Mathematical Model to Correlate Frost Heave of Pavements with Laboratory Predictions

A mathematical model of coupled heat and moisture flow in soils has been developed. The model includes algorithms for phase change of soil moisture and frost heave and permits several types of boundary and initial conditions. The finite element method of weighted residual (Galerkin procedure) was ch...

Full description

Bibliographic Details
Main Authors: Berg,R L, Guymon,G L, Johnson,T C
Other Authors: COLD REGIONS RESEARCH AND ENGINEERING LAB HANOVER NH
Format: Text
Language:English
Published: 1980
Subjects:
Ice
Online Access:http://www.dtic.mil/docs/citations/ADA084737
http://oai.dtic.mil/oai/oai?&verb=getRecord&metadataPrefix=html&identifier=ADA084737
Description
Summary:A mathematical model of coupled heat and moisture flow in soils has been developed. The model includes algorithms for phase change of soil moisture and frost heave and permits several types of boundary and initial conditions. The finite element method of weighted residual (Galerkin procedure) was chosen to simulate the spatial regime and the Crank-Nicholson method was used for the time domain portion of the model. To facilitate evaluation of the model, the heat and moisture fluxes were essentially decoupled; moisture flux was then simulated accurately, as were heat flux and frost heave in a laboratory test. Comparison of the simulated and experimental data illustrates the importance of unsaturated hydraulic conductivity. It is one parameter which is difficult to measure and for which only a few laboratory test results are available. Therefore, unsaturated hydraulic conductivities calculated in the computer model may be a significant source of error in calculations of frost heave. The algorithm incorporating effects of surcharge and overburden was inconclusively evaluated. Time-dependent frost penetration and frost heave in laboratory specimens were closely simulated with the model. After 10 days of simulation, the computed frost heave was about 2.3 cm vs 2.0 cm and 2.8 cm in two tests. Frost penetration was computed as 15 cm and was measured at 12.0 cm and 12.2 cm in the two laboratory samples after 10 days. (Author)