Song, Han, and Uy 2010
Experimental Study, Results and Discussion
Three concrete filled steel tube column to steel beam joints with reinforced concrete slabs are tested under combined loading and fire during the heating and cooling phases. The main parameters of the test were heating time (which ranged from 30-48 minutes), and the thickness of the fire protection coating. The joints were designed with strong-column weak-beam behavior. Eight gas burners and one exhaust were used to heat the RC slab during the heating phase, which followed ISO-834 standard fire, and the furnace temperature was reduced throughout the cooling phase. A hydraulic jack was used to apply loading onto the column, and two mechanical jacks applied load onto the beams. Displacement transducer were located at three different locations and were used to measure axial deformation of the column and vertical deformation of the two beam segments. The loading was applied during phase 1 (the ambient phase), and the loads were maintained during phases 2 and 3 (heating and cooling phases). During phase 4 (the post-fire phase), the joints return to room temperature, and the load on the columns is kept constant, whereas the loads on the beams increase simultaneously until the joint fails. The first specimen was heated during the heating phase until failure to determine the overall fire resistance time (48 minutes), whereas the fire resistance time of the last two specimens were reduced to 30 minutes. The peak temperature in the CFST column is significantly higher than that of the two points in the beam. During the cooling phase, the temperatures in the steel tube reduce much faster than the concrete core, which begins cooling a significant amount of time into the cooling phase. The temperatures in the joint zone are lower than in the non-joint zone both at the heating time and 60 minutes after heating, before becoming uniform after a significant amount of time. The temperatures of the steel beam are much higher than that of the CFST column. All three joints failed of local buckling due to the steel beam compressive flange. The specimens with the 30 minute fire exposure time only failed during the post-fire phase, due to the increased vertical loading.
A finite element analysis model was created to simulate the experimental test, which analyzed the temperature distribution of the columns throughout both the heating and cooling phases. The finite element analysis ultimately verified the experimental results. This analysis concluded that the temperature in the upper flange of the beam is lower than that of the lower flange within the joint zone.
Song, T.-Y., Han, L.-H., and Uy, B. (2010). “Performance of CFST column to steel beam joints subjected to simulated fire including the cooling phase.” Journal of Construction Steel Research 66 (4), April, pp. 591-604. doi:10.1016/j.jcsr.2009.12.006