Scholarlink

<p>Predicting the failure behavior of brittle rocks under overload is crucial for evaluating structural stability in engineering applications. Material heterogeneity and hidden flaws make the initiation and growth of cracks highly uncertain, so purely stress-driven models often produce very different macroscopic failure modes. In this work, we developed an acoustic emission (AE) data-informed numerical approach that narrows the range of possible fracture paths and improves the robustness of failure process simulation. Rock heterogeneity was represented by a Weibull distribution that had been calibrated against uniaxial compression tests, while AE information recorded from 10 to 90% of peak stress were gradually introduced into the evolving model. Over one thousand simulations were carried out to explore the statistical variability of fracture development. Without AE input, cracks occurred at the same positions in only about 16% of the realizations, leading to widely scattered fracture patterns. With adequate AE input, however, certain cracks reappeared in more than 90% of runs and controlled the overall pattern convergence. A laboratory comparison indicated that incorporating AE raised the agreement of predicted fracture geometry with actual tests to nearly 89%. Large AE events influenced only a millimeter-scale region, whereas the broader damage evolution was governed by the accumulation and interaction of many moderate events. Stress-driven damage dominates at the early stage, whereas AE-informed damage progressively constrains fracture evolution as critical cracks emerge. These results show how AE information can link laboratory observations with numerical modeling and help explain the shift from random microcracking to repeatable macroscopic failure.</p><p>Highlights:<UnorderedList Mark="Bullet"> <ItemContent> <p>An AE data-informed model for rock failure simulation is developed.</p> </ItemContent> <ItemContent> <p>Identified critical microcracks as AE data sufficiency indicators.</p> </ItemContent> <ItemContent> <p>Enhanced reliability of damage evolution prediction in heterogeneous rocks.</p> </ItemContent> <ItemContent> <p>Reduced fracture randomness and improved failure prediction accuracy.</p> </ItemContent> </UnorderedList></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

Failure Process Simulation of Rock Using an Acoustic Emission-Informed Model

  • 期刊论文

  • 发表日期：

    2025年11月6日

查看全文
  • Jingren Zhou,
  • Jiong Wei,
  • Xiang Lu,
  • Kai Guan,
  • Jiankang Chen,
  • Huajin Li

摘要

Predicting the failure behavior of brittle rocks under overload is crucial for evaluating structural stability in engineering applications. Material heterogeneity and hidden flaws make the initiation and growth of cracks highly uncertain, so purely stress-driven models often produce very different macroscopic failure modes. In this work, we developed an acoustic emission (AE) data-informed numerical approach that narrows the range of possible fracture paths and improves the robustness of failure process simulation. Rock heterogeneity was represented by a Weibull distribution that had been calibrated against uniaxial compression tests, while AE information recorded from 10 to 90% of peak stress were gradually introduced into the evolving model. Over one thousand simulations were carried out to explore the statistical variability of fracture development. Without AE input, cracks occurred at the same positions in only about 16% of the realizations, leading to widely scattered fracture patterns. With adequate AE input, however, certain cracks reappeared in more than 90% of runs and controlled the overall pattern convergence. A laboratory comparison indicated that incorporating AE raised the agreement of predicted fracture geometry with actual tests to nearly 89%. Large AE events influenced only a millimeter-scale region, whereas the broader damage evolution was governed by the accumulation and interaction of many moderate events. Stress-driven damage dominates at the early stage, whereas AE-informed damage progressively constrains fracture evolution as critical cracks emerge. These results show how AE information can link laboratory observations with numerical modeling and help explain the shift from random microcracking to repeatable macroscopic failure.

Highlights:

An AE data-informed model for rock failure simulation is developed.

Identified critical microcracks as AE data sufficiency indicators.

Enhanced reliability of damage evolution prediction in heterogeneous rocks.

Reduced fracture randomness and improved failure prediction accuracy.