数学与统计学院学术报告 007 号
报告题目: A Highly Scalable Newton-Krylov-Schwarz Solver for Massively Parallel Cardiac Mechanics Simulations
报告人： 蒋毅 助理教授 （中国科学院深圳先进技术研究院）
报告内容：Cardiac dynamics relate to several fundamental physical natures across a broad range of temporal and spatial scales, among which the continuum elastics is crucial. In this work, we consider the numerical simulation of the left ventricle of the human heart by a hyper-elastic fiber reinforced transversely isotropic model, and propose a fully implicit overlapping domain decomposition method on unstructured meshes aiming at efficiently solving the discretized systems on large scale supercomputers with a large number of processor cores. The algorithm is constructed within the framework of Newton-Krylov methods with an analytically computed Jacobian, and the Krylov subspace solver is preconditioned with a restricted additive Schwarz method. We show numerically that the algorithm is highly parallel and robust with respect to the material coefficients, the large deformation, and the fiber reinforcement. The proposed approach is first verified on a benchmark problem and then applied to a patient-specific left ventricle. Numerical experiments show that the algorithm scales well on a supercomputer with more than 8,000 processor cores for a mesh with $0.2$ mm resolution.