By David Stevenson, P.Eng., Yolles
Structures: The Pentagon & 9/11Engineering
At 9:38 a.m. on September 11, 2001 a hijacked Boeing 757-200 crashed into the first storey of the Pentagon in an act of terrorism. A portion of the building was severely damaged in the attack, and 19 ...
At 9:38 a.m. on September 11, 2001 a hijacked Boeing 757-200 crashed into the first storey of the Pentagon in an act of terrorism. A portion of the building was severely damaged in the attack, and 19 minutes after impact it collapsed. Sixty-four people on board the aircraft and 125 occupants of the Pentagon were killed.
Within hours of the attack, the American Society of Civil Engineers (ASCE) had assembled a building performance study team of experts. The experts included structural, fire and forensic engineers, and their purpose was to evaluate the structural performance of the building under the initial impact as well as during the subsequent fire. In January 2003, the ASCE published their findings in the The Pentagon Building Performance Report.
The Pentagon, located near Washington, D.C., is the headquarters for the United States Department of Defense. It is one of the largest office buildings in the world, with approximately 6.6 million square feet of floor space on five storeys. Five-sided in plan, the building consists of five rings of office space on the upper three storeys (each separated by light wells) and two larger rings of office space at the first two levels.
Originally completed in January 1943, the building was undergoing a major renovation at the time of the terrorist attack. The renovation of “Wedge 1,” which happened to coincide with the location of the aircraft impact, had been essentially completed by September 11, 2001. It is believed that this is one reason why the damage to the building was not more extensive. While most of the renovations had been for items such as new elevators, mechanical and electrical equipment and utility tunnels, the exterior walls and windows had been upgraded to increase their resistance to extreme pressure. The upgrades included the addition of structural steel framing on the inside of the existing exterior wall, the replacement of the existing windows with 1,600 lbs. blast resistant windows, and the addition of a “geo-technical mesh” that was stretched and fastened to the inside of the structural steel framing. The purpose of the mesh was to prevent debris from projecting into the building and injuring the occupants in the event of an external explosion.1
The original cast-in-place reinforced concrete building structure had not been changed. It was designed in accordance with the American Concrete Institute’s standard ACI 501-36. The typical floor framing consisted of closely spaced columns (varying in size from approximately 21″ square at the base to 14″ square at the upper storeys) supporting short span beams, girders and one-way slabs. The 5.5″ thick slabs spanned 10′ between 14″ x 20″ beams that in turn typically spanned 10′ to 20′ to supporting girders. The girders, which were parallel to the exterior walls, spanned 20′ between supporting columns.
The roof framing was similar in concept to the floors. The perimeter exterior walls were clad in 5″ thick limestone and backed with 8″ thick unreinforced brick infill. Most of the remaining exterior walls (those adjacent to the light wells) were constructed of 10″ concrete. The design live load for the building was 150 psf.
The performance of the structure is said to have been enhanced by key features that were inherent in the original design and detailing:
Approximately 50% of the bottom beam reinforcement was extended through the columns, resulting in continuous bottom reinforcement.
All columns that supported more than one level were reinforced with spiral type horizontal reinforcement.
Hit and impact
Through the use of security camera photographs, eyewitness accounts and on-site investigations, the building performance study team has been able to assemble a credible description of the events that transpired from just before the aircraft hit to the ultimate collapse of the structure. Their understanding of the events and the extent of the damaged areas observed on site seem to correlate closely with predicted responses of the structure to the impact and resulting fire.
The Boeing 757-200 aircraft, (carrying approximately 5,300 gallons of fuel) weighed an estimated 181,500 lbs. and was travelling at approximately 530 m.p.h. when it hit the building. Eyewitnesses reported that the aircraft was flying only a few feet above the ground and that it rolled slightly to the left immediately before impact. In fact, it is believed that the right wing hit an emergency generator located outside the Pentagon and the left engine hit the ground. As a result, it appears that the wing tips may have broken off outside the Pentagon perimeter wall.
The slight roll to the left resulted in the right wing coming up and impacting the second floor slab and the left wing passing down below the second floor slab and hitting the first storey columns. It is believed that this impact at the building perimeter resulted in many of the columns being destroyed as well as the remainder of the aircraft wings and tail section. The fuselage penetrated the building below the second floor slab, slid across the slab-on-grade and continued into the building to a maximum depth of approximately 310 (about twice the length of the aircraft). As the fuselage and remainder of the aircraft plowed through the first storey, additional damage was inflicted on many of the interior columns.
Since a number of perimeter columns were destroyed on impact, a portion of the structure was left unsupported and deflected downwards 18* to 24*. The structure was able to remain in this deformed position for 19 minutes before all five levels collapsed.
Before the damaged areas were demolished, the study team carried out two inspections of the site. They made several observations:
The exterior of the building showed clear signs of the intense fire that occurred within the building. The fire damage observed was similar to other serious office building type fires.
