Scholarlink

<p>This study proposes predictive model for the crack depth (CDp) of anisotropic shale from elastic and tensile strength&#xa0;under varying load directions—parallel and perpendicular to bedding planes—within a Freeze–Thaw range of − 20&#xa0;°C to 50&#xa0;°C – a rarity in previous research. In this study, six hundred (600) shale samples were subjected to unconfined compressive strength (UCS) and splitting tensile strength (TS) tests while laser technology was used for crack depth measurement. Experimental results show that the saturated shale (subjected to 12 FT-cycles in the temperature range of − 20&#xa0;°C to 50&#xa0;°C showed a reduction in UCS, TS, small strain elastic modulus (E<sub>o</sub>) and secant elastic modulus (E<sub>50</sub>) by 2.19, 3.11, 5.46 and 2.32-times and increased crack density (CD) and crack depth (CDp) by 2.26 and 1.55 times respectively for load direction perpendicular to the bedding plane. Similarly, for loading direction parallel to the bedding plane, the reduction in UCS, TS, E<sub>o</sub> and E<sub>50</sub> by 2.05, 4.9, 3.55, and 3.85 times and an increase in CD and CDp by 2.52 and 5.57 times, respectively, was observed. Under these conditions, Artificial Neural Network (ANN) models predicted the crack depth of shale rocks. Results for loading parallel (MSE: 82.83, RMSE: 9.1, MAE: 5.01 and R<sup>2</sup>: 0.99938) and perpendicular (MSE: 9.43, RMSE: 3.07, MAE: 1.76 and R<sup>2</sup>: 0.99921) to the bedding planes showed that the ANN performed 64.8% superior in predicting crack depth with load application perpendicular to the bedding plane. This study can be applied for preventive safety assessments and to prioritize repair work in areas susceptible to FT-weathering of shale rock.</p> Graphical abstract <p></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

-20 Celcius to 50 Celcius Freeze–Thaw: Anisotropic Shale Characteristic Evaluation for Crack Depth Prediction with Parallel and Perpendicular Loading Using ANN

  • 期刊论文

  • 发表日期：

    2025年8月27日

查看全文
  • Muhammad Ali,
  • Ayesha Zubair,
  • Zubair Masoud

摘要

This study proposes predictive model for the crack depth (CDp) of anisotropic shale from elastic and tensile strength under varying load directions—parallel and perpendicular to bedding planes—within a Freeze–Thaw range of − 20 °C to 50 °C – a rarity in previous research. In this study, six hundred (600) shale samples were subjected to unconfined compressive strength (UCS) and splitting tensile strength (TS) tests while laser technology was used for crack depth measurement. Experimental results show that the saturated shale (subjected to 12 FT-cycles in the temperature range of − 20 °C to 50 °C showed a reduction in UCS, TS, small strain elastic modulus (Eo) and secant elastic modulus (E50) by 2.19, 3.11, 5.46 and 2.32-times and increased crack density (CD) and crack depth (CDp) by 2.26 and 1.55 times respectively for load direction perpendicular to the bedding plane. Similarly, for loading direction parallel to the bedding plane, the reduction in UCS, TS, Eo and E50 by 2.05, 4.9, 3.55, and 3.85 times and an increase in CD and CDp by 2.52 and 5.57 times, respectively, was observed. Under these conditions, Artificial Neural Network (ANN) models predicted the crack depth of shale rocks. Results for loading parallel (MSE: 82.83, RMSE: 9.1, MAE: 5.01 and R2: 0.99938) and perpendicular (MSE: 9.43, RMSE: 3.07, MAE: 1.76 and R2: 0.99921) to the bedding planes showed that the ANN performed 64.8% superior in predicting crack depth with load application perpendicular to the bedding plane. This study can be applied for preventive safety assessments and to prioritize repair work in areas susceptible to FT-weathering of shale rock.

Graphical abstract