Five full-scale fire experiments were conducted to observe the performance of a two-level apartment-style structure constructed of mass timber. Each level consisted of a one bedroom apartment, an L-shaped corridor, and a stairwell connecting the two levels. One of the primary variables considered in this test series was the amount and location of exposed mass timber. The amount of mass timber surface area protected by gypsum wallboard ranged from 100% to no protection. For each experiment, the fuel load was identical and the fire was initiated in a base cabinet in the kitchen. In the first three experiments, the fire reached flashover conditions, and subsequently underwent a cooling phase as the fuel load from combustible contents was consumed. The first three experiments were carried out for a duration of up to 4 h. In the fourth experiment, automatic fire sprinklers were installed. Sprinklers suppressed the fire automatically. In the fifth experiment, the activation of the automatic fire sprinklers was delayed by approximately 20 minutes beyond the sprinkler activation time in the fourth experiment to simulate responding fire service charging a failed sprinkler water system. A variety of instrumentation was used during the experiments, including thermocouples, bidirectional probes, optical density meters, heat flux transducers, directional flame thermometers, gas analyzers, a fire products collector, and residential smoke alarms. In addition, the experiments were documented with digital still photography, video cameras, and a thermal imaging camera. The experiments were conducted in the large burn room of the Bureau of Alcohol, Tobacco, Firearms and Explosives Fire Research Laboratory located in Beltsville, Maryland, USA. This report provides details on how each experiment was set up, how the experiments were conducted, and the instrumentation used to collect the data. A brief summary of the test results is also included. Detailed results and full data for each test are included in separate appendices.