Several columns were severed completely by the impact.
Several columns experienced significant lateral deformations as a result of the impact.
Several columns exhibited signs of severe fire damage.
In many areas, the floor slabs and beams also exhibited signs of severe impact and fire damage.
The recently reinforced portions of the exterior facade performed reasonably well in the areas adjacent to the impact zone.
Figures in the report give details of the overall extent of the damage to the first storey columns, the extent of the area that collapsed, as well as areas where the second floor slab was affected. It is evident that the collapse extended to a relatively small portion of the area that suffered significant damage.
Following the detailed investigations of the building performance study team, it became evident that three key issues required further analysis: first, the effect of the impact of the aircraft on the spiral reinforced columns; second, the ability of the floor framing system to remain in place after the removal of a significant number of the supporting columns below; third, the effects of the fire loading that eventually resulted in the partial collapse of the structure 19 minutes after the aircraft impact.
Spiral reinforced columns
With respect to the ability of the reinforced columns to withstand impact, it was determined that their spiral reinforcement not only had a positive influence on their ductility (i.e. their ability to undergo significant lateral deformations without failure), but also that the columns had a much higher shear capacity than similar columns detailed with conventional horizontal column ties. In fact, the analysis indicated that if conventional column ties were provided, all columns in the affected area would have been destroyed and likely a larger portion of the structure would have collapsed. In addition, the vertical reinforcement in the columns was extended so that there was sufficient anchorage to develop the full capacity of the vertical bars. As such, none of the column failures were a result of the vertical bars being “pulled out” of the columns above.
The study team did a simplified analysis of the floor framing system in an attempt to determine its ultimate load carrying capacity and its ability to remain in place after the removal of many of the supporting columns. The analysis indicated that the type of floor system used had a significant inherent strength (based in part on the presence of continuous bottom reinforcement) and that likely it would have remained intact in the absence of the additional fire load that existed.
In order to assess the effect of the fire loading on the structure, the team had to make several assumptions. The assumptions included the fire fuel load (i.e. the amount of fuel that entered the building at the time of impact), the ventilation factor, the size of the area in which the fuel was consumed and the amount of combustibles in the building. Based on the assumptions made, it was estimated that a maximum temperature close to 1,560F (850C) would have been reached about 30 minutes after the fire started.
The thermal analysis also considered the effect of elevated temperatures on the performance of the floor beams and supporting columns. Two different conditions were investigated. The first assumed that no damage or loss of concrete cover occurred to either the beams or columns. The second assumed that the impact damage to the beams and columns resulted in the complete loss of the concrete cover. With the cover removed the steel reinforcement would be directly exposed to fire and the load carrying capacity of the structural system could deteriorate more quickly.
The analysis indicated that in the first condition with the cover intact, the time until the steel reinforcement reached the critical temperature of 932F (500C)– the impending yielding temperature of the reinforcement according to the Eurocode — varied from 100-130 minutes for the beams, and from 125-155 minutes for the columns. The range in time for each element depends on the time temperature curve that is used in the analysis. In the second condition with the cover removed, the time to reach the critical temperature reduced drastically to 12 to 20 minutes for the beams and 25 to 50 minutes for the columns. These reduced times correlate very closely to the actual time of collapse after the initiation of the fire.
Conclusions and recommendations
The building performance study team found through site investigations and analysis that the impact of the aircraft destroyed approximately 30 columns. They found that it had also severely impaired the load carrying capacity of approximately 20 additional columns at the first storey, as well as approximately six perimeter columns at the second storey.
The fire damaged a large area of the first storey, a significant area of the second storey and limited areas above the third storey. It is believed that the damage caused by the initial impact as well as the effect of the intense fire, caused the localized collapse of the building.
Although the number of columns damaged was extensive, the portion of the structure that collapsed was comparatively small. The study team attributed the limited collapse to several factors. These included: the short spans between columns; the redundancy in terms of alternative load paths inherent in the two-way beam system; the continuity of bottom reinforcement in the beams and girders; the reserve load carrying capacity in the structural elements; and the energy-absorbing capacity of the spirally reinforced concrete columns.
The study team’s report suggests that these factors should be considered in the design of structures that are required to resist progressive collapse.
The team also identified the need for more research in several key areas. These include: consolidation of information on the prevention of progressive collapse, investigation of the effects of extreme column deformations on their load-carrying capacity, and investigations into the energy-absorption capacity of concrete columns. The team also suggested more research on the ability of structures in general to withstand extreme impacts.CCE
David Stevenson, P.Eng., is a partner and the technical director of Yolles, structural engineers of Toronto.
The images are all taken from the Pentagon Building Performance Report, published by the American Society of Civil Engineers, January 2003, www.asce.pubs.org. Reproduced with ASCE permission.