Using a Novel Instrumented Roller to Estimate Soil Dry Density During Compaction

Abstract

AbstractCapturing evolution of density or void ratio during the compaction of geomaterials (soils and unbound granular materials) is essential for improved performance. This study developed a framework where the density evolution during compaction can be estimated using advanced instrumentation. The framework’s suitability was validated using a simulated large-scale soil box (dimensions: \(7.5\,{\rm m}\,\times\,4\,{\rm m}\,\times\,0.8\,{\rm m}\)) experiment mimicking the field conditions. Well-graded sand was compacted in 5 layers of 125 mm using a 1.5-tonne mini roller instrumented with Light Detection and Ranging (LiDAR) systems and a total station tracking system for positioning.The sand’s moisture content was homogenised at 8% (w/w) using a concrete truck. The in-situ sampling for measuring density was carried out using Nuclear Density Gauge (NDG) and sand cone test. The data from sensors were collected using a Data Acquisition (DAQ) system connected to a laptop. The measurement of the deformation in real-time provided an opportunity to estimate the density in real-time, and it was estimated using a machine-learning artificial neural network (ANN) model. The estimated density from deformation measured and NDG at the end of compaction shows that estimated density NDG density with an R = 0.9 for one layer, and for other layers, R was more than 0.8. This novel instrumentation allows the density to be measured during compaction with very high accuracy, which has a massive advantage over conventional approaches and contribute to the true Intelligent Compaction (IC) with an advancement of automation in construction.KeywordsCompactionLight detection and rangingNuclear density gaugeDensityIntelligent compaction