A different approach to design road pavements PERFORMANCE TESTS ON BITUMEN AND BITUMEN MIXES

Transcription

1 A different approach to design road pavements PERFORMANCE TESTS ON BITUMEN AND BITUMEN MIXES This article takes starting point from the technical article published on the Italian magazine STRADE & AUTOSTRADE, n , "Le prove prestazionali sui bitumi e sui conglomerati bituminosi" (17) according to EN Standards, written by Vincenzo Buccino** and by Pietro Ferrari*. The author illustrates the topic by referring to American Standards, ASTM and AASHTO. In fact the "different approach to design road pavements" observed in the present article pursues an identical aim. Therefore it seems convenient to reproduce, within quotation marks, the opening and closing parts where considerations of common problems appear. **At that time Engineer responsible for Bitumen Development and Innovation at eni Refining & Marketing. Pietro Ferrari* In this article, the Author intends to highlight how it is now possible to have a more engineering-based approach to design and construct bituminous road pavements, intervening both in definition of the bitumen mix and in choice of the binder which is its active component. The empirical tests used until now - and which have given excellent results over time, keeping under control groupings of requisites obtained from the experience gained over the years - are no longer sufficient for supporting the leap in quality of paving. Roads now must have increasingly extreme and durable performance over time, due to the continual development of road traffic (in terms of frequency, weight and load time). Test methods and the relative laboratory equipment currently exist, all complying with standards, which allow to identify the performance requisites which are necessary to guarantee extreme and durable quality of the surfacing. This approach, which has been used in the USA for over two decades through the SuperPave system, as part of the SHRP research program, should hopefully completely adopted and consolidated by other countries. As well known, SuperPave (Superior Performing Asphalt Pavements) is a product of the SHRP (Strategic Highway Research Program) asphalt research. The system, as Asphalt Institute SP-2 mentions, "incorporates performance -based, asphalt material characterization with design environmental conditions to improve performance by controlling rutting, low temperature cracking and fatigue cracking." Dynamic tests: set-up for the four-point bending fatigue test (Standard ASTM D AASHTO T321) Three levels of interpretation may be identified in this article. The first, which is the most practical, consists in identifying the limits imposed on the requisites of certain tests, generally related to technical contract specifications, performed on bitumen and its mix, as a function of the minimum and maximum operating temperature, the type and frequency of traffic of the specific road. Deterioration of road is, in fact, mainly attributable to three distresses: ruts (permanent deformation due to temperature), fatigue cracks (at moderate temperatures) and cracks due to rigidity at low temperatures. Certain performance requisites of bitumen and the mix correspond with these phenomena and must therefore be kept under control. The second level of interpretation consists in 1 understanding and interpreting the laboratory tests conducted and the engineering-based use of the results obtained. The third level of interpretation, more specialistic and directed towards laboratory technicians, is useful for anyone who must conduct tests in accordance with the texts of existing standards and the equipment currently available. However, the author will only make a brief mention of this in the boxes, since the principal purpose of this article is to provide, in a simple, and therefore incomplete manner, obviously for reasons of space, an interpretation for designers, firms and works managers. We will discuss the two topics (mixes and bitumen) separately below, each to be interpreted with the three methods of interpretation indicated above.

2 Bituminous mixtures The final intention of SuperPave is that the specifications of bitumen mixes are given in terms of fundamental properties, identified on the basis of product performance during use. SuperPave takes into consideration bituminous mixes as HMA (hot mix asphalt), WMA (warm mix asphalt) and RAP (reclaimed asphalt pavement). At the present time, it is reasonable to take note that in many countries, due to different level of experience, asphalt concrete is defined according to two approaches: - An empirical approach, such as Marshall test and Hveem test,where the mix is specified and classified with constitutive formulas, with clearly defined dosages and with constituent material requisites and final product requisites. These final product requisites have been identified, following decades of experience, in correlation with basic "performance related" engineering characteristics, in order to be able to predict product performance; - A fundamental approach, where the mix is specified directly in terms of the requisites of basic "performance based" engineering characteristics (such as stiffness, visco-elastic recovery fatigue strength and resistence to permanent deformation), which allow product performance to be predicted. The restrictions on composition and the constituent materials are limited, so a greater degree of freedom in product formulation is possible. However the two approaches share several general requisites, of which the most significant ones are: the granulometric composition, the percentage of air voids in the compacted mix, and susceptibility to deform, tested for instance using the "Hamburg wheel tracking test" method (AASHTO T324), which is linked to performance. machines which operate both under load control and strain control and which are contained in a thermostatic cabinet. It should be emphasised that the bitumen mix, as a viscoelastic material, behaves in a manner heavily dependent upon temperature and load in terms of load type, frequency and application time; these parameters must therefore be well-defined in the tests. Resilient modulus, Dynamic modulus and phase angle When it is stressed by dynamic loads, a compacted bituminous mix is characterised, within the limits of deformation, temperature, loading conditions, by mostly linear viscoelastic behaviour, although always accompanied by a plastic component, at the operating ambient temperature. For small deformations, of the type to which road paving is subjected, the material, under repeated loads - which may be applied in an alternate time sequence, for example a sinusoidal waveform- tend, on release from the load and irrespective of its entity, to return to the original configuration, although with a time delay. This behaviour is dependent upon the temperature and the load application time. At high temperatures, a plastic-type component of deformation, i.e. permanent, is accentuated, and the same applies with prolonged times at low load application frequencies. With dynamic application of the load, which may be the sinusoidal type, simulating the repetitive loads of traffic in the laboratory, at ambient temperatures, the Complex Modulus E* is determined, which represents the relationship between the unit load applied and viscoelastic deformation. E Fundamental Tests ("performance based") The specificity of the approach is provided by the introduction of tests which measure the physical magnitudes directly linked with the performance characteristics. The most important are: - Resilient modulus - Dynamic modulus and phase angle - Resistance to fatigue - Resistance to permanent deformation - Low temperature cracking These dynamic tests are conducted by automatic 2 Complex Modulus E* (bitumen mixes) With representation of the Complex Modulus on a graph, with Cartesian axes, for immediate evidence, the elastic component E 1 = E* cos (φ) is placed on the x-axis and the viscous component E 2 = E* sen (φ) on the y-axis. Angle φ, the phase angle, represents the time delay caused by the viscous component E 2. Complex

