Abstract: The compaction quality of cemented sand, gravel and rock (CSGR) is controlled by the VC value and the rolling parameters. The mix ratio parameters of raw materials have a direct impact on the VC value. Through orthogonal test (including four factors and three levels) in consideration of interactions, the test results of VC values were processed via statistical methods such as range analysis, variance analysis, multiple linear regression analysis and residual analysis. The sensitivity of VC value of CSGR to four factors, i.e., unit water consumption (W), unit cement consumption (C), unit fly ash consumption (F) and sand ratio (S) was examined. Results demonstrate that the contribution rates of W, C, F, S and test error to VC value were 75.55%, 6.73%, 17.00%, 0.08% and 0.64% respectively. The influence of sand ratio on VC value is almost negligible. Besides, W should be strictly controlled and the VC value should be monitored in real time, so as to adjust construction measures and control construction quality. In line with the research results, a prediction model of VC value is proposed, which is of guiding significance for rationally adjusting the mix ratio parameters to control VC value and improving construction quality.

Key words: cemented sand, gravel and rock dam, VC value, working performance, construction quality, sensitivity analysis, prediction model