Soutenance mémoire
Pavement structure
The basic function of a pavement is to reduce the stresses induced by traffic at the subgrade level. The latter would indeed be unable to bear the stresses by the repeated traffic loading.
The basic parts of a flexible pavement structure are:
Surface layer: It is the top layer that comes in contact with traffic loads. It may be composed of one or several different HMA sublayers (Pavement interactive, 2010). It can be divided into three layers: the HMA wearing course, the HMA binder course and the HMA base course. The HMA wearing course is in direct contact with the tires and it ensures the security and the comfort of the users because it helps to provide a well bonded surface free from loose particles which might endanger people. The HMA base course protects the granular layer underneath by providing mechanical protection (load transfer), thermal (attenuation of the temperature variations), and hydrous (water proofing). The composition of the HMA mix in the binder or base course can be modified by including RAP to up to 40%. Between those two course, a binder course is often used for ease of construction;
Base course: The base course is the lowest asphalt layers and it’s composed of granular materials, treated or not. It is subdivided in two layers: base and subbase, its essential function is to distribute the loads induced by traffic to be compatible with the subgrade (Laveissiere, 2002). The subbase layer is used in areas where the subgrade soil is extremely weak. The subbase course functions like the base course. The material requirements for the subbase are not as strict as those for the base course since the subbase is subjected to lower load stresses. The subbase consists of stabilized or properly compacted granular material (The constructor, 2015). The binder course is located between the surface layer and the base course and its use for its bearing capacity;
Subgrade: The subgrade is the natural soil under the pavement. Usually, the top of the subgrade is graded and compacted to rectify heterogeneities and to increase the bearing capacity.
Asphalt mixtures
The asphalt mixtures are composed of various selected aggregates (around 95% by mass) bound together into a cohesive mass with asphalt cement (around 5% by mass). According to Di Benedetto and Corté (2004), there are different compositions of asphalt mixtures that can be used according to their function and location within the pavement structure. Nevertheless, it is always the aggregate skeleton and the asphalt that will determine their mechanical and functional characteristics (Di Benedetto et Corté, 2004). Asphalt cement binds the aggregate particles together, enhancing the stability of the mixture and providing resistance to deformation under induced tensile, compressive and shear stresses. Bitumen materials are viscoelastic and their mechanical behaviour is dependent on both the temperature and rate of loading. At low temperatures and short loading times, asphalt cements behave as elastic solids, while at high temperatures and long loading times they behave as simple viscous liquids. At intermediate temperatures and loading times, the behaviour is more complex. A medium temperature range from 15 to 30°C is most suitable for fatigue cracking analysis and pavement fatigue life prediction (Deacon et al., 1994).
Recycling rate of recycled asphalt mixtures
The recycled asphalt mixture is mainly defined by the mass of RAP that they comprise. This mass is commonly called “percentage of RAP”. Recycled mixtures containing up to 20% of RAP are said to be “low rate of recycling”. This denomination is opposed to the asphalt mixtures containing between 40% and 65% of RAP which are then indicated by the term “high rate of recycling”. This arbitrary division does not preclude of recycling at intermediate rates (Berthier, 1982). Since there is no strict definition of low and high rate, we will present the following proposed definition which is taken from the French experiment of recycling at high rate.
Low level recycling
At low level of recycling, the properties of a bituminous mix are barely influenced by the composition (gradation, binder content and binder grade) of the RAP (Moneron et Measson, 2004; Li et al., 2008). Therefore, on a regulatory level, it is not necessary to take into account the composition of RAP and its influence on the produced recycled mixture manufactured when less than 10% of RAP is used (Gandil et Vesseron, 2001). This rate is increased to 20% in the case of binder courses. This provision of law specified in circular 2001-39 of June 18, 2001 aims to encourage using RAP to reduce the volume of waste from road industry.
Recycling rate higher than 10% – 20%
At higher rates of recycling, the composition of RAP must be taken into account because of the presence of stiffer (aged) binder in the RAP material and its influence on hot mix asphalt.
