Laser-Doppler Analysis of the Separation Zone of a Ground Vortex Flow

Laser-Doppler Analysis of the Separation Zone of a Ground Vortex Flow André R. R. Silva 1, Diamantino F. G. Durão 2, Jorge M. M. Barata 3, Pedro Santos 4, Samuel Ribeiro 5 1: Aerospace Sciences Department, Universidade Beira Interior, Covilhã, Portugal, andre@ubi.pt 2: Universitdade Lusíada, Lisbon, Portugal, durao@lis.ulusiada.pt 3: Aerospace Sciences Department, Universidade Beira Interior, Covilhã, Portugal, jbarata@ubi.pt 4: Aerospace Sciences Department, Universidade Beira Interior, Covilhã, Portugal, pedroteixeirasantos1@hotmail.com 5: Aerospace Sciences Department, Universidade Beira Interior, Covilhã, Portugal, samuelavionico@gmail.com Abstract Laser Doppler measurements are presented for a highly curved flow generated by the collision of plane wall turbulent jet with a low-velocity boundary layer. The experiments were performed for a wall jetto-boundary layer velocity ratio of 2, and include mean and turbulent velocity characteristics along the two normal directions contained in planes parallel to the nozzle axis. The results, which have relevance to flows encountered by powered-lift aircraft operating in ground effect, quantify the structure of the complex ground vortex flow resulting from the collision of a wall jet with a boundary layer. The results revealed the existence of a very low-frequency instability. The source of this low frequency unsteadiness is probably associated with a small vortex located near the separation point. In the central zone of the upwash flow where the maximum values of the vertical velocity component occurs, additional distinct high frequency peaks were also identified. 1. Introduction A primary design consideration for V/STOL aircraft is the flow environment induced by the propulsion system during hover with zero or small forward momentum. Ground effect phenomena may occur and change the lift forces on the aircraft, cause re-ingestion of exhaust gases into the engine intake and raise fuselage skin temperatures. An important source of each is the ground vortex (Fig.1) which forms far upstream of the impinging jet when the resulting radial wall jet meets a crossflow (Barata et al., 1986, 1987, 1991a; VanDalsem et al., 1987; Cimbala et al., 1987; Knowles and Bray, 1991). Measurements of this type of flow are very scarce, and have only been reported in the context of a secondary flow within the impinging jet flow problem with relatively different configurations and operating conditions. Most of the studies published so far with relevance for the V/STOL problem used small impinging distances (h/d<8) and high jet-to-crossflow velocity ratios (V R = U jet / U CL >1). Some information relevant to the flow beneath a V/STOL aircraft in ground vicinity has been provided for some limiting cases such as h/d=.4, and without the presence of a crossflow (Saripalli, 1983, 1987). Others include the effect of the crossflow with a solid surface at the jet exit plane to simulate the underside of the aircraft fuselage and wings (Barata et al., 1996; Barata et al., 1991). Among the studies published so far without the presence of the surface at the jet exit there is some agreement that the flow includes large scale, probably coherent, unsteadiness, although there is not a consensus as to their causes. Cimbala et al. (1991) report frequency spectra obtained with hot-wire measurements that revealed broadband humps indicating very low frequency unsteadiness (f=4hz for h/d=3 and V R =1) that were attributed to the large-scale puffing oscillation (low-frequency pulsating behavior) of the ground vortex, and results in a significant variation in size of the ground vortex. This phenomenon was found to do not correlate with disturbances either in the crossflow, jet wake of the jet tube, the crossflow or any oscillations in the flowfield. The low frequency oscillations were, therefore, attributed to the gross features of the ground vortex flowfield itself that included some irregularities as its growth and break-up. The reported unsteadiness was found to - 1 -

