Examination of robustness and accuracy of CFD simulations for external aerodynamics of commercial vehicles
Ladda ner
Typ
Examensarbete för masterexamen
Master Thesis
Master Thesis
Program
Applied mechanics (MPAME), MSc
Publicerad
2019
Författare
Josefsson, Erik
Modellbyggare
Tidskriftstitel
ISSN
Volymtitel
Utgivare
Sammanfattning
Commercial vehicles account for approximately 25% of the CO2 emissions from road vehicles in the EU. In order reduce the emissions levels new emission limits have been proposed. Since the aerodynamic drag is one of the dominating resisting forces acting on a truck at highway speed there is a need to improve the aerodynamic design of trucks. Given the limitations of experimental techniques, such as the large blockage created by trucks in available wind tunnels, numerical methods are often used for analyzing the ow field. However, in order to use numerical methods in the aerodynamic design process it is important to understand the accuracy and limitations of the method. This thesis investigates different methods for improving the correlation between numerical and experimental data. This is done by focusing on three different aspects. Firstly, different turbulence models are investigated. As a starting point a simplified truck model is used and a large number of different turbulence models are investigated. From this study a number of turbulence models will be identified, which are further evaluated using a fully detailed truck and compared to wind tunnel tests. It was found that the accuracy of the drag prediction is heavily dependent on the yaw angle, regardless which turbulence model is used, where a better correlation is achieved for larger yaw angles. Additionally it was found that the turbulence model greatly influences major ow structures such as the length and recirculation pattern of the trailer base wake. Finally, the benefits of using an IDDES compared to a RANS method will be highlighted. Comparing the yaw weighted average error of the drag prediction the IDDES methods resulted in approximately half the error of the RANS methods. Secondly, the modelling of the wind tunnel environment is investigated. The sensitivity of the predicted drag coefficient to the converging nozzle geometry, non-uniformity in the velocity distribution and test section turbulence levels are evaluated. It was found that the changes to the converging nozzle or turbulence levels did not alter the predicted drag significantly. Non-uniformity in the velocity field is investigated by adding a swirling component. There it was found that the swirling component acted similarly to an equivalent change of the yaw angle. A mesh study is performed in order to investigate and improve the level of mesh independence. By identifying regions with large gradients per cell a number of refinements are created and evaluated. The predicted drag was sensitive to refinements in areas such as the trailer base wake as well as in the tractor-trailer gap. However, no significant differences in drag prediction were obtained for refinements to regions such as the underbody and the rear-view mirrors.
Beskrivning
Ämne/nyckelord
Transport , Grundläggande vetenskaper , Hållbar utveckling , Strömningsmekanik , Transport , Basic Sciences , Sustainable Development , Fluid mechanics