United States Department of Transportation
i
Measurement of Temperature and Soil Properties for Finite Element Model Verification [2010]
-
2010-08-01
[PDF-26.14 MB]
Details:
-
Creators:
-
Corporate Creators:
-
Corporate Contributors:
-
Subject/TRT Terms:
-
Resource Type:
-
Geographical Coverage:
-
Corporate Publisher:
-
Abstract:In recent years, Alaska Department of Transportation and Public Facilities (AK DOT&PF) personnel have used TEMP/W, a commercially available two-dimensional finite element program, to conduct thermal modeling of various embankment configurations in an effort to reduce the thawing of ice-rich permafrost through thermally stable embankment designs. The overall goal of this study is to ground-truth the thermal modeling results produced by TEMP/W. To achieve this goal, we collected soil samples and installed thermistor strings to measure ground temperatures at two research locations: Richardson Highway mile post (MP) 113, and Dalton Highway 9 Mile Hill. Each site was chosen based on ongoing AK DOT&PF work in the area, fine-grained foundation soils, and the presence of permafrost. We conducted both standard AK DOT&PF laboratory testing and specialized testing (i.e., thermal conductivity, dry unit weight, volumetric water content) on the collected soil samples. At each research location, we installed thermistor strings at the shoulder and toe of the embankment and in an undisturbed area, forming a cross-section to the highway centerline. The thermistor strings extended to 30 ft below the ground surface. For the Richardson Highway location, analysis of the measured temperatures indicated that the foundation soils are not frozen directly below the highway embankment to a depth of at least 30 ft. In the undisturbed area, the active layer thickness is 6 ft, and the average permafrost temperature is 31.0°F. Analysis of the Dalton Highway location temperature data indicated that the foundation soils are frozen immediately below the highway embankment. The active layer in this area ranges between 6 and 12-ft thick. In the undisturbed area, the active layer is approximately 2-ft thick, and the average permafrost temperature is 30.1°F. For each site, we developed a thermal modeling mesh based on a measured surface profile and the drilling data, and ran several iterations of the model for each site with different input parameters. For the initial models, input parameters were chosen from published values, calculated from the AK DOT&PF laboratory results, or retrieved from the internet. In subsequent models, we systematically changed the input parameters, such as replacing the estimated thermal conductivity values with the results from the specialized testing, and replacing the air temperature boundary function derived from historical data from the internet with measured air temperatures at each site. The results of this study indicate that air temperature is the most critical input parameter. The results of the initial model iteration for each location indicated that the modeled temperatures were several degrees Fahrenheit colder than the measured temperatures. Using the air temperatures measured at each research site resulted in much better agreement between the measured and modeled temperatures, with differences typically less than 2°F at a depth of 30 ft. The modeled temperatures in the upper 10 ft of soil at each location demonstrated more deviation from the measured temperatures. The magnitude of these deviations was reduced by adjusting the n-factors used in the surface boundary condition; however, it is important to keep in mind that while fine-tuning the model by adjusting the n-factors was possible for this research project, typically site-specific soil temperatures are not available for comparison with model results.
-
Format:
-
Collection(s):
-
Main Document Checksum:
-
Download URL:
-
File Type:
