Electromagnetic flowmeters employed for industrial purposes generally achieve their good steady-state performance and base-line stability by employing low-frequency non-sinusoidal excitation of the magnetic field. This limits the dynamic performance achievable from such flowmeters since the measurement is being made by a magnetic field which samples the original flow signal and thus suffers the limitations of the Shannon sampling theorem. Output signal processing and filtering of the flow generated signal may further reduce the dynamic performance of the flowmeter and thus rapid changes or high frequency pulsations in the flow are not faithfully represented on the output of the flowmeter. This paper examines a model of the electromagnetic flowmeter operating under time varying flow conditions; why limitations are placed on the frequency of the magnetic field; the implications of these limitations in terms of the errors which may occur under dynamic flow conditions and how these limitations can be overcome. Soon after Michael Faraday proposed the use of the electromagnetic principle to measure the flow of conducting liquids it was recognised that it was necessary to provide a time varying magnetic field, rather than a dc field to be able to distinguish between flow and electrochemically generated signals on the electrodes. Figure 1 shows in schematic form the typical elements of an industrial electromagnetic flowmeter. The flowing liquid is contained within a non-magnetic pipe having an insulating liner. The coils provide the time varying magnetic field and the electrodes pick-up a time varying signal whose amplitude provides an encoded measurement of the flowrate. The signal processing amplifies the signals picked-up on the electrodes and processes the signal to extract the flow signal from the amplified output in the presence of unwanted signals such as rate induced signals, line frequency interference and amplifier and flow induced noise.
展开▼
