A review of out-of-ground-effect propulsion-induced interference on STOVL aircraft

ARTICLE IN PRESS

Progress in Aerospace Sciences 41 (2005) 175–191 www.elsevier.com/locate/paerosci

A review of out-of-ground-effect propulsion-induced interference on STOVL aircraft A.J. Saddington, K. Knowles Aeromechanical Systems Group, Department of Aerospace, Power and Sensors, Cranfield University, RMCS, Shrivenham, Swindon, Wiltshire, SN6 8LA, UK Available online 14 June 2005

Abstract This paper provides a review of the out-of-ground-effect propulsion-induced interference on the aerodynamics of jetlift short take-off and vertical landing (STOVL) aircraft. Two main flight regimes are discussed; hover and transition wherein the main fluid dynamics phenomena that cause the interference are presented. Where possible, an engineering assessment is made of the effect of jet and/or airframe configuration parameters on the observed interference effects. r 2005 Elsevier Ltd. All rights reserved.

Contents 1. 2. 3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . The free jet . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hover out of ground effect . . . . . . . . . . . . . . . . . . 3.1. The effect of jet decay rate . . . . . . . . . . . . . 3.2. The effect of nozzle pressure ratio (NPR) . . . 3.3. The effect of nozzle length and projection . . 3.4. The effect of planform shape and position . . 3.5. The effect of off-axis nozzle positions. . . . . . Transition out of ground effect . . . . . . . . . . . . . . . 4.1. The free jet in a cross-flow . . . . . . . . . . . . . 4.1.1. Correlation parameters . . . . . . . . . . 4.1.2. Jet trajectory . . . . . . . . . . . . . . . . . 4.2. Jet-induced interference effects . . . . . . . . . . 4.2.1. The effect of velocity ratio, V e . . . . 4.2.2. The effect of area ratio, An S . . . . . . 4.2.3. The effect of nozzle geometry . . . . . 4.2.4. The relative location of wing and jet

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

Corresponding author.

E-mail addresses: a.j.saddington@cranfield.ac.uk (A.J. Saddington), k.knowles@cranfield.ac.uk (K. Knowles). 0376-0421/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.paerosci.2005.03.002

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . .

176 176 177 177 179 180 180 181 181 181 181 182 183 184 184 184 185

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

176

qn qx r rn S T u¯ c Ve Vn V1 x z x¯ a; b dn D y k r¯ e rn ra

Nomenclature An dn d ne ¯ D i k1 , k2 L m _n m _x m max M Mn M1 n NPR pa Pc Pertotal

nozzle exit area nozzle diameter equivalent nozzle diameter angular mean planform diameter index (see Eq. (3)) constants in Eqs. (2), (4) and (7) lift constant in Eq. (10) nozzle exit mass flow rate mass flow rate at a streamwise distance x maximum pitching moment nozzle exit Mach number freestream (cross-flow) Mach number constant in Eq. (10) nozzle pressure ratio ambient static pressure settling chamber total pressure total jet perimeter

4.2.5. The effect of nozzle vector angle, dn 4.2.6. Lift improvement devices (LIDs) . . . 4.3. Jet and intake-induced interference effects . . 5. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction For over 40 years engineers have struggled with the concept of giving a jet-powered fixed-wing aircraft the capability to hover and transition to and from wingborne flight. From the tens of designs studied, only about 20 jet-powered STOVL aircraft made it to the flight test stage and of these, only two (the Harrier and Yak-38) became operational, a testimony to the difficulty in designing a successful STOVL aircraft. With development of the Joint Strike Fighter (JSF) nearing completion it is perhaps poignant to reflect on the lessons learnt throughout the development of these aircraft. This paper presents a review of the main out-ofground-effect propulsion-induced aerodynamic interference encountered by jet-lift STOVL aircraft. It will draw from several earlier surveys [1–6] as well as presenting more recent research results.

2. The free jet Many generalised studies of axisymmetric free jets have been made and their characteristics are well

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

nozzle exit dynamic pressure dynamic pressure at a streamwise distance x radius nozzle radius planform area thrust mean jet centreline velocity effective velocity ratio (see Eq. (9)) nozzle exit velocity freestream (cross-flow) velocity streamwise distance from nozzle exit distance perpendicular to nozzle exit ¼ x=rn constants in Eq. (10) nozzle vector angle incremental angular measurement (see Eq. (8)) constant in Eq. 1 ¼ ra =rn nozzle exit static density ambient static density

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

186 186 186 187 187

documented [7–9]. Studies focussing on particular aspects of free jets such as the self-similar region downstream of the nozzle [10–12], the effect of initial conditions [13–16], the effect of a co-flowing stream [17] and large-scale structures within the jet [18,19] have also been made. With regard to STOVL aircraft applications highspeed (in particular under-expanded) free jets are of relevance and these too have been well documented [20–24]. For a circular convergent nozzle, three variations of the flow pattern are possible, depending on the nozzle pressure ratio (NPR, the ratio of settling chamber total pressure, Pc to ambient static pressure, pa ). The three variations are the subsonic jet, the under-expanded jet and the highly under-expanded jet [23]. Note that with a convergent-divergent nozzle an additional flow pattern exists, that of the overexpanded jet. This occurs when the nozzle is operating off-design at too high a NPR [25]. Jet spreading rate and entrainment characteristics have been the subject of a number of studies. Anderson and Johns [20] considered a range of jet exit Mach numbers from 1.4 to 3.53. The jets were found to have a mean cone half angle of 4:6
with this value being independent of NPR. Later studies by Moustafa [24]

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

and Lau et al. [26] indicated that jet spreading rate is in fact strongly influenced by NPR and jet exit Mach number. Nozzle geometry has also been found to influence spreading characteristics with elliptical, rectangular and lobed nozzles all showing greater spreading rates than jets from a circular nozzle [27]. Ricou and Spalding [28] summarise results for the mass entrained by a jet, based on measured velocity profile data and augmented with analytical solutions, showing that for given jet exit conditions air is entrained into a jet at a constant rate. Witze [29] making use of Kleinstein’s [30] theoretical formulation and the results of 13 experimental studies, derived an empirical formulation (Eq. (1)) for the decay in the centreline velocity for varying jet exit Mach numbers. ! 1 u¯ c ðxÞ , (1) ¯ ¼ 1  exp kxð ¯ r¯ e Þ0:5  0:7

177

or wing, and exhaust vertically downwards, a small lift loss is induced by the entrainment action of the jet(s). For a simple circular planform with a single central round jet having a uniform exit velocity distribution (Fig. 2), the reduction in lift is about 2% of the installed thrust T, for a nozzle area to planform area ratio An =S ¼ 0:01 [35,36]. This lift loss reduces steadily, but non-linearly, as the ratio An =S increases, becoming only about 0.5% for An =S ¼ 0:11 and negligible if An =S is greater than 0.25 [35,36]. The lift losses on the plate arise from the entrainment action of the jet, which mixes with the surrounding air and sets up a cross-flow over the bottom of the plate, thereby inducing a suction pressure. The four-jet configuration of Otis [37] indicated that lift losses of between 3 and 4 per cent could be encountered for typical multiple-jet configurations.

3.1. The effect of jet decay rate

where k is a proportionality constant given by k ¼ 0:08ð1  0:16M n Þðr¯ e Þ0:22 k ¼ 0:063ðM 2n  1Þ0:15

ðM n o1Þ,

ðM n 41Þ.

The influence of temperature on the characteristics of free jets has also been widely investigated [29,31–33], particularly with respect to spreading rates [31] and entrainment rates [33].

Gentry and Margason [38] made lift loss measurements using a circular and a rectangular plenum chamber. The latter was designed to fit inside the fuselage of a wind tunnel model and was smaller than ideally desired. A comparison of the loads induced on a circular plate by a single nozzle fitted to the circular and rectangular plenum chambers (Fig. 3(a)), indicated that the loads induced on the circular plate mounted on the

3. Hover out of ground effect The basic flow-field associated with a jet-lift STOVL aircraft hovering out of ground effect is illustrated in Fig. 1 [34]. In all cases, the lifting jets issuing from the aircraft mix with ambient air to generate a complicated three-dimensional flow-field. In general, when one or more jet-lift engines are installed in an aircraft fuselage

∆L

Intake flow Entrained air

Fig. 1. The flow field surrounding a V/STOL aircraft in hover [34].

