Analysis of residential building performance in tornadoes as a function of building and hazard characteristics

Tornadoes damage and destroy many thousands of buildings per year in the United States leading to significant economic losses and long-lasting disruption to communities. In an effort to better understand the relationship between tornado wind speeds and resulting damage, multiple data sources from the May 22, 2011 tornado in Joplin, MO were mined to create a detailed Joplin Tornado Residential Damage Database. Methodologies were devised to partially automate data collection from several existing ground-based and aerial imagery sources and to quantify pre-tornado building characteristics and post-tornado observed damage to single family residential structures in a standard, repeatable manner. The database also includes information from a Texas Tech University post-tornado damage investigation and other data such as building information from the tax assessor and degree of damage estimates. Tornado speeds were estimated using results of a wind field model based on tree fall patterns developed at National Institute of Standards and Technology and enhanced at the University of Illinois Urbana-Champaign. This model provided an estimate of wind speeds throughout most of Joplin that was independent of building damage. Data was also collected on the performance of residential storm shelters hit by the May 20, 2013 tornado in Moore, Oklahoma in collaboration with the Federal Emergency Management Agency. That data was used to characterize basic tornado shelter design and installation information, determine how the structures were used during the event, and if there were any failures. The economics of having a storm shelter in high-risk areas was also considered. Although it is known from the literature that some building characteristics are likely significant determining factors for the level of residential damage caused by a tornado, quantification of how much specific parameters contribute to damage is still widely unknown. The empirical damage data combined with the modelled tornado speeds was used to create fragility curves for specific building characteristics to help fill in this gap in knowledge. The fragility curves provide some quantification of probabilities of damage as a function of tornado wind speed for features including roof geometry, building plan shape, age, and attached garage door size. Analysis of building damage data from past tornadoes proved very useful in understanding the influence of specific building and tornado characteristics on building performance. Both the methods used to collect and quantify building and tornado damage and the Joplin Tornado Residential Damage Database itself are tools that will enable tornado researchers to conduct more such analyses for Joplin and other tornadoes. The results from this dissertation have the potential to lead to significant improvements in the damage-based Enhanced Fujita (EF) Scale and help mitigate future damage through building designs that address vulnerabilities identified through such research. Additionally, fragility curves could be used to predict community economic performance after a tornado and help further figure out ways to increase societal function after an event.

Name thesis:degree_name

Doctor of Philosophy

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Wind Science and Engineering

Grantor

Texas Tech University

Year dc:date.issued

2021

Author dc:creator

• Scott, Pataya

Chair dc:contributor.committeechair

• Ewing, Bradley

Committee members dc:contributor.committeemember

• Liang, Daan
• Levitan, Marc

• Tornado Damage
• Joplin
• Wind Engineering
• Fragility Curves
• Storm Shelter

Language dc:language.iso

eng

Handle dc:identifier.uri

https://hdl.handle.net/2346/87580

OAI identifier oai:identifier

oai:ttu-ir.tdl.org:2346/87580