CFD-based multi-objective aerodynamic optimization applies to car rear diffusers.

• Ano de defesa: 2021
• Autor(a) principal: Peixoto, Denis Ramon dos Santos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Aerodinâmica; Computational fluid dynamics; Dinâmica dos fluidos computacional; Drag; Lift; Multi-objective optimizatio; Otimização multiobjetivo; Vehicle aerodynamics
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/3/3150/tde-04062021-111211/

This dissertation covers the aerodynamic optimization of a car geometry to minimize drag and lift coefficients. The lift coefficient plays an essential role in high-performance cars like race cars to improve the overall downforce and consequently maximize the car grip. More grip can provide more safety during car handling and increase the lateral acceleration of the vehicle during the corners, which is a factor that contributes decisively to reduce the overall lap time. By reducing the drag coefficient it is possible to achieve higher top speed and improve the car\'s fuel efficiency.The main objective of this thesis is to optimize the car geometry with respect to the aerodynamic performance through the CFD simulations. The geometry optimization entails the implementation of a rear diffuser because it is a highly efficient aerodynamic device. Due to the ground effect, the rear diffuser can generate a high downforce level without a substantial drag penalty.The research was divided into two parts. In the first one, simulations were performed using the Ahmed body model due to its simplicity and flow features. The Ahmed body was used as a benchmark to validate the numerical modeling and simulation methodology by comparing them with experimental results in the literature and developing all the optimization processes. The second part was based on a car geometry using the same techniques applied in the Ahmed body study. An optimization loop was built from the studies performed, integrating the MultiObjective Optimization (MOO) with CFD simulations to obtain the best geometry through an automated and reliable process.