Purpose -A prototype gas prover was constructed to serve as the Italian primary standard for gas flow rates in the range 0.1 ml/min to 2 l/min. The new prover is used to calibrate high-quality industrial standards, as well as the MFCs used in microelectronic fabrications and preparation of reference gas mixtures. Design/methodology/approach -The prover measures gas volume transfers caused by displacements of a 120 mm dia. motor-operated piston, which is introduced into a temperature-controlled chamber containing up to 3 l of the required working gas at near ambient conditions. Gas delivery is made at constant rate, whereas possibly variable incoming flows are measured at constant pressure. Displacements of the piston are measured by an optical interferometer. Findings -The analysis shows that standard uncertainty ranges between 0.013 and 0.03 percent. Owing to the very accurate control and measurement of both pressures and temperatures, these figures refer equally to volume and mass flowrate. Experimental comparisons with similar national standards at LNE-France and NIST-USA confirmed the consistency of measurement results in the three Nations.Research limitations/implications -The gas prover should be used with inert gases only. Practical implications -The national industrial gas standards and the best flow transducers can now be calibrated accurately down to unprecedented flowrate values. Originality/value -The need for measurement of extremely low gas flows is quite recent, therefore possibly less than ten primary national standards are available today worldwide. Several completely different principles and designs have been developed; description of design and performance of each instrument is important to assess their respective merits. The described apparatus is innovative as regards measurement range, accuracy and control techniques.

This paper presents the first results of a systematic experimental intercomparison of the metrological performances of seven Coriolis mass flowmeters that are commercially available in Italy. Each of them was mounted and tested with similar operating conditions in the IMGC primary flow calibration facilities, Which enabled the authors to vary some of the most significantly influential factors. The test fluid was water at room temperature, and the maximum flow rates were ranged between 2.8 and 4.5 kg/s, according to each manufacturer's specifications.

A compact version, designated MM1, of the HG5 primary standard manometer of the Istituto di Metrologia "G. Colonnetti" (IMGC) is described. The instrument was constructed in order to exploit thoroughly, in the range of pressures up to 5 kPa, the potential accuracy of the measurement principle and instrumentation already adopted in HG5 -the same optical components, instrument hardware and software can be used on either of the manometers. The improvement in measurement accuracy of the MM1 instrument over that of HG5 in the same range was largely obtained by reducing by an order of magnitude the lengths of both mercury and gas columns. This provision can reduce the uncertainties associated with the computation of aerostatic heads in the U-tube and with the correction of interferometer readings for gas refractivity, as well as reduce the detrimental effects of residual temperature gradients inside the manometer body. In addition, the systematic adoption of purpose-built cat's-eye optical devices allows the reflection of laser beams to be made directly on mercury menisci, whereas the massive, all-metal body construction of the manometer ensures an improved mechanical stability. The paper includes the uncertainty budget, with evaluated standard uncertainties ranging between 12 mPa at 1 Pa and 53 mPa at 5 kPa, in the relative or gauge measurement mode.

This paper is the final report of an evaluation and research project on the metrological performances of six different mass flowmeters of the Coriolis type which were extensively tested, under identical operating conditions, at the IMGC flow measurement laboratory, making use of the Italian national standard calibration facilities. The first results of the evaluation project, already detailed in a previous paper, are also briefly reviewed.evaluate the metrological characteristics of several commercially-available CMFs. The main aims of this investigation are: 8B making a comparative evaluation of as many CMFs of different design as possible, by testing them at an independent laboratory in identical conditions of practical interest to users; improving the knowledge of the metrological characteristics exhibited by a representative sample of CMFs, in order to allow a basic assessment of the performances of the whole class of such instruments.The test programme and the first results of this research were reported in Cascetta et al (1992) and are here briefly recalled. Further, this paper presents the results of the second part of the project, which consisted of on the influence of fluid density and fluid temperature. The extensive set of data given in both papers, which summarise many thousands of measurements, provides a further contribution to the documented characteristics of this type of flowmeter. The metrological evaluation programmeThe tests, as agreed with each manufacturer, include calibration of each meter:~ in water, by varying pressure (up to 400 kPa) and temperature (up to +80°C); in a test fluid (MIL-C-7042-B type II) in substitution for JP4 kerosene (in the following text called simply 'JP4 k~rc~sene'), whose density and viscosity at 20°C are respectively 768.5 kg/m3 and 1.~'l mm2/s; in a two-phase fluid, namely water mixed with a small amount (up to 10% in volume) of air bubbles. It must be stressed that this experimental evaluation was performed on a single instrument for each manufacturer. Therefore, the test results do not provide justification for a quality judgement on other flowmeters, and even less to a whole product line of instrumentation. It should also be noted that, by now, many improvements have been incorporated in most CMFs available today, owing to the fast development of this comparitively young technology.The model that each company was asked to provide was chosen in view of comparing meters with similar measurement flow ranges: more precisely, with full scale (FS) values as close as possible to each other.These FS values, for use in water at room temperature and low line pressure, ranged between 2.8 kg/s and 4.5 kg/s. All the CMFs chosen are equipped with DN 25 flanges.All flowmeters but one were tested by measuring their DC output across a standard resistor: this choice, according to previous experiences at the IMGC flow laboratory, is thought to be fully adequate and more reliable with respect to electromagnetic noise (Wade and Dandridge, 1988). Just one flowmeter, instead, was t...

No abstract