Search Results

Numerical and modeling errors in computational aerodynamics consist of multiple components. Previous investigations at Volvo have shown that low Reynolds k-ε models generally give better levels in pressure over the rear base area of the car than the corresponding wall function based model. However, these computations were carried out on car shapes without wheels. This paper presents numerical simulations of the flow field around three versions of the Volvo validation car series (VRAK). The geometry is a typical car with flat floor and simplified tires. The three car models differ by their rear shape. The configurations are: one with a nearly flat base, a fastback with a sloping rear window, and a car with a roof wing. The influence of the near wall formulation of the standard k-ε model on drag and lift is investigated. The performance of the low Reynolds number version of the cubic k-ε model by Suga [7] is also investigated.

Validation studies often contains uncertainties due to boundary condition differences between computational fluid dynamics, CFD-, and experimental fluid dynamics, EFD-, in the setup for the cases involved. The present study is in two parts where the first part is a 2D parametric study of two geometrical parameters and their variation with blockage. The parameters were radius in the transition region from the windscreen to the roof and the angle of attack for the windscreen. A mesh blending technology has been used where an interpolation between two meshes is conducted on a vertices level for the creation of a mix between the two grids. Many CFD simulations have been done of which some are presented. The windscreen angle was more sensitive to blockage in the sense that trends in drag from two configurations may be erroneously predicted if large blockage is encountered.

The accuracy of computational fluid dynamics, CFD, has improved considerably over the years but still, large errors are present and vehicle parameters such as drag and lift are often poorly predicted. The current work is investigating how transient CFD would cope with a very complex flow structure around a surface mounted cube. A transient Reynolds averaged Navier Stokes model, RANS model, is presented together with a large eddy simulation model, LES model. Furthermore, two “industrial like” test cases have been simulated using a transient RANS model.

The demand for more energy efficient vehicles is driven by environmental considerations and alternative engine technology. In order to reduce fuel consumption on future vehicles the power needed to propel the car has to be lowered. Hence, considerable efforts are needed to improve the aerodynamics. For a modern vehicle the potential for further improvements on drag is mainly to be found in the underbody region, Howell (1991). This requires more knowledge of the underbody flow and the flow around the wheels. In the present work the flow in the underbody region has been studied using a combination of experiments and calculations to obtain a more comprehensive database. The model chosen for this work was the so called ASMO model from Daimler Benz, which is a well known geometry that is available for the public on the internet. A simple model was preferred since the goal was to study the basic mechanisms behind drag generated by the underbody flow.

The numerical and modeling errors in computational aerodynamics consist of several components. The errors have been shown to be large at the front in the stagnation region and in the rear in the base region, see Ramnefors et. al. (1996). This paper presents results from numerical simulations of the flow around a simplified Volvo FH truck. Experimental data have been created for comparison in the Volvo Wind-tunnel. Simulation of the flow around the truck using several turbulence models and numerical schemes are compared to experimental measurements of drag, surface pressure distribution and total pressure distribution in the wake. It was found that the numerical scheme is very important for the base pressure. Several turbulence models have been tested. The models range from simple linear eddy viscosity models, EVMs, to more complex models such as non-linear EVMs and Reynolds stress transport models. The errors are shown to be surprisingly large along the roof of the truck.

The numerical and modeling errors in computational aerodynamics consist of several components. The errors are particularly large at the front in the stagnation region and in the rear in the base region. This paper presents results from numerical simulations of the flow around the Volvo ECC, see figure 1, and a generic vehicle shape. Several possible sources for the base pressure error are investigated. It was found that boundary conditions and mesh resolution affected the base pressure. Several turbulence models have been implemented and tested. The models range from a simple linear eddy viscosity model, EVM, to more complex models such as non-linear EVMs, explicit algebraic Reynolds stress models and Reynolds stress transport models.

Computations of the flow around a basic car shape, the Volvo ECC, have been done using different turbulence models: two lowRe k-ε model, the standard k-ε model and the IP and Gibson-Launder version of the Reynolds Stress Model (RSM). Results from the different computations are compared to windtunnel measurements of drag, pressure and velocity. The effect of the turbulence model on the accuracy of predicted pressure and drag is studied. To provide a basis for this work grid refinement studies have been performed to study the effect of mesh resolution and mesh quality on the error in pressure and drag. It is shown that mesh quality is important to preserve the order of the difference schemes but that the grid related error may be virtually eliminated. Further the computations show that a lowRe k-e model yields the best prediction of experimental data.