Search Results

Open jet wind tunnels are normally tuned to measure “correct” results without any modifications to the raw data. This is an important difference to closed wall wind tunnels, which usually require wind tunnel corrections. The tuning of open jet facilities is typically done experimentally using pilot tunnels and adding final adjustments in the commissioning phase of the full scale tunnel. This approach lacked theoretical background in the past. There is still a common belief outside the small group of people designing and using open jet wind tunnels, that - similar to closed wind tunnels, which generally measure too high aerodynamic forces and moments without correction - open jet wind tunnels measure coefficient too low compared to the real world. The paper will try to show that there is a solid physical foundation underlying the experimental approach and that the expectation to receive self-correcting behavior can be supported by theoretical models.

The paper describes the aerodynamic development and optimization process of the three different new models of the Audi A6/A7 family. The body types of these three models represent the three classic aerodynamic body types squareback, notchback and fastback. A short introduction of the flow structures of these different body types is given and their effect on the vehicle aerodynamic is described. In order to achieve good aerodynamic performance, the integration into the development process of the knowledge about these flow phenomena and the breakdown of the aerodynamic resistance into its components friction- and pressure drag as well as the induced drag is very important. The paper illustrates how this is realized within the aerodynamic development process at Audi. It describes how the results of CFD simulations are combined with wind tunnel measurements and how the information about the different flow phenomena were used to achieve an aerodynamic improvement.

Helmholtz-resonators are discussed in technical acoustics normally in conjunction with attenuation of sound, not with amplification or even production of sound. On the other hand everybody knows the sound produced by a bottle, when someone blows over the orifice. During the investigation of the sound produced in body gaps it was found that the underlying flow physics are closely related to the Helmholtz-resonator. But different from the typical Helmholtz-resonator generated noise – as for example the blown bottle or, from the automotive world, the sun roof buffeting – there is no fluid resonance involved in the process. For body gaps the random pressure fluctuation of the turbulent boundary layer is sufficient to excite the acoustic resonance in the cavity. The sound generation is characterized by a continuous rise in sound pressure level with increasing velocity, the rise is proportional to U with varying exponents.

Scaling laws - for example the variation of sound pressure with wind speed - are a key to the physical understanding of aeroacoustic phenomena. Aeroacoustics in Automotive applications differs from other fields of aeroacoustics: It is limited to low Mach numbers, the flow field is dominated by separated flows and the radiation into the far field is typically not of primary interest. On the other hand there are of course many common problems and findings shared with other fields in aeroacoustics. Therefore it is important to identify common areas with other, probably more advanced directions in aeroacoustics. But this has to be done without forgetting the practical demands of automotive application. Main sources for interior wind noise in vehicles are leakage noise, cavity noise and the noise generated by separated flows at the outer surface. All three of these noise sources will be investigated in this paper. Of special interest will be the dependence on the wind velocity.

Comparability of measured aerodynamic properties is important in the car industry. The understanding of test section interference is a key factor when testing ground vehicles in different wind tunnels. Special publications (SP) on this topic were prepared by SAE sub-committees in the past, where both closed wall and open jet test sections were covered. Although the information in these SP's is not outdated, a lot of new investigations have contributed to the existing knowledge. The aim of this paper is to give a brief survey of the available new literature, which was published afterwards. In order to give easy access to the information, the new literature is clustered in the paper according to the different approaches to the analysis of test section interference. The available benchmark data is ordered according to their origin (full scale, model scale, CFD).

Cooling drag is the increase in the total drag due to the internal flow in the cooling system. Because of the high flow resistance in the heat exchanger the momentum of the fluid needed for engine cooling usually is dissipated nearly completely. The resulting drag penalty can be approximated by the so called ram drag. For ground vehicles the cooling drag is typically lower than this approximation due to positive interference of the cooling flow with the general flow around the vehicle. Different mechanisms for the positive interference have been described in the literature. Inlet interference as well as outlet interference can result in significant reduction of the share of the cooling drag. Positive outlet interference is obtained, when the remaining kinetic energy of the cooling flow contributes significant thrust to the overall momentum balance.

