Küt cisimlerde sürükleme kuvvetinin azaltılması için kilitlenmiş girdap oluşumlarının deneysel yöntemlerle araştırılması

thumbnail.default.placeholder
Tarih
1997
Yazarlar
Erkara, Şeref
Süreli Yayın başlığı
Süreli Yayın ISSN
Cilt Başlığı
Yayınevi
Fen Bilimleri Enstitüsü
Institute of Science and Technology
Özet
Bu tez çalışmasında, iki boyutlu yan silindirik bir cismin toplam sürüklemesinin azaltılması, cismin arkasına ayırıcı levha ve dik plakaların eklenmesi durumunda deneysel olarak incelenmiştir. Yan silindirik cismin arkasına konan ayırıcı levha ve bu levhaya değişik uzaklıklarda dik olarak yerleştirilen plakalarla yakın iz bölgesi değiştirilmeye çalışılmış ve yan silindir, ayırıcı levha, dik plaka arasında oluşan boşluk içindeki akım yapısı incelenerek kilitlenen girdaplar hakkında araştırmalar yapılmıştır. Tezin 1. Bölümünde konunun önemi ve literatür çalışmaları hakkında bilgiler sunulmuş, 2. Bölümde; model boyutlan, kullanılan tünellerin akım şartlan ve yardımcı üniteler hakkında bilgi verilmiştir. 3. Bölümde ise, kantitatif bilgiler ve deneysel çalışmalar sayısal sonuçlarla birlikte yorumlanmıştır. Çalışmalarda kullanılan programlar, bilgisayar kontrollü sıcak tel ve LDA(lazer dopler anemometresi) ölçme sistemiyle ilgili detaylı bilgiler ek olarak çalışmanın sonunda verilmiştir. Deneysel çalışmalar, akım görünürlüğü deneyleri ile başlamış ve bu çalışmalarda, duman ve yağ metodları kullanılarak kalitatif sonuçlar elde edilmiştir. Deneylere değişik geometrik düzenlemelere ait detaylı basınç ölçümleriyle devam edilmiş taban,ayırıcı levha ve dik plaka üzerindeki basınç katsayıları değişimleri araştırılmıştır. Ayrıca, iz bölgesinde sürükleme analizi ve hız profillerinin incelenmesi ile değişik geometrik düzenlemelerin etkinliği araştırılarak, sayısal ve L.D.A. (Lazer Dopler Anemometresi ) ile yapılacak çalışmalar için ön bilgilerin yanında en uygun geometrik düzenlemelerin incelemesi yapılmıştır. Girdap hareketleri ve en uygun geometrik düzenlemenin tespit edilmesi amacıyla incelenen 18 boşluk oluşumu içinden optimum geometrik düzenleme ortaya çıkartılmıştır. Yapılan çalışmalarda, duman ve yağ metodu ile yapılan akım görünürlüğü deney sonuçlarına göre, kullanılan üç ayn yükseklikteki dik plakadan en yüksek boya sahip (0.43 7D) olan için kilitli girdabın varlığına dair kalitatif sonuçlar elde edilmiştir. Yan silindirik cismin tabanından,ayırıcı levhadan ve dik plaka üzerinden alınan statik basınç değerlerinin basınç katsayısına çevrilmesiyle, kilitli girdabın yarattığı basınç katsayısı dağılımları, en etkili düzenlemenin 0.437 D yükseklikli dik plaka ile ve tabandan 0.68 D geride olduğunu göstermiştir. Ayrıca bu düzenlemeyle yapılan iz bölgesi(2D mesafe geride) sürükleme analizi sonunda, sürükleme katsayısı diğer dik plakalara göre(0.375D, 0.31 2D) yaklaşık olarak %45 oranında daha düşük çıkmıştır. Sıcak tel ile yapılan çalışmalar, ayırıcı levhadan itibaren 3D mesafede yapılmış, bu çalışma sonunda elde edilen hız profillerinden, optimum geometrik düzenlemenin etkinliği tekrar tesbit edilmiştir. Zira U(%RMS) değerlerinin diğer boyuttaki dik levha durumlarına göre %50 oranında azalması bu görüşü desteklemiştir. LDA(Lazer Doppler anemometresi )ölçümleri sonunda, kilitli girdabın optimum geometrik düzenlemedeki(x5 nolu istasyon ve "BU" dik plakası)durumuyla diğerlerinin mukayese edilmesini sağlamış(girdap şiddeti I' =0.377m2/sn olup en yakın değerden %20 daha fazladır), çizdirilen hız vektörleriyle gerek büyüklük gerek şiddet olarak bu sonuçlar doğrulanmıştır. 
