Scholarlink

<p>Fracture networks are systems of complex mechanical discontinuities that significantly impact the physical behavior of rock masses. In two dimensions, these networks can be approximated as collections of straight-line segments using the mathematical framework of marked point processes. Unlike most approaches that model fractures as independent entities, a few models that include interactions between fractures have been proposed, but parameter inference has remained elusive. This paper proposes a new model that captures essential aspects of fracture network geometry and organization using pairwise geometric interactions between fractures, together with an approximate Bayesian methodology for estimating model parameters. The approach is demonstrated through the estimation of model parameters from a specific fracture network observed in the Oman Mountains. The development of this model, combined with the parameter inference, paves the way for predictive stochastic simulations of fracture networks.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Fracture Network Characterization Using Gibbs Marked Point Processes and Bayesian Inference

  • 期刊论文

  • 发表日期：

    2026年5月28日

查看全文
  • François Bonneau,
  • Radu Stefan Stoica,
  • Guillaume Caumon

摘要

Fracture networks are systems of complex mechanical discontinuities that significantly impact the physical behavior of rock masses. In two dimensions, these networks can be approximated as collections of straight-line segments using the mathematical framework of marked point processes. Unlike most approaches that model fractures as independent entities, a few models that include interactions between fractures have been proposed, but parameter inference has remained elusive. This paper proposes a new model that captures essential aspects of fracture network geometry and organization using pairwise geometric interactions between fractures, together with an approximate Bayesian methodology for estimating model parameters. The approach is demonstrated through the estimation of model parameters from a specific fracture network observed in the Oman Mountains. The development of this model, combined with the parameter inference, paves the way for predictive stochastic simulations of fracture networks.