3 Modulus E* is represented by the vector resulting from the two components E 1 and E 2 and the absolute value of its Modulus E* represents stiffness value "S". A common characteristic of this performance-based test is the possibility of application of the principle of superimposition of the stiffness/frequency curves, at a fixed temperature (isotherms) or the stiffness/temperature curves, at fixed frequency or constant load application time (isochrones). Thanks to this possibility, master curves may be created which allow the operator, once a limited number of tests have been conducted under different frequency and temperature conditions, to perform continual assessment of the modulus within a relatively wide range of temperatures and load application times, by shifting the experimental curves by a sperimental log a T parameter, in a manner to prolong one of the experimental curves, used as a reference (shifting). The stiffness/temperature master curve Resilient modulus Indirect Tension Test for Resilient Modulus of Bituminous Mixtures - ASTM D Cylindrical specimens 4" and 6" diameter are tested by indirect tension repeated loading at some different ambient temperatures and different vertical loading frequencies, within a loading range compatible with visco-elastic behavior, and specimen deformation data are used for resilient modulus calculation. The loading cycle pulse consists of sinusoidal (haversine) loading phase followed by a rest period. The method consists, generally, of a test series conducted at 5 C, 25 C and 40 C at one or more loading cycle frequencies, for instance 0.33, 0.5 and 1.0 Hz. Horizontal and vertical specimen deformation should be measured by installed LVDT 's pairs. More simply, in case of test at 25 C, horizontal deformation only can be measured by assuming a Poisson's ratio default value of Two resilient modulus are calculated: the instantaneous resilient modulus that considers the recoverable deformation that occurs immediately during the unloading phase of one cycle and the total resilient modulus that includes in the calculation both recoverable deformation and the time dependent delayed recoverable deformation during the unloading and rest- period phase of one cycle. E RI = P(ν RI +0.27)/ H I E RT = P(ν RT +0.27)/ H T where: E RI = instantaneous resilient modulus of elasticity, psi or MPa. E RT =total resilient modulus of elasticity, psi or MPa. ν RI and ν RT instantaneous and total resilient Poisson's ratio may have 0,35 default value The test is useful for ranking compacted bituminous mixtures in term of effects of temperature, loading rate, rest periods and other effects for pavement design or evaluation. Resilient modulus of Bituminous Mixtures by Indirect Tension test, ASTM D7369 The test is similar to ASTM D4123, but the substantial difference consists of the displacemen transducer configuration that is on sample type. This makes more feasible the measurement of Poisson s ratio and hence of the instantaneous and total resilient modulus.. Indirect tensile test jig according to ASTM D Dynamic Modulus and phase angle Determining Dynamic modulus of Hot Mix Asphalt (HMA) -AASHTO T 342 and AASHTO TP 79 Cylindrical 4"diameter specimens of compacted bituminous mixture are subjected to sinusoidal loading

4 Dynamic modulus of Hot Mix Asphalt (HMA) according to AASHTO T 342 and AASHTO TP 79. Resistance to fatigue The fatigue strength test consists generally of assessment of the number of pulses, usually of the sinusoidal type, necessary to achieve a specific reduction in resistance of the specimen subjected to repeated flexural bending, according to ASTM D7460 or AASHTO T321. at different frequencies generally ranging from 0.1Hz to 25 Hz and at different temperatures ranging from 14 to 130 F (-10 C to 54.4 C). Specimens are obtained by overcoring and cutting 6" diameter samples compacted by gyratory compactor. Vertical deformation is measured by on-sample displacement transducers. For each temperature the Dynamic modulus E* and phase angle φ are determined in order to achieve interdependent correlation between Dynamic modulus with test temperature and test frequency. It is so possible to elaborate the master curve(dynamic modulus versus temperature and frequency) and Black diagram ( phase angle versus temperature and frequency). These results in engineering units are a basilar tool for pavement design and bituminous mix analysis. The tests to AASHTO TP79 and to AASHTO T342 are rather similars. There is some difference. In AASHTO TP 79, "Determining the Dynamic Modulus and Flow Number for asphalt Mixtures Using The Asphalt Mixture Performance Tester" (AMPT), the test temperature starts fron 4 C and the specimen can be tested in unconfined or in confined configuration by mean of confining pressure and membrane in triaxial cell; the loading of the sample is controlled at target strain (100 microstrain). In AASHTO T 342 "Determining Dynamic Modulus of Hot Mix Asphalt", the specimen is tested unconfined only and the dynamic load should be adjusted to obtain axial strains between 50 and 150 microstrain, and minimum temperature is -10 C. Θ(ω)= Θ ε Θ δ E*(ω) = δ / ε Where: Θ(ω) = Phase angle between applied stress and strain for frequency ω, degrees Θ ε = average phase angle for all strain transducers, degrees Θ δ = stress phase angle, degrees E*(ω) = dynamic modulus for frequency ω δ = stress magnitude, kpa (psi) ε =average strain magnitude Normalized complex modulus multiplied by cycles versus number of cycles. Dissipated energy versus number of cycles. 4PB, 4 Point Bending test set 4

