Mikele Cândida Sousa de Sant'Anna^*, Rodrigo Antônio Pinto de Melo ^2†, Gabriel Francisco da Silva ³⁺ , Ricardo de Andrade Medronho ⁴⁺ , Sergio Lucena ⁵⁺

^*Universidade Federal de Pernambuco
Av. Prof. Moraes Rego, 1235- Cidade Universitária. Recife/PE. CEP 50670-901

e-mail: mikelecandida@gmail.com

^2† Universidade Federal de Pernambuco
Av. Prof. Moraes Rego, 1235- Cidade Universitária. Recife/PE. CEP 50670-901

e-mail: rodrigo.apmelo@gmail.com

³⁺Universidade Federal de Sergipe

Cidade Universitária Prof. José Aloísio de Campos. Av. Marechal Rondon. São Cristóvão/SE. CEP 49100-000

e-mail: gabriel@ufs.br

                                                                     ⁴⁺Universidade Federal do Rio de Janeiro

Av. Athos da Silva ramos, 149. Cidade Universitária, Ilha do Fundão. Rio de Janeiro/RJ CEP 21941-909.

                                                             e-mail: medronho@eq.ufrj.br

ABSTRACT

The gas-solid systems can be presented in different regimes depending on operating parameters such as gas velocity, particle characteristics and bed geometry [1]. Due to various forms of contact, different models are needed to predict the behavior of the bed.

The dynamic behavior of fluidized particulate mixed determines whether or segregate. The bubbles provide an excellent gas-solid contact and high rates of heat and mass transfer.  In order to optimize the processes in recent decades intensified the use of computational techniques such as Computational Fluid Dynamics, CFD (tool for numerical simulation) that solves the equations of conservation through the discretization of the finite volume method.

In the simulations of the computational system package ANSYS FLUENT 15.0 and simulation system involving air and san. 2D geometry was constructed to test the choice of turbulence model, the loop iteration and time were carried out.

The interaction between the phases was determined by the equation Syamlal-O'brien [2] and the value of the coefficient of restitution particle-particle was equal to 0.9. The viscosity of the particles was calculated by Kinetic Theory Granular. The equations comprise problem was soved in a segregated manner using the Phase Coupled SIMPLE method was used spatial discretization of second order for all equations, except for the volume fraction which was discretized using the QUICK method. The time step was fixed on 5.10-5 seconds with 100,000 iterations per time step, giving a value of 5s. The inlet velocity of the air was varied to 0.03, 0.10; 0.38; 0.46 and 0.51 (meters per second), according to [3].

Through the analysis of charts bed expansion and pressure drop, coupled with analysis of the volume fraction and pressure profiles, it was decided to be used in the following simulations the model k-ε because it has  concordant results with less computational effort, compared with the model k-ω.

Aiming to validate simulations of bubbling fluidized bed we used the experimental values collected [3]. For the other parameters studied there were no significant differences, illustrated that the results are agreement with the data compared.

At the beginning of the simulation the bed is still fixed at all speeds presented here. With the evolution of simulation time the bed reaches a bubbling regime and this remains, simulation becomes pseudo-stationary from the time of 2.5s. The regime continues bubbling, the expansion of the bed remains stationary, but the bubbles differ in position, size and shape. Just as in the real experiment. The velocities of 0.03 m/s and 0.10 m/s are not able to promote the bubbling fluidization in the bed. From the speed 0,38m/s becomes bubbling bed.

References

[1] D. Kunii and O. Levenspiel, Ind. Eng. Chem. Process Design Devel. 7, 481 (1968)

[2] M. Syamlal, T. J. O'Brien. Computer simulation of bubbles in a fluidized bed. A.I.Ch.E. Symposium Series 85, 22-31 (1989)

[3] F. Taghipour,  N. Ellis, C. Wong. Experimental andcomputational study of gas–solid fluidized bed hydrodynamics. Chemical Engineering Science.  60, 6857 – 6867. (2005)