Hence, because of the influence of aging, the HMA mixture with RAP may become vulnerable to durability cracking and premature failure. The composition of RAP is then studied in order to determine the characteristics of the RAP aggregate skeleton and the RAP asphalt cement (Daniel et Lachance, 2005).
Designing HMA with high percentage of RAP
In the middle of 1990s, the implementation of the Superpave method of mix design started. The original specifications for Superpave did not include guidance on how to include RAP into the new mix design system. Interim recommendations were developed through the FHWA Asphalt Mixture Expert Task Group (FHWA Superpave Mixture Expert Task group, 1997) based on experience and performances of Marshall mixes with RAP. The specifications were changed in 2002 after the results of an NCHRP research project (Incorporation of Reclaimed Asphalt Pavement in the Superpave System) became available (Mc Daniel et al., 2000). AASHTO Standards MP2 (now M323) (AASHTO, 1999), Standard Specification for Superpave Volumetric Mix Design for Hot Mix Asphalt, describe how to design HMA with RAP.
The guiding principle of the AASHTO standard is that mixtures with and without RAP should satisfy the same requirements. The aggregates provided by the RAP are included in the determination of the gradation of the mixture and in the consensus properties (coarse aggregate angularity (CAA), fine aggregate angularity (FAA), flat and elongated particles (F&E particles) and the sand equivalent value which is waived because of the inability to test. The bitumen contained in the RAP is regarded as part of the total binder content of the mixture.
The Federal Highway Administration (FHWA) and its Superpave Expert Task Groups developed a guide (draft) for RAPs usage based on past experience. These guidelines established a tiered approach for the use of RAP. Under those guidelines, up to 15% of RAP can be used without changing the virgin binder grade (McDaniel et al., 2000). When between 15 and 25% RAP is added, the high and low temperature grades of the virgin binder are both reduced by one grade. For more than 25% RAP, blending charts are used (McDaniel et al., 2000).
Extraction and recovery of asphalt binder
It is important to determine asphalt content, properties of RAP binder, and aggregate gradation, in the design of mixtures containing RAP. The only way to get the above information is to separate the asphalt cement from RAP. The most widely used methods are solvent extraction and ignition oven, which can determine both binder content and aggregate gradation (Pavement interactive, 2015).
In the solvent extraction method, a solvent such as trichloroethylene or ethylene chloride is used to dissolve and separate the binder from the aggregate. Asphalt cement is then calculated from the mass difference before and after extraction. In the ignition method, the mix sample is heated to around 540°C for about 45 minutes until all the asphalt is burned off.
The difference in mass before and after ignition is determined as the asphalt content. There are some disadvantages of the solvent extraction method such as it has a high standard deviation of test results (Brown et Murphy, 1994). Peterson et al. (1999) showed that the amount of binder content extracted differs by 0,3% to 9,5% when comparing different extraction methods using solvent. However, the solvent extraction method is banned in many countries because Trichloroethylene is hazardous to both man and environment. According to Kandhal et al (1995), the ignition method is accurate and precise. However, the ignition extraction method causes degradation of aggregate because of combustion of RAP in the oven which changed the properties of the aggregate (Prowell et Carter, 2000). This degradation can also lead to erroneous estimates of the binder content with some aggregates, especially for RAP sources with unknown correction factors. Therefore, ignition oven should be allowed only if it is calibrated with clean aggregate. Abson recovery method (ASTM D1856-09, Standard Practice for recovery of asphalt from solution by Abson method (ASTM. 2009)) and Rotavapor method (ASTM D5404/D5404M – 12 (ASTM. 2012)) can be used for recovery of asphalt cement from solvent.
Binder test and aging
The performance of the mixture containing RAP is known to be dependent on the change in the properties of binder and RAP due to aging. The aging is reflected in the change in the rheological properties of asphalt.
