9th Conference

code ISAP 9th Conference Titles & Abstracts
09000 9th International Conference on Design of Asphalt Pavements – Table of Contents

A listing of the Sessions, the titles of papers, and the authors.

Results of the Rilem Round Robin Test on Binder Rheology
A. Vanelstraete and D. Sybilski

Road and Bridge Research Institute, Warszawa, Poland ABSTRACT: In 1998 the RILEM Technical Committee TC 152 PBM-Task Group TG1 "Binders", which is now under the responsibility of the new RILEM TC PEB "Performance Testing and Evaluation of Bituminous Materials" initiated a second round robin test on binder rheology. The objective of this round robin test is to compare the results of dynamic rheometer (DSR or equivalent) and bending beam rheometer (BBR) measurements of different laboratories and on different binders. The aim of the test is to determine the repeatability and reproducibility of the test methods, to give recommendations for improvements in order to come to a reliable standard testing method. DSR – and BBR-measurements were performed on four different binders: one straight run bitumen, two SBS – modified binders (one low and one high modification) and an EVA – modified binder. In the first phase, the tests were performed on the original binders. In the second phase of this round robin test, the influence of RTFOT-ageing and long term ageing was studied by DSR and BBR measurements. Measurements of the softening point of the binders were also performed before and after ageing. DSR-measurements were also performed on a reference fluid, in order to find out if this material was suitable for calibration purposes. Detailed instructions were given concerning handling of the binders, calibration of equipment and testing. Questionnaires with details on the precise handling and testing procedures were distributed among the different laboratories as well. This paper concerns the analysis of the results of both phases of the round robin: the measurements on the binders prior to ageing and the measurements after RTFOT and RTFOT + PAV – ageing.


Long-term Ageing – Simulation by RCAT Ageing Tests
A. Verhasselt

The ageing of bituminous binders in service is an important factor in the long-term performance of bituminous materials. Relevant simulation by accelerated ageing can be achieved if the test is conducted at a temperature below 100°C. Comparison between the RCAT (Rotating Cylinder Ageing Test) developed by BRRC and the PAV (Pressure-Aging Vessel) test recommended by SHRP has shown that the reactions involved in these two ageing techniques are very similar (as opposed to what is observed with the RTFOT method). In this comparison, seven bituminous binders (two bitumen’s and five polymer-modified bitumen’s) were examined and results obtained for IR spectra, technological characteristics and rheological properties (DSR and BBR) were considered. Ageing time equivalence between the two test methods is, however, not a constant. The origin of this seems to lie in strong migration to the surface in the case of a static test and/or in degradation of the polymer; these types of behavior were demonstrated in a PAV-like test under nitrogen. With RCAT, a homogeneous sample of about 450 g of aged binder can be produced in one test run, and used for subsequent characterization and/or for further tests on asphalt mixes. In the marketed version of the RCAT apparatus, it is also possible to run, on a 550-g sample in the same device, a short-term ageing test at 163°C similar to the RTFOT test. The next step is to establish the equivalent ageing time for this application; this is now under way. With this apparatus, it will thus be possible to perform either the short-term ageing test and the long-term ageing simulation on separate samples of about 500 g of bituminous binder, or to perform the short-term ageing test followed immediately by long-term ageing simulation on the same sample. Enough material can be recovered after ageing to make some tests on asphalt mixes prepared with aged binder.


Determination of Asphalt Binders Viscosity From Other Rheological Parameters
M.O. Marasteanu and D.A. Anderson

The dynamic modulus of asphalt concrete is an essential parameter in any pavement design method. However, laboratory determination of the dynamic modulus of asphalt concrete is very expensive and time consuming. The new AASHTO 2002 pavement design method addresses this issue by using asphalt binder viscosity as one of the parameters in predicting asphalt concrete dynamic modulus. In addition, many studies indicate that asphalt binder viscosity is a better parameter in evaluating the rutting potential of asphalt concrete. The Superpave asphalt binder specification provides a direct method for measuring the viscosity of asphalt binders at mixing and compaction temperatures only. Currently there is no accepted method to determine the viscosity or zero shear rate viscosity at service temperatures. This paper investigates a number of different approaches of calculating viscosity from other viscoelastic parameters and proposes the use of a simple method based on rheological master curves.


Project for Developing Performance Related Standards in Europe; Evaluation of Test Methods to Characterise Bituminous Binders
A. Stawiarski, D. Jamois, W. Teugels, A. Madella, B. Lombardi and C. Robertus

The paper presents the results of the Eurobitume evaluation of candidate test methods for assessing the following binder properties, which are linked to the key asphalt pavement performance requirements:
– rheological property at elevated service temperature, in relation with permanent deformation,
– combined rheological and failure property, in relation with low temperature cracking,
– short term and long term binder ageing behaviour,
– binder stiffness modulus, in relation with asphalt structural strength,
– binder fracture behaviour, in relation with fatigue cracking.

