Making the Case for Improved Structural Design ... - Semantic Scholar

Downloaded from ascelibrary.org by Oregon State University on 10/22/12. Copyright ASCE. For personal use only; all rights reserved.

Making the Case for Improved Structural Design: Tornado Outbreaks of 2011 DAVID O. PREVATT, PH.D., M.ASCE; J OHN W.

VAN DE

L INDT, P H .D., M.ASCE;

E DWARD W. B ACK , P H .D., M.ASCE; A NDREW J. G RAETTINGER , P H .D., M.ASCE; S HILING P EI , P H .D., M.ASCE; W ILLIAM C OULBOURNE , M.E., M.ASCE; R AKESH G UPTA , P H .D., M.ASCE; D ARRYL J AMES , P H .D.;

AND

DUZGUN AGDAS, PH.D.

ABSTRACT: A total of 1,625 tornadoes occurred in the United States in 2011, resulting in economic losses that exceeded $25 billion. Two tornado outbreaks stand out because they caused more than half of those losses. The tornadoes that cut through Tuscaloosa, Alabama, on April 27 and Joplin, Missouri, on May 22 were responsible for a combined 223 fatalities and more than 13,000 damaged buildings in the two cities. Although the economic losses associated with tornado damage are well documented, the writers argue that the overall impact should encompass longer term, broader considerations such as the social disruption and psychological effects that impact communities. This paper examines observations by tornado damage assessment teams led by the first author in these two medium-sized cities and suggests that the evolution of building codes and past approaches to construction have led to conditions that made this extent of damage possible. The authors outline a multidisciplinary path forward that incorporates engineering research and social and economic studies into a new design paradigm leading to building code changes and social practices that will improve resistance and mitigate future losses at a community level from tornadoes. OCTOBER 2012

254

Leadership and Management in Engineering

Leadership Manage. Eng. 2012.12:254-270.


T

he year 2011 was one of the deadliest for tornadoes on record in the United States, with 1,625 tornado occurrences confirmed by the National Weather Service (NWS); Storm Prediction Center, National Oceanic and Atmospheric Administration (NOAA) (Storm Prediction Center 2011). One report estimated the total economic loss at $25 billion (Impact Forecasting 2011), and 553 fatalities were reported. Two tornado outbreaks alone, on April 22–28 and May 21–27, were responsible for 527 fatalities, $13.2 billion in damages, and impacts on 1.2 million homes (Impact Forecasting 2011). This paper discusses the findings of recent damage surveys after the two most devastating tornadoes of the 2011 season: the one that struck Tuscaloosa, Alabama, on April 27, and the one that struck Joplin, Missouri, on May 22 (Prevatt et al. 2011a, c). A team comprising structural engineers, wood scientists, and researchers investigated the damage left in the aftermath of the events to gather perishable data on damaged structures, paying particular attention to the performance of wood-frame residential homes. Despite substantial progress in our ability to model both the buildings and extreme loads acting on them, for over half a century residential and low-rise commercial buildings have been designed and built in essentially the same way. This class of structures, representing between 85% and 90% of the current U.S. building stock, are typically site-built, using

wood sheathing (e.g., plywood and/or oriented strand board) nailed onto light-frame wood and, more recently, cold-formed steel framing and covered with asphalt shingles on the roof and some form of cladding on the walls. Within the same period, tornadoes and hurricanes have periodically caused substantial losses to residential buildings, and damage surveys repeatedly associate the damage to inadequate connection details, an incomplete understanding of structural load paths, and inadequate construction practices.

BACKGROUND ON THE TORNADOES The NWS rated the Tuscaloosa and Joplin tornadoes EF4 and EF5, respectively, using the Enhanced Fujita (EF) scale (McDonald and Mehta 2006). The Tuscaloosa tornado cut across the city of Tuscaloosa (population 93,000), traveling in a northeasterly direction, and produced a damage swath that was 0.8 km (1=2 mile) wide by 10 km (5.9 mile) long. The Joplin tornado followed an easterly path, bisecting the city of Joplin (population 50,000) and creating a 1.1 km (3=4 mile) wide damage swath for approximately 10 km (6 miles). Figs. 1 and 2 show the Tuscaloosa and Joplin tornado paths. These two events caused 223 fatalities and $5 billion in economic losses [Alabama Center for Real Estate (ACRE) 2011; NOAA 2012]. In less than an hour, the tornadoes destroyed or caused major damage to more than 13,000 residential structures combined, or 12.6% and 30.2% of the housing stock in Tuscaloosa and Joplin,

Figure 1. Location map for Tuscaloosa tornado Leadership and Management in Engineering

255

OCTOBER 2012



Figure 2. Location map for Joplin tornado

respectively (Table 1). The initial estimate of the insured loss was $3 billion in Tuscaloosa and $2 billion in Joplin, while the overall economic losses are likely much higher, as nearly 40% of the damaged homes were either uninsured or underinsured (Associated Press 2011). In addition to the building catastrophe, the populations in Tuscaloosa and Joplin suffered other serious disruptions; schools and businesses were damaged, and many families became homeless and/or lost personal possessions.

LIFE EXPECTANCY AND DURABILITY OF RESIDENTIAL CONSTRUCTION According to the National Association of Home Builders (NAHB), the average life expectancy of a typical wood-frame house is more than 80 years—a lifetime (Economics Group, NAHB 2007). However, Table 1. Summary Statistics of the Tuscaloosa and Joplin Tornadoes (2011) Variable Economic loss Number of homes destroyed Percentage of homes destroyed Fatalities

Tuscaloosa tornadoa

Joplin tornadob

$3 billion 1,240

$2 billion 6,954

12.6%

30.2%

64

162

a

Data from Alabama Center for Real Estate 2011. Data from National Oceanic and Atmospheric Administration 2012.

b

OCTOBER 2012

such long life spans may be achieved only if the house is regularly maintained. Many systems in houses (i.e., plumbing and electrical accessories, windows, bathrooms, kitchens) become functionally obsolete long before that time. In reality, the building components and systems that protect the structural framing (e.g., roofing materials, siding, paint, plywood, joint sealant) do not last more than 15–30 years, even when properly maintained. Wood framing deteriorates rapidly in the presence of water or high humidity and when subjected to extreme environmental loads and insect infestation. Once a house is occupied, it is difficult and sometimes expensive to ascertain the structural condition of the wood wall framing that usually is hidden between the interior drywall and exterior siding. In many cases (as discussed below), homes in Tuscaloosa and Joplin, especially in low-income housing areas, were not well maintained and had structural flaws that would have reduced their resistance to even moderate tornadoes or loads. According to the 2005 American Housing Survey by the U.S. Census Bureau (2005), the U.S. housing stock consisted of more than 124 million homes with a median age of 32 years. The housing stock in Tuscaloosa and Joplin mirrored this national average; over 65% of the houses were at least 30 years old (Fig. 3). Most structures within the two cities were single-story, single-family homes and light commercial buildings built between the 1950s and 1970s (U.S. Census Bureau 2010a, b). The structural systems of these older buildings showed signs of structural distress, including deteriorated and rotting wood

256




Figure 3. Age distribution of housing stock in Tuscaloosa, Alabama, and Joplin, Missouri (data from U.S. Census Bureau 2010a, b)

framing, corroded nails and fasteners, and waterdamaged sheathing and framing members.