Fig. 2. The lift loss generated on a flat plate by a single nozzle [35].

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

178

0

Cylindrical plenum, clean nozzle Cylindrical plenum, restriction in nozzle Original rectangular plenum Modified rectangular plenum

∆L/T

-0.01

∆L/T

0

-0.02 Single jet Four jet Eight jet Four slot jet

-0.03

-0.01

-0.04

-0.02

-0.05 0

1

2

3

4

(a)

5 6 √(S/An)

7

8

9

10

-0.03 1.2 1.4

(a)

1.6 1.8 NPR

2.0

2.2

2.4

Cylindrical plenum, clean nozzle Cylindrical plenum, restriction in nozzle Original rectangular plenum Modified rectangular plenum

qx/(Pc -pa)

0.6 0.4

0.0

1.0

0

1

2

3

(b)

0.8

4

5 6 x/dne

7

8

9

10

Fig. 4. (a) Induced load and (b) jet decay for single-jet and several multiple-jet configurations ðNPR ¼ 1:89Þ [38].

0.6 0.4 0.2

(b)

0.8

0.2

1.2

0.0 0

Single jet Four jet Eight jet Four slot jet

1.0 1.2

qx/(Pc -pa)

-0.04 1.0

1

2

3

4

5 6 x/dn

7

8

9

10

Fig. 3. The effect of plenum chamber and nozzle configuration on (a) induced load on several plate shapes and (b) on jet decay ðNPR ¼ 1:64Þ [38].

rectangular plenum were three to four times as large as those induced on the same plate-nozzle arrangement with the circular plenum chamber. Measurements of the velocity profile obtained from the rectangular plenum indicated a distorted velocity distribution with a deficit at the centre, whereas the flow from the circular plenum had a uniform ‘top hat’ velocity profile. The flow through the rectangular plenum was found to be very turbulent and it was suspected that the high turbulence levels were causing the higher induced loads. To confirm the hypothesis, fairings and baffles were added to the rectangular plenum chamber to improve the flow quality and a strut was inserted in the circular plenum to degrade the flow quality. This also produced a much greater distortion of the exit velocity profile than was present on the original rectangular plenum chamber. Fig. 3(a) shows that these changes increased the lift loss for the

circular plenum chamber and greatly reduced the losses for the rectangular plenum chamber. The lift losses did not correlate with the exit velocity distribution, however, they were found to correlate with the rate of decay of the jet with distance downstream from the nozzle exit (Fig. 3(b)). These curves show the peak non-dimensional dynamic pressure in the jet at a non-dimensional distance downstream of the nozzle. The circular plenum chamber gave the smallest lift loss and had the lowest decay rate. The original rectangular plenum produced the greatest lift loss and had the highest decay rate. The modifications made to the circular and rectangular plenum chambers produced jets with similar decay rates, giving similar lift losses. The correlation was also found to hold true for multiple-jet configurations. Fig. 4 compares the lift loss and decay rates for three different jet arrangements (all using the modified rectangular plenum chamber). An empirical equation was developed to predict the lift loss per unit thrust based on the jet decay rate and model geometry [38]. rffiffiffiffiffiffi DL S (2) k 1 k2 , ¼ T An where k1 ¼ 0:009 and vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi u qðq =q Þ u x n u t @ðx=d ne Þ max , k2 ¼  ðx=d ne Þi

(3)

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

3.2. The effect of nozzle pressure ratio (NPR) Gentry and Margason observed that the measured lift losses showed a reduction with increasing NPR for a

0

∆L/T

-0.01 -0.02 Single jet, small scale Four jet, small scale Single jet, J85 Four jet, J85

-0.03 -0.04 -0.05 0

1

2

3

4

(a)

5 6 √(S/An)

7

8

9

10

1.2 Single jet, small scale Four jet, small scale Single jet, J85 Four jet, J85

qx/(Pc -pa)

1.0 0.8 0.6 0.4 0.2 0.0 0

1

2

3

4

(b)

5

6 7 x/dne

8

9 10 11 12

Fig. 5. Comparison of (a) induced load and (b) jet decay for the small-scale results of Ref. [38] and a J85 turbojet [41].

1.2 Circular nozzle (70°F) Circular nozzle (1200°F) Single slot nozzle (70°F) Single slot nozzle (1200°F) Four slot nozzle (70°F) Four slot nozzle (1200°F)

1.0 0.8 qx/(Pc -pa)

where x is the distance along the nozzle axis, the origin of which is at the exit plane. The lift loss is therefore proportional to the square root of the maximum rate of decay of the jet divided by the distance to the point at which this occurs. The correlation showed good agreement with the test data. It is clear that the use of rapid mixing nozzles, if installed on a STOVL aircraft, will have a detrimental effect on the aircraft’s hover performance. These nozzles, which are currently being developed for reduced radar and infra-red signatures [39] and for minimising ground erosion problems, are designed to promote high decay rates of the jet potential core and may employ techniques including: geometric modifications such as lobes, notches or castellations; excrescences such as tabs or vortex generators and fluidic techniques such as swirl or shear layer excitation. An extensive literature review of jet mixing enhancement is given by Wong [40], to which the reader is referred. When multiple jets are dispersed over the planform, the percentage lift loss is greater than for the corresponding single jet of the same equivalent diameter ratio, d ne [35]. This is partly due to the increased mixing rate, but also to the additional pressure drop produced on the lower surface between the jets because of the constricting effect of the jets themselves on the entrained flow. This does not appear, however, to be a strong effect and the lift loss seems unlikely to exceed 5% unless the jets are arranged in rows or elongated narrow slots, thus tending to enclose a significant amount of the planform area. Using a Pratt and Witney J85 engine equipped with two alternative jet pipes; one for a single jet investigation, and one with a special exhaust ducting to produce a four-jet configuration, McLemore [41] ran tests with three different sizes of rectangular plates to produce three different ratios of planform area to jet area ðAn =SÞ. The comparison of these real engine data with the model results of Gentry and Margason [38] shows good agreement (Fig. 5). Using data from Fleming [42] and Higgins et al. [43,44], Kuhn and McKinney [45] (and later Hammond [46]), present a comparison of experiments conducted using jets of different temperature. These suggest that temperature does not appear to have an appreciable effect on the jet decay rate except for a circular nozzle (Fig. 6). The studies on jet decay rate demonstrate that if accurate lift loss data are to be obtained from wind tunnel models then the decay rate of the simulated jets should accurately match those of the full size installation.

179

0.6 0.4 0.2 0.0 0

2

4

6

8 x/dne

10

12

14

16

Fig. 6. The effect of temperature and nozzle configuration on jet decay [45].

constant nozzle to planform area ratio [38]. At the time a theoretical model was not available to explain this. NPR effects were looked at, however, in an extensive study of jet-induced interference effects undertaken by Shumpert and Tibbetts [47]. Tests were conducted on a 14 per cent scale model of a jet VTOL aircraft with a fuselagemounted lift system at pressure ratios of 1.4, 1.7, 2.0 and 2.3. It was discovered that the lift loss did not follow the correlation proposed by Gentry and Margason (Eq. (2)). The results showed that there was a measurable

