Keywords

1 Introduction

In general, the structure of asphalt pavement is made up of layers laid one over the another. Bituminous materials of many types can be used to create the layers. For both private and commercial public users, the main purpose of the road pavements is to offer a strong, comfortable, and secure traveling surface in all weather circumstances. Over the course of its design life, a well-built pavement structure should be able to deliver these structural and functional qualities with the least maintenance.

Numerous researchers have shown the impact of interlayer bonding on pavement durability through experimentation and numerical simulations [1, 2]. To keep the asphalt structure monolithic and extend its service life, adequate bonding between the asphalt layers is required. The service life of the pavement is shortened if the connection between the layers is insufficiently strong and the pavement layers work independently. However, the pavement may instantly fail due to heavy traffic. The RILEM Technical Committee 206-ATB [3] and [4] state-of-the-art reports indicate that the following factors could affect the bonding between road asphalt layers:

  • the climate at the time the pavement is cast. As a general rule, the difference in temperature between the top and lower layers plays an important role;

  • reduced interlayer bonding can be caused by contamination, the presence of dirt as dust or oil from milling or construction traffic on top of the bottom layer [4, 5];

  • the water flow between the layers; in comparison to dry conditions, water flow in the pavement causes a decrease of interlayer resistance due to surface cracking [6, 7];

  • absence of bituminous emulsion at the interface of the two layers [8].

The connection between pavement layers can be strengthened by a variety of approaches. Chemical bonding (tack coatings, emulsions) and mechanical bonding (milling, tinning, etc.) are two ways to improve bonding [9].

As shown in Fig. 1, the function of each layer in the flexible pavement structure is to reduce and dissipate the stresses to a level that the layer below can support.

Fig. 1.
figure 1

Relative strength of the vertical stresses caused by the load at different depths [10].

Full size image

A wide range of approaches have been proposed by numerous studies for evaluating the strength of the link between pavement layers. In the 1970s, the interface bond strength between pavement layers containing penetration-grade bitumen as a tack coat at various temperatures was measured using a direct shear test known as an interface shear mold [11]. Uzan et al. [12] proposed a systematic test method (direct shear test) for evaluating the bond shear strength between the layers having stress-absorbing interlayers. Swiss Federal Laboratories for Materials Testing and Research produced a standard technique (Swiss Standard SN 671 961) for obtaining bond strength using a Swiss LPDS tester throughout Europe [13]. Torsion tests were also recommended by Roffe et al. [13] as a method of assessing interface bond strength. Using the Superpave Shear Tester, it was examined the interface between layers using different types of tack coats, application rates and temperature conditions [14]. A basic direct shear device was created by the Florida Department of Transportation (FDOT) and can be tested with Marshall Stability apparatus or a Universal Testing Machine (UTM) [15]. To evaluate the shear properties of the interface under various surface and temperature conditions, the Ancona Shear Testing Research and Analysis (ASTRA) equipment was developed in Italy [16].

Finite element (FE) software systems, such as Abaqus, introduced in the mid-1990s, provide additional options for pavement simulation. It has been demonstrated that Abaqus can be used to solve challenging pavement issues.

2 Case-Study

The road sector analyzed is a county road in Romania with climate type I, hydrological regime 2a, and foundation soil type P5. Table 1 shows the layers of the road structure of the investigated road sector, namely the deformability characteristics of the asphalt mixtures and the foundation layers; E - dynamic modulus of elasticity and µ - the Poisson coefficient and the bulk density of each layer, according to the Romanian norm PD 177 [17]. Considering the mentioned characteristics, the pavement subjected to design is a flexible pavement designed as a multi-layer system.

The analyses presented herein show the importance of the connection degree between the layers of the road structure by examining its reaction to traffic loads. The traffic loads have an imposed value of 0.90 m.s.a. The numerical investigations were employed to examine two dimensioning hypotheses: (i) the hypothesis of perfectly bound layers and respectively (ii) the hypothesis of semi-bonded layers.

Table 1. Pavement subject to design.
Full size table

3 Numerical Model

To model the road section, a 3D model was built by using the finite elements program Abaqus, in order to follow, with more precision, the distribution of the strains in the entire pavement layers (especially in the bituminous layers), in different bound hypothesis. The dimensions of the modelled elements (in plan) are the following: 1000 mm length × 500 mm width × 1500 mm height (for the complete model). Figure 2 shows the general geometry of the pavement layers used in the analyses.

The bottom surface of the foundation layer and the sides of the other layers are fixed tying the nodes in horizontal and vertical directions. Modelling the interaction between the layers of the road structure was done by considering a surface-to-surface-contact interaction.

