Real Life Car Aerodynamics

Abstract

Vehicles have become essential in our lives and they need to satisfy the ever – increasing demands of our society. Fuel efficiency has become the most important factor leading the companies to build light cars as hybrid power units are still in experimental phase. This causes the cars to be more sensitive to aerodynamic loads. Moreover, aerodynamic forces and distributions are obtained in wind tunnel facilities where the conditions are controlled and differ from the real life ones. Another equally important aspect, is the fact that there is a variety of used numerical methods with the most complicated ones being the modelling of the rotation of wheels. To address this topic, the differences between real – world conditions and wind tunnel ones using different wheel modelling techniques are assessed. The detailed Fastback DriVaer geometry is used and transient simulations utilizing Improved Delayed Detached – Eddy Simulation scheme are conducted in ANSYS Fluent. Only a handful of previous studies involved investigations at realistic cruise speed (100km/h) while the used wheel modelling methods in the project are considered to be state-of-the-art. In detail, body forces are obtained using three wheel models with the simplest one being the rotating boundary condition. Flow structures are studied utilizing the Multiple Reference Frames method and the Sliding Mesh technique. The implementation of each model is discussed as certain limitations apply to them. The differences between these models are underlined while the latter two seem to have more similarities even though all results are comparable. Regarding the real-world conditions, statistical data related to highway average vehicle speed provided by the Department of Transport of United Kingdom was taken into account. Also, a real life crosswind gust is implemented with special attention paid to the turbulent properties. Initially, three meshes were generated and a grid sensitivity study was undertaken. The overall results showed that the wheel modelling methods affect all parts of the car while the main body is strongly affected by each method accordingly. Differences are obtained at flow structures generated at the wheels while the effect of the models is decreased when real world conditions are applied. At these conditions, an increase of 20% is noticed in the drag while the lift is affected immensely and is increased by 80% approximately. Significant differences are obtained at the flow structures close to the car, such as A pillar vortices, with the overall flow moving to the direction of the gust.