Modeling of frost heave and surface temperatures in roads

Abstract: Temperature and moisture are very essential parameters when describing the condition of a pavement. In most cases, a high moisture content involves a decreased bearing capacity and, consequently, a shorter durability of the pavement. A frozen pavement has a greater bearing capacity than the corresponding construction in spring or late autumn. However, the freezing itself also implies strains to the pavement, as it heaves to different extent and in different directions in connection with the frost heave. The properties of an asphalt concrete pavement vary dramatically according to temperature. Cold asphalt concrete is hard, stiff and brittle, and therefore, cracks easily occur, whereas its bearing capacity decreases at high temperatures as softening progresses. Emphasizing the asphalt concrete, a numerical model has been developed for calculation of temperatures for summer condition, by means of recorded values for solar radiation, air temperature and wind velocity. Further, in order to also model temperatures and other conditions, occurring in the pavement during winter, a frost heave module has been developed and included in the model. The aim of this is to gain a better insight into the freezing process of a road structure. The model also provides an efficient tool for a better understanding of important factors related to frost depth and frost heave. A modified version of the model, is tested for falling weight deflectometer analysis. Input here, is a series of measured pavement surface temperatures and the output is calculated temperature distributions for the asphalt layer. Measuring equipment, developed at VTI, has been used to, in the field, automatically record frost heave and pavement temperature distribution. Furthermore, equipment for freezing tests in laboratory has also been developed. Experiences from such tests and field measurements have been used when developing the numerical model for freezing of pavements. At the laboratory freezing tests, a special interest has been devoted to water intake rate and heave rate related to cooling rate. The experiences, obtained from both the laboratory tests, as well as the field observations, have been compared to what has been reported in literature. Temperatures obtained from the numerical model for summer and winter temperatures have turned out to correspond well to measurements of pavement temperatures at all test sections studied, 12 in US and 3 in Sweden. The freezing tests in laboratory have shown, that a strong frost heave can exist without addition of external water to the samples. The natural water content is, consequently, sufficient to provide enough water for the heave. This "in-situ" water can be redistributed in the structure, thus providing water to the frozen portion of the profile to cause significant frost heave. Frost heave caused by a process like this, is not bound to uptake of external water, which normally is assumed in the relevant literature. Frost heave in freezing tests is often explained by 10 % volume expansion of the freezing water, which is sucked up by samples during the test. The freezing tests in laboratory have also shown, that the ratio (heave/water uptake) increases with frost susceptibility, i.e. the most frost susceptible soils require comparatively less added water for the heaving. The freezing tests in laboratory and field tests, have also shown that the relation between heave rate and heat extraction rate at the frost front, is weak. This is also in contradiction to what normally is stated in literature. The weak relation is found during longer periods of continous frost penetration. However, when heat extraction rate is varied at short time intervals, a strong dependency between heave rate and heat extraction rate is found.