The 2016 edition of ASCE 7 “Minimum Design Loads and Associated Criteria for Buildings and Other Structures” will, for the first time, provide some guidance on the use of performance-based methods to design fire protection. Meanwhile, the ASCE/SEI Fire Protection Committee is in the process of developing its first guideline, “Structural Fire Engineering,” to present best practices for structural engineers working with fire protection engineers. Performance-based fire engineering (PBFE) relies on advanced analysis tools to predict how the structure will behave during a fire. Anticipating the release of these new standards, Thornton Tomasetti has developed the analysis capabilities to offer PBFE as a new service for buildings and bridges.
A Flexible Approach to Fire Safety
Performance-based fire engineering is an alternative to the traditional prescriptive method of design by qualification testing (DQT), which is used to fireproof almost all buildings in the United States. DQT relies on testing structural components in furnaces (Figure 1), where a standard temperature is imposed over time (Figure 2) to determine how long the component can survive in a fire environment. The survival time in the furnace becomes the fire rating of the component, expressed in hours. Fire protection systems, such as spray-on fireproofing or intumescent paint, are also assigned a fire rating by performing the same tests on protected components. The fire engineer for a given structure can then select for each member a protection system with the fire rating required for the project. The typical fire rating of residential and commercial buildings varies from 1 to 3 hours.
Figure 1 – Furnace for standard fire test (source: Furnace Construction Co. LTD.)
Figure 2 – Standard fire temperature curves (source: NIST TN 1681)
Design by qualification testing has been successful at preventing or delaying structural collapse during fires, but the method has some limitations that PBFE seeks to overcome. First, DQT does not allow for designing at a level of performance higher than life safety. It may be of interest to make a structure repairable or even immediately usable after a fire, but increasing the fire rating in DQT only guarantees that structural failure does not occur for a longer time. Another limitation of DQT is that it focuses on individual components and seldom considers the structure as a system, where load redistribution could help the structure weather the weakening or loss of some structural members. In addition, the fire performance of the structure is not tied to any real fire scenarios in terms of nature, intensity, distribution and propagation of the fire, nor to other factors, such as air flow, that can affect the outcome. Finally, the fire rating achieved through DQT is typically uniform across the structure, even though some components are more critical or located closer to potential fire sources. A more focused application of fire protection could lead to more resilient structures for a lower cost.
Instead of systematically protecting each structural member to achieve a prescribed fire rating, the engineer using PBFE is granted considerable freedom in the implementation of fire protection. For example, a major column in a parking garage may be covered with thicker fireproofing, while the exposed structure of an atrium containing little combustible material can be left unprotected for architectural reasons. In exchange for this design flexibility, PBFE requires the engineer to demonstrate through rigorous analysis that the structure will perform as required in fire conditions, and a peer review may be required. The forthcoming edition of ASCE7 and other standards for PBFE focus on the procedure to demonstrate fire performance.
The first step of PBFE is to establish a list of fire scenarios, each of which specifies the location and intensity of a fire. The fire locations are determined through a combination of most-likely and worst-case assumptions, so that each fire scenario has either a high probability of occurrence, some serious consequences or both. The intensity of a fire is typically measured as a heat released rate (HRR) in megawatts, which depends on the amount of fuel and oxygen available at the fire location. Building fires are fueled by the contents of one or several rooms, with an oxygen input dependent on doors, windows and ventilation. For bridge fires, the fuel typically consists of vehicles with their gas and cargo, and the oxygen input is assumed unlimited. The HRR may be determined analytically, but standard values are also available. For example, Eurocodes provide the HRR curve for building fires shown in Figure 3, with parameters summarized in Figure 4. The number of scenarios depends on the criticality and complexity of the structure, as well as the computational resources available and the anticipated number of design iterations.
Figure 3 – Eurocode HRR curve for building fire
Figure 4 – Parameters of Eurocode HRR curve for building fire
Each fire scenario must be associated with a performance objective. A typical objective is to keep the structure standing long enough for the occupants to evacuate and the fire to be put out, but a key benefit of PBFE is the flexibility to specify quantitative objectives beyond the minimum life-safety requirement. One can, for example, set a limit on the number of structural members damaged during a fire to ensure that a post-fire repair is possible and cost-efficient. The performance objectives can also be varied with the likelihood of the scenarios so that each scenario represents consistent risk for the structure owner or insurer.
Once the fire scenarios and performance objectives have been formulated, the engineer develops a fire protection design and implements advanced analysis methods to evaluate the design against them. In this evaluation, each fire scenario is simulated separately and in two steps: a thermal analysis followed by a structural analysis.
The objective of the thermal analysis is to determine the temperature in the structure during and after the fire scenario. The analysis is performed on a multi-physics model of the building to include different forms of heat transfer: convection and radiation in the air, and conduction in the structure and nonstructural components. The two-layer model (Figure 5) is specifically designed to analyze fire-driven air flows in buildings. A more realistic, flexible but expensive alternative is to use full-fledged computational fluid dynamics (CFD) (Figure 6). In both cases, the heat eventually flows into a solid model of the structure where a conduction analysis takes over. The thermal analysis is the most critical step of PBFE, and will be optimized to obtain sufficiently accurate estimates of the temperatures in a time deemed acceptable for design iterations.
Figure 5 – Temperatures obtained through two-layer analysis
Figure 6 – Temperatures (top) and smoke transport (bottom) obtained through CFD analysis
The second step of a fire simulation is to perform a structural analysis with the temperatures obtained from the thermal analysis. The increase in temperature has two effects on the structure: it lowers the strength and stiffness of the materials, and it develops thermal stresses in the members that cannot freely expand. Both effects can be modeled in traditional structural analysis software (Figure 7), where the thermal loads are typically combined with the dead load and a fraction of the live load. A nonlinear analysis must be used to predict the large deformations that occur during fires, taking into account all of the possible failure modes of the critical components. The analysis must also be able to detect global instabilities. Since the structural analysis uses the results from the thermal analysis, the challenge is to model each step with appropriate amount of detail and to facilitate the flow of information between the thermal and structural domains. From the analysis results, one can determine whether the structure satisfies the performance objective associated with the scenario analyzed.
Figure 7 – Structural analysis at high temperature
The thermal and structural analyses are repeated for each fire scenario, and the fire protection design is considered acceptable if the performance objectives are always met. Otherwise the design must be revised and analyzed again, in which case the analysis results provide useful guidance for the next design iteration.
Performance-based fire engineering benefits our built environment in several ways. By focusing on critical fire locations and structural components, it leads to an optimized fire protection that can be both more effective and economical than a prescriptive design. In addition, by considering the inherent fire resilience of each particular structure, it can also allow for shapes and finishes that the traditional approach would render impractical and therefore open the door for new architectural expressions.
– By Ali Ashrafi, vice president, Pierre Ghisbain and Jenny Sideri, senior engineers, New York-Wall Street, and Reza Imani, senior engineer, San Francisco
Follow these topics: