Correlation Verification Experimental Investigation Of High-railway Subgrade Compaction Quality Test Value

Compaction quality of subgrade is a crucial link to ensure its engineering performance and long-term stability as well as reduce the post-construction settlement.Though there are requirements in specifications for the test methods, acceptance criteria and evaluation system of the compaction quality, thereare still difficulties to control the compaction quality of subgrade due to the inhomogeneity of filling material and the limitations of test andevaluation methods. These kinds of disadvantages which may further influence the riding performance as well as reduce the service life of the road and then will result in great economic loss. As a matter of fact, in order to guarantee the compaction quality, many inspect methods were employed in high-speed railway construction in China to guarantee the compaction quality, which may delay the construction process schedule and are not adaptable to the requirement of modern construction including the compaction quality controlling of subgrade. These disadvantages are more evident when the coarse filling material is used in practice.In order to solve these problems, the ministry of railway proposed the specification (Technical Code for Continuous Compaction Control of Railway Earth Structure Fill Engineering)in2011. Besides the compaction continuous control in construction process, it’s also important to test correlation. Based on three test subgrades in Shanghai-Kunming high-speed railway, Loudi, Shaoshan and Zhijiang, this paper developes the correlation test of this specification. The research results can be concluded in four parts:(1) In the correlation test between continuous compaction index and conventional index, the abnormal values will have effects on correlation coefficient and the target value of compaction index. To screen these values, based on the analysis of the cause of abnormal values, this paper propose two effective and accurate methods, average statistical distance method and neighborhood evaluation method. Based on the methods, a planar abnormal distinguishing model was established and the applicability of this model was examined based on relative compaction experiment. Finally, the screen principle based on the methods was proposed.(2) In the correlation test between continuous compaction index and conventional index, the non-uniformity of the sub-base stiffness will have effects on correlation coefficient and the target value of compaction index. To screen these values, based on the calculation results, the multiple linear regression estimation model is established to propose the compacting value which is more suitable for the first five layers of sub-base.(3) In the correlation test between continuous compaction index and conventional index, the difference of the area represented by compaction index will have effects on correlation coefficient and the target value of compaction index. The correlation test was examined based on point-to-point method in the specification. To make it more precise, this paper proposed the point-to-averge value and averge value-to-averge value methods to test the correlation, the results show that these two methods are more accurate than the point-to-point method and offer an practical method to test the correlation.(4) In the correlation test between continuous compaction index and conventional index, compaction indexes will have effects on correlation coefficient and the targetvalue of compaction index. This paper established a multiple regression model between CMV and compaction indexes based on regress function in MATLAB toolbox, then the model was applied to examine the applicability of this model based on relative compaction experiment. The results have some guiding significantly in constrcution control.