Dallas+Cowboy+Indoor+Practice+Facility+Collapse

=Dallas Cowboys Indoor Practice Facility Collapse (May 2, 2009) = //Michael Stremski, BAE/MAE, Penn State, 2012 //

//Keywords: Tensile membrane structures, fabric covered structures, Dallas Cowboys, Summit Structures, collapse, buckling, internal pressure, fabric bracing //

toc

Introduction
On the afternoon of May 2, 2009, the indoor practice facility used by the Dallas Cowboys of the National Football League collapsed during a severe thunderstorm. Designed and constructed in 2003 for approximately $4 million, the facility consisted of a fabric-covered tubular steel frame structure which was later upgraded in 2008 with additional framing and a new roof covering. (Battista, 2009) (Gross, 2010, v) Immediately before collapse, wind speeds were estimated to be in the range of 55 mph to 65 mph, well below the design speed of 90 mph at the time of design. (Gross, 2010, 5) Since the collapse, e-mails and other documents have surfaced indicating that the design company, Summit Structures LLC, knew of potential problems and dangers associated with the structure, but concealed the issue. Twelve people were injured as a result of the collapse, two seriously. (AP, 2012)

Structure Description
The Dallas Cowboys indoor practice facility at Valley Ranch was a tensioned fabric-covered steel frame structure with a gable roof as seen above in Figure 1. The structure was 204 ft x 406 ft in plan and measured 86 ft high at the ridge of the gable roof, corresponding to a mean roof height of 66.4 ft. Due to the exterior fabric as well as fabric liner on the interior, the structure was classified as a membrane-covered frame structure per International Building Code (IBC) 2000. Twenty-eight steel truss gable frames spaced at 15 ft on center were used to frame the practice facility with the aforementioned fabric covering and fabric liner on the exterior and interior, respectively. The inner and outer truss chords were formed with 5" diameter cold-formed steel tubes with truss webs consisting of hot rolled single or double angles. (Gross, 2010, 3) Frame nomenclature is illustrated further in Figure 2.



The structure's lateral resistance was provided through cross and sway bracing in addition to purlins. As seen in Figure 3, this steel cable cross bracing was attached to the inner truss chords; however, no cross bracing was provided between the keystone sections of each frame. (Gross, 2009, 4) As mentioned earlier, upgrades were made to the facility's structure in 2008 through addition of purlins, a new roof covering, and extra reinforcements to various members. (Gross, 2010, 4)

Collapse
On the afternoon of May 2, 2009, a heavy line of thunderstorms hit the Irving, Texas area. When the severe weather hit the area, approximately 70 players, coaches, staff members, and media members occupied the practice facility during a rookie mini-camp practice. According to video evidence as well as eyewitness accounts, lights suspended from the roof began to aggressively swing back and forth, and only a few seconds later, the structure collapsed. (McLaughlin, 2009) Video of the collapse can be seen below and a view of the collapse structure after the fabric covering had been removed can be seen in Figure 4. Twelve people were injured as a result of the collapse, including a scouting assistant who suffered a severed spinal cord and is now paralyzed from the waist down and a coach who suffered a broken neck. (McCann, 2009) (McLaughlin, 2009)

media type="youtube" key="U3fs_pjA9A4" width="425" height="350"


 * //Warning: This video contains graphic language.//**

//Wind Environment//
In January of 2010, approximately 8 months after the collapse, the National Institute of Standards and Technology (NIST) released a final report on the collapse following its investigation of the incident. One major conclusion drawn as a result of the investigation regarded wind speeds at the time of collapse. For the Irving, Texas area, the American Society of Civil Engineers (ASCE) Minimum Design Loads for Buildings and Other Structures Standards 7-98 and 7-05 require design wind speeds of 90 mph. However, analysis of the weather data performed by the National Oceanic and Atmospheric Administration (NOAA) National Severe Storms Laboratory concluded that wind speeds at time of collapse were in the range of 55 mph to 65 mph, well below the design speed of 90 mph. Based on analysis of National Weather Service data, it was also concluded that the winds at time of collapse came primarily from the west. (Gross, 2010, 26-27)

