On-site boundary conditions are often desired for lattice Boltzmann simulations of fluid flow in complex geometries such as porous media or microfluidic devices. The possibility to specify the exact position of the boundary, independent of other simu
lation parameters, simplifies the analysis of the system. For practical applications it should allow to freely specify the direction of the flux, and it should be straight forward to implement in three dimensions. Furthermore, especially for parallelized solvers it is of great advantage if the boundary condition can be applied locally, involving only information available on the current lattice site. We meet this need by describing in detail how to transfer the approach suggested by Zou and He to a D3Q19 lattice. The boundary condition acts locally, is independent of the details of the relaxation process during collision and contains no artificial slip. In particular, the case of an on-site no-slip boundary condition is naturally included. We test the boundary condition in several setups and confirm that it is capable to accurately model the velocity field up to second order and does not contain any numerical slip.
We simulate cluster formation of model colloidal particles interacting via DLVO (Derjaguin, Landau, Vervey, Overbeek) potentials. The interaction potentials can be related to experimental conditions, defined by the pH-value, the salt concentration an
d the volume fraction of solid particles suspended in water. The system shows different structural properties for different conditions, including cluster formation, a glass-like repulsive structure, or a liquid suspension. Since many simulations are needed to explore the whole parameter space, when investigating the properties of the suspension depending on the experimental conditions, we have developed a steering approach to control a running simulation and to detect interesting transitions from one region in the configuration space to another. The advantages of the steering approach and the restrictions of its applicability due to physical constraints are illustrated by several example cases.
