Bracing For A Storm
UF engineers say lessons learned from hurricanes can also be applied to tornadoes
by Donna Hesterman
If there is anything more surreal than the sight of a tornado tossing slabs of concrete and steel into the sky like bits of fluff, it’s the post-apocalyptic tableau of the aftermath.
Wind-battered houses leveled or shifted on their foundation; toys and furniture strewn across fields and lodged in treetops; stunned survivors sifting through rubble in search of their belongings.
That was the scene in April 2011 when a team of forensic engineers led by a University of Florida researcher arrived in Tuscaloosa, Ala., on the heels of a deadly tornado.
The twister had stayed on the ground for almost 9 minutes wreaking havoc on an 8-mile stretch of densely populated town. It killed 65 people, injured at least
a thousand more and caused billions of dollars in property damage. No one could have believed it at the time, but just one month later, Mother Nature would deal an even bigger blow to the people of Joplin, Missouri. That twister earned the title of deadliest single tornado in U.S. history, killing 117 people and causing an estimated $2.8 billion in damage.
The losses in both cases were staggering – the worst seen in U.S. history from tornadoes. But much of the damage could have been avoided, says David O. Prevatt, assistant professor of civil and coastal engineering at UF. He led a rapid response survey team sponsored by the National Science Foundation that analyzed construction damage caused by the tornadoes.
In Joplin and Tuscaloosa, Prevatt’s team found that the winds were survivable along a significant portion of the tornadoes’ paths, and that a few basic modifications in how homes were engineered could have dramatically changed the outcome of the storm.
“We’ve understood for a long time the physics behind keeping a structure together in wind pressures from less-severe tornadoes,” says Prevatt. “But that knowledge hasn’t been translated into building designs or guidelines for builders.”
Designing a home to withstand a tornado strike has largely been off the radar for builders because the chance of it being hit seemed too remote to warrant the extra cost. But as the U.S. population grows and cities spread to cover more of the landscape, homes and businesses are statistically more likely to experience a tornado strike.
Consider the case in Joplin, he says. If the twister had touched down just two miles farther south, it would have gone over farmland and no one would really be talking about it. But since it tore through a city, the damage was catastrophic – a loss comparable to nearly 3 percent of the state’s entire gross domestic product for the year.
Urban sprawl is upping the odds that tornadoes will strike where people live. Communities can’t afford to ignore the risks any longer, Prevatt says.
Five days after the tornado ripped through Tuscaloosa, Prevatt and graduate student David B. Roueche led a team of researchers from six universities, the construction industry and the non-profit Applied Technology Council to survey the damage.
“There was some damage in isolated areas on the outskirts approaching Tuscaloosa – trees down here and there. But in town, the devastation was truly chilling,” says Prevatt. “Whole structures looked like they had been simply pushed over, as if some massive finger had given them a little shove.”
Once at the site, he coordinated with emergency management crews and organized his group into smaller teams to cover the area. All told, they took more than 3,000 photos and collected information describing in detail all the ways that hundreds of wood-framed structures had failed.
They identified the structural connections between the gable-end walls and roof trusses as a key point of failure in many homes that were destroyed. Simple toe-nailed connections gave way when pressures increased beyond their capacity.
Generally, he says, the quality of construction was poor, in both older and newer homes. Wood wall plates weren’t bolted to foundations, brick facades weren’t tied to sheathing properly, and roof attachments were inadequate – all common practices in housing construction that seems adequate until the structures are subjected to high winds.
The findings were no different in Joplin one month later.
In both cities, Prevatt’s team used digital cameras synced to a GPS database to photograph the damage. The data was then downloaded into a digitized map with information embedded at various points that associate wind speed estimates to the details of each building’s destruction.
The database can be used by weather scientists studying the inner-workings of tornadoes or by engineers who want to know how various materials and construction methods bear up against a storm. But perhaps the most illuminating aspect of the project was that it revealed a distinct pattern in the damage.
Once mapped, the data showed that most of the winds that would be considered “un-survivable” were confined near the tornadoes’ center paths.
“Just 100 to 200 yards to the right or left of the main path, wind forces dropped down into a range that you see in less-intense tornadoes,” Prevatt says.
The materials and techniques needed to keep a home safe in those conditions are already available. Florida builders have been incorporating them into new construction since 1992 when Hurricane Andrew blew through and left more than 200,000 people homeless.
The Category 4 storm packed winds in excess of 175 miles per hour that would have leveled almost any structure in its path. But as with the monster tornadoes of 2011, much of the destruction was caused by lesser winds whirling within the larger system.