//Collapse Analysis//
Based on a survey of the collapsed structure, NIST determined the most likely series of events which led to the collapse. As mentioned previously, the predominant wind direction was west-to-east, and as expected, the practice facility collapsed toward the east. However, the collapse pattern for the north and south ends of the practice facility differed, and both are discussed below. (Gross, 2010, 31)

In regards to the northern portion of the structure, illustrated in Figure 5, it is believed that a kink formed in the frame on the east side due to buckling of the inner chord along the straight section of roof. The frame then began to sway eastward after both of the frame's knees failed. After this, it is likely that the east keystone web failed in compression, triggering fracture of the outer and inner keystone chords at the ridge due to tension. Finally, through load redistribution and loss of lateral stability, these individual frame failures spread throughout the structure, thus leading to complete collapse. (Gross, 2010, xxiii, 33)



The collapse sequence of the southern portion of the structure, on the other hand, varied from that of the northern portion. As seen in Figure 6, failures of the keystone and west knee triggered rotation about the west support, leading to the west half collapsing onto the field. These failures, combined with a failure at the east knee, led to additional rotation about the east support. Eventually, the east wall collapsed and this portion of the roof flipped upside down, as also seen above in Figure 4. (Gross, 2010, 32)

//Wind Design Analysis//
In the 2003 design, Summit Structures, LLC used IBC 2000 as the building code and the 1998 edition of the American Society of Civil Engineers (ASCE) 7 Standard for Minimum Design Loads for Buildings and Other Structures (ASCE 7-98) to determine design loads. For the 2008 facility upgrade, the ASCE 7-05 Standard was used in determining design loads. One main discrepancy between the design wind load calculations and those reproduced by NIST during the investigation was the enclosure classification and how it relates to internal pressures. Documents from the 2008 upgrade to the structure state the following: "This structure should be maintained closed at all unused times to prevent the possibility of an internal pressure build-up. As such, the structure is considered as enclosed for the purpose of internal pressure calculations." This rationale corresponds to an internal pressure coefficient of +/- 0.18 per Chapter 6 of ASCE 7-05. This enclosure condition, however, cannot be assured during an extreme loading event such as that seen on May 2, 2009. As a result of displacements and a variety of other factors, closed doors can open and the fabric covering can tear, both of which cause higher internal pressures on the structure. Not to mention, the practice facility had vents on the north and south ends as well as the roof, which also qualifies the structure as "partially enclosed." Had the structure been more accurately classified as "partially enclosed" per Chapter 6 of ASCE 7-05, an internal pressure coefficient of +/- 0.55 would have been utilized, effectively tripling the design wind loads due to internal pressure as seen in Figure 7. (Gross, 2010, 39-42)

In addition, there were other portions of the provided wind load calculations which did not prove accurate. As in the 2008 upgrade documents, the structure was also classified as "enclosed" in the original 2003 design. However, the design calculations provided by Summit Structures LLC did not include any internal pressures on the structure. Also, in determination of external wind pressures, the design company used Figure 6-4 of ASCE 7-98, which is only valid for structures with mean roof heights less than or equal to 60 ft. (Gross, 2010, 45-48)

**//Exterior Fabric Covering//**
Throughout the investigation and aftermath of the collapse, discussions have arose regarding the stiffness of fabric coverings and whether or not they can provide bracing for chord members. Section 4.3.1.1 of the ASCE/SEI 55-10 Standard for Tensile Membrane Structures states,

//"Where the support structure relies on the tension membrane structure to provide stability to components or individual members of the support structure, the designer of the support structure must insure that a local failure of the tension membrane structure does not cause collapse of the tension membrane structure and support structure by either; (a) loss of capacity of the individual member of the support structure or (b) excess movement of the tension membrane structure."//

The design documents provided by Summit Structures LLC for the 2003 design state that "the outer chords of the truss frames receive additional lateral support from the fabric roof sheets on them." In addition, the "exterior fabric is an integral part of the structural system, removal or alteration without prior authorization is prohibited." However, the actual amount of lateral support that the fabric assumed to provide was never clearly indicated. Other factors must also be taken into consideration such as potential tearing due to debris or deterioration. (Gross, 2010, 51)

