İki elemanlı bir profil sisteminde yüksek basınç gradyantlı flap üstü akım alanının incelenmesi

Abstract

Sınırlı pist uzunlukları nedeniyle kalkış ve iniş hızlarının küçük olması ve buna bağlı olarak da tutunma kaybı sınırı yakınlarında uçulması zorunluluğu, uçaklar için bu uçuş rejimlerini kritik hale sokmaktadır. Bu hallerde kanatların performanslarını, slat ve flap adı verilen aşın-taşıma düzeneklerini açarak iyileştirmek mümkündür. Ancak aşın taşıma şartlarında kanat etrafındaki akım alam hayli karmaşıktır ve bu konuya yönelik çalışmalar teorik ve deneysel olarak devam ettirilmektedir. Bu tez çalışmasında flaplı bir NLR 7301 profili etrafındaki akım alam teorik ve deneysel olarak incelenmiştir. Teorik incelemelerde profil etrafmdaki gerçek viskoz akımın bir sürtünmesiz akım/viskoz akım birleştirme yöntemiyle hesaplanması amaçlanmıştır. Bu amaçla, tek elemanlı profiller için geliştirilmiş olan bir yöntem bu çalışma kapsamında çok elemanlı profiller için uyarlanmış, ayrıca yöntemin sürtünmesiz akım hesaplamalarıyla ilgili kısmım oluşturan panel yöntem de yine bu çalışma kapsamında kompleks düzlem kullanılarak çok elemanlı profiller için geliştirilmiştir. Genelleştirilen bilgisayar kodu analitik kökenli tek ve çok elemanlı kanat profilleri üzerinde geniş bir şekilde test edilerek başarısı kanıtlanmıştır. Deneysel çalışmalar kapsamında NLR flaplı profilinin modern malzeme ve tekniklerle üretilen bir modeli üzerinde elektronik çoklu-basmç-ölçer kullanılarak basınç ölçmeleri ve sıcak-tel anemometresi kullanılarak hız ölçmeleri gerçekleştirilmiştir. Ayrıca bu çalışmalarda kullanılmak üzere bir çok bilgisayar programı ve donanım geliştirilmiştir. İlk safhada, yüzey basınç dağılımım belirlemek üzere flap açısının 20° değerinde ve değişik hücum açılarında (0°-14°), 726.000 Reynolds sayısmda yapılan deneylerde, küçük hücum açılarında gerek potansiyel akım ve gerekse gerçek akım hesaplamalarının sonuçlarının deneysel sonuçlara hayli yakın olduğu, ancak gerçek akım sonuçlarının viskoz etkilerin de hesaba katılması nedeniyle deneysel sonuçlara daha yakın olduğu tespit edilmiştir. Bu deneyler potansiyel akım çerçevesinde desteklenerek sonuçlar üzerindeki rüzgar tüneli duvar etkileri incelenmiş ve düzeltilmiştir. İkinci grup deneyler olarak model sisteminin flapüstü bölgesinde sıcak- tel anemometresiyle değişik istasyonlarda hız ve türbülans ölçmeleri gerçekleştirilmiştir. Elde edilen sonuçlardan, ana kanat izinin flap üzerindeki gelişimini görmek mümkün olmaktadır. Özellikle türbülans profilleri, ana kanat iziyle flap üzerindeki sınır tabakanın kuvvetli bir karışım göstermediğini ortaya koymaktadır. Flap firar kenarı bölgesine yaklaştıkça karışımın bir miktar arttığı dikkati çekmektedir. Sonuçlar, incelenen akım alam için anlamlı hız ve türbülans haritalarının ortaya konulabileceğini gösterir niteliktedir.

