Open Access

ARTICLE

Improved State-Based Peridynamic Lattice Model Including Elasticity, Plasticity and Damage

Siavash Nikravesh^{1, *}, Walter Gerstle¹

1 Department of Civil, Construction and Environmental Engineering, University of New Mexico, Albuquerque, NM, USA.

* Corresponding Author:Siavash Nikravesh. Email:

Computer Modeling in Engineering & Sciences 2018, 116(3), 323-347. https://doi.org/10.31614/cmes.2018.04099

Abstract

In this study, a recently developed peridynamic lattice model called the “State-based Peridynamic Lattice Model” (SPLM) is improved and demonstrated. In the SPLM, rather than as a continuum, solids are simulated using a close-packed lattice of peridynamically interacting particles. The new SPLM approach advances the SPLM model by improving the damage and plasticity models. Elasticity, plasticity and damage are coupled in this approach. A robust method for damage initiation is developed. A new damage model called the “two-spring damage model” allows damage to localize to a single lattice particle, thus allowing highly localized damage (cracks) to emerge in a realistic manner. A plasticity model that includes hardening, softening, and damage due to plasticity is proposed and demonstrated. Peridynamic boundary effects are modeled efficiently and reasonably. The improved SPLM method is then employed to simulate three common concrete laboratory tests: Uniaxial tension, uniaxial compression, and the Brazilian split cylinder test. The SPLM results are then compared with results from the earlier SPLM model, with simplified classical predictions, and with laboratory results. By solving the same benchmark problems using various lattice rotations and lattice spacings, the approach is demonstrated to be sufficiently objective to be a useful engineering tool to predict the essentially random behavior of concrete laboratory specimens. The improved SPLM demonstrates significant improvements over the previously published version and is found to simulate concrete structures accurately and efficiently using far less computational effort than comparable computational simulation methods.

---

Keywords

---