Correlation of roof loads to wind speed and direction on a post-frame building in real time

In 2004 a post-frame structure with plastered straw bales as an in-fill wall system was built at the University of Manitoba. Load cells installed at the top and bottom of ten eave wall posts were intended to measure the tributary load transferred from the roof structure into the supporting posts. In 2011 wind speed and direction were measured adjacent to the structure and correlated to simultaneous load data. A linear regression model relating load to wind speed within four directional quadrants revealed that load behaviour was inconsistent with design-level loading prescribed by the National Building Code of Canada (2005). A second regression model with both speed and direction as independent variables did not determine any statistically significant relationships. This research concluded that the initial assumptions made in 2004 required additional scrutiny, including the conditions under which the load cells were calibrated, and the structural contribution of the walls' plastered skins.