Sally A. Shoop

Moisture migration during freeze and thaw of unsaturated soils: modeling and large scale experiments

Abstract A coupled heat flow and moisture flow model (FROSTB) was used to simulate large-scale freeze-thaw experiments to assess its ability to predict soil moisture conditions. The experimental data consist of temperature and soil moisture profiles measured during freeze-thaw cycles in a 1-m layer of frost-susceptible silty sand over roughly 2 m of gravelly sand. Two experimental conditions were modeled: (1) where the soil was fairly wet and the water table was shallow (1 m below surface), and (2) where the soil moisture was lower than specific retention and water table was deep. Overall, the model predicts the frost penetration and heave quite well; however, it tends to overpredict ice formation. The additional ice in the modeled frozen soil then causes a slower thaw. Matching the (total) moisture contents in the drier soils results in underpredicting the heave. The discrepancies stem from the model representation of the physical processes involved in freezing (and frost heave) of unsaturated soil. We propose improvements through using a “pseudo” three-phase flow potential and calculating volumetric segrated ice content starting at 90% of saturation. The effects of changing the constants related to hydrologic properties are also discussed.

Effect of test method on winter traction measurements

Abstract Traction on winter surfaces was measured using three instrumented vehicles, each designed to measure traction for a different purpose: vehicle mobility research (CRREL instrumented vehicle), commercial tire testing (Uniroyal-Goodrich traction tester), and airport runway safety (Saab friction tester). The traction measured with each method is comparable but there are systematic differences due to the effects of the surface materials and test and analysis techniques. This comparison serves as the basis for collaboration between the various traction testing communities and illustrates the need for well documented, standard test and analysis procedures for traction testing and evaluation.

All-Season Virtual Test Site for a Real-Time Vehicle Simulator

Abstract : A virtual, all-season test site for use in real-time vehicle simulators and mobility models was constructed of an Army firing range in Northern Vermont. The virtual terrain will mimic the terrain of our Virtual Data Acquisition and Test Site (VDATS) at Ethan Allen Firing Range (EAFR). The objective is to realistically simulate on- and off-road vehicle performance in all weather conditions for training and vehicle design for the US Army. To this end, several spatial datasets were needed to accurately map the terrain and estimate the state-of-the- ground and terrain strength at different times of the year. The terrain strength is characterized by terramechanics properties used in algorithms to calculate the forces at the vehicle-terrain interface. The performance of the real vehicles will be compared to the simulated vehicle performance of operator-in-the-loop and unmanned vehicles for validation of the simulations. Real vehicles are instrumented and perform maneuvers at the test site to develop and validate mathematic models describing vehicle behavior in all-season conditions, including snow, ice, frozen and thawing ground.

Thawing soil strength measurements for predicting vehicle performance

Abstract The CRREL Instrumented Vehicle (CIV), shear annulus, direct shear and triaxial compression devices were used to characterize the strength of thawed and thawing soil. Strength was evaluated in terms of the Mohr-Coulomb failure parameters c ′ and φ′, which can be used in simple models to predict the tractive performance of vehicles. Use of an instrumented wheel (like those of the CIV) is proposed for terrain strength characterization for traction prediction because the conditions created by a tire slipping on a soil surface are exactly duplicated. The c ′ and φ′ values from a portable shear annulus overpredict traction because of the curved nature of the soil failure envelope in the region of low normal stress applied by a portable annulus. Of all the tests, the direct shear test yielded the highest φ′ value, due to its slow deformation rate and drained conditions. The triaxial test produced results closest to those of the instrumented wheel. For all methods, φ′ increases with soil moisture but decreases rapidly beyond the liquid limit of the soil. The φ′ measured with the vehicle was also found to be strongly influenced by the freeze-thaw layering of the soil.

Rapid stabilization of thawing soils: field experience and application

Abstract Thawing soils can severely restrict vehicle travel on unpaved surfaces. However, a variety of materials and construction techniques can be used to stabilize thawing soils to reduce immobilization problems. The US Engineer Research and Development Centers Army Cold Regions Research and Engineering Laboratory (CRREL) and the Wisconsin National Guard evaluated several stabilization techniques in a field demonstration project during spring thaw at Fort McCoy, Wisconsin, in 1995. Additional tests on chemical stabilizing techniques were conducted at CRRELs Frost Effects Research Facility. The results of these test programs were reduced to a decision matrix for stabilizing thawing ground, and used during the deployment of US troops in Bosnia during January and February of 1996. The soil frost and moisture conditions expected during this time frame were predicted using MIDFROCAL (MIDwest FROst CALculator). This paper is an overview of the stabilization techniques evaluated and their recommended application based on the expected soil frost conditions and traffic requirements. Although the experiments were performed with military vehicles in mind, the techniques are suitable for many civilian applications such as forestry, construction, mining, and oil exploration.

