Category WATER IN ROAD STRUCTURES

The permeability of asphaltic mixtures is controlled by the size and interconnec­tion of the void space. To illustrate this, some recent data is given in Fig. 5.1. This presents results for various hot-mix asphaltic specimens as described in Table 5.1. Figure 5.1 shows that permeability values of intact asphaltic materials are typically in the 0 to 40 x 10-6 m/s range. It is apparent that permeability is insignificant at less than approximately 7% air voids but can then rapidly increase. Probably this is be­cause interconnection of voids becomes possible at these high air void ratios and be­cause mixtures that exhibit such air void proportions may be inadequately prepared leaving permeable fissures in the material’s structure. Other authors (Zube, 1962; Brown et al., 1989 and Santucci et al., 1985) conclude that a limit of 8% air voids should be adopted to avoid rapid oxidation and subsequent cracking and/or ravelling and to keep permeability low.

Furthermore, higher permeability values are associated with asphaltic mixtures having larger voids. Larger voids are found both in fine-grained mixtures having high in-situ air void contents or in coarser grained mixtures at lower void contents. For example, Table 5.2 summarises the air void content at which a threshold be­tween essentially non-permeable and permeable behaviour was observed in-situ,

together with the permeability coefficient at that point (Cooley et al., 2001). The laboratory derived data of Fig. 5.1 tells a similar story although with a threshold void content of 8-9% for a permeability of 10 x 10-6 m/s. Both data sets reveal that the coarse, low-fines mixtures are least well-performing. In Fig. 5.1, Mixture 3, the poor performance is seen in the rapid increase in post-threshold permeability, while in Table 5.2 the threshold air-void content at which permeability increases is much lower.

Andrew Dawson[7], Niki Kringos, Tom Scarpas and Primoz Pavsic

Abstract Pavement surfaces provide a key route of ingress of rain water into the pavement construction. Thus, permeability of asphaltic materials and the water ingress capacity of cracks in the pavement are very important. A range of equipment exists to determine the permeability of asphaltic mixtures both by in-situ and labo­ratory testing. Sometimes porous asphalt surfacing is provided to deliberately allow water into the pavement to limit spray from vehicles and to limit tyre-pavement noise generation. These porous surfaces can become clogged with fines, but reha­bilitating without causing premature damage is a challenge. Except for this planned acceptance of water into the pavement, water is generally undesirable as it often causes ravelling (stripping) of the asphalt whereby aggregate and binder separate. The mechanisms behind this separation are becoming better understood due to ad­vances in computational engineering and mechanical and physio-chemical testing.

Keywords Asphalt ■ cracking ■ infiltration ■ stripping ■ ravelling ■ porous asphalt ■ permeability

5.1 Introduction

The topmost layer of most pavements is comprised of a bound layer. The vast major­ity of pavements have an asphaltic surface. A far lower proportion have a Portland cement concrete (PCC) surfacing. Whilst it is usually a design aim of these surfaces that they provide an impermeable covering to all the lower pavement layers, water does penetrate such surfacings. It may do so either through intact, but not imperme­able, bound material or through cracks and joints in the surfacing. Although the em­phasis of this book is on water movement and its impact in the unbound material and subgrade layers of the pavement, some information on the movement and response in the upper, bound layers is indispensable. Apart from any other consideration, any
complete analysis of water in the road and foundation structure must consider the input conditions – which are significantly affected, perhaps even controlled, by the surface layer. In this chapter the emphasis is on asphaltic mixtures, their permeabil­ity and the damage that they suffer from water. Consideration is also given to the ingress of water through joints and cracks. Some of this may also be applicable to the ingress through construction joints and cracks in Portland concrete surfacings. In addition, some consideration is given to porous asphalt mixtures – surfacing that is designed to prevent runoff flowing over the surface.

Researchers interested in frost action in soils agree on the description illustrated in Fig. 4.3 on how ice lenses grow and cause frost heave. Nevertheless, when it
comes to the degree of water saturation of the unfrozen soil below the freezing front there are two different conceptions. Some researchers believe that the un­frozen soil is fully saturated while others believe it is unsaturated. Of course, these discrepancies in understanding lead to different explanations of the driving force of the capillary rise of water as well as different opinions on how to run laboratory experiments. Andersland and Ladanyi (2004), for example, Konrad and Morgenstern (1980) and Nixon (1991) give equations where it is obvious that full saturation is assumed for the unfrozen soil. Accordingly, Konrad (1990) refers to experiments where the specimen freezes from below and free access to water is permitted at the top. This, of course, gives full saturation of the unfrozen soil. In contrast to this view, Miller (1980) discusses frost heave as freezing of unsaturated soil and references experiments where water is fed in at the bottom of the specimen and ice lenses are fed through capillary rise in unsaturated soil. Accordingly, Penner (1957) freezes unsaturated soil and Hermansson and Guthrie (2005) present laboratory ex­periments where freezing and frost heave takes place at a height more than 0.5 m above the level of the water supply. It should also be noted that Hermansson and Guthrie (2005) describe testing where the specimen heaves significantly without addition of any external water at all. This, of course, contradicts the assumption that the soil below the freezing front is fully saturated. The expansion without addition of water is suggested to be an effect of air entering the soil.

