Modeling and field evaluation of the strength of surface soils for vehicle mobility
Date
2019
Authors
Pauly, Matthew J., author
Scalia, Joseph, advisor
Niemann, Jeffrey D., advisor
Green, Timothy R., committee member
Butters, Gregory, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Surficial soil strength is a critical variable in vehicle mobility and terrain trafficability analysis and varies substantially in time and space with soil moisture and texture. Fine-resolution (5-30 m grid cell) patterns of soil strength and soil moisture are necessary for routing of off-road vehicle operations and must be estimated for applications when direct measurement is too expensive, labor-intensive, or dangerous. Rating cone index (RCI) is the in-situ method typically used in mobility applications to empirically evaluate the strength of surficial soils. The RCI method provides one simple parameter to evaluate soil trafficability, but in doing so fails to separately characterize the various mechanisms (compressibility, stress independent shear strength, stress dependent shear strength) that govern soil behavior in relation to vehicle traffic. Alternatively, the Bekker soil strength framework, which encompasses pressure-sinkage and shear strength soil properties, offers a mechanics-based representation of soil behavior and has received increased interest from the terramechanics community in recent years. However, because RCI has been the focus of the terramechanics community over several decades, predictive relationships to estimate Bekker parameters using basic spatially- and temporally-variable input data (soil moisture and soil composition) do not exist. The objective of this study is to develop and evaluate a framework for prediction of Bekker parameters (cohesion and friction angle) as a function of soil moisture and soil texture (percentage of sand and clay). A model, termed the Strength of Surface Soils (STRESS) model, is introduced to estimate shear strength of surface soils using soil moisture, pedotransfer functions based on soil texture, and unsaturated soil mechanics. The STRESS model is paired with an existing soil moisture downscaling model, the Equilibrium Moisture from Topography, Vegetation, and Soil (EMT+VS) model. The pre-existing EMT+VS model includes two untested simplifications that make the model inconsistent with the STRESS model framework, so two previously neglected soil-related hydrologic considerations are introduced to the EMT+VS model: runoff and residual water content. The impacts of runoff and residual on soil moisture downscaling performance and spatial patterns of soil moisture are assessed at a test region in northeastern Colorado called Drake Farm with measured soil moisture data for model calibration and evaluation. The additions are successfully included in the EMT+VS model but the assumptions made in the pre-existing EMT+VS model are shown to be adequate for soil moisture downscaling. After assessing EMT+VS model additions, the STRESS model is applied to Drake Farm to produce spatial patterns of estimated friction angle and cohesion. Model estimates are compared to measured shear strength using a human-powered shear strength bevameter to evaluate the predictive capability of the STRESS model. The model is found to underpredict friction angle and overpredict cohesion at Drake Farm due in part to the use of class-average effective shear strength parameters that do not appear to adequately reflect the properties of surficial soils. Finally, the design and construction of two bevameters are summarized for field and laboratory measurement of Bekker parameters. The results of laboratory tests on the human-powered shear strength bevameter used in STRESS model evaluation are compared to traditional geotechnical strength testing to validate field-testing results and ensure repeatability of measurements. Additionally, the design and construction of a fully automated, laboratory-focused bevameter device with pressure-sinkage and shear strength testing capabilities are described, but this bevameter is not used for testing in this study.
Description
Rights Access
Subject
soil strength
unsaturated
terramechanics
soil moisture