Scholarlink

<p>This study introduces a novel methodology for predicting hydraulic conductivity (<i>K</i>) from standard penetration test (SPT) <i>N</i>-values, addressing the critical challenges of conventional field measurements that result in sparse <i>K</i> data. The research objectives were to: (1) establish empirical correlations between <i>N</i> and <i>K</i>, (2) develop a robust prediction model with quantifiable bounds, and (3) demonstrate practical applications for enhanced subsurface characterization. Analysis of 3508 boreholes across South Korea revealed a statistically significant negative correlation between <i>N</i> and <i>K</i> in sandy soils. Quantile regression enabled prediction of both point estimates and percentile ranges. Evaluation of six empirical equations for <i>K</i> estimation identified the Chapuis equation as optimal, which was integrated with field measurements to strengthen the regression model. For weathered rocks, a consistent <i>K</i> range was established. The methodology’s novelty lies in combining readily available SPT data with advanced statistical techniques to generate high-resolution 3D <i>K</i> domains, as demonstrated through kriging. Despite relatively low <i>R</i><sup>2</sup> values, the methodology achieves practical accuracy with most predictions falling within one order of magnitude of measured values. This approach significantly enhances spatial and depth-wise characterization of subsurface <i>K</i>, offering a practical solution for groundwater flow modeling and geotechnical design with improved resolution.</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

Enhanced characterization of hydraulic conductivity via standard penetration test for sandy soils and weathered rocks

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月2日

查看全文
  • Wanhyuk Seo,
  • Eomzi Yang,
  • Tae Sup Yun

摘要

This study introduces a novel methodology for predicting hydraulic conductivity (K) from standard penetration test (SPT) N-values, addressing the critical challenges of conventional field measurements that result in sparse K data. The research objectives were to: (1) establish empirical correlations between N and K, (2) develop a robust prediction model with quantifiable bounds, and (3) demonstrate practical applications for enhanced subsurface characterization. Analysis of 3508 boreholes across South Korea revealed a statistically significant negative correlation between N and K in sandy soils. Quantile regression enabled prediction of both point estimates and percentile ranges. Evaluation of six empirical equations for K estimation identified the Chapuis equation as optimal, which was integrated with field measurements to strengthen the regression model. For weathered rocks, a consistent K range was established. The methodology’s novelty lies in combining readily available SPT data with advanced statistical techniques to generate high-resolution 3D K domains, as demonstrated through kriging. Despite relatively low R2 values, the methodology achieves practical accuracy with most predictions falling within one order of magnitude of measured values. This approach significantly enhances spatial and depth-wise characterization of subsurface K, offering a practical solution for groundwater flow modeling and geotechnical design with improved resolution.