Modification of the U.S. Army Corps of Engineers Model 158 for Prediction of Frost–Thaw Profiles in Northern New England

Abstract

Knowledge concerning frost–thaw profiles is important in the design of pavements for low-volume roads located in seasonal frost areas. In addition, because these roads are highly susceptible to damage during the spring thaw–weakened period, this knowledge is helpful in making decisions about seasonal load restriction (SLR) policies. Direct measurement of frost depth is expensive and usually limited in scope. Therefore, several highway agencies have expressed the need for a prediction model to help them in estimating frost–thaw profiles beneath roadways. A report from the U.S. Army Corps of Engineers New England Division examined frost prediction models. One of the equations in that report was Model 158, which was similar to the modified Berggren equation. Model 158 uses air temperature indexes as well as pavement material properties to integrate heat flow into the calculation of maximum seasonal frost depth. That model has been modified slightly and programmed in an Excel spreadsheet to predict frost and thaw depths on a daily basis. This paper presents an overview of the modified Model 158 and compares predictions obtained from that model with measured frost–thaw profiles obtained at several test sites located in northern New England during a period of five freeze–thaw seasons. Results suggest that the model shows much promise although it generally tended to slightly underpredict maximum frost depths. In tracking the thaw process for SLR posting, the model tended to be conservative in estimating end-of-thaw dates (especially during rapid thawing events); in many instances, estimated end-of-thaw dates fell after measured dates.