The following factors are reported to contribute to age hardening of asphalt during mixing and/or during the service life (Robert et al., 1996):
Oxidation through diffusive reaction of oxygen in the air with asphalt binder; Volatilization through evaporation of the lighter components from asphalt binder. It usually does not contribute to long term aging in the pavement and is primarily a function of temperature; Polymerization through chemical reaction of molecular component; Thixotropy due to the formation of a structure within the asphalt binder over a period of time; Syneresis due to the exudation of lighter constituents of the binder; and Separation through the removal of oil constituents and resins by absorptive aggregates.
The rheological behaviour of aged binder will differ from virgin materials as asphalt binder reacts and loses some of its components during the aging process which effects on the PG grade. Thus, designing a mix with RAP should be done with care. If the old binder is too stiff, the blending of old and virgin binders can’t function as envisaged. With small percentages (up to 20%), an aged binder does not influence the properties of the blend of virgin and RAP binder significantly (Kennedy, Tam et Solaimanian, 1998).
It was found that as the amount of RAP binder increases, the stiffness of the mix increases (Souparth, 1998). Lee, Terrell and Mahoney (1983) conducted a study on the evaluation of the mechanical and rheological properties of asphalt binders containing binders from RAP. In their study, two virgin binders (PG 58-28 and PG 64-22) were blended with 0, 10, 20, 30, 40, 50, 75, and 100% reclaimed asphalt binders obtained from two different stockpiles. RAP and virgin binders were tested using dynamic shear rheometer (DSR) according to AASHTO TP5-98 (AASHTO, 1998). It was observed that the RAP binder is stiffer than the virgin binder for about 10 times more. Also, it was found in the same study that the RAP binder obtained from one asphalt plant RAP stockpile were 10 times stiffer than those for a RAP binder obtained from a different RAP stockpile located in the same region. Moreover, there was variability in stiffness of a RAP binder from the same source but with different magnitude.
|
Table des matières
INTRODUCTION
CHAPTER 1 BITUMINOUS MIXTURES USED IN PAVEMENT STRUCTURE AND RECYCLING
1.1 Pavement structure
1.1.1 Flexible pavements
1.2 Asphalt mixtures
1.2.1 Aggregate skeleton
1.2.2 Asphalt binder
1.2.3 HMA volumetric properties
1.2.4 Benefit of recycling asphalt pavement
1.2.5 Asphalt recycling methods
1.3 Recycled mixtures
1.4 Composition of recycled mixtures containing RAP- definition
1.5 Recycling rate of recycled asphalt mixtures
1.5.1 Low level recycling
1.5.2 Recycling rate higher than 10% – 20%
1.6 Designing HMA with high percentage of RAP
1.7 Extraction and recovery of asphalt binder
1.8 Determining RAP aggregate properties
1.8.1 RAP aggregate gradation
1.8.2 RAP aggregate specific gravity
1.8.3 Coarse aggregate angularity (CAA)
1.8.4 Fine aggregate angularity (FAA)
1.8.5 Flat and elongated particles
1.8.6 Hardness/wear
1.8.7 Cleanliness
1.9 Binder test and aging
1.10 Aging methods of HMA
1.11 Production of hot mix asphalt
1.11.1 Recycling at drum mix plants with parallel-flow or counter-flow drying (cold addition)
1.11.2 Recycling at a batch plant
1.11.3 Special plants
CHAPTER 2 MECHANICAL PROPERTIES OF ASPHALT MIXTURES: BEHAVIOUR AT SMALL DEFORMATION
2.1 Introduction
2.2 Solicitation on bituminous roadways
2.2.1 Traffic effect
2.2.2 Temperature effect
2.3 Bituminous materials behaviour classification
