Uncertainty quantification and race car aerodynamics

Abstract

Car aerodynamics are subjected to a number of random variables which introduce uncertainty into the downforce performance. These can include, but are not limited to, pitch variations and ride height variations. Studying the effect of the random variations in these parameters is important to predict accurately the car performance during the race. Despite their importance the assessment of these variations is difficult and it cannot be performed with a deterministic approach. In the open literature, there have been no studies dealing with this uncertainty in car racing aerodynamics modelling the complete car and assessing the probability of a competitive advantage introduced by a new geometry. A stochastic method is used in this work in order to predict the car downforce under stochastic variations and the probability of obtaining a better performance with a new diffuser geometry. A probabilistic collocation method is applied to an innovative diffuser design to prove its performance with stochastic geometrical variations. The analysis is conducted using a complete three-dimensional computational fluid dynamics simulation with a k–ω turbulence closure, allowing the performance of the physical diffuser to be more accurately represented in a stochastic real environment. The random variables included in the analysis are the pitch variations and the ride height variations in different speed conditions. The mean value and the standard deviation of the car downforce are evaluated.