The analysis and modeling of pressure fluctuations in a fluidized bed

Abstract

The objective of this research was to evaluate several different pressure probe techniques used to measure the pressure fluctuations in a fluidized bed, to determine the effect of bed parameters on the power spectrums from the corresponding pressure fluctuations, and to develop a second-order model that describes the power spectrums of these pressure fluctuations. The motivation for this work is to increase the knowledge of fluidized bed pressure phenomena and consequently further the understanding of such fluidized bed research areas as similitude, chemical processes, and the use of pressure fluctuations as a diagnostic tool. Pressure fluctuations in a fluidized bed were measured with four pressure probe techniques to determine if differences existed amongst them. These techniques included static-absolute, dynamic-absolute, static-differential, and dynamic-differential pressure probe techniques. The pressure fluctuation data were analyzed using spectral analysis techniques and graphed on Bode plots. This testing showed that the absolute and differential pressure probe technique produced drastically different results under certain conditions. Comparison of the dynamic and static techniques showed that the results were very similar in most situations, but that the amplitude of the pressure fluctuations measured with the static technique was lower than that of the dynamic technique in a few situations. The position of the probe arm in the bed was also determined to be an important testing parameter for all techniques as several frequency peaks in the power spectrum were a function of position in the bed. Pressure fluctuation data for a range of bed heights, fluidization velocities, particle sizes, particle densities, and bed temperatures were taken to determine each parameter's effect on the power spectrum. The relationship between the bed height and location of the dominant frequency was shown to agree with the literature, but secondary peaks were shown to be a function of position in the bed (e.g., distance above the distributor) and not affected by increasing the bed height above this position (e.g., by adding sand to the bed). Through the use of the Bode plot, the shifting of frequency peaks with an increase in fluidization velocity was documented to be a continuous growth. Large-diameter, high-density particles