5 Determining the Fatigue Life of Compacted Hot-Mix Asphalt (HMA) Subjected to repeated Flexural Bending, AASHTO T321, ASTM D7460 Once determined the initial stiffness close to 50 cycles and selected a microstrain level by which a minimum of load cycles are requested to reduce the stiffness to 50% of initial stiffness, it is possible to run the test. Parameters that can be calculated: -Maximum tensile stress (Pa) σ t =(0.375P)/bh 2 where: P = applied load (N) b= average specimen width (m) h = average specimen height (m) Maximum tensile strain(m/m) E t = (12δh)/(3L 2-4a 2 ) where: δ=maximum deflection at center of the beam (m) a= space between inside clamps(m) L= length of beam between outside clamps(m) Flexural Stiffness(Pa) S= δ t / E t -Phase angle (deg) φ= 360 fs where: f= load frequency (Hz) s= time lag between P max and δ max in seconds -Dissipated Energy( J/m 3 ) per cycle D = π σ t E t sin(ψ) i=n -Cumulative Dissipated Energy, J/m 3 = Σ D i I=1 where: D i = D for the i th cycle -Initial Stiffness (Pa) S= Ae bn where: e = logarithm base e A= constant b= constant Other criteria to determine the end of the test is to plot the normalized complex modulus, multiplied by cycles, versus cycles. The test has to be prolonged beyond the peak of the diagram until the flexural beam stiffness reduces to about 40% of the initial beam stiffness. It could also be required to estimate the resistance to fatigue on the basis of cycles and dissipated energy. Fatigue lines of bitumen mixes. A characteristic of these performance-based test is being able to construct "Fatigue lines", diagrams which represent decay of the fatigue strength value as the deformation set increases. The specimen consists of a prismatic beam of compacted bituminous mix, generally at 7% air voids, 380 mm (14.96 ) long by 50 mm (1.97 ) thick by 63 mm (2.48 ) wide.in the case of dynamic tests conducted with preset deformation (controlled strain), the criteria consists in counting the pulses, obviously lower than the ultimate tensile stress, necessary to reach a given decrease usually 50% of the stiffness modulus compared with the initial one of the test. 5

6 Recent empirical-machanistics studies have highlighted and rationalised, through fatigue tests, the phenomenon of healing of the micro-cracks which form in the compacted mix. This contributes to better understanding of development of damage from fatigue and to bringing it closer to the real behaviour of the paving. Flow zones of permanent deformation. Resistance to Permanent Deformation Determining the Flow Number for Asphalt Mixtures Using the Asphalt Mixture Performance Tester - AASHTO TP 79; NCHRP 9-29 (1). Well known AASHTO TP 79 is related to AMPT, Asphalt Mixture Performance Tester, versatile system that can perform more than one type of performance based test. Beyond the Dynamic Modulus and phase angle determination, that has been used for the rutting criteria in MEPDG program, Mechanist Empirical Pavement Design Guide, the second section of AASHTO TP 79 is dedicated to the permanent deformation characteristics of paving materials by the "Flow Number Test". A cylindrical sample, 4" diameter * 6" height, is subjected to pulse load cycles consisting each of 0,1s load followed by 0,9 s rest period, in confined or unconfined configuration, where the magnitude of vertical load, or deviator stress in confined configuration, is in the order of 10 to 30 psi (30 and 200 kpa). The test temperature can range from 100 F and 130 F (37,4 C and 54 C). The permanent deformation characteristic of a compacted bituminous mixture is indicated as flow number. Flow number is defined as the number of cycles at which the incremental ratio of strain versus cycle number is minimum. It is positioned at the end of secondary flow zone. Beyond, the tertiary flow zone begins and strain increases rapidly. In the data processing of permanent deformation resistance it has to be taken count of recoverable strain component. In NCHRP 9-29 (National Cooperative Highway Research Program) beyond the "Flow Number Test", it is described the "Flow Time Test". A cylindrical sample, 4" diameter * 6" height, is subjected in confined or unconfined configuration to a static axial load at specified temperature. The flow time is the time corresponding to the minimum rate of change of deformation versus time. The flow time falls along the 2nd stage of the creep curve, where the creep rate is constant approximately. The test is useful for judging and ranking bituminous paving mixtures about resistance to permanent deformation. (2) 6 Flow time Flow number Hamburg Wheel -Track Testing of Compacted Hot-Mix Asphalt (HMA) - AASHTO T324 By this test, the rutting and moisture susceptibility of a compacted hot mixed asphalt specimen is measured. A specimen slab, or a 150 mm diameter cylinder specimens pair, submerged in thermostatic bath in the temperature range from 40 C to 50 C, is subjected to a repetitive loading by a 705 ± 4.5 N loaded steel wheel, by mean of a relative reciprocating movement, at the rate of 50 passes/minute. The depth of the produced impression is measured continuously and plotted versus the number of wheel passes, and give indication to rutting tendency of the compacted mix. The rut formation testing on submerged samples allows also to evaluate the moisture susceptibility of hot mixed asphalt (HMA).

7 HMA samples 150 mm diameter are compacted by a Super pave Gyratory Compactor (SGC) and cut into specimens from 38 to 100 mm thick. Slab specimens can be prepared by a laboratory linear slab compactor with kneading action or equivalent. Slab dimensions are 320 mm long x 260 mm wide with thickness varying from 38 mm to 100 mm. Besides laboratory compacted specimens also specimen taken from field can be tested. Cylinder specimens can be wet cored and cut at desired depth and slab specimens can be taken from field by wet saw-cutting. The curve of the rut depth versus number of passes is generally characterized by two stages. The first one is the first steady -state portion of the curve, the second one is the second steady state portion of the curve. By the plot, values of slope and intercept of first steadystate portion of the curve with y axis and values of slope and intercept with x axis are obtained. The Stripping Inflection Point (SIP)= Intercept (second portion) - Intercept (first portion)/slope (first portion) - Slope (second portion) means the point, in term of number of passes, at which the specimen collapses due to aggregate stripping caused by moisture induced damage. The intercept of the second steady state portion of the curve with x axis means the Failure rut depth. Low temperature cracking Determining the Creep Compliance and Strength of Hot-Mix Asphalt (HMA) Using the Indirect Tensile Test Device. AASHTO T322 The test configuration consists of an indirect tensile loading machine which acts on a 150 mm diameter compacted hot asphalt cylindrical specimen. Besides load is measured together with vertical and horizontal deformation, measured by on sample displacement transducers. From the test it is possible to evaluate the tensile creep compliance, defined as time-dependent strain divided by the applied stress, the Poisson's ratio and the tensile strength. In "thermal cracking analysis", the creep compliance and strength testing are usually conducted at low temperatures: 0 C, -10 C and -20 C. The static load is fixed and is applied to the specimen for 100 ± 2 s in order to obtain a horizontal deformation between mm and mm ( that is within the linear visco-elastic range) and the Poisson's ratio, "ν", between 0.05 and 0.5. The creep compliance D(t) can be measured (kpa -1 ). After the fixed load application, an increasing load is applied at the rate of 12.5 mm / minute of the loading ram. The load has to be recorded, along with vertical and horizontal deformations, until it starts to decrease and the tensile strength is determined. In "fatigue cracking analysis" the indirect tensile load is applied at 50 mm/minute rate at temperatures that start from 20 C or less (for Superpave mixture analysis, 0 C, -10 C and -20 C are required). The load has to be recorded until it starts to decrease along with vertical and horizontal deformations. The knowledge of Poisson's ratio is necessary to determine the failure time t f,n and to give indications for fatigue cracking analysis. By plotting y-x (vertical deformation minus horizontal deformation) versus time, it is possible to evaluate the "first peak time", when the difference y-x is maximum, to be identified as t f,n. By continuing the load the difference y-x decreases by consequence. Associated to t f,n is P f,n failure load at time t f,n. Correlated tensile strength S t,n of specimen n can be calculated on the basis of P f,n failure load at time t f,n. Hamburg wheel tracker 7

8 AMPT and relevant tests AMPT, Asphalt Mixture Performance Tester, is a compact servo-hydraulic testing machine developed through a research programme developed by the National Cooperative Highway Research Program (NCHRP) Projects 9-19 Superpave Support and Performance models Management, and Project 9-29, Simple Performance Tester for Superpave Mix Design. The project criteria aimed to realize a compact, fully integrated testing machine at a reasonable cost able to cover the main critical demand to grant the necessary performance in bituminous paving mixtures. To do this, Dynamic Modulus test, Flow number test and Flow time test( to NCHRP projects only) can give to the designer or supervisor a tool useful for evaluating the majority of Hot Mix Asphalt (HMA) critical aspects. Dynamic Modulus test, as conceived in AMPT, simplifies the development of master curves, that are needed for pavement structural design; Flow Number test enables to quickly evaluate the mixture's rutting resistance. Both tests are standardized currently by AASHTO TP 79 "Determining the Dynamic Modulus and Flow Number for Asphalt Mixtures using the Asphalt Mixture Performance Tester (AMPT). AMPT is characterized by high performance electro-hydraulic actuator for compression and tension tests, built in refrigeration and heating unit, by integrated triaxial cell with compressed air-driven confining pressure system, by environmental chamber with temperature control unit and by a digital control with optimized data oversampling and data acquisition system. AMPT is also characterized by notable versatility. In fact beyond Dynamic Modulus test, Flow number test and Flow time test, AMPT is suitable to perform other new tests that are developed to better adapt to mechanistic empirical asphalt pavement modeling, such as: Fracture Energy test. In AASHTO TP 105 Standard Method of Test for Determining the Fracture Energy of Asphalt Mixtures Using the Semicircular Bend Geometry (SCB) and ASTM D Standard test method for determing Fracture Energy of Asphalt-Aggregate mixtures using the Disk-Shaped compact tension geometry. The following parameters can be measured at fracture point: Fracture Energy G f,, Fracture Toughness KIC and Stiffness S. Conditioning and testing temperatures can be 10 C above or 2 C below the PG lower limit of the mixed binder. The specimen consists of semicircular section 150 mm diameter cut from gyratory compactor specimen having a notch cut along the diameter. The deformation measuring device is applied on both size of the notch. Crack initiation test. Crack initiation and propagation is a problem that becomes more present due to the tendency to increase the use of recycled bituminous pavement (RAP). The introduction of hardened binder of RAP increases the tendency to cracks formations. Tendency that has to be controlled by optimizing the dosage binder renewing agents and by blending with additional suitable binder. SCDUF test (Simplified Continuum Damage Uniaxial Fatigue Test) is dedicated to this type of damage. The test is currently illustrated by Christensen, D., and R. Bonaquist. "Analysis of HMA Fatigue Data Using the Concepts of Reduced Loading Cycles and Endurance Limit," Journal of Association of Asphalt Paving Technologists, Vol. 78, 2009, pp SCDUF test is a dynamic tension test that can be performed by AMPT by means of dedicated accessory and software. SCDUF test has been forwarded to AASHTO to future publication. Overlay test Overlay Test is related to the investigation on the reflecting cracks of a new asphalt layer that is laid down on an old pavement with cracks or construction joints. A dedicated accessory and software installed on AMPT allows to assess the suitability of the bituminous mixture to deal with cracks of the support layer. ASTM WK has developed a testing procedure. ASTM WK26816 New Test Method for Determing the Susceptibility of Bituminous Mixture to Replicative Cracking using the overlay tester. Viscoelastic fatigue test S-VECD (Simplified Viscoelastic Continuum Damage) test considers fatigue damages by the viscoelastic behavior of compacted bituminous mixtures, differently from the classic fatigue tests that deal with the fatigue damages by the point of view of crack initiation and propagation. S-VEDC test examines the complex behavior of the bituminous mix related to non linear response to dynamic compression/testing and to associated healing effects by ciclic Push- Pull fatigue approach. S-VECD test can be performed by AMPT by mean of adequate testing accessory and software. The test has been widely studied and validated by University of Michigan. The subject has been treated in NCHRP 9-19 project. 8

9 TSRST test set AMPT - The creep curve for back calculation of the rheological parameters by tensile creep test (TCT); - The fatigue resistance test by combination of cryogenic and mechanical loads by uniaxial. Cyclic tension stress (UCTST). Standard Test Method for Thermal Stress Restrained Specimen Tensile Strength, AASHTO TP 10 (3). A cylindrical or prismatic sample of compacted bituminous mix is epoxy glued restrained by a pair of platens of a dedicated testing frame that is housed in climatic cabinet. As much as the temperature decreases the shrinkage effect into the sample increases and the developed tensile stress is measured until breaking of sample occurs. The thermal induced curve is plotted and its solpe calculated as ds/dt, where DS is the change in stress versus time along the linear portiion of the curve before the breaking point (now AASHTO TP10 included TSRST test, as mentione hereunder It could be interesting to mention a development of some related parameters that can be processed according to the analogous European Standard EN "Test methods for hot mix asphalt. Part 46: Low temperature cracking and properties by uniaxial tension tests". They are: - The tensile strength versus temperature, by uniaxial tension stress test (UTST) - The failure by thermal stress restrained specimen test (TSRST); - The tensile strength reserve test, by combination of TSRST and UTST; - The relaxion time test (RT); 9

10 Sample preparation. Specimens are prepared by following a volumetric mix design approach by means of gyratory compactor, to AASHTO M 323 "Superpave Volumetric Mix Design". Trial aggregate - asphalt binder blends, each with different asphalt binder content, are compacted according to the required grade of compaction. The result of the gyratory compacted tests will be closer to the actual in service behaviour of the mix then those compacted with impact method. The slab specimens that are mainly used for wheel tracking test (Hamburg test) are obtained by compaction of slab or prisms by mean of linear kneading compactors or equivalent compactor such as roller or wheel compactors or shear box compactor. Prisms for flexural fatigue resistance tests to AASHTO T321 and ASTM D7460 can be prepared by sawing the compacted slabs or prisms. Overview on some reference points. Regional agencies and bituminous mixture designers often use the MEPDG guide (Mechanistic - Empirical Pavement Design Guide) to asses the performance of pavement with HMA mixtures. This can be realized thanks to the comprehensive and complex calculation program and to the wide-ranging prediction models that have been calibrated by test sites such as LTPP (8) (9) (10) (Long- Term Pavement Performance) test sections with database located throughout the United States, Shear box compactor MnROAD (11) (Minnesota Department of Transportation), WesTrack (12) (Nevada Accelerated Pavement Testing ), FHWA ALF (13) (Federal Highway Administration Accelerated Loading Facility) and other test sites. Stress/ strain laboratory data (E* and Fn), traffic loading, grade of binder, environmental data are inputs to MEPDG guide software for the eleboration of distress prediction, such as rutting and cracking in the pavement. The program uses three levels of quality of input parameters, in hierarchical order, according to the importance of the road and traffic data.* Stell Roller Compactor Steel Roller Compactor Principle In Level 1, the input parameters are measured directly. Beyond the environmental data and traffic data,the primary material property input is the measured Dynamic modulus E*. By means of few measurements of E* at different temperatures and frequencies the E* modulus master curve is calculated and is available for evaluation procedures. * Traffic data are generally expressed in ESALs (number of million of axels equivalent to 80 kn that it is foreseen to pass during the life of the road; therefore they are level E1 1 million, level E3 3 millions, level E10 10 millions, level E30 30 millions ESALs. Other traffic parameters that can be used are AADT, Annual Average Daily traffic, WIM, Weigh In Motion, AVC, Automatic Vehicle Classification and vehicle count. 10

