Improving urban wind-flow prediction in complex built environments using a Cut-cell Cartesian grid CFD model

Abstract

This research explores the implementation of a Cut-Cell Method (CCM) within a Cartesian-grid CFD framework to mitigate the geometric distortions of building boundaries not aligned with the grid. By avoiding excessive grid refinement, CCM offers an efficient alternative for high-fidelity urban wind modeling. Performance was validated against AIJ wind-tunnel experimental data for Niigata, covering 80 points across 16 inflow directions. Comparisons with the conventional Stair-Step Method (SSM) demonstrate that CCM significantly enhances prediction accuracy. Quantitatively, the domain-averaged Index of Agreement increased by 18%, while RMSE and Mean Bias decreased by 18% and 55%, respectively. Detailed analysis reveals that while SSM creates artificial eddies and constricts street canyons, CCM more realistically captures building corners and spacing. However, in convergence and reattachment zones near tall and mid-rise buildings, CCM tends to overpredict reattachment lengths and enlarge secondary vortices, leading to localized wind speed underestimation. Despite these specific deviations, the method successfully brings all evaluated statistical indicators within recommended ranges. Overall, CCM provides a superior representation of complex urban morphology, proving essential for urban ventilation assessment, wind-corridor planning, and pollutant dispersion analysis. Future research should further evaluate this method under thermally stratified conditions to broaden its practical applicability.

 

Acknowledgments

This study was carried out with the support of 'R&D Program for Forest Science Technology '(Project No. "RS-2025-25404070")' provided by Korea Forest Service(Korea Forestry Promotion Institute).

Key words: CFD model, AIJ wind tunnel validation, cut-cell method, stair-step method