Category WATER IN ROAD STRUCTURES

8.3.1 Frost Heave — Introduction

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. The frost heave typically causes an uneven road surface and results in reduced travelling speed and comfort. The main problem though usually arises upon thawing when ice lenses involved with the frost heave melt and result in high water content in the pavement. The increased water content often means reduced bearing capacity and spring-thaw load restrictions are imposed to avoid severe pavement deterioration.

Granular pavement layers normally show a substantial decrease in stiffness with increasing values of moisture. Once thawing commences in the spring season, the granular layers often reach a state of near-saturation that substantially reduces the load carrying capacity. During the winter, short thawing periods have similar effect, especially on granular base courses. Therefore, seasonal changes cause a significant variation in the ability of a pavement to support traffic loads. During the thawing

period, water is melted from the ice lenses and since the layers where the ice lenses are formed have high fines content, the stiffness can drop dramatically. Since the road thaws primarily from the surface downwards, the free water can not drain through the still frozen underlying layers. Another effect of frost heave is that the road layer where the ice lens was formed loses its compaction (density) which is gradually regained under traffic load.

Determination of frost-sensitivity of soils is generally carried out using frost heave tests, such tests are mainly used to classify soils, according to their frost – sensitivity. However, prediction of the mechanical behaviour of such soils in pave­ments (heave during the frost period or loss of bearing capacity during thawing) is much more complex, because it depends on the climatic and moisture conditions, and on the characteristics of the whole pavement structure.

The principal characteristic of stiffness variation due to environmental effects is given in Fig. 8.7. During winter short thawing periods can cause temporary de­creases in the aggregate base and sub-base stiffness. If the thawing penetrates down to the subgrade it also loses its stiffness. As the freezing starts again both layers regain stiffness. During spring thaw, the stiffness of the granular base and sub-base again lowers but the bearing capacity regains soon after the spring thaw period is over. However, if the subgrade has high fines content, it can take longer for the bearing capacity to recover.

Calculated stiffness values, based on measured deflections under loading of a pave­ment surface, for a thin pavement structure are given in Fig. 8.6, along with the water content. One can see that the spring-thaw period started in early April as the water content at the three probes increased from 4%-7% to 12%-16% in a very short period of time. When the water content in the lower part of the granular base reached its maximum value (15.2%), the stiffness of that layer reached its minimum value. As the water content during the summer period gradually decreased to 11%, the stiffness increased to its maximum value. The same trend was mainly true for the subgrade as well. The water content of the subgrade though reached its lowest value much later than the granular base and the recovery went on during the whole summer. This is probably due to the subsoil having much higher fines content than the base and the sub-base and, therefore, it takes much longer time for the water to dissipate from the subgrade.

ra

CL

«

з

3

■o

о c

0)

Increased use of dielectric sensors (see Chapter 3, Section 3.2.2) have permitted moisture assessments to be continued during cold-climate winters. By this means it has been observed that complete freezing of all pore moisture doesn’t necessarily occur in all the granular pavement layers, even though they are, nominally, within the frost-affected depth.

Accelerated load testing of pavements was done with the HVS-NORDIC at VTI in Sweden in 1998 (Wiman, 2001). Figure 8.5 shows the rut depth measurements for a weak pavement comprising a 49mm thick asphalt layer over a bitumen sta­bilised granular base of thickness 89mm over a sand subgrade 2.5 m thick (mean thicknesses).

After 500 000 passes the increase in rut depth was constant and only 0.88 mm/100 000 passes. Then it was decided to increase the test load from 60-80 kN

and the tyre pressure from 800 to 1000 kPa. The rut propagation increased but only to 1.03 mm/100 000 passes. The next step was to weakening the sub grade by adding water to the sand to bring the water table to a level 300 mm below the surface of the sub grade – the highest level permitted in the Swedish specifications when constructing new pavements. The test load was at the same time reset to 60 kN and a tyre pressure of 800 kPa. Now the rut propagation increased to 4.16 mm/100 000 passes and the first cracks could be seen at the pavement surface.