One of the most critical maneouvers of an aircraft is always the take off or the landing. Because of the insufficient speed to create the necessary lift of aircraft results in high angles of attack near the stalling conditions. For military aircaft, the requirement to operate from shortened runways, places a greater emphasis on the performance at take-off and landing. For civil aircraft, improving the airfield performance has always been recognized as an important ingradient in achieving the best overall performance. The lifting capability of a wing is impoved by deploying devices such as slats and flaps which increase the camber and chord of the aerofoil. These devices distributes the large pressure rise over the upper surface of the aerofoil associated with the high lift coefficients into several parts, thus a fresh boundary layer is generated to compensate for the pressure gradiant and separation is less likely to occur than for the corresponding pressure drop on a single aerofoil. However the flow about a two- dimensonal high-lift aerofoil contains several complicating features in comparison with a single aerofoil. These include: a) the region of flow is multiply connected, which makes even the calculation of the inviscid flow difficult. b) the wakes from upstream elements interact with the boundary layers on downstream elements. c) initially the wake from upstream elements develop in the strong pressure gradients induced by the downstream elements. d) the viscous flow region over the flap is very thick with the curvature of the flow producing significant pressure gradients across the region. e) the flow curvature has a marked effect upon the development of the turbulent viscous layers. f) the slat and flaps must fit snugly back to form the cruise aerofoil with the result that when the devices are deployed there are coves on the lower surfaces of the slat and the wing and the flow in these regions is separated. Xll g) for a multi-element airfoil, there is a small region of supersonic flow on the upper surface of the slat even at low freestream Mach numbers (say 0.2). h) for large deflections of the flap the flow on the upper surface of the flap is usually separated throughout the incidence range. All the features of this flow may be captured by a solution of the 'Reynolds-averaged' Navier-Stokes equations with a suitable turbulence model. Although solutions of the Navier-Stokes equations can now be produced on supercomputers in under one hour for a wing, the turbulence models are not sufficient to represent all the features of the flow for a high-lift aerofoil such as the changes in the turbulence structure induced by the flow curvature. For the aircraft designer the adoption of a Navier-Stokes solution would require access to a supercomputer; the development of a suitable turbulence model; an algorithm for which the numerical method does not introduce a high level of artificial diffusion and a fast algorithm to meet the tight schedules of aircraft design. With these problems in mind, the computation of the flow about high-lift aerofoils has relied upon viscous-inviscid interaction methods. Improved schemes describe accurately more of the features of the flow. While notable progress is being made in the development of methods of computing flows around single airfoils, including flow separations at subcritical speeds, methods that compute the flow around high-lift systems of multi-element airfoils, including realistic representations of complex shear flows, do not widely exist, and development efforts are currently being made. The complex geometric configurations and the resulting complexity of the physics of the flow mechanisms of such systems make the task very difficult if details of the shear flows are to be modeled. At present, due to the scarcity of experimental data that contain sufficient details of the flowfield, it is not even clear what features need to be modeled and to what level of sophistication. Among the published data on multielement airfoil flows, few of them appears to be extensive and detailed, revealing a number of features that draw the attention of computational modelers. They give limited information about small- and large-scale separations, complex interacting shear layers, highly curved boundary layers, and near wakes, although in most high-lift cases, the viscous-inviscid interaction appears to be strong. It is clear that experimental data are needed that contain the detailed information including turbulence characteristics of shear flows about realistic multielement airfoils under realistic flight conditions. The theoretical and experimental works in this thesis summarized below contain primarily: - a viscous / inviscid coupling method to calculate the real flow around multi-element airfoils, xm - detailed investigations of the flowfield around a flapped NLR 7301 airfoil model at low Reynolds number. Theoretical methods The theoretical work has been focused mainly on the calculation of the real flow around multi-element airfoils by a viscous / inviscid coupling technique. However, a two-dimensional panel method developed in complex plane was used to support the experimental works, particularly in order to investigate the wind tunnel wall effects and to correct them. The coupling method originally developed for single element airfoils by B.R. Williams has been adopted for multi-element airfoils in the frame work of this thesis. The potential flow part of the method also reconsidered in the complex plane to obtain a formulation for multi-element airfoils. Experimental Arrangement Wind Tunnel In this thesis, the experiments have been realized in the low speed closed circuit wind tunnel of ITU Aeronautical and Astronautical Faculty (Gümüşsüyü), having a test section of 80x1 10x160 cm. The flow quality in the test section was improved before the experimental works, by introducing some vortex producing devices on the walls at the entreance of the diffuser to remove the flow separations on its side walls. The vortex generators diminish the flow Auctions in the test section from 4% to 1.5%. The turbulence level in the test section is about % 0.6. Multielement Airfoil Model The two-element airfoil