مشخصات پژوهش

The current paper aims to thoroughly study the compaction properties of cemented sand-gravel (CSG) mixtures by both experimental and computational approaches. Compaction properties of many mixtures with different soil types covering wide ranges of grain size and plasticity and with different cement contents are measured here. The main factors governing the compaction specifications of CSG mixtures are then determined by the statistical analysis via an artificial neural network (ANN). The results showed that the cementing agent improved the maximum dry density (MDD) of cemented soils by up to 6%. The greatest increase of MDD with cement occurred in CSG mixtures with no fine-grained particles. Besides, the cement slightly decreased the optimum water content (OWC) of CSG mixtures. If CSG mixtures contained non-plastic fine-grained particles, OWC increased with cement. According to the ANN analysis, mean grain size, PI, Gs, fine content, and gravel content have the highest influence on the MDD of mixtures, and the cement content was of the lowest relative importance. In contrast, PI and fine content had the highest impact on the OWC of the CSG mixtures. The parametric analyses indicated that the compaction of mixtures was primarily governed by the type and content of the fine-grained particles in the cemented mixture.