11 Superpave Gyratory Compactor (7) Gyratory compactor has been standardized by ASTM D6925 and AASHTO T312 "Standard test methods for Preparation and Determination of the Relative Density of Hot Mix Asphalt (HMA) Specimens by means of the Superpave Gyratory Compactor". It is well known that the gyratory test defines the mix by its volumetric characteristics. A given mass of asphalt mixture, heated to its optimum compaction temperature, is poured into a cylindrical steel mold and then subjected to two simultaneous actions by the compactor: vertical static compression and horizontal gyratory rotation with an axial offset, creating an internal gyratory angle of The rotational frequency and vertical pressure are 30 gyrations per minute and 600 kpa. By means of installed transducers and software, as the sample decreases in height, the following volumetric parameter can be calculated in real time cycle by cycle: percentage of theoretical density (ASTM D2041), density of sample, V% (air void of compacted mix), VMA ( percentage voids of mineral mix), VFB (percentage of voids filled by binder). An approximately straight line curve is also drawn in real time showing the density versus number of cycles in log scale on the horizontal axis. Gyratory compactor plays a key role in the Superpave volumetric mix design. In the compaction curve there are three compaction levels that correspond to the three characteristic number of gyrations, that is N initial, N design and N max. N init corresponds to density after 10 gyrations; N max, or final, corresponds to a compaction of 98% of theoretical density, meaning that a max value of density is reached beyond which the densification curve no longer follows a straight line; N design is defined as the number of gyrations corresponding to a density reproducing the one to be obtained by the roller compactor in a new asphalt pavement. N design is positioned therefore along the compaction curve with a lesser percentage of theoretical density than at N max. The initial density at 10 gyrations, N init, corresponds approximately to the density of a laid down bituminous mix just behind the finisher machine. These three gyration numbers are linked by the following equations(as per Superpave): Log N initial = 0.45 log N design Log N max = 1.10 log N design Some considerations: a steep compaction curve indicates high compactibility of the mix. By appropriate dosage of binder and graded aggregates with high angularity, the bituminous mix will absorb a high degree of compaction energy, or intensive in-situ rolling; by consequence, a good resistance to permanent deformation due to good interlocking of aggregate can be achieved. On the opposite, densification curves with low slope are characterized by poor compactibility and poor resistance to permanent deformation. The aggregate interlocking development during gyratory compaction creates attrition between particles that develops a shear force. Some research models of gyratory compactors are provided by transducers that measure gyratory shear force. Measurement of gyratory compaction shear is a useful tool that can refine the search of the most suitable bituminous mix. During compaction a max value of shear force, expressed in kn/m 2, is reached and its positioning along the compaction curve is an indication that characterizes the mix. Mixes with relatively elevated V% (air void%) and low VFB (%aggregate mix voids filleed by binder), show the highest value of shear at N design or N max, while mixes with low V% and high VFB, show a drop of the shear between N design and N max and the max value is positioned within the first half of the curve. This means that as much closer the particles are, the lubricating effect of binder prevails and shear decreases. The gyratory compactor (according to ASTM D AASHTO T312), shear force measurement included. 11

12 Giratory Compaction density versus cycles, shear versus cycles and shear versus density. According to the information requested and the performance models included in the software, MEPDG can predict the following pavement distresses: Rut depth for HMA layers, unbound aggregate layers, and subgrade. Transverse thermal cracking. Alligator cracking due to bottom - initiated fatigue. Longitudinal wheel path cracking due to surface initiated cracking Reflection cracking (overlay) Roughness Ruth depth As deeply explained in NCHRP report 580, Dynamic Modulus E*, measured to AASHTO TP79 or AASHTO T342, is used to predict the rut depth of the bituminous mix by means of the related master curve as per AASHTO PP69, and to compare the predicted rut depth of the layers with the allowed specified ruth depth. The dynamic modulus testing has to be run on four hours 12 at 135 C aged specimens, compacted by gyratory compactor, to the expected air void content of the pavement. The component E* /sinφ is the parameter used in the calculation and acceptance according to rut depth criteria. The user can adjust the design of HMA and structure to comply with the rut limit according to the predicted indication. It is to point out that the user can choose to use the more simplified E* AMPT Criteria Program when just he needs to compare the predicted rut depth with the allowed specified (3) (14) rut depth. By the E* AMPT Criteria Program the rut resistance is also determined by the "Flow Number" or " Flow Time" input. Fatigue Cracking On the basis of required information input by the user, MEPDG performs a complete analysis and processes the predicted distress of bottom- up (alligator) fatigue cracks, in terms of percent of total area, and of top - down (longitudinal) cracks, in term of feet per lane mile. MEPDG software uses the Dynamic Modulus E*master curve for fatigue cracking prediction. Thermal cracking Once received the needed data, included climate data in term of longitude and latitude, MEPDG provides prediction of transverse thermal cracking with time, expressed as feet of cracks per lane mile. MEPDG software needs input of VMA value of surface layer as well as of low temperature creep compliance and tensile test data on aged specimens according to AASHTO T322. Reflection cracking MEPDG software empirically predicts the time of appearance of reflective cracks information based on thickness of HMA and condition of underlying pavement. Roughness MEPDG software predicts the evolution of IRI value (International Roughness Index) by processing data based on initial IRI value and other processed and related parameters. Durability distresses such as moisture damage, raveling and scuffing are not included in MEPDG models. It is supposed that such distresses will be minimized by complying with a proper HMA mixture design. Likewise the change of skid resistance with time and traffic is not predicted.

