TOWARDS EFFICIENT SCALE-RESOLVING SIMULATION FOR URBAN POLLUTANT DISPERSION: A COMPARATIVE STUDY OF RANS AND OPTIMIZED LES SOLVERS

This study compares RANS and LES models for urban pollutant dispersion. While RANS struggles with unphysical wake regions, LES offers high accuracy but at a prohibitive computational cost. To address this limitation, we evaluate an implicit pressure-velocity block-coupled solver for LES. By solving equations simultaneously and allowing larger time-steps, this approach reduces simulation time by 30% to 43% compared to standard segregated methods like SIMPLE. This optimization makes high-fidelity scale-resolving simulations significantly more cost-effective and viable for industrial applications, effectively balancing computational speed with physical accuracy.