Abstract: A method of incorporating effects of solid dynamic objects into each discrete time step of a fluid simulation in a high quality fashion on Cartesian grids. The method relies on modifying the fluid velocity field within regions internal to the solid objects prior to the calculation of advection and pressure effects such that the sub voxel details of the solid objects are accurately represented. The modification of the velocities is based on allowing maximal freedom of fluid movement under the constraints the occlusions impose. Additionally, the solid objects are optionally represented in a unified level set fashion such that the computations required for modifying the fluid velocity are efficient. The overall result of this improved method rivals the quality of that achieved with the more complex curvilinear grid approach.

Abstract: An explicit definition of a geometric object is converted into a level set which implicitly defines the geometric object. The explicit definition is converted to a form for efficient ray casting. Magnitude and sign values are assigned to a grid of voxels using ray casting in which the magnitude value represent a minimum distance from the voxel to the object and the sign value defines whether the voxel is inside or outside the object. The magnitude and sign values may be removed of discontinuities and smoothed before an accurate level set implicitly defining the object is achieved. Run length encoding may be applied to define a sparse level set in which consecutive voxels having identical values are compressed into a single value including a marker to indicate a quantity of the consecutive voxels.

Abstract: A method of incorporating effects of solid dynamic objects into each discrete time step of a fluid simulation in a high quality fashion on Cartesian grids. The method relies on modifying the fluid velocity field within regions internal to the solid objects prior to the calculation of advection and pressure effects such that the sub voxel details of the solid objects are accurately represented. The modification of the velocities is based on allowing maximal freedom of fluid movement under the constraints the occlusions impose. Additionally, the solid objects are optionally represented in a unified level set fashion such that the computations required for modifying the fluid velocity are efficient. The overall result of this improved method rivals the quality of that achieved with the more complex curvilinear grid approach.

Abstract: An explicit definition of a geometric object is converted into a level set which implicitly defines the geometric object. The explicit definition is converted to a form for efficient ray casting. Magnitude and sign values are assigned to a grid of voxels using ray casting in which the magnitude value represent a minimum distance from the voxel to the object and the sign value defines whether the voxel is inside or outside the object. The magnitude and sign values may be removed of discontinuities and smoothed before an accurate level set implicitly defining the object is achieved. Run length encoding may be applied to define a sparse level set in which consecutive voxels having identical values are compressed into a single value including a marker to indicate a quantity of the consecutive voxels.