2.4 Application to the pavement structure design
2.5 Linear viscoelastic behaviour
2.6 Measurement of the modulus of asphalt mixtures in the frequency domain: complex modulus
2.6.1 Definition and measurement’s principle
2.6.2 Complex modulus tests
2.6.2.1 Tension-compression test on cylindrical sample
2.6.3 Alternative methods of analyzing complex modulus test results
2.6.3.1 Isothermal and isochronal curves
2.6.3.2 Cole – Cole plane (complex plane)
2.6.3.3 Black space
2.6.3.4 Time temperature superposing principles and shift factor
2.6.4 Mechanical models used in linear viscoelastic
2.6.4.1 Presentation of the 2S2P1D model
2.7 Fatigue phenomenon of bituminous mixtures
2.7.1 Introduction
2.7.2 Hot mix asphalt fatigue
2.7.3 Laboratory fatigue test
2.7.4 Fatigue test analysis methods
2.7.4.1 Criterion based on the evolution of the stiffness modulus (classical method)
2.7.4.2 End of phase II criterion
2.7.5 Determination of the initial modulus E0∗
2.7.6 Approaches used to predict failure in fatigue
2.7.6.1 Fatigue curve or Wöhler curve
2.7.6.2 Fatigue cracking prediction modelling
2.7.7 Method of analysis in terms of damage (the DGCB approach)
2.7.8 Performance of RAP mixtures in laboratory: complex modulus tests
2.7.9 Fatigue performance of RAP mixtures in laboratory
2.8 Chapter conclusion
CHAPTER 3 RESEARCH METHODOLOGY AND EXPERIMENTAL WORK
3.1 Introduction
3.1.1 Research Methodology
3.1.2 Task 1 – literature review
3.1.3 Task 2 – Materials and Mix design and Preparation
3.1.3.1 Design of mixture with no RAP
3.1.3.2 Design of mixture with RAP
3.1.4 Task 3 – Laboratory experimental testing
3.1.4.1 Complex modulus test
3.1.4.2 Fatigue resistance test
3.1.5 Task 4: Improvement of the basic fatigue model .
3.1.6 Task 5: Fatigue life prediction performance of RAP mixtures using the new models
3.2 Experimental work
3.2.1 RAP sampling
3.2.2 Volumetric properties of the mixtures
3.2.3 RAP insertion in mixes
3.2.3.1 Hot recycling with cold RAP
3.2.3.2 Hot recycling with reheated RAP
3.2.4 Mixing temperature
3.2.5 Mix selection
CHAPTER 4 MATERIALS CHARACTERIZATION FOR MIX DESIGNS AND LABORATORY TESTING
4.1 Introduction
4.2 RAP material characterization
4.2.1 RAP material preparation
4.2.2 Sieve analysis before extraction
4.2.3 RAP aggregate specific gravity
4.2.4 RAP binder content by ignition oven
4.2.5 RAP binder content from solvent extraction
4.2.6 Comparison between the ignition oven test results and solvent extraction test
4.2.6.1 Gradation analysis
4.2.6.2 Asphalt binder content analysis
4.2.7 Recovered RAP binder (RRB)
4.2.8 Virgin aggregate properties
4.2.9 Properties of virgin binder
4.3 Recycled mix design
4.3.1 Reference mix design
4.3.1.1 Mixing procedure
4.3.2 Mix specification
4.4 Slab preparation
4.4.1 Manufacturing
4.4.2 Coring
4.4.3 Sawing and resurfacing
4.4.4 Compact of tested specimens
4.5 Tension-compression test: complex modulus and fatigue
4.5.1 Principle of tension-compression test on cylindrical sample
4.5.2 Implementation of tension-compression test
4.5.2.1 Gluing the specimen to the testing platens
4.5.3 Procedure for complex modulus
4.5.4 Procedure for fatigue test
4.5.5 Data Acquisition and measured parameters
4.5.6 Data processing
4.5.7 Quality of the mechanical test
CHAPTER 5 ANALYSIS OF COMPLEX MODULUS RESULTS AND LVE MODELLING USING 2S2P1D MODEL
5.1 Introduction
5.2 Complex modulus test results
5.2.1 Isothermal and isochronal curves
5.2.2 Cole-Cole plane and Black space diagram
5.2.3 Master curves
5.3 Modelling the behaviour of asphalt mixtures using 2S2P1D model
5.3.1 LVE Simulation results for the virgin mixture (RAP028)