In the last ten years, significant progress has been made in the measurement of in-situ water contents in pavements, using in particular TDR probes (see Chapter 3, Section 3.2.2). These measurements have shown that, often, significant amounts of water infiltrate in to pavements through the pavement surface and from the shoulders.

Low traffic pavements are particularly exposed to water infiltration. Examples of moisture measurements on a typical flexible pavement (6 cm thick bituminous surfacing and granular base) are shown in Figs. 8.3 and 8.4. Figure 8.3 shows that the daily variations of water content in the granular base and in the clayey sub­grade (near the pavement edge) are important and strongly related with the rainfall. Figure 8.4 shows average water contents measured in the granular base, at different locations, near the centreline of the pavement and near the edge. The critical zone is clearly the pavement edge where the water content is about 2 percentage points higher than near the centreline. In this pavement, subjected to a mild oceanic cli­mate, seasonal variations of water content are low, but they can be more important with more continental climates.

Thick bituminous or cement-treated pavements are less permeable, and water infiltrates mainly when cracking develops, thus accelerating the deterioration. In such pavements, protection against water infiltration, by proper maintenance (crack sealing, renewal of the surface course) is one of the main concerns.

Changes in water content, especially excess moisture, in pavement layers com­bined with traffic loads and freezing and thawing can significantly reduce pavement service life. Failures associated with moisture are detected on roads all over the Europe. There is some evidence to suggest that water has less impact on thick and well-construced pavements than it does on thinner ones (Hall & Crovetti, 2007). It appears that in thicker pavements the effect of water may be more indirect than in thinner ones, reducing material stiffness leading to later distress.

To minimize the negative effects of moisture on pavement performance, first we have to identify the sources of infiltration of water. There are many different possible sources of water infiltration in pavement systems. Mainly, the presence of water in pavement is due to infiltration of rainwater through the pavement surfaces through joints, cracks and other defects, especially in older, somewhat deteriorated, pave­ments and shoulders. An important source is also migration of liquid water upwards to the freezing front. Water may also seep upward from a high groundwater table due to capillary suction or vapour movements, or it may flow laterally from the pavement edges and side ditches.

The significance of the routes of infiltration depends on the materials, climate, and topography.

Many pavement failures are the direct result of water entering the pavement courses and/or the subgrade. Water entry in the compacted unsaturated material will increase water pressure or decrease suction, and in turn, reduce the effective stress (see Chapter 9). Hence, the strength and the elastic and plastic stiffnesses of the pavement material and the subgrade will be reduced. The rate of traffic-induced deterioration of the road will increase during this time. The loss of strength and stiffness can lead, in the extreme, to rutting and other forms of surface deformation or, more commonly, to pavement edge failures. The worst situation occurs with poorly-compacted granular material (as a result of shear strength reduction) with frost susceptible soils or with cohesive swelling clays (as a result of damaging vol­ume changes).

Water seeping through the pavement can also transport soil particles and cause erosion and pumping (i. e. transport) of fines as well as leaching of many materials (see Chapter 6). Moisture entry can also affect the performance of the surface course by causing stripping of bitumen from aggregate, layer separation between bound courses, and pothole formation (see Section 5.5). A probable mechanism of pothole formation is illustrated in Fig. 8.1. A somewhat similar failure mechanism has also been observed in slabbed concrete pavements (Roy & Johnson, 1979): [21]

(ii) Stage 2 Water travels along the interface between the asphalt (AC) and the compacted aggregate base

(iii) Stage 3 Asphalt lifts to allow pressure to dissipate, but also allows more water to be pumped into the crack

Fig. 8.1 Layer separation and pothole formation (from Gerke, R. J. (1979), cited by Lay (1986)). Reproduced by permission of M. G. Lay

• the fast-moving water eroded the finer materials;

• at their edges, the concrete slabs progressively lost support from the underlying layers as material was washed away;

• the deflections of the slab became greater and the erosion more rapid eventually leading to cracking of the unsupported concrete at edges and corners of slabs; and

• dirty water was seen to squirt from the joints in the pavement when the slabs were trafficked.

Pavement damage, associated with water, can be divided into moisture-caused and moisture-accelerated distresses. Moisture-caused distresses are those that are primarily induced by moisture, while moisture-accelerated are those that are initi­ated by different factors, but the rate of deterioration is accelerated by presence of water. Most commonly observed damages due to the moisture are:

• surface defects;

• surface deformations; and

• cracking.

Generally, water threatens the stability of soil and it is a particular problem for pavement structures since they are built through areas of changeable moisture quantities.

It has been generally established that the increased subgrade water content during spring results in increased deformability, i. e. a decreased bearing capacity of the pavement. These changes in bearing capacity are, in particular, obvious for silty and clayey materials.

Water contents contained in materials under flexible pavements are influenced by the amount and intensity of rainfall. Periods of long rainfall of low intensity can be more severe than concentrated periods of high intensity, since the amount of moisture absorbed by the soil is greatest under the former conditions. Also, the combined effects of rainfall and freezing temperatures determine, in part, the extent of pavement damage.

Water can penetrate into the pavement structure in several different ways as fol­lows (Fig. 8.2):

• Seepage from the elevated surrounding soil, which depends on the hydraulic gradient and soil permeability coefficient.

• Rise and fall of the phreatic surface, which depends on the climatic circum­stances and soil composition (e. g. following heavy rains there is an increase in the sub-surface water level in permeable strata, while water remaining on the surface of impermeable ground will drain away or evaporates and there is no risk of a rapid rise in sub-surface water level).

• The penetration of water through damaged pavement surfaces causes high lo­cal concentration of water in penetration areas. Under heavy traffic load, such a condition can result in significant damage if the subgrade made of changeable material is in contact with water. Even worse pavement damage can result from freezing of the structure and subgrade when soaked with water.

• The penetration of water through shoulders (if the shoulders are permeable or their surface is deformed in such a way as to allow the retention of water), which depends on the material permeability, compactness of the surface, inclination and the drainage from pavement surface. The effect is similar to the effect of water penetration through a damaged pavement surface.

RAINWATER INFILTRATION

THROUGH SHOULDER

V V

CONDENSATION

CAPILLARY WATER

RISING

і LM lit

OF WATER I ABLE

Fig. 8.2 Possibilities of water movements into the pavement zone (from Moris & Gray, (1976) cited by Lay (1986)). Reproduced by permission of M. G. Lay

• Capillary rise from the foundation soil. The water rises from the foundation soil through the fine-grained soil up to the pavement structure.

• Evaporation of water from the foundation soil and its condensation under the pavement structure if the pavement structure is colder than the soil.

Moisture behaviour in a pavement can be considered as occurring in three phases:

• an entry phase – which occurs quite rapidly;

• a redistribution phase when water moves within the material in response to suc­tion and gravity; and

• an evaporative phase when water, as vapour, leaves a material or moves to other layers. Water vapour movements can occur under temperature gradients with the water vapour travelling from a warm to cool area when it then condenses.

Water content under road pavements will vary seasonally, annually and over longer periods. Seasonal variations in water content are commonly located in the upper 1-2 m, and in the metre or so of pavement width at the edge of the pave­ment surface (the outer wheel path is clearly the critical zone). Significant moisture changes will only occur immediately after rainfall if very permeable layers exist.

The behaviour of the pavement related to moisture should be considered in ref­erence to the climate, type of soil, groundwater depth and moisture concentration in the soil.

