.. only:: html .. figure:: dem_am.gif :width: 60% :align: center **单位(cm)**
Note
~10万颗粒,~13万时步,16核并行(Intel Xeon E5-2650),耗时 ~3小时 完成计算。
构建了一个典型离散元数值模拟(DEM)试验,对比与构造物理模拟(AM)和差异与相似点。
DEM与AM模拟结果较为一致,基本反映了AM中石英砂的变形行为。AM中材料选取有限(如硅胶、粘土、微玻璃珠等),而DEM材料选取范围大,但其模拟结果依赖参数选取。AM和DEM作为两种独立的方法,有很好的互补性 [Li2021] 。
数值模拟分两步进行:
ex8_dem_am_gen.py
生成颗粒,沉积,构建初始模型。ex8_dem_am_push.py
给定颜色、粘结参数,挤压。
表 1 颗粒微观参数表. [Li2021]
d(mm) | k(N \cdot m^{-1}) | \rho (kg·\cdot m^{-2}) | g(m·\cdot s^{-2}) | f | \mu | \eta (N \cdot s \cdot m^{-1}) | \upsilon (m·\cdot s^{-1}) |
---|---|---|---|---|---|---|---|
0.6 | 7.5e3 | 1.3e3 | 10.0 | 0.4 | 0.3 | 0.04 | 0.04 |
The particle packing consists of four particle sizes, with diameters and quantity ratio of 0.2 mm, 0.4 mm, 0.5 mm, and 0.6 mm and 2:2:1:1, respectively. :math:`d , largest particle diameter. \rho , particle density. g , gravitational acceleration. :math: f , safety factor of the time step. k , stiffness of the contact. \mu , friction coefficient. \eta , dynamic viscosity. \upsilon , velocity of the mobile wall.`
ex8_dem_am_gen.py
.. literalinclude:: ex8_dem_am_gen.py
ex8_dem_am_push.py
.. literalinclude:: ex8_dem_am_push.py
计算结束后,将得到以下结果:
参考文献
[Li2021] | (1, 2) LI C.S., YIN H.W.*, et al. (2021) Calibration of the discrete element method and modelling of shortening experiments. Front. Earth Sci. 9:636512. |