HISTORY OF TORNADO DAMAGE INVESTIGATIONS The dynamics of the tornado vortex and theoretical models are available in the literature (Bengtsson and Lighthill 1982; Davies-Jones 1986; Lewellen et al. 1979), and Lewellen (1993) provided a thorough review of these and other models. Knowledge of tornado interaction with structural systems has been improving in recent years through the efforts of scientists collecting data during the passage of tornadoes (Kosiba and Wurman 2010; Kosiba et al. 2008; Lee and Wurman 2005; Kosiba et al. 2005) and from related surface observations of winds and forensic investigations of the damage following these events (Karstens et al. 2010; Coulbourne 2008; Building Performance Assessment Team 1999; Minor et al. 1978; Kikitsu and Sarkar 2010). Although tornado occurrences are most associated with weather conditions in the midwestern United States, they are a global natural hazard occurring on most continents (Chmielewski et al. 2011; Bienkiewicz 2008; Budek et al. 2006). Posttornado Damage Survey Investigations Engineers have routinely performed damage investigations after tornadoes, providing valuable insight to the engineering community. An early paper in Transactions of the American Society of Civil Engineers published 116 years ago detailed tornado damage Leadership and Management in Engineering

and load estimates “with special reference to wind bracing for high buildings” in St. Louis, Missouri (Baier 1897). In the modern era, Minor et al. (1978) and Fujita and Pearson (1976) were further instrumental in establishing methodologies of tornado damage assessments. Fujita also proposed a tornado intensity rating scale that became known as the Fujita or F-scale, which provides a means to estimate wind speeds based on damage observations in the tornado path. In 2003, NOAA published an F-scale Assessment Guide detailing how to use the F-scale and some recommendations for forming teams, selecting equipment, and postprocessing the data (Doswell 2003). A classification of a tornado’s damage potential is currently defined using the Enhanced Fujita scale created by Texas Tech University researchers (McDonald and Mehta 2006). The current EF rating scale, which has six levels, estimates wind velocity using observations of damage and construction quality. Still, direct correlations of wind speed and structural damage in the EF ratings are only estimates, as there are very few reliable wind speed measurements to validate the ratings (Karstens et al. 2010; Wurman et al. 2007). The EF scale is organized around observed damage according to a set of 28 different indicators. Each indicator has unique degrees of damage that are correlated to specific ranges of wind speeds. Each range consists of upper and lower bound values to account for circumstances that may cause the actual wind speed to deviate from the expected value. Such circumstances primarily relate to construction quality. Normal conditions imply appropriate construction materials and traditional construction. Low construction quality or a less than robust (or even total lack of a) building code would push the wind speed estimate toward the lower bound. The damage survey teams used the EF scale in assessing damage and estimating wind speeds for the Tuscaloosa and Joplin events. In the days immediately following the Tuscaloosa and Joplin tornados, the NWS reported damage assessments within 2 to 3 days and rated both tornadoes as EF4. Subsequently, the Joplin tornado rating was adjusted upward to EF5. Our damage survey team assigned similar ratings, EF4 for the Tuscaloosa tornado and EF4 for Joplin, indicating that both were extremely violent tornadoes. This variation in the NWS rating is indicative of the subjectivity of the rating system, as described in a previous report by Prevatt et al. (2011b).

257

OCTOBER 2012



Experimental Studies on Tornadic Fluid-Structure Interactions Several tornado simulators have been developed over the past decade, including the VorTECH at Texas Tech University and the Tornado Simulator at Iowa State University. These large research devices are capable of producing tornado-like single-celled or two-celled vortices of sufficient size to conduct wind load studies on model-scale buildings. Fig. 4 presents a surface pressure profile from a simulation in the VorTECH that matches the surface pressure profile of the June 2003 Manchester, South Dakota, tornado as reported by Lee et al. (2004). Tornado simulators have also been able to match tangential velocity data from tornadoes. Tornado wind loads on a structure have been found to be highly variable depending on the structure’s location with respect to the tornado vortex, translational velocity of the tornado, size of the tornado, orientation of the building, and shape of the building (Wang 2002; Fouts et al. 2003; Mishra et al. 2005; Sarkar et al. 2006; Mishra et al. 2008a, b; Sengupta et al. 2008; Haan et al. 2010; Hu et al. 2011). Preliminary results indicate that tornadoinduced wind loading differs markedly from loads associated with straight-line atmospheric boundary layer (ABL) flows, and in some cases it can be several times larger for the same wind velocity. This fundamental understanding of how tornado winds interact with and damage buildings is necessary for engineers to develop design load and detailing recommendations that will increase the safety of both engineered and nonengineered residential structures, the latter arguably more important because they make up most

Figure 4. Comparison of surface pressure profiles from a VorTECH simulation and the Manchester, South Dakota, tornado of June 2003 (data from Lee et al. 2004); the geometric scale in VorTECH was approximately 1∶125. Vertical axis shows minimum recorded pressure (TTU-TVSII = Texas Tech University VorTECH tornado simulator) OCTOBER 2012

of the building inventory, so improved safety would have a greater impact. Research continues in determining, for instance, locations on a structure where the greatest deviation from traditional ABL loading occurs in order to develop specific strength (and performance) enhancements and the effect of topography on the type of tornado-like vortex formed. Physical simulation of tornadic fluid-structure interaction is the only reliable, repeatable, safe method to obtain data on building performance in tornadoes. Physical simulation will continue to improve as inflow and near surface data from atmospheric scientists are incorporated.

DATA COLLECTION APPROACH The damage assessment teams were formed and mobilized within days of the tornadoes, and they consisted of university engineering faculty and students, structural engineers, wood scientists, and local professional engineers. Investigations were performed in under a week, beginning 5 to 7 days after each event. Prior to each deployment, the extents of the damage swaths were identified and mapped using available satellite and aerial photography from many sources. Street maps of the towns were acquired and overlaid on the damage swaths, providing investigators with a view of the magnitude of the task. The teams were independent but interfaced routinely with disaster mitigation leadership at the Federal Emergency Management Agency (FEMA) in Washington, DC, and local FEMA representatives. Building officials in both cities provided logistical support. The survey teams used global positioning system (GPS) enabled cameras and video cameras, notebooks, data collection forms, and standard engineering equipment (measuring tape, annotated maps, GPS units) to conduct rapid forensic evaluations of the structural damage to buildings. The survey teams were divided into two- or three-person units that were assigned to collect structural data on a specific street or location, called transects. To complete the data collection, the damage area was divided into several linear transects running across the swath and located approximately 0.8 km (1=2 mile) apart (Fig. 5). On each transect, teams collected information for a number of residential structures on damage to roof cover and siding and roof structure, major component failure of exterior and interior walls, and damage to trees and outbuildings. Although we looked at buildings on the edges of the damage path to establish the limits of the damage, these undamaged structures