The paper gives recommendations on which tests could be selected for standardisation as common European methods and for their possible use in future binder specifications. This work represents a contribution from the European bitumen industry to the project in progress, under the umbrella of CEN TC 336 Bituminous binders, for developing in Europe the next generation of performancerelated standards for bituminous binders.


Uncertainty and Appropriate Use of CBR Design for Airfields
D.R. Alexander, R.B. Freeman and C.R. Gonzalez, L.N. Lynch

The performance of recent pavement test sections instigated a review of the origin and use of the California bearing ratio (CBR) airfield pavement design procedure. The study involved inspection of the uncertainty associated with design solutions, given the uncertainty in the empirical equation itself. Variability in design thickness solutions, as produced by uncertainty in criteria, is reasonable. Variability is largest for relatively thin structures and/ or those that are subjected to relatively high load repetitions. Use of the CBR design equation for predicting remaining pavement life (i.e., load repetitions), given a pavement structure and traffic characteristics, leads to much higher uncertainty. The CBR equation evolved from research aimed at producing a thickness design tool and was not specifically developed for the purpose of evaluating pavements. It is best used for the purpose for which it was originally intended.


Asphalt Complex Moduli Determination Via FWD Test
R. Al-Khoury, A. Scarpas and C. van Gurp

This paper addresses the development of a technique capable of identifying (viscous) pavement parameters in elastic-viscoelastic layered systems via non-destructive tests and techniques of wave propagation. The spectral element technique is adopted for determination of the stress-strain relationship in viscoelastic materials like asphalt pavement structures. The mathematical formulation of spectral elements for Burgers viscoelastic material model is highlighted. Some aspects of the forward calculating model and the inverse technique are presented. The first model can be used for proper assessment of deflections, stress, strains, etc, whereas the inverse technique is suited for detailed analysis of deflection histories and the identification of the material’s Young’s moduli and complex moduli for layer systems. Examples are provided. It is shown that the proposed methodology leads to robust and computationally efficient forward and inverse calculations.


Structural Design of Airport Asphalt Pavements on Soil with High Water Table
Y. Hachiya, O. Takahashi, H. Kanazawa and K. Akimoto

Many airport projects have been constructed on ground reclaimed from the sea. At these sites, not only the subgrade, but also the pavement can be submerged because the water table generally remains high. When heavy loads are repeatedly applied to this pavement, severe damage to the pavement can result. The following three topics were studied to develop a structural design method for airport asphalt pavement at such sites: the condition of the pavement at an airport constructed on reclaimed ground, the influence of water on the pavement, and methods for constructing airport asphalt pavement at these sites. As a result, water-resistant airport pavement can be constructed if the following measures are used: reduce the design CBR of the subgrade, use asphalt stabilized material for the base, and install a drainage system surrounding the pavement.


Measuring of Dynamics Wheel Loads
E. Doupal, R.Calderara, R. Jagau

This presentation will provide a brief information on using high-accuracy weigh-in-motion (WIM) traffic data for the evaluation of the dynamic axle loading influence affecting the pavement deformation and rutting. The measuring method is based on quartz crystal WIM sensors. This technology enables not only accurate axle load determination, but also evaluating the dynamic wheel loads separately. WIM data are used for a wide variety of applications. One of the most important applications is the estimation of the current and future loading of bridge constructions and road pavements. This paper presents several practical applications of such measurements, data reliability and recommendations for their use. The WIM measuring systems equipped with long term stable and temperature independent sensors collect the required pavement loading and dynamic effects data in real time. These data can be used for experimental checking of pavement modeling methods and for their improvement.


New Guidelines for the Design of Flexible Pavements for Australian Conditions
G. Jameson, K. Sharp and D. Potter

This paper summarises the work undertaken to revise the guidelines for the design of flexible pavements for the 2002 edition of the Austroads Pavement Design Guide, which has recently been issued as a final draft for public comment prior to its release in 2002. The 2002 edition of the Pavement Design Guide incorporates substantial improvements that have resulted from Australian and overseas research into materials characterisation, traffic assessment, pavement performance studies and design methodology.


