Search Results

This paper presents an integrated numerical and experimental approach to take best possible advantage of the common development tools at hand (1D, CFD and wind tunnel) to determine the cooling air mass flow at the different vehicle development stages. 1D tools can be used early in development when neither 3D data nor wind tunnel models with detailed underhood flow are available. A problem that has to be resolved is the dependency on input data. In particular, the pressure coefficients on the outer surface (i.e. at the air inlet and outlet region) and the pressure loss data of single components are of great importance since the amount of cooling air flow is directly linked to these variables. The pressure coefficients at the air inlet and outlet are not only a function of vehicle configuration but also of driving velocity and fan operation. Both, static and total pressure coefficient, yield different advantages and disadvantages and can therefore both be used as boundary conditions.

The reliability of electronic components is of increasing importance for further progress towards automated driving. Thermal aging processes such as electromigration is one factor that can negatively affect the reliability of electronics. The resulting failures depend on the thermal load of the components within the vehicle lifetime - called temperature collective - which is described by the temperature frequency distribution of the components. At present, endurance testing data are used to examine the temperature collective for electronic components in the late development stage. The use of numerical simulation tools within Vehicle Thermal Management (VTM) enables lifetime thermal prediction in the early development stage, but also represents challenges for the current VTM processes [1, 2]. Due to the changing focus from the underhood to numerous electronic compartments in vehicles, the number of simulation models has steadily increased.

Cooling air flow is an important factor when it comes to vehicle performance and operating safety. In addition, it is closely linked to vehicle aerodynamics. In recent years more and more effort is being spent to optimize the losses generated by the flow through the vehicle. Grille shutters, better sealing and ducting are only some examples for innovations in this field of work, resulting in a lower contribution of the cooling air flow to overall drag. When investigating those effects, both experiments and numerical simulations are commonly used in the automotive environment. Still, when comparing the results from both methods, differences in the effect of cooling air flow can often be observed. To better understand the effects of cooling air flow, the ECARA Subgroup CFD decided to establish a common design for a generic open source vehicle model with a detailed underhood compartment to lay the foundation for a common investigation model.

In this article the unsteady aerodynamic properties of a 25% scale DrivAer notchback model as well as the influence of the wind tunnel environment on the resulting unsteady aerodynamic forces and moments under crosswind excitation are investigated using experimental and corresponding numerical methods. Research Institute of Automotive Engineering and Vehicle Engines Stuttgart (FKFS) swing® (side wind generator) is used to reproduce the essential properties of natural stochastic crosswind in the open jet test section of the Institute for Internal Combustion Engines and Automotive Engineering (IVK) model scale wind tunnel (MWK). The results show that the test environment of an open jet wind tunnel alters the amplitudes of side force and yaw moment under crosswind excitation when compared to an ideal environment neglecting wind tunnel interference effects.

Cooling air flow is an important factor when it comes to vehicle performance and operating safety. In addition, it is closely linked to vehicle aerodynamics. In recent years more and more effort is being spent to optimize the losses generated by the flow through the vehicle. Grille shutters, better sealing and ducting are only some examples for innovations in this field of work, resulting in a lower contribution of the cooling air flow to overall drag. But cooling air not only affects the internal flow of the vehicle but also changes the flow around it. This paper will show changes in the flow field around the generic DrivAer model resulting from cooling air flow, especially in the wake behind the car and in the region around the front wheels. The results were gathered using PIV measurements, multi-hole-probe measurements and pitot tube measurements in the 1:4 model scale wind tunnel of IVK University of Stuttgart.

In this paper the influence of different turbulent flow conditions on the aerodynamic drag of a quarter scale model with notchback and estate back rear ends is investigated. FKFS swing® (Side Wind Generator) is used to generate a turbulent flow field in the test section of the IVK model scale wind tunnel. In order to investigate the increase in drag with increasing yaw, a steady state yaw sweep is performed for both vehicle models. The shape of the drag curves vary for each vehicle model. The notchback model shows a more pronounced drag minimum at 0° yaw angle and experiences a more severe increase in drag at increasing yaw when compared to the estate back model. Unsteady time averaged aerodynamic drag values are obtained at two flow situations with different turbulent length scales, turbulence intensities, and yaw angle amplitudes. While the first one is representing light wind, the second one is recreating the presence of strong gusty wind.

As a consequence of reduced fuel consumption, direct injection gasoline engines have already prevailed against port fuel injection. However, in-cylinder fuel homogenization strongly depends on charge motion and injection strategies and can be challenging due to the reduced available time for mixture formation. An insufficient homogenization has generally a negative impact on the combustion and therefore also on efficiency and emissions. In order to reach the targets of the intensified CO2 emission reduction, further increase in efficiency of SI engines is essential. In this connection, 0D/1D simulation is a fundamental tool due to its application area in an early stage of development and its relatively low computational costs. Certainly, inhomogeneities are still not considered in quasi dimensional combustion models because the prediction of mixture formation is not included in the state of the art 0D/1D simulation.

