Models For Vehicle Aerodynamics

Project Information

MOVA

Grant agreement ID: BRPR980624

Project closed

Start date 1 April 1998

End date 30 September 2001

Funded under

• Specific research and technological development programme in the field of industrial and materials technologies, 1994-1998

Total cost

No data

EU contribution

No data

Coordinated by

Technische Universiteit Delft
Netherlands

CORDIS provides links to public deliverables and publications of HORIZON projects.

Links to deliverables and publications from FP7 projects, as well as links to some specific result types such as dataset and software, are dynamically retrieved from OpenAIRE .

Exploitable results

Models for Vehicle Aerodynamics (MOVA)

The prospects of designing a car or other vehicle shape using solely computer simulation without performing wind tunnel experiments on a scale model, has long been a challenge for car manufacturers. Nowadays many, if not all, car manufacturers use Computational Fluid Dynamics (CFD) packages. However, although viewed as potentially a major tool for design and optimisation in many branches of industry, CFD has not yet reached the level of reliability and confidence where it could be used exclusively as a design tool. One of the major difficulties is the proper simulation of turbulence. Because of the great complexity of the car geometry and the still limited computational resources, numerical simulations by the car industry are based on solving the averaged equations of fluid motion (Reynolds-averaged Navier-Stokes, "RANS", approach), which involve different approximations ("modelling") of various flow and turbulence phenomena. In the automotive industry, expensive and time consuming large-scale wind tunnel measurements are still unavoidable for the final shaping of the automobile body, whereas the CFD assists in the preliminary development phase and for qualitative exploration of effects of various design parameters. A reliable CFD tool, if available, would reduce both the costs and time of the research and design, and accelerate the appearance of new designs on the market. Hence, the further improvement of CFD is of primary interest to all automobile industries. The goal of the MOVA project (Models for Vehicle Aerodynamics) was to develop a computational turbulence model (or models) for the road-vehicle industries, which should make it possible to perform accurate and efficient computations of external flow over road vehicles. With such a model, the CFD codes are expected to become a reliable tool giving relatively inexpensive and fast computations in the design process as well as to opening a route for interactive CFD - wind-tunnel optimisation. The objectives were pursued by the joint efforts of six partners, three from academia and three from industry. The research tasks consisted of developing, refining and validating advanced RANS models, supplemented with new experiments and Large Eddy Simulations (LES). The optimal turbulence model was sought among three model classes: nonlinear eddy viscosity models (NEVM), three-equation elliptic-relaxation eddy-viscosity models (ERM) and the differential second-moment (Reynolds-stress) closure models (DSM). Some new simple improvements of the standard linear eddy-viscosity model (EVM), as well as a new hybrid EVM/DSM (HTM) have also been derived and tested. Besides model refinement, new wall functions were also developed, which should permit the use of high-Reynolds-number model versions in complex flows. The focus of the development was especially the rear end of a car body, which greatly affects the vehicle drag. Model refinements were validated in several generic test flows relevant to vehicle aerodynamics where experimental data exist. New experimental data were also generated in two dedicated experiments: detailed mean flow and turbulence measurements in flow over a generic car afterbody (Ahmed car model) in a laboratory wind tunnel, and velocity and pressure measurement in flow over a simplified 1/5 scale car model (PSA model) in an industrial wind tunnel. The final validation of the developed models in the computation of flows over both simplified car models (Ahmed and PSA models) showed notably improved agreement with experiments, as compared with the standard linear eddy viscosity models. In parallel to the RANS model development, a new unstructured finite-volume solver for Large Eddy Simulations (LES) was developed and used to simulate flow over a simplified car mirror. The LES code is envisaged as an additional tool for optimising the shapes of car mirrors and other vehicle protrusions, as well as to generate input for computing aerodynamic noise.