For the derivation of the coefficients of the
deviation function in the range from 0 to 400 C,
we used the data obtained by comparison with a
SPRT in water, oil, and salt baths which are
given in Table 3. The coefficients of the deviation
function derived by the method of least square
were added to those of the reference function to
give the coefficients of the calibration function for
the thermocouple when used in the range 0–400
C.
The coefficients of the polynomial derived by
the calibration of the Pt/Pd thermocouple in the
range from 0 to 400 C are
b0 ¼ 0:000000; b5 ¼ 2:012523e 11;
b1 ¼ 5:419459; b6 ¼ 1:268514e 14;
b2 ¼ 2:380234e 3; b7 ¼ 2:257823e 17;
b3 ¼ 3:640398e 6; b8 ¼ 8:510068e 21:
b4 ¼ 2:992243e 8;
The deviations of the measured EMFs in the range
from 0 to 400 C from the corresponding values
calculated by using our reference function are
given in Table 3 and shown in Fig. 3.
Fig. 4 shows the difference between the measured
EMFs in the range from 660 to 1064 C and
the corresponding values calculated by using the
reference function. From the figure it is clear that
the deviation at the freezing points of Al and Ag is
zero.
Figs. 3 and 4 show that the difference between
the experimental values and the calculated values
by using the polynomial suggested by Burns et al.
[1], is in the order of 2 lV.
The inhomogeneity of the thermocouple was
checked by moving the thermocouple upward
from full immersion position inside the freezing
point cells to a height of 19 cm when the temperature
of the aluminum and silver freezing plateaus
was constant. The results of that experiment are
not given in the text. The uncertainty due to inhomogeneity as given in Tables 4–6, had been
estimated.
A detailed analysis of the different components
of uncertainty is given in Tables 4–6.