Model Glue

Computational fluid dynamic modeling cements life safety and constructability, setting the table for another Smithsonian smash hit

By Mike Rzeznik, P.E., New York Principal, and Andrew Bowman, P.E., Chicago Principal, Gage Babcock Associates, Chantilly, Va. May 1, 2002

The most visited museum in the world-the Smithsonian Institution’s National Air and Space Museum in Washington, D.C.-has more than 10 million visitors cross its threshold annually to view some of aviation’s most prized treasures.

Such impressive numbers testify to the appeal of viewing and experiencing actual air and spacecraft up close, which has apparently captured the imaginations of everyone from small children to dedicated aviation buffs.

Considering the size and popularity of the museum’s exhibition, it is amazing that only 10% of the Smithsonian’s aviation collection has been able to be displayed at one time. The rest of the collection has been hidden from view in various hangers throughout the United States. In an attempt to make these aircraft more accessible-and in order to better care for and maintain these treasures-the museum decided that a single, central location for restoration and display was in order. This new air and space museum, known as the Steven F. Udvar-Hazy Center, is under construction at Dulles International Airport.

Tale of the tape

Constituted of two large hangar-like structures-interconnected and located at right angles to each other, similar to a lowercase ‘t’-the museum will consist of approximately 760,000 sq. ft. of space. Nearly 400,000 sq. ft. will be used for the display of aviation-related artifacts, with the remainder of the complex housing a mix of restoration, education, public amenities, storage and administrative support functions.

The larger display area is referred to as the Aviation Hangar, a semi-cylindrical structure approximately 250 ft. wide by 1,000 ft. long, with a floor-to-ceiling height of approximately 100 ft. at the center of the space. Its counterpart, the 260-ft. long by 180-ft. wide Space Hangar, will predominantly house artifacts associated with man’s journey into the outer atmosphere.

The design concept represents the merging of two very dissimilar occupancies: hangars and places of assembly. Hangars, in general, are extremely large, but sparsely populated. The new museum, however, may accommodate thousands of people at a time. Unfortunately, the code requirements covering hangars do not account for large populations of people, and the requirements for places of assembly do not adequately address the immense magnitude of the new museum. This presented a major design challenge for many of the building’s systems-including fire protection, life safety and security, areas Smithsonian officials expressed notable sensitivity toward. In fact, the institution employs its own in-house fire protection and security branches to conduct ongoing maintenance and even review outside design.

Project architect HOK, St. Louis, enlisted Gage-Babcock & Associates, Inc. (GBA), Chantilly, Va., for the cause. In addition to designing all fire-suppression, detection, alarm and security systems, GBA was required specifically to review the overall design to ensure that it was compliant with applicable codes.

The first critical task was to mitigate and prevent costly design changes down the road. This involved determining what aspects of the building were going to be outside the ordinary design realm due to the magnitude of the proposed structure. That early analysis determined that the most notable challenge involved the cavernous volume of the exhibit hangars. While thorough engineering could certainly resolve some problems, such as acoustical issues, achieving code compliance could not initially be resolved through traditional approaches. Again, this is due in large part to building code requirements being more geared toward commonly constructed building types-an approach that generally satisfies most situations, but does so without providing flexibility for architecturally progressive and unique designs.

Furthermore, the Smithsonian uses NFPA 101 , Life Safety Code (LSC) and BOCA’s National Building Code to govern life-safety requirements. These requirements include relatively simple issues, such as limits on dead-end corridors and the number of required exits-as well as more complicated analyses such as horizontal exiting solutions, exit discharge and exit arrangement. In the case of the new museum, travel distances to exits-typically one of the most fundamental and mundane life-safety requirements-came to the forefront.

A bridge too far?

The expansive area of the Aviation Hangar meant that occupants located near the center of the area had a large amount of ground to cover before getting to an exit.

Further complicating this issue was the proposed two-level elevated walkways or ‘skyways.’ One of the most visitor-friendly features of the Aviation Hangar, the skyways effectively serve two purposes: First, they were arranged to give visitors unique perspectives and access to aircraft exhibits located both on the museum floor and suspended two levels from the hangar roof assembly. Second, their gentle slope efficiently serves visitors that have physical impairments.

As effectively as these elements serve the space-both through form and function-their extensive lengths created extreme maximum travel distances to the exits-650 ft. in the most extreme case. This far exceeded the maximum 200-ft. travel distance prescribed by the applicable building code. It was clear that in order to safely incorporate this critical design feature, a thoughtful and prudent assessment of the life-safety threat would be required. In fact, justifying the extensive travel distance was one of the more intriguing and appealing facets of the job.

Because of the sheer height of the space, one could intuitively sense that permitting travel distances in excess of those allowed by the prescriptive code might be reasonable. This, of course, would depend on many factors, including the anticipated fuel load for a reasonable worst-case fire and the expected egress time from the most remote point.

At the time that this issue was being tackled, the National Institute of Standards and Technology (NIST), Gaithersburg, Md., broke through with a technology that proved to be most helpful to the situation. Specifically, NIST had begun applying their newly developed computational fluid dynamics (CFD) model, Large Eddy Simulation (LES) to non-research situations. This model-subsequently renamed Fire Dynamics Simulator (FDS)-deals with smoke movement in large volume spaces and represented an accumulation of years of research that, when coupled with advances in computer processing speed, led to the development of a rare, PC-compatible CFD model. Because the model had been verified against applications in aircraft hangars, the Udvar-Hazy Center was a perfect opportunity for its introduction into real-world building design. With the encouragement of the Smithsonian’s fire-protection engineering department, GBA proceeded to partner with NIST on the first implementation of FDS in performance-based design.

