Investigating the Ultrasonic Flowmeter Error in Conditions of Distorted Flow Using Multipeaks Salami Functions

Abstract

Paper is devoted to the problem of improving the accuracy of ultrasonic flowmeters installed in conditions that differ from their calibration conditions. The authors highlighted high sensitivity of ultrasonic flowmeters to distortions of flow structure and large additional error of flowmeters caused by flow distortions. The computer simulations using empirically-derived functions of distorted flow velocity of professor Salami are used to investigate the additional error of flowmeters. The analytical formulas of two- and multi-peaks Salami functions are used in this work to reproduce the distorted flow velocity profile. The authors proposed to determine the calibration factor of the ultrasonic flowmeter using the undistorted component of Salami functions. The values of the calibration factor for the number of acoustic channels from 1 to 6 were calculated using four numerical integration methods: Chebyshev (equidistant location of acoustic paths), Gauss, Gauss-Jacobi, OWICS met­hod. This made it possible to realize the flowrate equation for multi-path ultrasonic flowmeters and to determine their additional error for different location of the acoustic paths. The average flow velocity along each path is calculated based on the flow velocity profile in the pipe cross section. Six two- and multi-peaks Salami functions of velocity are used to calculate the velocity profile of the distorted flow caused by typical local resistances. According to the research results the recommendations were developed for choosing the number of the acoustic paths of the ultrasonic flowmeters and for using the methods for determining the location coordinates of the acoustic paths. We determined the required minimum number of chordal acoustic channels of the ultrasonic flowmeter, which makes it possible to reduce the investigated error to defined limits, even in the presence of complex flow distortions reproduced by Salami multi-peaks functions.