An interesting discussion, however, is the collapse of the northern portion of the structure. As indicated previously, the triggering mechanism was found to be buckling of the inner chord, which would not have received any lateral bracing from the outer fabric covering anyway. Dr. Louis F. Geschwindner, Co-Chairman of the ASCE Tensile Membrane Structures Standard Committee and Professor Emeritus at The Pennsylvania State University, discussed the collapse in a personally conducted interview. Although questions arose about the structural integrity of tensile membrane structures after the collapse, Dr. Geschwindner maintained that the structure type is not the problem. "There is nothing wrong with tensile membrane structures," he stated. "There was just something wrong with this one." The exterior fabric, he says, can be designed and constructed to provide a certain amount of lateral bracing to the frame. (Geschwindner, 2012) As long as proper connections are made during construction based on assumptions made in the design, the structure should not experience any inaccuracies related to lateral bracing from the exterior fabric. Nonetheless, Dr. Geschwindner believes the biggest issue in the collapse of the practice facility lies in the aforementioned wind load calculations and design rather than the bracing provided by the exterior fabric. (Geschwindner, 2012)

//Findings//
In its investigation of the collapse, NIST listed what it believed to be the four main contributors to the practice facility's collapse:

1.) Wind Loads - Perhaps the most important contributor was the discrepancy between wind loads used in the original and upgrade designs compared to those calculated using the provisions of the ASCE 7-98 and ASCE 7-05. As mentioned above, this included internal and external wind pressure discrepancies as well as misuse of pressure coefficients pertaining to frame zones in certain areas of the facility.

2.) Frame Member Capacities - The second main reason for collapse is attributed to considerable differences in frame member capacities when comparing the 2003 design to those calculated by NIST in its investigation. This was due to not only the exterior fabric lateral bracing assumption, but also the use of 0.5 as the effective length factor for inner and outer chord members. In the event that the exterior fabric covering actually did provide full bracing to the outer chord, an effective length factor of 1.0 is much more appropriate to use than the ideal fixed-fixed effective length factor of 0.5.

3.) Frame Knee Details - The third item that NIST listed in its report dealt with details of joints at frames' knees. These particular details led to large flexural and shear forces in certain members of the frame which were not accounted for in the design of the structure.

4.) Inadequate Upgrades - The final contributor to the collapse was inadequacy of reinforcements in the 2008 upgrade. These particular reinforcements only enhanced the compressive strength of certain members. Even so, the most critical members were not upgraded at all. (Gross, 2010, 83-84)

Ultimately, the structure was not designed properly from the beginning and the findings above presented by the National Institute of Standards and Technology have been accepted to accurately describe the failure. John Gross, leader of the NIST study for the collapse of the practice facility, summed up the collapse by saying, "Our investigation found that the facility collapsed under a wind load that a building of this type would be expected to withstand." (Engineers Edge, 2009)

Recommendations & Prevention
In addition to the causes of failure and collapse analysis presented in NIST's report, a series of recommendations was also provided for increasing the safety of similar-type structures. Emphasized in the report is the fact that NIST deems these particular recommendations feasible in regards to execution as well as time. As stated in the final report, the recommendations read as follows:

"NIST recommends that fabric-covered steel frame structures be evaluated to ensure the adequate performance of the structural framing system under design wind loads. Of particular concern are (1) the use of fabric covering to provide lateral bracing for structural frames, (2) determination of the appropriate enclosure classification in the calculation of internal pressures for design wind loads, and (3) the ability of the structural system, including the lateral bracing, to maintain overall structural integrity." (Gross, 2010, 88)

These recommendations essentially remind the designer of particular issues that should be focused on when designing these types of structures. In essence, if the structure is designed and constructed properly and thorough engineering is performed throughout, safety can be expected. Had the calculations been performed correctly and the structure designed appropriately, it is very likely that the Dallas Cowboys indoor practice facility would still be standing today.

