No abstract

A new finite element procedure for the solution of the electromagnetic flow meter weight function is presented. The weight function represents the relative contribution of the fluid velocity at a particular spatial location in the flow cross section to the output signal from the flow meter. The solution of the weight function is important for inferring the velocity profile from measured potential differences within the electromagnetic flow meter. First, a numerical simulation model was constructed with COMSOL Mutiphysics [1]. Next, a study was undertaken to compare the numerical simulation model with Shercliff's [2] analytical solution for the weight function. The finite element methodology was found to be correct for solving the weight function. Based on this result, the numerical simulation model was used to compute the weight function for two flow meter geometries. For both cases the electrodes were located on the internal circumference of the flow pipe and the line joining the electrodes was normal to the imposed magnetic field. In the first case the separation of the electrodes was equal to one pipe diameter but in the second case the electrode separation was less than one pipe diameter. The methodology presented in this paper represents the first stage in the development of an image reconstruction technique which could be used to obtain the liquid velocity profile from boundary voltage measurements obtained in flows with highly skewed velocity distributions.

The measurement of the distribution of magnetic flux density in an electromagnetic (EM) flow meter is important in obtaining an accurate flow rate. In this paper, Helmholtz coils are used in the EM flow meter to generate a homogenous magnetic flux density across the cross section of flow pipe. The finite element method analysis software, COMSOL, was used to build a model of the magnetic flux density, which was then verified using experimental testing. A 2D magnetic camera system containing an 8 £ 8 array of 64 Hall sensors is used to detect the magnetic flux density. The magnetic flux density data is then measured using a national instruments PCI 6255 data acquisition card. The results from the simulation model and experimental measurements are presented and compared in the paper. The study is useful in designing the electrode sensor arrays of an EM flow meter. In addition to this, the distribution of magnetic flux density is useful in improving the accurate flow rate measurement of the EM flow meter.