Date of Award


Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Master of Science (MS)


Civil Engineering


Dana N. Humphrey

Second Committee Member

Thomas C. Sandford

Third Committee Member

Willem Brutsaert


Pavement performance resulting from current installation practices for shallow cross pipes (culverts) has been problematic for departments of transportation in northern climates, such as Maine. Shallow cross pipes are located within the seasonal frost zone, typically ranging from 1.2 to 2.4 m (4 to 8 ft) below ground surface in Maine (Floyd, 1978). Differential heave of shallow cross pipes beneath Maine's highways can result in premature pavement failure reducing serviceability and quality of ride. The type of heave developed at cross pipes can also become a safety issue for the traveling public. In some cases, the heave remains after thawing, resulting in permanent displacement of the road surface. The cause of differential heave is generally attributed to discontinuities within the subgrade although variable access to water and exposure to the sun also contribute. Discontinuities such as the installation of shallow cross pipes create differences in the soil properties and/or thermal regime between the cross pipe trench and the adjacent soil. The objective of this study was to evaluate the performance of shallow cross pipes insulated with extruded polystyrene insulation as a method to distribute differential heave over larger distances. Research consisted of field testing, instrumentation, and numerical modeling of insulated and uninsulated cross pipes at five selected sites across Maine. The effectiveness of the installation methods with extruded polystyrene insulation placed above and below the cross pipe were evaluated. Cross pipe installations were evaluated based on: influence of insulation, effect on pattern of frost penetration, surface frost heave, influence of soil type, and regional climatic effect. Experimental cross pipes were installed between May and October 2004 along Routes 11 in Nashville Plantation, 27 in Eustis, 117 in Buckfield/South Paris, 150 in Cornville, and 166 in Castine/Penobscot. A total of 15 experimental cross pipes were selected, reconstructed, instrumented, and monitored. Four types of test methods were constructed: (1) underlying insulation, (2) overlying insulation, (3) overlying insulation for very shallow cross pipes, and (4) control. Experimental cross pipes were instrumented with thermocouples, to measure the temperature profile in the vicinity of the cross pipe and frost penetration, and frost-free elevation benchmarks, to assist with measuring vertical displacement of the wearing surface. The thermocouples indicated that sections with insulation above the cross pipe significantly reduced the frost penetration into the underlying soils. The reduction in freezing temperatures limited the amount of frost action and subsequent total heave directly over the cross pipe. However, from the heave surveys, most pipes that develop heave have greater total heave away from the cross pipe than over the pipe resulting in greater differential heave. Results showed for sites that experienced significant total heave, the placement of insulation over the cross pipe increased the both the magnitude and abruptness of the differential heave. This showed that the length of the transition insulation on each side of the cross pipe was inadequate. The maximum differential heave was observed to occur at the end of the freezing season and start of the spring thaw.