为定量研究盾构掘进复合地层时掘进参数对碴土颗粒级配特征的影响，以试验研究和统计分析为手段，得到掘进参数和颗粒特征之间的量化规律。首先，通过开展现场掘进试验，在不同有效推力和刀盘转速下实地采取碴样颗粒； 然后，对颗粒筛分结果进行统计分析，提出碴土颗粒级配预测方程； 最后，对级配预测方程的系数进行多元回归预测和影响性分析，定量研究方程系数和掘进参数之间的相关性。研究结果表明： 1）当盾构刀盘转速或盾构有效推力增大时，碴样特征粒径减小，碴土中的粗颗粒含量降低，细颗粒含量增大，碴样粒度分布范围增宽； 2）碴土颗粒的粒径存在上限，RosinRammler分布函数更加适用于盾构破岩颗粒级配特征分析； 3）在一定掘进参数范围内，盾构碴土颗粒级配方程系数的预测模型预测精度较高，能够较好地反映工程实践中掘进参数对方程系数的影响。

In order to study the quantitative relationship between shield tunneling parameters in composite strata on the particle characteristics of muck, the experimental study and statistical analysis are conducted. Firstly, the muck sample particles are collected in field shield tunneling test under different effective thrust forces and rotation speeds of cutterhead. And then the statistical analysis of particle screening results is carried out, and the prediction equation of particle grading of muck is proposed. Finally, the multiple regression prediction and influence analysis are carried out for the coefficients of the grading prediction equation, and the correlation between the coefficients of the equation and the tunneling parameters is quantitatively studied. The results show that: (1) When the rotation speed of cutterhead or effective thrust force of shield increases, the characteristic particle size of muck sample decreases, the coarse particle 〖HJ2.7mm〗content in muck decreases, the fine particle content increases, and the particle size distribution range of muck sample widens. (2) The particle size of muck particles has an upper limit, and RosinRammler distribution function is more suitable for analyzing the grading characteristics of shield broken rock particles. (3) Within a certain range of tunneling parameters, the prediction model of particle grading equation coefficients of shield muck is more accurate, which can better reflect the influence of tunneling parameters on the equation coefficients in engineering practice.