Evaluation of repeatability measurement procedures in gravimetric volume calibrations of glassware

Project Description

Following the cooperation between EURAMET TC-F and ASTM Committee E41 in the revision of the standard "E542 - Standard Practice for Calibration of Laboratory Volumetric Apparatus" it was decided to investigate the different approaches regarding repeatability measurements of glassware calibration performed by different National Metrology Laboratories, accredited laboratories and manufactures around the world and its influence to the determined volume and uncertainty.

Suitable repeatability estimates are needed for evaluating measurement results and for determining the measurement uncertainties.
Lack of repeatability agreement within a set of measurement results can lead to significant problems with the operating characteristics of the volumetric instrument.

The technical protocol can be downloaded here>>.

Final Report 2024-03-31

The project has been completed and the report can be downloaded here>.

This project had the purpose to investigate the different approaches regarding repeatability measurements of glassware calibration performed by different National Metrology Laboratories, accredited laboratories and manufactures around the world and its influence to the determined volume and uncertainty. Six laboratories have participated and two sets of flasks of the following 3 nominal volumes were calibrated:100 mL, 500 mL, 1000 mL.

OWM NIST has used the procedure described in NIST GMP 3 (Option A: setting bottom of the meniscus between upper and lower portion of graduation line ellipse), this may lead to lower volume values that the laboratories that use the ISO 4787 meniscus adjustment procedure. This led to some inconsistent results.

OWM NIST used the repeatability of three runs, five runs or ten runs and compared to historical process standard deviations where available with control charts and standard deviation charts with F-test used to assess repeatability for each flask compared to an accepted value.

Also, the use of fewer repetitions leads to higher uncertainty values.Therefore, it is advisable to use at least 10 repetitions, EURAMET cg 19 and ISO 4787 annex E, a). In general, the NMIs have better uncertainty than the accredited laboratories.

Regarding reproducibility results, it was verified that the laboratories had good agreement within instruments and with similar uncertainties.

Progress Report 2023-03-08

Several USA accredited laboratories have joined this research project. Measurements are almost completed. Only one participant is missing. All data has been sent for evaluation. The data for the official participants will be dealt separated than the ones from the USA laboratories.

Progress Report 2022-04-19

Measurements were already performed by IPQ, NORMAX, INRIM, MIRS, ZMK and Artel. The instruments are now at OWM (USA).