中南大学学报（自然科学版）2025，Vol.56Issue(4)：1588-1603,16.DOI:10.11817/j.issn.1672-7207.2025.04.028
红黏土干缩裂隙表面形态量化下的深度计算模型
Calculation model of depth under quantification of surface morphology of dry shrinkage cracks in red clay
摘要
Abstract
Clay desiccation cracking is a common natural phenomenon that compromises soil integrity,reducing its strength and permeability,and it is a key factor in geotechnical failures such as landslides and dam instability.Understanding the variation in desiccation crack depth is crucial for investigating the impact of crack depth on slope infiltration patterns and determining slope stability.Laboratory model experiments can reveal the relationship between surface and depth changes during crack formation,and explore the effect of surface roughness on the desiccation cracks in saturated red clay.Regression analysis and machine learning algorithms were used to establish the depth prediction models.The results show that based on surface morphology,the effective desiccation cracks are generally classified into three types,i.e.linear,curved,and kinked,with linear and curved types accounting for 54％and 29％of the total,respectively.With the increase of soil surface roughness,the propagation speed of surface cracks increases by 75％,and the probability of turning increases by 8％,resulting in the variation characteristics of effective crack morphology at the same depth.The Pearson correlation coefficients between the surface area,length,and spacing of cracks and their depth in saturated red clay are 0.90,0.93 and 0.87,respectively.Using surface geometric parameters,depth prediction models are established via multiple nonlinear regression and gradient boosting regression tree algorithms,with prediction errors controlled within 15％,which can provide reference for preliminary crack depth assessment.关键词
红黏土/干缩裂隙/表面形态/裂隙深度/计算模型Key words
red clay/dry shrinkage crack/surface morphology/crack depth/calculation model分类
建筑与水利引用本文复制引用
董辉,王威,姚英杰,刘高志,徐平..红黏土干缩裂隙表面形态量化下的深度计算模型[J].中南大学学报（自然科学版）,2025,56(4):1588-1603,16.基金项目
国家自然科学基金资助项目(42272304) （42272304）
湖南省教育厅科学研究项目(22C0046)(Project(42272304)supported by the National Natural Science Foundation of China （22C0046）
Project(22C0046)supported by the Education Department of Scientific Research Program of Hunan Province) （22C0046）
