Beam and Vaulted Ceiling Fire Tests: Investigation of Sprinkler Activation under Sloped Ceilings

This page may contain links to non-U.S. government websites. What this means to you »

The purpose of this study was to first determine the effect of ceiling geometry, including beamed, sloped, and sloped beamed ceilings, on the activation time of a residential sprinkler, and secondly, to compare predicted results from the National Institute of Standards and Technology (NIST) computational fluid dynamics fire model Fire Dynamics Simulator (FDS) with the actual test data obtained.

A series of 72 experiments was conducted to compare the effects of beamed, sloped, and sloped beamed ceilings on the activation times of a quick response residential pendent sprinkler. Six different geometries were studied. The geometries included a smooth horizontal ceiling, a horizontal beamed ceiling, a smooth ceiling with a slope of 13', a beamed ceiling with a slope of 13', a smooth ceiling with a slope of 24', and a beamed ceiling with a slope of 24'. For each configuration, the fire source, a computer controlled methane gas burner, was placed in three different locations. Additionally, for each burner location, the flow of methane gas to the burner was supplied in such a way as to give two different fire growth rates. For each ceiling configuration and fire growth rate two experiments were performed. Measurements taken include the time to sprinkler activation, temperature and velocity of the ceiling jet at the sprinkler of activation, and temperatures at various other locations and elevations within the fire compartment. Additionally, all geometries were modeled using the NIST computational fluid dynamics fire model Fire Dynamics Simulator (FDS) to compare predicted results with the test data obtained. It was found in a majority of cases that simply sloping the ceiling to an angle of 13' or 24' decreased the activation time of the sprinkler when compared to a smooth horizontal ceiling. However adding beams to the ceiling caused an increase in sprinkler activation time in all but three cases. For the Fire Dynamics Simulator model, the best prediction was the beamed ceiling sloped at 24' where model predictions were within an average of 4 % of measured times. The worst case for model prediction was the smooth ceiling sloped at 13'; in these cases Fire Dynamics Simulator predicted activation times within an average of 26 % of measured times.