This thesis describes a series of 5 tests that were conducted at Carleton University Fire Research Laboratory to assess the contribution of Cross Laminated Timber (CLT) panels to the development, duration and intensity of room fires. The tests were conducted in rooms constructed from 105 mm thick 3-Ply CLT panels and measured 3.5m wide by 4.5 m long by 2.5 m high. Propane and furniture fires were used with the CLT panels in protected and unprotected configurations. Data was collected on Heat Release Rate (HRR), room temperatures and charring rates. In protected configurations, no noticeable contribution was observed from the CLT panels, however in unprotected configurations, the CLT panels contributed to the fire load and increased fire growth rates and energy release rates. When charring advanced to the interface between the CLT layers, the polyurethane based adhesive failed resulting in delamination. Delaminated members contributed to the fire load and exposed uncharred timber which increased the intensity and duration of the fire. When delamination occurred, the fire in unprotected rooms continued to burn at high intensity well after the combustible contents in the room were consumed by the fire. These fires were extinguished as they could have resulted in structural failure of the test rooms.
Most of the previous work focused on fire behavior of non-combustible construction. However, few investigations have systematically addressed fire development and window ejected flame based on large-scale light timber frame construction (LTFC). This paper conducted a large-scale natural fire experiment to explore the fire development of wooden buildings and the ejected flame behavior by a two-layer light timber frame construction (LTFC). The experimental LTFC included two compartments, with four façade walls consisted of external and internal linings, within 5.1 m height, 3.6 m long and 2.4 m width, and weight of 1480.1 kg. The room temperature, mass variation in burning, radical temperature profiles outside the openings-façade wall, and ejected flame dimension were measured and analyzed. The results were summarized as follows: In LTFC, the room temperature and heat release rate (HRR) would show a second rapid rise, as if “twice flashover” occurred in fully burning stage. This phenomenon is obviously different from the traditional compartment fire development of buildings. Besides, after flashover, the ejected flame height continuously increased until the fire turned into decay stage, whereas the horizontal ejection distance would maintain a steady stage and increased as the openings broken extremely. Furthermore, the region outside the openings façade wall could be divided into three regions, ejected flame region (including continuous and intermittent flame) (Tr > 180 °C), buoyant plume region(150 °C > Tr > 60 °C) and heated air region(60 °C > Tr > T8). A modified function was proposed to predict the temperature profile at different heights for the openings-ejected flame. The data of this paper will enhance the comprehension for fire development of timber buildings and provide some useful information to assess the thermal behavior of window-ejected flame of façade wall.
This thesis studies the fire behaviour of Cross Laminated Timber (CLT) panels in partially protected rooms. A one-dimensional heat transfer model was developed to determine the fire resistance of CLT floor and wall panels. During this study, three room fire tests were conducted at Carleton University Fire Research Laboratory to determine the maximum percentage of unprotected CLT surface area that will yield similar results to that of a fully protected room. The rooms had a single opening and were constructed entirely using 3-ply, 105 mm thick CLT panels. A non-standard, parametric fire using furniture and clothing as fuel was used and 2 layers of gypsum board were used to cover the ceiling and the protected walls. The Heat Release Rate, temperature, charring rate and gypsum falloff time of each test was collected. The results obtained from the room test were then compared to the numerical heat transfer model to evaluate its accuracy.