The Design, Creep and Fatigue Performance of Stone Mastic Asphalt
G. Stephenson and F. Bullen

Stone Mastic Asphalt (SMA) is recognized as an ideal surface layer in heavy traffic situations, due to good rut resistance properties which is attributed to the stone on stone contact of the mix skeleton. Methods of ensuring that the stone skeleton is not overfilled with mastic were investigated in this research. In Queensland, Australia, thin surfacing layers (<40mm for SMA10) are used, often over old, weak pavements that exhibit high deflection. Under such conditions, fatigue failure of SMA, rather than rutting, may be the limiting design factor. This paper reports on an investigation into using an applied vacuum as a confining pressure in the dynamic creep test. The fatigue performance of typical Queensland SMA mixtures was investigated using UMATTA equipment. The fatigue properties of SMA incorporating flyash, hydrated lime and ultra fine dust as the mineral filler and a SBS polymer modified binder are also investigated. The study included Queensland dense graded asphalt as a reference for quantifying any benefits obtained by using the various SMA mixtures. Some fatigue relationships for the SMA mixtures studied are proposed and the benefits of using SMA in combined creep-fatigue situations investigated. The testing programme demonstrated the effect of the type of filler and binder on the stiffness and fatigue life which has significance for generic specifications where a range of fillers are permitted.


A Damage Approach for Asphalt Mixture Fatigue Tests
D. Bodin, C. de la Roche, J -M. Piau and G. Pijaudiet-Cabot

Fatigue performances of asphalt pavement material are assessed using laboratory fatigue tests. It is proposed to apply damage mechanics to these tests to describe the stiffness decrease induced by microcrack development in the material samples. An isotropic nonlocal damage model is chosen as well as a cyclic time integration scheme for finite element implementation. Parameter identification is performed using tension compression fatigue test results. Validation on bending tests leads to good results in terms of damage field evolution and well describes the stiffness decrease before the macrocrack formation.


Determining Asphalt Concrete Fatigue Property
J -P. Michaut and J. Bilal

Assessing the mechanical characteristics of bituminous materials required to design pavement calls for arduous, long and expensive tests. In addition, the results at the end of the tests – when compared to specifications – are not certain. When the mechanical characteristic called fatigue performance is required, the designer must start by studying different mixes. Once the formula has been selected, fatigue tests are carried out to verify if the formula fulfils the requirements. When the results do not comply, the entire study must be redone, which is a loss of time and money. To simplify and facilitate mix design studies, equations have been determined to ensure – in our mind – sufficiently precise fatigue deformation values for the bituminous materials. This value can then be used in pavement design models. This article covers the elaboration of these equations drawn up from a fatigue data file compiled by the company and from other published results. These equations take account of the elements that are used by every mix designer, e.g., binder and compaction ratio. Backed by these equations combined with sound knowledge of the parameters required with these types of asphalt mix, the designer is now able to carry out fatigue tests that comply with the required specifications without having to worry about the results.


Laboratory Evaluation of the Effects of Aggregate Gradation and Binder Type on Performance of Asphalt Mixtures
H.J. Lee, J.Y. Choi, Y. Zhao and Y.R. Kim

This paper presents findings from a laboratory study on the performance of seven different asphalt mixtures. These mixtures were developed by systematically changing aggregate gradations and binders to investigate the effects of these mixture variables on performance. Dynamic Shear Rheometer (DSR) and Bending Beam Rheometer (BBR) tests were conducted on the binders to obtain temperature susceptibility. Axial creep, triaxial repeated permanent deformation, and uniaxial tensile fatigue tests were performed on the mixtures. Fatigue data were analyzed using the viscoelastic, continuum damage approach. The results from all the experiments were investigated together to evaluate the effects of different gradations and asphalt modifications. It was found that the Styrene- Butadiene-Styrene (SBS) modified binder improved the fatigue cracking and rutting resistance of the mixtures regardless of the aggregate gradation. The ranking of the performance among gradations was dependent on the distress type. In general, however, the dense gradation showed better cracking resistance and the coarse Superpave gradation showed better rutting resistance.


WAM-Foam, Asphalt Production at Lower Operating Temperatures
B.G. Koenders, D.A. Stoker, C. Robertus, O. Larsen and J. Johansen

Reducing asphalt production temperatures has significant benefits from both cost and environmental perspective. The hot mix asphalt industry has been aware of this for many years. The biggest challenge however has always been to achieve adequate asphalt mix quality at lower or ambient operating temperatures. Recently, new production processes at temperatures between 80 – 120°C have attracted much interest because of the possibility to approach or even obtain hot mix quality. In the Warm Asphalt Mixture (WAM) process using foamed bitumen this region is explored using conventional production and laying equipment. The WAM-Foam process has been tested in the laboratory and in road trials. The results show that for dense asphalt wearing courses the properties and the performance can be comparable with hot mixture. Additionally the benefits to the environment were confirmed by measured reductions in energy consumption, emissions and fume exposure.