used in the experiments is an NLR 7301 flapped airfoil originally developed in NLR. The coordinates of main element of this flapped airfoil is the same of the original NLR 7301 airfoil up to 94,36% on the upper surface and up to 60% on the lower surface. The rest of the main element is modified to obtain a good flow characteristics with no separation under the effect of the flap. Aerodynamic characteristics of the flapped airfoil have been invested in various researches. The results of these investigations indicate how carefully the airfoil coordinates have been selected, giving a very big advantage for experimental investigations on complicated problems. Models have been produced in slices by a kind of hard plastic material KESTAMID used videly in the industry to produce some machine parts. The slices have been manufactured in a CNC machine. Slices have than been connected with centralized pins and long bolts. There are fifty pressure holes on the middle slices of the models (34 on the main element and 14 on the flap). It has been a great advantage of using a plastic material in order to dril holes for pressure ports and to connect these ports to the manometer situated and out of the wind tunnel. XIV Circular lids of runnel test section are used in model mounting. Models have been fitted to these lids and angle of attack of the models could be changed under control by turning the lids. Measurement Systems The measuring devices used in the experiments are an electronic pressure scanning system of S CANI VALVE, a constant temperature hot-wire anemometer of DISA, an electronic micromanometer of FURNESS CONTROL, a three dimensional traversing mechanism and a data acquisition system of PCL 818 connected to a PC 386. Electronic pressure scanning system has 32 pressure sensors, with a 2x32 channels capability and a high sampling rate. For the efficient use of the system; a data acquisition system having digital output functions and some pneumatic flow control elements with high pressure air supply is needed to select operational modes. For the boundary layer measurements a DISA 55M CTA hot-wire anemometer system have been used. During this investigation the possibility of producing an anemometer by the researcher himself also was searched, a pilot work was started and a prototype was produced. Unfortunately, there was no time during the thesis to test this anemometer. The data acquisition methods has been effectively used in the hot-wire anemometry. The traversing mechanism was produced locally within the frame work of this thesis, by the support of NATO/ AG ARD Fluid Dynamics Panel through the Support Project T-76. The mechanism have three axis, one of them controlled by a step motor while two other are controlled manually. The necessary software to control the traversing mechanism also have been developed locally during this thesis. A FURNESS CONTROL 's electronic micromanometer was used to measure the free stream dynamic pressure in the test section. Thus the little effect on the measurements of the flow fluctuations in the test section also removed by non- dimensionalizing all the measured data with this real time measured dynamic pressure. A PC-Lab 's PCL-818 interface board having a 12 bit resolution on 16 channel with analog/digital and digital/analog conversion capability, digital input/output and counter/timer functions was used for all data acquisition requirements during the experiments. The necessary software have been developed locally in the frame of this thesis in QBASIC 4.0 language. The programs allow the users to see the results of measurements with the graphics plotted on the computer screen during the experiments. Thus all the evaluations during the experiments can be judged. Measurements Although around the main aim of the experimental work in this thesis is to analyse the flow field the NLR 7301 airfoil 's flap under large pressure gradient, pressure measurements on the airfoil surface are also realized to understand the overall XV characteristics of the airfoil system. Before the experiments in flapped case some initial measurements are made on the main element to test the model, measurement system, softwares etc. and to compare the results with potential flow calculations and real flow calculations. The wind tunnel wall effects was investigated before producing the models as to decide on the chord length and a length of 400 mm was found convenient. However some corrections are needed on the experimental results. The flow around the flapped airfoil was visualized by tufts to check the effects of test section side walls on the two dimensionality of the flow. A system to suck the boundary layer around the model side wall connection regions is developed. However there was no need generally to use this system, since the experiments were not covering the separated flow cases. The free stream velocity was about 26-29 m/s and the based on the chord length Reynolds number was about 0.7-0.75 million during the pressure and hot-wire measurements. The flap angle was choosen as 20° to give high pressure gradient on the flap with no separation. Angle of attack was changed between 0° to 14° by 2° of steps for single element case. 14° angle of attack excluded for flapped case, since the airfoil was stalled. Results The experimental results both for isolated and for flapped cases deviate from the potential flow calculations and indicate the importance of viscous effects at the current Reynolds number. The real flow calculations in isolated case coincide well with the experimental results indicating the success of the theoretical approach. The mean velocity and the turbulence profiles obtained from the hot-wire measurements in the wake of the main element on the upper surface of the flap are given on graphs to ve an insight about the flow field. The development of the wake of the main element on the upper surface of the flap can easily be seen from the results obtained. Especially the turbulence profiles show that the wake of the main element and the boundary layer on the upper surface of the flap do not mix considerably. Results can be used to obtain the maps of meaningful velocity and turbulence in the flowfield examined.