Moisture movement in highway pavement structures coupled with soil-atmospheric fluxes

The overall performance of a standard highway pavement structure depends on moisture characteristics of the underlying soil layers. Increase in the moisture content of the sublayers caused by moisture fluxes across the soil-atmosphere interface, decreases the strength of the pavement structure, which may cause premature failure of the structure. Therefore, the ability to evaluate the surface fluxes may be helpful in understanding mechanisms, which may enhance or degrade highway pavement performance. This research evaluates the application of the soil-atmosphere modelling software VADOSE/W as a tool for predicting the movement of moisture in highway pavement structures. VADOSE/W is a two-dimensional transient finite element program that simulates coupled heat and moisture migration in unsaturated soils with particular focus on fluxes across the soil-atmosphere interface. A typical standard highway pavement structure in Saskatchewan was chosen to evaluate coupled heat and moisture interactions between highway pavement structures and atmosphere, and the impact that design features may have on moisture movement. A laboratory testing program was established to characterize the material properties of hot mix asphalt (HMA), which is used as a surface layer for the driving lane as well as on occasion for the shoulder of highway pavement structures. HMA was characterized by its saturated hydraulic conductivity, soil-water characteristic curve, vapour flux rate and air permeability. The saturated hydraulic conductivity defines the maximum rate at which water can infiltrate HMA in the absence of cracks. Drastic changes to the saturated hydraulic conductivity of HMA can significantly increase or decrease the amount of infiltration during critical storm durations. The volumetric water content of HMA decreases rapidly at relatively low values of suction suggesting that HMA is either relatively hydrophobic or contains cracking of the internal structure such that it demonstrates very low air entry values. The 'pore spaces' of the HMA are likely only partially filled with water following drainage. The vapour flux rate of HMA defines the maximum evaporation rate through HMA in the absence of cracks. HMA produces a negligible amount of evaporation during the summer period compared with the amount of infiltration. The measured and calculated air permeability results for HMA were quite different indicating that problems might have occurred during the testing process. Some of the possible problems include air bubbles in the manometer, air leakage, and not allowing the flow meter to come into equilibrium. A numerical modelling component evaluated the mechanisms of coupled heat and moisture flux into the pavement structure when using six different design features, which have the ability to either enhance or degrade performance. The six design features include: varying the fluxes on the HMA surface; changing the shoulder conditions from unpaved to paved shoulders; changing the steepness of the sideslope; using both good and poor vegetation conditions; varying the initial suction conditions; and varying the snow removal process during the winter season. Each of the eight numerical simulations was simulated for twelve years in order to reach long-term equilibrium conditions. The results of annual cumulative boundary flux for each of the numerical simulations indicates that long-term equilibrium conditions for the highway pavement system have yet to be established after the twelve years of simulation. However, overall the cumulative flux rates are slowly changing from year to year. Comparing each of the three divisions of the top boundary indicates that the flow through the driving surface (i.e. driving lane and shoulder) is quite negligible. The granular area of the sideslope is generally moving water into the system due to the granular materials along the surface. The organic area of the sideslope (i.e. ditch area) is generally moving water out of the system for all eight cases. The modified numerical simulation design features can cause a positive or negative impact on the moisture and heat fluxes of highway pavement structures. The design feature that has a positive impact is applying a vapour diffusion rate to the paved surface (Case 5). Paving the shoulder (Case 2) also has a positive effect. These features produce less moisture to accumulate in the subgrade layer over the entire climatic season. The design features that create a negative impact include applying poor vegetation conditions to the sideslope and ditch areas (Case 3), steepening the sideslope (Case 6) and allowing snow to accumulate along the shoulder (Case 8). These three design features cause more moisture to accumulate within the subgrade layer in comparison to that of the base simulation case. Further studies are required before the research is extended to engineering practice. The six most important recommendations include: 1. Addressing the convergence and water balance issues within the numerical models themselves. 2. Enhancing the mesh to provide more realistic highway pavement geometry. 3. Improving the initial conditions for a more realistic hydrogeologic setting. 4. Refining subgrade layer to include: (a) a difference in the recompacted layer from the compacted layer and (b) a separate upper layer that will be affected by freeze-­thaw. 5. An indepth investigation into the ability to apply actual HMA properties from laboratory studies along with pertinent climate information required within VADOSE/W. 6. Laboratory testing of the pavement sublayers (i.e. base, subbase and subgrade layers)