It is well known that a scorched odor is emitted from combustible materials undergoing oxidative pyrolysis and/or combustion reactions. In the present study, 15 kinds of domestic woods were selected consisting of six kinds of broad-leaved trees, eight kinds of conifer trees, and one kind of bamboo, and those were used as test samples with the pure cellulose as the reference. TG-DTA system was employed as the device that heats the sample up under the specific conditions. The sample was heated under the air (G3) condition from ambient temperature to 500°C at the rate of 2K/min to clarify the relation between the change in quality of odors and the temperature during oxidative pyrolysis. The generated gas was collected in a sampling bag which was attached to the exhaust port of a TG-DTA system, and the quality of odor was measured using electronic noses. The weight of the wood materials decreased approximately 95% in the range of approximately 200°C to 500°C due to thermal and oxidative decomposition. The ratios of similarity indices ((aldehyde)/(ammonia) and (aldehyde)/(hydrogen sulfide)) showed rapid increase during the progression of combustion and/or oxidative thermal decomposition, and which would be the important signals for the fire detection system. This property did not take the influence of a water content and kind of woods.
The effect of sample size on the characteristics (spread rate, temperature) of fire spread upward or downward in a packed bed of cylindrical pellet was experimentally examined. Ten kinds of wood (Western hemlock) pellets of which the dimension differed were used : aspect ratio 1 and 2.5, diameter 6.5~30 mm, length 6.5~75 mm. The sample size was evaluated in equivalent diameter, for example, volume equivalent diameter Dve and hydraulic equivalent diameter. It was found that the fire spread characteristics were strongly influenced in sample diameter and opening (or void fraction). Spread rate of fire Vb increased to Dve=14~16 mm linearly with sample diameter, then Vb was constant or decreased within Dve =30~35 mm regardless of the spread direction of fire. However, Vb increased in downward fire spread of the bottom side opened, when the sample diameter increased further.
This paper is concerned with experimental study on what is the minimum requirements for design of a low cost residential sprinkler system and how effect this system will be for fire suppression. At first, several basic experiments at different water pressure conditions are carried out to investigate the minimum requirements for the fire suppression. Next, fire extinguishing tests on a residential sprinkler experimental system have been taken place at a real residential house. These experimental results have pointed out several important facts. The additional water tank with electric pump will be required if the pressure at the sprinkler head is lower than 0.1 MPa. Further, if the output pressure at the head is lower than 0.02 MPa or the flow is less than 20 L/min, the fire extinguishing will be probably failed. On the other hand, if the output pressure at the head is higher than 0.1 MPa and the flow is larger than 50 L/min, the initial fire extinguishing will be succeeded. Finally, a compact residential sprinkler fire extinguishing system has been designed based the experimental results.