lead to larger fluctuations in the height of the vortex which reaches more that 8 jet diameters for V R =2, with an inverse variation of the frequency which tends almost linearly to zero when V R increases. Saddington et al. (27) have also observed a distinct frequency oscillation for the case of a fountain flow resulting from two compressible impinging jets without an upper plate or crossflow and nozzle pressure ratios, NPR, from 1.5 up to 4, and impinging heights, h/d, of 4.4. The studies for the highest velocity ratios with a wall at the jet exit can be found in Barata et al. (1986, 1987, 1991a, 1991b), Barata (1996a, 1996b), and Barata and Durão (24), for single, twin, and three jets configurations. These studies report numerical and experimental results obtained with LDV for velocity ratios, V R =3, 42, and 73, and impingement heights, of h/d=3, 4, and 5, for the case of a confined crossflow. Their work includes an extensive analysis of the turbulence structure of the impinging and ground vortex zones. Nevertheless, their results did not exhibit any bimodal LDV histograms of discrete frequencies for single or multiple jets that could be associated with any instabilities or oscillations. These results could be considered somewhat surprising at first sight, but it should be pointed out that all the unsteadiness of the ground vortex has been reported for unconfined impinging jet configurations only. An extrapolation of the results of Cimbala et al. (1991), that present a comprehensive set of data on oscillations for the unconfined case, to the situation of a confined crossflow of Barata et al. (1987) would correspond to ground vortex puffing frequencies of about 1.2Hz for V R = 3, but were not noticeable from the LDV measurements using collision zone Fig.1 Jet impinging on a surface through a low-velocity crossflow - 2 -

samples of 1, values and data rates of about 1Hz. Also, the corresponding range of unsteady vortex height would be h/d=6 (from 3.5 to 9.5) which is larger than the available distance between the upper and the lower plates for the crossflow. Barata and Durão (24) further analyzed the ground vortex flow resulting from an impinging jet in a confined crossflow, and found that the shape, size and location of the ground vortex were dependent on the ratio between the jet exit and the crossflow velocities, and two different regimes were identified. One is characterized by the contact between the ground vortex and the impinging jet, while another is detached upstream the impinging zone. They also report that the crossflow acceleration over the ground vortex, resulting from the blockage effect due to confinement, was directly connected with the jet exit velocity, and the influence of the upstream wall jet was not limited to the ground vortex but spread vertically upwards by a mechanism not explained so far. The quantitative and the visualization results did not revealed any distinguishable oscillation of the ground vortex which size and location seems to be only dependant on the velocity ratio, V R, and impinging distance, h/d, for the case of the confined crossflow. These results seemed to indicate that the confinement may avoid or dissemble any instabilities of the ground vortex, but the unlike relevance of the jet exit velocity, U j, reported by Barata et al. (1986) and Cimbala et al. (1991) could be an indication that the jet-to-crossflow velocity ratio could be also another important parameter. The present research program is dedicated to the identification of the parameters and relevant regimes associated with instabilities and other secondary effects of a ground vortex flow. To avoid the influence of the impinging region a plane wall jet is produced independently using a configuration already used to study two-dimensional upwash flows (see Gilbert, 1983). The wall jet collides with the boundary layer produced using a conventional wind tunnel giving rise to a highly curved region, which can be studied for different velocity ratios between the wall jet and crossflow. This paper presents a detailed analysis of a ground vortex flow resulting from the collision of a wall jet with a boundary layer, and follows that of Barata and Durão (25), which has detected a small recirculating zone, located upstream the separation point not yet reported before for this type of flows. 2. Experimental Method and Procedures The experimental facility designed and constructed for conducting laser-doppler velocimeter measurements on wall jet to boundary layer collision flows is diagrammed in Fig.2. With this facility the three-dimensional effects created by skewing of pre-existing spanwise vorticity are eliminated, and makes our data particularly interesting to assess the turbulent or transient effects near the separation point of the ground vortex where the transverse velocity component is null. The recommendations of Metha and Bradshaw (1979) for open circuit wind tunnels were followed throughout all the design process especially for the boundary layer part of the flow. A fan of 15KW nominal power drives a maximum flow of 3 m 3 /h through the boundary layer and the wall jet tunnels of 3 x 4mm and 4 x 4 mm exit sections, respectively. The facility was built to allow variable heights of the wall jet exit from 15 up to 4mm, but in the present study a Y Fig.2 Diagram of the ground vortex facility. X - 3 -

