Failure of wood-framed low-rise buildings under tornado wind loads Nikhil Kumar, Vinay Dayal ⇑ , Partha P. Sarkar Department of Aerospace Engineering, Iowa State University, Ames, IA, USA article info Article history: Received 14 April 2009 Revised 5 February 2012 Accepted 6 February 2012 Available online 22 March 2012 Keywords: Wood-frame Tornado Stress analysis Failure analysis abstract Buildings in the ‘‘tornado alley’’ of the United States, are built to withstand 3-s wind speeds of 90 mph (40.2 m/s), whereas 90% of the tornados are of F2 or lesser intensity that generate anywhere from 40 to 157 mph (17.9–70.2 m/s). At the same time, these codes are based mostly on studying the effects of straight line winds and not on tornado type winds, especially on low-rise, wood framed buildings which make up majority of structures in the United States. Previous research at Iowa State University (ISU) includes extensive testing on a scaled down low-rise gable roof building model (1:100) to understand tor- nado induced loads as the tornado sweeps past the building. In the present work, Finite Element (FE) models were developed using ANSYS for full-scale numerical simulation of the gable roof buildings with three different roof angles (13.4°, 25.5° and 35.1°). The nail is modeled as a non-linear element but the wood is assumed to be linear. The tornado-induced wind loads recorded in the laboratory were scaled up and applied to the models to determine the detailed stress distribution in the structure. This numer- ical study was performed using the same parameter as in the laboratory experiments such as those listed earlier. The deterministic FE model incorporated the damage criteria to assess the damage potential due to tornadic forces. The stress distribution, pattern of failure, the order of failure and the type of failure have been studied as the tornado sweeps past the building at different angles to the building centreline. Ó 2012 Elsevier Ltd. All rights reserved. 1. Introduction Tornados are violently rotating columns of air, extending from a thunderstorm to the ground. Though they occur in many parts of the world, they are found to occur most frequently in the United States. There are around a thousand tornados reported annually in the US, causing around 60 fatalities, thousands of injuries, Grazulis [1], and resulting in damage of over a billion dollars. Though tornados have occurred in all fifty states, they are concen- trated in what is known as the ‘‘tornado alley’’, located in the cen- tral region of the country. According to the current design codes, low-rise buildings are built to withstand only up to 3-s gust of 90 mph (40.2 m/s) of straight-line winds, while 90% of the tornados are of F2 or lesser intensity that generate anywhere from 40 to 157 mph (17.9–70.2 m/s) fastest 1 = 4 -mile wind speed. At the same time, these codes are based on studies involving the effects of straight line winds and not the tornado type winds. Also, the prop- erty damages that occur due to tornados are significant due to wind-borne debris similar to the direct effect of high speed wind on them. It is therefore necessary to assess the wind damage potential of buildings as a function of distribution of local wind speed and map the generation of wind-borne debris from the buildings. Extensive wind tunnel tests have been performed on these types of structures under tornado type winds to obtain the forces acting on the structure. These tests were performed on low-rise building models with a variety of commonly used roof an- gles and shapes. It is needless to say that even though the buildings are designed for 90 mph wind speeds, the probability of it being subjected to tornado loads is small. Nevertheless, one fatality is too many and hence the building codes are developed to minimize such losses. The main objective of this research work was to apply these tor- nado-induced wind loads, obtained in the laboratory using a scaled model of a low-rise building, to a numerical Finite Element model (FEM) of the building to assess its damage potential. Loads are applied in the quasi-static manner, i.e. the experimental loads are applied to the building at discrete intervals and analysis per- formed to calculate the internal stresses and failure criteria is applied to determine the integrity of the elements. The model is assumed to be deterministic and the probabilistic nature of load- ing, material properties, nail pull-out have not been included. 1.1. Previous work A lot of work has been done to deal with 3-D performances of timber framed buildings. One of the first analytical models was developed by Tuomi and McCutcheon [2] which assumes linear elastic behavior of nails. The nail deformation here is defined by 0141-0296/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.engstruct.2012.02.011 ⇑ Corresponding author. Tel.: +1 515 294 0720; fax: +1 515 294 4848. E-mail addresses: vdayal@iastate.edu (V. Dayal), ppsarkar@iastate.edu (P.P. Sarkar). Engineering Structures 39 (2012) 79–88 Contents lists available at SciVerse ScienceDirect Engineering Structures journal homepage: www.elsevier.com/locate/engstruct