Foamed Bitumen Mix: Soil or Asphalt?
A. Nataatmadja

The research described herein involved characterisation of a range of foamed bitumen mixes in the laboratory. The test results suggest that although foamed bitumen mixes can have some engineering properties which are comparable to those of hot-mix asphalt concrete or other stabilised soil mixes, they have some unique and distinct characteristics. With its low bitumen content and high air void content, the resilient modulus of foamed bitumen mix is affected by curing methods and moisture. Furthermore, it is also shown that while the stiffness of foamed bitumen mixes may be as high as that of typical asphalt mixes, it is much less sensitive to temperature variation. Similar to that of conventional asphalt mixes, however, the resilient modulus of foamed bitumen mixes varies with the applied strain.


The Use of a Designed Foamed Bitumen Stabilised RAP in an Urban High Street
K. Khweir, D. Fordyce and G. McCabe

An analytical design approach and performance-based specification was used with a roadway refurbishment in Scotland. The approach was taken as the roadbase used was a foamed bitumen stabilised RAP. In the refurbishment the mixture was fully compacted and production variables had only a limited impact on material performance.The most significant factor influencing material performance was found to be material moisture content. It has been suggested that the load spreading capability of such stabilised RAP mixtures can be compared to conventional hot applied bitumen mixtures through the use of a single equivalence factor. The data from this work would suggest this is not the case and the use of a single factor should be used with caution.


Performance Prediction of Cold Foamed Bitumen Mixes
K.J. Jenkins, M.F.C. van de Ven, A.A.A. Molenaar and J.L.A. de Groot

As the use of foamed bitumen mixtures in road pavements continues to grow on a global scale, so the need for performance functions that may be used in the design of pavements incorporating these materials becomes increasingly important. The challenge of modelling the behaviour of these mixes is complicated by the variety of foamed mixes that are produced and the range of properties that prevail. This paper focuses on the performance of cold foamed bitumen mixes that have low binder contents (generally less than 2,5%) and a model that could be used in pavement design. Mixes with and without active filler (cement) have been considered. Due to the lightly bound nature of these materials, the shear properties of the foamed mix were adjudged to be critical parameters for the prediction of permanent deformation under repeated loading. A model was developed on the basis of triaxial testing carried out on foamed mixes comprising a range of aggregate types, including marginal and recycled materials. This included monotonic and dynamic triaxial testing on large foamed and granular specimens (300mm phi x 600mm high) and intermediate specimens (150mm phi x 300mm high). The model was then validated using accelerated pavement tests (APT) with the Model Mobile Load Simulator MMLS on a layer treated with foamed bitumen using Cold in-place Recycling (CIPR). Through triaxial testing, the performance properties of the foamed mixes in terms of permanent deformation under repeated loading, can be related to the shear properties of the same material. This provides a link between laboratory mix design and field performance. Stress dependent models incorporating shear parameters, can be used to define the resilient foamed mix behaviour under dynamic loading. The ratio of deviator stress at failure under monotonic loading and the deviator stress in the pavement structure has been identified as a critical parameter for rut prediction. Using finite element methods that incorporate non-linear elements, these models can be applied in pavement analysis. The applicability of the rutting performance model has been verified using the NOLIP finite element analysis programme and guidelines for (lightly bound) foamed mix layer design now exist.


Influence of the Surface Energy Status of Aggregates On the Performance of Asphalt Hot Mix
O. Harders and I. Noesler

Aggregate selection for asphalt production is mostly governed by economic factors. This includes transportation costs, availability, capacity and seasonal effects. The storage time of the aggregates after crushing can vary remarkably. The surface energy of freshly broken minerals is changing over the time due to reorganisation of polar component and absorption of water and dust at its surface. This process takes some months to stabilise. The knowledge about the actual status of this energy is most important for application of bitumen emulsions as it has an influence on the breaking time and absorption. It can also have an influence on the performance of asphalt hot mix formulations. A work program included the investigation into the influence of different sources of minerals and granular sizes as well as the general behaviour of asphalt compositions. The interaction with different sources of bituminous binder was another topic of the investigation. The different surface activities were addressed by the methylene blue (MB) test and the performance of asphalt compositions was evaluated with a rolling wheel rut tester at 50°C under water.


