Fire Safe implementation of visible mass timber in tall buildings – compartment fire testing

Five real scale fire tests of compartments constructed of cross-laminated timber (CLT) and glued laminated timber, compliant with product standards specified in current US model building code, were performed. Four of the tested compartments were designed to result in a representative and severe fire scenario in a residential fire compartment, using a probabilistic approach. The other tested compartment had additional openings and a greater opening factor, which was aimed to be representative of buildings designed for business occupancy. The interior of the compartments had surface areas of exposed mass timber that varied from approximately the area of the floor plan to approximately two times the area of the floor plan. The tests included measurements to study the internal compartment exposure, the temperature development at gypsum protected surfaces, the temperature development in the structural timber, oxygen concentrations at locations of interest and exposure to exterior surfaces of the wall and façade above the openings. The fire in the compartment with a greater opening factor had two layers of fire-rated gypsum board protection on the back wall and all other surfaces of CLT and glued laminated timber exposed. Despite having the highest peak combustion rate, this compartment fire had the least severe internal and external fire exposure. The fire decayed relatively quickly after flashover and continued to decay until the test was stopped at 4 hours after ignition. This fire resulted in less structural damage than the fires in compartments with fewer and smaller openings. The compartments with fewer and smaller openings had similar temperatures for approximately the first 10 minutes after flashover. The compartment with only the ceiling (including the glued laminated timber beam) exposed started to decay after 22 minutes of post-flashover fire and continued to decay until the end of the test at 4 hours after ignition. The other three tests had, in addition to the ceiling, significant areas of exposed wall and column surfaces. To accommodate for the extended fire duration that was expected in these configurations an extra layer of gypsum board protection was applied to the protected surfaces. The additional exposed surface areas of walls led to an increase of the fully developed fire duration by 6 - 9 minutes. One of the compartments included corners where two exposed walls intersect. Significantly increased damage was observed in the lower part of these wall corners, and an overall higher radiative exposure in the test with such corners. After more than three hours of decay, surface flaming developed on the walls in that test. The fires in the tests without such corners exhibited continual decay for the full 4-hour test duration. Post-test analysis showed that the structural damage was lower in exposed ceilings than at the bottom of the exposed walls for all tests. After the tests, remaining smoldering and hot spots were reduced using relatively small amounts of water mist. Overnight measurements to study the thermal wave going through the loadbearing structure indicated no post-test reduction of structural capacity