Impact of ventilation on the fire dynamics of an open‐plan compartment with exposed timber ceiling and columns: CodeRed #02

Abstract

AbstractThe desire by developers and architects to build mass timber buildings using cross laminated timber (CLT) and glulam has significantly increased globally in the last decade due to its benefits with regards to sustainability as well as other architectural and commercial drivers. This paper presents novel experimental evidence from CodeRed #02, the second in a series of large scale fire experiments carried out inside a purpose‐built, open‐plan compartment to capture fire dynamics in large compartments with exposed timber. The experiment used a continuous wood crib (6 × 29 m) as a controlled movable fuel load. The aim of the CodeRed #02 experiment was to study the impact of reduced ventilation on fire dynamics by keeping all other parameters the same as CodeRed #01 (Kotsovinos et al., 2022), with the exception of ventilation area which was reduced by almost half. The reduction in ventilation was found to significantly impact the fire dynamics by slowing the fire spread and burning rate. The reduced ventilation led to an increased fire duration by 4 min 30 s, which corresponds to 20% longer duration compared to CodeRed #01. The reduced ventilation had a greater overall impact on the rate of flame spread across the CLT (−23%) than the crib (−8%) compared to CodeRed #01. While the maximum temperature and incident heat fluxes inside the compartment were approximately the same as in CodeRed #01, their evolution in time and space were significantly different. The external flames were higher than in CodeRed #01 (3–3.5 m compared to 2.5–3 m) and protruded further laterally (up to ~4–5 m compared to ~1 m) outside of the compartment from the large end openings. Unlike CodeRed #01, the external flaming from CodeRed #02 pulsated between visible flames and dark soot with significantly greater frequency. This was caused by the limited ventilation resulting in incomplete combustion in CodeRed #02. The peak heat release rate of CodeRed #02 was estimated to be 16.5% lower than CodeRed #01, despite having the same fuel load. The average char depth in the ceiling, as measured at the end of the experiment, was 28 mm. This was 11% greater than in CodeRed #01 and was likely due to the increased duration of the fire. After the extinction of the flaming, smouldering occurred for CodeRed #02 as it did for CodeRed #01. One of the glulam columns lost its restraint and fell to the ground because smouldering spread through the thickness of the member. The findings from this work can help engineers quantify the hazard and therefore identify appropriate solutions for the fire safety design of exposed mass timber buildings.