DOES THE STANDARD FIRE TEST ALWAYS PRODUCE A SAFE BUILDING DESIGN? — A NUMERICAL STUDY ON THE BEHAVIOR OF STEEL BEAMS EXPOSED TO LOCALIZED FIRES
Chao Zhang and John L. Gross
Current building design practice in the US is intended to insure that walls, columns, floors, and other building members do not fail under fire exposure conditions. The fire resistance rating for a building component is determined by a fire endurance test conducted on an isolated member subjected to the standard fire exposure such as ASTM E-119 or ISO 834. Although the standard fire bears little resemblance to a real fire, it is used because it is considered to represent a worst-case fire such that, if a component can survive the standard fire exposure, it can survive a real fire. This assumption is valid when the failure of a building component is caused by the loss of material strength and stiffness at elevated temperature or when the load capacity of a component is controlled by the maximum temperature achieved in uniform fire conditions. However, when the failure of a component is caused by the thermal gradient, design based on the standard fire may be unsafe.
This reported work is about a numerical study on the behavior of steel beams exposed to localized fires. Coupled thermal-mechanical simulations were conducted to obtain the steel temperature distribution, the deformation mode and failure temperature of restrained beams, and the lateral torsional buckling resistance of simply supported beams. The heat fluxes from the localized fire to the exposed surfaces of the beams were calculated according to correlations derived from localized fire tests. The standard fire and uniform heating condition were also considered for comparison. The study found that,
· The temperature distributions within the steel beams subjected to localized fires are highly non-uniform both across and along the beams. Along the beam length, temperatures near the fire source can be several hundred degrees (Celsius) higher than those far from the fire source.
· The deformation mode for a restrained steel beam subjected to localized fires may be significantly different from that of a beam subjected to the standard ASTM or ISO fire. The failure temperatures for restrained steel beams subjected to localized fires may, in some cases, be lower than those for restrained beams subjected to the standard fire.
· The critical temperature, defined as the temperature when the beam fails in lateral torsional buckling, in a localized real fire may be hundreds of degrees lower than that in the standard fire.
· Reliance on the standard fire test may lead to an unsafe design if there is a potential for a real fire to be localized.