180

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

discrepancy between the lift losses at different NPRs for a constant nozzle area and a modification to Eq. (2) was proposed. rffiffiffiffiffiffi DL S (4) ðNPRÞ0:64 k1 k2 . ¼ T An The constant k1 was found to be aircraft configuration dependent and for the tests conducted, varied between 0:01 and 0:016. The parameter k2 was unchanged (see Eq. (3)). As this correlation suggests, an increase in NPR is seen to result in a decrease in percentage lift loss. This effect was found to be similar for all configurations. The NPR effect described above is explained by the jet entrainment rate of different NPR jets. At an axial distance, x downstream of the nozzle exit plane, the ratio of mass flow rate in the jet to the mass flow rate at the nozzle exit is given by [28] rffiffiffiffiffi _x m x ra ¼ 0:32 . (5) _n m d n rn Eq. (5) shows that as NPR is increased (and hence the jet density), the mass flow rate of the jet at a given streamwise location is reduced. The decay rate of a high NPR jet will, therefore, be less than that of a lower NPR jet. Comparison with data collected by Schwantes [48] and reported by Ransom and Smy [49] suggests that although Eq. (5) gives the correct trend for jet entrainment, the magnitude calculated may not be relied upon. Shumpert and Tibbetts also conducted experiments at constant thrust, i.e. the nozzle area was reduced as the NPR was increased. Varying NPR at constant thrust had no measurable effect on lift loss [47]. A more direct, easier to use method for estimating hover lift losses is presented by Kotansky [50]. Correlation of data from various single and multiple jet configurations [51] resulted in the following equation: rffiffiffiffiffiffi DL S ¼  0:0002528 T An

 Pertotal 1:581 . ð6Þ  ðNPRÞ0:64 d ne Eq. (6) implicitly accounts for the higher decay rate of multiple jet configurations in terms of equivalent nozzle diameter but does not account for higher decay rates caused by jet exit conditions involving high entrainment rates i.e. rapid mixing nozzles. If higher than normal turbulence and decay rates are involved, Kotansky suggests that Eq. (4) should be used [50]. 3.3. The effect of nozzle length and projection Gentry and Margason [38] investigated the effect of nozzle length and projection on induced loads for a single round jet with the modified rectangular plenum

chamber described in Section 3.1 and with rectangularplanform plates. In these experiments, nozzle length was defined as the distance from the bottom of the settling chamber (the start of the nozzle profile) to the nozzle exit plane. Nozzle projection, by contrast, was the distance the nozzle exit plane extended below the planform surface. The findings are summarised in Table 1 for NPRs of 1.5 and 2.0. Extending the nozzle length caused a slight reduction in the lift loss for the case of zero projection. Measurements of the jet decay rate for the different nozzle lengths showed that the shortest nozzle had the highest decay rate. A further reduction in lift loss was obtained by projecting the nozzle through the circular plate distances of 0:5d n and 1:0d n . This reduction is due to the increased distance between the plate and the free surface of the jet which is entraining air. 3.4. The effect of planform shape and position In order to determine the effect of planform shape on the induced lift loss, Gentry and Margason tested four plate shapes on their circular plenum chamber [38]. The four shapes were circular, square, rectangular (aspect ratio 1.52) and triangular (equilateral). In all cases the ratio of nozzle exit area to plate area, An =S was 0.0144 and the centre of the nozzle was located at the centroid of the plate. There was no measurable effect of planform shape on the lift loss. The effect of wing height was also investigated. As would be expected, a low wing configuration had more lift loss than a mid-mounted wing which in turn performed worse than a high wing. This is due to the relative proximity of the wing to the mixing region of the jet. The effect of planform lip shape was investigated by Ing and Zhang [52]. An experiment was carried out to study the parameters affecting single-jet ground-effect hover lift loss and to identify the cause behind the large discrepancies in lift loss between the experiments of Wyatt [53] and Corsiglia et al. [54], commonly known as

Table 1 Lift loss for various nozzle lengths and projections, An =S ¼ 0:0105 [38] Nozzle length ðd n Þ

Nozzle projection ðd n Þ

DL=T ðNPR ¼ 1:5Þ

DL=T ðNPR ¼ 2:0Þ

0.248 0.910 1.410 0.910 1.410 1.410

0.0 0.0 0.0 0.5 0.5 1.0

0.0187 0.0163 0.0163 0.0130 0.0133 0.0116

0.0175 0.0155 0.0148 0.0129 0.0126 0.0107

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

the ‘Wyatt anomaly’. The cause of the discrepancies was traced to a single geometrical parameter, namely the planform plate edge geometry, which significantly affected the flow separation underneath the plate. The out of ground effect lift loss was, however, found to be insensitive to the edge geometry.

Viscous effects on wing and control surfaces

Intake mass flow

3.5. The effect of off-axis nozzle positions The effect of positioning a nozzle away from the fuselage centre-line was part of a study carried out by Welte [55]. Tests on a model of the VTOL transporter, Dornier Do-31, indicated that a centrally located jet will produce more lift loss than a peripherally located jet. The podded lift engines installed at the wing tips induced a lift loss of 2.2 per cent compared with 3.6 per cent for the lift/cruise engine installed below the inner wing section. Welte postulated that the constant k1 in Eq. (3) is universal and an additional term could be introduced into the equation to take into account configuration effects. The following equation was developed [55]. ¯ D DL ¼ ðNPRÞ0:64 k1 k2 , d ne T

181

(7)

¯ is the angular mean planform diameter and is where D given by  Z 2p  d ne ¯ ¼1 dy, (8) rðyÞ  D p 0 2 where k1 ¼ 0:009 and k2 is unchanged from Eq. (3). It was found that for the lift/cruise engines, both Eqs. (4) and (7) predicted the lift loss quite well, 4.0 and 3.4 per cent, respectively. With the peripherally located lift engines at the wing tip, Eq. (4) failed to produce a satisfactory result, again predicting 4.0 per cent (as would be expected) compared with 2.4 per cent for Eq. (7).

4. Transition out of ground effect The transition from hover to wing-borne flight is a very important design consideration for STOVL aircraft. Fig. 7 shows the principle aerodynamic phenomena affecting a STOVL aircraft in transition out of ground effect [56]. During the transition from hovering to conventional flight, the efflux from the lifting jet(s) is swept back by the freestream and rolled up into pairs of vortices. These vortices, along with the entrainment action and blockage effect of the jet(s), induce suction pressures on the bottom of the configuration beside and behind the jet(s) and a smaller region of positive pressure ahead of the jet(s). In most cases, these induced pressures result in a nose-up pitching moment and a loss of lift on the aircraft.

Jet blockage and entrainment

Fig. 7. The flow field about a STOVL aircraft in transition out of ground effect [56].

4.1. The free jet in a cross-flow The jet in cross-flow has received considerable research attention because it is a fundamental fluid dynamic phenomena with many applications such as smoke plumes from chimneys, the dispersal of liquids in streams, jet engine combustors and reaction control jets on rockets and missiles. The most widely studied, and reviewed, application, however, is related to STOVL aircraft [57,58]. A jet exhausting into a cross-flow generates a complex flow-field with several distinguishable features. When the jet efflux exits the nozzle it is deflected by the freestream to follow a curved path downstream while its cross-section changes. For the case of a circular jet, there are stagnation points upstream and downstream and minimum pressures at the lateral edges. As a consequence, the jet spreads laterally into an oval shape. At the same time the cross-flow shears the jet fluid along the lateral edges downstream to form a kidney-shaped cross-section. At increasing distances along the jet path this shearing folds the downstream face over itself to form a vortex pair which dominates the flow. Associated with the vortex pair is the flow induced into the wake region of the jet from the freestream. Fig. 8 shows the topology of a jet in cross-flow [58]. 4.1.1. Correlation parameters Early jet in cross-flow investigations usually characterised the test condition by the ratio of freestream velocity to jet exit velocity, V 1 =V j . It was soon recognised, however, that this ratio was not appropriate for hot jets. Williams and Wood observed that nondimensional increments in lift, drag and pitching moment were primarily, but not solely, a function of jet to freestream velocity ratio [35]. The influence of jet Reynolds number (based on nozzle diameter) and model Reynolds number (based on wing chord) on jet

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

182

interference effects were secondary. They showed that the influence of temperature and compressibility effects could better be accounted for by the use of an effective velocity ratio (Eq. (9)), which is still widely used. Ve ¼

V1 Vn

rffiffiffiffiffi rffiffiffiffiffi ra M 1 pa ¼ . rn M n pn

(9)

For STOVL applications the magnitude of V e for transition to wing-borne flight depends on the pressure ratio of the lifting propulsion device. For high-pressure ratio jet engines, transition will occur with V e o0:2; for the Harrier, transition will occur with V e o0:3; and for lift fans with low pressure ratios, transition will occur with V e o0:5 [58].

