A couple months back, I have been asked to solve a macro shot involving a single drop of blood falling on a surface of sand. I have conducted a thorough RnD process, from studying the scale ratios, to the behavior of sand particles and their collisions, shapes and sizes. After generally determining scale and density ratios, I have proceeded to simulation. The project got cancelled in the meantime, but I thought the results were interesting enough to show.
~58K sand particles(counting only the ones that use dynamics, not all particles in the shot)
~100hrs simulation time (not including early tests, trial and error, and stacking simulation time)
-Houdini dynamics used for stacking sand particles(random RBD sand particles emitted, and stacked using Bullet dynamics)
-Final fluid+dynamics simulated in Realflow