Thermal explosion with linear time dependent boundary condition was investigated from the exact numerical analysis. Critical condition for ignition was determined by means of a new ignition criterion. This criterion was based on the distortion of the temperature profile from the stable parabolic shape. The critical condition for ignition was indicated by the curved surface in δ-B-σ space (σ is the dimensionless heating rate). Further the empirical relationship for criticality was derived from the exact numerical solution and connected with Gray's results for the traditional theory with reactant consumption admitted. Also the numerical analysis was in good agreement with the experimental results.

View full abstract

Paper phenolic laminates are widely used as insulation boards for printed and integrated circuits and for battery switches. The insulation is sometimes affected by electrolytic corrosion under polluted conditions. The tracking in organic insulators under wet polluted conditions has been studied by many investigators. However, no satisfactory analysis regarding the behavior of metals which are dissolved and deposited from copper electrodes on an insulation surface under low dc or ac voltage has been reported.
This paper describes the correlation between deterioration of insulating material and the behavior of dissolved and deposited metals from copper electrodes at the interface between an electrolyte and paper phenolic laminate using a test method similar to the IEC (International Electrotechnical Commission) tracking test under application of dc or ac voltage. The paper also describes the results of an analysis of the dissolved and deposited metals obtained by scanning electron microscopy (SEM) and X-ray diffraction.
In the case of dc voltage application, the metal was dissolved from the positive electrode in the first stage of applying drops of electrolytic solution. In successive stages metals were deposited from the negative electrode. When the deposited metals bridged both electrodes, the test solution was evaporated by Joule heating caused by the leakage current flowing through the deposited metals produced at the insulation surface. Then, scintillation appeared at the edge of the positive electrode. From the results of the analysis of dissolved and deposited metals obtained by SEM and X-ray diffraction, it was found that the dissolved metallic compounds from the copper electrode were Cu(OH)Cl and Cu₂(OH)₃Cl. On the other hand, the deposited metal originating from negative electrode was Cu. Furthermore, the deposited metals after scintillation appeared were Cu(OH)Cl, CuCl, Cu₂O and Cu.
In the case of ac voltage application, the test solution was evaporated by Joule heating caused by the leakage current in the test solution in the first stage of applying the drops, and then the metal was dissolved from both electrodes. It was observed that the dissolved metallic compounds were Cu(OH)Cl and Cu₂(OH)₃Cl. In successive stages scintillation appeared at the center between electrodes, and Cu(OH)Cl, Cu₂O, CuCl and Cu were produced. After scintillation disappeared, these deposited metals diffused to both electrodes. In spite of repeat of these processes, tracking was not formed.
From these experimental results, the relation between the processes and mechanism of production and diffusion of copper metallic compounds and the deterioration of insulating material were discussed.

View full abstract

Wet water has been known as having an effective extinguishing action for the fires of solid fuels, but it has scarcely been used. It is because the effects of wet water is not appreciated precisely.
For this reason, small scale extinction tests of tire and wood crib fires using wet water and plain water are conducted, and the total amount of water required for a complete fire extinction, the useless water drained off, the effective water soaked into a burning solid and the other items were measured in detail.
Thus, it was found from the results that the difference of fire suppression effect between plain and wet water is not so large for wood fire, although striking for tire fire, and that the effect of wet water is its durability in comparison with plain water, because of its large soaking ability.

View full abstract

In this paper, we consider of fire spread between wooden buildings. The relation between burnt area (A[m²]) and time (x[min]) is approximately S-curve. Former theories are not able to estimate real fire spread condition correctly. As the most appropriate expression, we examine the polynominal and logistic curve. By the method of MAICE, degree of the optimal polynominal is shown three. But the polynominal is not fit for real fire spread phenomenon because it is not always monotone function. On the other hand, we suggest that the function of logistic curve is good enough to explain the fire spread phenomenon. And by modifying the logistic curve, real fire spread phenomenon is shown more correctly. By using this modified logistic curve we can explain not only fire spread phenomenon of one building but also the phenomenon of some buildings by summing up the burnt area of each building.
For determination of the parameters of logistic curve, strictly we should take the method of non-linear least square. However, in view of the accuracy of data, the convergence condition and so on, we can put to practical use of two expedient methods, log-linear least square method and finite difference graphical method. Comparison to former method, latter one is more useful because it is able to estimate the value of the parameter G.

View full abstract

Most studies on mathematical simulation for enclosure fires have been conducted by using energy and mass balanse equations. The terms of generation and consumption of heat or mass in these equations determines the worth of the models. Especially, that of heat generation in an enclosure seems to be troublesome to formulate. It was indicated that the dependence of heat capacities and chemical equilibrium compositions on temperature have great influences on the evaluation of heat generation in an enclosure. Reaction efficiency, which is related to the degree of mixing of fuel and entrained air, have also great effects on this evaluation. In this paper, the variances of gas temperature in fire were calculated by using the estimated degree of mixing. And the effects of heat capacities of chemical shemical species and mixing of fuel and entrained air on the evaluation of heat generation in an enclosure were discussed.

View full abstract

There are little theoretical works on the reignition of water-extinguished wood. Usui³⁾⁴⁾ performed the experiments on the reignition time and obtained relations among the char-depth, water content and environmental temperature.
The author rearranged his data and drew an interesting simple conclusion.
By assuming the reignition time, τ (min), is the time required for total evaporation of the soaked water, the following theoretical equation must hold:
Lx = φσ (T ⁴ − T₀⁴) · τ
where,
L  :   heat required for evaporation of water (≈ 620 kcal/kgH₂O)
x  :   water content (kg/m²)
φ  :   overall heat absorption coefficient
σ  :   Stephan Boltzman coefficient (kcal/m² · °K⁴ · min)
T  :   environmental temperature (°K)
T₀ :   surface temperature of the wet extinguished wood (≈ 373°k)
because T ⁴ » T₀⁴ , τ is rewritten simply as,
τ = (L /φσT ⁴) x (min)
Using φ ≈ 0.7, the time calculated agreed well with Usui's experimental results.

View full abstract