The Characterization of Pavement Layer Interfaces
S.A. Romanoschi and J.B. Metcalf

The primary objectives of the research were to derive constitutive models for the interfaces of flexible pavement layers, and to determine the effects of interface condition on pavement performance. A simple constitutive model was derived for the asphalt layer interfaces using data provided by laboratory direct shear tests at several normal load levels. In the model, the shear displacement is proportional to the shear stress until the shear strength of the interface is reached. When tack-coat is not applied between asphalt layers, the interface shear strength and stress-displacement relationship depend on the magnitude of the normal stress acting at the interface. Field shear tests at several levels of normal load were used to study the bond between asphalt surface layers and granular or soil-cement bases. The interface constitutive models were integrated into the ABAQUS finite element model to prove the significant impact of the interface condition on pavement life.


The Importance of Good Bond Between Bituminous Layers
B.A. Hakim

Deflection testing using the Falling Weight Deflectometer (FWD) and the Deflectograph are often carried out on new pavements as a quality control measure to ensure that the pavement can carry the future traffic loading. Analysis of FWD data has indicated low bituminous stiffness and low residual life for some newly constructed pavements. Therefore, theoretically the pavement was not fit for purpose and strengthening by an overlay was needed. However, de-bonding between bituminous layers was noted in the majority of cores extracted from the pavement, where the layers were separated during coring under the torque applied by the rotary core barrel. Lack of bond between pavement layers will produce higher deflections under surface loading, since the layers will act independently in the absence of shear continuity at the interfaces. Lower bituminous layer stiffnesses are usually predicted from FWD test results if poor bond exists between different sub-layers. Therefore the back-analysed stiffness is an apparent value for the combined bituminous layers, which reflects the in-situ behaviour accounting for poor bonding, and is not a unique property reflecting material condition. The FWD data were re-analysed using a new method to predict both the pavement layer stiffness and the bonding condition between bituminous layers. The bond stiffness at the interface is considered as a variable affecting the FWD deflections and is therefore backanalysed in a similar manner to a layer stiffness. The results indicated better bituminous materials but with poor bond between the layers, and a good quality foundation. This was consistent with the laboratory Indirect Tensile Stiffness Modulus results carried out on the bituminous core samples. The implication of lack of bond between bituminous layers on pavement deterioration mechanisms and residual life prediction are discussed in this paper.


Experimental Investigation of Fracture and Healing of Bitumen at Pseudo-Contact of Two Aggregates
F. Hammoum, C.De La Roche, J -M. Piau and C. Stefani

The study presented here aims at highlighting the bitumen part in the fatigue of bituminous mixtures, due to cohesive failure in binder. In order to do so, a specific test Repeated Local Fracture of Bitumen has been designed [de La Roche and al., 1999]. It aims at understanding cohesive failure and healing phenomenon of binder. This article reports the analysis of the test results, regarding crack initiation and growth within the sample. Such events appear as steps in a load versus displacement curve. These observations are confirmed with the help of acoustic emission. Successive tests performed after different rest period duration show that binder healing depends on temperature and rest period duration. The developed test seems a very promising tool to study bitumen part in fracture and healing of bituminous mixtures.


Healing Characteristics of Asphalt Mixture Under High Temperature Conditions
K. Uchida, T. Kurokawa, K. Himeno and T. Nishizawa

In order to evaluate the healing characteristics of asphalt binders and asphalt mixtures, new test methods were developed. These methods include two procedures, one is to measure the bonding strength between asphalt surfaces and another is to measure a flexural strength of a specimen which is healed after artificially cracked and then estimate the healing potential which is defined as a ratio of the flexural strength of the cracked and healed specimen to that of a virgin specimen. In the latter procedure, loading time, temperature and degree of ageing can be variable. Using the methods, effects of loading time, temperature and degree of age on the healing characteristics were investigated. Also, cracking process at the surface of asphalt mixtures was observed using a modified wheel tracking test and discussed in terms of the aging and healing of asphalt mixtures.


Fracture of Bitumen Films
J.A.F. Harvey and D. Cebon

This paper investigates the adhesive properties of bitumen films. Tensile tests on bitumen butt joints and double cantilever beam specimens, over a wide range of test conditions are described. The test results are treated using a fracture mechanics approach. It is found that the failure strain of bitumen is generally independent of equivalent strain rate, in both the brittle and ductile regimes, whereas the ‘normalised toughness’, G/2h is rate independent in the brittle regime and rate dependent in the ductile regime (G is the strain energy release rate, 2h is the film thickness). The toughness of a typical paving bitumen is quantified, both for brittle fracture and ductile failure, and a mechanism map classifying the failure mechanisms as a function of temperature and strain rate is presented.


