The hot gas spurting from a window of a burning concrete house has a tendency to flow from horizontally to vertically because of its buoyancy due to higher temperature than that of surroundings. The position of central axis of jet gives considerable influence upon the temperature of rising hot current ; when the hot air current rises apart from a wall above a window, it diffuses more widely than the current rising closely to the wall. In the former case, therefore, the temperature of the current is lower than that in the latter.
The position of a rising hot gas in various window shapes is discussed in this paper.
Results are as follows :
1. When the space above a window is free.
Using the velocity at a window as an initial condition, the equation of trajectory of the spurting gas can be obtained from the equation (1) and (6), where the upward acceleration due to the buoyancy is given.
Assuming that the rising air current is spurting from a line source placed at x_o inside a window plane, the trajectory shows a cubic parabola expressed in the equation (10) ; the parabola depends upon the window height, but has little connection with the air temperature in a burning room.
The parameter x_o was obtained from model experiments.
The above results were confirmed in model and full-scale experiments.
2. When a wall is above a window.
In this case the rising air current comes nearer to the wall than in the former case 1, depending upon the ratio of window width to window height ; when the window width is greater than the window height, the spurting air current comes nearer to the wall, and its temperature is higher, even if the spurting heat quantity per unit time is equal.

View full abstract

In petroleum-fire fighting, destruction of air foam mainly caused by the radiant heat from the flame and the contact with the high temperature petroleum (abbreviated h. t. petrol. in the following).
Although the destruction due to radiant heat has been studied by many investigators, destruction due to h. t. petrol. has been scarcely studied and author has known that J. F. French’s paper³⁾ had concerned only with that destruction in detail.
J. F. French and co-worker adopted the drained liquid volume as a measure to compare the resistance of air foam against h. t. petrol.
Author considered that the French’s method would not be suitable for that determination because the drainage was not proportional to the collapsed foam volume and there were foam compounds, whose foams drained rapidly but were destroyed a little by h. t. petrol.
A small scale test method for the resistance of foam to h. t. petrol. was given, which determined the collapsed foam volume directly and velocity of destruction of foam layer, that obtained from the former.
Experiments were conducted on three commercial foam compounds, in which two were keratine hydrolyzates and one unknown protein hydrolyzate, and on eight petroleums, in which five were petroleum products and three were crude oils.
The range of temperature of petroleums was 45～110°C.
From the experiments, the following results were obtained :
(1) Velocity of destruction of foam layer increased exponentially with the temperature of petroleums in range of the used experimental condition.
(2) Velocity of destruction was mainly governed by the contact with h. t. petrol. in the case of the used two keratine hydrolyzates and by the radiant heat of flame in the case of the unknown protein hydrolyzate.
From these facts, both the determinations of resistance to destruction of h. t. petrol. and of radiant heat could not be disregarded as the test for foam compounds.
(3) Considering fire risk and velocity of destruction, it would be suitable for the determination of resistance of foam to destruction of h. t. petrol., that the test method which presented by author in this paper, was conducted on gas oil at 70°C.

View full abstract

Fire detectors are the most important part of automatic fire alarm systems, and diaphragms are the essential part of the detectors.
Ordinarily they are made of thin phosphor bronze or brass strips. Because the heat treatment is not commonly applied to the strips, after they are punched and corrugated, they show considerable secular change which reduces the reliability of fire detectors.
The authors made diaphragms of extremely thin beryllium copper strips of 0.03mm thickness. And after repeating heat-treatment-tests at various temperatures, the tempering at 300°C for 1～1¹/₂ hour followed by some auxiliary treatments to remove the thermal striction was found most effective for decreasing secular changes of diaphragms. ln the course of these investigations, the authors found also what kind of treatments are most suitable for decreasing the secular changes in the case of phosphor bronze diaphragms.