For many years FKFS has operated the full-scale aeroacoustic wind tunnel of University of Stuttgart. To keep this wind tunnel as one of the most modern ones of its kind, it has again been upgraded significantly. The upgrade improved the aerodynamic as well as the aeroacoustic performance and accelerated the operational processes. Additionally, new innovative features have significantly enlarged the test capabilities. A new patented, modular belt system (FKFS first®) allows high performance measurements for race cars in a 3-belt mode as well as efficient measurements for production vehicle development in a 5-belt mode. The belt system is accompanied by a new, larger turntable and a new under-floor balance which enables high-accuracy measurements of forces and moments also for a high resolution in time. For the elimination of parasitic forces generated at the wheel drive units, a specific correction procedure has been implemented, which is patented, too (FKFS pace®).

In this paper the effect of aerodynamic modifications that influence the unsteady aerodynamic properties of a vehicle on the response of the closed loop system driver-vehicle under side wind conditions is investigated. In today's aerodynamic optimization the side wind sensitivity of a vehicle is determined from steady state values measured in the wind tunnel. There, the vehicle is rotated with respect to the wind tunnel flow to create an angle of attack. In this approach however, the gustiness that is inherent in natural wind is not reproduced. Further, unsteady forces and moments acting on the vehicle are not measured due to the limited dynamic response of the commonly used wind tunnel balances. Therefore, a new method is introduced, overcoming the shortcomings of the current steady state approach. The method consists of the reproduction of the properties of natural stochastic crosswind that are essential for the determination of the side wind sensitivity of a vehicle.

With its recent wind tunnel upgrade, FKFS installed the first interchangeable three-belt / five-belt-system (FKFS first®) in a full scale automotive wind tunnel. With the five-belt system, which today is a state-of-the-art ground simulation technique, the system is ideally suited for production vehicle development work. The five-belt system offers high flexibility, quick access to the underfloor and vehicle fixation, and setting the vehicle’s ride height by the restraint device. The first results of the five-belt system have already been published in SAE 2015-01-1557 [1]. The three-belt system on the other hand, offers a much more sophisticated ground simulation technique which is necessary especially for sports and racing cars. For such vehicles with very low ground clearances, it is important to have a more accurate ground simulation, in order to capture the same aerodynamic modes of action and response as on the road.

The DrivAer model, a detailed generic open source vehicle geometry, was introduced a few years ago and accepted widely from industry and academia for research in the field of automotive aerodynamics. This paper presents the evaluation of the aerodynamic properties of the 25% scale DrivAer model in both, CFD and in wind tunnel experiment. The results not only include aerodynamic drag and lift but also provide detailed investigations of the flow field around the vehicle. In addition to the available geometries of the DrivAer model, individual changes were introduced created by morphing the geometry of the baseline model. A good correlation between CFD and experiment could be achieved by using a CFD setup including the geometry of the wind tunnel test section. The results give insight into the aerodynamics of the DrivAer model and lead to a better understanding of the flow around the vehicle.

The aeroacoustic development of vehicles is still mainly carried out in wind tunnels under steady flow conditions, although the real situation is different. However, as discussed in several earlier publications, a vehicle experiences unsteady, turbulent flow on road, which results for example from natural wind, wakes of other vehicles, or obstacles at the roadside in combination with side wind. The resulting temporal variations of the wind noise inside the cabin affect the passengers’ comfort and safety through fatigue. To be able to also consider the unsteady aeroacoustics in the vehicle development process, a comprehensive method has been developed that is presented in full for the first time in this paper. The on-road situation is simulated in a realistic and reproducible manner in the full-scale wind tunnel of the University of Stuttgart by means of an active turbulence generator, developed by FKFS.

The wind tunnel is the standard tool in the development and improvement of vehicle aerodynamics. Usually, automotive wind tunnels contain an open test section, which results in a shear layer developing on the edge of the jet. This shear layer brings instabilities that can lead to resonance effects in the wind tunnel influencing the pressure distribution in the test section. To investigate the resonance effects, the classic wind tunnel corrections were applied to averaged drag measurements recorded in a resonance and nonresonance configuration of the model scale wind tunnel of the University of Stuttgart. The Mercker-Wiedemann-Method shows good compensation for the differing pressure gradients. Pressure measurements on the surface of the DrivAer Notchback model show different separation points on the rear window for measurements in resonance and nonresonance configuration. This means that the resonance effects can influence the separation significantly.