Alternative Bitumens for Road Building
A. Inman and E.E. Larsen

The rheological properties of bitumen deteriorate with time, this phenomenon is commonly called ageing. The main cause of this deterioration is oxidation. The oxidation of bitumens has long been a known problem in road building. The main factors that affect the rate at which bitumens deteriorate are the asphalt mix properties, exposure to elements, temperature, bitumen film thickness, and the bitumen ageing resistance capability. We have over a period of time investigated this phenomenon in connection with asphalt recycling projects in Africa and the Middle East.

The investigations carried out were based on the calculation of the Viscosity Index of the bitumens from asphalt mixes where recycled asphalt and rejuvenators were added. The bitumens were artificial aged by employing the Thin Film Oven Test (TFOT, ASTM D 1754) for 5 hours and a modified test of TFOT for 24 and 30 hours. The modified test method for TFOT for 24 and 30 hours was developed in cooperation with a Danish research laboratory.

The Investigations indicated that the long term ageing of these bitumens was dramatically less than conventional bitumens and this lead us to seek alternative bitumens for road building. In 1999 a road was designed in Tanzania where a Propane Precipitated bitumen was to be employed. The bitumen is supplied by a lubrication oil refinery. This paper presents the findings of the above mentioned investigations along with results of the bitumen investigation for the Propane Precipitated bitumen.


A Constitutive Material Modelling Methodology for Asphalt Mixtures
G.D. Airey, S.T. Dunhill and A.C. Collop

This paper presents a methodology for the experimental characterisation and modelling of asphalt mixtures using an elastic-visco-plastic constitutive model. The recommended methodology has been successfully used in the paper to characterise two standard UK wearing course mixtures. The characterisation of the mixtures is undertaken through a series of monotonic uniaxial compression and tension tests. Due to the significant influence of strain rate and temperature on the response of asphaltic materials, the tests have been carried out at three temperatures and three displacement rates. Two different test configurations have been used for the compression and tension tests comprising cylindrical and necked specimens respectively. The basic model parameters required for the constitutive model have been calculated solely from the results of the monotonic uniaxial test data. Relationships are presented to describe the material parameters for the two standard asphalt mixtures as functions of peak strength, strain rate, temperature and plastic work.


Viscoelastic Rutting Model with Improved Loading Assumptions
R. Blab and J.T. Harvey

A 3-dimensional FE model of a road pavement is presented, in which the temperature- and loaddependent performance of flexible pavements is characterized by a generalized Maxwell model and improved loading assumptions are used based on measurements of tire-road contact stresses. A method has been developed that permits the derivation of the required viscoelastic model parameters from dynamic shear tests. The model was evaluated using simulation calculations for a specific test structure on which rutting tests had been performed with a Heavy Vehicle Simulator (HVS). This evaluation demonstrated good agreement between the deformations predicted by the theoretical model and the deformations actually measured.


Design Aspects for Wearing Courses on Orthotropic Steel Bridge Decks
T.O. Medani, A. Scarpas, M.H. Kolstein and A.A.A. Molenaar

Preliminary analyses have shown that linear elastic theory is not applicable for the analysis of surfacings on orthotropic steel deck bridges. It is believed that an improved nonlinear material models need to be implemented. The ACRe material model developed at Delft University of Technology can successfully describe the different aspects of the behavior of asphaltic materials (e.g. elasticity, visco-plasticity, cracking). In this contribution the model parameters for a mastic asphalt mix have been estimated. There are tools, which can be used to better understand the interaction between the different components of the structure and the influence of specific structural material parameters on the response. These tools comprise accurate non-linear material models and smart and powerful finite element based programs. Furthermore, it is expected that such tools, after the necessary simplifications, will lead to more useful procedures that can not only be used efficiently at the design phase but also at the construction phase (quality control). Delft University of Technology, the Netherlands


Permanent Deformation of Asphalt Concrete Pavements: Development of a Nonlinear Viscoelastic Model for Mix Design and Analyses
F. Long, S. Govindjee and C.L. Monismith

Truck loading is increasing worldwide, resulting in more permanent deformation of asphalt concrete pavements. It is therefore necessary to ensure pavements can withstand this loading without rutting, which requires improvements to mix design and analyses. This paper discusses the development of a nonlinear viscoelastic constitutive model to describe the rutting behaviour of asphalt concrete. The model focuses on those properties that dominate asphalt concrete deformation; viscoelasticity and shear deformation. The model was formulated with laboratory test data, and includes both strain and temperature dependence using time-temperaturestrain superposition. To validate the model, finite element simulations of a laboratory test and of a pavement were compared to real data. The model predicts the laboratory test and pavement behaviour fairly well. The model is also used to evaluate the SHRP performance-based mix design procedure. The effects of different tyre types were found to have a significant effect on the development of rutting.