Aerodynamic drag results from various sources such as skin friction, pressure, interference and lift induction. Drag reduction technology is important especially for military and civilian operators. For example a 10% drag reduction on a large military transport aircraft is estimated to have the potential to save up to 13 million gallons of fuel per aircraft over its life time [1]. Therefore studies on drag reduction technologies can be considered highly crutial. The aerodynamic forces experienced at the surface of a vehicle may be either tangential or normal to the surface. The tangential force is the viscous skin friction which is due to the development of the boundary layer over the surface. The production of the normal forces which are due to normal pressures is more complex. They can arise from a number of contributors: 1) Pressure field modification due to displacement thickness of the boundary layer and possible formation of separation regions, 2) pressure forces arising from the formation of vortices in the wake which influence the flow over the body (This is also termed as the vortex drag), 3) if the compressibility effects are present there are additional pressure forces due to compressibility and presence of the waves in the flow, 4) Due to lifting conditions there is a strong component of the lift dependent vortex drag which is called the induced drag. All of these drag sources contribute to the total drag by different relative amounts for different types of aircraft [1,2]. The drag of bluff bodies consists mainly of pressure drag while the skin friction drag being only a small part of the total drag. At subcritical Re numbers the flow over bluff bodies is characterized by a large wake and periodic, alternate vortex shedding. The separated shear layers feed vorticity to these alternating vortices. For Re numbers higher than the critical value,a transition occurs in the separated shear layer and the flow reattaches to the body as a turbulent boundary layer. This reattached flow continues along the body and eventually separates and forms a turbulent wake much smaller in size than the wake at subcritical Re numbers. Therefore in order to obtain the low drag the vortex- shedding must significantly be suppressed. If there are abrupt changes in the boundary conditions the turbulent wall layers indicate a local modification such as a loss or increase in momentum, a change in inner region and outer region properties and scales. There are several methods to alter the turbulence production process in the outer region on the boundary layer and reduce the turbulent skin friction drag[3]. Generally, there are two methods for the reduction of skin friction drag. The first is to delay transition as much as possible. This is called laminar flow control method [1]. The second one involves modifying or interacting with the turbulent structures. Since the friction drag is lower for laminar boundary layers it is important to delay the transition of boundary layers from laminar to turbulent. This can be done mainly by three ways: 1) Shaping the surface to provide long runs of favorable pressure gradients, 2) Providing more stable boundary layers through suction, 3) Providing more stable boundary layers by surface cooling. Diffusion without separation is accomplished either by suction or by tangential blowing just ahead of and through the region of adverse pressure gradient. In the aircraft industry fuselage shaping to increase the extent of natural laminar flow has not received much attention in the literature. Fuselages generate about 50% of the xii total profile drag for the all turbulent airplanes. The delay of laminar to turbulent transition by about 27% of the body length reduces the body drag coefficient by as much as 30% [4]. One method for laminar flow control is to use wall suction. The suction is not sufficient to suppress any existing turbulence, but serves to modify the curvature of the laminar velocity profile which in turn reduces amplification of any instability waves in the boundary layer that lead to the formation of turbulence. Another method for drag reduction is to use wall cooling. A reduction in surface skin temperature is known to lead to significant increases in the minumum critical Reynolds number. This is not because the kinematic viscosity goes up but because viscosity distribution across the boundary layer is modified to an increased level of stabilities [1]. Fuselage drag reduction or flow control using thermal means is becoming more popular but delaying transition by surface temperature control is difficult at fuselage Reynolds numbers as high as 300 million. For bluff bodies, the separation delay from partial wall cooling of the afterbody is approximately the same as the fully cooled body. The partial wall-cooled body has the additional benefit of lower frictional drag increase due to wall cooling. Therefore, partial wall cooling of the afterbody may have applications in reducing the separation of bluff bodies and the associated pressure drag [5]. If turbulence can not be avoided then another turbulence skin friction reduction approach may be either passive as in the case of the riblets and large eddy break up devices or active as in the case of the synthetic boundary layers. Changing the surface geometry with micro-grooves works as spatially locking the near wall structures of the turbulent boundary layer. If the geometry is right these grooves alter the momentum transport characteristics and reduce the skin friction of the order of 10% [1, 6]. This is accomplished