constant value of 16mm was used. The origin of the horizontal, X, and vertical, Y, coordinates is taken near the visual maximum penetration point. The X coordinate is positive in the wall jet flow direction and Y is positive upwards. The present results were obtained at the vertical plane of symmetry for a wall jet mean velocities of U j =13.7m/s and mean boundary layer velocity of U o =6.9m/s, corresponding to a velocity ratio, V R, of 2. Fig. 3 Laser Doppler velocimeter. The velocity field was measured with a two-color (two-component) laser-doppler velocimeter (Dantec Flowlite 2D), which comprised a 1 mw He-Ne and a mw diode-pumped frequency doubled Nd:YAG lasers. Bragg-cell frequency shifting at f o =4MHz was used in both channels to detect the flow reversals. The half-angle between the beams was 2.8 o and the scattered light was collected in backward scattering mode with a focal lens of 4mm. The probe volume with calculated axis dimensions at the e -2 intensity locations of 135x6.54x6.53µm and 112x5.46x5.45µm was positioned at the required location by use of a computer remotely driven X-Y-Z traversing unit with a precision of ±.mm. The horizontal, U, and the vertical, V, mean and turbulent velocities together with the shear stress, u ' v', were determined by a two-channel Dantec BSA F6 processor. The principal characteristics of the laser-doppler velocimeter are summarized in Table 1. The seeding of the flow was obtained with a smoke generator with particles of.1-5µm. The number of Table 1. Principal characteristics of the Laser-Doppler velocimeter. - Wave length, λ [nm] 633 (He-Ne) 532 (Diode Laser) - Focal length of focusing lens, f [mm] 4 4 - Beam diameter at e-2 intensity [mm] 1.35 1.35 - Beam spacing, s [mm] 38.87 39.13 - Calculated half-angle of beam intersection, θ 2.78 o 2.8 o - Fringe spacing, δ f [µm] 6.53 5.45 - Velocimeter transfer constant, K [MHz/ms -1 ].153.183-4 -

the individual velocity values used in the measurements to form the averages was always above 1,. As a result, the largest statistical (random) errors were 1.5% and 3%, respectively for the mean and variance values for a 95% confidence interval following the analysis of Yanta and Smith (1978). 3. Results and Discussion Experimental visualization studies were first performed using a direct digital photography and a smoke generator to produce the tracer particles. The visualization results of the present complex flow were used to provide a first insight into the nature of the flow and to guide the choice of quantitative measurement locations. The wall jet collides with the boundary layer and is strongly Y boundary layer wall jet X U o U j Fig.4 Example of the visualization of the collision zone, and the coordinate system used. deflected backwards giving rise to an extremely complex flow, which includes a small secondary vortex flow near the separation point, probably due to the roll up of the vorticity of the boundary layer. More detailed visualization studies that confirmed the existence of this vortex were reported in Barata et al. (28) for V R =1.7. It was found to be highly unstable with its shape, size, and location varying almost constantly. The behavior of this small vortex was found to be quite similar to the puffing of the ground vortex as reported by Cimbala et al. (1991). First the vortex is very small, but growing. The lower part of the boundary layer with anticlockwise vorticity seems to merge into the growing vortex. As the small vortex continues to grow it becomes higher than the boundary layer, and breaks up suddenly while is convected upwards in the direction of the curved flow. Then, a new small vortex appears and starts to grow, and the cyclic process repeats itself with a frequency of about 8.33Hz. Cimbala et al. (1991) attributed the vortex growth to the shear layer vortices, which convect with the wall jet, and merge into the ground vortex. Barata et al. (28) found that a different mechanism should be present for the case of higher wall jet-to-boundary layer velocity ratios, because the secondary vortex cannot merge into the deflected flow resulting from the collision of the wall jet with the boundary layer, since the vertical velocity component is always positive above the vortex. So, probably the unsteadiness reported before is due to an additional small vortex upstream of the ground vortex that due to its extreme small size could not be observed so far, as in the case of high jet-to-crossflow velocity ratios (Barata et al., 1986, 1987, 1991a; Barata et al., 1991b; Barata, 1996a, 1996b; Barata and Durão, 24). In the present paper the velocity ratio, V R, was further increased to 2, which according to Cimbala et al. (1991) would correspond to a puffing frequency of about 18 ± 1 Hz. Fig. 5 shows contours of the mean horizontal, U mean, and vertical, V mean, velocity components, which confirm the above description of the flow and quantify the mean flow characteristics of the collision zone. Streamlines - 5 -

