Abstract
This paper presents the requirements in order to simulate drifting snow around buildings in a wind tunnel. The threshold friction velocity ratio and the angle of the repose of the particles should match in both the wind tunnel model and the prototype situations. The real snow particles and the large particles with artificial snow in the wind tunnel did not follow the flow around the building model. Stokes parameter in the wind tunnel should be less than one in order to follow the flow around the building model. If the model particles were very fine, the terminal velocity ratio was not necessarily matched for both the model and prototype. By the large turbulence generated by buildings, particles were easy to transit from the saltation motion to the suspension motion at a low wind speed of 5-6m/s. As a result, the saltation motion was not the principal mechanism of the particle transport process near buildings. The Froude number requirement was then relaxed from the wind tunnel similarity pa-rameter.
