GPU-Accelerated Lagrangian Advection for Viscous Incompressible Fluids on Adaptive Grids

This paper presents a novel approach to simulating two-dimensional viscous incompressible fluids using a GPU-accelerated Lagrangian advection method. The simulation is based on the Navier-Stokes equations, solved on an adaptive grid to optimize computational efficiency. A key feature of this implementation is the adaptive grid, which refines the simulation in areas of high fluid dynamic activity, allowing for detailed simulation where it matters most. The method’s effectiveness is demonstrated through simulations showcasing fluid interactions with static objects. Performance is evaluated in terms of computational speed and visual fidelity, with a focus on real-time applications in interactive environments. Results highlight the benefits of adaptive grid refinement in reducing computational cost while maintaining visual realism.