Fracture analysis employing equivalent domain integral method and nodal integration techniques based on reproducing kernel particle method

Satoyuki Tanaka, Akihiro Takata, Ming Jyun Dai, Hanlin Wang, Shota Sadamoto

Research output: Contribution to journalArticlepeer-review

7 Citations (Scopus)

Abstract

A novel technique to evaluate fracture mechanics parameters is investigated employing the equivalent domain integral (EDI) method and nodal integration (NI) techniques. Galerkin-based meshfree method is adopted. Reproducing kernel (RK) is chosen for the meshfree interpolant. Stabilized conforming nodal integration (SCNI) and sub-domain stabilized conforming integration (SSCI) are adopted for numerically integrating the stiffness matrix. Voronoi diagram is employed to compute volume of each NI domain. The EDI method is addressed to evaluate the fracture mechanics parameters, i.e., energy release rate and stress intensity factors (SIFs). Because the displacement and its derivatives are computed based on SCNI/SSCI, the EDI can be discretized by summing up the physical quantities and volume of each cell/sub-cell. No special quadrature rule is required. To separate the energy release rate into the mixed-mode SIFs, interaction integral method is chosen. Efficient and accurate fracture parameter computation is achieved. Some numerical examples are demonstrated for mixed-mode fracture parameter evaluation and crack propagation analysis. Accuracy and effectiveness of the presented approach are studied.

Original languageEnglish
Pages (from-to)1265-1278
Number of pages14
JournalComputational Particle Mechanics
Volume9
Issue number6
DOIs
Publication statusPublished - 2022 Nov

All Science Journal Classification (ASJC) codes

  • Computational Mechanics
  • Civil and Structural Engineering
  • Numerical Analysis
  • Modelling and Simulation
  • Fluid Flow and Transfer Processes
  • Computational Mathematics

Fingerprint

Dive into the research topics of 'Fracture analysis employing equivalent domain integral method and nodal integration techniques based on reproducing kernel particle method'. Together they form a unique fingerprint.

Cite this