Powered by FirstSource
Products Articles
Advanced Search | Search Tips


Structural engineers are discovering new ways to deal with this deadly phenomenon in the wake of the 9/11 attacks

By Gordon Wright, Executive Editor

Reprinted courtesy of Building Design & Construction, 11/01/2003.

The destruction of the World Trade Center towers has heightened concern about the progressive collapse of buildings. Ironically, engineers disagree on the definition of this phenomenon, and many argue that the factors surrounding the 2001 terrorists attacks were too extraordinary to classify the Twin Towers' failure as progressive collapse.

"Progressive collapse is a nebulous term," says engineer Shankir Nair, SVP with Chicago-based architect Teng. "Any collapse is progressive in that one thing fails, something else fails, and so on." Nair says that a significant problem results only if the subsequent failure is "disproportionate."

Jon Magnusson, CEO of structural engineer Magnusson Klemencic Associates, Seattle, recalls that attendees at a 2002 workshop sponsored by the National Institute of Science & Technology had difficulty agreeing on a definition of progressive collapse. "The only thing that everyone could agree on was that the World Trade Center performed very well in resisting progressive collapse," he says, noting that one of the towers held despite the loss of two-thirds of its columns on the side impacted by a jetliner. "Its collapse was progressive, but it didn't fit the traditional definition," he says.

Peter DiMaggio, senior associate with structural engineer Weidlinger Associates, New York, N.Y., says research into progressive collapse gained momentum after the 1995 bombing of the Murrah Federal Office Building in Oklahoma City.

Because the exact nature of an event that might impair a structural system typically cannot be known, the best approach is an "alternate path" design that will resist failure if a primary column were removed, according to DiMaggio.

He says that design criteria for progressive collapse is less restrictive than for conventional loading for service loads (dead, live, or wind). For example, if a floor were to sag several feet as the result of structural damage, building occupants should be able to safely reach an egress route. "We usually allow a building to crack and bend, as long as it doesn't collapse," he says. "We do this for cost-effectiveness, and to maintain the building's architectural intent."

A variety of techniques are available to analyze structural integrity. DiMaggio sees a trend toward the use of finite element analysis, which can model a beam/column connection and provide an accurate representation of forces and deflections that would be difficult to detect with less sophisticated software. "You're able to more efficiently resist the type of loads you would see if you were to remove a column or have a particular local failure, so you end up with more-efficient designs," he says.

DiMaggio says that in recent years, designs to resist blast and progressive collapse are now more subtly integrated aesthetically.

Magnusson notes that the U.S. General Services Administration's basic requirement for resisting progressive collapse assumes that a structural system will be able to redistribute the load to other columns if a perimeter column is removed. He says some approaches being considered to increase structural integrity when a single column is impaired might actually be detrimental if damage is more extensive. In the event of a major fire, for example, several columns could simultaneously reach a temperature that could cause their failure.

Statistically speaking, high-rise office buildings have an excellent performance record under fire conditions. Only one fire death occurred in office buildings of more than seven stories during the period 1991 to 1998, according to data compiled by the National Fire Data Center, Emmitsburg, Md., and the National Fire Protection Association, Quincy, Mass.

The Federal government has implemented blast-resistant designs for some overseas projects for as long as a quarter century, and had a plan in place for their application to domestic buildings before 9/11, according to DiMaggio.

The growth in concern about structural integrity has come from private developers, who are trying to determine if their properties could be damaged even if they were not a primary target for terrorists. Given the extensive damage inflicted on the buildings surrounding the World Trade Center, this is hardly an idle concern. A survey by BOMA International found that building owners spent 14% more for security last year than they did in 2001.

Magnusson emphasizes that a determination of the most likely hazard the structure might face is the first step for optimizing a design that resists progressive collapse. "We're not going to create real safety until the hazard definition has been addressed," he says. "Then we can figure out the smartest ways to prevent collapses."

One approach may be to harden the first two or three stories of a building to make it blast-resistant, "rather than spending millions on the whole building trying to design for something that may never happen," he says.