In agreement with the laboratory experiments Hermansson (2004) described a field study where the depth to the groundwater table is 6 m. Under such a thoroughly drained condition, it is reasonable to assume that the soil is far from saturated. De­spite this Hermansson reported 80 mm of frost heaving over a period of 2 months.

The conclusions from these studies are twofold,

• Firstly, frost heave does not require full saturation; and

• Secondly, even a well drained soil might experience a significant frost heave.

In addition to the different understandings about the importance of saturation, there are also two different schools when proposing equations to describe frost heave (Hansson, 2005). One school neglects the liquid water pressure and the other one neglects ice pressure. The first school, characterised by “Miller-type” models, de­velops models describing the frost heave on a microscopic scale while the second school, characterised by “hydrodynamic” models, handles equations for the redis­tribution of water up to the freezing front, supplying the frost heave. No computer code is known that handles both processes realistically.

3.5 Conclusions

Heat transfer in soils involves convection, radiation, vapour diffusion and con­duction. For pavements, conduction is the most important factor. During warm and sunny summer days though, natural convection should not be neglected. The heat transfer is closely associated with water movements – evaporation pulls water through the soil to the evaporation surface. Freezing also drives water movement as water is drawn to the freezing front in soils which have moderate pore sizes and moderate permeability.

Frost susceptible soils always experience frost heave at freezing even if there is no saturation. Drains will lower the heave by reducing the water content but a frost susceptible soil will always hold enough water for a significant heave. Chapter 13 describes some drainage techniques that can help to address these prob­lems. Interested readers are also directed to the reports on frost and drainage, mostly in the context of seasonally frost affected roads, available from the ROADEX project (Berntsen & Saarenketo, 2005; Saarenketo & Aho, 2005).

Frost heaving of soil is caused by crystallization of ice within the larger soil voids. Ice lenses attract water to themselves by the, so-called, cryo-suction process, and grow in thickness in the direction of heat transfer until the water supply is depleted or until freezing conditions at the freezing interface no longer support further crys­tallization, see Fig. 4.3. As the freezing front penetrates deeper into the pavement, the growth of ice lenses ceases at the previous level and commences at the new level of the freezing front. At some point the heat flow will be reduced so that further freezing slows or the weight of overlying construction will impose sufficient stress to prevent further ice lens growth.

Fundamentally, the so-called cryo-suction process results from the consequences of the phase change of pore water into pore ice and the associated energy changes in the remaining pore water. As water arrives at the point of freezing, the soil skeleton expands to accommodate it while consolidation of the adjacent, unfrozen, soil may occur as water is pulled from that. The particular characteristics of the process are strongly affected by soil porosity, soil-water chemistry, stress conditions at the point of ice formation, temperature, temperature gradients in the adjacent soil(s), water availability, etc. Coussy (2005) gives a more detailed outline of the mass flow and heat flow elements that combine in the processes that are active at the freezing front.

Capillary water is crucial for frost heaving. The pavement damage range depends on the rapidity of freezing, i. e. if freezing occurs rapidly, ice lenses are distributed over a greater mass of soil, which is somewhat more favourable compared to slow freezing where the capillary inflow of water will cause a high concentration of ice lenses. Frost heave primarily occurs in soils containing fine particles (i. e. frost – susceptible soils). Clean sands and gravel are non-frost susceptible because they cannot hold significant pore suctions, so water cannot be drawn to the freezing front through them. Silty soils represent the greatest problem. For even finer soils, like clays, the pores are very small. These small pores lead to a low coefficient of per­meability which does not allow water to travel through the pores at a speed that would allow a fast capillary rise, hence ice lens formation is less.

3.4.1 Frost Heave and Spring Thaw

Frost heave occurs in roads having fine graded, so-called frost-susceptible, mate­rial, at a depth to which the freezing front reaches during the winter. A well-built road of consistent materials and cross-section can be expected to heave relatively evenly. When inconsistencies or inhomogeneities are found in the construction of the affected subgrade, then frost heave is likely to be uneven and may well cause an uneven road surface that results in reduced travelling speed and comfort.

