Final report on key comparison CCM.P-K4 of absolute pressure standards from 1 Pa to 1000 Pa

Abstract

A key comparison of low absolute pressure standards, organized under the auspices of the Consultative Committee for Mass and Related Quantities (CCM), was carried out at seven national metrology institutes (NMIs) between March 1998 and September 1999 in order to determine the degrees of equivalence of the standards at pressures in the range 1 Pa to 1000 Pa. The primary standards, which represent two principal measurement methods, included five liquid-column manometers and four static expansion systems. The transfer standard package consisted of four high-precision pressure transducers: two capacitance diaphragm gauges to provide high resolution at low pressures, and two resonant silicon gauges to provide the required calibration stability. Two nominally identical transfer packages were used to reduce the time required for the measurements, with Package A being circulated among laboratories in the European region (Istituto di

Final report of key comparison CCM.P-K3: absolute pressure measurements in gas from 3 × 10−6Pa to 9 × 10−4Pa

“…(6) to define p jU as the mean gauge pressure reading, as these NMIs had only one calibration cycle. For the pilot laboratory (NIST), following the analysis of [13] and [14], a single value of p jU was calculated as the arithmetic mean of the four cycles. For PTB, a single value of p jU was calculated as the arithmetic mean of the two cycle values:…”

“…Prior to the comparison, NIST had no data on long-term performance of these two artifacts. Past international comparisons using vacuum gauge artifacts have also used repeated measurements at the pilot laboratory during the course of the comparison to characterize the uncertainty due to long-term shifts [11,13,14,15]. One method [11,15] assumes that the pilot laboratory value of a i , at the time of the participant measurement of a i , would be midway between the pilot laboratory value taken before and after the participant measurement.…”

“…In [13] and [14], a constant model was used to characterize the changes in the artifact behavior, which is appropriate if the changes are believed to be random. As the statistical sample of the pilot laboratory in [13] and [14] was limited (three), a Type B evaluation was used to estimate u LTS of each gauge.…”

“…As the statistical sample of the pilot laboratory in [13] and [14] was limited (three), a Type B evaluation was used to estimate u LTS of each gauge. The variation due to long-term shifts was modeled as a normal distribution, with the standard uncertainty due to long-term shifts equal to one-half the difference between the maximum and minimum values of the pilot laboratory.…”

“…(23), giving p jU that is referred to as the mean gauge pressure reading 13 . For NIST and PTB, a single value of p jU was calculated as the arithmetic mean of the multiple cycles: (24) 13 Although there is no averaging occurring for NPL, NPLI, and KRISS in p jU , we refer to the term as the mean gauge pressure reading for consistency with its definition for the SRGs.…”