The Application of Shakedown Approach to Granular Pavement Layers
G. Arnold, A. Dawson, D. Hughes and D. Robinson

The concept of shakedown was developed for describing the material and structural response to the repeated application of a cyclic load. In this paper the concept is applied to describing the behaviour of unbound granular materials in the repeated load tri-axial permanent strain test. Behaviour was categorised into 3 possible shakedown ranges A, B or C, where A is a stable shakedown response and C is incremental collapse while B is intermediate. From this data test stresses near or at the boundary of shakedown range A and B were determined to define a shakedown limit line stress boundary in p (mean normal stress) – q (principle stress difference) stress space. The shakedown limit line was applied as a yield criteria in the finite element model of the field trial to predict whether or not shakedown occurs in the UGM for a range of asphalt cover thicknesses. For comparison the lower and upper bound shakedown theorems were also applied to the field trial cross-sections.


Modelling of Granular Layers in Pavement Constructions
S. Werkmeister, R. Numrich and F. Wellner

The shakedown approach was used to characterise the deformation behaviour of Unbound Granular Materials (UGM) in pavement constructions. In this paper it is shown that the application of the shakedown concept to UGM as used in pavement construction is possible, although adaptations have to be made. The essence of a shakedown analysis is to determine the critical shakedown load for a given pavement. The Dresden material law was introduced for modelling the permanent and resilient deformation behaviour of UGM-layers in pavement constructions under consideration of the shakedown ranges. A design method has been developed that utilizes test results from the Repeated Load Triaxial Tests to establish the risk level of permanent deformations in the UGMlayers. The non-linear resilient material law was implemented into the Finite Element Program FENLAP for the UGM-layer. These Finite-Element-Program was used to check, using a part of the German pavement design guideline, if the critical stress level for the UGM-layer (between stable and unstable behaviour) is exceed.


The Effect of Mix Design Technology on the Rutting Characteristics of Asphalt Pavement
G.A. Huber, W. Pine, J.F. Corte, F. Moutier and P. Langlois

Permanent deformation of asphalt pavements has been the focus of mix design systems since the early part of the twentieth century. Today, one of the main objectives of mix design remains the prevention of rutting. Several main types of mix design are in use in the world. Each method approaches the problem of rutting using a different set of criteria that include type and effort of compaction, and selection of volumetric criteria and asphalt content. After 1950, the Marshall method of mix design was used widely throughout the world. In the 1960’s and 70’s, LCPC developed an approach to mix design significantly different than the Marshall method. In the 1990’s the Superpave method of mix design was developed and is being implemented in the United States. The Province of Quebec developed a hybrid mix design system using Superpave equipment but applying LCPC design principles. In 2000, mix designs were independently done on the same set of materials using the four design methods and were built at the LCPC test track at Nantes, France. The sections were loaded during the summer of 2000 and rutting was monitored. This paper gives an overview of the mix design methods and a summary of the independently performed designs.


Effects of Fine Aggregate Properties on Rutting Resistance
J. W. Button, D-W. Park, A. Chowdhury, Dallas N. Little, E. Masad

The objective of this laboratory study was to compare rutting resistance of hot mix asphalt (HMA) containing fine aggregates with different angularities. Permanent deformation was measured using the Asphalt Pavement Analyzer (APA). HMA mixtures were designed following the Superpave specifications, to the extent possible, and prepared using the Superpave gyratory compactor. A few mixes did not meet all Superpave specifications. Six different mixtures were prepared using the same gradation. In all mixtures, crushed limestone was used as the coarse aggregate. Six different fine aggregates were used: crushed granite, crushed river gravel, crushed limestone, sub-rounded natural sand, blend of 85% granite with 15% natural sand, and blend of 70% limestone with 30% natural sand. Angularity of the fine aggregates was measured using Superpave fine aggregate angularity (FAA), direct shear test, compacted aggregate resistance (CAR), and different image analysis techniques. FAA varied from 39.0 to 48.0; the other measured values of angularity varied similarly. Findings from APA testing indicate (1) HMA mixtures containing natural sand with a FAA of 39 and those containing river gravel fines with a FAA of 44.3 yielded statistically equivalent rut depths, (2) HMA mixtures containing crushed limestone fines with a FAA of 43.5 and those containing granite fines with a FAA of 48.0 yielded statistically equivalent rut depths, (3) HMA mixtures containing crushed limestone fines with a FAA of 43.5 yielded significantly lower rut depths than similar mixtures containing river gravel fines with a FAA of 44.3, (4) HMA mixtures containing a fine aggregate blend with a FAA of 41.9 gave essentially the same rut depth as similar mixtures containing a fines blend with a FAA of 46.8, (5) FAA is not sensitive to rut resistance of HMA mixtures. Further, certain fine aggregates with FAA values lower than 43, but with relatively high particle surface texture, can produce mixtures with relatively good rut resistance, (6) Angle of friction derived from direct shear tests and some of the image analysis parameters showed very good correlations with rut depth.