In closed wall test sections the total correction to the measured drag usually consists of several parts: solid blockage corrections related to the displacement of the model, horizontal buoyancy corrections due to empty tunnel gradients and the wake blockage corrections, which are necessary to handle effects created by the displacement effect of the wake. The latter will be investigated in more detail in the paper. The wake blockage correction usually consists of two parts: a correction to the measured dynamic pressure (q-correction) and a gradient correction, the so-called wake induced drag increment. Both corrections are directly dependent on the source strength which is equivalent to the displacement effect of the wake. Therefore the displacement of the wake is analyzed in more detail.

For the aerodynamic development of an aircraft the induced drag is an important quantity and it has a significant impact on the design of the wing. The induced drag corresponds to the power requirement of the wing to generate the necessary lift. In many cases this is the dominant source of drag for aircraft. In ground vehicle aerodynamics the concept of induced drag up to now has attracted much less attention. This is partly due to the fact, that vehicle aerodynamicists usually optimize the vehicles to generate little or no lift. The second reason is that it is much more difficult for a ground vehicle to separate the total drag into the different contributions. During wind tunnel tests of vehicles with and without ground simulation some astonishing results were found, especially when comparing results for different rear end shapes.

In open jet wind tunnels with high blockage ratios a sharp rise in drag is observed for models approaching the nozzle exit plane. The physical background for this rise in drag will be analyzed in the paper. Starting with a basic analysis of the dependencies of the effect on model and wind tunnel properties, the key parameters of the problem will be identified. It will be shown using a momentum balance and potential flow theory that interaction between model and nozzle exit can result in significant tunnel-induced gradients at the model position. In a second step, a CFD-based investigation is used to show the interaction between nozzle exit and a bluff body. The results cover the whole range between open jet and closed wall test section interaction. The model starts at a large distance from the nozzle, then moves towards the nozzle, enters the nozzle and is finally completely inside the nozzle.

A region with floor boundary-layer suction upstream of the vehicle to remove the oncoming boundary layer is often used in automotive wind tunnels. These suction systems inevitably change the empty-tunnel pressure gradient. In this paper, the empty-tunnel pressure gradient created by the use of boundary layer suction and its effect on measured drag are investigated. By using excess suction - more suction than necessary to remove the floor boundary layer – it was possible to show experimentally that the major part of the drag increase due to boundary layer suction is created by unintended gradient effects. Only a minor part of the drag increase is due to the increased flow velocities at the lower parts of the vehicle, or in other words, due to the improved ground simulation. A theoretical model, using the concept of horizontal buoyancy to predict the gradient effect, is proposed. The model is compared to the experimental results as well as to CFD calculations.

The reference pressures are determined in automotive wind tunnels by measurement of pressures and pressure differences at upstream positions along the wind tunnel nozzle. For closed wall wind tunnels usually the so called nozzle method is used, where the volume flux is calculated from a pressure difference measured at the nozzle contour and a calibration factor determined in the empty test section. For open jet wind tunnels a choice is available between nozzle and plenum method. For the plenum method the reference static pressure is taken from the plenum chamber and the dynamic pressure also refers to the plenum conditions. The static reference pressure in closed wall tunnels is calculated by subtracting the dynamic pressure from the total pressure in the settling chamber. In this paper, the definitions and the differences between the two methods are discussed in detail.

The limited size of a wind tunnel test section can result in serious errors in the measured results. Therefore corrections for blockage effects are quite commonly used in automotive wind tunnel tests. A second type of test section boundary interaction apart from blockage is a biased flow angle due to lift of the test object. This wind tunnel effect is typically also corrected for in aerospace applications, but not in automotive tests. The reason for this is, that lift corrections are normally negligible for ground vehicles. The general validity of this statement will be demonstrated in this paper. In addition, two special cases will be discussed, where lift corrections are in fact of interest. The first example is the correction of measurements under yawed conditions. The lift correction is transformed into a side force correction in this case. In the second example, the effect of the test section height to width ratio on the lift correction for race cars is studied.

