共 50 条
Recovering elastic inclusions by shape optimization methods with immersed finite elements
被引:13
|作者:
Guo, Ruchi
[1
]
Lin, Tao
[1
]
Lin, Yanping
[2
]
机构:
[1] Virginia Tech, Dept Math, Blacksburg, VA 24061 USA
[2] Hong Kong Polytech Univ, Dept Appl Math, Kowloon, Hong Kong, Peoples R China
关键词:
Inverse problems;
Elasticity systems;
Inclusions reconstruction;
Discontinuous Lame parameters;
Shape optimization;
Immersed finite element methods;
INVERSE CONDUCTIVITY PROBLEM;
LEVEL SET METHODS;
ELECTRICAL-IMPEDANCE TOMOGRAPHY;
SENSITIVITY-ANALYSIS;
RECONSTRUCTION;
IDENTIFICATION;
DESIGN;
REGULARIZATION;
COEFFICIENTS;
SPACES;
D O I:
10.1016/j.jcp.2019.109123
中图分类号:
TP39 [计算机的应用];
学科分类号:
081203 ;
0835 ;
摘要:
This article presents a finite element method on a fixed mesh for solving a group of inverse geometric problems for recovering the material interface of a linear elasticity system. A partially penalized immersed finite element method is used to discretize both the elasticity interface problems and the objective shape functionals accurately regardless of the shape and location of the interface. Explicit formulas for both the velocity fields and the shape derivatives of IFE shape functions are derived on a fixed mesh and they are employed in the shape sensitivity framework through the discretized adjoint method for accurately and efficiently computing the gradients of objective shape functions with respect to the parameters of the interface curve. The shape optimization for solving an inverse geometric problem is therefore accurately reduced to a constrained optimization that can be implemented efficiently within the IFE framework together with a standard optimization algorithm. We demonstrate features and advantages of the proposed IFE-based shape optimization method by several typical inverse geometric problems for linear elasticity systems. (C) 2019 Elsevier Inc. All rights reserved.
引用
收藏
页数:24
相关论文