by Jos G.M. van der Grinten, Arnthor Gunnarsson, Mijndert van der Beek, and Bodo Mickan
A recently performed intercomparison between three laboratories using the piston provers as a primary reference, was re-evaluated using PTB’s turbine meter model. The two DN100 transfer packages that were calibrated, were also utilized in a previous intercomparison and were calibrated at PTB using atmospheric air.
The unprocessed data show that most of the piston prover data lie within the 95% contours of the fit. Most of the data from the previous intercomparison also lie within the 95% contours of the fit of the piston prover results, which proves the consistency of the two intercomparisons. Only the air calibration results, which were measured at much lower Reynolds numbers show a different course.
In the PTB turbine meter model the meter deviation is split up into three contributions: the bearing friction, a high-speed Mach effect and the remaining flow forces. When the first two contributions are corrected for, a much better connection with the air calibration data is achieved. The consistency between the piston prover data and the data from the previous intercomparison remains. This makes the turbine meter model promising for future intercomparisons that are performed in a wider Reynolds number range.