2005 年 69 巻 4 号 p. 332-340
The non-stationary, hot-wire method has been developed for use with an insulated, coated probe in order to determine thermal conductivities of metals. This development was carried out using both simulation and experimentation. The simulations of the temperature increase (ΔT) of the heater were carried out using various solid samples (Al, V, Ag and W). Experiments were carried out using hot-strip and hot-wire probes to determine ΔT from the voltage change (ΔV) using the four terminal method. Measurements of ΔT on solid Fe, Ni and Ti were made using hot-strip probes with coatings of (i) silica (ca. 5 μm thickness) and (ii) mica (20-300 μm thickness). Similarly, experiments were carried out on liquid Hg and Ga using hot-wire probes with coatings of (i) silica (ca. 5 μm thickness) and (ii) alumina-based material (170-470 μm thickness). The results of simulations and experiments have shown:
(1) The following equation applies to dΔV/d ln t (t: time) obtained using an identical probe:
dΔV/d ln t=A (I3•αT•R273•XT/4π)•(1/λ)+B
where dΔV/d ln t is an average slope for the time period 1-2 s, λ is the thermal conductivity at a certain temperature (T), I is the current supplied to the heater, αT is the temperature coefficient of electric resistivity of the heater at T K, R273 is the resistance between the potential leads for the four terminal method of the heater at 273 K, XT is the resistance per unit length of the heater at T K, and A and B are probe constants.
(2) The probe constants are independent of temperature. Thus equations for other temperatures (T1) can be obtained by replacing temperature-dependent terms αT and XT in the above equation by those for T1 as follows:
dΔV/d ln t=A(I3•αT1•R273•XT1/4π)•(1/λ)+B
Using these relations, the thermal conductivity of liquid Ga was determined over the temperature range (310-500 K). It has also been found that determination of the thermal conductivity is unaffected by coating thickness providing that the thickness of the insulating layer is <ca. 300 μm.