Integrated Evolutionary Algorithms/Computational Fluid Dynamics for Drag Reduction in Highway Design

Abstract

Reducing ground vehicle emissions can have substantial economic and environmental benefits. A possible way to promote this goal is by redesigning highway infrastructure to reduce aerodynamic drag on vehicles. This paper introduces an integrated computational framework that couples an evolutionary algorithm with computational fluid dynamics (CFD) to redesign highway noise barrier walls to include aerodynamic drag reduction. An existing highway section is captured via terrestrial laser scanning, which is taken as the reference geometry for targeted modifications, in the form of a periodic array of lateral slabs. The aerodynamic drag is quantified through CFD. The minimization of both the aerodynamic drag on moving vehicles and the construction cost of the proposed modifications are considered as the objectives of the optimization. The approach iteratively evolves an initial population of constrained designs toward an optimal Pareto set. A potential improvement of aerodynamic drag between 0.4% and 1.3% is shown to be achievable for a cost ranging between USD 43,000 and 47,000 per kilometer. Ultimately, this could beget a reduction of CO2 emissions of up to 0.87% per unit distance traveled.