AIJ guidelines on the applications of large-eddy simulation to pedestrian wind environment: Recommendations and validation benchmarks

This study introduces recent efforts of the Architectural Institute of Japan (AIJ) to develop guidelines for large-eddy simulation (LES) of pedestrian wind environments (PWEs). Reynolds-averaged Navier–Stokes (RANS) models have been widely used for predicting urban wind environments following best practice guidelines (BPGs) by Franke et al. [Int J Environ Pollut 44 , 1–4 (2011)] and Tominaga et al. [J Wind Eng Ind Aerodyn 96 (10–11), 1749–1761 (2008)]. Although RANS models can predict mean wind velocity and some turbulence statistics based on empirical assumptions, LES provides higher accuracy in resolving transient turbulence structures larger than the grid scale. With increasing urbanization, understanding instantaneous complex wind and wind-related phenomena around buildings is essential for ensuring pedestrian wind comfort and safety. However, LES applications face challenges owing to a lack of BPGs. This study outlines key recommendations for simulation setups and post-processing, including domain size, building modeling, grid generation, boundary conditions, turbulence modeling, discretization, convergence criteria, and reliability evaluation. Additionally, new benchmark cases are provided to support validation for PWEs. The AIJ working group systematically evaluated LES performance across urban scenarios to ensure practical applicability while balancing computational costs. These guidelines aim to enhance prediction reliability, thereby contributing to the standardization of LES applications for PWE and advancement of computational wind engineering. • The AIJ working group introduces LES guidelines for pedestrian wind environments. • The guidelines aim to describe key recommendations for simulation setups and post-processing. • The AIJ working group systematically evaluated the LES performance across various benchmarks. • Additional experimental cases are presented as benchmarks for verification and validation. • The database of appropriate inflow turbulence for several benchmarks is shared.