scholarly journals The aerofoil in a wind tunnel of elliptic section

The lift and drag experienced by an aerofoil in a wind tunnel differ from the lift and drag experienced by the same aerofoil under free air conditions. These differences, which are the induced effects due to the walls of the enclosure, can be determined by the aid of general considerations laid down by Prandtl. In a closed tunnel, that is, a tunnel with rigid walls, the necessary boundary condition is that the velocity normal to the walls shall be zero. In an open tunnel, or free jet, the condition is that the pressure is constant over the boundary. Assuming that trailing vortices spring from the aerofoil and extend downstream without distortion, Prandtl has shown that the problem can be converted into one dealing with the flow in a section of the wake far behind the aerofoil, the necessary boundary condition being that the velocity potential is constant over the trace of the open tunnel. Prandtl ( loc. cit .) himself has investigated the interference experienced by an aerofoil in a tunnel of circular section for an elliptic distribution of lift across the span. Glauert, to whom a considerable extension of the theory is due, found approximate values of the induced drag in a rectangular tunnel when the span of the aerofoil is indefinitely small. Terazawa modified Glauert’s method and obtained the exact solution for an aerofoil with uniform distribution of circulation in a rectangular channel. Rosenhead obtained exact results for uniform and elliptic distributions both in circular and rectangular tunnels. More recently, in connection with the building of a wind tunnel of elliptic section, Glauert was led to reconsider the general problem of wind tunnel interference, and his conclusions are embodied in three valuable papers. In the first of these he pointed out that the problem discussed by previous investigators is that in which the lift distribution is prescribed to be the same as that in free air, and the aerofoil is twisted in the tunnel to a position in which this distribution is maintained. In general, if the aerofoil is not twisted in this way, there is a change in the distribution of circulation. If this change is taken into account, Glauert has shown for a tunnel of circular section “that the formulæ derived from the assumption of elliptic distribution of lift are sufficiently accurate for all conventional shapes of aerofoil, but that those derived from the assumption of a uniform distribution over-estimate the effect of increasing span of the aerofoil.”

The effect of the walls of the enclosure on the measured values of the lift and drag experienced by an aerofoil is quite appreciable and it has been known for a considerable time that correction must be applied to wind tunnel result before they can be applied to free air conditions. Prandtl* investigated the effect on an aerofoil in a free jet or circular tube both in the case where there is a uniform lift distribution, and in the case where there is an elliptic distribution of circulation. The elliptic distribution is of importance because it gives the minimum drag for a given lift. Glauert by means of an approximate method found the induced drag and lift in a rectangular channel when there is a uniform distribution of lift. Terazawa modified Glauert’s method and obtained the exact solution for an aerofoil with uniform distribution of circulation in a rectangular channel. It is The object of this note to extend these results and to obtain the induces drag and lift in a rectangular channel when there is an elliptic distribution of lift. In addition, the discussion of Prandtl is briefly gone through because Prandtl’s results are usually given as the first few terms of an infinite series, and it has not been noticed that the result can be obtained exactly. Glauert’s work on the effect of plane barries is briefly re-examined because, in his analysis, approximate expression were summed over an infinite series of points, and at first glance it appeared that this would introduce some error of the same order as the result. In this note the summation is carried out rigorously and the approximations to the actual values. The small divergences from Glauert’s result obtained by Terazawa in two numerical cases are, in effect, the result of a slightly more accurate approximation. From the practical point of view the results of this paper add little to what is known already, for the major corrections are given by the results of the approximate methods, but this note should fill in some small gaps in the theory of wind tunnel interference.


It is advantageous from many points of view to make test models as large as possible. One method of doing this is to measure the characteristics of half the model in existing wind tunnels. One half of the aerofoil is mounted horizontally on a vertical reflexion plate and the plate is placed in a suitable position in an open jet which, in the undisturbed state, is of circular section. The contour of the jet is distorted, especially with models of large semi-span, but this distortion is neglected in the analysis. The correcting factor associated with ‘uniform’ distribution of lift is worked out exactly and that associated with ‘elliptic’ distribution approximately. The effect of the induced downwash on the distribution of lift is ignored. The results are given in suitable tables and figures. Throughout the working range of normal experiments the correcting factor is of the same order of magnitude as that obtaining when a full model is tested in a jet of circular section.


