Research on real-time quality evaluation method for intelligent compaction of soil-filling

How to evaluate the compaction quality in real-time, i.e. how to establish the relationship between acceleration and dry density, is the key issue in the field of intelligent compaction. The response of acceleration to dry density involves two different physical mechanisms, where the relationship between force and acceleration follows the laws of kinematics and the relationship between force and void ratio (i.e. dry density) follows the constitutive law of soil. The current approach is to directly correlate dry density with acceleration, by which it is difficult to distinguish the role played by these two mechanisms. Moreover, the empirical formula established based on limited field data may not be in the correct mathematical form and may lead to invalid prediction in some cases. Firstly, the formula for peak acceleration and peak impact was obtained based on the kinematic equation, and the relationship between the acceleration and peak impact stress is further acquired. Additionally, the compaction envelope equation was established by the theoretical analysis of the compaction process under lateral restriction conditions, which describes the relationship between the peak impact stress and void ratio. The real-time calculation formulation for dry density in terms of peak acceleration was gained by coupling the above two equations based on the corresponding relationship between void ratio and dry density. Considering that there is lateral deformation during the actual compaction process, the proposed dry density formulation is approximated, and the calculation deviation caused by the difference between the actual constraint conditions and the ideal conditions can be reflected by adjusting parameters. Subsequently, the proposed compaction envelope equation was verified by the agreement between the existing laboratory experimental results and the prediction results. Finally, the real-time formulation for dry density was applied to the prediction of two sets of field tests, demonstrating that the formulation proposed in this paper is capable of predicting reasonably the compaction quality of soil.