a) 1 1 8. 6.6 5.2 3.8 2. 1.1 -.2-1.6-3. b) 1 1-1 - 1 3. 2.563 2.1 1.688 1.2.813.3 -.63 -. -1-1 Fig.5 Contours of the mean velocity characteristics for V R = 2: a) horizontal component, U mean ; b) vertical component, V mean. computed from the measured values are also plotted together with velocity vectors. The mean vertical velocity component is negative near the wall in the boundary layer side for Y<4mm, but is always positive elsewhere. This confirms the existence of the secondary vortex, but it also reveals that if it is unstable it will be simply swept upwards by the curved flow resulting from the collision between the wall jet and the boundary layer, and no puffing mechanism is observed. From the measured velocities it can be concluded that the center of the deflected flow corresponding to the maximum vertical velocity component is located in the wall jet side for X -2mm. This figure also indicates that the center of the secondary vortex flow is located upstream the separation point (with its center near X=+45mm) but probably somewhere before the so-called maximum penetration point. This result indicates that this secondary vortex flow may also be present for other situations, depending not only on the velocity ratio V R, but also on the relative size of the clockwise vorticity zone of the wall jet and the counterclockwise vorticity of the boundary layer. Figure 6 shows the turbulent velocity characteristics of the collision zone and deflected wall jet flow. The peaks of u ' 2 (Fig. 6a) are larger than the corresponding peaks of v ' 2 (Fig. 6b) in the 2 2 collision zone giving rise to high levels of anisotropy with u ' v' 2. 5.The maximum values of the horizontal velocity fluctuations are observed in the collision zone where the mean horizontal velocity component is zero giving rise to extremely high local turbulence intensity values of u' 2 U mean greater than 1%. For vertical velocity fluctuations the maximum values only coincide with - 6 -

the zero values of the mean vertical velocity component close the ground plane, and the local turbulence intensity are much smaller. These results are misleading to some extent because although the LDV measurements were obtained with a sufficiently high data rate to detect the possible low frequencies characteristic of these type of instabilities (18 ± 1 Hz according to Cimbala et al., 1991), the total time to obtain the 1, measurements needed to keep the statistical errors sufficient low (1.5% and 3%, respectively for the mean and variance values for a 95% confidence interval; Yanta and Smith, 1973) allow the averaging of about 2 cycles. And, as a consequence, such low frequency instabilities might be being treated as turbulence. a) 1 1 4. 3.6 3.2 2.8 2. 2.1 1. 1.3 1. b) 1 1-1 - 1 2. 1.81 1.62 1.43 1.2 1.6.8.68. c) -1-1 1 5. 4. 3. 2. 1.. -1. -1-1 Fig.6 Contours of the turbulent velocity characteristics for V R = 2: a) horizontal normal stress, 2 u ' ; b) vertical normal stress, 2 v ' ; c) Reynolds shear stress, u ' v' - 7 -

