An apparatus which produces highly intense beam of radiation with the maximum intensity of 220 cal/cm²/sec has been constructed (Fig. 1 and Fig. 2). Radiant flux from the positive crater of a 2.4kw D.C.arc is condensed by an ellipsoidal mirror with diameter of 14 inches and focused on the plane of a shutter. Then the diverging beam from the plane is received by another ellipsoidal mirror with diameter of 17 inches and converged in the window of a target. The effective diameter of the hole of the target is about 9mm. The shutter, which also has been built by the author, is situated in the focal plane of the first ellipsoidal mirror. The shutter has two thin alminum blades the surface of which are highly polished (Fig. 3 and Fig. 4). The blades are released manually or automatically by means of an electro-magnet and then driven by helical springs. The exposure time of the shutter is controlled by an electrical circuit associated (Fig. 5 and Fig. 6). It may be varied from 0.1 to several seconds. The blades open or close in about 0.01 sec (Fig. 7). By the use of apparatus, many substances were irradiated by intense beams of radiant energy from 15 to 220 cal/cm²/sec. And time required for ignition of those substances was measured at each intensity. Several results are shown in Figure 11. Then using those results, total radiant energies applied to substances until ignition occurred were obtained. While the total energies are almost constant with medium radiation intensities from 5 to 50 cal/cm²/sec, they increase in value with radiation of both higher and lower intensities (Fig. 12). Values of minimum radiant energy for ignition of various substances are shown in Table 1.

View full abstract

Author has been engaged in research on above-captioned subject continually with field work data and found the following facts through the analytical study. The conflagrations, in which burnt down more than 500 houses, were much influenced by an extraordinary dry strong wind, blowing at the time of break out of fire.
The nature of these winds can be classified into the following four points :
1) Strong winds connected with conflagrations were almost dry and over the with scale 3 (velocity over 5m/sec) and blowing about over 5～20 hours.
2) The conflagrations occurred in “San-in & Hokuriku” regions which broke out from April to October were very much influenced by cyclones from the Asian continent and the Typhoons passing through the “Kyushu” region from south to north and those Typhoons moving to NNE direction in Japan Sea.
They have some tendency to cause the Foehn and found that these phenomena have great relation to the occurrence of conflagration in “San-in & Hokuriku” regions.
3) Conflagrations in winter season (from Nov. to Mar.) have been greatly influenced by the winter monsoon blowing from Siberian region. Distribution of atmospheric pressure, on that occasions, were almost high in the south and low in the north.
4) At the time the fire broke out, there was no relation on the time of beginning of blowing of strong wind but was found that almost same time when the wind velocity reached the peak and humidity relatively low. And also it was almost at the same time that the time of extinguishment of the fire and wind direction change and wind velocity became weak.

View full abstract

A number of experiments on this theme have been made with small model houses whose windows are of various sizes. According to the experiments, temperature distribution of ejected air current along the downstream is much similar to that of the upward current from a rectangular heat source.
The writer compared these two cases carefully and has come to the conclusion : As far as the temperature distribution is concerned, the window, which is in vertical plane and ejects the hot stream produced by the indoor fire, can be replaced by the rectangular heat source which is placed in the horizontal plane at the same level of the top of the window. If we assume that there is a high vertical wall above the window, when the horizontal length of the window is shorter than the vertical one, the length of one side of the replaced rectangular heat source is the same as the horizontal length of the window, but the length of the other side must be one-half of the vertical length of the window. But when the horizontal length of the window is longer than the vertical one, the size of the replaced rectangular heat source is the same as that of the window. The above facts are explained physically in this report.
The temperature distribution of the upward current from a rectangular heat source is already reported by the same writer¹⁾. When the result of this study is applied, the temperature distribution of the ejected current from a rectangular window can be gained if only the size of the window and the quantity of heat ejected from the window per unit time are given.

View full abstract

1) Fire Ratings of Safes and Insulated Containers in U. S. A.
Underwriters’ Laboratory in U. S. A., which has been engaging in fire tests for long, has established fire ratings of safes and insulated containers. They are given in Table 1. The fire test consists of three tests :
(a) Fire resistance test ; a heating of a standardized fire, and allowable interior temperature during the test is 350°F.
(b) Explosion test ; a sudden heating of 2,000°F
(c) Impact test ; an impact from 30ft after being heated and reheating in the inverted position after the impact.
2) Actual Result of Labeled Safes.
Actual results of safes which have been in a fire are given in Tables 2 and 3. According to the Table 3, 88% of labeled safes retained usability of papers, but only 60% of uninspected safes did.
3) Fire Tests of Safes in our Institute. We made fire tests of more than ten safes. Thermal insulating materials of these safes were mainly made of foam concrete, but No. 8 was of rockwool, and No. 15 and No. 16 of gypsum. No. 12 was reheated three days after the fire test, and next day a wet towel was put in it and heated again.
Desirable factor of an insulating material of safe is that the material has not small thermal conductivity but large heat capacity. Foam concrete is suitable for the insulating material of safe because of its much water content.
It was obvious that the insulating material of a safe varied in thickness ; door was very thin, but too thick in both walls.
It is necessary to take out papers from safe as soon as possible after it is involved in fire.

