The influence of a radial velocity profile on the velocity measured by noise analysis

Abstract

For fluid-velocity measurements based on noise-analysis techniques, the result of the measurement is usually only one value which is an averaged value of the more or less developed velocity profile in the flow-area. In what follows, we investigate two problems: 1. (i) What kind of average value we measure, when measuring the fluid velocity in a flow-area with a fully-developed velocity profile. 2. (ii) Is it possible to explain the second transit time, often observed in BWR measurements in the upper part of the core, by assuming a certain velocity profile in the bundles. The investigation is performed by using synthetic signals. These are digitalized analog signals of a noise generator which are delayed numerically according to a given profile, and weighted by arbitrary weighting functions in the time and frequency domain to model the different effects influencing the signals' structure as determined by the detector field of view or the two-phase flow-pattern. The results of these investigations show that: 1. (a) The fluid velocity measured by noise analysis does not agree with any mathematically-defined average value of the profile. The deviation of the measured value from the profile average depends on the ratio between minimum and maximum velocity in the profile and is usually in the order of ± 10% or even more. 2. (b) The influences on the signal structure arising from the detector field of view or from the two-phase flow-pattern, do not seem to have a significant effect on the measured velocity. 3. (c) With some types of profiles, a second or even a third transit time (velocity) appears. The occurrence of sidepeaks in the cross-correlation function depends much more on the shape of the profile than on any other parameter (e.g. detector field of view, two-phase flow-pattern etc.).