The primary objective of this collaborative project between Texas Tech University (TTU) and Iowa State University (ISU) was to conduct a fundamental study to characterize, model and analyze uncertainty of tornado winds and their loading effects on buildings. The project tried to address critical gaps in the current knowledge about spatial and temporal distributions of wind near the ground level and its interaction with natural and built environments. The TTU project focused on studying (a) the transient aerodynamic force of tornado wind; (b) the uncertainties in building performance subject to tornado wind; and 3) improvement to building design to tornado wind.

The tornado-induced pressure distributions on low-rise frame were strongly influenced by storm’s location, tending to be more uniformly distributed when inside of the core. As the result, the building frame responses were much larger than those estimated using the wind load specified in the current building code. The equivalent static wind load (ESWL) based on gust response factor approach gave reasonably estimation of tornado-induced peak responses. The translation of tornado led to a delay of occurrence of maximum time-varying mean and standard deviation (STD), and reduction in maximum STD (thus peak response). It also made the energy distribution of pressure fluctuations being shifted to higher frequencies with a broader power spectrum.

For examining the performance of building envelope against tornado-induced debris impact, full-scale finite element (FE) models were developed for metal roof decking assemblies in LS-DYNA. A mesh sensitivity analysis served as basis for selecting optimal mesh sizes. The FE models were then calibrated and validated with testing results with respect to metal deck residual deformation and failure mode. A reasonable level of agreement between simulation and experiment was achieved, allowing comprehensive parametric study involving different debris impact locations and missile velocities as well as varying material properties.

Previous studies showed that the wind-induced response of the base isolation system could be smaller than that of the fixed-base system when the hysteretic damping generated by yielding of base isolation system was noticeable. The yielding also caused the responses of base isolation system and structure to have non-Gaussian probability distributions. The research results indicated that the Gaussian linearization approach was able to give quite accurate estimations of building top displacement and base shear force in a wide range of wind speed or ductility factor of base displacement. As compared to fixed-base building, the displacement and acceleration of base-isolated building at higher wind speeds could be significantly reduced thanks to the effect of hysteretic damping. While the hysteretic damping resulted in reduction of the fluctuating response around the time-varying mean component, the increase in the mean response led to total inelastic alongwind response noticeably higher than that of the corresponding linear system in general, especially in the case of lower second stiffness. In general, crosswind response was larger than alongwind response of tall buildings. Therefore, building design could consider permitting crosswind response in inelastic range while the alongwind response remains almost elastic.

This project resulted in a better understanding of underlying physics of the dynamic flow field of tornado wind and its interaction with buildings that will lead to a more accurate prediction of wind damage to buildings and improved building design in the future.  This project has raised public awareness of wind hazards and benefited graduate, undergraduate and K-12 students, research scholars as well as public in general through improved curricula, laboratory demonstrations, student projects, publications and news media outlets.

Last Modified: 09/19/2019
Modified by: Daan Liang