Fig. 2.
figure 2

Road pavement geometry.

Full size image

The layers of the analyzed road structure with high dynamic modulus of elasticity values were considered as master surfaces. Those with lower values of dynamic modulus of elasticity were considered slave surfaces. The interaction between layers was influenced by a friction coefficient considering a unitary value in the hypothesis of perfectly bonded layers and 0.5 in the hypothesis of semi-bonded layers between the wearing course and the binder course.

The load considered in the analyses is introduced according to the Romanian standard 115 kN axle, which has the following characteristics:

  • static load applied on twin wheels: 57.5 kN;

  • contact pressure: 0.625 MPa;

  • radius of circular area equivalent to the tyre-to-road contact area: 0.171 m.

The analysis of the road section at standard axle stress involves the calculation of specific deformations and stresses at critical points of the road complex, characterized by a maximum stress. The following verification steps were considered in the design:

  • verification of the structure in terms of vertical specific strain (εr) at the base of bituminous layers;

  • verification of the structure in terms of vertical specific compressive strain (εz) at the formation level.

The 10-node quadratic tetrahedron (C3D10) finite element with reduced order numerical integration, is used to model every layer of the pavement. The solid element (C3D10) has three degrees of freedom at each node and can represent large deformation, geometric and material nonlinearity. Figure 3b shows the total mesh model. Global mesh size was 25 for each meshed layer and a maximum deviation factor with 0.1 value. Figure 3 shows the boundary conditions used in the analysis and the circular contour of the pressure load as well as the mesh FE before loading.

Fig. 3.
figure 3

(a) Load and boundary conditions; (b) Mesh model.

Full size image

4 Numerical Results

4.1 Perfectly Bound Layers Hypothesis

The FE results for perfectly bound layers are shown in Fig. 4, Fig. 5 and Table 2. Units in the figures are in MPa. The admissible values (R.D.O.adm and εzadm) were calculated using PD 177 [17]. R.D.O. is the ratio between the design traffic in m.s.a. and the number of permissible stresses, in m.s.a., that can be supported by the bituminous layers, corresponding to the deformation state at their base.

All the FE values were compared with the values resulting from the dimensioning of the road structure analyzed by considering the Romanian road structure dimensioning program, Calderom 2000. And it appears that the deformations and stresses were identical. The results show that under the assumption of perfectly bonded layers, the checks required by the Romanian norm PD 177 [17] are fulfilled.

Fig. 4.
figure 4

Radial deformation at the base of bituminous layers.

Full size image
Fig. 5.
figure 5

Vertical specific compressive strain at the foundation ground level.

Full size image
Table 2. Numerical results.
Full size table

4.2 Semi-bonded Layers Hypothesis (BA 16 50% Bonded to BAD 22.4)

For the present study, a partial bond between the BA 16 wearing course layer and the BAD 22.4 binder course was considered. This 50% bond was applied in the FE software by considering a friction coefficient of 0.5. The interaction between the other layers was “tie”. Figure 6, Fig. 7 and Table 2 show the FE modelling results.

Considering the road structure according to the above condition, it was found that the R.D.O. value resulting from the modelling with semi-bonded layers does not check for the design traffic of 0.90 m.s.a. The resulting vertical strains εz are higher than the admissible values εzadm, The εzadm value is calculated as a function of the design traffic Nc. In these conditions, only with a designed traffic load of 0.22 m.s.a. will satisfy the criterion. This means that the initial traffic should be reduced by approximately 76%. Thus, in order to obtain an adequate design, either the design traffic should be reduced or the stratification should be thicker.

Fig. 6.
figure 6

Radial deformation at the base of bituminous layers.

Full size image
Fig. 7.
figure 7

Vertical specific compressive strain at the foundation ground level.

Full size image

5 Conclusions

The aim of this study was to determine whether the road pavement with deficiencies of adhesion between asphalt layers can take over the calculated traffic load considered in the study, or the proportion of design traffic load that can be taken over.

The paper shows that the layer adhesion of road asphalt layers directly affects the bearing capacity of the road structures and, as a result, its service life as the number of standard axle crossings. The results show that the reduction of the connection degree between the asphalt layers to 50% leads to a reduction of the design traffic by approx. 76% as proven by the present case-study. All the values resulting from the dimensioning with semi-bonded layers check for a reduced design traffic of 0.22 m.s.a., traffic determined for a service life of 3.67 years, whether its evolution shows a consistent increase over the whole service life.

In the context of the presented study, it is of interest to evaluate other degrees of interactions, by considering different values for the friction coefficient and different type of loads (ex. Moving load together with braking force).