Modelling downburst velocity fields in relation to Main Wind Force Resisting Systems

• Contributions of Kareem and Solari to study of downburst winds • The velocity slowly-varying part is expressed as vertical and time variation • A sine wave models the time variation, OBV and Wood & Kwok model the height variation • Models’ application to large-scale experimental downburst signals at WindEEE Dome • Models suitable to reproduce downburst interactions with ABL, storm motion, roughness Over the past two decades, wind engineering has focused on non-synoptic wind storms, which exhibit greater spatio-temporal complexity than synoptic scale winds. Here we focus on the modelling of downburst velocity fields in relation to the way these models can be used to determine structural responses to downburst winds. Two approaches have been defined in relation to the Main Wind Force Resisting Systems (MWFRS): (i) the Gust Front Factor (GFF) from Professor Ashan Kareem’s group at Notre Dame University in USA and (ii) the Thunderstorm Response Spectrum Technique (TRST) from the late Professor Giovanni Solari’s team at the University of Genova in Italy. Both methods decompose the downburst mean wind field into a vertical profile and a time variation. Here we focus on the modelling of downburst velocity fields in terms of spatial and time characterizations including the effects of translation, atmospheric boundary layer (ABL) superposition and surface effects. Herein, we focus on analysing how various analytical models, that include both time and space velocity variations, can be adapted to fit an experimental database of downburst-like flows (DLFs), generated at WindEEE Dome at Western University under the project THUNDERR. The calibration is conducted through the superposition of stationary mean flow fields with the effects of translation, background synoptic wind and surface roughness. Analytical models for the profile variation, the Oseguera-Bowles-Vicroy (OBV) and Wood & Kwok models, along with the sine wave models for temporal variation, are tested against this extensive database. The resulting adapted proposed models provide a potential frame on DLFs to be applied in the context of MWFRS approaches.