258




Figure 5. Transects through the Tuscaloosa tornado path

indicated only the extent of wind-induced forces strong enough to cause damage; undamaged buildings could not be used to provide information about construction quality. In Tuscaloosa, the survey team took and geotagged 6,016 damage photographs, 1,569 of which were given EF damage ratings. Multiple photographs were taken of many houses, so less than 50% of the photographs were of unique buildings. In Joplin, approximately 1,350 geotagged, EF-rated photographs were taken, with far fewer repeats, so 80–90% of the Joplin photographs were of unique buildings. The variation in damage perpendicular to the tornado path captured at numerous locations was of particular interest to the teams. Data were collected via both active and passive modes. Active data collection consisted of taking photos, writing text descriptions, making hand sketches, and generating ground-based light detection and ranging (LIDAR) point clouds. Passive data collection consisted of taking ground-based and vehicle-based photos linked to GPS tracks. A methodology was developed for tracking the time and location of all measurements that enabled rapid assembly of a robust spatial-temporal data set that can be displayed in a geographic information system (GIS). Photos and Leadership and Management in Engineering

GPS tracks were downloaded from field equipment at the end of each day and processed to create a progress map of the survey. The available use of GIS systems provided a significant improvement in data collection and dissemination methods from previous tornado damage surveys, to the extent that preliminary information was available on the Web within 24 hours of the surveys in map form. The damage to each building was assessed and rated in accordance with the Enhanced Fujita scale, described earlier. These results were displayed using color-coded icons on the GIS maps (Fig. 6). Case Studies and Data Dissemination During each survey, a few structures were chosen for more detailed assessment. Team members used case study sheets to evaluate these structures and gather details pertaining to vertical and lateral load paths in the structure; failure mechanisms; materials and condition; presence, size, and spacing of anchor bolts and roof-to-wall connections; and sizes of key structural components. The individual EF rating for each structure was combined with the spatial location of the rating from the GPS data to generate EF contours for the damage swaths (Figs. 7 and 8).

259

OCTOBER 2012



Figure 6. Locations of rated photos and case studies with Enhanced Fujita ratings from geographic information system Web sites (reprinted with permission from the Center for Advanced Public Safety at the University of Alabama)

Figure 7. Contour maps of Enhanced Fujita wind speed estimates for Tuscaloosa, Alabama

The approach significantly improved the efficiency and accessibility of data from these damage investigations. A daily blog was also used to post live updates OCTOBER 2012

of the findings, which can be accessed at http://www .davidoprevatt.com. Currently the information is available for further analysis and can be obtained from online

260




Figure 8. Contour maps of Enhanced Fujita wind speed estimates for Joplin, Missouri

servers (http://esridev.caps.ua.edu/tuscaloosa_tornado/ and http://esridev.caps.ua.edu/JoplinTornado/). These GIS platforms can selectively display information from the survey’s specific layers, including damage boundary, satellite views, rated photo locations, damage rating, detailed case studies, and EF contours. The case study reports include before and after photographs of the structures, which are useful to gain insight into the power of the tornado and structural vulnerability of the structure in question. (For further details, see Prevatt et al. 2011a, d). Summary of Damage Patterns and Typical Failure Mechanisms The damage survey results from the 2011 tornadoes exposed many key elements in light-framed wood construction that failed. Numerous tornado damage reports (Building Performance Assessment Team 1999; Martin and Readling 2008; Kikitsu and Sarkar 2010) have identified similar findings, such as failure of toe-nailed truss-to-wall connections, poor attachment to foundations, horizontal “hinge” failure at the gable end truss-to-wall top plate connection, and inadequate structural wall sheathing panels. It should be mentioned that these issues have also been identified in hurricane damage surveys, and subsequent improvements to the building codes in hurricane-prone regions have significantly improved Leadership and Management in Engineering

building performance in those events (Gurley et al. 2006). Structural damage also occurred when homes were lifted off their foundations. In too many structures, the connections of the structure to its foundation and the strength of the foundation itself are inadequate, again underscoring the need to ensure structural adequacy throughout the complete vertical (and lateral) load paths. Failure occurs at the weakest link of the chain, and in many cases this is the foundation and its connection to the ground. It is concerning that design load standards such as ASCE 7 (2010) do not include guidelines for tornadoresistant structures, although design provisions for hurricane-resistant design have been around for decades. This limitation has occurred because despite the costly consequences of a tornado strike, particularly to densely populated areas, the most damaging tornadoes have relatively low probability for a sitespecific occurrence. Also, the original Fujita scale had incredibly high wind speed estimates, as high as 318 mi/h (Fujita 1971) and thus was an impractical measure resulting in economically unfeasible residential building designs. Despite the low probability of the event, the potential for catastrophic loss when a tornado touches down in a large urban community of traditional residential construction makes the current status quo of uncontained structural damage untenable. Statistically, over 90% of all tornadoes

261

OCTOBER 2012



are rated EF3 or lower, and they account for nearly 75% of the economic losses (Walker 2007). Estimates of tornado wind speeds suggest that the wind velocity of EF0 to EF2 tornadoes is within a similar range as used in current hurricane-prone regions, although this estimate does not account for the strong suction pressures below the tornado’s vortex. This increase in forces due to the tornadic winds needs further work to enable a better understanding over the next decade. EF contour maps developed for these damage surveys indicate most of the region impacted by EF4 or EF5 tornadoes did not experience EF4 or EF5 wind speeds (Fig. 9). The data from the Tuscaloosa and Joplin tornadoes are in good agreement with damage contours previously developed following the devastating F5 tornado that struck Moore City, Oklahoma, on May 3, 1999 (Building Performance Assessment Team 1999). Given that most of the region affected by a tornado of any size will likely experience intensities of EF2 or less, it is worthwhile to consider improvements to residential structures against these lower wind speeds (i.e., instead of designing for the higher EF levels) in order to improve communities’ overall resilience.

have not changed significantly in these areas since they were first implemented. One cannot expect different results from the same actions, and mitigation of unsustainable wind-related disasters through improved building performance cannot be expected unless a new design paradigm is implemented. Van de Lindt et al. (2011) developed a dual objective–based design philosophy based on observations of failures in wind-damaged buildings within the destructive path of the tornado. According to this approach, there are two considerations or design objectives for tornado design: damage (D) and life safety (L). The dual design approach can be achieved using three philosophies (see Table 2):

DUAL OBJECTIVE–BASED TORNADO DESIGN METHODOLOGY The weak roof-to-wall connections, inadequate wallto-foundation attachment, and gable end failures are all issues that have been noted for decades, yet construction methods in most parts of the country

1. Damage can be controlled at lower levels of the Enhanced Fujita scale wind speeds (i.e., EF0 and EF1) through the use of engineered connectors, design ensuring continuous vertical uplift load paths, horizontal load distribution and load paths, and better shingles and reinforced garage doors. This is handled typically at the component (C) design level (i.e., connectors, single load paths). 2. For wind speeds corresponding to EF2 and EF3, both component- and system-level loading must be considered to enable better performance. System (S) performance is related to load sharing among wall lines and distribution of the lateral load path as a whole throughout the building as a structure is racked by wind and windborne debris.

