基于堆积颗粒几何特征的多尺度渗透注浆扩散半径预测

Multiscale penetration grouting radius prediction based on geometric characteristics of stacked particles

为精准预测渗透注浆在粗粒土中的扩散半径,基于堆积颗粒的几何特征,提出微尺度颗粒几何特征识别与中尺度注浆通道重建算法,并结合大尺度注浆扩散模型,构建多尺度渗透注浆扩散半径预测模型。首先,基于粗粒土CT断面图像,采用颗粒几何特征识别算法(GIM)识别颗粒的椭圆度、粗糙度、长轴倾角以及粒径参数;而后依据这些识别参数,结合孔隙率和级配,通过中尺度注浆通道重建算法(GRM)建立渗流通道模型;基于该模型,联合运用多场耦合有限元(FEM)与蒙特卡洛法(MC),获得具有统计意义的中尺度渗透注浆模型;通过渗流速度将中尺度渗透注浆模型与大尺度注浆扩散模型耦合,建立多尺度渗透注浆扩散半径预测模型。通过与试验结果对比,验证预测模型的合理性(误差小于10%)。最后,采用该模型系统研究椭圆度、粗糙度、粒径、孔隙率、长轴倾角、流性指数和浆液平均黏度对注浆扩散半径的影响,大小关系为:粒径>浆液黏度>长轴倾角>流性指数>孔隙率>椭圆度>粗糙度,对应的皮尔逊相关系数依次为0.5,-0.374,-0.238,-0.188,0.161,-0.036,-0.019;通过对浆液扩散过程分析,发现堆积颗粒几何特性对浆液扩散的影响机制主要包括“孔喉约束”、“惯性阻力”、“通道宽度”以及“流固界面”效应。

In order to accurately predict the penetration grouting radius in coarse-grained soil,a method for identificating geometrical characteristics of microscale particle and resconstructing mesosceale grouting channels based on the geometrical characteristic of stacked particles was put forward.And combining these two methods and the largescale grouting diffusion model,a multiscale penetration grouting radius prediction model was established.First,based on the mesoscale CT sectional images of geomaterials and employing the geometric characteristics identification method(GIM),the ellipticity,roughness,long-axis inclination and size of the stacked particles were identified.Then,employing the grouting channel reconstruction method(GRM),the penetration channel model was reconstructed according to the obtained parameters,porosity and gradation.According to the reconstructed model,the mesoscale penetration diffusion model with statistical significance was obtained using the multi-field coupled finite element method(FEM)and Monte Carlo method(MC).Coupling the mesoscale penetration diffusion model and the largescale grouting diffusion model according to the seepage field results in the multiscale prediction model of penetration diffusion radius.Compared to the existing experimental results,the rationality of the prediction model is verified(the error is less than 10%).Finally,based on the proposed method,this paper systematically studied the influences of ellipticity,roughness,particle size,porosity,long-axis inclination,time-varying factor and average slurry viscosity on grouting diffusion radius.The affecting relationship in reducing order is:particle size,average slurry viscosity,long axis inclination,time-varying factor,porosity,ellipticity and roughness,and the Pearson coefficient is 0.5,-0.374,-0.238,-0.188,0.161,-0.036,-0.019,respectively.Through the analysis of the grouting diffusion process,it is found that the influence mechanism of the geometric characteristics of the stacked particles on grouting diffusion mainly includes the“pore-throat constraint”,“inertia resistance”,“channel width”and“fluid-solid interface”effect.