5.4 Statistical analysis
5.5 Complex modulus results’ repeatability
5.6 Analysis and Discussion
5.6.1 Effect of RAP content on complex modulus
5.6.2 Effect of other parameters on complex modulus
5.7 Shift factors
5.8 RAP coefficient of evolution CRCE∗
5.8.1 Variation coefficient evolution between replicates of complex modulus specimen’s(CVCE ∗)
5.8.1.1 Comparison between CVCE ∗ obtained from experimental data or using 2S2P1D model
5.8.1.2 Summary of the results corresponding to the repeatability of complex modulus
5.8.2 Effect of RAP content
5.8.3 Influence of binder grade
5.8.4 Effect of RAP aging
5.8.4.1 Sensitivity of RAP conditioning evolution values (CRCE – 2S2P1D∗) of mixes made with 25% RAP
5.9 Normalized curves
5.10 Chapter conclusion
CHAPTER 6 ANALYSIS OF FATIGUE TEST RESULTS
6.1 Introduction
6.2 Tested mixes
6.3 Experimental campaign
6.4 Results and analysis
6.4.1 Interpretation of fatigue test results
6.4.1.1 Evolution of temperature
6.4.1.2 Evolution of the heating on the surface of the specimen
6.4.1.3 Curves of the quality index
6.4.1.4 Evolution of the norm of complex modulus
6.4.1.5 Evolution of axial deformation
6.4.1.6 Axial Strain’s signal centered value
6.4.1.7 Stress amplitude
6.4.1.8 Center of stress’s signal
6.4.1.9 Phase angle evolution
6.4.1.10 Presentation in the black space
6.4.1.11 Dissipated energy evolution
6.4.1.12 Fatigue test validation
6.5 Correction of the test results based on air voids of the specimens
6.6 Comparison of fatigue test results
6.6.1 Comparison of initial values of the dynamic modulus
6.6.2 Comparison of fatigue tests results performed on the same asphalt mixtures at different levels of deformation
6.6.3 Comparison of fatigue tests at the same imposed strain level for different recycled mixtures
6.7 Evaluation of the fatigue life
6.7.1 Relationship between the fatigue life and the amplitude of deformation: fatigue lines
6.7.1.1 Relationship between the fatigue line parameters and RAP content of a mix
6.7.2 Measurement of the correlation between the Nf predicted by the basic model and the measured Nf
6.8 Effect of the addition of RAP
6.8.1 Comparison between recycled asphalt mixtures
6.8.2 Summary of the results
6.9 Prediction of the fatigue life according to the DGCB method
6.9.1 Calculation of the corrected damage at the end of phase II: DIIIc
6.9.2 Level of damage leading to rupture (DIII)
6.9.3 Evolution of the corrected damage leading to rupture (DIIIc)
6.9.4 Effect of air voids content on the critical damage (DIIIc)
6.10 Predicting the number of cycles to failure when considering only the 300 000 first cycles (DGCB N fII / III)
6.11 Summary of the results
CHAPTER 7 PROPOSED MODIFICATIONS TO THE BASIC FATIGUE MODEL TO TAKE INTO ACCOUNT THE EFFECT OF RAP CONTENT ON THE FATIGUE CRACKING RESISTANCE OF HMA MIXTURES
7.1 Introduction
7.2 Statistical approach: goodness-of-fit
7.3 Modification of the basic fatigue model based on phenomenological approach
7.3.1 Evaluation of the Accuracy of the newly developed model by the phenomenological approach
7.3.2 Determination of the A parameter value acting in the slope by the optimization process
7.3.3 Specific validation of the general fatigue model
7.3.4 Conclusion
7.4 Statistical fatigue model
7.5 Prediction of fatigue life at the pavement structure level
7.5.1 Determination of the fatigue lives values at the pavement level
7.5.2 Comparisons between the predicted pavement fatigue lives (structure level) and the predicted fatigue lives of recycled asphalt mixtures (material level)
7.6 Chapter conclusion
SUMMARY AND CONCLUSION
Télécharger le rapport complet
