated. when separation has occurred in the inlet. In the preced- ing discussion, Figs. 5c and 7c were presented to show Lip separation and diffuser separation can be the static pressure distribution in a fully separated inlet, observed in the midlength inlet by comparing the distri- i.e. an inlet which separates at the lip. In these cases, the butions of Mtoc(X) at the limiting values of a and at inlet exhibits a normal (unseparated) type of pressure angles greater than the limiting values. Two such distribution at an angle-of-attack of only about one distributions are shown in Fig. 9 for the highest flow degree less than the value associated with lip separation. condition. In Fig. 9a the distribution is shown for a = In other words, the lip separation occurs rather abruptly 28.3° which corresponds to a peak value of Mtxx:_,t_x of after some limiting value of a is reached. This type of 1.51 (Fig. 8b). The inlet Mach number distribution separation generally occurred at the high corrected mass shown in Fig. 9b corresponds to a slightly higher a of flow operating condition. For purposes of selecting an 29.9° and indicates full separation emanating from the lip experimental lip separation criterion, this limiting value of the inlet. This separation is evident from the relatively of a was assumed to be the angle-of-attack where the fiat distribution of Mach number over the entire inlet. highest value of inlet local Mach number is obtained. On the basis of this illustration for the midlength inlet, This criterion was also applied to the results obtained at the maximum value of Mtoc_Ax(a ) appears to represent the low mass flow rates to determine when separation a proper criterion for determining the limiting angle-of- occurred in the diffuser. Experimental support for these attack for the prevention of separation (lip separation in assumptions will be provided below. this illustration). First, consider the distributions of MLoc_t_x(a ) for The second type of separation; namely, diffuser the midlength inlet in Fig. 8. Two sets of distributions separation, occurs when the peak inlet Mach number is are provided to show (1) the influence of inlet rakes on less than 1.5 but a peak value appears in the Mt.oc_x(a) the results, and (2) the reduction of Mt_c_o. x with an distribution. Such acondition is apparent in the low flow increase in ,, after the peak Mach number is reached. results shown in Fig. 8b. The inlet Mach number distri- The influence of the rakes can be noted by comparing butions for three angles-of-attack at and above the Figs. 8a and 8b. The distributions of Mt_c._o_x(a ) are limiting value are shown in Fig. 10. In Fig. 10a the inlet essentially the same except for a slight increase in the Mach number distribution is shown for a = 28.4° which limiting value of a when the rakes were installed. The corresponds to a peak value of Mtoc._o.x of 0.80 (Fig. presence of the rakes, located 3.0 inches upstream of the 8b). Separation is not evident because the local Mach propeller face, can increase the separation angle-of-attack number continuously decreases without regions of in two ways. First of all, the blockage from the rakes constant Mach number which would suggest separation. changes the diffusion rate in the inlet which may delay In Fig. 10b the distribution is shown for a = 30.1°. Two the separation. A second factor concerns the turbulence changes become apparent. First of all, the peak Mach generated by the rakes which energizes the diffuser number has decreased. Secondly, aregion (-5 < X < -4) boundary layer and delays separation. All of the results, of constant Mach number, representing separation, with the exception of the inlet total pressure contours, appears in the diffuser. The extent of this separation in- were obtained with the rakes removed. creases with increasing angle-of-attack until finally, the inlet becomes fully separated. This condition, shown in Fig. 10c, occurs at an angle-of-attack of 32.4°. Predicted Setmratlon Ba_¢fl on Simple Model In Figs. 9and 10, separation was determined on the The two-part separation model shown in Fig. 1was basis of a region of constant Mach number in the inlet. incorporated in the panel code to provide estimateg of Additional support for selecting the separation angle on the separation angle-of-attack. Results for the four k,lets the basis of a peak value in the MLOC.M_X(a) distribution are shown in Fig. 17 where the separation angle-of-atl ack can be provided if one considers the reason for the drop is plotted as a function of the specific flow at the propel- in MLOC.MAwXhen separation ispresent. Before the onset ler face. The region under the solid line for each inlet of inlet separation, at a given angle of attack, the high represents the conditions for separation-free inlet opzra- rate of flow curvature around the tip reduces the static tion. The dashed line is an extension of the predi,'ted pressure indicating high values of surface Mach number. diffuser separation (C_ = 0). This line would repre:;ent As the inlet angle-of-attack is increased, tip