Fig. 6c shows contours of the turbulent shear stress, u ' v', that are generally consistent with the direction of the mean flow. The shear stress is positive along the vertical direction of the centre of the collision zone (X=) suggesting that faster moving elements of the wall jet (u > ) tend to move upwards with the deflected upper side of the boundary layer (v > ). Similarly, the shear stress along the wall jet side of the deflected flow (-11<X<-6mm and 12<Y<mm) is negative because the forward movement of fluid particles corresponds to negative vertical velocity fluctuations (v < ). However, the location of the zero values of the shear stress occur near X=, and do not coincide with the central zone of the deflected flow, which is associated with the maximum of the vertical V velocity component where =. Far from the wall (Y > mm) with the approach of the X U separation point (X = ), increases in the wall jet side (X < ) and decreases in the boundary Y V U layer side (X > ). Near the wall and are the most important shear strains, and the y X magnitude of the peak of the shear stress decreases. This is because the flow in this region is subjected to strongly stabilizing curvature that reduces the shear stress more than the turbulent kinetic energy. The largest positive values of the shear stress occur near the maximum center velocity of the upwash flow for Y>12mm with surrounding values both positive in the wall jet or the boundary layer sides. The particular ordered sequence that was identified from the preliminary visualization studies for the small recirculation zone that appears near the separation point can also be interpreted as an oscillation of the separation zone or of the virtual deflected flow origin, and can be confirmed by the bimodal histograms of the horizontal and vertical velocity measurements obtained in this zone. The histograms were classified in different types for the horizontal and vertical velocity components, and the results are plotted in Fig.7. For the horizontal velocity component, U, four different types of histograms were identified (Fig. 7a): bimodal histograms with symmetric peaks in the central region of the deflected flow; bimodal histograms with a larger positive peak in the wall jet side; bimodal histograms with a larger negative peak in the boundary layer side; and near- Gaussian histograms away from the collision zone. As shown in Fig. 7b for the vertical velocity component five different types of histograms were identified with the bimodal pattern occurring in a slightly different region around the upwash flow. In spite of the apparent organized sequence of the turbulent structure of the collision zone, the power spectra of the horizontal and vertical velocity components do not exhibit any accentuated particular peak for the same locations (Fig. 8). Three types of power spectra occurring in very well defined regions were identified, and are represented in Fig.8 for both velocity components. In the region of the collision zone where the bimodal histograms were obtained, the power spectra exhibit broadband humps with center frequencies between approximately 4 and 15Hz, which is an indication of the low frequency unsteadiness already mentioned before. Another type of spectra was found in the central zone of the upwash flow near the location of the maximum values (V mean 3 m/s). The broadband hump with a maximum value of about 15Hz can be observed, but distinguishable high frequency peaks have also appeared. The source of the low frequency unsteadiness is probably associated with a small vortex located upstream the separation point. As already mentioned before, this secondary vortex has a similar very low broadband pulsating behavior by expanding and contracting observed in some impinging jet configurations with ground vortex flows. The secondary vortex growth cannot be attributed to the shear layer vortices, which convect with the wall jet, since it cannot merge into the deflected flow resulting from the collision of the wall jet with the boundary layer, because the vertical velocity component is always positive above the vortex. The unsteadiness of the ground vortex - 8 -

a) 1 1-1 - 1 b) 1 1-1 - 1 Fig.7 Typical histograms at the collision zone for V R =2: a) horizontal velocity component, U; b) vertical velocity component, V - 9 -

1 a) 2 3 4 5 6 7 8 9 1 1 Spectrum [LDA2] [x²/ Hz] -1-1 b) Spectrum [LDA2] [x²/hz] 2 3 4 5 6 1 1 2 3 4 5-1 - 1 Fig.8 Typical power spectra at the collision zone for V R =2: a) horizontal velocity component, U; b) vertical velocity component, V - 1 -

reported before for the case of impinging jets in unconfined crossflows may also be associated with an additional small vortex upstream separation point, but due to its extreme small size could not be observed so far, as in the case of large jet-to-crossflow velocity ratios. The present results are in agreement with the results of Barata et al. (28) and Cimbala et al. (1991) as far as the low frequency unsteadiness is concerned but more information of its source and relation to the high frequency now observed is still needed. Spectrum [LDA1] [x²/ Hz] 1E-2 1E-3 1E-4 1E-5 4. Conclusions Laser-Doppler measurements of the velocity characteristics of a two dimensional ground vortex flow resulting from the collision of a wall jet with a boundary layer were presented and discussed with visualization results for a wall jet-to-boundary layer velocity ratio of 2. 1E-6 1, 1, 1, Frequency [Hz] Fig.9 Spectra of the vertical velocity component, V with high frequency peaks at (X,Y,Z)=(-2,18,) for V R =2. The results revealed the existence of a very low broadband humps in the frequency spectra indicating concentration of unsteady turbulent energy. The source of this low frequency unsteadiness is probably associated with a small vortex located upstream the separation point. The particular ordered sequence that was identified from the visualization studies for the small recirculation zone that appears near the separation point can also be interpreted as an oscillation of the separation zone or of the virtual deflected flow origin, and can be confirmed by the bimodal histogram of the horizontal velocity measurements obtained in this zone. In spite of the apparent organized sequence of the turbulent structure of the collision zone, the power spectra of the horizontal velocity component does not exhibit any pronounced particular peak for the same location. In the central zone of the upwash flow where the maximum values of the vertical velocity component occurs, additional distinct high frequency peaks were also identified. Aknowledgements The present work has been performed in the scope of the activities of the Aeronautics and Astronautics Research Center - AeroG of the University of Beira Interior. The financial support of the FCT-Fundação para a Ciência e Tecnologia of the Portuguese Ministry of Science under Contract nº PTDC/EME/-MFE/64493/26 is gratefully acknowledged. References Barata, J.M.M., Durão, D.F.G., Heitor, M.V. (1986) Experimental and Numerical Study on the Aerodynamics of Jets in Ground Effect. Tenth Symposium on Turbulence, September 22-24, Rolla, Missouri. Barata, J.M.M., Durão, D.F.G., Heitor, M.V. (1987), The Turbulent Characteristics of a Single Impinging Jet Through a Crossflow. Sixth Symposium on Turbulent Shear Flows, September 7-9, - 11 -