Jet CL V

Vortex pair Horseshoe vortex Wake vortex street

Vortex curve

Separation line Fig. 8. Sketch of the vortex systems associated with a jet in cross-flow [58].

4.1.2. Jet trajectory One property of jets in cross-flow which has received a lot of research attention is the jet trajectory [59–63]. The most common parameter for defining the trajectory is the locus of maximum velocity; however, other definitions are sometimes used depending on the preference of the researcher. These trajectory definitions include: the locus of maximum stagnation pressure; the locus of maximum dynamic pressure; the locus of maximum stagnation temperature and the line of maximum vorticity. Since none of these definitions will give the same trajectory, it is necessary to exercise great care when comparing data from different sources. Nearly all the work done to establish a jet trajectory has used a subsonic jet in a subsonic cross-flow. A widely used empirical equation for the jet centreline trajectory is [58],

m x z z ¼ aV ne þ b cotðdn Þ. (10) dn dn dn The z and x origins are taken as the centre of the nozzle exit. When the nozzle vector angle is 90
(i.e. perpendicular to the cross-flow), the second term in Eq. (10) becomes zero. The empirical approaches correlate the experimental data using similarity laws to obtain the jet trajectory. Table 2 summarises the coefficients and exponents for jets with a 90
vector angle [64–67] and for jets with varying vector angles [68–70]. Eq. (10) has been plotted using an effective velocity ratio, V e of 0.25 which covers the range of applicability for most of the researchers. Fig. 9 shows the jet trajectories for a nozzle vector angle of 90
while Fig. 10 shows the trajectories for a vector angle of 60
. The equations derived by the researchers show good agreement with each other. There are slight variations in the initial depth of penetration and the rate of jet deflection. These can probably be accounted for by variations in the initial jet and cross-flow conditions in the experiments from which the empirical equations are derived. In particular the turbulence intensities of the jet and cross-flow, which will affect the mixing rate between the two, will have a strong influence on the jet

Table 2 Jet trajectory equation terms and applicability Author(s)

a

n

m

b

Ve

dj (deg)

Bradbury [64] Kamotani and Greber [65] Chaissaing et al. [66] Fearn and Weston [67] Ivanov [68] Shandorov [69] Margason [70]

1.08 1.38 ð0:59 þ V e Þ2:6 1.08 1.00 1.00

2.70 2.61 2.60 2.68 2.60 2.00 2.00

3.00 2.78 2.60 2.95 3.00 2.55 3.00

n/a n/a n/a n/a 1.00 ð1 þ V e Þ2 1.00

0.08–0.43 0.13–0.26 0.16—0.42 0.10–0.33 0.03–0.71 0.21–0.71 0.10–0.85

90 90 90 90 60–120 45–90 30–180

1=4 sin2 dj

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

z/dn

10

5 Kamotani and Greber [65] Chaissang et.al [66] Fearn and Weston [67] Bradbury [64]

0 0

5

10 x/dn

15

20

Fig. 9. Comparison of predicted jet in cross-flow trajectories ðdn ¼ 90
; V e ¼ 0:25Þ.

z/dn

10

183

thrust. With forward velocity the wing also develops lift and, in the absence of interference effects, lift from these two sources could be added together to produce the solid curve. The jet-induced effects, however, cause a loss of lift and the actual lift measured in a wind tunnel is shown by the symbols. The difference between the calculated and measured curves is the interference lift loss, which is generally independent of angle of attack. The rolling up of the jet wakes into contra-rotating vortices is the primary cause of the interference effects in transitional flight [83]. The vortices change the flow field in the near-field and induce additional suction pressures on the lower surface of the fuselage and wing of the aircraft. The interaction between the jet and the cross-flow means that positive pressures are generated on surfaces ahead of the jet and negative pressures behind [84]. Spreeman [85] measured negative pressures as high as 3 to 4 times the freestream dynamic pressure. The pressures diminished with distance from the jet but extended 10 to 15 jet diameters downstream and 5 to 10

5 1.5

Ivanov [68] Shandorov [69] Margason [70]

0

0

5

10 x/dn

15

Calculated Measured

Power-off lift

20 1.0

Jet lift

L/T

Interference lift loss (∆L)

Fig. 10. Comparison of predicted jet in cross-flow trajectories ðdj ¼ 60
; V e ¼ 0:25Þ.

0.5

0.0 -10

0

(a)

10 α (deg.)

20

30

Measured power-on moment Interference moment (∆M)

M

penetration and curvature. Zhang and Hurst [71] have observed that with supersonic jets, where the turbulence intensity in the jet may be high (due to the breakdown of the shock structure), the jet centreline trajectory will ultimately lead to less penetration than for a subsonic jet. Integral methods, e.g. [72–77] for jet in cross-flow research have also been developed which consider the deflecting mechanisms of a jet by taking into account either a pressure force or cross-flow momentum entrainment or both. More recently the use of computational fluid dynamics (CFD) has given researchers an addition numerical tool, e.g. [78–82].

0 Power-off moment

4.2. Jet-induced interference effects In general, the jet-induced interference effects in transition are characterised by a loss in lift and a noseup pitching moment. Fig. 11(a) illustrates typical trends for a tail-off configuration [83] and shows the lift divided by the thrust as a function of angle of attack. Fig. 11(b) shows pitching moment as a function of angle of attack. In hover, the jets produce a lift which is equal to the net

-10 (b)

0

10 α (deg.)

20

30

Fig. 11. Jet interference in transition flight (tail off) [83]. (a) Lift, (b) pitching moment.

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

4.2.1. The effect of velocity ratio, V e The lift loss on a typical STOVL aircraft in hover has already been discussed and is of the order of 2 to 3 per cent of the installed thrust. Early experiments conducted by Williams and Wood [35] on a simple rectangularwing model with a centrally located jet established that the lift increment, DL=T falls steadily below its static value as the effective velocity ratio V e is raised from zero. The lift loss was accompanied by the expected nose-up pitching moments, again increasing as V e was raised. The observed interference effects are due partly to the steady growth of the downward load from jet interference on the lower surface behind the jet and partly from the rearward movement of the centre of this jet interference load. Additionally, the normal component of thrust will be reduced because of the jet deflection. As the velocity ratio is increased further, the lift loss reduces but the nose-up pitching moment continues to increase. This may be due to two main factors; firstly, the configuration being tested may be generating aerodynamic lift which will offset the lift loss, and secondly, the very high jet deflection angles involved may put the jet interference close to the trailing edge of the wing. The jets may then act as a crude jet-flap on the wing lower surface and thereby generate extra lift by super-circulation. For practical STOVL aircraft, effective velocity ratios below about 0.3 are of primary concern during transition with jet-lift schemes although values as high as 0.5 could be of importance with high by-pass ratio engines or lift fan arrangements. Fig. 12 shows typical results for a four-jet configuration [37]. 4.2.2. The effect of area ratio, An =S The interference lift loss is strongly sensitive to the ratio of nozzle area to planform area, An =S. With the configuration mentioned above, Williams and Wood observed that the lift loss increases as the area ratio is decreased and the relationship is approximately linear for a simple wing/jet geometry [35]. A larger planform provides more area over which the interference pressures can act thereby increasing the lift loss. At very low area ratios, An =So0:0016, L=T actually fell to zero. For more practical area ratios (between 0.01 and 0.04 for a

0.4

∆M/Tdne

diameters to each side of the jet [85]. The combination of positive pressures ahead of the jet and negative pressures behind gives a nose-up pitching moment. The following sections aim to describe parameters which have been observed to influence the jet-induced interference effects. Due to the complex nature of the flow fields surrounding STOVL aircraft in transition, it is difficult to isolate the effects which different parameters may have on any particular configuration and as such only general trends can be identified.

