A series of experiments has been undertaken to investigate the behaviour and the performance of a clamp-on transit-time ultrasonic flowmeter in two-phase air/water flow. The results show the performance of the liquid ultrasonic meter to be seriously affected by the presence of free gas in a manner that is dependent on the actual flowrate of the gas and the flow regime. The data presented here covers water/air flow under slug flow regime and an evaluation of transit-time ultrasonic meter and its suitability for two-phase flow measurements. The factors affecting performance are discussed and a closure method using a transit-time ultrasonic flowmeter is introduced.
The closure method for the liquid flowrate employs a mathematical relationship, under slug flow conditions, between the estimated liquid flowrate Q_(Estimated); the indicated liquid flowrate Q_(Indicated); a velocity profile correction factor, K₂(Re); the partial-filled area of the pipe; the cross-sectional area of the pipe and a height correction factor K₁(h). This closure method for the gas flowrate uses measurements from the clamp-on transit-time ultrasonic flowmeter, an axial differential pressure transducer and a vertical differential pressure transducer.
The height of the water is obtained by using the vertical pressure transducer and applying the hydrostatic pressure equation. The height of the film reflects at the same time the functional condition of the ultrasonic flowmeter whether it is in a functional state or not. The acceptable height which keeps the clamp-on transit-time ultrasonic flowmeter to continue to function is above 50% of the pipe diameter. An axial differential pressure sensor gives the pressure drop along the slug body. The flowrate of the liquid is constant throughout the tests; the gas flowrate was increased over the range of tests.
Two main correction factors are implemented in the closure model. The Velocity Profile Correction Factor (VPCF), which is a function of Reynolds number K₂(Re) for fully developed pipe flow for an assumed Reynolds Number, and the height of the water film correction factor K₁(h). The K₁(h) correction factor relates the velocity measured along the beam to the mean flow velocity over the entire cross-section of the flowing medium under partially filled pipe conditions.
It is found that the clamp-on transit-time ultrasonic flowmeter can be successfully applied to a horizontal slug flow regime with the application of this closure method to the indicated ultrasonic flowrate Q_(Indicated). The average error of the clamp-on ultrasonic flowmeter measurements in two- phase flow were reduced by 50% to 75% and that is dependent on the height of the water, gas flowrate, and the slug flow characteristics. The average error of the estimated flowrate of the ultrasonic flowmeter is ± 3.5% to ± 6.5% relative to the reference turbine flowmeter. Applying a pressure drop closure model at a given liquid flowrate provides an air flowrate measurement with an accuracy of better than ±10%.

The work refers to the vortex flow meter optimisation. Searching of the optimal geometry of the meter became the fundamental task for designers. For many years their attention has been focused mainly on the bluff body as well as on the sensor designing. Numerous experiments made by authors of the paper confirm that not only the bluff body shape but also geometry of the duct impacts the vortices development. Duct walls stabilize the vortex shedding and its development process. Hence the conception of flow duct tapering in the vortices development zone. On the basis of the numerical simulation it is concluded that due to the pipe cross-section contraction (causing the flow velocity increase) the vortex rotation energy enhancement as well as vortex life-time increase has been attained.

PTB’s new "Hydrodynamic Test Field", which represents a high-accuracy water flow calibration facility, will serve as the national primary standard for flow measurands: volumetric and mass flow rate, respectively, and total flow measurement, i.e. the quantity of fluid (volume or mass) passing a flowmeter. As most accuracy determining component parts, it comprises three different-size dual-balance gravimetric reference systems: 300 kg, 3 tons and 30 tons. This type of gravimetric references were realized as a combination of a strain gauge based and electromagnetic force-compensation load cell based balance, each. Though each of these two weighing principles fulfils the individual accuracy requirements that have been derived from the total measurement uncertainty budget of the calibration facility as a whole, the electromagnetic force-compensation load cells reveal several advantages concerning linearity error, hysteresis error and sensitivity.

A new design concept for improving the sensitivity of a Coriolis flowmeter is proposed. The differences between the developed flowmeter and current commercial products lie in adoption of acceleration sensors and reinforcement frames. These features impart an interesting characteristic: sensitivity improves with increasing drive frequency. A flow tube has been fabricated on the basis of the new design concept, and subjected to basic experiments in order to confirm the characteristics of the measurement system. Although some problems still remain, the characteristics have been partially confirmed.

The theory for the current-sensing electromagnetic flowmeter was newly developed. The current-sensing flowmeter has a high temporal resolution so that it can be applied to measure the flows with fast transients like two-phase flow. The signal prediction and the calibration of the current-sensing flowmeter in two-phase flow and the measuring of the characteristics of two-phase flow are the major concerns.
To do this, using a finite difference method, the three-dimensional virtual potential distributions for the electrodes of finite size were calculated for single-phase flow, annular flow and slug flow. With the gradient of the virtual potential, the rectilinear weight functions for the single-phase flow and the annular flow which were the main parameter for the conventional voltage-sensing flowmeter were deduced and compared with existing analytic solutions for the point-electrode. There was a reasonable correspondence between the present and existing results. Particularly the axial weight function and the radial weight function for single-phase flow, annular flow and slug flow were newly defined and computed by taking the gradient of the virtual potential.
The flow pattern coefficient f was introduced to simplify the calibration process for two-phase flow. It was calculated from the solved virtual potential distributions of single-phase and two-phase flow. For annular flow, the coefficient was well-fitted with two decaying exponential functions of the normalized film thickness δ*, f = 1 + 1.3 exp(-7 δ*) + 7.3 exp(-31 δ*). For the slug flow, it was provided as a function of the normalized film thickness δ* and the normalized position of a slug bubble tail L*. The coefficient by the numerical simulation was compared with experimental results obtained by Frequency Response Analyzer (FRA) and Potentiostat/Galvanostat. The comparison clearly showed the agreement between the numerical and the experimental results.