[交流] 【讨论】自编程序中正则系综的体系温度为什么不能保持恒定？

Originally posted by zeoliters at 2009-11-8 18:17:
既然使用了周期性边界条件，如果体系为正则系综，体系的数目应该不会变化的，如果体系数目减少，我怀疑是不是程序出现了错误。

Originally posted by dellus at 2009-11-9 11:20:
周期性边界条件用的不对，
不要把原胞范围外的原子放回原胞，只需要在计算相互作用时把座标折算到原胞内就可以了

Originally posted by lionel770 at 2009-11-9 15:56:
具体情况没有看你的运行情况不好说，如果是你说分析的问题，可以考虑减小每步的时间步长试试。具体的解决方案还是在程序的BUG判断及处理上下功夫。

Originally posted by sheath at 2009-11-9 17:20:

我的程序在计算原子位置和相互作用力都使用了周期性边界条件。

Originally posted by sheath at 2009-11-9 17:20:

"The wraparound effect of the periodic boundaries must be taken into account in both the integration of the equations of motion and the interaction computations. After each integration step the coordinates must be examined, and if an atom is found to have moved outside the region its coordinates must be adjusted to bring it back inside."