0.2

0 0

0.1

0.2

0.3

0.2

0.3

Ve

(a) 0

∆L/T

184

-0.2

-0.4 0 (b)

0.1 Ve

Fig. 12. The effect of wing planform on lift and pitching moment ðS=An ¼ 38Þ [37]. (a) Pitching moment, (b) lift.

typical STOVL aircraft) the lift loss and accompanying nose-up moments are less severe, a maximum lift loss of 25 per cent was recorded with an area ratio of 0.01. 4.2.3. The effect of nozzle geometry Extensive studies of the effect of nozzle geometry were carried out by Volger [86]. A VTOL model was equipped with various arrangements of interchangeable single or multiple, round or slotted jets, with and without jet deflection. All configurations showed interference lift losses that increased with velocity ratio. For most jet arrangements, nose-up pitching moments due to jet interference occurred and increased with velocity ratio. The configurations which suffered the least interference effect on the lift and pitching moment were the single longitudinal slot jets, which was thought to be due to their more streamlined shape in cross-section and the smaller planform area behind the jet. An ejector powered STOVL model, equipped with various arrangements of slot nozzles, was tested by Volger and Goodson [87]. The aerodynamic lift of the fuselage, which was almost square in cross-section, at 0
angle of attack was zero. As velocity ratio was increased

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

ratio 6, unswept, untapered wing-fuselage model equipped with a 30 per cent chord slotted Fowler flap was used. Two jets, one on either side of the fuselage, were positioned at about 25 per cent semi-span and at the various longitudinal and vertical positions shown by the ‘plus’ marks in Fig. 13. The jets were mounted independently of the wing so that only the aerodynamic forces and interference effects were measured on the wing. The data show that with the jet exits on the wing chord plane, considerable jet interference was experienced even with the jet as far as four wing chords ahead of the wing. Favourable interference effects, however, were encountered with the jets beneath and behind the 50 per cent chord point of the wing and the interference effects are most favourable for positions closest to the flap. These favourable interference increments were believed to be due to the action of the jet in helping the flap achieve its full lift potential. Another slightly different configuration [88], this time with jets operating simultaneously both in front of and behind the wing, indicated an overall favourable interference lift effect, again indicating the importance of configuration geometry on the jet interference lift and moment characteristics. These findings agree well with those of Margason and Gentry [112], Schultz and Viehweger [89] and Nangia [90]. Winston et al. [91] report on a configuration which was tested to determine the optimum jet exit location from an induced lift standpoint. The model was tested with the lift/cruise jets in three longitudinal positions as shown in Fig. 14 and illustrate the lift improvement obtained by successive aft movement of the jet exits towards the wing trailing edge. The Harrier II, with its new wing and large slotted trailing edge flap, initially suffered from interference problems whereby the forward nozzle flow interacted

from 0 (hover) to 0.4, all configurations showed a reduction in the combined aerodynamic lift plus thrust. Moving the slots outward reduced the affected model area directly behind the jets, thereby reducing the interference effects between the model, jets and freestream. Yawing the slots either inward or outward at the forward end resulted in a reduction in lift and an increase in nose-up pitching moments compared with the unyawed slots. Yawing the slots outward was less detrimental than inward. The effect of splaying the jet 20
away from the plane of symmetry produced some lift improvement and a more nose-down pitching moment as a result of less interference between the freestream and the outward spreading jets. Any indicated benefit from splaying the jet would, however, be partially offset by the 6 per cent lift loss resulting from the inclination of the thrust axis. A combination of 10
outward yaw of the slots and 10
outward splay gave less lift loss than configurations with or without angular deflections. Yaw appeared to affect the results more than inclination. 4.2.4. The relative location of wing and jet As regards jet position relative to the wing, rearward location of the jet exit naturally tends to alleviate the lift loss and the nose-up pitching moments arising from jet interference, since the surface extent aft of the jet exit is reduced. Williams and Wood found evidence to suggest that with multiple jets the velocity ratio range over which the lift loss occurs can be much reduced and the subsequent lift augmentation much increased by the adoption of a span-wise row arrangement towards the rear of the wing instead of a close cluster [35]. A generalised study of jet positions from several wing chord lengths ahead, to several wing chord lengths behind, an unswept wing was reported by Hammond and McLemore [88]. In this investigation, an aspect

+

+

+ + + + + +

δn = 90° Ve = 0.18

+ + ++++ + + + +++ + ++ + + + +

0.2

∆L/T

0 -0.2 -0.4 -0.6

-4

-3

-2

185

-1 0 x/chord

1

2

3

Fig. 13. The effect of varying chordwise jet location relative to a nearby wing on the induced lift [88].

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

186

3

of the trailing vortex flows. This was consistent with the apparent negligible effect of the fences at low velocity ratios, where the rate of deflection of the jet is inherently small. The effect of the fences on a simple fuselage model was excellent, reducing maximum transitional lift loss from 27% to 7%.

Calculated (no interference)

L/T

2

4.3. Jet and intake-induced interference effects 1

0 0

0.1

0.2

0.3

0.4

0.5

Ve Fig. 14. The effect of chordwise location of deflected lift/cruise jets relative to the wing trailing edge on the total lift for a supersonic combat lift plus lift/cruise V/STOL aircraft configuration [91].

with the wing undersurface and inboard pylon at transitional forward speeds causing a flow separation upstream of the flap. Removal of the scarf directed the forward jet to a position further below the wing and produced a lift coefficient increment of approximately 0.2 [50]. 4.2.5. The effect of nozzle vector angle, dn Wind tunnel tests carried out on a one-sixth scale model of the Kestrel (XV-6A) measured the effect of vectoring the nozzles on the incremental lift and pitching moment at a range of velocity ratios and angles of attack [92]. At 0
angle of attack, increasing nozzle vector angle to a more vertical position, increased the lift loss and nose-up pitching moment. This correlates well with similar works [93,94]. 4.2.6. Lift improvement devices (LIDs) Spreeman [85] noted that in one full-scale flight investigation, deflecting the trailing-edge flaps reduced the losses in lift and nose-up pitching moments. The beneficial effect of the flaps on this configuration can be attributed to positive pressures being built up in front of the flap on the lower surface. The alleviation of unfavourable jet interference effects was shown possible by Williams and Wood with the addition of stream-wise fences to the lower surface of the airframe, along the sides of the jet exit [35]. The geometry of the fence was found to be important. The depth of the fence should be at least one jet diameter and the stream-wise length at least two jet diameters so as to extend forward of the jet exit as well as bounding the sides. Flow visualisation suggested that the fences reduced the initial curvature of the jet, thus increasing the penetration of the jet plume and delaying the growth

During wind tunnel testing of jet-lift STOVL aircraft, it is usual to simulate the jet efflux but not the intake flows. The intakes, which are commonly faired over or are unpowered, are generally tested in separate wind tunnel experiments. This may lead to a lack of integration between intake [95,96] and nozzle design [97]. Separate jet and intake testing ignores any mutual interferences between them and airframe, but is adopted due to the complexity of building a fully powered, metric STOVL model [49,98]. This is in addition to the complex tunnel interference problems encountered during transition testing [99] necessitating careful planning of test procedures [100,101]. Early research using simple wind tunnel models showed only minor interference effects due to intake flows [45,112,102,103], which were easily accounted for. This lead researchers to conclude that when testing models in the transition phase of flight the intake interference effects were generally not sufficiently important to warrant modelling the intake flows [64]. Nevertheless some research projects have shown significant jet/intake interference effects. An investigation carried out by Mineck and Margason [103], and Mineck and Schwendemann [104] determined some interference effects between the jets and intakes on a STOVL aircraft. Comparison of the force and moment data for the intake open and intake closed (using elliptical plugs) configurations, indicated that closing the intakes decreased the aerodynamic angle of attack by approximately 2
. Data for the VAK 191B described by Haftmann [105] showed a consistent discrepancy between the lift loss measured in flight and the lift loss measured on a wind tunnel model. The full-size aircraft had a transitional lift loss 4–10% greater than the wind tunnel model and a nose-up pitching moment 5–8% higher. This was probably due, in part, to the use of a free flow intake on the model with no suction simulated. As more complex STOVL aircraft designs were developed, it became increasingly apparent that the argument for separate jet and intake testing on STOVL aircraft was not valid. The term ‘close-coupled aerodynamics’ has become increasingly popular in describing the engine/airframe designs and operating conditions in which the air intake and nozzle system flows create significant interferences with respect to each other and/ or which, when combined, create interferences to airframe surfaces differing from the sum of the

