Characterization on the modes of flow-patterns of smoke induced by building fire have been pursued.
Smoke-patterns have been characterized into the following processes ;
I ) Evolution-rate of smokes in the combustion and flame-propagation.
II) Flow-patterns through the corridors and staircases.
III) Separation of particle motion in the smoke-flow.
IV) Final diffusion process.
Studies has been conducted so far for the smoke-evolution woods and synthetic polymers in correlation with their mode of combustion and flame spread in laboratory furnaces by one of the authors^{1, 2)}.
The present investigation have been primarily concerned with the flow-patterns of smoke involved in the hot-current from the fire.
The bump of smoke current upon the ceiling has been explained in terms of Bernoulli eq. with respect to staircases, which determines the thickness of smoke layer creeping along the ceiling through the corridor in the next floor.
The flow-patterns of smoke current as interrupted by smokestops on the ceiling through the corridors have been numerically computed by I. B. M. 1620 computer on the bases of particular solution of Navier-Stokes’ eq. with or without the presence of the temperature profile in the vertical and translational flow.
The presence of Karmann Vortex has been admittedly considered from the velocity profile in the numerical solution with appropriate parameters for the smoke current interrupted by the smoke-stops.
Follow up-job will be concerned mainly on the clarification of items III) and IV) in the succeeding reports.
1) T. Handa et al ; Bull. of the Fire Prevention Society of Japan Vol. 17 No. 2 (Feb. 1968).
2) T. Handa et al ; Bull. of the Fire Prevention Society of Japan Vol. 18 No. 1 (Sept. 1968).

View full abstract

Studies have been pursued on the spontaneous ignition of woods as a source of fire, when heated for the long period at relatively low temperature. Experimental checks have been conducted on the possibilities for the spontaneous ignition of newly-constructed fresh wood-pillar which had been partially surrounded by the plaster and heated periodically by the neighboring chimney at relatively low temperature for four months.
The above studies have been pursued for the objective to elucidate the suspicion on the spontaneous combustion of wood-pillar as a cause of fire at Happo-kan Hotel in Tokyo.
For the purpose‚ systems of the same size and proportions with the real one have been revisualized for the experiment in the large-scale and measurements have been conducted to obtain the changes in the temperature-profile of the system concerned during the passage of long-period heating. As results, stationary states with the heating-media have been obtained in both periodic and continuous heating cases and no symptoms have been observed for the temperature-jump as the cause of fire. Checks have been made on the basis of Fourier’s equation involving non-linear Arrhenius-term for the balance between the heat-loss by the conduction to the circumference and gain of heat evolved in the system concerned by the thermal degradation or smouldering combustion at relatively low temperature for the long periods. Numerically computed temperature profile by I. B. M 1620 computer on the basis of the observed ones in the plaster-part of the system have been fairly coincided with the experimental ones at the center of the wood-pillar in the system concerned. Checks have also been made for the possibilities of spontaneous ignition by the contribution of the Arrhenius-terms based on activation-energies among new wood-panel‚ dead wood and saw-dust‚ taking into consideration the effect of heat-conduction by the circumferences.
As results‚ we have reached to our conclusion that there has been no reason to commit the Possibilities of the spontaneous combustion of wood-pillar as the cause of fire in the system concerned except the false case for the presence of the very fragile wood-pillar made from saw-dust or dead-wood to support the newly-constructed building.

View full abstract

In this paper we report the experimental study on the flow of smoke and air in the model anteroom in the case of upper smoke outlet and lower air inlet openings, and the comparison of this result with the field experiment of similar openings. Finally the disagreement between the model and the field experiments reported in part 1 are discussed.
The model and the apparatus used in this experiment are almost the same as in part 1, except some improvements and additions to temperature and smoke optical density measurements.
Condition of similarity assumed between model and prototype in this case is the same as part 1, that the ratios of supply air and in flow smoke volumes to anteroom volume are equal.
The scale of model is one tenth of the prototype, so that the flow l/s for the model corresponds to m³/s for the prototype.
In this experiment following results are obtained.
(1) Flow characteristics of smoke and air in the anteroom, under constant smoke optical density and inflow smoke volume, change continuously with supply air volume.
The classification of flow characteristics to types in A, B, C, D and E given in part 1 is also valid in this case, but some supplements and corrections are made on the method and on the details.
(2) The limits of supply air volume keeping up to two layer flow of types A, B and C are almost the same as in part 1, however, the limits of type B which has perfectly transparent lower layer, are much wider than part 1 at lower supply smoke volume.
(3) The relations between mean boundary height and air smoke ratio in types A, B and C show good agreement.
The mean boundary height is larger about one seventh of the ceiling height than part 1.
The mean boundary height determined by observation agrees fairly with height estimated from temperature profile，and very well with height estimated from optical density profile.
For designing smoke tower, the recommended values to the limits of supply air volume to anteroom are from 0. 6 to 5. 2 m³/s and air smoke ratio larger than 2.
(4) Good agreement is shown between mean boundary heights in the model and the field experiments with similar openings.
(5) According to conclusions (2) and (3), position of smoke outlet openings is better at upper part of the wall than fully opened from ceiling to floor as in part 1.
(6) Characteristics of temperature profile for types O, A, B, C, D and E do not differ markedly so that the classification of types by temperature profile is very difficult, however, distinct classification is possible by optical smoke density profile.
(7) Reexamination on part 1 showed that, in the absence of subwall under the ceiling, both flow types of field and model experiment are the same and types E and no direct connection of smoke from emergency opening to smoke outlet are observed.

View full abstract