A Bivariate Statistical Model of Large Pressure Peaks Measured in Wind Tunnel Tests
摘要
A statistical approach for the treatment of large negative peak pressures is discussed. It is based on spatial correlation so that these large peak values, which are difficult to handle in cladding design, could be smoothed out by invoking the characteristic size associated with these events. The proposed method hinges on a mixture model of bicubic translation processes. Development of the method is summarized and applied to the computation of the area-average pressure on the building façade of a model tested in wind tunnel facilities at Politecnico di Milano.