Wolftrap+Performing+Arts+Center+-+Weld+Failure

Wolftrap Performing Arts Center - Weld Failure //Christopher Barlow, MAE/BAE, The Pennsylvania State University Class of 2017 // toc
 * Fairfax, VA - January 1985 **

Introduction
Through the peaceful countryside, over grassy hills and tall forests, the sound of music resonates through Vienna in summertime. But this is not Vienna, Austria. This is Vienna, Virginia, a small town thirty miles west of Washington, DC. And of course this is not Julie Andrews and the beloved Von Trapp Family. This is the //Wolf//trap Performing Arts Center. For nearly fifty years, the Wolftrap Performing Arts center has dazzled its patrons with hundreds of musical guests including legendary world-renowned musicians of every genre as well as today's pop stars. Central to this cultural hub is the Filene Center, a large wooden indoor and outdoor auditorium with seating for 7,000. Truly, the experience offered by this venue is unique and inspiring.

However, this peaceful sanctuary was also once the source of a tremendous amount of panic and turmoil for those involved in its construction. Not long after its completion and totally without warning, a monstrous 8-foot crack appeared in a major steel roof girder above the Filene Center, resulting in a 9.5 inch deflection (Figure 1) (Feld & Carper 1996). The park was immediately closed and a lengthy investigation commenced to determine the source of the failure. The drama intensified as the architects, engineers, steel suppliers, fabricators, and contractors involved in the project waited anxiously to read the report that would determine who was at fault and who would pay the hefty cost for repairs. This wiki will summarize various theories regarding the cause of the failure, the final conclusion of the official report, subsequent legal ramifications, as well as lessons learned for future construction of large-scale outdoor venues such as this.

Analysis & Material Testing
Thorough structural analysis of the member alleviated the original designers of fault in the failure. Despite running the numbers numerous times through several load cases, a maximum stress could not be generated that exceeded the allowable stress for the member (Kaminetzky 1993). The highest tensile stress generated by the structural analysis, which was computed for the load case of dead lead plus snow load and temperature load, was only 164 MPa (Table 1). This number does not even come close to the tensile properties provided in the ASTM A572 Gr. 50 Specifications. Therefore, it could be reasonably concluded that the fault was not in the original design of the structure. Given the lack of theoretical evidence for the cause of failure, a close examination of the actual structure was carried out next. The girder was made from welded steel plates forming a large rectangular box (Maranian 2010). Typically, steel behaves in a highly elastic manner, undergoing a considerable amount of strain before rupturing. However, the severity of the crack and its rapid propagation seemed to suggest a brittle failure of the material (Figure 1). Flaws in the material were sought out to explain this bizarre behavior. Indeed, a metallurgical examination of the girder “revealed that the backup steel plate used for the full penetration weld was itself not welded continuously,” (Kaminetzky 1993). Further testing of the material revealed a small crack in the backing bar approximately 5.59 mm deep. The fracture pattern expressed by the girder could also be linked with this defect. Therefore, it could be reasonably determined that insufficient welding a major factor involved in the failure.

Weld Failure
Welding is a complex procedure requiring a high level of skill that many designers often take for granted. The figure to the left shows the many different types of cracks that can occur as a result of insufficient welding (Figure 2). In the case of girder G-1, the initial defect was determined to be a fusion-line edge-type crack in the weld between the web plate of the box member and the backing bar (Kaminetzky 1993). The welding type was designed to be a full penetration weld; however, the backing bar used to create the full penetration weld was not itself continuously welded to the inside of the box, which likely introduced the flaw into the member.

**Investigation & Cause of Failure**
As is the case with most building failures, there was not one definite cause but a number of mistakes or oversights made throughout the design and delivery process that ultimately triggered the failure. The official report of this failure cited three main causes: "weld flaws, extremely cold temperatures, and poor metallurgy (steel with insufficient fracture toughness," (Feld & Carper 1996). For a thorough discussion of the welding flaws, see above. Unusually low, sub-zero temperatures also played a role in the failure. The theater, used exclusively during the summer, was designed to be open to the elements and unheated. Investigators found that this factor was not taken into account by designers of the girder as the weld material used was not capable of maintaining its required toughness in this environment (ENR Mar. 1985).