In a wind tunnel with simulation of the rotating wheels the measured lift is affected by pressure forces acting on the surface of the mini belts used to drive the wheels. The belt surface is part of the weighted model and the pressure forces therefore have to be subtracted to obtain the correct lift result. For stationary wheels this pad correction can be determined experimentally, but this is not possible for rotating wheels. Therefore a computational investigation was carried out using the CFD code FLUENT. The calculation is first verified for the case of stationary wheels and is then extended to rotating wheel case. The influence of a number of parameters as pad dimensions, position on the pad, vehicle shape and tire dimensions is studied. An approximate correction both for rotating and stationary wheels is given for the wind tunnel in question.

The classical theory of wind tunnel corrections calculated from potential flow theory is revisited. In this context a flow model uniformly valid for all types of test sections is developed for the correction of drag in automotive wind tunnels. To define and size the singularities setting up the flow model only geometrical properties of the model and measured force coefficients will be used. To achieve a correct representation of the flow about a vehicle body a number of improvements to the classical approach are proposed. Based on the uniformly valid flow model, correction formulae for closed wall, open jet and slotted wall test sections are given. For the open jet and slotted wall case it is shown, that the presented formulae are still incomplete, whereas for the closed wall case the correction is ready to use. The correction approach is validated step by step by comparison with appropriate experimental data.

Low-frequency pressure fluctuations which occur at certain flow speeds are an undesired feature in many open jet wind tunnels. This so called ‘wind tunnel pumping’ affects the aerodynamic quality of the flow and thus the quality of the measured data. In this paper a novel approach is presented to control the pulsation phenomenon by active damping of the acoustic resonant modes of the wind tunnel circuit. The acoustic mechanism of the resonance effect was investigated using a 1/20 scale pilot wind tunnel with a complete and detailed representation of test section, ducting, turning vanes and fan. The newly devised Active Resonance Control (ARC) System essentially consists of a microphone which picks up the pressure fluctuations in the plenum, a loudspeaker which is mounted in the tunnel wall and a time delay to adjust the phase relation between the microphone signal and the loudspeaker output.

Audi's new full scale aeroacoustic wind tunnel is under full operation now. The new facility is designed for full scale automotive testing of aerodynamics and aeroacoustics for vehicles up to 3 m2 frontal area at wind speeds up to 300 kph. The highlights are the unique ground simulation system with boundary layer suction and a 5-belt-system, and the extremely low background noise of only 60 dB(A) at 160 kph. First the background of the project is illustrated and the need for the special features of the tunnel is deduced form the industrial requirements. Then an overview of the facility design is given with a detailed description of the key technical components. The calibration of the self-correcting test section will be discussed and the physical background for it will be examined more closely. For the calibrated wind tunnel the results of two correlation tests including open jet as well as closed wall wind tunnels show a reasonable conformity.

This paper deals with the correction of aerodynamic interference effects taking place between the nozzle of an open jet wind tunnel and a test model. In order to deduce correct aerodynamic coefficients these interference effects have to be allowed for in the determination of the correct wind tunnel speed. In open jet wind tunnels basically two different methods are used to determine the tunnel speed. One is the so-called nozzle-method, utilizing the pressure difference down the nozzle to determine the nozzle exit velocity or tunnel speed. The other procedure is the so-called plenum-method, where the pressure difference between the settling chamber and the surrounding plenum chamber of the test section is measured and used. In this paper it is shown that both methods yield a systematic error, since the velocity distribution in the nozzle differs from the velocity distribution in an unbounded stream measured at the same distance from the model.

The present paper reveals the design concept as well as results of experimental investigations, which were conducted in the early design stage of the planned AUDI Aero-Acoustic Wind Tunnel. This low-noise open-jet facility, featuring a nozzle exit area of 11 m2 and a top speed of approximately 60 m/s, enables aerodynamic as well as acoustic testing of both, full-scale and model-scale ground vehicles. Ground simulation is provided by means of a moving-belt rig. The surrounding plenum is designed as a semi-anechoic chamber to simulate acoustic free-field conditions around the vehicle. Fan noise will be attenuated below the noise level of the open jet. The work reported herein, comprises 1/8-scale pilot-tunnel experiments of aerodynamic and acoustic configurations which were carried out at the University of Darmstadt.