Vol.3 No.1 2010

35/110

Research paper : Improving the reliability of temperature measurements up to 1550 ℃ (M. Arai et al.)−32−Synthesiology - English edition Vol.3 No.1 (2010) of approximately 0.05 °C (standard deviation). By using this device, thermocouples could be calibrated with the expanded uncertainty of 0.79 °C (level of confidence of approximately 95 %) at the melting point of palladium. 4.2 Element 2: Technique for fabrication of stable Pt/Pd thermocouples4.2.1 Drift and inhomogeneity of thermocouplesWhen transferring the temperature standard from the temperature fixed point to a thermocouple, the stability of the thermocouple itself is the largest component of uncertainty. Figure 9 shows the property changes of a new thermocouple when the thermocouple is inserted into a high-temperature furnace and the measuring junction is exposed to high temperature for a long time at a fixed insertion length. S is called the Seebeck coefficient of the thermocouple, and for simplification to consider, here, it is assumed that S has no temperature dependency. The E in the figure shows the electric field generated at the wire in the temperature gradient region, and the integration of E along the thermocouple wire is the emf that is actually observed using a voltmeter. E and S has the relationship E = S dT/dx[9][10]. Here, x is the positional coordinate along the wire of the thermocouple.When the new thermocouple is inserted into the furnace, in the beginning, the Seebeck coefficient S shows a constant value regardless of position x (this is called homogeneity), as shown by the solid line in Fig. 9(c). When the measuring junction is exposed to high temperature with the fixed insertion length of the thermocouple, the Seebeck coefficient of the part exposed to high temperature gradually changes due to the compositional and structural changes of the thermocouple wire and may become inhomogeneous in some parts, as shown by the dashed line in Fig. 9(c). With such changes in the Seebeck coefficient S, the electric field E changes as shown in Fig. 9(d), and the emf change is observed as a result. When the measuring junction is exposed to high temperature with the thermocouple fixed to a certain position, the emf change called the drift is observed. The tendency and magnitude of this emf change vary greatly depending on the type of thermocouples.In the case where the temperatures of the measuring junction and the reference junction are constant, wherever the temperature gradient is on the position of the wire, the emf shows the same value regardless of the insertion depth, because the emf is determined by the temperatures of the measuring junction and the reference junction of the thermocouple that has homogeneous Seebeck coefficient S, as seen in the new thermocouple in Fig. 9(c). On the other hand, in the thermocouple that is exposed to high temperature for a long time in the furnace and a drift has been observed, the emf changes when the insertion length of the thermocouple Fig. 9 Emf change by the exposure of thermocouple to high temperature.The solid line indicates the Seebeck coefficient and the electric field of the new thermocouple, and the dashed line shows the values after exposure to high temperature.Temperature distribution(a)(b)(c)(d)MeasuringjunctionThermocoupleCopper wireVoltmeterReferencejunction(0 ℃)FurnaceTSChange in Seebeck coefficient SEChange in electric field Exxx000New productAfter exposureFig. 10 Emf change when the insertion length of the thermocouple into high temperature furnace is altered.The solid line shows the Seebeck coefficient and the electric field when the thermocouple exposed to high temperature is inserted, and the dashed line shows the values when it is withdrawn.Temperature distribution(a)(b)(c)InsertionWithdrawalTSChange in Seebeck coefficient SEChange in electric field Exxx000

※このページを正しく表示するにはFlashPlayer9以上が必要です