Figure 9. Comparison of Enhanced Fujita intensity distribution by area for the Tuscaloosa, Alabama; Joplin, Missouri; and 1999 Oklahoma City tornadoes OCTOBER 2012

262



Table 2. Design Objectives and Philosophy Considered as a Function of Wind Speed Methodology proposed


Tornado design approach Design objective Damage (D) Life safety (L) Philosophy considered Component (C) System (S) Alternative (A)

EF0 (65–85)

EF1 (86–110)

EF2 (111–135)

EF3 (136–165)

EF4 (166–200)

EF5 (>200)

D

D

D/L

D/L

L

L

C

C

C/S

S

S/A

A

Note: EF = Enhanced Fujita scale wind speed range (3-s gust wind speed at 10 m in open terrain exposure, in miles per hour).

3. In tornadoes with wind speeds corresponding to EF4 and EF5, the major issue becomes system effects and other alternatives (A) to provide life safety to the building occupants assuming total devastation of the main structure. These alternatives are safe rooms, underground shelters, and basements.

the norm in nearly every U.S. city or community. The consequence of this reality is that most communities do not consider housing to be an element of the public sector’s comprehensive infrastructure; that is, housing is not generally financed, insured, constructed, or maintained by the public.

Table 2 presents the design objective and philosophy aligned with each EF level. The dual objectives of damage and life safety must be used simultaneously in building design, and therefore so should the three philosophies that drive the design toward these objectives. This alignment will ensure minimization of financial losses when possible and protection of life safety for building occupants in the worst case.

Housing Risks Undeniably, housing is susceptible to risk events, and individual homeowners and other residential property investors are wise to use mitigation strategies such as property insurance and similar measures to protect their investment. As a general rule, risk events normally impact few structures. A fire may affect several homes or even one or more city blocks. Similarly, localized thunderstorm damage or flooding may affect only a few dozen homes in a defined geographic area. However, major or extreme storm events such as hurricanes, large-scale floods, and tornadoes have unprecedented impacts on the community’s overall housing supply. The sudden loss of housing leaves cities with unexpected consequences that can only be reactively addressed.

VULNERABILITY OF RESIDENTIAL COMMUNITIES TO TORNADO DAMAGE Because the Tuscaloosa and Joplin tornadoes inflicted most of their damages on residential housing, it is worthwhile to consider the nature of the housing market in order to identify the parties who would be responsible for implementing these new design philosophies and mitigation of future damages. Housing is generally a private sector investment. Whether through individual home ownership or single or multifamily investment, residential property is largely financed and maintained by citizens who recognize their responsibility as personal investors. The provision of adequate housing within a community is largely a function of free enterprise, with the implication that local governments perceive housing supply or availability as a response to local demand within various demographic constructs. A market-driven, supply–demand approach to housing provision is Leadership and Management in Engineering

Response to Housing Damages Following a Disaster It is clear from the aftermath of the storms of the last few years that housing supply may be the forgotten infrastructure within a community. Its sudden loss results in transference of risk to the local public sector. The inability of the private sector to respond to a critical housing need at a macro level, when thousands of homes are required in a very short time frame, reveals a public sector or social vulnerability that until now has been unappreciated or even unrecognized by many engineers and social scientists. Communities need to

263

OCTOBER 2012



consider the structural resiliency of the housing supply at the macro or community level if they are properly mitigating risks likely to occur with extreme storm events. One reason that the private sector cannot adequately respond to housing needs in the recovery phase of major storm events is that often the ability to recover is to some degree dependent on the financial strength of the individual community. Poor or less affluent housing areas are much more likely to have difficulty recovering from a disaster, in part due to the higher level of uninsured or underinsured property conditions. For instance, within Alabama, 36 of the 42 counties that FEMA declared eligible for individual disaster assistance following the Tuscaloosa tornado had poverty rates higher than the national average (Alabama Possible 2011). Lower-income families generally live in housing that is less structurally resilient, leaving the occupants vulnerable to higher levels of personal injury and other forms of physical and emotional stress. Impact of a Tornado Although the focus of the preceding discussion has been on two tornadoes in particular, now consider the entire outbreak of tornadoes across the state of Alabama. The Alabama tornados of April 27, 2011, had a significant impact on overall housing supply within the state, and particularly the county of Tuscaloosa. The events of that day evidence clearly the need to consider housing supply at the macro or public sector level when assessing social vulnerability and structural resiliency. The impacts were greatest in Tuscaloosa along a short stretch of densely populated urban area within the 80-mile path of the tornado on the ground. One approach that may be considered when designing buildings is to adjust the Building Risk Classification in ASCE 7-10 (ASCE 2010). This Building Risk Classification results in a design based on larger loads for buildings that have greater importance or that place a greater number of lives at risk. However, the approach is calibrated on a buildingby-building basis. An analogy can be made that a community’s risk, determined by its longer expected life and greater population at risk, can be assigned by aggregating individual building risks for a group of buildings in a dense urban neighborhood, since the collective exposure of the neighborhood to tornado damage is higher than for buildings constructed in isolated locations. With the building classification procedure in the current ASCE 7-10 engineering design procedure, OCTOBER 2012

it is possible that designers and owners will more frequently come to an agreement that a building built in an area with frequent tornado activity should be built to a higher risk category. However, if the design uses the wind speed maps in ASCE 7, even the basic design wind speeds for Category III/IV buildings (i.e., a basic wind speed of 120 mi/h) are well below the wind speeds expected from tornadoes. Direct Impacts—Building Damage and Loss of Contents An assessment by the Alabama Center for Real Estate (ACRE 2011) found that approximately 14,000 housing units were directly impacted by the storms in Alabama on that day. Of that total, 18% were located in Tuscaloosa County (Department of Homeland Security 2011). This finding indicates that about one-third of the occupied housing at the time of the storm was likely to have the financial resources to rebuild. Of the total FEMA applicants for assistance within the state to date, over 50% indicated that their housing was uninsured. Within the city of Tuscaloosa, the damage to the housing supply was severe. The tornado traveled through numerous residential neighborhoods. All housing types were affected, including owneroccupied homes of varied economic value; rental units, including a high percentage accommodating university students; and multifamily units that principally housed low-income families. Before the storm, 40,872 housing units existed within the city of Tuscaloosa, according to U.S. Census Bureau data of 2010 (ACRE 2011). More than 5,100 housing units were impacted directly by the tornado of April 27, 2011, representing 12.6% of the city’s total housing supply. A very preliminary estimate, based on market analysis only, indicates a financial need approaching $300,000,000 for replacement of housing that was lost on that single afternoon. Of the housing units affected by the tornado, approximately 24% were completely destroyed; 17% of the destroyed units were apartments, and 15% provided public or low-income housing. More than 1,400 housing units, or approximately 28% of the total, were not destroyed but were rendered uninhabitable until significant repairs were completed (ACRE 2011). Such data are not particularly surprising. Catastrophic storms consistently have a significant impact on housing supply. The same was true when evaluating the impact of Katrina; Alabama experienced more than $512,000,000 worth of housing

