Design and Evaluation of the Effectiveness of a Scaled Single-Compartment Experimental System

Abstract

Introducing a combustible material like mass timber will increase the fuel loading of the compartment, where this added fuel can lead to ‘non-standard’ fire behavior that changes with ventilation conditions. The National Building Code of Canada guidelines lack design provisions that address this ‘non-standard’ fire behavior, and there are no elaboration provided on how the thermal boundary conditions and the fire induced environment changes with the introduction of exposed mass timber for different ventilation conditions. To validate the impact of changing the design parameters such as introducing exposed mass timber or altering the ventilation conditions, then it is crucial to describe the thermal boundary conditions and the fire induced environment. However, to describe the thermal boundary conditions and the fire induced environment of a compartment fire, single-compartment fire tests with variety and high resolution data are necessary. As a part of the ‘Compartment Fire Dynamics in Timber Structures Under Differing Ventilation Scenarios’ research project, the focus is to collect and review existing experiments and data on the behavior of fires in compartments within mass timber buildings under realistic fire exposure. In addition, the work undertook to design and characterize an instrumented scaled single-compartment experimental system in which design parameters, such as the ventilation conditions, location and percentages of exposed mass timber could systematically be varied is outlined. Compartment fire tests with different loads and fuels are tested within the experimental system to establish a suitable realistic fire exposure for the scaled compartment. The tests aimed to utilize established realistic fire exposure for further testing and to investigate the effectiveness of the experimental system to capture differences in the thermal boundary conditions and the fire induced environment between a reference case and one with 5% exposed CLT. For the instrumented scaled single-compartment experimental system, the design fire determined as suitable to achieve the thermal boundary conditions required for ignition and charring of the exposed CLT was a 36 stick wood crib fire with an average fuel loading of 130 MJ/m². The comparative plots between the other fuel loads and types reveal that the wood crib could achieve gas temperatures and heat fluxes above the threshold values (500ᵒC and 20 kW/m²) most consistently for the longest duration of time. A broad summary of the compartment fire results present the high resolution of thermocouples and DFT heat flux gauges inside the compartment and the instrumentation in the compartment opening of the test compartment are determined to be effective. The effectiveness of the high resolution of thermocouples instrumented inside the test compartment was demonstrated by the creation of detailed gas temperature results, which captured the contribution of the added 5% exposed CLT to the compartment fire environment in comparison to the reference case. Capturing the increased gas temperatures both globally and locally to the walls in proximity to the burning CLT. Sufficient data was obtained to estimate the HRR in the compartment opening, model the time evolution of the NP height throughout the fire, and model the mass inflow and outflow across the compartment opening. Details of small changes in the NP height, mass outflow, and mass inflow due to the introduction of 5% exposed CLT were captured. Estimations of opening HRR revealed that the majority of combustion occurred inside the test compartment even with the introduction of 5% exposed CLT, but further testing is needed to confirm if this would hold for different ventilation conditions. In addition, these comparisons have revealed that the mixing in the compartment opening and fluctuations in the NP height could influence estimations of opening HRR. Future testing is needed with changes in the percentages of exposed mass timber and ventilation conditions to determine if the fluctuations in NP height or mixing in the compartment opening worsens such that estimation of the opening HRR is no longer feasible. The sensitivity of the thermocouples inside the compartment was demonstrated by its ability to capture that the ignition of the CLT occurred within a range of gas temperatures between 423ᵒC to 481ᵒC, and that the CLT flameout occurred within a range of gas temperatures between 573ᵒC to 588ᵒC. The effectiveness of the high density of heat flux gauges was highlighted by its ability to capture that the exposed CLT increases the severity of heat flux exposure in different locations through a fire test. The high resolution of O₂, CO₂, and CO effectively captured local and global differences with the introduction of 5% exposed CLT. These differences included the overall greater consumption of oxygen globally, the creation of locally oxygen-limited state that affects both the hot layer and the wood crib, increased global production of CO₂, and decreased local production of CO. The furniture calorimeter captured that even with 5% exposed CLT it could significantly increase the fire size and severity. The high variety of instrumentation captured that the introduction of 5% exposed CLT decreased the combustion efficiency of the wood crib fuel. Furthermore, the combination of in-depth thermocouples and the ability to perform cross-sectional cuts of the reduced scale CLT specimen provided insight into the impact compartment fire exposure has on charring and the thermal propagation of heat. Overall, the instrumented scaled single-compartment experimental system is effective in capturing differences in thermal boundary conditions and fire induced environment generated in fires with and without a small percentage of exposed CLT. However, recommendations are presented on how to increase the robustness of data collected by test compartment and to improve the effectiveness of test compartment in capturing differences in the fire induced environments. Recommendations were also advanced for future work with different design parameters that will provide further insight into the test compartments effectiveness and address current knowledge gaps.