//"Real Sports"//
On January 19, 2010, HBO's //Real Sports with Bryant Gumbel// aired an examination of the Dallas Cowboys indoor practice facility collapse. HBO's account of the collapse, entitled "Blue Star Fallen," was nominated for a Sports Emmy for Outstanding Sports Journalism and conducted interviews with various key players in the design and collapse of the structure. Among those interviewed were Rich Behm, a scouting assistant who was paralyzed from the waist down due to injuries suffered in the collapse, as well as Joe DeCamillis, a special teams coach who suffered a broken neck. (Real Sports, 2010) The examination also consulted Dr. Geschwindner for background on structural behavior as well as code interpretation. (Geschwindner, 2012) Dr. Geschwindner explains that, when non-conservative assumptions are made in the design and analysis of a frame, they can build up to the point where the entire structure is non-conservative, making the facility prone to failure. (Real Sports, 2010)

//Real Sports// also discussed the Cowboys' decision to have the facility upgraded in 2008. After learning of the collapse of a similar structure designed by Summit Structures' Canadian parent company Cover-All Building Systems, the Cowboys notified Summit Structures that they were aware the indoor practice facility was a safety hazard. Summit Structures subcontracted the upgrade work to another engineering firm, and it was subsequently discovered that the employee who designed much of the upgrades was not a professional engineer and had not obtained a college engineering degree. (Real Sports, 2010) It was also noted that at least five other facilities designed by Summit Structures had collapsed since 2002, a number that was later claimed to be at least 14 after further examination of court records and reports. (Real Sports, 2010) (AP, 2012)

**//Lawsuits//**
Over one year after the collapse, the two Dallas Cowboys employees who were seriously injured reached a settlement with Summit Structures LLC and Cover-All Building Systems. Rich Behm settled with Summit Structures LLC for approximately $19.5 million, and Joe DeCamillis settled for approximately $4.5 million. Due to workers' compensation laws, Mr. Behm and Mr. DeCamillis were not eligible to sue the Dallas Cowboys for negligence. However, the land that housed the practice facility was owned by a third party, a company Cowboys owner Jerry Jones was involved with. (Real Sports, 2010) As a result, both men received an additional $5 million each in cash and benefits provided by companies held by Dallas Cowboys owner Jerry Jones. (Formby, 2010)

//Controversy//
In the summer of 2012, reports surfaced that Summit Structures LLC knew the practice facility may have been under-designed, but did not take action. E-mails, handwritten notes, and other documents which were not released to the public, suggested that the designer knew that portions of the structure were "too slender and long" and "prone to buckling." Documents also indicated that the Cowboys were concerned about the structure, and Summit Structures LLC planned to supply the Cowboys with calculations that would conceal the problem. (AP, 2012) Both companies filed for bankruptcy in March of 2010, not before recommending that all its structures be checked for structural stability and adequacy (AP, 2012) (Geschwindner, 2012).

Similar Structures
The collapse of the Dallas Cowboys indoor practice facility spurred questions about the structural integrity of similar facilities. Close attention was given to tensile membrane structures following the collapse due to the multitude of applications including industrial and agricultural facilities, casinos, storage facilities, military installations, aircraft hangars, and also other sports facilities. (Collapse of the Dallas Cowboys Indoor Practice Facility, 2009) One of these structures was the McFerrin Athletic Center on the campus of Texas A&M University. Completed in 2008, the McFerrin Athletic Center was also designed by Summit Structures LLC and consisted of an indoor practice facility as well as a running track. (Click here for more details.) The structure made headlines in November of 2009 after Haynes Whaley Associates, a structural engineering firm in Houston, released analysis expressing concerns about the structure's ability to withstand the design wind speed of 90 mph. In response to the analysis, Summit Structures LLC adequately strengthened the structure by adding cables before filing for bankruptcy the following year. (AP, 2009)

Conclusion
Great care and responsibility must be taken in the design of long-span structures, particularly fabric-covered steel frame structures. Assumptions made during the design, such as enclosed vs. partially enclosed for internal wind pressures, are vital and can make the difference between a structure standing or collapsing. Unfortunately, events like the Dallas Cowboys indoor practice facility collapse do take place, often leading to the injuries of innocent people. Through a combination of the recommendations provided by NIST and greater design care by practicing structural engineers, it is hopeful that these types of collapses can be limited or even eliminated for the greater good of the public. As Dr. Geschwindner noted in his interview with //Real Sports//, "Safety of the public is the engineer's ultimate responsibility."

//Fabric Roof Failures//
Hubert H. Humphrey Metrodome

//Stadium/Arena Failures//
Louisiana Superdome Hartford Civic Center University of Washington Football Stadium