Tornado Impacts on Communities
Abstract
Low-rise wood-frame residential buildings, which constitute most buildings affected by tornadoes, are highly vulnerable to tornado-induced damage. Evidence indicates that the wind velocities impacting these buildings during tornadoes are often comparable to hurricane winds, for which effective design provisions and mitigation measures exist. In this research, pressures were measured at the WindEEE Dome during interactions of tornado-like vortices (TLVs) with a scaled model of part of the Dunrobin community, struck by an EF3 tornado in September 2018. Three swirl ratios were tested with TLVs moving along various paths and translation velocities. Measured loads were compared with design loads specified in Chapter 32 of ASCE/SEI 7-22 for different internal pressure scenarios simulated using the multiple discharge equations (MDE). An iterative method was developed to estimate tornado-velocity gust factors from wind tunnel data. Results showed that uplift forces on the entire roof in scenarios with one dominant opening were 44\% to 63\% higher than in the distributed leakage scenario, emphasizing the importance of maintaining the envelope's integrity. Revised values for internal and external pressure coefficients or correction factors could enhance the standard’s ability to provide safer design loads. ASCE/SEI 7-22 uses pressure coefficients from atmospheric boundary layer (ABL) wind tunnels to evaluate tornado-induced loads. This approach is not supported by evidence, as most studies suggest pressure distributions differ between tornadoes and straight-line winds. However, this study's comparison of ABL and TLV-induced pressure coefficients revealed that for radial distances larger than 0.6 times the core radius, TLV-induced peak pressure coefficients are lower in absolute value than those from ABL winds. A 1.4 safety factor applied to ABL coefficients sufficiently accounts for most loading scenarios at these distances. Finally, a method was developed to stochastically simulate the pressure field induced by TLVs on the model buildings. This approach could reduce testing costs by enabling data extrapolation for Monte Carlo simulations. While the method shows promise in predicting the empirical extreme value distribution (EEVD) of minimum pressures, further work is required to validate and improve it.