Article by Oldric Svec, Jan Skocek, Henrik Stang, John Forbes Olesen and Peter Noe Poulsen, Technical University of Denmark.
Abstract
To correctly predict the casting process of a fiber reinforced self compacting concrete on a structural level is a challenging task since the distribution and orientation of fibers influence the global flow pattern and vice versa. In this contribution, a modeling approach capable to represent accurately the most important phenomena is introduced. A conventional Lattice Boltzmann method has been chosen as a fluid dynamics solver of the non-Newtonian fluid. A Mass Tracking Algorithm has been implemented to correctly represent a free surface and a modified Immersed Boundary Method (IBM) with direct forcing is used to explicitly represent individual fibers in the fluid. A novel variable time sub-stepping algorithm for dynamics of immersed rigid particles ensuring stability of simulations has been developed which, together with the IBM, provides an efficient, yet accurate way to simulate flow of suspensions.
In the following, the developed model is used to simulate flow of the fiber reinforced self compacting concrete. Fibers are modeled as slender rigidcylinders using thepreviouslydeveloped correction for particlesofa sub-grid size. Alubrication force correction term and collisions with Coulomb friction between fibers and between fibers and boundary conditions are introduced. Several simulation scenarios are presented and compared to experimental data to validate the proposed approach. Further an extension of the model to the structural scale is discussed and demonstrated on an example modeling the final dispersion and orientation of fibers during a real casting process.
Keywords: fluid mechanics, concrete, multiscale problem, numerical analysis, structural mechanics.
Warsaw_2011_DTU (662 KB)