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

0

-0.02 ∆Cl (root)

individual interferences. Harris et al. [106] suggest that aircraft designs, in which the distance from the intake face to the front nozzle relative to the diameter of the compressor face is less than 3, will undoubtedly feature close-coupled aerodynamics. Whereas for designs in which this ratio is greater than 8 the aerodynamics can generally be considered to be uncoupled. Recent research has concluded that during low speed transition ðV e o0:1Þ mutual jet/intake interference effects do exist for close-coupled propulsion systems [107–109]. A generic jet-lift STOVL aircraft equipped with powered intakes was tested in the RMCS low-speed wind tunnel (Fig. 15). The model was not fully metric and so airloads were inferred from static pressure tappings on the wing. The model was tested with jet and intakes powered separately and simultaneously— the difference between the airloads giving an indication of the mutual interference effect. This is shown in Fig. 16, which illustrates the mutual interference as an increment of wing root sectional lift coefficient for the rearmost nozzle position. One would expect that the solution to the complexities of building a STOVL wind tunnel model would be to use computational fluid dynamics (CFD) tools. The combination of viscous interactions, broad range of flow velocities and extensive regions of separated flow has meant that CFD methods tried so far have been unable to adequately predict overall airframe forces and moments [110]. Although CFD is undoubtedly useful in some aspects of STOVL aircraft development, designers are still reliant on exhaustive wind tunnel tests backed up by large historical databases. Even in the 21st Century semi-empirical correlations are still being developed for propulsion-induced interference effects [111].

187

-0.04 NPR=1.586 NPR=2 NPR=3 NPR=4

-0.06

-0.08 0

0.02

0.04

0.06 Ve

0.08

0.1

0.12

Fig. 16. Mutual interference between propulsive jet and intake flows shown as an incremental change in wing root sectional lift coefficient ðdn ¼ 60
Þ.

5. Conclusions This paper has provided a review of the out-ofground-effect propulsion-induced interference on the aerodynamics of jet-lift short take-off and vertical landing (STOVL) aircraft. Two main operational environments encountered by STOVL aircraft have been discussed: hovering out of ground effect and transition out of ground effect. The former has included the fluid mechanics of an isolated free jet and the interference effects caused by such a jet when combined with simple and more complex fuselage/wing arrangements. On the subject of transition out of ground effect, the discussion has followed a similar format. Where possible, an engineering assessment is made of the effect of jet and/or airframe configuration parameters on the observed interference effects. Although some parametric trends are observed in simplified experiments, jet-induced interference effects for a complete airframe are difficult to predict and must therefore be assessed through rigourous tests of the chosen configuration. The evaluation process is still predominantly experimental since computational fluid dynamics models of complete STOVL aircraft do not yet exist with sufficient fidelity to provide a reliable prediction of overall airframe forces and moments.

References

Fig. 15. Front view of the jet/intake model in the RMCS 1:5 m  1:1 m open-jet wind tunnel: intake suction pipe is seen behind and to the right of the model; compressed air supply is via the light-coloured pipe above the fuselage.

[1] Margason RJ. Review of propulsion-induced effects on aerodynamics of jet/STOL aircraft, Technical Note D-5617, NASA, 1970. [2] Barche J. Jet-lift problems of V/STOL aircraft. In: AGARD conference proceedings CP-143, fluid dynamics panel symposium, Delft, Netherlands, 1974. p. 16.1–18.

188

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

[3] Knott PG, Hargreaves JJ. A review of the lifting characteristics of some jet-lift V/STOL configurations. In: AGARD conference proceedings CP-143, fluid dynamics panel symposium, Delft, Netherlands, 1974. p. 24.1–12. [4] Leyland DC. Aerodynamics of V/STOL aircraft—performance assessment. AGARD Report R-710, 1984. p. 11.1–26. [5] Kuhn RE. The induced aerodynamics of jet and fan powered V/STOL aircraft. In: Krothapalli A, Smith CA, editors. Recent advances in aerodynamics, New York : Springer; 1986. p. 337–73. [6] Margason RJ. Propulsion-induced effects caused by outof-ground effects. In: Proceedings of the international powered lift conference, Santa Clara, CA, USA, 1987. p. 31–57, SAE 872307. [7] Abramovich GN. The theory of turbulent jets. Cambridge: MIT Press; 1963. [8] Schlichting H. Boundary-layer theory, 6th ed. New York: McGraw-Hill Book Co.; 1968. [9] Rajaratnam N. Turbulent jets. Amsterdam: Elsevier; 1976. [10] Gelb GH, Martin WA. An experimental investigation of the flow field about a subsonic jet exhausting into a quiescent and a low velocity air stream. Can Aeronaut Space J 1966;12:333–42. [11] Wygnanski I, Fiedler H. Some measurements in the selfpreserving jet. J Fluid Mech 1969;38(3):577–612. [12] Agrawal A, Prasad A. Properties of vortices in the selfsimilar turbulent jet. Exp Fluids 2002;33(3):565–77. [13] Oosthuizen PH. An experimental study of low Reynolds number turbulent circular jet flow. In: Proceedings of the ASME applied mechanics, bioengineering and fluids engineering conference, Houston, TX, USA, 1983, ASME 83-FE-36. [14] Otu¨gen MV, Kim S, Vradis GC. The near field structure of initially asymmetric jets. In: Proceedings of the 34th AIAA aerospace sciences meeting and exhibit, Reno, NV, USA, 1996, AIAA 96-0202. [15] Antonia RA, Zhao Q. Effect of initial conditions on a circular jet. Exp Fluids 2001;31(3):319–23. [16] Xu G, Antonia RA. Effect of different initial conditions on a turbulent round free jet. Exp Fluids 2002;33(5): 677–83. [17] Antonia RA, Bilger RW. An experimental investigation of an axisymmetric jet in a co-flowing air stream. J Fluid Mech 1973;61:805–22. [18] Champagne FH, Crow SC. Orderly structure in jet turbulence. J Fluid Mech 1971;48:547–91. [19] Dimotakis PE, Maike-Lye RC, Papantoniou DA. Structure and dynamics of round turbulent jets. Physics of Fluids 1983;26:3185–92. [20] Anderson AR, Johns FR. Characteristics of free supersonic jet exhausting into quiescent air. J Jet Propulsion 1955;25(1):13–5. [21] Love ES, Grigsby CE, Lee LP, Woodling MJ. Experimental and theoretical studies of axisymmetric free jets. Technical Report R-6, NASA, 1958. [22] Adamson Jr TC, Nicholls JA. On the structure of jets from highly underexpanded nozzles into still air. J Aerospace Sci 1959;26(16):16–24.