View full abstract

In our previous papers, we have introduced theoretical formulae which render it possible to calculate the pressure produced in the pressure-responsive-chamber of a pneumatic tube type fire detector, if the temperature variation of the tube be given.
But it is necessary to know how the temperature of the tube fluctuates with the time in the case of an actual fire or in the case of room heating, in order that these formulae may be utilized properly.
In this report, we give particulars of our experimental investigations on the fluctuation of the temperature of a pneumatic tube.
As a practical installation, a long single tube of copper is commonly used and is set (in a coil) so as to extend to all the peripherical edges of ceilings of several rooms. In our investigation, our attention was firstly directed to one portion of this tube and the local temperature of this portion was discussed. From this, a conception of the efficiency of a pneumatic tube for detecting a fire was established. Such efficiencies are shown in Table 1 by the reference letter η.
Next, we dealt with the problem of the mean temperature of the whole pneumatic tube, and obtained the following experimental formula which gives the relation among the mean temperature, the size and the structure of the room, and the magnitude of the heat source :
θ₀＝k_tS_ε^-0.057L
where L denotes the length of a pneumatic tube ; S the surface area of a pan which contains burning methanol ; and the coefficient k_t approximately 0.014 for concrete building, while 0.011 for a wooden building ; θ₀ the rate of temperature-rise (°C/min) ; and ε the base of the natural logarithm.
This experimental formula is applicable to the case where the pneumatic tube is in contact with the periphery of the ceiling of a room and the pan containing burning methanol is located at the center of the room which is nearly square in plan, with a height (from floor to ceiling) of 3 to 3.5m. If the shape of the room is rectangular and appreciably deviates from square shape, θ₀ will be a little higher than what is predicted by the above formula. This formula can be applied to the case where the setting of the tube is in a usual way, provided the tube is not painted at all after it has been installed. But, if the tube is painted after having been fixed to the ceiling, θ₀ will be lower than that given by this formula by some 20 to 30%.

View full abstract

There are various methods to measure the vapor density of liquid on a surface. The writer worked out a way to measure the vapor density of liquid except water by means of making improvement on a wet and dry type psychrometer.
The principle of a wet and dry type psychrometer is that the difference of vapor pressure is proportional to the temperature difference between dry and wet parts. When the temperature difference increases and the temperature of wet part decreases to the dew-point of water vapor in air, the temperature difference between dry and wet parts is not proportional to the difference of vapor pressure, because water vapor becomes to condense around the surface of wet part. Now, if wet part should be heated and its temperature become equal to ambient air temperature, water vapor in air would not condense around the surface of wet part.
Fig. 1 shows the measuring element produced for trial in accordance with the consideration mentioned above.
The distribution of vapor density of benzene and ethyl alcohol in vessel was measured with this element, in this case, the diameter of vessel, the height of vessel wall on liquid surface and the hole area of vessel cover were varied.

View full abstract

Position of fire front for 30 minutes in conflagration in city area have been surveyed and well illustrated especially in recent years by Mr. K. Kamei. ln this case, the fire front was defined as a line connecting places where buildings were simultaneously burnt down by a spreading fire. Measuring the fire front shifting at the points of intersection on mesh drawn on maps for seven cities respectively, the following results were obtained.
(1) Velocity of fire front shifting immediately after the outbreak of fire was extraordinarily great. In the extreme case, it reached 1670 meters per hour. Its direction coincided closely with the prevailing wind direction at that time. Therefore, shape of the fire front in the first one hour formed something like a long tongue.
(2) ln the following one hour from the outbreak of fire, the velocity of fire front decreased sharply and its direction, on the contrary, spread widely. Accordingly, the fire front seems to have a tendency that their velocity at the respective direction gradually becomes almost equal. These circumstances are shown in Fig. 1 A and B.
(3) Hourly change of the velocity of fire front showed remarkable characteristics that on usual great fires the 2nd maximum appeared, compared with in greater conflagrations the 2nd and also the 3rd maxima, crused by powerless prevention against the fires. As shown in Fig. 2, the occurrence time of those maxima depends upon the wind velocity and the velocity of fire front shifting, the latter of which is given by meters per hour in brackets in Fig. 2.
(4) Decreasing rate of the velocity of fire front shifting has a close relation to the own absolute velocity : Fig. 3 represents this relation that the higher the velocity of fire front, the greater the decreasing rate.

View full abstract