Most actual pavement design methods are based on the same principles. Linear elas­tic calculations are used to determine the stresses and strains in the pavement layers, under a reference traffic load, and then the calculated stresses and strains are com­pared with maximum allowable values, depending on the nature and characteristics of the pavement materials. Most usual design criteria are the following:

• For bituminous layers: a fatigue criterion, often limiting the tensile strain at the bottom of the bituminous layer to prevent upward cracking in thin bitumi­nous layers. For thicker asphalt layers, where the cracking may be from the top-downwards due to aging, traffic and thermal effects, fatigue relationships will need correlating to actual cracking performance if this calculation route is selected;

• For cement-treated layers: a fatigue criterion, limiting the maximum tensile stress at the bottom of the treated layer; and

• For subgrades; a criterion limiting the vertical elastic strain at the top of the layer, to avoid risks of rutting.

Thus, pavement design is based on elastic calculations and on the application of a limited number of design criteria (mostly fatigue criteria). Calculations are generally performed with constant material properties (corresponding to the initial characteristics of the materials after construction) and a design traffic, defined by an equivalent number, NE, of standard axle loads (ESALs). For thicker asphalt pavements and pavements containing cement-treated layers, unless there has been a specific calibration between the causes of pavement deterioration, such as age­ing and deterioration of materials and climatic effects (temperature and moisture variations, frost), then these factors will need assessment before the safety of the road pavement and embankment can be secured. The effect, either way, is likely to increases the road cost by over-design of the structure.

Design against frost is generally based on the evaluation of the frost sensitivity of the subgrade (by frost heave tests, swelling tests, or on the basis of empirical classifications). When the subgrade is sensitive to frost, a thermal propagation model is often used to determine the thickness of protective material needed to reduce penetration of frost into the subgrade. However, swelling of frost-sensitive soils during freezing, or loss of bearing capacity during the thaw period are generally not taken into account, except in some countries (some examples are given in the Annex).

8.2.1 Different Types of Road Structures Versus Sensitivity to Water

Road pavements are multilayer structures (see Fig. 1.5) generally comprising a sur­face course and one or more asphaltic or granular base layers, resting on a pavement foundation. Chapter 1 introduced the major pavement layers – the foundation, the sub-base, the pavement base and a surfacing (see Section 1.4.2).

Water permeability should normally increase from the top of the pavement (the asphalt or concrete layers) downward until about 0.7 m depth (see Chapter 5). Oth­erwise water would accumulate onto the low permeability layer and keep the upper layer wet; freezing of the accumulated water might then unbind the upper layer. This would decrease it’s bearing capacity and service life. An exception to these rules are the porous asphalt surfaced pavements described in Chapter 5, Section 5.7, which are designed to carry water within their thickness.

Pavement structures can be divided in four main groups:

• Thin bituminous pavements, which consist of a relatively thin bituminous sur­face course, resting on one or more layers of unbound granular materials. These are typically used for carrying low traffic levels.

• Thick bituminous pavements consisting of a bituminous surfacing, over one or two bituminous layers/asphaltic concrete (AC) (base) then an aggregate (sub-base). Their application is, typically, for high traffic levels. They may be considered as flexible, but they are much stiffer then the preceding pavements.

• Semi-rigid pavements comprising a bituminous surfacing over one or two layers of materials treated with hydraulic binders (e. g. concrete). This type of pavement is also appropriate for high traffic levels.

• Portland cement concrete (PCC) pavements which consist of a Portland ce­ment concrete slab (15-40 cm thick), possibly covered with a thin bituminous surfacing, resting on a sub-base (bound or unbound), or directly on the foun­dation. The concrete slab can be continuous with longitudinal reinforcement, or discontinuous. Once again, this type of pavement is also appropriate for high traffic levels.

These various pavement types present different types of mechanical behaviour, and different deterioration mechanisms. However, for all structures, water plays a major role in pavement deterioration.

In thin bituminous pavements, high stresses are transmitted to the unbound granular layers and to the subgrade; and lead to permanent deformations of these layers. Because unbound layers and subgrades are sensitive to water content, the performance of these pavements is strongly dependent on variations of moisture conditions. This can lead, in particular to “edge failures”: water infiltrates from the pavement shoulders, under the edge of the pavement, leading to subsidence at road edges. As these pavements are very sensitive to moisture, impermeability of the surface course and good drainage are very important for their performance.