264




damage during Katrina, most of which, approximately $378,000,000, was experienced by homeowners (Gerhart 2011). Some argue that a potential gap of over $300,000,000 in recovery funding for homeowners still remained 6 years after Katrina made landfall. Indirect Tornado Losses—Social Dislocations and Psychological Effects Clearly, the housing supply is the most vulnerable structurally to damages from natural disasters, including tornadoes. Further, any meaningful concept of structural resiliency must include the goal of protecting the most vulnerable. Therefore, efforts to improve resiliency should explicitly address residential communities that are likely to be more vulnerable due to such issues as average age of structure or overall structural integrity, average age of occupants, community financial strength, level of poverty, and likelihood of insurance and other risk mitigation factors. Sociologists Fothergill and Peek (2004) concluded that while poor individuals are more likely to perceive hazards as risky, they are less likely to prepare for hazards, buy insurance, or respond to warnings. They are also more likely to die, suffer injuries, and experience psychological trauma; have proportionately higher material losses; and face more obstacles during the phases of response, recovery, and reconstruction. These factors should be particularly important to engineers in light of the fact that over 15% of Americans live below the federal poverty threshold (U.S. Census Bureau 2010c). Economic hardship can be an impediment to mitigation, particularly when straightforward engineered tornado-resilient designs do not exist. There is a need for engineers to develop designs that can be integrated with local, state, and federal resources to meet the needs of the most vulnerable within a community. Without focused outreach, planning, and assessment at the macro level, entire communities will fail to recognize their vulnerability, and the risk will remain invisible to public sector administrations and organizations. A better understanding of this issue is emerging. In Alabama, a housing task force within the state government was established to address postdisaster housing issues by coordinating and facilitating the delivery of state, local, and federal resources to assist in the rehabilitation and reconstruction of destroyed and damaged housing, whenever feasible, and the development of other new accessible, permanent housing options. Similar efforts should be undertaken in other communities as well. Leadership and Management in Engineering

ESTABLISHING THE TRUE COST OF LARGE-SCALE TORNADO DAMAGE The true cost of large-scale disasters is of profound interest to researchers in many fields. A fundamental difficulty in assessing the impact of damage to civil infrastructure is the connectivity among systems as they pertain to the functionality of the society as a whole. While it is feasible to collect data on immediate hazard mitigation and disaster recovery efforts, the true long-term, indirect costs are more difficult to conceptualize and analyze. Melcher (2003) classified the impact of natural disasters in three major categories— economic, humanitarian, and ecological—and further stratified these categories with respect to the scale of the analysis of the event under investigation, as well as the duration of the impact considered. In the case of the 2011 tornadoes, the $25 billion estimate represents only the economic cost related to removing debris and replacing the building infrastructure and lost possessions of the affected populations. The figure does not include so-called humanitarian costs due to fatalities and injuries during the event, long-term underemployment, and elevated medical and psychological problems the community will need to address for months or years to come. In the engineering context, the feasibility of a structural retrofit program or building replacement is established using benefit–cost analysis (BCA), a methodology that includes the lifetime costs and benefits of disaster-related impacts to support the decision-making process. BCA generally assumes a useful life of the structure and a discount rate in order to aggregate long-term cash flows, taking the time value of money into account. BCA analysis is meaningful only when all the aggregated cash inflows and outflows are discounted back to the future data and a simple ratio of inflows (benefits) and outflows (costs) is calculated. Kramer (1995), Venton and Venton (2004), and Dedeurwaerdere (1998) discussed shortcomings of the BCA approach when applied to large-scale disasters like a tornado impacting a densely populated city. Aspects of BCA pertinent to the Tuscaloosa and Joplin tornado disasters are system boundaries for analysis, reliability of data, assumptions, and interdependence of the infrastructure systems. System Boundaries for Analysis A true economic assessment should be comprehensive enough to incorporate all the costs associated with natural disasters, including the immediate economic losses and the social and humanitarian costs. If the

265

OCTOBER 2012



BCA boundaries are extended to include the indirect costs of disasters, such as long-term health and economic implications, it becomes necessary to assign a dollar figure to these noneconomic factors. In the past, indirect costs may not have been considered, perhaps limited by the difficulties in conducting such a comprehensive assessment. A clear example of the ripple effects of large-scale disasters is provided in a report on Hurricane Katrina’s impact on children’s mental health (Mailman School of Public Health 2010). One of the significant findings was that children displaced because of Katrina were 4.5 times more likely to have a serious emotional disturbance compared with the national average. Despite the importance of these findings, the question remains how to successfully incorporate these long-term and indirect effects into a quantitative analysis. Future damage assessment surveys should include data collection that captures both the economic and noneconomic impacts of large-scale disasters. Reliability of Data Consistency and accuracy of input data are crucial for reliable analyses. In the case of large-scale disasters, the extents of damage and the multiple scales of factors to be assessed make compiling reliable data a significant hurdle. For the focused damage assessment presented in this paper, the research team comprised structural engineers and researchers with backgrounds in the construction industry and engineering. Other damage survey teams consisted of meteorologists from the National Weather Service, and still others may have come from the health services and FEMA. Generally, these groups collect and disseminate their information separately, with little overlap or collaboration. There was some coordination for the Tuscaloosa and Joplin damage surveys among various academic research teams, scientists from the National Institute of Science and Technology, and FEMA first responders. Daily teleconference briefings and debriefing meetings at the end of the survey periods enabled researchers to exchange preliminary information. A centralized location that aggregates information from these multiple teams would be a valuable resource for future study, as the fragmented data collection and multiple formats require a substantial amount of secondary data processing to enable aggregation. Assumptions While the simplicity of the BCA approach makes it an attractive decision support tool, by its nature as a deterministic assessment, it may be overly dependent OCTOBER 2012

on the accuracy of original, sometimes spurious assumptions. In the context of disaster management, the BCA result may have long-term implications, even though it may be based on oversimplified analyses. To overcome this overdependence, probabilistic analysis methods such as Monte Carlo simulation have gained favor, but these are also very sensitive to the assumptions made, and possible inherent inaccuracies would compromise the reliability of the analysis. Thus, substantial effort is required to justify all assumptions prior to the analysis during the initial planning phase of a BCA study. Interdependence of the Infrastructure Systems Despite current knowledge and the sophistication of available analysis tools, little is known about the true interdependence of the infrastructure systems in developed countries due to the inherent redundancies that exist (Vermeiren et al. 1998). A redundancy exists when multiple subsystems can provide the same functionality if the main subsystem fails to do so. Thus, except for large-scale extreme events, impact assessments might not be very accurate as the loss of functionality of one subsystem can be compensated for by other components, and functionality reduction in substitute system components is generally overlooked. In the case of large-scale disasters, the level of damage and the complex interrelationships among subsystems make it difficult to accurately account for the overall impact of the natural disaster.

