针对洞庭湖生态修复工程中钻探抽砂工艺引发的沉积层突发性塌陷预测难题,基于自行设计的取砂可视化试验装置,开展了不同级配砂土、取砂深度和水深下的概化物理模型试验,分析砂土变形发展过程,并建立了钻探取砂诱发地表塌陷预测模型,实现钻探取砂引起地表变形宽度和最大塌陷深度的预测。研究结果表明,砂土流动区边界呈现椭圆形,其离心率随特征粒径D₉₀或取砂深度增加而减小,随水深的增加而增大,此外,水深增加还将加快砂土塌陷速度;在Peck沉降曲线与地表塌陷曲线的拟合优度中,当土体损失率低于20%时,拟合优度偏低,呈现离散分布(R²=0.78~0.96)。随着取砂过程的持续,Peck曲线与地表塌陷曲线的拟合优度逐步增加,并在55%左右达到最大(R²=0.98),此后,塌陷曲线逐渐向倒三角形态进行演变,与Peck曲线拟合优度也随之降低;建立的地表塌陷预测模型在塌陷深度<15 cm存在较好的适用性。该研究成果可揭示洞庭湖生态疏浚工程中沉积层塌陷规律并预测塌陷范围,为安全施工与灾害风险提供理论参考。

[Objective] This study addresses the challenge of predicting sudden subsidence of sedimentary layers caused by drilling and sand extraction in the Dongting Lake ecological restoration project. By analyzing the development process of soil deformation, a prediction model for surface subsidence induced by drilling sand extraction was established to achieve rapid prediction of surface subsidence deformation width and maximum subsidence depth. The aim is to reveal the collapse patterns of sediment layers in the ecological dredging project of Dongting Lake and predict the collapse areas, providing theoretical references for safe construction and disaster risk management. [Methods] Based on a self-designed sand extraction visualization test apparatus, physical model experiments of drilling sand extraction were conducted under different sand gradations, extraction depths, and water levels. The sand soil collapse process was recorded using a high-speed camera, and the data were processed using PIV (Particle Image Velocimetry) technology to obtain the collapse curve and quantify the soil deformation process. [Results] Experimental results revealed:(1)The collapse zone exhibits an elliptical shape, with its eccentricity decreasing as the characteristic particle size D₉₀ or extraction depth increases, while increasing with greater water depth.(2)In terms of the goodness-of-fit between the Peck curve and the surface collapse curve, when the soil loss rate is below 20%, the goodness-of-fit is relatively low, showing a scattered distribution (R²=0.78-0.96). As the extraction process continues, the goodness-of-fit of the Peck curve to the surface collapse curve gradually increases, reaching a maximum at around 55% (R²=0.98). Thereafter, the collapse curve gradually evolves into an inverted triangular shape, and the goodness-of-fit of the Peck curve correspondingly decreases.(3)The developed collapse prediction model has good applicability within the range of collapse depth <15 cm, with a collapse depth prediction R²=0.73 and a surface collapse width prediction R²=0.65. [Conclusions] This research demonstrates that:(1)Water depth will increase the range of sand soil collapse and accelerate the speed of sand soil collapse.(2) The eccentricity is affected by sand soil particle size, extraction depth, and water depth.(3)The Peck curve is suitable for the range of moderate sand soil loss and needs to be modified when the sand soil loss rate is small or large.