13 By input a mix design with related laboratory test results, layers structure, environmental data, traffic data and allowed contractual distress limits, pavement design engineers and contractors can draw essential directions on the consistency of their structural solutions and on adjusting them to complying. MEPDG software does not supply mix designs, or structure designs, but assessment of the consistency with performance requirements. Brief mention on adoption of E* modulus and Flow number to determine rutting bituminous mixture performance and QC/QA criteria by Michigan DOT. (15) E* modulus based acceptance criteria. 13

14 E*modulus Concerning E* modulus tests for rutting assessment, they are conducted at various temperatures ( -5 C, 4 C, 13 C, 21.3 C, 39.2 C) and at various frequencies ( 0.1Hz, 0.5 Hz, 1 Hz, 5 Hz, 10 Hz, 25 Hz) on 15 standard mixtures with different maximum size and related to different Traffic levels: E1, E3, E10, E30. Master curves have been processed by keeping E* modulus test at -5 C as reference curve, for rutting assessment. There is a correlation between the laboratory E* results with field rutting, with possibility of plotting graph for various traffic levels. By the graph it is evident the criteria of "acceptance" or "rejection" of E* modulus laboratory result on the basis of a bituminous mixture performance rut depth. Other method consists of constructing a master curve of a designed mix by testing it at 3 temperatures from -5 C to 40 C and at 5 different frequencies (ranging from 0.1Hz to 25 Hz), and comparing it with the master curve of the same mix but constructed with Minimum Required Dynamic Modulus, drawn from DOT database, and checking if the experimental master curve is higher, as suggested. Example of tabular data of minimum acceptable E* values for some mix types drawn from related master curve of minimum required dynamic modulus at 13 C: Flow number Concerning the Flow Number determination for each mixture, the tests have been conducted by shortening procedure (Stepwise Approach) consisting of measuring at 45 C the secondary flow curve slope and processing the Fn by following relation: Flow number= x Fnslope^ Flow Number is defined as the minimum point of strain rate versus load cycle number. On the basis of field rutting performance and TPRLI, of two year rutting warranty period and of Flow Number Slope mixture types ranking, acceptance criteria of Maximum Flow Number Slope and Minimum Flow number according to the traffic level and compaction grade (V%) have been developed. Average flow number slope of mixes. Flow number and flow number slope based 14

15 Bitumen Asphalt binder Performance Grading. As known, bitumen is the active binder of the bituminous mix. It is present on average as 5% of the mixes and determine their rheological behavior to a large extent. In the past, up to about twenty years ago, asphalt binders were classified by mean of traditional empirical tests. The requisites that were given by the tests could grant an experience based performance and give certain guaranty on constancy of source of asphalt binder. On the purpose, AASHTO M20 and ASTM D946 ruled the "Penetration Grade Verification" where the following traditional standard tests were defined: Penetration (ASTM D5), Flash point ( ASTM D92), Ductility (ASTM D113), Solubility (ASTM D2042), Thin Film Oven Test, TFOT (ASTM D1754,), Retained penetration after TFOT (ASTM D243), Ductility (ASTM D113) after TFOT. As already said, this system is no longer commonly used in the United States. The penetration test only is still alive as valuable indicator of consistency of source and, some time, as commercial denomination. As mentioned in AASHTO M 20, five penetration grade of asphalt cement are identified: 40 to 50, 60 to 70, 85 to 100, 120 to 150, and 200 to 300 dmm penetration at 25 C. Other empirical classification has been given by AASHTO M226 "Viscosity Grade Verification" that assesses asphalt binder by the viscosity point of view. The system includes the following tests: Absolute Viscosity (ASTM D2171, AASHTO T202), Kinematic Viscosity ( ASTM D2170, AASHTO T201), Penetration, Flash Point, Solubility, Thin Film Oven Test (TFOT), Absolute Viscosity after TFOT, Ductility after TFOT. This system was extensively used prior the development of Superpave, and does not address low temperature properties. Superpave, since more than two decades, has defined a mechanistic empirical system that includes some empirical tests and fundamental tests where the results are expressed in engineering units at a defined temperature. The system, "PG classification" has been standardized as AASHTO M320 and ASTM D6373. Binder is classified according to two critical temperatures, where the classification assigns a performance grade delimited by two temperatures that correspond to the average seven day maximum temperature of the year and to the minimum temperature of the year of the site where the binder has to be used. The tests adopted are: Flash Point, Rotational Viscosity (ASTM D4402 and AASHTO T316), DSR (dynamic shear rheometer test,( ASTM D7175/ASTM D7405 and AASHTO T315), Rolling Thin Film Oven Test (RTFOT, ASTM D2872 and AASHTO T240), DSR after RTFOT, Pressure Aging Vessel (PAV, ASTM D6521 and AASHTO R28), DSR test after PAV, Bending Beam Rheometer test (BBR, ASTM D 6648 and AASHTO T313) after PAV and Direct Tension test, DTT ( AASHTO T314 and ASTM D6723) after PAV. The aim of PG classification is to grant the necessary characteristics of binder for road construction. The specifications listed in the "table 1"of AASHTO M320 and ASTM D6373 cover original binders and modified binders. A brief illustration of "table 1" structure (see page 16). Seven main grades of binder are defined according to "Average seven-day maximum design temperature C ": 46 C, 52 C, 58 C, 64 C, 70 C, 76 C, 82 C. Each grade includes some "minimum pavement design temperature C" values, which are 6 C spaced : 46 grade includes three temperatures from -34 C to -46 C; 52 grade includes seven temperatures from -10 C to -46 C; in 58 grade there are five temperatures from -16 C to -40 C; in 64 grade, six temperatures from -10 C to -40 C; in 70 grade, six temperatures, from -10 C to -40 C; in 76 grade, five temperatures, from -10 C to -34 C; in 82 grade, five temperatures, from -10 C to -34 C. Therefore thirty-seven types of binders are considered whose performance grade is defined by two limit temperatures e.g. PG It should be noted that the seven performance grade ranges of binders overlap each other partially and are not equally spaced ; this implies that the variety of combinations of PG grades should cover the performance complying with a complete overview of climatic situations. As per "table 1" (see page 16), the testing requirements are included in three sections. First section includes two empirical requirements of original binder. 15

