Examining Computational Fluid Dynamics with Electrical Capacitance Tomography
The University of Cambridge and the University of Canterbury have used ITS’ Research & Education tomography packages for the purpose of this study. Utilising Electrical Capacitance Tomography, the two universities collaborated to explore the accuracy of data provided by computational fluid dynamics when examining fluidised bed applications.
Computational Fluid Dynamics (CFD) is making huge strides in the design and optimization of fluidised beds, as virtual experiments facilitate better scale-up, increase accuracy and minimise risk. The most commonly used CFD model for fluidised bed applications is the Two Fluid Method (TFM). Only TFM can be used to simulate industrial scale fluidised beds in a reasonable simulation time, so it is widely considered the most promising approach for aiding the design and operation of fluidised beds.
TFM however, requires the use of closure models to describe collisions in the particle phase and the interaction of the fluid. In order to use TFM to predict the performance of industrial reactors it is essential that these closure laws are tested and authenticated. To do this, the researchers decided to collect readings from two tangible fluidised beds (one circulating and one dense) and then compare the data collected to that of TFM simulations. To thoroughly test the validity of TFM readings of fluidised beds, the researchers devised two different methods of TFM testing: the modified Gibilaro model and a model based on the Energy Minimisation Multiscale (EMM) criterion.
This study successfully utilised ITS’ very own M3c electrical capacitance tomography system to provide precise capacitance measurements on a circulating fluid bed; which in turn allowed the researchers to contrast the results with the measurements from the two simulated models.
The measurements gathered reported that the revised Gibilaro model can be used to predict the voidage and particle velocity distribution in the dense fluidised bed with a high degree of accuracy, however when used to simulate the circulating fluidised bed it underestimates the solids circulation rate by approximately a factor of 4. The EMM drag model is able to predict the solids circulation rate in the circulating fluidised bed to within about 50%, however it does not predict the correct voidage distribution in the dense fluidised bed.
In summary, this study provided evidence that neither TFM model could accurately simulate the entire range of fluidisation conditions, both providing measurements far below the standard set by the m3c system. However, when persisting with TFM, and computational fluid dynamics, the results indicated that the riser should be simulated using the EMM model, conversely the U-valve and possibly the base of the riser may be more accurately described using a model such as the revised Gibilaro model.
If you’re interested in the applications of Computational Fluid Dynamics, then request more information today! You can reach us through our website, via email – firstname.lastname@example.org or by giving us a call at +44 (0) 161 832 9297.