Application of Adjoint Methods on Drag Reduction of Current Production Cars

Automotive manufacturers are facing stronger and stronger pressure to optimize all aspects related to fuel consumption of cars, and aerodynamic drag makes no exception, due to increasing government enforcing rules for the reduction of the emissions and the increasing influence of aerodynamic performance on fuel consumption with WLTC and RDE driving cycles. Nowadays, CFD simulation is a common tool across automotive industries for the assessment and the optimization of vehicle resistance in the design phase. The full power of these numerical methods of studying many design variants in advance of experimental testing, however, can be fully exploited when coupled with optimization techniques, always keeping into account constraints and aesthetical demands. On the other hand, a massive use of CFD optimization can lead to unaffordable computational efforts or a limitation of the design exploration space. The exploration of new time-saving CFD and optimization techniques are necessary in order to overcome these limitations; for this reason, a gradient-based optimization using adjoint method applied to road vehicle aerodynamic drag reduction has been evaluated. The main advantage of this method is the independence from the number of analyzed design variables, allowing to explore a potential infinite number of geometrical parameters with turn-around times compatible with product development timetables. In this paper, methodology and application of the adjoint solver on full production vehicles are presented, showing how surface sensitivity on drag generation can drive design decisions and can be coupled with optimization workflows.


  • English

Media Info

  • Media Type: Digital/other
  • Features: Figures; References; Tables;
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 01687304
  • Record Type: Publication
  • Source Agency: SAE International
  • Report/Paper Numbers: 2018-37-0016
  • Files: TRIS, SAE
  • Created Date: Nov 29 2018 9:51AM