Another alternative is to embed steel cables in the floors at the building perimeter. The cables would redistribute gravity loads to other columns if an exterior column were to be severely damaged.

Nair believes the most important step in preventing disproportionate collapse centers on tying structural components together. Noting the inherent advantages of steel (greater ductility) and cast-in-place concrete (greater continuity, since it is poured simultaneously), he cites three factors that reduce the potential for progressive collapse: greater local resistance (to prevent initial failure); redundancy (providing alternate load paths); and continuity (tying components together).

Larry Griffis, president of the Structures Division of Houston-based engineer Walter P. Moore Engineers and Consultants, recalls that progressive collapse was a much-discussed topic years ago when the American Concrete Institute's ACI 318 code first included provisions to address the issue. It specified the incorporation of "integrity steel" at column locations in order to prevent a floor from collapsing as a result of column failure.

Griffis anticipates that both concrete and steel codes will adopt new provisions related to progressive collapse. The 2005 edition of the American Institute of Steel Construction's Specification for Steel Buildings will, for the first time, have a chapter that covers designing for fire conditions.

Due to its function of resisting wind loads, the structure of a high-rise building also contributes to its ability to support gravity loads — a characteristic of increasing importance as building height increases, says Griffis. Tall buildings have substantial redundancy simply because they are designed for wind loads that are unlikely to occur simultaneously with a building emergency, he says.

Griffis foresees a trend in the U.S. toward the use of performance specifications that would supplant the current prescriptive approach which dictates, for example, how to achieve a three-hour rating for columns and a two-hour rating for beams. A performance specification would more closely relate fire resistance to the amount of gravity load a member could support at elevated temperatures.

"I think there will be a move toward treating fire more as a structural load, rather than the prescriptive approach now used," Griffis says. "The Europeans are way ahead of us on this." He cites Spanish and German projects he has worked on that utilized a wide-flange column infilled between flanges with concrete, which acts as a heat sink by drawing heat away from the steel.

Exiting strategies get closer scrutiny

The deaths of six people in a Chicago high-rise office fire last month has put a renewed focus on evacuation procedures.

To more intelligently engineer the egress capabilities of major buildings, designers can turn to sophisticated computer simulations. Egress evaluation systems with visualization capability have been available for more than a decade, says Brian Meacham, principal fire and risk consultant with Arup's Fire Engineering Design Group in Westborough, Mass. They became more prominent in the early 1990s as designers sought tools to help them perform evacuation analyses more quickly.

For most buildings, exiting requirements are usually spelled out by the local building code. But complex buildings may require the use of calculations to establish factors such as maximum time to evacuate a building. Computer models allow these calculations to be made much more quickly. Programs with visualization capability graphically simulate the movement of building occupants during an evacuation.

"It's a way to do the calculations more quickly, with the added benefit of having visualization to get an idea of how people move and where the choke points are in the exiting process," Meacham says. "This is not as easy to understand when you do a calculation using just a spreadsheet or calculator. The animation allows you to see how people move, and where they queue up."

These programs not only point out inadequacies in the exiting design, but also may demonstrate where codes are overly conservative. For example, the code may specify a maximum dead end or travel distance because it assumes a certain amount of time will be required for people to move through an area. A computer analysis may show that people can move through the area at a faster rate, thereby documenting that a longer travel distance than required by code is warranted.

While the code prescribes the number of exits required, says Meacham, the analysis can show how many are needed to get people out safely.

Arup typically performs egress analyses in conjunction with fire-effects modeling to indicate the expected spread of fire and smoke. "What you're trying to show is that an egress system design allows for enough people to get out in the event of a fire," he says.

An assessment of the expected path of smoke will indicate the type of HVAC system that is needed for smoke extraction or to maintain an area that is already clear of smoke.

Jeff Tubbs, an associate with Arup, says egress analysis systems with visualization capability, which are marketed by several companies, can readily identify "pinch points" in the exiting system, such as a narrow doorway. He believes that the trend toward the use of performance-based specifications will make designers more receptive to the kinds of alternatives these studies can identify.