Using GPGPU to Enhance Simulation of the Functionalized Cahn-Hilliard Equation

摘要

The Functionalized Cahn-Hilliard (FCH) Equation describes the dynamics of pore network formation in a functionalized polymer/solvent membrane. During the casting of such membranes, processes occur on very diverse timescales requiring long computationally intensive simulations. Our scheme is designed to be unconditionally gradient stable allowing for adaptivity in time stepping while preserving the discrete energy law. We present a novel iteration for solving the nonlinear equation with a Fourier spectral method, and numerical results are shown with a 100x speedup using a Graphics Processing Unit (GPU).