The design fire

As with any computer model, a critical aspect of obtaining valid results is providing the proper input, not the least of which is the design fire. Selection should be contingent on what fuel load, in which location, and under what set of reasonable circumstances, will result in the worst credible result in terms of impact on life safety.

Further, the list of variables can quickly become extensive, so selection can be difficult. For instance, in a typical atrium, a wide range of fuel loading is conceivable. Fortunately, the combustible loading at the museum was a known entity. The Smithsonian had already identified which aircraft were to be displayed and had also developed a specific layout for each aircraft.

Further, because the Smithsonian had no plans for vendor kiosks or combustible loading of any kind on the museum floor-other than the aircraft displays-ascertaining the combustible fuel loading that would make up the worst-case credible design fire was far simpler than more undefined spaces.

For the most part, the aircraft on display are of non-combustible construction, with the exception of minor interior and operational components. Some, however, are constructed of combustible wood and synthetic materials. Again, the museum had a wealth of knowledge on the display location, size and materials of these aircraft. Based on this information, the largest contiguous combustible loading was determined to be represented by a World War II British fighter plane, the Mosquito.

Using data provided by the museum, the dimensions, type of wood and orientation were used to develop a reasonable fuel load for the model. Also, data from national testing laboratories about various wood fuel sources with similar arrangements were evaluated so that a reasonably similar fuel load mimicking a fire involving the Mosquito was developed.

Of course, additional safety factors with regard to the fire were considered. Since the volume of smoke increases exponentially with increased ceiling height, the location of the aircraft itself was assumed to be in the center of the main hangar floor. This assumption resulted in the largest possible volume of smoke generated by the fire. Further, although the building is fully sprinklered, modeling assumed that the system was not functioning. Therefore, once the fire started, it would be allowed to burn out. This was substantiated by calculations indicating that the likelihood of radiant ignition of nearby fuel sources was negligible.

Ultimately, the results of the modeling had to be compared with the results generated from egress models for the space. The rate of occupant evacuation relates to a number of factors, including the density of people using the egress path, the width of the egress path, type of egress component (stair, door, ramp, corridor, etc.), as well as the mobility level of the occupants. Additionally, delays relating to detection, alarm and human factors had to be considered. All of these factors have to be accounted for to obtain a reasonable result with regard to expected egress time.

The majority of these variables for egress path were relatively easy to determine, since the most remote point was clearly the elevated walkway with distance to an exit of 650 ft., and the choice of exit routes was limited. Once the geometry and slope of the walkway were accounted for, the population density and mobility level of the occupants had to be considered.

It was critical to determine the population density not only for the number of occupants that the museum would routinely expect, but also for the greatest occupant load that the museum anticipated in the foreseeable future-10 years or so. Fortunately, a programming consultant had been retained by HOK, and produced this key information as part of their services.

Armed with this information, the largest one-day anticipated occupant load was developed. As a safety factor, it was assumed that the entire anticipated occupant load would be in the exhibition hangars only during a fire emergency.

Rate of travel is a critical component of modeling egress from a space. To be conservative, a reasonable factor of safety was incorporated. Therefore, as part of the egress analysis, mobility was based on an evaluation of safe exit time using criteria from the Americans with Disabilities Act. These guidelines are much more restrictive than those required for able-bodied egress. In fact, the guidelines require the assumption that impaired occupants will rest for two-minute periods for every 100 ft. traveled. Using this approach, exit times from the most remote point approached 30 minutes.

Programmatic limits

As with most performance-based analyses, positive results must be tempered by the limitations of the analysis. Primarily, the assumptions used in the analysis must be maintained for the life of the building. Therefore, factors such as anticipated fuel loading, configuration and location must be maintained at levels at or below those assumed in the analysis. Any future reconfiguration of the means of egress routes, increased population sizes and other factors may necessitate re-evaluation of the analysis. Fortunately, the Smithsonian has an extremely competent fire-protection division that can monitor the impact of changes and determine if re-evaluation of the analysis is necessary. No one way

Once the results of the egress model and FDS were evaluated, it was clear that the worst reasonable case fire in the Aviation Hangar would not have a significant impact on the level of life safety in the space.

The cavernous volume of the hangar would literally allow it to act as a smoke and heat sink for the products of combustion. Additionally, the sloped walkways would lead exiting occupants in a downward direction away from the harmful upper layer as they move toward the closest exit. In fact, the results showed that even without sprinklers operating, the hangar was so large that the entire fuel load of the design fire could be fully consumed without having a detrimental effect on life safety.

In recent years, such analyses have become more commonplace. Especially intriguing is how clearly the results of this project call into question the validity of one of the prescriptive code’s most fundamental requirements. The sheer difference in the evaluated travel distance-650 ft.-and the maximum permitted travel distance-200 ft.-demonstrates that strict compliance with prescriptive code requirements is only one of many perfectly valid ways to safely design a building.