文章摘要
张季如,钟思维.四参数随机生成法重构土体微观孔隙结构的分形特征[J].水利学报,2018,49(7):814-822
四参数随机生成法重构土体微观孔隙结构的分形特征
Fractal behaviors of microscope pore structure of soil reconstructed by quartet structure generation set
投稿时间:2018-03-20  
DOI:10.13243/j.cnki.slxb.20180237
中文关键词: 四参数随机生成法  重构  微观孔隙结构  质量分维数Dm  表面分维数Ds
英文关键词: quartet structure generation set (QSGS)  reconstruction  microscopic pore structure  mass fractal dimension Dm  surface fractal dimension Ds
基金项目:国家自然科学基金项目(41072214,41272334);武汉理工大学研究生优秀学位论文培育项目(2016-YS-044)
作者单位
张季如 武汉理工大学 土木工程与建筑学院, 湖北 武汉 430070 
钟思维 武汉理工大学 土木工程与建筑学院, 湖北 武汉 430070 
摘要点击次数: 165
全文下载次数: 112
中文摘要:
      四参数随机生成算法(QSGS)生成的微观结构模型,已被广泛用于研究土的微观孔隙结构并预测土的水力学性质,但模型能否真实反映实际土体的孔隙结构特征,目前尚缺乏实验证据及定量对比分析。利用扫描电镜(SEM)对3种不同孔隙率的土样制备SEM图像,以与QSGS模型进行定量对比分析。采用数字图像技术分析QSGS模型与土体中孔隙的形态特征和分布规律,基于实测数据估算孔隙的质量分维数Dm和表面分维数Ds。结果表明:QSGS算法生成的微观结构由孔隙度P、生长核的分布概率Pd和方向生长概率Pi等参数控制,其中Pd对孔隙结构的影响更为显著。Pd ≤ 0.01时所生成的QSGS模型与实际土体具有相似的微观孔隙形态和分布规律,以及相同的分形特性和相近的分维数值。P越小则Dm越大,DmP存在较为显著的线性回归关系。孔隙轮廓愈不规则,Ds愈大,各孔隙的Ds分布符合总体正态分布形式。研究结果揭示了模型参数对QSGS算法生成的微观结构的影响,为合理选取参数提供了科学依据。
英文摘要:
      The microstructure model generated by a multi parameter random generation-growth algorithm, termed as the quartet structure generation set (QSGS), has been widely used to study the soil pore struc-ture and predict the hydraulic properties of soil. However, there is still lack of experimental evidence and quantitative comparative analysis whether the model can truly reflect the soil pore structure characteristics. The samples of three soils different in porosity were scanned with a scanning electron microscope (SEM) producing SEM images for comparing quantitatively with the QSGS model. The morphological characteristics of pores and their distributions in the model and soil were analyzed using digital image technology. The mass fractal dimension Dm and the surface fractal dimension Ds of pores are estimated from the measured data. The results show that the microscopic pore structure generated by the QSGS algorithm is controlled by parameters such as porosity P, core distribution probability Pd, and directional growth probabilities Pi, among which Pd has more significant influence on the pore structure. The QSGS model generated when Pd ≤ 0.01 has similar microscopic pore morphology and pore distribution with the soil, as well as the same fractal properties and close fractal dimension values. The lower the porosity is, and the larger the Dm is. The Dm displays a significant linear regression relationship with the porosity. The larger the Ds the more irregular the soil pore profile. The distribution of Ds is found in agreement with a total normal distribution in the pore. The research results reveal the influence of model parameters on the microstructure generated by QSGS algorithm and provide scientific basis for the rational selection of model parameters.
查看全文   查看/发表评论  下载PDF阅读器
关闭