Logo of the Physikalisch-Technische Bundesanstalt

How exactly can temperatures be measured?

03.07.2015

Fig. 1: The worldwide oldest operational triple-point-of-water cell (Physikalisch-Technische Reichsanstalt (PTR), 1928).

Within the scope of a cooperation project with the National Institute of Metrology (NIM, China), researchers succeeded at PTB in improving the uncertainty in the realization of the SI base unit "kelvin" via the triple point of water to less than 20 µK at PTB. In addition, it was possible to render proof – by means of direct comparison measurements with a triple-point-of-water cell generated at the NIM within the scope of a research visit of a Chinese guest researcher – of the fact that the realizations of the kelvin in China and Germany are in agreement, with relative deviations of smaller than 5 x 10-8 within 15 µK (k=1).

For that purpose, very precise characterizations of various triple-point-of-water cells, smaller uncertainties of the electrical instrumentation (temperature measuring bridges), and extensive investigations of the influence of glass corrosion on the concentration of impurities in water were required. Among other things, the long-term drift of the worldwide oldest triple-point-of-water cell was determined. This cell originates from the Physikalisch-Technische Reichsanstalt (PTR) and dates back to the year 1928. The deviation from the current realization of the kelvin at PTB is only 1.2 mK. Furthermore, several standard cells of the Office for Standardization, Metrology and Quality Control (ASMW) of the former German Democratic Republic (GDR) manufactured in 1979 were investigated. Thereby, very small drift rates of only 7 µK/year for cells made of quartz glass and of 15 µK/year for cells made of borosilicate glass were determined.  

In a key comparison carried out in the year 2005 (CCT-K7), the deviations between the kelvin realizations of the 21 participating institutes and the reference value were still in the range of -76 µK to +95 µK; the uncertainty indicated by PTB was at 57 µK (k=1).

Contact:

S. Rudtsch, 7.42, e-mail: Opens window for sending emailSteffen.Rudtsch(at)ptb.de