Lattice Boltzmann simulation and direct observation of pore fluid flow around irregularly shaped grains

An accurate velocity model of porous flow plays an important role in the prediction of the ground water pollution.To clarify the behavior of porous flow passing through irregularly shaped grains, we have been performed CFD simulation and direct observation based on the Lattice Boltzmann Method and the LAT-PIV visuallization technique respectively. The Lattice Boltzmann simulator, which works on a graphics processing unit(GPU), is employed to evaluate the pore fluid velocity distribution in an accurate three dimensional digital model involving Toyoura sand. From the simulation results, the pore fluid velocity distributions converge into a unique non-gaussian distribution under various Reynolds numbers ranging from 2 to 10. The features of the non-gaussian distribution are summarized as follows: (1)It has a long tail until sextuple of the mean velocity magnitude. (2)It has a peak frequency close to zero velocity magnitude. (3)It slightly contains negative velocities. The LAT-PIV visuallization technique, a kind of laser slicing visualization method combined LAT(Laser-Aided Tomography) and PIV (Particle Image Velocimetry), visualizes both grain edges and pore fluid behavior inside specimen which is composed of crashed glass grains and specially blended silicone oil. The pore fluid velocity distributions captured by the LAT-PIV images indicate a similar tendency compared with those measured by the LBM simulations. This supports that the LBM simulation has sufficient ability to predict the pore fluid flow even if the porous medium is composed of irregularly shaped grains. GPU accelerated LBM simulation for Toyoura sand model Pore fluid velocity magnitude distributions for several Reynolds numbers