Prediction of slurry compressibility using an artificial neural network

The compressibility of fine-grained slurries is important to develop appropriate dewatering schemes for mine tailings, dredged materials and cutoff walls. Severe geoenvironmental issues can result owing to the high water-holding capacity of such geomaterials. Laboratory investigation and field monitoring programmes to assess large-strain consolidation of slurries are expensive and time-consuming. Based on artificial neural networks, this chapter develops a modelling approach to predict slurry compressibility using the basic soil properties data of 71 fine-grained materials. Results indicate that the parameters A (corresponding to an effective stress of 1.7 kPa) and B (scaling exponent) as well as the void ratio are successfully estimated (R2 > 0.9). Finally, grain size distribution and liquid limit were found to govern the dewatering behaviour of fine-grained slurries.