Toulouse. Barata, J.M.M., Durão, D.F.G., Heitor, M.V. (1991a) Turbulent Energy Budgets in Impinging Zones. Eighth Symposium on Turbulent Shear Flows, September 9-11, Munich. Barata, J.M.M., Durão, D.F.G., Heitor, M.V. (1991b) Impingement of Single and Twin Turbulent Jets Through a Crossflow. AIAA Journal, Vol. 29, No. 4:595-62. Barata, J.M.M. (1996a) Ground Vortex Formation with Twin Impinging Jets. International Powered Lift Conference. SAE Paper 9627, November 18-2, Jupiter, Florida. Barata, J.M.M. (1996b) Fountain Flows Produced by Multiple Jets in a Crossflow. AIAA Journal, Vol. 34, No. 12: 23-3. Barata, J.M.M., Durão, D.F.G. (24) Laser-Doppler Measurements of Impinging Jets Through a Crossflow. Experiments in Fluids, Vol.36, No.5:117-129. Barata, J.M.M., Durão, D.F.G. (25) Laser-Doppler Measurements of a Highly Curved Flow. AIAA Journal, Vol. 43, No.12:2652-2655. Cimbala, J.M., Billet, M.L., Gaublomme, D.P., Oefelein, J.C. (1991) Experiments on the Unsteadiness Associated with a Ground Vortex. Journal of Aircraft, Vol. 28, No. 4:261-267. Gilbert, B.L. (1983) Detailed Turbulence Measurements in a Two-Dimensional Upwash. AIAA 16 th Fluid and Plasma Dynamics Conference, AIAA paper 83-1678, July 12-14, Danvers, Massachusetts. Knowles, K., Bray, D. (1991) The Ground Vortex Formed by Impinging Jets in Crossflow. AIAA 29th Aerospace Sciences Meeting, AIAA Paper 91-768, January 7-1, Reno, Nevada. Metha R.D., Bradshaw P. (1979) Design Rules for Small Low-Speed Wind Tunnels. Saddington, A.J., Knowles, K., Cabrita, P.M. (27) Flow Visualization and Measurements in a Short Take-off, Vertical Landing Fountain Flow. 45 th AIAA Aerospace Sciences Meeting and Exhibit, January 8-11, Reno, Nevada. Saripalli, K.R. (1987) Laser Doppler Velocimeter Measurements in 3D Impinging Twin-Jet Fountain Flows. Turbulent Shear Flows, edited by Durst et al., Springer-Verlag, Berlin, 5:147-168. Saripalli, K.R. (1983) Visualization of Multijet Impingement Flow. AIAA Journal, 21:483-484. Van Dalsem, W.R., Panaras, A.G., Steger, J.L. (1987) Numerical Investigation of a Jet in a Ground Effect with a Crossflow. International Powered Lift Conference, SAE Paper 872344, December 7-1, Santa Clara, California. Yanta, Z., Smith, R.A. (1973) Measurements of Turbulent-Transport Properties with Laser- Doppler Velocimeter. AIAA Paper 73-169, 11 th Aerospace Sciences Meeting, Washington. - 12 -