Although such heave can be problematic, a much greater problem usually arises in spring-time when the ice that has formed in the road construction, which was instrumental in causing the frost heave, melts and results in a very high water content in the pavement and subgrade. The increased water content often means reduced bearing capacity. For this reason many countries impose spring-thaw load restrictions on low volume roads to avoid severe pavement deterioration.

The thermal diffusivity, a (m2/s), is the ratio between thermal conductivity (X) and thermal capacity (c):

a = X/c (4.3)

It, thus, measures the ability of a material to conduct thermal energy relative to its ability to store thermal energy. Soils of large a will respond quickly to changes in their thermal environment, while materials of small a will respond more sluggishly. From a physical point of view the thermal diffusion of a medium is indicative of the speed of propagation of the heat into the body during temperature changes. The higher the value of a, the faster propagation of heat within the medium.

For example, during sunny days the pavement surface temperature will show strong daily oscillation and in soils and pavement materials with a high thermal diffusivity this oscillation penetrates to a greater depth.

Thermal capacity characterises the ability of a material to store or release heat. It is the important property that relates to the delay in heat transfer. The thermal capacity of water is approximately twice as high as that for most minerals and for ice, while the thermal capacity of air is negligible.

The thermal capacity of saturated soils ranges between 800 and 1000 J/(kg°C) – that is between 2000 and 2400 J/(m3°C) – while dry soils exhibit values of between 300 and 1600 J/(m3°C).

Because the thermal conductivity of minerals is much higher than that of water and air, thermal conductivity of soil decreases with increasing porosity.

4.3.2 Degree of Water Saturation, Sr

The thermal conductivity of air in a soil or aggregate’s pores is negligible but the conductivity increases with increasing degree of water saturation.

Fine soils generally have a high porosity and a low quartz content and, conse­quently, the thermal conductivity of dry clay and silt is low. However, the fine pores of these soils more easily hold a higher amount of water and fine soils typically deliver thermal conductivities in the same range as other soils.

4.3.3 Temperature, T

The thermal conductivity of ice is four times higher than that of water and con­sequently the thermal conductivity of soils with a high degree of water saturation increases dramatically at or below freezing. Here it should be kept in mind, that fine soils at temperatures below 0°C can still hold a large amount of unfrozen water and that thermal conductivity increases, therefore, progressively with decreasing temperature. Coarse soils typically have low degrees of water saturation and thermal conductivity does not increase significantly at freezing.

A typical relationship between thermal conductivity and water and ice content is shown in Fig. 4.2. The relationships shown assume only ice or only water, respectively.

For pavements, conduction of heat is the most important factor for heat transfer. During warm and sunny summer days though, the temperature of a pavement base layer under a thin asphalt concrete, may reach high values and natural convection in a fairly permeable base layer should not, then, be neglected.

4.3 Thermal Conductivity, X

Mineral content, porosity, degree of water saturation and temperature affect the ther­mal conductivity of soils. The total conductivity is a function of the conductivity of each soil phase, solid grains, water and gas. Various equations for these mixtures have been proposed by Keey (1992) and Krischer (1963). Thermal conductivity values range between 1 and 4 W/m°C for saturated soils, and from 0.2 to 0.4 W/m°C for dry soils.

4.3.1 Mineral Content

Because thermal conductivity of quartz is 3-4 times higher than that of other min­erals the quartz content of a soil greatly affects the thermal conductivity. Typically, cohesive soils have a low quartz content while the quartz content of a fine sand is normally high.

4.2.5.1 Temperatures Below 0°C

The transfer of heat by conduction is the dominating factor at temperatures below 0°C (Sundberg, 1988). In the small pores of frost susceptible soils though, due to freezing point depression, some water remains unfrozen at temperatures below 0°C. This allows convection caused by so-called cryo-suction effects (see Section 4.6.2 below) and a small amount of heat transfer at temperatures below 0°C.

4.2.5.2 Temperatures Above 0°C and Below Approximately 25°C

At this temperature range, conduction of heat is still the dominating factor (Sundberg, 1988). In highly permeable soils there may be more forced convection – like groundwater flow that is natural or caused by water abstraction. High tempera­ture gradients in permeable soils may also cause significant natural convection.

4.2.5.3 Temperatures Above Approximately 25°C

For the lower temperatures in this range, conduction is still the dominating factor (Sundberg, 1988). At higher temperatures and relatively low water content, vapour transport gets successively more important. At saturation, heat conduction is always the dominating factor. Again, high temperature gradients in permeable soils may cause more natural convection. In coarse soils at high temperature, there will be more radiation but it will still play a minor role.