A finite-strain viscoelastic-damage numerical model for time-dependent failure and instability of rocks

Cheng, Xiaofeng; Tang, Chun’an; Zhuang, Duanyang

doi:10.1016/j.compgeo.2021.104596

Cited by 12 publications

(2 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Considering that rocks exhibit a pronounced time-dependent mechanical behaviour (e.g., the deformation and failure of rocks evolve gradually over time instead of occurring instantaneously), which is supported by evidence from several field studies [182][183][184], the rheology of rocks has become one of the essential topics in geomechanics. Recently, within the framework of irreversible thermodynamics, a finite-strain viscoelastic-damage model was developed by Cheng et al [185] using FEM to investigate the long-term strength and timedependent failure and instability of rocks. Within the model, the heterogeneity of materials was considered on the mesoscale, and a continuum damage constitutive relationship was employed to characterize the material damage process.…”

Section: Failure and Instability Analysis Of Viscoelastic Rocksmentioning

confidence: 99%

A Review on Modelling of Viscoelastic Contact Problems

et al. 2022

View full text Add to dashboard Cite

Approaches to solving viscoelastic problems have received extensive attention in recent decades as viscoelastic materials have been widely applied in various fields. An overview of relevant modelling approaches is provided in the paper. The review starts with a brief introduction of some basic terminologies and theories that are commonly used to describe the contact behaviour of viscoelastic materials. By building up the complexity of contact problems, including dry contact, lubricated contact, thermoviscoelastic contact and non-linear viscoelastic contact, tentative analytical solutions are first introduced as essential milestones. Afterwards, a series of numerical models for the various types of contact problems with and without surface roughness are presented and discussed. Examples, in which computational tools were employed to assist the analysis of viscoelastic components in different fields, are given as case studies to demonstrate that a comprehensive numerical framework is currently being developed to address complex viscoelastic contact problems that are prevalent in real life.

show abstract

Section: Failure and Instability Analysis Of Viscoelastic Rocksmentioning

confidence: 99%

A Review on Modelling of Viscoelastic Contact Problems

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Several works study the effects of material heterogeneity and its relation with strain localization [ 65 , 66 , 67 ]. The introduction of material defects and material heterogeneity in the problem description or numerical discretization are common in geomechanics [ 68 , 69 , 70 ], these approaches remain very pragmatic and do not question the nature of the material heterogeneity. Indeed, the micro-macro issue of localization onset and its relation with inhomogeneity are still under intense research [ 71 , 72 , 73 , 74 ].…”

Section: Introductionmentioning

confidence: 99%

Predicting the Non-Deterministic Response of a Micro-Scale Mechanical Model Using Generative Adversarial Networks

Argilaga

Zhuang

2022

Materials

Self Cite

View full text Add to dashboard Cite

Recent improvements in micro-scale material descriptions allow to build increasingly refined multiscale models in geomechanics. This often comes at the expense of computational cost which can eventually become prohibitive. Among other characteristics, the non-determinism of a micro-scale response makes its replacement by a surrogate particularly challenging. Machine Learning (ML) is a promising technique to substitute physics-based models, nevertheless existing ML algorithms for the prediction of material response do not integrate non-determinism in the learning process. Is it possible to use the numerical output of the latest micro-scale descriptions to train a ML algorithm that will then provide a response at a much lower computational cost? A series of ML algorithms with different levels of depth and supervision are trained using a data-driven approach. Gaussian Process Regression (GPR), Self-Organizing Maps (SOM) and Generative Adversarial Networks (GANs) are tested and the latter retained because of its superior results. A modified GANs with lower network depth showed good performance in the generation of failure probability maps, with good reproduction of the non-deterministic micro-scale response. The trained generator can be incorporated into existing multiscale models allowing to, at least partially, bypass the costly micro-scale computations.

show abstract