Since the introduction of the DrivAer model, an increasing amount of aerodynamic research and CAE method development activities are based on this detailed generic car body. Due to the Open Access nature of the model, it has not only been quickly adopted by academia but also by several automotive OEMs and CAE software developers. The DrivAer has delivered high quality experimental data to permit validation of existing aerodynamic CAE capabilities and to accelerate the development of new sophisticated numerical methods. Within the last decades, the registration number of SUV, especially in Europe, has increased significantly. Among other things, a large cross-sectional area, an increased ground clearance and larger wheels characterize this kind of vehicle. The DrivAer is not capable of depicting this vehicle category. Therefore, there is a demand for an expansion of this generic vehicle concept.

Numerical methodologies for aeroacoustic analyses are increasingly crucial for car manufacturers to optimize the effectiveness of vehicle development. In the present work, a hybrid numerical tool based on the combination of a delayed detached-eddy simulation and a finite element model, which relies on the Lighthill’s acoustic analogy and the acoustic perturbation equations, is presented. The computational aeroacoustics is performed by the software OpenFOAM and Actran, concerning respectively the CFD and the FEM. The aeroacoustic behavior of the SUV Lamborghini Urus at a cruising speed of 140 km/h has been investigated. The main aerodynamic noise phenomena occurring in the side mirror region in a frequency range up to 5 kHz are discussed. The numerical simulations have been verified against the measurements performed in the aeroacoustic wind tunnel of the University of Stuttgart, operated by FKFS.

Typical automotive research in wind tunnels is conducted under idealized, stationary, low turbulence flow conditions. This does not necessarily reflect the actual situation in traffic. Thus, there is a considerable interest to simulate the actual flow conditions. Because of this, a system for the simulation of the turbulence intensity I, the integral linear scale L and the transient angle of incidence β measured in full-scale tests in the inflow of a test vehicle was developed and installed in a closed-loop, closed test section wind tunnel. The system consists of four airfoils with movable flaps and is installed in the beginning of the test section. Time-series of the flow velocity vector are measured in the empty test section to analyze the system’s envelope in terms of the turbulence intensity and the integral length scales.

Natural wind, roadside obstacles, terrain roughness, and traffic influence the incident flow of a vehicle driven on public roads. These transient on-road conditions differ from the idealized statistical steady-state flow environment utilized in CFD simulations and wind tunnel experiments. To understand these transient on-road conditions better, measurements were performed on German public highways and on a test site. A compact car was equipped with a measurement system that is capable of determining the transient airflow around the vehicle and the vehicle’s actual driving state. This vehicle was driven several times on a predefined 200 km long route to investigate different traffic densities on public highways in southern Germany. During the tests the transient incident flow and pressure distribution on the vehicle surface were measured.

Automotive vehicles operate in complex, transient aerodynamic conditions that can potentially influence their operational efficiency, performance and safety. A moving-model facility combined with a wind-tunnel is an experimental methodology that can be utilized to model some of these transient aerodynamic conditions. This experimental methodology is an alternative to wind-tunnel experiments with additional crosswind generators or actively yawing models, and has the added benefit of modelling the correct relative motion between the vehicle and the ground/infrastructure. Experiments using a VW Golf 7 were performed with a 1:10 scale model at the moving-model facility at DLR, Göttingen and a full-scale, operational vehicle at the BMW Ascheim side-wind facility.

A computational study of the vehicle aerodynamics influenced by the wake of the rotating wheel taking into account a detailed rim geometry is presently performed. The car configuration corresponds to a full-scale (1:1) notchback configuration of the well-known ‘DrivAer’ vehicle model, Heft et al. [1]. The objective of the present work is to investigate the performance of some popular turbulence models in conjunction with different methods for handling the wheel rotation – rotating wall velocity, ‘multiple reference frame’ and ‘sliding grid algorithm’. The specific focus hereby is on a near-wall RANS eddy-viscosity model based on elliptic-relaxation, sensitized to resolve fluctuating turbulence by introducing a specifically modeled production term in the scale-supplying equation, motivated by the Scale-Adaptive Simulation approach (SAS, [2]), proposed by Krumbein et al. [3].

After being in operation since 1989, the 25% / 20% model scale wind tunnel of University of Stuttgart received its second major upgrade in 2016. In a first upgrade in 2001, a rolling road with a 5 belt system from MTS was installed. This system includes a steel center belt to simulate the road underneath the vehicle and four FKFS designed rubber belts for wheel rotation. The recent upgrade now enables the wind tunnel to be used not only for standard, steady state aerodynamic measurements but also for measurements of unsteady aerodynamic effects. This enables the use of the FKFS swing system as a standard measurement technique. Therefore, the former balance was replaced by a balance manufactured by AND with a high Eigenfrequency and the ability to sample the measurement data at up to 1000 Hz. The second large part of the upgrade was the replacement of the control system. With the new Wind Tunnel Control System (WCTS), control system.