Platinum Metals Rev., 1964, 8, (4), 145
Calibrating Platinum Thermocouples
A Practical Method Based on the Freezing Points of Base Metals
The use of platinum: 10 per cent rhodium-platinum and platinum: 13 per cent rhodium-platinum thermocouples is widespread throughout industry for the continuous measurement of temperatures up to 1400°C and for short periods up to 1650°C. Both rhodium and platinum can be refined to very high states of purity, which enables the accuracy and reproducibility of new thermocouples to be guaranteed. But when industrial processes require precise temperature control the thermocouples must be recalibrated periodically because, even in the absence of contaminants, there is some slight drift from calibration after prolonged use, caused by rhodium migration from the alloy to the pure limb of the thermocouple.
Three methods are in general use for the recalibration of thermocouples after use. These are: comparison with a thermocouple of known calibration; the wire bridge method, involving calibration at the gold point(1063°C), the palladium point(1552°C), or the platinum point (1769°C); or taking cooling curves of pure metals with known freezing points, for example, aluminium (660°C) and silver (960.8°C).
However, works laboratories do not always possess high temperature furnace facilities which would enable them to recalibrate accurately at the gold, palladium or platinum points, and to meet these conditions. W. Heyne (1, 2) of the Deutsche Amt fur Messwesen und Warenpriifung, Berlin, has now reported the use of three relatively easily obtained fixed points in a modification of the freezing point method. In this method the thermal e.m.f.-temperature relationship of the thermocouple under test is determined by reference to the freezing points of copper (1083°C), zinc (419.5°C), and aluminium. The method is independent of any standard thermocouple and is therefore unaffected by the stability of another instrument.
Heyne claims that his apparatus permits the calibration of platinum : 10 per cent rhodium-platinum thermocouples from o to 1300°C. Values of the thermal e.m.f. E(t) obtained at the freezing points enable the constants to be determined in the equation:
which governs the thermal e.m.f.-temperature relationship over this range. From this equation calculation gives the e.m.f.s at 50°C stages up to 1300°C. Linear interpolation between the points so derived is not more than 1 μV in error, which represents an insignificant error in the temperature reading obtained from the thermal e.m.f.
It will be noticed that Heyne’s upper fixed point is only 1083°C, which is well below the 1300°C he claims as the upper limit of calibration by this method, but he shows that the maximum error obtained during extrapolation of results to cover this region is only ± 2° C. Systematic error between 300° and 1300°C is barely 0.1°C which is well within the uncertainty limit of 0.5°C.
Heyne’s apparatus, developed for this work, consists of a graphite crucible to contain the molten metal specimens, which are copper, aluminium or zinc of a known high degree of purity. These metals are prevented from oxidising, which would affect their freezing points, by maintaining a reducing atmosphere in several graphite chambers above the crucible. The thermocouples to be tested are protected by surrounding them with a gas-tight tube of pure sintered alumina before insertion in the apparatus. Three thermocouples can be tested simultaneously by using a graphite block as the crucible with three equally spaced holes for the molten specimens.
Heyne points out that his method, although similar in many respects to previous work, offers advantages over the Russian method using antimony in place of aluminium and is simpler than the NBS method using four fixed points (3).
This method of calibration for thermocouples appears to offer a convenient way of checking their accuracy after prolonged service at high temperatures when some drifting is suspected. The apparatus is simple and the freezing points of copper, aluminium and zinc are not difficult to obtain.
- 1 W. Heyene, Feingeratetechnik,, 1964, 13, ( 7 ), 311 315
- 2 W. Heyene, Physik,, 1964, 12, 2 90
- 3 NBS Tech. News Bull., 1961, (March), 44 – 48