separation the predicted separation if the separation model did not begins to occur causing a reduced rate of curvature and contain the Mach 1.5 limitation. Diffuser separation was results in lower levels of Mach number. As the angle-of- the dominant limiting factor at the lower values of attack continues to increase, the separation region grows specific flow. The calculations suggested that the highest until there is very tittle or no tip acceleration. In addi- angles-of-attack could be achieved with the short shroud tion, the inlet separation produces a total pressure loss inlets (hybrid and plug). Predictions for the midlength which, for a given fan speed line, reduces fan corrected and conventional inlets showed a difference of only about flow and further reduces the inlet surface Mach number. 2° in diffuser separation angle-of-attack with the midle- ngth inlet predicted to yield slightly higher value:, of Distributions of MLOC_AX(a) are shown in Figs. 11 separation angle. Similar trends are apparent in the to 14 for the four inlets with the inlet rakes removed. curves representing the Mach 1.5 limitation; how_:ver Each figure shows the distributions at the low and high these tines indicate a rather pronounced reduction in corrected flow rates. A maximum in these distributions separation angle with increasing flow rate. occurs at the high mass flow rates in all of the inlets. This maximum was not reached with the short shroud Predicted vs. Ex_rimental Separation inlets operating at low flow rates because of the 35° angle-of-attack limit of the test stand (Figs. 12 and 14). Fig. 18shows the results of experimental separa_ ion A maximum was not obtained for the conventional inlet superimposed on the results of the predicted separation operating at the low mass flow rate (Fig. 11) because of which were shown in Fig. 17. The simple separation high propeller strain levels possibly caused by diffuser model yields reasonable predictions of the separation for separation. the conventional, hybrid, and midlength inlets (Figs. :iSa, 18b, and 18c, respectively) with a general tendency to Inlet Total Pressure Contours. Midlength Inlet underpredict the experimental separation angle. The results for the plug inlet, shown in Fig. 18d, tend tc be Inlet total pressure contours for the midlength inlet the most conservative, particularly at the highest flow are shown in Figs. 15 and 16 for the lowest and highest rates. One of the reasons concerns the Mach 1.5 limita- corrected mass flow rates, respectively. Very tittle total tion at the high flows. As noted earlier, the short shr.)ud pressure deficit can be observed at the low flow condi- inlets operated with peak Mach numbers as high as 1.65. tion, even at an a of 31.2° which is beyond the angle of If the short shroud results for the hybrid and plug inlets peak Mach number (refer to Fig. 8a). It will be shown are compared, the panel code with the separation model later that separation in the diffuser was predicted for suggests that higher angles should be achievable with the these low flow, high a conditions. The results in Fig. 16, hybrid inlet (Fig. 17). However, if the experime:atal obtained at the high flow rate, show tittle total pressure results are compared for these inlets (Figs. 18b and 18d) deficit at angles up to 29.9°; however, a pronounced tittle difference can be detected. The differences in the distortion is apparent at a slightly higher a of 31.3°. As calculated results must be related to the difference: in shown in Fig. 8a, these angles correspond to the peak spinner geometry. The panel code indicated sligtatly Mach number condition and a condition where MLoc_o. x higher separation-free angles-of-attack with the hybrid has experienced a rapid drop. inlet which contained the short spinner. However, the experiment indicated tittle effect of the spinner on the shroud pressure distributions. The results in Fig. 18 suggest that the separation A panel method involving a series of codes was used angle-of-attack might be represented exclusively by the to predict the inlet pressure distributions and separation calculated results based on the boundary layer analysis angle-of-attack. Limited calculations indicated that the since the data and the analysis both reveal only modest panel code results for the inlet static pressure distribution reductions in the separation angle with increasing specific were in good agreement with 3-D Euler predictions at flow. On the other hand, the lip separation curves based angles-of-attack approaching the separation value. The on a limiting Mach number of 1.5, indicate a rapid panel code was used in conjunction with a simple two- decline in the limiting angle-of-attack at high specific part model to predict separation. This model, consisting flows. This criterion provided conservative estimates of of a boundary layer separation calculation (Ct = 0) and lip separation for the inlets of this investigation. It is a limiting Mach number criterion, yielded good estimates recognized that the Mach 1.5 limit was based on an of the