Um método front-tracking completamente adaptativo para a simulação de escoamentos tridimensionais bifásicos

Abstract

This thesis presents a computational framework for simulating three-dimensional two-phase flows based on adaptive strategies for space and time discretization. The method is based on the fronttracking method of Unverdi and Tryggvason (1992), and the discretization of the Eulerian domain is based on the SAMR strategy of Berger and Colella (1989). The time integration algorithm is based on the IMEX scheme, and the time step is calculated based on CFL criteria. The implementation of the Lagrangian framework relied on the GNU Triangulated Library (GTS), which provides a complete data structure and supporting functions for data access, remeshing tools and data output. The memoryless simplification algorithm of Lindstrom and Turk (1998) is used for surface remeshing, preserving the volume and shape of the interface. Nevertheless, additional tools for volume recovery were implemented, motivated by the non-conservative behaviour of the advection of the Lagrangian interface. This process may also induce some non-physical undulations on the Lagrangian interface, which was circumvented with the implementation of the TSUR-3D algorithm of Sousa et al. (2004). The methodology was applied to a series of rising bubble simulations, in which a single bubble rises in an initially quiescent liquid, and validated against experimental results (BHAGA; WEBER, 1981). The validation process consisted in comparing the terminal shape and Reynolds number, as well as the topology of the streamlines downstream of the bubble. Finally, the algorithm was applied to the simulation of two cases of bubbles rising in the wobbling regime, which is characterized by the continuous change in the bubble shape and complex patterns of vortex shedding. The use of adaptive mesh refinement strategies led to physically insightful results, which would not be possible in a serial code with a uniform mesh.