(186i) New Method to Compute Local Fluxes and Stresses at the Bubble Surfaces in Multiphase Flow Simulation

When present, bubbling determines many important features of fluidized bed reactors, including reaction, rates of heat and mass transfer, gas-solid mixing, solids elutriation, wall erosion, etc. A detailed description of bubbles can be found by CFD, which can provide insight useful for scale-up, design, or process optimization for reliable commercial plants. Solution to the Navier Stokes equations by way of two-fluid method (TFM) provides discrete solutions for velocity and pressure throughout the computational domain. To be useful to engineers, the data must be simplified and described in macroscopic terms, for example, bubbles, etc. Thus, obtaining more accurate bubble properties from post-processing, both numerical and experimental data, is crucial for minimal uncertainties in later scale-up. The aim of the present work is to develop an algorithm useful for the analysis of bubbling fluidization. First, a comparative analysis is performed between the current practice of data post-processing ‘binarization’ and a newly developed ‘contouring’ method by Uddin et al.,¹ to obtain bubble properties from CFD simulation. Commonly preferred bubble shapes with analytically known areas are discretized to compute areas numerically applying binarization and contouring methods. Predefined, well-characterized shapes include a composite formed by a circle and a line, a semi-circle, a semi-ellipse, and a composite formed by discounting semi-ellipse from semi-circle. Error analysis is conducted for various sizes of those bubble shapes and for different discretization schemes. Second, both methods are applied to a previously validated CFD simulation¹ data to calculate the bubble sizes. The average bubble size is then compared with experimental data² and semi-empirical correlations³. Finally, an algorithm is developed to compute normal vectors on the bubble surfaces which allows for calculation of fluxes, local stresses as well as local geometric characteristics such as curvature. Even though this study is focused on evaluating bubble dynamics in gas-solid fluidized beds, the algorithm itself can be easily applied and extended for detecting bubbles, drops and clusters in other areas of multiphase flows for industrial applications.

 

References

1. Uddin MH, Khan MAH, Coronella CJ. 3-D face-masking detection and tracking algorithm for bubble dynamics: Method and validation for gas–solid fluidized beds. Powder Technology. 2017;313:88-98.

2. Velarde IC, Gallucci F, van Sint Annaland M. Development of an endoscopic-laser PIV/DIA technique for high-temperature gas–solid fluidized beds. Chemical Engineering Science. 2016;143:351-363.

3. Darton R. A bubble growth theory of fluidized bed reactors. Trans IChemE. 1979;57:134.