Naeim Roshan, Hamzeh Hajiloo
To elevate glass fiber–reinforced polymer (GFRP)-RC fire safety confidence to levels comparable to steel-RC, this study validates numerical models and design codes by quantifying practical design options. The experimental program tested (1) varying GFRP anchorage lengths (100, 150, and 200 mm) in cool zones under a 50% load ratio (Slabs S2–S4); (2) a midspan lap splice under a 30% load ratio protected by a water-based intumescent coating (Slab S5); and (3) reinforcement cutoffs under a 30% load ratio (Slab S1). A distinctive feature is the use of realistic, monolithic concrete supports. Unlike previous studies that rely on external insulation to create cool zones, this approach simulates natural heat transfer, thereby eliminating the need for artificial boundary conditions. All slabs sustained service loads through 3.5 h of the standard fire exposure. Loads were then increased to 93% of design capacity for the anchorage slabs and 62% for the others. Slab S2 (100 mm anchorage) failed at 4 h 5 min due to GFRP pullout, while the other slabs had not failed. The test was stopped at that time for safety reasons, as Slab S2's failure created a gap allowing flames to escape the furnace. Although the intumescent coating delayed reinforcement heating for 60–90 min, thermal analysis indicates the GFRP-RC slab with splices (Slab S5) would have survived without protection under the applied load. These findings confirm that specific detailing enables GFRP-RC structures to achieve fire-safety reliability equivalent to that of steel-RC systems.