despite the increased wetted area. The depth of the grooves for the flight conditions encountered by a commercial transport aircraft is in the order of 0.002 - 0.004 inches [7]. Caram and Ahmed [8] investigated the effect of the grooves in the near wake region. They found that for a symmetrical profile at Re=2.5*105 the total drag decreased by %13.3 Another promising concept for the reduction of turbulent skin friction is the use of plates or fences inserted into the boundary layerfl], [9]. Friction reductions on the order of 20% have been recorded. Since the suggested devices are to break up the large scale stuructures they are called the large eddy breakup devices (LEBU). Here there is a device drag penalty, but a net drag reduction can be achieved. The best configuration for these devices is the thin airfoil shape to minimize the drag of the device. They must be on the order of the local boundary layer thickness in streamwise extent and located at about 80% of the boundary layer thickness from the wall. Tandem devices also perform well and geometrical characteristics are critical for good performance. A tandem configuration within the boundary layer can yield local skin friction reduction of up to 35% [9]. Hough [10] investigated the effects of placing a parallel plate as a turbulence manipulator in a boundary layer using flow visualization and hot wire techniques. The skin friction coefficient distribution was calculated from the streamwise change in the fluid momentum. The wall region was seen to be dominated by bursting events, i.e., the ejection of low speed fluid from the wall. Measurements indicate that bursts are triggered by the sharp acceleration and inflexional profile associated with the sweep of high speed fluid from the outer flow field causing the growth of the turbulent boundary layer as a mechanism for surface drag. By altering or removing the intermittent outer structure in the turbulent boundary layer the bursting can be interrupted. This will also inhibit the interaction of high and low speed fluid near the wall thereby leading to reduction in the local shear stress at the surfaces [10]. Howard and Goodman [11] investigated the effect of shoulder radiusing and grooving the afterbodies to reduce the base drag at low speeds. The results indicate that xiii increasing the shoulder radius to two body diameters can reduce the drag levels to those of a streamline body having 67% greater fineness ratio. For the relatively sharp shoulder case using circumferential or longitudinal grooves gave 50% and 33% drag reductions. Bluff bodies have a low fineness ratio usually on the order of 3 or less. It is possible to eliminate most flow separation to obtain low drag by moving the maximum body thickness The major drag reduction mechanism of the circumferential grooves is substituting several small regions of separation for a larger separated flow region. The V-grooves are the most effective when the after body shoulder radius is zero. Afterbody shoulder radiusing is seen to be more efficent technique than circumferentional or longitudinal grooves [11]. These are referrred to as micro-air bearings with an implication that small vortices circulate in the cavities providing low shear stress to the external flow at the lip of the cavity. The separation of flow at or near the base of a bluff body moving through a real fluid leads to the development of a free shear layer. This creates a trailing stagnation point or region due to the recirculation of flow behind the base. This recirculation region is usually referred to as the near wake of the body and has a significant influence on the base drag, base heat transfer, and the configuration of the far wake.The variation of base suction with Re number shows remarkably good correspondence with the variation of the Strouhal number [12]. Williamson and Roshko also report that there İs a relation between with the variation in suction and the physical modes of wake formation [12]. The drag of slant based three dimensional bodies such as the unswept rear fuselage of airplanes and car shapes can be high. Generally, this high drag is associated with the powerful longitudinal vortices in the flow direction. Small changes of incidence may produce large changes of drag near the critical angle, and very simple modification to the base alter the drag significantly by influencing the formation of the vortices [13]. Colin and Alkorn [14] measured the drag and base pressures for ogive cylinders slanted base and found that when slant angle was 45 degrees, there was a dramatic change in aerodynamic characteristics corresponding to the wake structure. The strength of wake vortices is influenced by the state of the forebody boundary layer. Direct base suction was also used to reduce the drag because wake region was reduced. If the geometry of the body allows the flow in the afterbody region to be attached, then base suction flow control concepts are not necessary [1]. Base mass transfer has also a great influence on the location of the rear stagnation point. The thickness of the approaching boundary layer has a dominant effect on base pressure [15]. The base drag can be reduced by weakening the vortex shedding or moving displacing the vortex formation position further downstream. The base cavity can also be used as a drag reducing mechanism and is investigated. The vortices penetrate partially into the cavity for at least a portion of the shedding cycle. The pressure rises in the low velocity region between the first vortex and the back of the cavity yielding