View full abstract

We investigated here the ignition temperatures of fibre materials such as cotton and staple fibre and their change according to their packing density and to the diameter of spherical vessel in which materials are packed.
The results obtained are as follows :
1) The materials were packed in a same vessel and the packing density was varied. In this case, the ignition temperature increased, at first, linearly with logarithm of density. But, this linearity failed to hold as the density became large.
2) The ignition temperature of several kinds of material were observed and raw cotton showed the smallest value of ignition temperature.

View full abstract

It has been known that foam had to be applied gently to the oil surface and if foam is allowed to fall violently on the oil, several factors will tend to prevent the foam being effective. But it could not be found in the literature about the dependence of expansion of foam and kinds of petroleum upon petroleum contamination and rate of coverage of foam layer in violent falling application. Experiments were carried out as that foam stream was permitted to fall from outlet of 75cm height in violent application and 20cm in gentle application, on non-burning petroleums floated on water which contained in test vessel of 48cm diameter. Air foam compound used was a FeSO₄-fortified keratine hydrolyzate of 3% type and petroleums were gasoline (0.94c, st, at 20°C), mineral turpentine class 2 (0.71), class 4 (1.3), gas oil (3.0), heavy oil (104) and mixtures of heavy oil and gas oil : 60 : 40 volume ratio (19.4) and 80 : 20 (45.1). Expansions of foam were 7, 10 and 15, and water rate 1.4 and 2.8l/m²/min. From photographs, percentage of coverage (percent. of foam-covered area to whole oil surface) and percent. of petrol. contamination (percent. of contaminated parts of foam layer to foam-covered area) were obtained by means of planimeter. At first it was found that from curves of percent. of coverage to time, rates of coverage of foam layer in violent application were not as low as rates in gentle application about any experimental conditions, that is, expansions of foam, water rates and petroleums used. Secondly, it was found that from curves of percent. of petrol. contamination to time, the following results obtained :
(1) Petrol. contamination so easily occurred that in falling application, extensive contamination was observed about foams of low expansion. (exp. 7) when height of discharge outlet was only 75cm.
(2) Contamination occurred especially about low expansion and petrol. of low viscosity.
(3) At high expansion (exp. 15), contamination was not observed, and medium expansion (exp. 10) considerable contamination occurred.
(4) Contamination was observed about viscous petrol. having viscosity as high as 19.4c.st., and not observed about heavy oil (104c.st.).
Author concluded that low effectivity of foam in violent application was due mainly to petrol. contamination and not to rate of coverage of foam layer. Lastly, the comparison of the rate of coverage in application along wall with rate in falling application was conducted. In application along wall, foam stream was falling along the brass plate. The height of outlet was 75cm in the former and 20cm in the latter, and both cases petrol. contamination not occurred. And it was found that application along wall was so effective that there was no contamination and rate of coverage was remarkable as compare with rate in falling application.

View full abstract

Recent buildings have a tendency toward large glass windows in their appearance. But it is an important problem to make clear the fire resistance of glass windows when a fire occurred.
The authors report the fire resistance of various kinds of glass windows.
1. Test Specimens
Kinds of glass used in this test are normal plain glass (3mm thickness, clear), wired glass (6.8mm thickness, hexagonal net, striped pattern) and wired glass (6.8mm thickness, diamond net, rift pattern).
Two types of sashes are tested, and the glass is attached to the sash by bead or putty.
2. Testing Method
Specimens are tested by JIS A 1301 “Testing Method of Outdoor II Class Heating (850°C)”. Temperatures at four points are measured.
3. Results
a. Normal plain glass is not fire resistive, a part of glass is destroyed in 5′30"-8′30".
b. There is no difference between two types of sash in the fire resistivity.
c. Wired glass cracks with the heat after 10 minutes, but the glass is not shattered and the flame does not shoot through the cracks.
d. The front putty smells something burning after 10 minutes, and expands slowly. Its surface changes to black porous material.
e. The glass attached by putty without clips is not fire resistive fails to pass for this test.
f. Wired glass pass for “Outdoor II Class Test”.
g. It is positively dangerous to keep an inflammable within 5cm from glass, and an inflammable within distance of 15cm is dangerous in case of longtime heating.

View full abstract