Accurate average flow velocity determination is essential for flow measurement in many industries, including automotive, chemical, and oil and gas. The ultrasonic transit-time method is common for average flow velocity measurement, but current limitations restrict measurement accuracy, including fluid dynamic effects from unavoidable phenomena such as turbulence, swirls or vortices, and systematic flow meter errors in calibration or configuration. A new spatial averaging method is proposed, based on flexural ultrasonic array transducer technology, to improve measurement accuracy and reduce the uncertainty of the measurement results. A novel two-dimensional flexural ultrasonic array transducer is developed to validate this measurement method, comprising eight individual elements, each forming distinct paths to a single ultrasonic transducer. These paths are distributed in two chordal planes, symmetric and adjacent to a diametral plane. It is demonstrated that the root-mean-square deviation of the average flow velocity, computed using the spatial averaging method with the array transducer is 2.94%, which is lower compared to that of the individual paths ranging from 3.65% to 8.87% with an average of 6.90%. This is advantageous for improving the accuracy and reducing the uncertainty of classical single-path ultrasonic flow meters, and also for conventional multi-path ultrasonic flow meters through the measurement via each flow plane with reduced uncertainty. This research will drive new developments in ultrasonic flow measurement in a wide range of industrial applications.

No abstract

The calculation of the averaged flow velocity along an ultrasonic path is the core step in ultrasonic transit-time flow measurement. The conventional model for calculating the pathaveraged velocity does not consider the influence of the flow velocity on the propagation direction of the ultrasonic wave and can introduce error when the sound speed is not much greater than the flow velocity. To solve this problem, a new mathematical model covering the influence of the flow velocity is proposed. It has been found that the same mathematical expressions of the pathaveraged flow velocity, as a function of the absolute time-of-flight (ToFs) of ultrasonic waves travelling upstream and downstream, can be derived based on either of the models. However, the expressions as a function of the time difference (the relative ToF) between the ultrasonic waves travelling upstream and downstream derived by the two models are completely different. Flow tests are conducted in a calibrated flow rig utilising air as flowing medium. Experimental results demonstrate that the pathaveraged flow velocities, calculated using either the relative or the absolute ToFs based on the new model, are much more consistent and stable, whereas those calculated based on the conventional model have shown evident and increasing discrepancy when the flow velocity exceeds 15 m/s. When the flow velocity is around 39.45 m/s, the discrepancy is as high as 0.38 m/s. As the relative ToF can be more accurately, reliably and conveniently measured in real applications, the proposed mathematical model has a great potential for the increase of the accuracy of the ultrasonic transittime flowmeters, especially for the applications such as the measurement of fluids with high flow velocities.

Precise control of ultrasonic (US) power is required during sonoporation to ensure pressure amplitude in the target tissue is maintained within the bounds of therapeutic efficiency and below the maximum threshold of safe use. Pressure mapping with a scanning tank and needle hydrophone (NH) represents a convenient means to evaluate US beam profiles and peak negative pressures (PNP) achieved by therapeutic arrays in the medium. However, the method is restricted to fluid media only, and lacks the dynamic capabilities required during the medical procedure. Software modelling can address these issues for dynamic determination of the optimum array driving parameters in relation to therapeutic requirements. This paper describes the correlation between measurements obtained with a scanning tank and the simulation results for four different experimental 1-D phased arrays. Two simulation frameworks were developed using OnScale (Onscale, CA, US) and cross-compared, with the main difference between them being the solver used for modelling US propagation in the medium. The first method used a finite element analysis (FEA) approach for the water and the second method, previously described in [1], relied on Kirchhoff time-domain extrapolation. Electrical cabling was first omitted then included in the model and the outputs were compared with the previous data. Results show that both frameworks led to a similar fit in PNP amplitude with the measurements and that the beam shape strongly agrees with the experimental data in both cases. However, the extrapolation-based method was less computationally demanding than FEA and allowed for modelling larger loads within available computing resources i.e. memory.