Measurement of Fracture Mechanics Parameters in Silty-Clay Soils

Soil mechanical strength is an important indicator of soil physical properties that áffect agricultural production and reflects the soil-cracking characteristics important to water management and salt leaching of clay soils as well. Building on the field work and theoretical developments of Lima and Grismer in 1992 and 1994, we apply existing laboratory methods for measurement of fracture mechanics parameters of cement and mortar mixtures to clay soils in order to estimate the stresses associated with crack extension as soil dries. A set of ASTM procedures (C 580, D 2166, D 5045, and E 399) developed from linear elastic fracture mechanics theory are used to determine the relationship between the soilwater stress, σ, or water content and stress intensity factor, K, from cylindrical and beam specimens. Four different aggregate size fractions (<75 μm, 75-500 μm, 0.5-1.0 mm, and >1.0 mm) of Imperial silty clay (hyperthermic Typic Torrifluvent) were used to form sample specimens to bulk densities ranging from 1.31 to 1.59 Mg/m3 and water contents ranging from 10 to 23% by mass. The effects of aggregate size and soil-water salinity on soil cracking are also considered in terms of the stress intensity factor, K. Of the 64 tests conducted, the greatest number of valid test results (96%) occurred with the samples formed from the 75 to 500-μm aggregate size fraction, whereas the smallest size fraction yielded valid results in 60% of the tests. Valid test results were assumed when the test satisfied elasticity criteria determined from the system compliance and specimen size criteria associated with minimizing the "process" or plastic deformation zone relative to the specimen thickness, height, and initial crack length. Results from test specimens using the two larger size factors were inconclusive because of failure to meet validity criteria. An increase in soil-water salinity to approximately 5800 mg/L resulted in slightly smaller K values compared with distilled water-treated soils at water contents up to 23% by mass; however, considering the salinity effects on the soil-water retention functions, the effects of soil salinity on soil cracking are likely to be more pronounced at the greater water contents common in the field. Finally, the test methods described here for characterizing strength properties of soils may be preferred to the commonly used methods for measuring Young's modulus and modulus of rupture because of the validity criteria incorporated in the testing procedures.