a higher base pressure for the cavity compared to the blunt based. However, in the experiments the presence of the base cavity was seen not to decrease the shedding frequency. Devices such as splitter plates and using base bleed decrease the interaction between the vortices, and increase the shedding frequency [16]. It was found that two dimensional bluff body wake was more stable and the effect of freestream turbulence was different for different blunt based bluff bodies. When longitudinal vortices are present, the addition of free stream turbulence slightly reduces the magnitude of the peak suctions on the base, but has little effect on the base, drag. The interaction between the wake and boundary layer is an important phenomenon in the study of turbulent flows. The study of the interactions between the boundary layer and the wakes of a streamlined body and a bluff body shows that for the same drag a bluff body causes faster interaction in the sense that the velocity becomes monotonic earlier. This is caused by the higher level of the fluctuating quantities behind xiv the bluff bodies. However as the velocity profiles becomes monotonic, the velocity gradients in the shear layer decrease, and the velocity fluctuations also decrease faster. Wing profiles with a blunt trailing edge may have many advantages in high speed flow compared to the conventional profiles with a sharp trailing edge. The high base drag of the blunt trailing edge is a great penalty. Therefore methods for reducing base drag have been the subject of numerous investigations. Tanner [17] gave the results from tests using broken trailing edges and splitter plates or splitter wedges. The base drag was decreased as much as 13%. With a splitter plate of a length equal to the base thickness approximetaly 50% reduction in the base drag was achieved. Using trailing edge slats the base drag was reduced about 16%. The greatest base drag reduction, about 64%, was obtained by using broken trailing edge [17]. At low subsonic Mach numbers and at low Reynolds numbers drag reductions as much as 56% were achieved using stepped after bodies as compared to identical bodies with flat bases [18]. Kentfield found that a stepped after body could be an effective device for fairing the bodies of revolution especially when it was necessary to minimize the length of the afterbody. The minumum drag attained using a stepped afterbody was only 33% greater than that was obtained using a conical afterbody of 2-4 times the length. It was also stated that And optimization of the configuration geometries could result in further improvements [18]. Kidd and etal. investigated the stepped bases for high subsonic, transonic and supersonic cases using free flight tests [19]. The results indicate that the addition of a stepped base can significantly reduce the aerodynamic drag over that of a flat base. Hovewer, they are not as efficient in reducing drag as a standard truncated boattail. Supersonicly, these stepped bases have no apparent advantage over that of the flat base [19]. Apelt and West investigated the flows past circular cylinders with wake splitter plates having lengths between two and seven diameters [20]. The pressure distributions and wake Strouhal numbers were measured and visualizations studies were carried out. Compared to the values for a plain cylinder, splitter plates were seen to cause a reduction in the drag, an increase in the base pressure, narrowing of the wake and a change in the Strouhal number for vortex shedding. The greatest effect on the drag coefficient and base pressure coefficient was observed for a splitter plate a diameter long. However very short plates caused considerable changes in the drag and the base pressures. It was concluded that for the splitter plates shorter than the cylinder diameter the separation point on the cylinder did not change considerably. For longer splitter plates the size of the wake is reduced. Splitter plates longer than 7 cylinder diameters do not have any significant change. Drag and vortex shedding is strongly affected for splitter plates with lengths from 2 to 5 diameters. With splitter plates longer than 5 diameters the flow reattaches to the surface at about 5 diameters downstream from the cylinder, regardless of the splitter plate length. Once the reattachment point is on the plate surface, there is no further significant change in the flow pattern [20]. The effect of splitter plates on the wake flow characteristics of a rectangular cylinder with fixed separation points is found to be similar in nature to a circular cylinder with an attached splitter plate. The base pressure values were seen to be inversely proportional to the distance of the vortex formation position from the cylinder. The overall drag coefficient of the cylinder can decrease about 50% depending on the gap and the length of the splitter plates. The wake characteristics downstream of a circular cylinder can be considerably altered by placing a splitter plate on the wake centerline. The effect of the splitter plate is maximum when its edge nearer to the cylinder is about 2 to 3 cylinders diameter downstream from the cylinder[21]. For the reduction of the drag associated with the separated flow of generic streamlined shapes, xv concepts such as the use of vortex generators have been in use for many years. A novel technique for flow control uses a disk in the wake region of a bluff body to lock a vortex in the wake. This gives rise to some pressure recovery on the afterbody which in turn reduces the total drag.. Drag reduction of