Property-based assessment of soil mineralogy using mineralogy charts

It is essential to determine soil mineralogy as soils containing high amounts of smectites pose a hazard to geotechnical structures and increase susceptibility to soil erosion and landslides. The everyday use of analytical techniques in geotechnical engineering practice is considered expensive and time consuming. Prior studies suggest that physico-chemical properties of soil can be used to estimate the soil mineralogy. The current study evaluated specific surface area and liquid limit as possible parameters to estimate soil mineralogy. Mineralogy charts are introduced to estimate the percentage of montmorillonite, kaolinite, and illite in a sample. The degree of accuracy is highly dependent on the method used to determine the soil properties. Mineralogy ChartSSA yielded as accurate results as artificial neural network based models. Prediction of soil mineralogy based on liquid limit of the soil yielded a wide range of correlation coefficients between the measured and calculated values.