Numerical analysis of full-scale structural fire tests on composite floor systems

Abstract

Recent experiments conducted at the NIST examined steel-concrete composite floor systems designed per U.S. practice under standard fire. The first experiment designed per prescriptive provisions for a 2-hour fire rating with 60 mm²/m of reinforcement, developed a central breach integrity failure after approximately one hour of exposure. The second and third experiments, designed with 230 mm²/m of reinforcement, with the third one omitting the fire protection on the central steel beam, showed no failure within 2 hours. This paper describes a numerical investigation to gain further insights into the fire behavior of the composite systems tested in these experiments. Nonlinear finite element models were validated against the tests. Simulation of the first test shows concrete damage and rebar fracture in the hogging moment area corresponding with the cracks observed experimentally. Simulation of the third test captures the development of tensile membrane action, confirming the redistribution from the unprotected secondary steel member to the floor reinforcing steel. A sensitivity analysis allows identifying the minimum reinforcement steel for protected and unprotected central beams configurations. The results can support improvements of the fire requirements in the U.S. codes as well as application of performance-based structural fire design for composite structures.