Full scale fire test was performed in Yokohama from 14th to 16th, Sept. 1964, with a view of finding the following factors :
(a) The difference of the development of the fire in rooms which were finished by different internal linings .
(b) Smoke discharge test for the smoke tower (18m×18m) which was installed adjoining to the stair-case.
(c) The development of the fire in a large office room whose opening is very small and whose ceiling is covered with treated acoustical perforated fiberboards.
The building used for the test was of 4 stories and was to be destroyed soon after the fire test.
The results :
(i) The difference of the development of the fire in rooms which were finished with different internal linings, were very similar to that of the fire in a small model compartment. In the room finished with untreated plywood board, the period from ignition to flash over was about 4 minutes and this was the minimum value of the flash over-time. In the room, finished with treated plywood board, this period was about 6 minutes and in the room finished with plaster board, the flash over could not be seen this time.
(ii) The smoke tower discharged the smoke very well from the stair-case this time. This is thought to be due to the following two facts :
(1) The air-current which contributed to lead the smoke into the tower, was made up by the air which came through the opening of the entrance, located at the lowest end of the stair case. This current went up along the stairs and entered into the tower. The importance of the existence of an inlet of the air in the building was recognized now.
(2) In the room next to the stairs, the door between the stair-case and this room, was located on the lower part of the wall, while the opening to the smoke tower was located on the upper part. This enabled to form the draught from the door to the opening in this room and contributed to lead the smoke to the tower.
Though the tower was very useful this time, further careful considerations will be necessary for the tower in a very tall building.
(iii) In the full scale fire test which we had in June 1961, we ignited in the room (32.7m×7.1m×3.6m) which was very similar to that of this building (21.0m×8.0m×2.5m), but whose openings (total area 21.6m²) were not so small as this room (total opening area 5.8m²). At that case, the fire spread in the whole of the room in 3.5 minutes after ignition and grew up to a very heavy fire which we had never expected. This time the fire was very powerless and we had to ignite again 25 minutes after the first ignition. After the second ignition, the fire front progressed very slowly toward the openings and when it drew very near the openings, the fire suddenly became active and the fire men started fire extinguishing.

View full abstract

The combustion of solid materials which contain no oxydant in themselves takes place only on their surface supported by the air supply from outside. For the first step of the study of the propagation of combustion of this kind, the downward propagation of smouldering (combustion without flame) along a vertical sheet of cardboard was investigated, because this proved to be the simplest case both experimentally and theoretically.
The fire front proceeds downwards with a velocity V (cm/s) making an angle θ with the vertical. The propagation velocity υ perpendicular to the fire front was observed in a vessel of temperature T_a with rectangular pieces of cardboard of various breadth and thickness, and the results for υ² were summarized as follows:
υ²=0.55 × 10^-2{1/b+1/d+0.53/bd}(1/(T_i-T_a)-1.55 × 10^-3)
where T_i is the vessel temperature which makes the cardboard ignite instantaneously and was found, in an electric furnace, to be 460°C.
In the theoretical part, the authors considered the heat balance in a part of length dX in the strip just in front of the fire front with the following assumptions.
(1) When a solid material in the open air ignites, combustion gases envelop it and prevent fresh air from getting to the burning surface. Combustion takes place when fresh air reaches the surface by natural convection. The heat dQ₁ produced in the part dX in the time dt is considered to be proportional to the velocity of the ascending convection current which was assumed to vary as the temperature difference (T-T_a) between dX and the ambient air.
(2) dQ₁ is also proportional to the velocity of the combustion reaction which depends upon the ambient temperature. Since υ should be infinite when T_a→T_i , dQ₁ was assumed to be proportional to 1/(T_i-T_a).
(3) dQ₁ is proportional, too, to the surface area dS of the part dX. When burning, however, the surface of the solid material is covered with a stagnant boundary layer of combustion gases and the fresh air is supplied by diffusion through this layer from outside. Therefore, the surface area of dX was assumed to be 2(b+d+ε)dX instead of (b+d)dX, where ε is about four times the thickness of this layer.
(4) Heat is dissipated through this stagnant layer of temperature T into the ambient field of temperature T_a. Thus the heat dissipated dQ₂ from dX in dt was assumed to be 2h(b+d+ε)dX(T-T_a)dt, where h is the heat transmission coefficient.
From these considerations, Fourier’s equation was derived as follows :
cρ (∂T/∂t)=k(∂²T/∂X²)cosec²θ+2(1/b+1/d+ε/bd)(q/(T_i-T_a)-h)(T-T_a),
where q is a constant depending upon the chemical properties of the cardboard.

View full abstract