Recent Highlights in Materials Reliability Division

The Structural Steel of the World Trade Center Towers

Materials Reliability Division Contacts:

and

Stephen W. Banovic, Richard J. Fields, Timothy J. Foecke, William E. Luecke, and Frank W. Gayle, Metallurgy Division

In terms of the steel used, the World Trade Center was very complex. Plans called for steels of fourteen different yield strengths between 36 ksi and 100 ksi. Many of these steels were proprietary and were not supplied to any existing ASTM specification. To further complicate matters, four different fabricators provided steel for the upper stories, and each used steel from more than one supplier. Most of the records that documented the steel actually used were either not preserved or destroyed in the collapse.

The MSEL Metallurgy and Materials Reliability Divisions have determined properties of the steel recovered from the WTC site and characterized failure modes associated with pre-collapse damage. This information has been provided for use in models of the building response to airplane impact and fire. In 2004, all major tasks were completed.

Task 1 — Collect and Catalog Physical Evidence

The Structural Engineers Association of New York (SEAoNY), later supplemented by NIST personnel, spent countless hours selecting steel in the New Jersey recycling yards for later forensic analysis. After shipment to NIST, the Metallurgy Division cataloged the 236 recovered steel components. Since many pieces were stamped and painted with unique serial numbers correlated with the building plans, it was often possible to identify the exact original location in the building.

As a second part of this task, we also researched contemporaneous construction documents and steel specifications. This task helped identify steel suppliers and fabricators, as well as material substitutions.

Task 2 — Categorize Failure Mechanisms Based on Visual Evidence

The damage to the steel components of the buildings represents an important check on the accuracy of the fire and impact modeling, with the caveat that the recovered steel could have been damaged before the collapse, during the collapse, or during the salvage efforts.

Image enhancement techniques were used to study photographs of the impact zones of the towers to establish the failure mechanisms of the perimeter columns. By making montages of many different images, it was possible to produce a nearly smoke-free, composite view of the columns that the aircraft struck. The columns failed by several different mechanisms. One important result was the demonstration that several key impact-zone columns in the NIST inventory were damaged primarily in the aircraft impact, and not in the subsequent collapse or recovery efforts. Figure 1 is a map of the impact zone of WTC 1 that illustrates the nature of the damage to the individual columns.

Figure 1: Enhanced image of the impact zone has allowed determination of failure modes in the perimeter columns. An outline of the plane is superimposed. Steel panels recovered for the investigation are highlighted in color.

Tasks 3 & 4 — Determine Steel Properties to Support Airplane Impact and Structure Performance Studies; Correlate Properties with Those Specified in the Construction

One important part of this task was to assess the quality of steel and determine whether it met the original specifications. In support of this goal, the group conducted more than 200 room-temperature tensile tests on samples from 39 different building components, which represented all the steels relevant to the investigation. Tests indicate that the steels generally met or exceeded the specified minimum strengths (Figure 2).

Figure 2: Yield strength — measured vs. specified minimum. The few anomalous data can be attributed to damage incurred in the collapse that removed the yield point behavior, or to the natural and accepted strength variability of structural steel.

Steel specifications also regulate the chemical compositions. More than 350 different specimens, representing nearly every relevant steel component in the building, were analyzed for chemical composition and all of the sections relevant to the investigation. The chemical analysis helped confirm archival information that an American steel mill supplied steel for parts of the perimeter columns. The analyses also support the conclusion that the steels in the buildings met the specifications called for in the building plans.

The strength of steel increases as the deformation rate increases. Accurately modeling these rate effects is essential for estimating the energy and momentum of the aircraft remnants which damaged the core columns. To establish the strain rate sensitivity of the relevant steels, we have conducted more than 100 tests on specimens from both the perimeter and core columns at rates up to 2000 s^–1 (200,000 % deformation per second).

Ultimately, the fires in the towers weakened the steel structure leading to collapse. We conducted more than 100 high-temperature tensile and creep tests to provide models of the deformation of structural steel at elevated temperatures for use in models of the building response to the fires.

Task 5 — Analyze Steel to Estimate Temperature

Just as the nature of the deformation and failure of the recovered steel reveals information about the impact and collapse, the steel can also contain evidence of its exposure to the elevated temperatures in the fire. The situation is made more complex because much of the steel was also exposed to extended fires in the rubble before recovery. Several different methods, both conventional and novel, were examined for estimating high-temperature excursions seen by the steel. Only one method proved to be robust and easy to implement: paint on steels that reached temperatures over 250 °C cracked from the difference in thermal expansion between the paint and the steel.

From photographs showing fire in windows, all recovered columns that may have been exposed to pre-collapse fire were identified and characterized using the paint crack technique. Figure 3 illustrates one of the columns that showed evidence of having exceeded 250 °C.

Outputs

Outputs for the project include parts of the June 2004 Progress Report on the World Trade Center Disaster (NIST SP 1000-5) and numerous memos for contractors that explain how to employ Project 3 data and deformation models. The final NIST investigation report will be released in the spring of 2005.

Back to top

MRD Home