Scholarlink

<p>Shallow foundations under inclined loading experience reduced bearing capacity and increased risks of sliding and overturning. Typical cases include wind turbine bases, offshore pier footings, and transmission tower foundations that are subjected to lateral wind or cable tension. This study investigates the behaviour of skirted footings on clay under inclined loads using finite element limit analysis with adaptive mesh refinement. The clay is modelled using the Tresca failure criterion to represent undrained conditions. A parametric study examines the effects of normalized undrained shear strength, skirt depth, load inclination angle, and soil-footing interface properties on foundation response. Results show the bearing capacity factor decreases significantly with increasing load inclination, dropping by up to 70% at 75°, while increasing skirt depth improves capacity, stabilizing beyond a depth ratio of 5. The soil-footing interface factor strongly influences bearing capacity, nearly doubling it when increased from 0.25 to 1.0. Machine learning regression models, especially Gaussian Process Regression which achieved a high prediction accuracy of 0.9999, proved highly effective in predicting the ultimate bearing capacity of foundations by accurately capturing the relationships between geotechnical inputs and the load bearing response.</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

Adaptive FELA and Machine Learning for Skirted Footing Performance Under Inclined Loading

  • 期刊论文

  • 发表日期：

    2025年11月27日

查看全文
  • Vinay Bhushan Chauhan,
  • Trisha Singh,
  • Sweta Verma,
  • Vikash Singh,
  • Suraparb Keawsawasvong

摘要

Shallow foundations under inclined loading experience reduced bearing capacity and increased risks of sliding and overturning. Typical cases include wind turbine bases, offshore pier footings, and transmission tower foundations that are subjected to lateral wind or cable tension. This study investigates the behaviour of skirted footings on clay under inclined loads using finite element limit analysis with adaptive mesh refinement. The clay is modelled using the Tresca failure criterion to represent undrained conditions. A parametric study examines the effects of normalized undrained shear strength, skirt depth, load inclination angle, and soil-footing interface properties on foundation response. Results show the bearing capacity factor decreases significantly with increasing load inclination, dropping by up to 70% at 75°, while increasing skirt depth improves capacity, stabilizing beyond a depth ratio of 5. The soil-footing interface factor strongly influences bearing capacity, nearly doubling it when increased from 0.25 to 1.0. Machine learning regression models, especially Gaussian Process Regression which achieved a high prediction accuracy of 0.9999, proved highly effective in predicting the ultimate bearing capacity of foundations by accurately capturing the relationships between geotechnical inputs and the load bearing response.