对各级颗粒含量进行等效分级 ( 按粒径由小到大分为四级 ) 并确定其破坏模式，依据颗粒破碎的 Markov 性建立模拟堆石料颗粒破碎的连续时间 Markov 链模型。根据室内三轴试验数据的统计分析确定稳态的判别模式，给出计算流程和仿真实例,得到了各级颗粒含量的动态变化曲线，并对模型的灵敏度进行了分析。结果表明：随着时间的增加，粒径较大的第三级和第四级颗粒的含量单调下降，下降的速率随时间的增加而减缓，其中第四级颗粒下降的幅度较大。第二级颗粒的含量先随时间增长，到第 15d 左右后随时间下降。通过对模型的灵敏度分析，该模型对试验参数的变化是稳定的。

Carrying on the equivalent graduation to all levels of particle content and determining its destruction mode, the author constructs a continual Markov chain model based on the Markov nature of the particle breakage to simulate the particle breakage of coarse aggregates. According to data of the indoor triaxial tests, the criterion of the steady-state is obtained and then a simulation example is given. The results show that, as the time increases, the particle contents of the fourth and third classes are monotonically decreased, and their rate slowed down with time, of which the fourth class is dropped largely, and that the particle content of the second class is first grown with the time and then declined after 15 days. Through the sensitivity analysis of the model, it is stable with the test parameters.