Experimental investigation of the influence of external vibrations on Coriolis mass flow meters

The Coriolis flow meter is basically a vibrating tube device; it is therefore potentially susceptible to disruption by external vibrations transmitted from the environment in which the meter is mounted. The paper reports the findings from a carefully structured experimental study in which the results of both analysis and numerical simulation studies were used to guide the choice of vibration frequencies and the directions and the spatial distributions of the vibrations. The total of eight different meters from five different manufacturers covered a wide range of meter geometries and drive frequencies. In addition to comparisons of the flow rates indicated by the meters with independent measures of the flow rate, all the tests involved the recording of raw signals from the displacement sensors so that the effects of using different techniques to extract the phase relationship between these signals could be investigated. All the tests were performed using cold (room temperature) water as the working fluid. The results of the study show that vibrations at the meter drive frequency caused errors in all meters. Vibrations at other frequencies also caused errors in several meters but these errors appear to be due to the algorithm (implemented in the meter electronics) used to extract the phase difference (the measurand) between the sensor signals. However, the complete study suggests that, by suitable choices of meter mechanical design and of the algorithm used to determine the phase difference, it is possible to make a meter which is unaffected by vibrations at any frequency other than the meter drive frequency (provided only that the meter tube motion produced by the vibration is smaller than that produced by the meter drive). For vibrations at the drive frequency the results show that (in general agreement with the analytical and numerical studies) the magnitude of the error depends on the phase relationship between the imposed vibration and the meter drive. Errors also depend on the spatial distribution of the vibration (e.g. the error is different for the same amplitude of vibration applied uniformly to a meter and applied to one end only of the meter). Methods for reducing drive frequency errors are discussed but it is concluded that it may not be possible to eliminate these errors completely.