Date of Award

Spring 2014

Degree Type


Degree Name

Master of Science (MS)


Civil Engineering

First Advisor

Maria C. Santagata

Second Advisor

Antonio Bobet

Committee Member 1

Darrell G. Schulze


The work presented in this thesis is part of a broad effort aimed at the characterization of the engineering properties, microstructure and mineralogy of soft fine-grained carbonatic soil deposit in Southwestern Indiana (Daviess Co.). The specific focus of this thesis is the characterization, through both field and laboratory measurements, of the small strain behavior of this soil, commonly referred to as a "marl."

Index tests indicate that the 3.5 m marl layer present at the site is characterized by two soil types alternating in small sublayers: a high plasticity silt (soil M) with wn ∼60% and CaCO3 ∼60%, mostly in form of shells, and a low plasticity clay (soil C) with lower water (∼45%) and CaCO3 (∼40%) contents, and no shells. Resonant column tests conducted on high quality samples are used to measure the shear modulus of isotropically consolidated specimens of both soil M and soil C as a function of stress level (70-650 kPa) and OCR (1-4) for shear strains between 10 -4 % and 0.1%. The Gmax data for the two soil types fall on distinct bands when normalized by void ratio, with the modulus of soil M consistently greater than that of soil C soil at any stress level. Differences are also observed in the stiffness degradation behavior, with soil C exhibiting greater non-linearity at the same stress level and OCR. Finally, measurements of G max over time provide values of the aging parameter N G , which for both soils falls in the range typically reported for clays.

The shear wave velocity (Vs ) profile of the site, obtained from two seismic cone penetration tests, indicates that the marl layer is characterized by values of Vs in the 110-160 m/s range, significantly lower than those of the layers above and below it. However, these measurements do not allow resolution of the C and M sub-layers. Values of Vs derived from these measurements are 25-30% greater than those measured at the same stress level and OCR in the laboratory, leading to Gmax values 55%-70% greater than the resonant column measurements. This difference can be ascribed primarily to differences in the age of the specimens. Consideration of the increase in modulus associated with the age of the deposit in the field yields a closer match between laboratory and field results, although it is found that values of the aging parameter (N G ) derived from laboratory measurements conducted following extended aging overpredict the increase in stiffness associated with geologic aging times.