Wind pressures on multi-level flat roofs of medium-rise buildings

Since there is limited information on design of wind pressures on multi-level flat roofs from previous studies and in wind loading codes, a wind tunnel experiment was carried out on 1:67 scaled models of a series of medium-rise buildings. Using an assembled cubic plexiglass model and uniformly distributed pressure taps, multi-level flat roofs with different configurations were easily fabricated and pressures on the roofs measured. Mean and fluctuating, local peak and area-averaged pressure coefficients were mainly investigated and stochastic characteristics of wind pressures on multi-level flat roofs were determined. Parameters of step geometry such as the effects of step dimension and configuration were considered. The results showed no significant difference between minimum values of negative peak pressures on simple flat roofs and on multi-level flat roofs, while negative pressure distributions and maximum values of positive peak pressures on the low-roof of multi-level flat roofs were highly dependent on step parameters. A comparison with ASCE/SEI7-10 indicates that negative values with small tributary areas are larger than those in related contents in the code, while negative values in other zones and positive values agree closely with the code. Gaussian and negative-skewed non-Gaussian PDFs, as well as positive-skewed non-Gaussian ones, were measured for multi-level flat roofs.

► Wind pressure on multi-level flat roofs was studied using the assembled plexiglass model.
► Minimum Cˇp is similar to a simple flat roof, and maximum C^p increases with step height.
► 5 zones for Cˇp and 2 zones for C^p are divided to evaluate the effect of step geometry.
► Comparisons show Cˇp,area with small areas larger than code, and C^p,area agrees with code.
► Gaussian, negative-skewed and also positive-skewed non-Gaussian PDFs were measured.