inlet separation over most of the operating range. average of the results from several inlets °4) which fell The results were generally conservative, especially at high within a Mach number band of about 1.5 ±0.15. Indeed, flow conditions. At high flow rates, the local Mach the present results for the short inlets indicated that the number limit of 1.5 used in the model, was too low for local Mach number can be as high as 1.65 without the short shroud inlets. Experiments for these inlets separation. However, as the specific flow increases, lip indicated separation-free operation with local Mach separation will eventually occur, and the rapid drop in numbers as high as 1.65. separation angle-of-attack represented by the Mach 1.5 limiting lines would be expected. Consequently, in References utilizing panel codes for the design of inlets for large ADP systems, which might require higher specific flows, 1. Hawk, J. D., and Stockman, N. O., "I'heoretical a conservative approach of maintaining the Mach 1.5 Study of VTOL Tilt-Nacelle Axisymmetric Inlet limit in the simple separation model is recommended. Geometries," NASA TP-1380, Jan. 1979. Concluding Remark_ . Boles, M. A., Luidens, R. W., and Stockman, N. O., "Theoretical Flow Characteristics of Inlets for The results for the hybrid and plug inlets, which Tilting-Nacelle VTOL Aircraft," NASA TP-1205, contained a short shroud (L/Dpgop = 0.2), revealed April 1978. surprisinglygood aerodynamic climb performance. These short inlets were able to support local Mach numbers as , Hwang, D. P., and Diedrich, J. H., "A Summary of high as 1.65 without incurring lip separation, whereas the V/STOL Inlet Analysis Methods," NASA longer conventional and midlength inlets were limited to TM-82725, Dec. 1981. local Mach numbers of about 1.5. The reason for these differences is still unclear; however, it is believed to be . Hwang, D. P., and Abbott, J. M., "A Summary of associated with the pumping effect of the propeller. The V/STOL Inlet Analysis Methods," NASA influence of the propeller would tend to become more TM-82885, Aug. 1982. dominant as the inlet becomes shorter. . Hwang, D. P., "A Numerical Analysis Applied to Summary High Angle-of-Attack Three-Dimensional Inlets," AIAA-86-1627, June 1986. An axisymmetric panel code was used to evaluate the performance of a series of ducted propeller inlets . Kao, H. C., "Some Aspects of Calculating Flows which were designed by P&W Aircraft and tested in the About Three-Dimensional Subsonic Inlets, NASA NASA Lewis 9- by 15-Foot Low Speed Wind Tunnel as TM-82678, July 1981. part of a joint program with P&W Aircraft. Four inlets, with ratios of shroud length to propeller diameter of 0.2 . Hess, J. L., Mack, D. P., and Stockman, N. O., "An to 0.5, were tested with the P&W 17-in. ducted prop/fan Efficient User-Oriented Method for Calculating simulator. A short and long spinner were used in various Compressible Flow in and About Three-Dimen- combinations with the shrouds. These spinners provided sional Inlets - Panel Method," NASA CR-159578, hub-to-tip ratios of 0.44 at the propeller face. The tests April 1979. were performed at a free stream Mach number of 0.2 and at angles-of-attack from 0° to 35°. 7 ° Hess, J. L., Friedman, D. M., and Clark, R. W., 13. Lieblein, S., and Stockman, N. O., "Compressibility "Calculation of Compressible Flow About Three- Correction for Internal Flow Solution," Journal 9f Dimensional Inlets with Auxiliary Inlets, Slats, and Aircraft, Vol. 9, April 1972, pp 312-313. Vanes by Means of a Panel Method," NASA CR-174975, June 1985. 14. Luidens, R. W., Stockman, N. O., and Diedrich, J. H., "Optimum Subsonic, High-Angle-of-Attack ° Smith, A. M. O., and Pierce, J., "Exact Solution of Nacelles," ICAS Proceedings 1980. 12th Congress the Neumann Problem. Calculation of Non-Circu- of the International Council 9f the Acr0nau_:cal latory Plane and Axially Symmetric Flows About or Sciences. Oct. 1980. Munich/Federal Repubiic of Within Arbitrary Bodies," ES-26988, Douglas Germany, edited by J. Singer and R. Staubenbiel, Aircraft Co., April 1958. AIAA, 1980, pp. 530-541. 10. Hess, J. L., and Smith, A. M. O., "Calculation of 15. Burley, R. R., "Experimental Investigation of Tan- Potential Flow About Arbitrary Bodies," gential Blowing Applied to a Subsonic V/STOL in Aeronautical Sciences, Vol. 8, D. A. Kuchema- Inlet," NASA TP-2297, April 1984. nn, ed., Pergamon Press, Elmsford, NY, 1967, pp. 1-138. 16. Herring, H. J., "PL2 - A Calculation Method for 11. Hess, J. L., "Calculation of Potential Flow About Two-Dimensional Boundary Layers with Cross,flow Bodies of Revolution Having Axes Perpendicular and Heat Transfer," Dynalysis of Princeton, R_port to the Free-Stream Direction," Journal of the No. 65, July 1980. Aerospace Science_, Vol. 29, No. 6, June 1962, pp. 726-742. 17. Ni, Ron-Ho, "A Multiple Grid Scheme for Solving the Euler Equations," AIAA-81-1025, June 1981. 12. Stockman, N. O., and Farrell, Jr., C. A., "Improved Computer Programs for Calculating Potential Flow in Propulsion System Inlets," NASA TM-73728, July 1977.