16 "Flash Point by Cleveland Open Cup" ( AASHTO T48, ASTM D 92) where the minimum acceptance value of 230 C of flash point grants that excessively high volatile components cannot deteriorate the quality of binder. The second testing requirement of original binder is viscosity according to AASHTO T316/ASTM D4402 Viscosity Determination of Asphalt Binder Using Rotational Viscometer. The maximum viscosity value at 135 C is 3 Pa*s. By this upper limit binder handling, pumping and spraying, is guaranteed. Complying with these two empirical requirements is a prerequisite to continue the subsequent tests to define the grading of binder. By the third test on original binder, the grading procedure starts with dynamic shear measurement, DSR test, according to AASHTO T315 Determining the Rheological Properties of Asphalt Binder Using a Dynamic Shear Rheometer (DSR). By DSR test, the well known dynamic shear modulus G*and phase angle δ are measured within the viscoelastic field. The upper temperature grading is defined, e.g. PG62-xx. The upper temperature grading defining parameter is G*/ sin δ. The minimum acceptance value of 1.00 kpa at 10 rad/s at upper grading temperature has to be obtained by trials by starting at about midrange intermediate test temperature and increasing or decreasing at 6 C increments. It means that the minimum elastic consistency, in the linear viscoelastic field, that a road asphalt binder has to provide at the highest operating temperature, is secured. The second section of "table 1" (see page 16) is dedicated to the requirement of binder subjected to RFTO test (RTFOT, ASTM D2872 and AASHTO T240). As known, RFTO test simulates the short term ageing of asphalt binder after heating, spraying, mixing with aggregate in the plant and laying by the finisher and roller. AASHTO M320 designation takes precaution against change of binder mass during the oxidative heating action of the Rolling Thin Film Oven Test by fixing a maximum limit of 1.00%. On the binder residue from RFTOT, the dynamic shear measurement, DSR test, with 25 mm plate and 1 mm gap at 10 rad/s test configuration, is performed at the test temperature used to determine the starting PG grade in order to verify that the minimum acceptance value of 2.20 kpa of G*/sin δ parameter, at the upper grading temperature, has been reached. Obviously the limit has been raised in comparison to 1,00 KPa limit of original binder, by considering the binder hardening after RFTO test. Binder performance at the possible highest operating temperature is definitely confirmed. In the third section of "table 1" (see page 16) the suitability of binder after long term climatic hardening is defined by using the PAV Pressure Ageing Vessel AASHTO PP1 and ASTM D6521 conditioning. PAV conditioning simulates long term ageing of asphalt binder in compacted road bituminous mix layers. In the PAV apparatus, thin film of binder is subjected to oxidative heating at 90 C, or 100 C, depending on the grade of binder (110 C in desert climates) ; the residue is collected and tested by Dynamic Shear test, DSR, and by Bending Beam Rheometer test, BBR. At intermediate operating temperatures, ranging from 4 C( PG46-46) to 40 C( PG 82-34), DSR test, 8 mm plate, 2 mm gap, 10 rad/s, is performed at 3 C increments, and the minimum acceptance value, in term of G*sin δ, has to be 5000 kpa. Term G*sin δ is defined as loss shear modulus and the value is related to the viscous component of the aged binder at the test temperature. An acceptable viscous consistency, in the linear viscoelastic field, is to be guaranteed notwithstanding the ageing hardening. By the BBR test to AASHTO T313 "Determining the Flexural Creep Stiffness of Asphalt Binder Using the Bending Beam Rheometer (BBR)", the significant parameters S, estimated creep stiffness, and m-value, absolute value of the slope of the logarithm of stiffness curve versus logarithm of time, are evaluated in 60s period. Two sub sections of BBR tests are conducted. One sub section on PAV residue by operating at 6 C increments and starting from the temperature by which G*sin δ, is <5000 kpa by the DSR test. Test temperature is increased until S, stiffness, value is 300 MPa and m- value is The second test subsection is conducted on fresh binder, by operating at 6 C increments and starting from the minimum pavement design temperature added by + 10 C after 24 hours conditioning. For instance, in case of a binder PG 64- xx, the starting temperature should be -40 C + 10 C = -30 C. The minimum test temperature by which the required values of S and m-value are reached, that is the second term of PG grading, is confirmed, e.g. PG

17 In certain cases, at a given minimum temperature, Stiffness value S could be above 300 MPa, in any case should be between 300 and 600 MPa, while m-value is It is then allowed to confirm the low temperature grading by the Direct Tension test, DTT (AASHTO T0314 and ASTM D Determining the Fracture Properties of Asphalt Binder in Direct Tension - DT)at the same minimum temperature and, if necessary, by increasing the testing temperature by 6 C increments, until the failure strain is 1.0%, at 1.0 mm/min test speed. From this third section of AASHTO M320 and ASTM D6373 the aim to guarantee a minimum of viscous characteristic of asphalt binder, at intermediate and lowest temperatures after long term ageing also, is evident. 17