DISCUSSION Tornadoes will always be with us, and an annual increase in damage and loss of life is increasingly likely as the U.S. population increases. These increasing damages are easily apparent in the visible stock, such as buildings, but are not quite as apparent in less visible consequences, such as unemployment experienced by affected persons, disruptions in schooling, and long-term psychological effects. One option to mitigate these damages is to improve the design and construction of the most vulnerable infrastructure, residential housing, by enabling it to resist these hazardous loads. Building practices in tornado-prone states are similar in that the design wind speed in most of those states is 115 mi/h for ASCE 7-10, Building Risk Category II. Some of these states do not have a statewide building code, and states or local governments that have adopted a building code may not be able to adequately enforce it. The survey results show that

266




builders typically are building for gravity loads only (some include snow loads) but not for high winds. Our surveys found causes of catastrophic building failures similar to those reported by Minor et al. (1978) and Fujita and Pearson (1976) over 40 years ago. That a similar level of damage occurred in 2011 should not come as a surprise, since the residential structural systems are essentially unchanged since that time. Given the age (> 30 years) and the lack of proper maintenance of most of the houses, they are more vulnerable to any EF level of tornado damage. New engineered structural systems may need to be developed for residential construction that can resist moderate tornado loads. Better structural enhancement options for existing buildings are also needed. Regular structural inspection of houses has been proposed as a means to identify and repair inadequate structures within a community. When homes are sold, many of them are inspected, but typically their structural capacity to resist natural hazards is not a consideration. If structural inspection reports on houses were made public, homebuyers would have important facts (or a rating) of the resilience of the structure. Further, such reports may be an incentive that encourages sellers to structurally retrofit their homes, knowing the value the public places on such upgrades.

CONCLUSION The destruction of large communities of homes by tornadoes highlights the need for acceptance of more resilient residential construction practices as the basis for viable housing in the 21st century. This paper summarized the damage investigation approach used in two posttornado building-damage surveys and discussed the vulnerability of the existing residential infrastructure, arguing for more resilient residential construction in the United States. The study described in the paper confirmed that reduced tornado intensities occur at a moderate distance from the centerline of the tornado path. This finding qualitatively indicates that most tornado-resilient structures may survive even in the most violent events. A dual objective–based design methodology, introduced in an earlier paper by the authors (van de Lindt et al. 2011), explains the design objectives and philosophy for a feasible design approach for tornadoresilient communities that would simultaneously reduce economic losses and fatalities. Modern GIS data-collection tools have enabled rapid dissemination and archiving of damage survey results and have made it possible to conduct ongoing analysis of the data by Leadership and Management in Engineering

a wide network of researchers. Recent laboratory programs have developed tornado simulators that may soon yield empirically based design guides for estimating tornado-induced structural loads on buildings. Studies are still needed to determine what design wind speeds can be economically feasible, but the economics must account not only for direct building and infrastructure losses, but also for social and psychological effects and the time value of money, including future losses.

ACKNOWLEDGMENTS The authors acknowledge the financial support of the National Science Foundation for the Tuscaloosa damage surveys under Grant No. 1139722. Simpson Strong-Tie, the Applied Technology Council, and the University of Alabama provided support through personnel and access to their facilities. The International Associations for Wind Engineering provided travel support to the Tuscaloosa team. The authors express their gratitude to the American Society of Civil Engineers/Structural Engineering Institute for supporting the Joplin damage survey and for identifying their local members who became part of the damage survey team. The authors thank the Federal Emergency Management Agency for providing logistical support on the ground and through FEMA head office staff, who coordinated daily briefings and provided other valuable information during both surveys. Finally, the authors express their appreciation for the dedicated help and enthusiasm of the many university students from the University of Florida, University of Alabama, Clemson University, Iowa State University, Oregon State University, and South Dakota State University who participated in the study and have been conducting the postsurvey data analysis of the results.

REFERENCES Alabama Center for Real Estate. (2011). “Tuscaloosa residential real estate: Short-term impact and analysis of April 27th disaster.” Univ. of Alabama, Tuscaloosa, AL. Alabama Possible. (2011). “Revised—High poverty areas hit hard by tornadoes: 36 of 42 counties on disaster list have above-average poverty.” Alabama Poverty Project, Birmingham, AL. ASCE. (2010). “Minimum design loads for buildings and other structures.” ASCE/SEI 7-10, Reston, VA.

267

OCTOBER 2012



Associated Press. (2011). “Insurers pay out $1.13 billion for Joplin tornado damage; total will be higher.” 〈http://www.columbiamissourian.com/ stories/2011/11/28/insurers-pay-out-113-billionjoplin-tornado-damage/〉 (Apr. 9, 2012). Baier, J. (1897). “Wind pressures in the St. Louis tornado, with special reference to the necessity of wind bracing for high buildings.” Trans. Am. Soc. Civil Eng., 37(1), 221–289. Bengtsson L., and Lighthill J. (1982). Intense atmospheric vortices, Springer-Verlag, Berlin. Bienkiewicz, B. (2008). “Lessons learned from structural damage investigations—A case study of 2003 Missouri–Kansas tornadoes.” Proc., Structures Congress 2008: Crossing Borders, ASCE, Reston, VA, 1–10. Budek, A., Zain, M., Qiao, L., and Phelan, R. S. (2006). “Validation of finite-element analyses for storm shelters.” J. Archit. Eng., 12(2), 64–71. Building Performance Assessment Team. (1999). “Midwest tornadoes of May 3, 1999: Observations, recommendations and technical guidance.” Rep. FEMA 342, Federal Emergency Management Agency, Mitigation Directorate, Washington, DC. Chmielewski, T., Nowak, H., and Walkowiak, K. (2011). “Tornado near Opole—Estimates of wind velocity and assessment of buildings performance.” Proc., 13th Int. Conf. on Wind Engineering, International Association for Wind Engineering, Atsugi, Kanagawa, Japan, 〈http://folk.ntnu .no/karpa/ICWE%2013/ICWE13%20Proceedings/ start/papers/032_8page_tornado_near_opole_ estimates_of_wind_velocity_and_assessment_ of_buildings_performance.pdf〉 (Mar. 27, 2012). Coulbourne, W. L. (2008). “Wind speed analysis of Greensburg, KS tornado.” Proc., Structures Congress 2008: Crossing Borders, ASCE, Reston, VA. Davies-Jones, R. P. (1986). “Tornado dynamics.” Thunderstorms: A social, scientific, and technological documentary, E. Kessler, ed., National Oceanic and Atmospheric Administration, Washington, DC, 197–236. Dedeurwaerdere, A. (1998). “Cost–benefit analysis for natural disaster management: A case-study in the Philippines.” CRED Working Paper 142, Centre for Research on the Epidemiology of Disasters, Brussels, Belgium. Department of Homeland Security. (2011). Housing Recovery Support Function: Initial housing impact assessment Alabama tornadoes, U.S. Department of Homeland Security, Washington, DC, 〈http:// OCTOBER 2012

