Skip to main content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Performance-based Design for Structures in Fire - Modeling and Validation

Summary

The National Fire Research Laboratory (NFRL) at NIST is a unique research facility that enables scientists and engineers to conduct research on the response of real­-scale structural systems to realistic fire and mechanical loading under controlled laboratory conditions.  This project will develop performance-based tools, methodology, and guidance for designing and evaluating structures for fire performance.  This will involve the development of:  (1) validated computational modeling approaches and tools; (2) temperature-dependent characterization of material and connector behavior; (3) quantification of uncertainty, including propagation of uncertainties from fire simulation to thermal analysis to structural analysis; and (4) design methodologies including a performance-based design framework and improved prescriptive design requirements.

Description

Modeling and Validation

Objective:
 To produce validated computational tools and technical guidance enabling the development of performance­-based standards for the cost-effective fire resistance design and assessment of structures.

What is the new technical idea?
The focus of this project is to develop a comprehensive approach to structural fire safety that will enable the development of performance­ based design tools and guidelines for structural systems exposed to fire, and alternate retrofit, design, and fire protection strategies. Performance-based design (PBD) methodologies to evaluate the fire performance of buildings and other structures are needed to move beyond the prescriptive procedures presently in use, which cannot be used to determine actual structural performance in fire. At present, buildings and other structures are designed for primary natural hazards and protected against fire effects. The proposed approach will, for the first time, consider fire as a design condition in the building design process, leading to risk-consistent, cost-effective designs. The proposed approach will make use of experimental data on the performance of structural members and systems, including connections, subject to realistic fires rather than controlled furnace conditions. Experimental data will be developed in the National Fire Research Laboratory (NFRL).  The data, developed within the project “Measurement of Structural Performance in Fire”, will be used to develop and validate computational models that are highly nonlinear due to the concurrent effects of temperature-dependent reduction of material strength and stiffness and thermally-induced load effects. Additionally, this project will take a risk and reliability­-based approach to the prediction and specification of the fire hazard, structural fire effects, and calculation of structural response.

What is the research plan?

This project aims at developing a performance-based methodology to evaluate the fire performance of building structures by incorporating knowledge concerning structurally significant fires and the material and structural system response to elevated temperatures.  In particular, the project will develop (1) validated computational modeling approaches and tools; (2) temperature-dependent characterization of material and connector behavior; (3) quantification of uncertainty, including propagation of uncertainties from fire simulation to thermal analysis to structural analysis; and (4) design methodologies including a performance-based design framework and improved prescriptive design requirements.  In addition, the project will support NFRL planning and testing activities for structural systems subject to fire, including pre­test design, planning, and predictions, and post­test validation of developed models.

The structure currently under consideration for testing in the NFRL is a two-story steel gravity frame two bays by three bays in plan, and employing a composite floor system.  The 6.1 m (20 ft) by 9.1 m (30 ft) test bay will be loaded hydraulically to simulate the gravity service load condition (see Project “Measurement of Structural Performance in fire”).  Composite floor systems were selected for this study because of their widespread use in construction and because of the significant challenges in modeling the complex response of the system under fire loading, including the behavior of gravity connections, concrete slab, metal deck, and shear studs.  These systems, which utilize costly fire proofing materials, could be vulnerable to fire effects due to significant reduction in strength and stiffness under elevated temperatures.  Issues that will be investigated through the experimental and analytical studies include symmetry in floor framing, geometry of floor plates, influence of connections, restraint of thermal expansion, and fire exposure.  Pre- and post-test modeling will be conducted for the various configurations with the objective of developing tools for engineers to enable performance-based design of structures under fire conditions.  Prior to testing the two-story steel gravity frame, five isolated, long-span composite beams were tested under fire effects to gain insight into the behavior and failure modes of the beams, composite floor slab, and their connections.

The research plan has four tasks:

  • Validated computational modeling approaches and tools:  In collaboration with the project ““Measurement of Structural Performance in Fire”, this project will conduct pre­test prediction and post­test validation of response of steel structures and composite floor systems subjected to fire exposure. The project will also develop tools and methods to couple or transfer data between fire­ thermal-structural models.  Pre-test prediction will be used to (1) help with planning of experiments and instrumentation, (2) help ensure safe execution of tests, and (3) evaluate accuracy of current state-of-the art modeling approaches to structural response and failure under fire.  Post-test analyses will be used to provide engineers with validated system-level modeling tools and quantification of uncertainties.
  • Temperature-dependent characterization of material and connector behavior:  To accurately model the structural response of structural systems to fire, temperature-dependent material and connector data are essential, yet data are still lacking.  This project will develop high-temperature test protocols for coupons, bolts, welds, etc. under a variety of thermal loading conditions.  The project will identify high-priority experimental needs, based on sensitivity studies using system-level and global models.  Testing will be conducted to develop a database of material and connector behavior under elevated temperatures.  Test results will be used to identify and quantify key sources of uncertainty.
  • Quantification of uncertainty: Analyses of structures exposed to fire are subject to many sources of uncertainties.  This project will identify dominant sources of uncertainty in thermal loading, thermal and mechanical material behavior, and boundary conditions (degree of restraint).  The project will also study propagation of uncertainties from fire dynamics modeling through structural response analysis, and develop load and resistance factors for design.
  • Design Methodologies: The project will develop performance-based tools and guidelines for the fire resistance design and assessment of structures, including a framework linking fire intensity measures with expected structural performance objectives for a variety of risk categories.  For cases where a performance-based design approach is not used, this project will develop improved prescriptive design requirements based on the results of advanced computational modeling.
Created March 11, 2016, Updated October 16, 2019