摘要
土体的密实程度对土体的工程性质影响很大。获得给定相对密实度下的初始颗粒堆积体,是利用离散单元法进行土体细观力学分析的重要前提。基于离散单元法,首先获得颗粒在自重作用下稳定后的最松散堆积体。其次提出采用减小压实过程中颗粒间摩擦系数的拟振动压实法来模拟现实中的振动压实作用,并验证了其可行性。当压实过程中颗粒间摩擦系数为0时获得最密实堆积体。研究了拟振动压实过程中颗粒间摩擦系数与压实稳定后的堆积体孔隙率的关系。最后根据该关系可估算出获得任意给定相对密实度堆积体所需的压实过程中颗粒间的摩擦系数。研究结果表明:颗粒的配位数与粒径近似满足指数关系;通过减小压实过程中颗粒间的摩擦系数能较好的模拟振动压实效果;压实过程中颗粒间的摩擦系数与堆积体孔隙率满足负指数关系;该方法可获得任意相对密实度下的颗粒堆积体,可以为土体细观力学的数值模拟提供初始颗粒堆积结构。
The relative density of non-cohesive soil has a great impact on the engineering properties of soil. For most studies of mesoscopic simulation by using distinct element method(DEM), the first step of the analysis is to create an initial granular assembly with a desired relative density. Based on the DEM, the loosest state is obtained by depositing particles under gravity firstly. Secondly, a quasi vibration and compaction method(QVC) is proposed to simulate the vibration and compaction method by reducing the friction between the particles during the compaction. And the feasibility of QVC method is validated. The densest state is obtained by reducing the friction to zero. The relationship between the friction and packing porosity is studied.Based on this relationship, the friction which is used to obtain the assembly with desired relative density can be estimated. The results show that the relationship between the coordination number of particle and the particle diameter is compatible with an exponential distribution. By reducing the friction between the particles, the vibration process can be well reflected. The relationship between the friction and the packing porosity agrees with the negative exponential distribution. Based on the proposed method, an initial granular assembly with any desired relative density can be obtained, which can be used for mesoscopic simulation of grained materials.
出处
《岩土工程学报》
EI
CAS
CSCD
北大核心
2015年第3期537-543,共7页
Chinese Journal of Geotechnical Engineering
基金
国家自然科学基金项目(51079039
51009053)
关键词
离散单元法
相对密实度
颗粒堆积体
摩擦系数
配位数
distinct element method
relative density
granular packing
friction coefficient
coordination number