ANN prediction of some geotechnical properties of soil from their index parameters

This paper presents artificial neural network prediction models which relate compaction characteristics, permeability, and soil shear strength to soil index properties. In this study, a database including a total number of 580 data sets was compiled. The database contains the results of grain size distribution, Atterberg limits, compaction, permeability measured at different levels of compaction degree (90–100 %) and consolidated–drained triaxial compression tests. Comparison between the results of the developed models and experimental data indicates that predictions are within a confidence interval of 95 %. To evaluate the effect of each factor on these geotechnical parameters, sensitivity analysis was performed and discussed. According to the performed sensitivity analysis, Atterbeg limits and the soil fine content (silt + clay) are the most important variables in predicting the maximum dry density and optimum moisture content. Another aspect that is coherent from the sensitivity analysis is the considerable importance of the compaction degree in the prediction of the permeability coefficient. However, it can be seen that effective friction angle of shearing is highly dependent on the bulk density of soil.