www.adeca.alabama.gov/C5/Housing/Document% 20Library/Alabama%20Housing%20Impact% 20Assessment%2020110811.pdf〉 (Mar. 27, 2012). Doswell, C. (2003). A guide to F-scale damage assessment, National Oceanic and Atmospheric Administration, Silver Spring, MD. Economics Group, National Association of Home Builders. (2007). Study of life expectancy of home components, National Association of Home Builders and Bank of America Home Equity, Washington, DC. 〈http://www.nahb.org/fileUpload_details.aspx? contentID=99359〉 (Mar. 27, 2012). Fothergill, A., and Peek, L. A. (2004). “Poverty and disasters in the United States: A review of recent sociological findings.” Nat. Hazards Rev., 32(1), 89–110. Fouts, L., James, D. L., and Letchford, C. W. (2003). “Pressure distribution on a cubical model in tornado-like flow.” Proc., 10th Int. Wind Engineering Conf., International Association for Wind Engineering, Atsugi, Kanagawa, Japan. Fujita, T. T. (1971). “Proposed characterization of tornadoes and hurricanes by area and intensity.” Research Paper 91, Satellite and Mesometeorology Research Project, Univ. of Chicago, Chicago. Fujita, T. T., and Pearson, A. D. (1976). U.S. tornadoes 1930–74, Univ. of Chicago, Chicago. Gerhart, M. (2011). Down home: Alabama housing recovery in the aftermath of Katrina and implications for the tornadoes of 2011, Equity and Inclusion Campaign, New Orleans. Gurley, K., et al. (2006). “Post 2004 hurricane field survey—An evaluation of the relative performance of the standard building code and the Florida building code.” Proc., 2006 Structural Engineering and Public Safety, ASCE, Reston, VA, 1–10. Haan, F. L., Jr., Balarmudu, V. K., and Sarkar, P. P. (2010). “Tornado-induced wind loads on a lowrise building.” J. Struct. Eng., 136(1), 106–116. Hu, H., Yang, Z., Sarkar, P., and Haan, F. (2011). “Characterization of the wind loads and flow fields around a gable-roof building model in tornadolike winds.” Exp. Fluids, 51(3), 835–851. Impact Forecasting. (2011). United States April & May 2011 severe weather outbreaks, AON Benfield, Chicago. Karstens, C. D., Samaras, T. M., Lee, B. D., Gallus, W. A., Jr., and Finley, C. A. (2010). “Near-ground pressure and wind measurements in tornadoes.” Monthly Weather Rev., 138(7), 2570–2588.

268




Kikitsu, H., and Sarkar, P. P. (2010). “Post investigation on damage to buildings by Ef5 tornado in Iowa, U.S. on May 2008.” J. Wind Eng., 35(3), 73–87. Kosiba, K. A., Trapp, R. J., and Wurman, J. M. (2005). “The 12 May 2004 Harper, KS tornado: Analysis of Dow radar observations of the low level wind field.” Proc., 11th Conf. on Mesoscale Processes and 32nd Conf. on Radar Meteorology, American Meteorological Society, Boston. Kosiba, K. A., Trapp, R. J., and Wurman, J. (2008). “An analysis of the axisymmetric threedimensional low level wind field in a tornado using mobile radar observations.” Geophys. Res. Lett., 35, 1–6. Kosiba, K., and Wurman, J. (2010). “The threedimensional axisymmetric wind field structure of the Spencer, South Dakota, 1998 tornado.” J. Atmos. Sci., 67(9), 3074–3083. Kramer, R. A. (1995). “Advantages and limitations of Benefit–Cost Analysis for evaluating investments in natural disaster mitigation.” Disaster prevention for sustainable development: Economic and policy issues, M. Munasinghe and C. L. Clarke, eds., World Bank, Washington, DC, 61–76. Lee, J. L., Samaras, T. M., and Young, C. R. (2004). “Pressure measurements at the ground in an F-4 tornado.” Proc., 22nd Conf. on Severe Local Storms, American Meteorological Society, Boston, 1–21. Lee, W.-C., and Wurman, J. (2005). “Diagnosing new details about tornado dynamics.” Bull. Am. Meteorol. Soc., 86(11), 1547–1550. Lewellen, W. S. (1993). “Tornado vortex theory.” The tornado: Its structure, dynamics, prediction, and hazards (Geophysical Monograph No. 79), American Geophysical Union, Washington, DC, 19–40. Lewellen, W. S., Teske, M. E., and Sheng, Y. P. (1979). “Wind and pressure distribution in a tornado.” Proc., 5th Int. Conf. on Wind Engineering, International Association for Wind Engineering, Atsugi, Kanagawa, Japan, II-3-I–II-3-II. Mailman School of Public Health. (2010). Legacy of Katrina: The impact of a flawed recovery on vulnerable children of the Gulf Coast, Columbia Univ., New York. Martin, Z., and Readling, B. (2008). “Midwest tornadoes—Performance of wood framed structures.” Wood Design Focus, 18(4), 3–7. McDonald, J., and Mehta, K. C. (2006). “A recommendation for an enhanced Fujita scale.” Wind Science and Engineering Center, Texas Tech Univ., Lubbock, TX. Leadership and Management in Engineering

