A computationally efficient approach to resolving vehicle-induced turbulence for near-road air quality

Abstract

Near-road air pollution is worldwide public health concern, especially in urban areas. Vehicle Induced Turbulence (VIT) has a major impact on the initial dispersion of traffic-related pollutants on the roadways, affecting their subsequent near-road impact. The current methods for high-fidelity VIT simulations using computational fluid dynamics (CFD) are often computationally expensive or prohibitive. Previous studies adopted the TKE method, which models VIT as a fixed TKE volume source and produces turbulence uniformly in the computational traffic zones. This paper presents two novel methods, namely the Force method and the Moving Force method, to generate VIT implicitly by injecting a force source into the computational domain instead of physical vehicles in the domain explicitly, thus greatly reducing the computational burden. The simulation results were evaluated against experimental data collected in a field study near a major highway in Las Vegas, NV, which included collocated measurements of traffic and wind speed. The TKE method systematically overestimated the turbulence produced on the highway by converting the drag force completely into turbulence. This indicates that the TKE method, currently being used to implicitly model VIT in CFD simulations, requires major improvements. In comparison, the proposed Force and Moving Force methods preformed favorably and were able to capture turbulence anisotropicity and fluid convection. The Force method was shown to be a computationally efficient way to simulate VIT with adequate accuracy, while the Moving Force method has the potential to emulate vehicle motion and it is impact on fluid flow.