
@Article{cmes.2023.026922,
AUTHOR = {Fei Han, Zhibin Li, Jianyu Zhang, Zhiying Liu, Chen Yao, Wenping Han},
TITLE = {ABAQUS and ANSYS Implementations of the Peridynamics-Based Finite Element Method (PeriFEM) for Brittle Fractures},
JOURNAL = {Computer Modeling in Engineering \& Sciences},
VOLUME = {136},
YEAR = {2023},
NUMBER = {3},
PAGES = {2715--2740},
URL = {http://www.techscience.com/CMES/v136n3/51816},
ISSN = {1526-1506},
ABSTRACT = {In this study, we propose the first unified implementation strategy for peridynamics in commercial finite element
method (FEM) software packages based on their application programming interface using the peridynamics-based
finite element method (PeriFEM). Using ANSYS and ABAQUS as examples, we present the numerical results and
implementation details of PeriFEM in commercial FEM software. PeriFEM is a reformulation of the traditional
FEM for solving peridynamic equations numerically. It is considered that the non-local features of peridynamics
yet possesses the same computational framework as the traditional FEM. Therefore, this implementation benefits
from the consistent computational frameworks of both PeriFEM and the traditional FEM. An implicit algorithm
is used for both ANSYS and ABAQUS; however, different convergence criteria are adopted owing to their unique
features. In ANSYS, APDL enables users to conveniently obtain broken-bond information from UPFs; thus, the
convergence criterion is chosen as no new broken bond. In ABAQUS, obtaining broken-bond information is not
convenient for users; thus, the default convergence criterion is used in ABAQUS. The codes integrated into ANSYS
and ABAQUS are both verified through benchmark examples, and the computational convergence and costs are
compared. The results show that, for some specific examples, ABAQUS is more efficient, whereas the convergence
criterion adopted in ANSYS is more robust. Finally, 3D examples are presented to demonstrate the ability of the
proposed approach to deal with complex engineering problems.},
DOI = {10.32604/cmes.2023.026922}
}