In thick bituminous pavements, the much lower flexibility of their bituminous base layers means that the stresses transmitted to the soil are much lower, and the risk of permanent deformations in the soil, as well as the sensitivity to the wa­ter content of the soil, are lower. The main mechanism of deterioration of these pavements is cracking due to the combined effects of traffic-induced strains and thermally-induced movements, causing high tensile stresses in the bound layers. Once the pavement is cracked, water infiltration accelerates the degradations, lead­ing to weakening of the subgrade, attrition at the lips of the cracks, and material chipping away to form potholes (see Chapter 5). Without maintenance, deterioration can lead rapidly to total ruin.

In pavements with layers treated with hydraulic binders, the main deteri­oration process is generally due to reflective cracking. Thermal contraction and shrinkage in the cement treated materials create transversal cracks. These cracks generally rise up through the surfacing and appear on the pavement surface at fairly even spacing (5-15 m). These cracks tend to deteriorate and split under traffic loads. Then again, water infiltration is a major problem, leading to a deterioration of the bonding between bound layers, a decrease of the bearing capacity of the subgrade in the cracked area, thus decreasing load transfer and favouring attrition of the crack lips. On these pavements, protection against water infiltration, by using relatively thick surface courses and by sealing the shrinkage cracks, is essential;

In concrete pavements, due to the high modulus of elasticity of concrete, only very low stresses are transmitted to the foundation. Thermal cracking in concrete structures is generally controlled by transverse joints, or by the longitudinal rein­forcement, producing only very fine micro-cracks. Two main types of damage are observed:

• Cracking created by excessive tensile stresses at the top or base of the slabs due to the combination of traffic loads and deformations of the slabs due to thermal gradients; and

• Reduction in bearing capacity around joints and cracks, leading to pumping phe­nomena. This reduction is essentially due to the presence of water at the interface between the slab and the sub-base. Under loading by vehicles, water at the inter­face is locally highly pressurised, high pressure gradients appear, inducing high water flow velocities which can erode the sub-base material (near a pavement edge crack or joint), reduce the bearing capacity of the support and reduce load transfer between the slabs. This is generally observed as edge or corner cracking.

For all pavements, freezing and thawing phenomena are also a major source of deterioration. In frost sensitive, fine grained soils, freezing leads to a concentra­tion of water near the frozen zone (due to the so-called cryo-suction process – see

Section 4.6). This leads to heaving of the pavement, and then loss of bearing capacity during the thaw period. It should also be noticed that in cold climates, where winter tyres with studs are used, the wear of the surfacing of the pavement and the consecutive re-paving is often faster than fatigue or deformation damage.

and Pavement Performance: Field Observations

Robert Charlier[20], Pierre Hornych, Mate Srsen, Ake Hermansson, Gunnar Bjarnason, SigurSur Erlingsson and Primoz Pavsic

Abstract This chapter presents a mechanical behaviour study, i. e. the bearing capacity as a function of the moisture degree. The field point of view is expressed and the chapter summarises a number of observations on road behaviour, in relation to variations of moisture. First, the road structure is recalled with respect to the mechanical analysis point of view. Then some observations onfield under temperate climate, humid, are given. In a second step, the specific case of frost and thawing are discussed.

Keywords Bearing capacity ■ moisture level ■ field measurement ■ stiffness ■ rutting ■ thawing

8.1 Introduction

This chapter introduces the study of mechanical behaviour, i. e. it seeks to describe bearing capacity as a function of the moisture degree. This description is based on a summary of road behaviour with respect to variations of water content as observed in-situ.