This paper arose out of a discussion with Mr. Glauert on the validity of certain results on wind-tunnel interference given in a recent paper by Theodorsen, and I am greatly indebted to him for several very valuable suggestions. In his paper Theodorsen investigated, on the lines of the approximate theory laid down by Glauert, the interference factors due to rectangular tunnels in the following five conditions: (1) tunnel entirely enclosed; (2) free jet; (3) horizontal boundaries only, the vertical sides being free surfaces; (4) vertical boundaries only, the horizontal sides being free surfaces; (5) bottom boundary only, the remaining sides being free surfaces. Apart from one or two minor errors, Theodorsen's paper has to face a more serious objection on the score that the approximate method employed is not always valid. This objection to the approximate method was raised rather briefly in a previous paper, but at that time all the cases discussed, using the approximate method, were those due to Glauert, and they were such that the approximate method did not come into conflict with the exact theory. The more complicated method was, however, of some practical importance as it enabled additional refinements to be introduced into tire numerical evaluation of the interference factors, and it enabled Glauert to show that there is an appreciable drop in the interference factor in the Duplex Tunnel (height/breadth equal to one half) as the span of the aerofoil increases. Some of the cases discussed by Theodorsen are such that they must be considered according to the exact theory. The crux of the matter is as follows:— The basis of the theory of wind-tunnel interference is due to Prandtl, and he investigated several cases himself. On the assumption that trailing vortices spring from the aerofoil and extend downstream without distortion, Prandtl showed that the whole problem can be converted into one dealing with the flow in a transverse section of the wake far behind the aerofoil, the necessary boundary conditions being that the velocity potential is constant over the surface of the free jet, and the stream function constant over the rigid boundary. The theory was extended by Glauert and others, and applied particularly to the case of the small aerofoil in a rectangular tunnel. When the aerofoil is very small the exact distribution of vorticity is not of extreme importance as, to a first approximation, they all give the same result.


Author(s):  
Alessandro Bianchini ◽  
Francesco Balduzzi ◽  
John M. Rainbird ◽  
Joaquim Peiro ◽  
J. Michael R. Graham ◽  
...  

Accurate post-stall airfoil data extending to a full range of incidences between −180° to +180° is important to the analysis of Darrieus vertical-axis wind turbines (VAWTs) since the blades experience a wide range of angles of attack, particularly at the low tip-speed ratios encountered during startup. Due to the scarcity of existing data extending much past stall, and the difficulties associated with obtaining post-stall data by experimental or numerical means, wide use is made of simple models of post-stall lift and drag coefficients in wind turbine modeling (through, for example, BEM codes). Most of these models assume post-stall performance to be virtually independent of profile shape. In this study, wind tunnel tests were carried out on a standard NACA0018 airfoil and a NACA 0018 conformally transformed to mimic the “virtual camber” effect imparted on a blade in a VAWT with a chord-to-radius ratio c/R of 0.25. Unsteady CFD results were taken for the same airfoils both at stationary angles of attack and at angles of attack resulting from a slow VAWT-like motion in an oncoming flow, the latter to better replicate the transient conditions experienced by VAWT blades. Excellent agreement was obtained between the wind tunnel tests and the CFD computations for both the symmetrical and cambered airfoils. Results for both airfoils also compare favorably to earlier studies of similar profiles. Finally, the suitability of different models for post-stall airfoil performance extrapolation, including those of Viterna-Corrigan, Montgomerie and Kirke, was analyzed and discussed.


2016 ◽  
Vol 2 (3(4)) ◽  
pp. 2
Author(s):  
Stanisław Ziętarski ◽  
Stanisław Kachel ◽  
Adam Kozakiewicz

Topics below are rather undesired, but important, outcome not yet completed research on the aircraft airfoils, turbine and compressor blades, parametric design of airfoils, establishing the relationships based on the results of experiments in a wind tunnel, developing databases for determining the relationships between airfoil parameters and lift and drag coefficients. Reliable database created as a result of the research work allows to simulate the wind tunnel. Very early on, however, was necessary to extend the developed specialized software for a new applications, and it meant the need for generalization of software, e.g. for gas turbine engines, propellers, etc. But after some time it turned out, that in order to achieve the required accuracy, the changes are needed in the underlying assumptions, set decades ago. In addition, coordinate measuring machines and systems, and associated software were not always as accurate as expected. Concepts how to solve it and develop software carrying out these tasks are presented in the article. It is like to withdraw from the old path and look for a new path that will lead to the reliable data base. Processes related to air or gas flow should be similarly defined in all the specialized software applications (e.g. aircrafts and turbine engines). Accuracy (10-9 mm) achieved in virtual measurements within the integrated system can be used to verify the results of CMM and other measuring systems, provided that an appropriate software has been developed.


2003 ◽  
Author(s):  
Jeffrey Hoffman ◽  
Bill Martindale ◽  
Stephen Arnette ◽  
Jack Williams ◽  
Stan Wallis

Author(s):  
Christopher L. Rumsey ◽  
Jan-Renee Carlson ◽  
Judith A. Hannon ◽  
Luther N. Jenkins ◽  
Scott M. Bartram ◽  
...  

2015 ◽  
Vol 12 (3) ◽  
pp. 261-270
Author(s):  
Albert Boretti

The paper proposes a study of a GT2 racing car with a computational fluid dynamic (CFD) tool. Results of STAR-CCM+ simulations of the flow around the car in a wind tunnel with movable ground and wheels are presented for different air speeds to assess the different contributions of pressure and shear to lift and drag over the speed range. The rear wing contributes more than 85% of the lift force and 7-8% of the drag force for this particular class of racing cars. When reference is made to the low speed drag and lift coefficients, increasing the speed from 25 to 100 m/s produces an increase of CD of more than 3% and a reduction of CL of more than 2%. The resultsuggests modifying the constant CD and CL values used in lap time simulation toolsintroducing the tabulated values to interpolate vs. the speed of the car.


Sign in / Sign up

Export Citation Format

Share Document