Comparison of Finite Element Model (FEM) Data and Single Point Layered Elastic Model (SPLEM) Data of a C130 Operating on a Frozen Runway Structure

A 3-D Finite Element Model (FEM) was constructed of a loaded C130 tire rolling over a frozen unpaved runway. The runway was constructed of two layers, one frozen and one unfrozen, each of varying thickness, but the combined thickness was fixed. The material model used in the dynamic FEM simulation represents a frostsusceptible soil, which was used in full scale unpaved road tests at CRREL’s Frost Effects Research Facility (FERF), and was calibrated with triaxial tests and validated against direct shear test data. The area of interest in both models is the interaction between the frozen (top) layer and the unfrozen (bottom) layer. Stress and strain data collected from the dynamic FEM simulation and a prior single point layered elastic (SPLEM) simulation were compared, along with the capabilities of each model. Initial findings show that the single point layered elastic model is much faster at obtaining results for a single point, but it can only solve problems with pure elastic material. The FE model can solve problems with any material and the results can be viewed at any location or point in time during the run. The FEM visually represents what is happening in the soil around the tire as it rolls along the surface, while the static SPLE model only predicts what is occurring directly below the tire.

CAP PLASTICITY MODEL FOR THAWING SOIL

A material model for soft, wet soil was generated to simulate the deformation behavior of thawing soil under vehicle loading on pave d and unpaved roads. The soil modeled, a frost-susceptible fine sand called Lebanon Sand, was subjected to a full suite of saturated and unsaturated triaxial testing duplicating conditions experienced during large-scale freeze–thaw testing. Material parameters were generated for a capped Drucker–Prager plasticity model with hardening. These were calibrated in triaxial test simulations using the commercial finite element code ABAQUS. The material model was then implemented in several three-dimensional finite element simulations for validation and robustness. THE PROBLEM OF SPRING THAW Spring thaw is a critical time of year for the deteri oration of roads and airfields. Deformation of the road surface (either paved or unpaved) during spring thaw is nearly always the result of defo rmation of the thawing layer, which is weakened though a reduction in density and an increase in moisture as a result of the freezing process. The deformation of the loose, weak, thawing layer is largely plastic, consisting of both compaction and shear. Because of the difficult and time -consuming nature of constructing full -scale test sections and field experiments on thawing ground, we aimed to create a finite element modeling capability, validated with experimental data, which would then be useful for performing computer experiments on a wide range of off-road, pavement, and

Effects of Military Vehicle Trafficking on Vegetated Soils

Numerous studies have investigated the effects of vehicle trafficking on terrain and how soils and soil conditions affect the mobility of military vehicles. The majority of these studies however were conducted on non-vegetated soils. The purpose of our study is to investigate the effects of heavy vehicle trafficking on vegetated soils and to assess the impacts of vegetation, specifically grass, on vehicle mobility. The research program includes a series of experiments assessing the effects of trafficking, mowing, and burning on vegetated soil strength. Three test sections were constructed and planted with perennial ryegrass: one section representing outdoor field conditions and two controlled indoor sections (sand and clayey-loam). Mobility parameters of motion resistance and traction were collected in each test section prior to trafficking by a large military vehicle (HEMTT). Before and after trafficking, each test section was characterized including soil strength, moisture content, soil density, and terrain disturbance. The results show that vegetation affects soil strength and thus the terrain impacts of trafficking. Additionally, the treatment of vegetation affects soil strength, especially in clay materials. This paper summarizes the soil condition, soil strength, and vehicle impact results of the first year of a four-year study. Future years will assess the recovery of the vegetation in the tested areas with the ultimate goal of making recommendations for the treatment of vegetated military training lands.

Modeling Heavy Wheel Loading on Frozen and Unfrozen Ground

Two ABAQUS 3-D Finite Element Models (FEM) were constructed with hyper-elastic rubber tires loaded on unfrozen and frozen ground conditions (0, 7.62, 15.24, 22.86, and 30.48 cm frost depths), one contained a statically loaded tire, and the other contained a tire rolling at 8.05 km/h. The ground was constructed of multiple layers, allowing each of the respective frost depths to be adequately modeled. This was accomplished by changing the material properties in the layers to match the given frost condition. The material models used in the FEM simulations are representative of a frost-susceptible soil, which was used in full scale unpaved road tests at CRREL. The interest in this modeling is to see how the stress and displacement change as a function of frost depth as a heavy tire rolls over the surface. Stress and displacement data collected from the ABAQUS FEM simulations are compared to experimental data. The results show similar trends between the ABAQUS and experimental data for vertical displacement and vertical stress below the contact patch.

Application of historical mobility testing to sensor-based robotic performance

The USA Engineer Research and Development Center (ERDC) has conducted on-/off-road experimental field testing with full-sized and scale-model military vehicles for more than fifty years. Some 4000 acres of local terrain are available for tailored field evaluations or verification/validation of future robotic designs in a variety of climatic regimes. Field testing and data collection procedures, as well as techniques for quantifying terrain in engineering terms, have been developed and refined into algorithms and models for predicting vehicle-terrain interactions and resulting forces or speeds of military-sized vehicles. Based on recent experiments with Matilda, Talon, and Pacbot, these predictive capabilities appear to be relevant to most robotic systems currently in development. Utilization of current testing capabilities with sensor-based vehicle drivers, or use of the procedures for terrain quantification from sensor data, would immediately apply some fifty years of historical knowledge to the development, refinement, and implementation of future robotic systems. Additionally, translation of sensor-collected terrain data into engineering terms would allow assessment of robotic performance a priori deployment of the actual system and ensure maximum system performance in the theater of operation.