[23] Donaldson CD, Snedeker RS. A study of free jet impingement. Part 1. Mean properties of free and impinging jets. J Fluid Mech 1971;45(2):281–319. [24] Moustafa GH. Further study of high speed single free jets. Aeronaut J 1993;97(965):171–6. [25] Anderson Jr JD. Modern compressible flow. McGraw Hill: London; 1990. [26] Lau JC, Morris PJ, Fisher MJ. Measurements in subsonic and supersonic free jets using a laser velocimeter. J Fluid Mech 1979;93:1–27. [27] Zaman KBMQ. Spreading characteristics of compressible jets from nozzles of various geometries. J Fluid Mech 1999;383:197–228. [28] Ricou FP, Spalding DB. Measurements of entrainment by axisymmetrical turbulent jets. J Fluid Mech 1961; 11(1):21–32. [29] Witze PO. Centreline velocity decay of compressible free jets. AIAA J 1974;12(4):417–8. [30] Kleinstein G. Mixing in turbulent axially symmetric free jets. J Spacecraft Rockets 1964;1:403–8. [31] Lau JC. Effects of exit Mach number and temperature on mean flow and turbulence characteristics in round jets. J Fluid Mech 1981;105:193–218. [32] Lepicovsky J. Total temperature effects on centreline Mach number characteristics of free jets. AIAA J 1989; 28(1):478–82. [33] Curtis P. Investigation into the behaviour of a single jet in free air and impinging perpendicularly on the ground. Technical Report BAe-KAD-R-RES-3349, British Aerospace, 1987. [34] Milford CM. Hot gas ingestion in V/STOL. In: Proceedings of the international powered lift conference, Santa Clara, CA, USA, 1987. p. 17–29. [35] Williams J, Wood MN, Aerodynamic interference effects with jet-lift V/STOL aircraft under static and forward speed conditions. Technical Report 66403, RAE 1966. [36] Wood MN, Jet V/STOL aircraft aerodynamics. In: Proceedings of the international congress on subsonic aerodynamics, New York, NY, USA, 1967. p. 893–920. [37] Otis Jr JH. Induced interference effects on a four-jet VTOL configuration with various wing planforms in the transition speed range. Technical Note D-1400, NASA, 1962. [38] Gentry Jr GL, Margason RJ. Jet-induced lift losses on VTOL configurations hovering in and out of ground effect. Technical Note D-3166, NASA, 1966. [39] Rao GA, Mahulikar SP. Integrated review of stealth technology and its role in airpower. Aeronaut J 2002; 106(1066):629–41. [40] Wong RYT. Enhancement of supersonic jet mixing. Ph.D. thesis, Cranfield University, RMCS, Shrivenham, UK, 2000. [41] McLemore HC. Jet-induced lift loss of jet VTOL configurations in hovering condition. Technical Note D-3435, NASA, 1966. [42] Fleming WA. Characteristics of a hot jet discharge from a jet propulsion engine. Research Memorandum E6L27a, NACA, 1946. [43] Higgins CC, Wainwright TW. Dynamic pressure and thrust characteristics of cold jets discharging from several

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191

[44]

[45]

[46]

[47]

[48]

[49]

[50] [51]

[52]

[53]

[54]

[55]

[56]

[57] [58]

[59] [60]

exhaust nozzles designed for VTOL downwash suppression. Technical Note D-2263, NASA, 1964. Higgins CC, Kelly DP, Wainwright TW. Exhaust jet wake and thrust characteristics of several nozzles designed for VTOL downwash suppression—tests in and out of ground effect with 70
F and 1200
F nozzle discharge temperatures, Contractor Report CR-373, NASA, 1966. Kuhn RE, McKinney Jr MO. NASA research on the aerodynamics of jet VTOL engine installations. In: AGARDograph 103, Specialists Meeting, Tennessee, TN, USA, 1965. p. 689–713. Hammond AD. Thrust losses in hovering for jet VTOL aircraft. In: NASA special proceedings, SP-116, 1966. p. 163–75. Shumpert PK, Tibbetts JG. Model tests of jet-induced lift effects on a VTOL aircraft in hover. Contractor Report 1297, NASA, 1969. Schwantes E. The propulsion jet of a VTOL aircraft. In: AGARD conference proceedings CP-91, propulsion and energetics panel symposium, Sandefjord, Norway, 1971. p. 15.1–10. Ransom ECP, Smy JR. Introduction and review of some jet interference phenomena relevant to V/STOL aircraft. In: AGARD Report R-710, 1984. p. 2.1–23. Kotansky DR. Jet flow-fields. In: AGARD Report R-710, 1984. p. 7.1–48. Kotansky DR, Durando NA, Bristow DR, Saunders PW. Multi jet induced forces and moments on VTOL aircraft hovering in and out of ground effect. NADC Report 77-229-30, Naval Air Development Centre, June 1977. Ing DN, Zhang X. An experimental and computational investigation of ground effect lift loss for single jet impingement. Aeronaut J 1994;98(974):127–36. Wyatt LA. Static tests of ground effect on planforms fitted with a centrally-located round lifting jet. Aeronaut Research Council C.P. No. 749, Ministry of Aviation, 1964. Corsiglia VR, Wardell DA, Kuhn RE. Small-scale experiments in STOVL ground effects. In: Proceedings of the international powered lift conference, London, UK, 1990. p. III.14.1–13. Welte D. Prediction of aerodynamic interference effects with jet-lift and fan-lift VTOL aircraft. In: AGARD conference proceedings CP-143, fluid dynamics panel symposium, Delft, Netherlands, 1974. p. 23.1–9. Agarwal RK. Recent advances in prediction methods for jet-induced effects on V/STOL aircraft. In: Krothapalli A, Smith CA, editors. Recent advances in aerodynamics, New York : Springer; 1986. p. 471–521. Hancock GJ. A review of the aerodynamics of a jet in a cross-flow. Aeronaut J 1987;91:201–13. Margason RJ. Fifty years of jet in cross-flow research. In: AGARD conference proceedings CP-534, fluid dynamics panel symposium, Winchester, UK, 1993. p. 1.1–41. Keffer JF, Baines WD. The round turbulent jet in a crosswind. J Fluid Mech 1962;1(31):481–96. Shaw CS, Margason RJ. An experimental investigation of a highly underexpanded sonic jet ejecting from a flat plate

[61] [62]

[63]

[64]

[65] [66]

[67] [68]

[69] [70]

[71]

[72] [73]

[74]

[75]

[76]

[77]

[78]

189

into a subsonic cross-flow, Technical Note D-7314, NASA, 1973. Stoy RL, Ben-Haim Y. Turbulent jets in a confined crossflow. Trans ASME J Fluids Eng 1973; 551–6. Ransom ECP, Wood PM. A literature review on jets in cross-flow. In: AGARD conference proceedings CP-143, fluid dynamics panel symposium, Delft, Netherlands, 1974. p. 26.1–7. Taylor P, Watkins DJ. An investigation of inclined jets in a cross-wind. In: AGARD conference proceedings CP-308, fluid dynamics panel symposium, Fundac- ao Calouste Gulbenkian, Lisbon, Portugal, 1981. p. 6.1–4. Bradbury LJS. Some aspects of jet dynamics and their implications for VTOL research. In: AGARD conference proceedings CP-308, fluid dynamics panel symposium, Fundac- ao Calouste Gulbenkian, Lisbon, Portugal, 1981. p. 1.1–26. Kamotani Y, Greber I. Experiments on a turbulent jet in a cross-flow. AIAA J 1972;10(11):1425–9. Chaissaing P, George J, Claria A, Sananes F. Physical characteristics of subsonic jets in a cross-stream. J Fluid Mech 1974;62(1):41–64. Fearn RL, Weston RP. Induced velocity field of a jet in a cross-flow. Technical Paper 1087, NASA, 1978. Ivanov Y. Shape of the centreline of an axisymmetric fantype jet in a cross-flow. Izv Vuz Aviotsionnaya Teknika (4) (1963). Shandorov G. Calculation of a jet axis in a drifting flow. Technical Translation F-10638, NASA, 1966. Margason RJ. The path of a jet directed at large angles to a subsonic freestream. Technical Note D-4919, NASA, 1968. Zhang X, Hurst DW. An investigation of a supersonic jet in a low speed cross-flow. Report 301896, Department of Aeronautics and Astronautics, University of Southampton, 1990. Wooler PT. Flow of a circular jet into a cross-flow. AIAA J Aircraft 1969;6(3):283–4. Snel H. A method for the calculation of the flow-field induced by a jet exhausting perpendicularly into a crossflow. In: AGARD conference proceedings CP-143, fluid dynamics panel symposium, Delft, Netherlands, 1974. p. 18.1–6. Siclari MJ, Migdal D, Palcza JL. The development of theoretical models for jet-induced effects on V/STOL aircraft. In: Proceedings of the AIAA/SAE 11th propulsion conference, Anaheim, CA, USA, 1975, AIAA 75-1216. Sucec J. Bowley WW. Prediction of the trajectory of a turbulent jet injecting into a cross-flowing stream. Trans ASME, J Fluids Eng (1976) 667–73. Adler D, Baron A. Prediction of a three-dimensional circular turbulent jet in a cross-flow. AIAA Journal 1978;17(2):168–74. Fearn RL. Progress towards a model to describe jet/ aerodynamic surface interference effects. In: Krothapalli A, Smith CA, editors. Recent advances in aerodynamics, New York : Springer; 1986. p. 407–34. Roth KR. Numerical simulation of a subsonic jet in a cross-flow. In: Proceedings of the international powered