Fracture toughness is an important property to be considered when designing any steel structure. A higher fracture toughness means a higher resistance to brittle fracture. Unfortunately, higher strength steel tends to have a lower fracture toughness. Therefore, designers must be careful with balancing these two properties. Simply designing a steel structural element using the highest strength steel available is never a good idea, especially when designing in cold environments. Unfortunately, in the case of the Wolftrap roof girder, there was no minimum requirement set for the fracture toughness of the material (Maranian 2010). This slight omission in the design process, in tandem with the weld flaws and cold temperatures, ultimately caused a failure.

Corrective Actions & Cost
The crack was first discovered during an inspection after a piece of copper flashing had fallen off the end of the girder (ENR Feb., 1985). Immediately following the discovery of the crack, scaffolding was placed around the structure to hold up the roof until a proper fix could be installed. According to the original designers of the building, the roof was in no danger of collapsing (ENR Feb., 1985). This is because the structure had a sufficient level of redundancy, allowing for the load to take alternate paths to the foundation (Maranian 2010).

The first fix introduced to the structure appointed the removal of all the plates which had been compromised by the crack. Next, new plates were added with more careful welding of the back up bars. Finally, to provide ultimate peace of mind, the entire girder was post-tensioned with cables (Feld & Carper 1996). The image to the left first shows the original design of the member followed by the fixes made after the failure (Figure 3). It is important to note the shop note provided, which states that the back-up bars must also be welded and continuous along the entire length of the beam.

Investigators also recommended that corrective work be applied to the remaining roof structure (ENR Mar. 1985). The roof structure consisted of two large inner girders, each carrying a third of the load, and two outer girders, each carrying a sixth of the load (ENR Feb. 1985). According to the firm coordinating the repairs, "There are many possibilities for similar flaws on the outside of the roof girders," (ENR Mar. 1985).

The final cost of repairs came to a staggering $1.5 million (Los Angeles Times 1985). The companies involved in the original construction include: Dewberry & Davis, the architect and engineers; Bethlehem Steel of Bethlehem, PA, the steel supplier; Globe Iron Construction of Norfolk, the beam fabricator; and G&C Construction of Vienna, the chief contractor. All four firms agreed to hire one forensic engineer, John W. Fisher, to determine the cause of the failure. Although Fisher did not explicitly place blame in his report, he did often gravitate toward the poor workmanship of the beam (Hockstader 1985).

Lessons
After deeply studying this costly failure, there are a number of important lessons to be learned to avoid future calamities of this nature. These are summarized below:


 * Include a requirement for minimum fracture toughness of all structural steel members, especially those constructed with large metal plates. Remember: "the tougher the steel, the more resistant it is to brittle fracture," (Kaminetzky 1993)
 * Avoid designing structural steel members that abruptly change from thick plates to thin pates (Kaminetzky 19993)
 * Reduce the potential for brittle failure in steel members by (Kaminetzky 1991):
 * Carefully specifying welds and fabrication methods that have a lesser chance of introducing small, undetectable fissures into the member
 * Maintain adequate temperatures for the welding below the ductile to brittle transition for the welded parts of the member
 * Avoid designing members with too much restraint via heavy-section thickness or connections
 * Avoid high rates of loading (impact)
 * Design the structure with adequate redundancy so that the failure of one member does not lead to a collapse
 * Inspect all structural members carefully during construction, especially welded areas and other critical details

Conclusion
On January 24, 1985, a large crack was discovered in a girder supporting the roof structure of the newly constructed Filene Center at the Wolftrap Performing Arts Center in Vienna, Virginia. Engineers quickly designed buttresses to support the sagging structure, which had deflected nearly 9.5 inches; however, they assured the owner that the building was in no real danger of collapse due to the high level of redundancy in the structure allowing for alternate load paths. The proceeding investigation, carried out by John W. Fisher, revealed that there were three primary causes for the failure: "weld flaws, extremely cold temperatures, and poor metallurgy (steel with insufficient fracture toughness," (Feld & Carper 1996). The resulting damages amounted to $1.5 million for fixes to the cracked girder.