The types of metal hurricane ties, clips and straps used to reinforce construction in hurricane-prone Florida could have made a significant difference in Tuscaloosa and Joplin, according to reports based on Prevatt’s data. Houses along the outer quarter of the tornado’s track tended to come apart at the seams where connections failed between walls, foundations and roofs.
Only at the center of the path were homes completely obliterated.
This is the sort of information people need when they are making decisions about how their homes are put together, Prevatt says. He and colleagues Forrest Masters and Kurt Gurley study weather impacts on man-made structures at UF’s Engineering School of Sustainable Infrastructure and Environment.
“Storms interrupt prosperity,” says Masters, an assistant professor of civil and coastal engineering. “Losing a house in a storm doesn’t just affect the individual who owned the home. It’s a shared loss that takes money out of public coffers that could be used in education programs, infrastructure and development.”
The researchers do their testing in a 9,000-square-foot metal building on campus called the Powell Family Structures and Materials Laboratory. The lab is tall enough to accommodate a 36-foot-tall crane and is packed with an assortment of steel-framed machines that pry, twist and pummel architectural structures in a variety of ways.
Masters’ team is currently testing a new composite panel that can be used for walls or roofs in residential structures. The panes are being developed by colleagues at the University of Alabama at Birmingham, but UF teams are testing them at the lab. The panels are lightweight, resilient, energy-efficient and perfect for modular construction, he says.
Similar products are already on the market, but the University of Alabama design team is interested in enhancing characteristics that will help prevent failure in a storm scenario.
Masters uses a Seussian-looking contraption called a pressure loading actuator to see what the panels can take. A bank of powerful centrifugal fans drives air through a series of valves and large yellow ducts that blast wind-driven rain into a sealed chamber holding an 8-foot square of the panel.
The actuator occupies a space inside the laboratory about the size of a two-story single-car garage, and it is the smaller but mightier version of its forerunner – a hurricane simulator known as Medusa. Medusa has been a fixture of the hurricane studies effort at UF for three years, but at one-tenth the size, the actuator delivers six times the wind pressure that its predecessor does and costs much less to operate.
During hurricane season, Masters leads a mobile storm-chasing unit that captures real-time data about storm dynamics and how they affect residential structures. Over the years, the program has served to fill critical gaps in what researchers know about storms and how wind velocity and pressure can damage a building. Masters has led the team since 2004.
“We can’t afford to wait around for real storm conditions if we want to keep innovation moving at a steady pace,” Masters says. “So we bring the storm to the lab.”
Back in the lab, Prevatt’s research team is testing a structural foam adhesive designed to increase a roof’s staying power. The product can be sprayed on the underside of existing roof sheathing for less than $5,000 in most cases, and can keep a roof intact at more than twice the wind speeds that traditional roof attachment methods can handle. His students test the adhesive using a pressure loading actuator to simulate suction pressure created by strong winds blowing across a rooftop.
“There is a mistaken belief that the weight of the roof is sufficient to keep it on during a storm – that you don’t need to reinforce attachment points between the roof and the rest of the structure,” says Prevatt. But that just isn’t the case.
It’s a problem that has occupied his mind since he was a young engineering student in Trinidad.
“A tropical storm had passed through and I went out to look at the damage,” he says. “It wasn’t too bad – mostly palm trees leaning and a few things stuck in walls like missile projectiles. But then there was a roof – a perfectly intact roof – lying upside down in the middle of the road.”
With the homeowner’s permission, he scoured the wreckage and quickly determined that the roof had not been bolted to the walls below.
“I was all business, analyzing the wreckage like a scientist and the homeowner stood by politely answering my questions,” he says. “But then as I was leaving I saw her face crumble as if the reality of it all was finally sinking in.
“It was a completely avoidable tragedy,” he says.
Fast-forward 20 years to a Tuscaloosa hotel conference room where Prevatt is debriefing his damage assessment team.
“I was making a list of our findings on the whiteboard, and then it hit me. For years we have rushed out to these sites to collect data and assign EF-ratings to these storms – and then we walk away as if we have nothing else to contribute,” he says. “That’s what needs to change.”
Prevatt and his UF colleagues want to pick up where engineers have left off in the past. They are adapting what they have learned from many years of hurricane research in Florida to develop new technologies for tornado-resilient homes. He recently received a pretigious National Science Foundation CAREER Award of $400,000 to fund the work.
But beyond the physics and engineering challenges, Prevatt says, there is the larger issue of getting those solutions out of the lab and into the economy. Retrofits and more expensive construction methods can be a tough sell for homebuyers. People need to know what engineers know so that they can make informed decisions.
“Because when a big storm hits,” he says, “we all pay.”
Assistant Professor of Civil and Coastal Engineering
(352) 392-9537 ext. 1505
David O. Prevatt
Assistant Professor of Civil and Coastal Engineering