Melcher, R. (2003). “Natural disaster risk and cost– benefit analysis.” Building safer cities: The future of disaster risk, A. Kreimer, M. Arnold, and A. Carlin, eds., World Bank, Washington, DC, 45–56. Minor, J. E., McDonald, J. R., and Mehta, K. C. National Severe Storms Laboratory Texas Tech University Institute for Disaster Research. (1978). “The tornado: An engineering-oriented perspective.” NOAA Technical Memorandum ERL NSSL-82, National Oceanic and Atmospheric Administration, Washington, DC. Mishra, A. R., James, D. L., and Letchford, C. W. (2005). “Comparison of pressure distribution on a cubical model in boundary layer and tornadolike flow fields.” Proc., 10th Americas Conf. on Wind Engineering, American Association for Wind Engineering, Fort Collins, CO. Mishra, A., James, D. L., and Letchford, C. W. (2008a). “Physical simulation of a single-celled tornado-like vortex, part A: Flow field characterization.” J. Wind Eng. Ind. Aerodyn., 96(8–9), 1243–1257. Mishra, A., James, D. L., and Letchford, C. W. (2008b). “Physical simulation of a single-celled tornado-like vortex, part B: Wind loading on a cubical model.” J. Wind Eng. Ind. Aerodyn., 96(8–9), 1258–1273. National Oceanic and Atmospheric Administration. (2012). “2011 tornado information.” 〈http://www.noaanews.noaa.gov/2011_tornado _information.html〉 (Mar. 27, 2012). Prevatt, D. O., Coulbourne, B., Graettinger, A., Pei, S., Gupta, R., and Grau, D. (2011a). The Joplin tornado of 2011—Damage survey and case for tornadoresilient codes, ASCE, Reston, VA. 〈ftp://gsd.asce.org/ SEI/Prevatt_et_al_(2011)_The_Joplin_Tornado _ASCE_Manuscript.pdf〉 (Mar. 27, 2012). Prevatt, D. O., et al. (2011b). “Building damage observations and EF classifications from the Tuscaloosa, AL and Joplin, MO tornadoes.” Proc., 2012 Structures Congress, ASCE, Reston, VA, in press. Prevatt, D. O., van de Lindt, J. W., and Graettinger, A. (2011c). Damage study and future direction for structural design following the Tuscaloosa Tornado of 2011, Univ. of Florida, Gainesville, FL. 〈http:// www.davidoprevatt.com/wp-content/uploads/2011/ 08/tuscaloosa-tornado-report-final.pdf〉 (Apr. 5, 2012). Prevatt, D. O., van de Lindt, J. W., Gupta, R., and Coulbourne, B. (2011d).“Structural performance— Tuscaloosa tornado.” Structure Magazine, July, 24–26.

269

OCTOBER 2012



Sarkar, P. P., Haan, F. L., Jr., Balaramudu, V., and Sengupta, A. (2006). “Laboratory simulation of tornado and microburst to assess wind loads on buildings.” Proc., 2006 Structural Engineering and Public Safety, ASCE, Reston, VA, 1–10. Sengupta, A., Haan, F. L., Sarkar, P. P., and Balaramudu, V. (2008). “Transient loads on buildings in microburst and tornado winds.” J. Wind Eng. Ind. Aerodyn., 96(10–11), 2173– 2187. Storm Prediction Center, National Oceanic and Atmospheric Administration. (2011). “Climatological or past storm information.” 〈http://www .spc.noaa.gov/climo/online/monthly/2011_annual_ summary.html#〉 (Mar. 27, 2012). U.S. Census Bureau. (2005). “American Housing Survey.” U.S. Department of Housing and Urban Development, U.S. Government Printing Office, Washington, DC. U.S. Census Bureau. (2010a). “Selected housing characteristics: 2006–2010 American Community Survey, Joplin, MO.” Washington, DC. U.S. Census Bureau. (2010b). “Selected housing characteristics: 2006–2010 American Community Survey, Tuscaloosa, AL.” Washington, DC. U.S. Census Bureau. (2010c). Small area income and poverty estimates, Washington, DC. van de Lindt, J. W., et al. (2011). “Dual objective design philosophy for tornado engineering.” Proc., 2012 Structures Congress, ASCE, Reston, VA, in press. Venton, C. C., and Venton, P. (2004). Disaster preparedness programmes in India: A cost benefit analysis, Humanitarian Practice Network, Overseas Development Institute, London. Vermeiren, J. C., Stichter, S., and Wason, A. (1998). Costs and benefits of hazard mitigation for building and infrastructure development: A case study in small island developing states, Organization of American States, Washington, DC. Walker, A. S. (2007). “Spatial and temporal analysis of tornado fatalities in the United States: 1880– 2005.” Weather Forecasting, 22(6), 1214–1228. Wang, H. (2002). “Fluid-structure interaction of a tornado-like vortex with low-rise structures.” M.S. thesis, Dept. of Mechanical Engineering, Texas Tech Univ., Lubbock, TX.

OCTOBER 2012

Wurman, J., Alexander, C., Robinson, P., and Richardson, Y. (2007). “Low-level winds in tornadoes and potential catastrophic tornado impacts in urban areas.” Bull. Am. Meteorol. Soc., 88(1), 31–46. David O. Prevatt is assistant professor of Civil & Coastal Engineering, Engineering School of Sustainable Infrastructure & Environment, University of Florida, Gainesville. He can be contacted at [email protected]. John W. van de Lindt is professor and Drummond Chair in Civil Engineering, Department of Civil, Construction and Environmental Engineering, University of Alabama, Tuscaloosa. Edward Back is associate professor and director of construction engineering and management, Department of Civil, Construction and Environmental Engineering, University of Alabama, Tuscaloosa. Andrew J. Graettinger is associate professor of civil engineering, Department of Civil, Construction and Environmental Engineering, University of Alabama, Tuscaloosa. Shiling Pei is assistant professor of structural engineering, Department of Civil and Environmental Engineering, South Dakota State University, Brookings. William Coulbourne is director, Wind and Flood Hazard Mitigation, Applied Technology Council, Rehoboth Beach, DE. Rakesh Gupta is professor of structural engineering, Department of Wood Science and Engineering, Oregon State University, Corvallis. Darryl James is professor of mechanical engineering, Department of Mechanical Engineering, Texas Tech University, Lubbock. Duzgun Agdas is lecturer, Construction Engineering and Management, Engineering School of Sustainable Infrastructure and Environment, University of Florida, Gainesville. LME

270



Making the Case for Improved Structural Design ... - Semantic Scholar

Making the Case for Improved Structural Design ... - Semantic Scholar

Suggest Documents

Making a case for introspection - Semantic Scholar

Hierarchical structural design for fracture ... - Semantic Scholar

An Improved Method for the Design of Variable ... - Semantic Scholar

Structural basis for improved efficacy of therapeutic ... - Semantic Scholar

making the case for design in the development sector - Squarespace

Developmentally Situated Design (DSD): Making ... - Semantic Scholar

Structural Conformance Checking with Design ... - Semantic Scholar

Multimedia Presentation Design on Structural ... - Semantic Scholar

Improved Adaptive Droop Control Design for ... - Semantic Scholar

Towards TechniTues for Improved OO Design ... - Semantic Scholar

THE SCENESCORE FOR IMPROVED PRE ... - Semantic Scholar

Making the Case for Ontology

The Optimum Structural Design for Spiral Case in Hydraulic ... - Core

Interactive Parametric Tools for Structural Design - Semantic Scholar

A reusable structural design for mobile ... - Semantic Scholar

Risk-Based Seismic Design for Optimal Structural ... - Semantic Scholar

Risk-Based Seismic Design for Optimal Structural ... - Semantic Scholar

Interactive Parametric Tools for Structural Design - Semantic Scholar

Improved Cost Function in the Design of Boolean ... - Semantic Scholar

Informing design for tangible interaction: a case for ... - Semantic Scholar

The Design of Implicit Interactions: Making ... - Semantic Scholar

The Boulevard of Inspiration: Making design ... - Semantic Scholar

The Design of Implicit Interactions: Making ... - Semantic Scholar

Structural Insights into the Molecular Design of ... | Semantic Scholar