blunt bodies by placing discs on spindles ahead of the body or for afterbody reduction was investigated [22,23]. Low drag regime was characterized by a relatively smooth vortex motion in the cavity formed between the body base and the trailing disc (locked vortex). Body contouring was accomplished using locked vortices. Little and Whipkey investigated the drag force, in addition smoke tunnel flow visualization and laser velocimeter measurements were used for the physical description of the flow over locked vortex afterbodies. It was demonstrated that in order to reduce the drag the disk or other device defining downstream boundary of the locked vortex cavity must be large enough to separate the wake back flow from the cavity flow so that a locked vortex could exist in the cavity. Futhermore, the cavity formed must have dimensions such that the locked vortex effectively fills the cavity. The cavity dimensions for a given flow must match the smooth stable vortex [22,23]. In this study the drag reduction of two dimensional bluff bodies were investigated using locked vortices. The vortices are locked behind the bluff body by using a splitter plate and vertical plates. The size of the locked vortices are determined by the height and the location of the vertical plates. The strength of the vortices produced should determine the level of the drag reduction. Therefore the optimum configuration was searched by the flow visulization, pressure and drag measurements. Velocity measurements were also performed using a Laser Doppler Anemometer (LDA). From these measurments vortex structure in the cavity was obtained by plotting the velocity vectors on top of the physical geometry. Details of the experimental results are given in Section 3. Flow Visualization Studies Smoke method was used to verify the existence of locked vortices in certain configurations. Although these experiments were made at a much lower Re number than that of pressure measurements, they can still show that the size of the wake is reduced by the presence of the vertical plates in certain configurations which provide relatively lower drag. In flow visualization by the oil method, the effects of the large scale vortices on the surfaces of the model (half cylinder-splitter plate-rectangular plate combination) were examined. For this purpose, a mixture of TİO2 and engine oil was applied on the surfaces of the model and the surface flow patterns were obtained by taking pictures with a camera. Results show that among three vertical plates with different heights, the heighest one provided strong evidence that the locked vortex was present in the cavity by examining the pattern of oil streamlines on the cavity surface. Pressure Measurements Pressure measurements were made on cavity surfaces and in the body wake to determine the cavity pressure distribution and the corresponding pressure drag. For the cavity surface, static pressures were measured on the splitter plate over 13x19 holes, on the base over 16x4 holes and on the vertical plate over 8x3 holes. All of these measurement locations lie in a central portion of the model so that the upper and lower walls of the wind tunnel do not have a significant effect. In the wake, pressures were measured by using pitot static tube at a distance of 2 diameters (D) downstream of the model. From these xvi measurements pressure coefficients and the pressure drag were obtained using Jones Method. Results show that a configuration with the vertical plate having a height of 0.437 D and located at a distance 0.68 D downstream of the base has minimum pressure drag, therefore considered as the optimum configuration. Pressure distribution on the cavity surface for the optimum configuration is found to be consistent with the other results such as flow visualization and the drag measurements. Hot Wire Measurements Hot wire measurements were made to obtain velocity profiles at a distance of 3 diameters (D) downstream of the model. From these measurements rms values of the streamvise velocity (U(%RMS)) were obtained. Note that these measurements were made at a further downstream location to check the development of the wake structure and to compare with the results obtained by the Jones Method. Results indicate that the U(%RMS) for the optimum configuration is 50% less than that of almost all other configurations. It is also seen that for the optimum configuration, the wake velocity profile was smoother than other configurations suggesting that the influence of the vertical plate on the wake was minimized by having the vortex locked largely in the cavity. LDA (Laser Doppler Anemometer) Measurements LDA measurements were made to obtain a detailed structure of the vortex inside the cavity. Using the LDA system mean values of the streamvise (U) and transverse (V) velocities were obtained and the resultant velocity vectors were plotted to visualize the vortex structure. For the optimum configuration, there is a clearly observable main vortex within the cavity and two small corner vorticies. For non-optimum configurations the main vortex is difficult to observe and the corner vortices are relatively bigger suggesting that the main vortex is not in a locked condition as opposed to the optimum configuration.
Açıklama
Tez (Doktora) -- İstanbul Teknik Üniversitesi, Fen Bilimleri Enstitüsü, 1997
Thesis (Ph.D.) -- İstanbul Technical University, Institute of Science and Technology, 1997
Anahtar kelimeler
Küt cisimler, Sürüklenme katsayısı, Uçak Mühendisliği, Blunt body, Elutriation coefficient, Aircraft Engineering
Alıntı