Ternary Diagram Modeling of Soil Texture Data for Predicting Subsurface Fracturing in Glacially Related Fine-Grained Materials

摘要

Fractures play an important role in the movement of water and contaminants through soil materials. The objectives of this research were: (I) to determine the best model for predicting fracturing, which could serve as a practical tool to anticipate andinvestigate fractures in glacially related fine-grained soil materials, and (2) to identify those soil textures most likely to support fracturing. A dataset based on Ohio soils observed in the field to be fractured was extended to cover a wider range oftextures through performing controlled laboratory fracturing experiments. The 30 data produced by these laboratory experiments were added to the 143 field data and together used to develop five statistical and graphical methods: the silhouette method, the girth method, the hexagonal field method, Weltje's method, and a non-statistical best-fit method. The best predictive model of the five was the non-statistical best-fit method. When plotted on a USDA ternary diagram, soils with <79% sand and >6% claywere predicted to support fracturing. All texture classes of soils were identified as able to sustain fracturing except the loamy sand and sand classes. This graphical model can be useful to explain how fractures are created in glacially related fine-grained materials and may be a practical tool allowing geologists and field engineers to anticipate fractures without resorting to more costly methods such as soil borings, soil pits, or excavations.