Measurement of a Media Density in a Jig Using Radiometric Density Meter with a Counter of Pulses

Abstract

Radiometric density meters are used to monitor many different technological processes. Most of them use gamma-ray absorption. Basic element of the meter is a measuring head that consists of a radiation source (137Cs) and a detector usually in the form of a scintillation counter. The output signal from the detector is the sequence of pulses which is always a stochastic signal with Poisson distribution, regardless of the character of the input signal. The series of pulses are often counted during the time ts in a counter. In that situation the counter should be considered as a kind of a low-pass digital filter whose parameter is a time of measurement ts. The longer the time of measurement ts, the higher the accuracy of the monitor in steady state. When density varies, the dynamic error of measurement increases with the excessive lengthen of measurement time ts. The main filtration problem is a selection of value of the measurement time to minimise the dynamic error of measurement according to accepted criterion. The measurement time can be determined experimentally by simulation when the shape of density changes is known. In that situation, the known shape of density should be treated as a reference signal. In case of coal separation process in a jig the radiometric densitometer can be used to stabilize the separation density and the shape of density dynamic changes. The paper presents the method of selection of the measurement time using the computed changes density of three following cycles in the jig. Problems, advantages and disadvantages associated with using only the counter for digital signal processing from the detector of radiometric densitometer are discussed. Model of the density changes, during three cycles of the separation process in a jig, was identified based on results of industrial tests and its equation is given in the paper. Selection of the counter parameter, that is the measurement time, was done by the simulation minimizing the value of accepted criterion. Simulation results were tabulated and presented in the graphic form.