颗粒系统的多尺度耦合方法

Multi-scale Coupling for Particle System

本文用随机模拟的思想产生了均匀分布与正态分布的颗粒系统,建立了大小系统之间的多尺度耦合关系,定义了反映不同尺度系统结构之间相似特征的量化参数-团聚体数目和空隙率,在此基础上建立了颗粒系统的多尺度放大。并且,本文还讨论了系统尺度、颗粒粒径、初始颗粒数以及统计微元的步长对多尺度放大效应的影响,其研究结果表明: 1)在实现关于团聚体数目的多尺度放大时,颗粒尺度较大、初始颗粒数较多,放大效果较好。 2)在实现关于空隙率的多尺度放大时,初始颗粒数较多,放大效果较好;而颗粒尺度对放大效果的影响不大;统计微元的步长对放大效果有一定的影响,对于颗粒分布比较均匀的介观结构, 可以取较大尺度(步长较大)的统计微元进行研究,而对于颗粒分布具有明显非均匀特性的介观结构,则要取较小尺度(步长较小)的统计微元进行研究。

It is presented that the multi-scale coupling relationship of different scale particle systems for uniform and normal distributions by using random simulation. The number of cluster and the voidage are defined, and they show similarities for different scale particle systems. Then the multi- scale enlargement concerning these similar variables is founded. Meanwhile, it is discussed how the multi-scale enlargement are influenced by the system scale, the particle diameter, the initial number of particles and the step of tiny statistic unit. Based on analyses, some results can be obtained as follows. 1） The number of clusters is a similar variable that shows the relation between large-scale and small- scale particle systems. Therefore, the multi-scale enlargement concerning the number of clusters can be achieved. When the particle diameter is larger or the initial number of particle is bigger, the effect of enlargement is better. 2） The voidage is also a similar variable that shows the relation between large-scale and small-scale particle systems. Therefore, it can also achieve the multi-scale enlargement about the voidage. The more initial number of particles is, the better effect of enlargement is. In addition, the particle diameter doesn＇t affect this kind of enlargement too much. The step of tiny statistic unit has certain effect to this kind of enlargement. When particle distribution is more even in microstructure, we can consider the larger-scale （larger-step） tiny statistic unit. On the contrary, when uneven character of particle distribution is obvious, the smaller-scale （smaller-step） tiny statistic unit should be considered.