Physical modelling of climate-soil-infrastructure interactions of paved roadways constructed in expansive soil

Abstract

Roadways founded on expansive soils are notorious for incurring damage in the months and years after construction. Soil swelling and shrinking manifest at the road surface as differential displacements of the pavement and eventually lead to linear cracks developing parallel with the direction of traffic flow. This climate-soil-infrastructure interaction occurs over daily, seasonal, and annual cycles that cause variations in moisture, temperature, relative humidity, and precipitation. Design recommendations often call for expansive soil to be compacted 1–2% wet of optimum to limit pavement damage regardless of the local climatic conditions or soil swell potential. Despite the prevalence of expansive soils found worldwide and the significant damage they cause to overlying infrastructure, very few studies appear in the literature that include field monitoring. To overcome this lack of physical data and examine the traditional design specification centrifuge modelling was undertaken to reproduce a stress-appropriate environment and control the moisture boundary conditions. These centrifuge experiments examine the effect of as-placed moisture content on the response of a two-lane paved road profile subjected to wetting–drying cycles. The centrifuge model results provide high temporal and spatial resolution displacement measurements using Digital Image Correlation. Cumulative subsurface strains are reflected at the road surface and cause detectable differential displacements of the model paved road. The relatively minor variation of initial moisture content in the three tests examined (varied from 29 to 34%) was found to have a significant impact on the hydraulic-mechanical response expressed in the isolated ground located beyond the influence of the road as well as the pavement-ground interface. The experimental results provide new insight into climate-soil-infrastructure interaction under a paved road, which would otherwise be unobservable using traditional discrete measurement techniques in the field. Interpretation of the high-density displacement measurements allowed for calculation of unique curvature / bending envelopes for the pavement under the different as-placed moisture conditions. The wet of optimum model displayed 5-times greater curvature / bending compared to the optimum and dry of optimum tests. Compacting wet of optimum may not be uniformly beneficial to limit pavement damage and construction specifications should consider local climate and swelling potential.