Abstract: Five factors (the water to binder ratio X1, ratio of cementitious material to sand X2, ratio of fly ash to lime X3, ratio of bentonite to water X4, and ratio of water reducing agent to cementitious material X5), and three levels were selected to conduct orthogonal experiments of inert slurry. The influence laws of various factors on the consistency (cone penetration P), setting time ts, and consolidation shrinkage rate ε, 28-day compressive strength f28 and bleeding rate η of inert slurry were analyzed. Partial least squares regression (PLS) was used to obtain regression formulas for various performance indicators of inert slurry based on five factors. The PLS regression values of various performance indicators were compared with the test values to verify the rationality of the regression formula. In the experiment, water reducing agents (X5=0, 0.002, 0.004, 0.006, 0.008) and CMC thickeners (dosage 0, 0.05%, 0.10%, 0.15%, 0.20%) were added to the reference inert slurry, and the cone penetration P, bleeding rate η, viscosity μ, and flowability F of the prepared inert slurry were measured. The modification effects of water reducing agents and CMC thickeners were analyzed, and the engineering geological applicability of inert slurry was discussed. The results show that the water to binder ratio has a significant impact on all indicators, the ratio of cementitious material to sand has a significant impact on the setting time and consolidation shrinkage rate, the ratio of bentonite to water and ratio of water reducing agent to cementitious material have a significant impact on the cone penetration and bleeding rate, and the overall impact of ratio of fly ash to lime is relatively small. The PLS regression formula can be used to optimize the proportion of inert slurry. When proportioning the slurry, an appropriate amount of water reducing agent can be added to reduce the water to binder ratio while ensuring the required consistency requirements. An appropriate amount of CMC thickener can also be added to improve the viscosity of the slurry.

Key words: tunnel engineering; inert slurry; partial least squares regression (PLS); mixing proportion

摘要： 选择5因素（水胶比X1、胶砂比X2、粉灰比X3、膨水比X4、减胶比X5）、3水平开展惰性浆液正交试验，分析了各因素对的惰性浆液稠度（锥入度P）、凝结时间ts、固结收缩率ε、28天抗压强度f28及泌水率η等5个性能的影响规律；采用偏最小二乘回归（PLS）得到基于5个因素的惰性浆液各性能指标的回归公式，将各性能指标的PLS回归值与测试值进行对比以验证回归公式的合理性；在基准惰性浆液中分别掺入减水剂（减胶比X5=0、0.002、0.004、0.006、0.008）和CMC增稠剂（掺量0、0.05%、0.10%、0.15%、0.20%），测得所制备惰性浆液的锥入度P、泌水率η、黏度μ和流动度F，分析了减水剂及CMC增稠剂的改性效果；讨论了惰性浆液的工程地质适用性。结果表明：水胶比对各个指标影响均较大，胶砂比对凝结时间、固结收缩率影响较大，膨水比与减胶比对锥入度、泌水率影响较大，粉灰比总体影响较小；PLS回归公式可用于惰性浆液配比的优化；浆液配比时，在保证稠度要求的前提下可加入适量减水剂以减小水胶比，也可添加适量CMC增稠剂来改善浆液黏度。

关键词: 隧道工程；惰性浆液；偏最小二乘回归；配比