190

[79]

[80]

[81]

[82]

[83] [84]

[85]

[86]

[87]

[88]

[89]

[90]

[91]

[92]

[93]

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191 lift conference, Santa Clara, CA, USA, 1987. p. 425–31, SAE 872343. Baker AJ, Snyder PK, Orzechowski JA. Three-dimensional characterisation of the near-field of a VSTOL jet in a turbulent cross-flow. Comput Methods Appl Mech Eng 1993; (102) 1–13. Chiu SH, Roth KR, Margason RJ, Tso J. A numerical investigation of a subsonic jet in a crossflow. In: AGARD conference proceedings CP-534, fluid dynamics panel symposium, Winchester, UK, 1993. p. 22.1–14. Chocinski D, Leblanc R, Hachemin J-V. Experimental/ computational investigation of supersonic jet in subsonic compressible crossflow. In: 35th AIAA aerospace sciences meeting and exhibit, Reno, NV, USA, 1997, AIAA 970714. Kennedy KD, Walker BJ, Mikkelsen CD. Comparisons of CFD calculations and measurements for a sonic jet in a supersonic crossflow. In: 36th AIAA/ASME/SAE/ASEE joint propulsion conference and exhibit, Huntsville, AL, USA, 2000, AIAA 2000-3313. Margason RJ. Jet-induced effects in transition flight. In: NASA special proceedings SP-116, 1966. p. 177–89. Fearn RL, Weston RP. Induced pressure distribution of a jet in a crossflow, Technical Note D-7916, NASA, July 1975. Spreeman KP. Induced interference effects on jet and buried-fan VTOL configurations in transition. Technical Note D-731, NASA, 1960. Volger RD. Interference effects of single and multiple round or slotted jets on a VTOL model in transition. Technical Note D-2380, NASA, 1964. Volger RD, Goodson KW, Low speed aerodynamic characteristics of a fuselage model with various arrangements of elongated lift-jets. Technical Note D-7299, NASA, 1973. Hammond AD, McLemore HC. Hot-gas ingestion and jet interference effects for jet V/STOL aircraft. In: AGARD conference proceedings CP-27, fluid dynamics panel symposium, Go¨ttingen, Germany, 1967. p. 8.1–27. Schultz G, Viehweger G. Detailed experimental and theoretical analysis of the aerodynamic interference between lifting jets and the fuselage and wing. In: AGARD conference proceedings CP-150, fluid dynamics panel symposium, Rome, Italy, 1974. p. 24.1–13. Nangia RK. Prediction & evaluation of vectored jetsinduced interference on aircraft configurations in lowspeed flight. In: Proceedings of the 19th congress of the international council of aeronautical sciences, Anaheim, CA, USA, 1994. p. 859–70. Winston MM, Weston RP, Mineck RE. Propulsioninduced interference effects on jet-lift VTOL aircraft. In: Proceedings of the AIAA/SAE 11th propulsion conference, Anaheim, CA, USA, 1975, AIAA 75-1215. Margason RJ, Volger RD, Winston MM. Wind tunnel investigation at low speeds of a model of the Kestrel (XV-6A) vectored thrust V/STOL airplane, Technical Note D-6826, NASA, 1972. Fu¨tterer H, Harms L. Jet interference measurements on a VTOL model with jet simulation by fans, In: AGARD

[94]

[95]

[96]

[97]

[98]

[99]

[100] [101]

[102]

[103]

[104]

[105]

[106]

[107]

[108]

conference proceedings CP-27, fluid dynamics panel symposium, Go¨ttingen, Germany, 1967. p 17.1–26. Krenz G, Barche J. Jet influence on V/STOL aircraft in the transitional and high speed regime. In: AGARD conference proceedings CP-27, fluid dynamics panel symposium, Go¨ttingen, Germany, 1967. pp. 2.1–25. Wiles WF. Jet-lift intakes. In: AGARDograph 103, specialists meeting, Tennessee, TN, USA, 1965. p. 559–86. Curtis P. The design and testing of short duct intakes for supersonic STOVL aircraft. In: Proceedings of the RAeS/ IMechE conference on engine/airframe integration, Nottingham, UK, 1992. p. 10.1–12. Tape RF, Hartill WR, Curry LS, Jones TJ. Vectoring exhaust systems for STOL tactical aircraft. Trans ASME J Eng Power 1983;105(3):654–62. McMichael T, McKay K, Walker MJ, Fielding C, Lockley G, Curtis P, Probert B, Lee CS, Moretti G. Aerodynamic technology—the role of aerodynamic technology in the design and development of modern combat aircraft. Aeronaut J 1996;100(1000):411–24. Carbonaro M. Interference problems in V/STOL testing at low speeds. In: AGARD conference proceedings CP-174, fluid dynamics panel symposium, London, UK, 1975. p. 40.1–21. Koenig DG, V/STOL wind-tunnel testing. In: AGARD Report R-710, 1984. p. 9.1–71. Harris AE. Test techniques for engine/airframe integration. In: AGARD conference proceedings CP-498, fluid dynamics panel symposium, Fort Worth, TX, USA, 1991. 1.1–17. Winston MM. Induced interference effects on the aerodynamic characteristics of a 0.16-scale six-jet V/ STOL model in transition. Technical Note D-5727, NASA, 1970. Mineck RE, Margason RJ. Pressure distribution on a vectored thrust V/STOL fighter in the transitionspeed range. Technical Memorandum X-2867, NASA, 1973. Mineck RE, Schwendemann MF. Aerodynamic characteristics of a vectored thrust V/STOL fighter in the transition-speed range. Technical Note D-7191, NASA, 1973. Haftmann B. Jet effects on forces and moments of a vstol fighter-type aircraft. In: AGARD conference proceedings CP-308, fluid dynamics panel symposium, Fundac- ao Calouste Gulbenkian, Lisbon, Portugal, 1981. p. 18.1–13. Harris AE, Wilde GL, Smith VJ, Mundell ARG, Davidson DP. ASTOVL model engine simulators for wind tunnel research. In: AGARD conference proceedings CP-498, fluid dynamics panel symposium, Fort Worth, TX, USA, 1991. p. 15.1–10. Saddington AJ. Mutual interference between jets and intakes in STOVL aircraft. EngD thesis, Cranfield University, RMCS, Shrivenham, UK, September 1997. Saddington AJ, Knowles K. Mutual interference between jets and intakes in STOVL aircraft. Aeronaut J 1999;103(1024):281–6.

A.J. Saddington, K. Knowles / Progress in Aerospace Sciences 41 (2005) 175–191 [109] Saddington AJ, Knowles K. Jet/intake interference in short take off, vertical landing aircraft. AIAA J Aircraft 2001;38(5):924–31. [110] Hunt DL, Bruce RJ. The role of CFD in predicting vectored jet interference for STOVL aircraft. In: Proceedings of the international powered lift conference, London, UK, 1998. p. 11.1–12

191

[111] Margason RJ, Correlation of STOVL jet induced effects with examples for configurations with supersonic convergent jets. In: Proceedings of the international powered lift conference, Williamsburg, VA, USA, 2002. AIAA 2002-5967. [112] Margason RJ, Gentry Jr GL. Aerodynamic characteristics of a five-jet VTOL configuration in the transition speed range. Technical Note D-4812, NASA, 1968.