Comparisons by PTB, NIST, and LNE-LADG in Air and Natural Gas With Critical Venturi Nozzles Agree Within 0.05 %

The PTB, NIST, and LNE-LADG compared gas flow measurement standards in air and in natural gas in the pressure range from 0.9 to 42 bar using four critical nozzles with ISO standard toroidal shape [1]. The four facilities generated calibrated flows through the four nozzles to determine curves of the discharge coefficient (Cd) versus Reynolds number (Re) spanning the range from 2 × 10 to 2 × 10 in Re. Three of the facilities used air; the fourth facility used natural gas. The data show the laminar-to-turbulent transition at throat Reynolds numbers in the interval 1 × 10 to 2 × 10. All of the laboratories’ discharge coefficients agreed within 0.15 % or less, well within the uncertainty of the comparison. Comparing the average Cd’s for the four nozzles, all of the laboratories agree within 0.05 %. The geometry of the nozzles was measured with coordinate measuring machines at NIST and PTB. The throat diameter measurements from the two institutes agreed within 4 μm or better. The dimensional measurements were used to generate theoretical Cd curves, for both the laminar and turbulent regimes, for air and natural gas. The theoretical Cd curves agree with the experimental measurements within 0.15 % or less. The results show that critical nozzles calibrated in air may be used as working standards in natural gas with an uncertainty less than 0.2 %, provided that gas’ properties are known with sufficient accuracy.