Initially, the road structure is presented from the point of view of a mechanical analysis. The mechanical and hydraulic specific behaviour of each subgrade layer is discussed. Next, the chapter briefly analyses water penetration and the water effect on the road structure layers before illustrating the water-induced mechanical effects. The change of water content over time is shown for specific locations that are sub­jected to a humid, temperate, climate. Then a consideration of the development of ruts, i. e. the strains and deformation, with time is given. A discussion of the stiffness of a road structure is also given as it relates to the moisture level. A clear decrease of the stiffness and an increase of the strain accumulation are observed when the moisture increases.

In the second part of the chapter, specific cases of frost and thawing action are discussed. In particular, the stiffness decrease during and after thawing is described. Field measurements of temperature, of layer moisture and of deflectometer stiffness are presented.

Despite most regulatory constraints being based on physico-chemical analysis, the hazard toward the natural environment represented by a contaminated solution or matrix cannot simply be assessed on the basis of the single analytical approach. The latter supposes that the contaminants can all be identified and are not too numerous (which is not always the case), but moreover, the chemical concentration does not provide any information about phenomena of synergy or antagonism between pollu­tants, and does not provide information about the toxicity towards living organisms (criterion H14 of the European Directive 91/689). Biological methods can do so (ADEME, 1999). The purpose of these methods is to assess the eco-toxicological danger of solutions and matrix. They are carried out in vitro on biological species chosen for their sensitivity to pollution (Ramade, 2000).

Eco-toxicity tests can also be carried out from solids thanks to extraction tech­niques (see Section 7.6.1 on leaching and percolation, above).

Biological test analyses range between classical tests on organisms measuring survival to tests on cells and enzymatic activity.

Water from different parts of the road pavements and embankments and their surrounding environment may be analysed for toxicity to plants, animals, fish and humans. The methods used for collection of the water for this purpose will be as for collection for analyses for chemical compounds. It is, however, especially important that the water quality does not change during the toxicity test. Therefore, it must be kept cool and in dark, and quickly transported to the laboratory (see Section 7.4.6).

The classical tests for deciding toxicity, biological degradation and bioaccumu­lation are tests according to international standards (OECD Guidelines, ISO). The tests use living micro and macro-organisms (plants, animals) or cell cultures to characterise the toxicity of tested single chemical compounds or mixtures of com­pounds. In vitro methods use cells or enzymes and proteins for the testing of single compounds or complex mixtures.

When assessing the environmental effects, the test solution is often subjected to several test organisms such as algae, crustaceans and fish to search for differences in the sensitivity of organisms at different trophic levels of the ecosystem.

Eco-toxicity tests (Table 7.4) may be classified as acute tests or chronic tests. Acute tests are tests with effects showing within a short time. A classical acute test is the measurement of the survival of organisms. The results are recorded as the concentration at which half the number of test organisms survive/die during the test period (LC50, Lethal Concentration). If the test period is 96 h the concentration referred to will be 96 h LC50. The chronic tests are conducted during a longer pe­riod at lower test compound concentrations. The end point is not death, but some secondary sub-lethal effect.

7.7 Concluding Remarks

This chapter presents a general overview of water and soil sampling and analy­sis in the road environment. The main principles of data collection and storage, and methodologies for sampling design are presented. Furthermore, water and soil sampling procedures as well as in-situ and laboratory measurements and analyses methods are described, with an elucidation about their usefulness, potentialities and fields of application.

It is intended that the information presented in this chapter, as well as the biblio­graphic material that is referenced at its end, can provide a sufficient and valuable base from which the reader can consider the best choices for contaminant sampling and analysis methodologies, accordingly to the purpose of his/her investigation, and considering the abilities of available methods and tools as presented above.

Historically, chemical analysis of water was achieved by titration methods. It is, practically, impossible to use these on water containing a pollutant at a low con­centration due to the need to collect a very large volume of water that can be con­centrated to permit a weighable amount of chemical to be obtained at the end of the procedure. Also, such procedures are very time-consuming and operator sensi­tive. Therefore, modern analysis is based on electrical, atomic and spectrographic techniques.

The most common analytical techniques suitable for determination of the pres­ence of selected pollutants are synthesised in Table 7.3.