Asphalt Mixture Design on the Criteria of Shear and Crack Resistance of the Pavement
G.N. Kiryukhin

An asphalt mix design on the criteria of shear and crack resistance calls for determining the main rheological and strength properties of the pavement asphalt concrete characterising its deformation and damage depending on its stressed-strained state and temperature-time conditions of loading, determinations being carried out under laboratory conditions. Algorithms for determining the asphalt concrete shear resistance at the maximum design temperature, residual deformation for a design service life of the pavement and its crack resistance due to the combined action of tensile stresses in the pavement are proposed. The design method makes it possible to choose an efficient composition of asphalt concrete on two criteria of shear resistance and a complex criterion of crack resistance of the pavement.


Use of Gyratory Compaction for the Design of Asphalt Mixes
L. Francken and L. Didier

After presenting the basis principles of the Belgian volumetric mix design method [BRRC, 1997], this paper describes experimental work carried out in order to improve the method and enlarge its field of application. Gyratory tests were carried to determine input data needed by a packing model used to evaluate the distribution of volumes within asphalt mixtures. The sigmoidal compaction curve proposed by Moutier [1996] was used to fit the experimental results obtained on different road mixes. The improvements proposed to the model take into account on one hand the grading curve and characteristics of the granular mix components and on the other hand the energy or rate of compaction. Results of this study will be incorporated in the PRADO [Francken & Vanelstraete, 1993] mix design software together with a set of tools for predicting mechanical characteristics and performance laws.


Use of the Concept of Pore Pressure in Unsaturated Soils for Evaluation of Rutting Potential of Asphalt Paving Mixes
R.B. Mallick, M.R. Teto and J.E. Haddock

Rutting is a common problem in hot mix asphalt (HMA) pavements. For characterizing HMA, a theory has been developed on the basis of the concept of pore pressure in unsaturated soil mechanics. It is hypothesized that in a mass of HMA, the mastic acts similar to water in a partially saturated soil mass. It is proposed that a measure of increase in lateral pressure that is generated in a mix during gyratory compaction can give a good indication of the generated pore pressure in the mix – the more the lateral pressure, the more is the generation of pore pressure. Results from tests with a variety of mixes show that there is an increase in pore pressure, and hence reduction in shear strength with increase in asphalt content and at air voids lower than 2 percent, and that compared to dense graded HMA, SMA mixes remain relatively insensitive to low air voids.


Provisional Validation of the New South African Hot-Mix Asphalt Design Method
A.T. Visser, F. Long, B.M.J.A. Verhaeghe and A. Taute

South African asphalt technology practitioners recognised the need for an updated and expanded hotmix asphalt (HMA) design system. This resulted in the launching of a three-year project aimed at developing new HMA design guidelines for South Africa. This new design system is centred on spatial design concepts and established knowledge of mix design, with additional validation of expected performance being attained through laboratory tests and prediction models. The aim of the paper is to present the initial results of a evaluation programme in which the new method was carried out in parallel with established practice. The paper presents the experimental programme and the results of the parallel testing. Aspects that were considered were rutting, durability, moisture sensitivity and permeability, as well as constructability issues. The new method provides the designer with greater insight and flexibility to overcome problems such as tender mixes, difficulties with construction and inability to achieve density and impermeability.


Repeated Simple Shear Test for Mix Design: A Summary of Recent Field and Accelerated Pavement Test Experience in California
J. Harvey, I. Guada, C. Monismith, M. Bejarano and F. Long

The repeated simple shear test at constant height (RSST-CH) is a laboratory test to evaluate mix rutting performance. The test was developed as part of the SHRP program, which ended in 1993. The test is designed to isolate the primary mechanism for rutting of asphalt mixes: permanent shear deformation at elevated temperatures under repeated loading. The RSST-CH is a candidate for use as a simple performance test and performance prediction test. Since its development, the RSST-CH has been used for a large and varied number of field projects and for mixes used in accelerated pavement test sections in California. The projects include mix designs and forensic investigations. Representative examples presented in this paper include:

– Mix designs for a warranty pavement for a state highway in the mountains near Redding California

– Comparison of RSST-CH results with rutting performance under the HVS for asphalt-rubber and conventional overlays

– Mix design (conventional and polymer modified binders) binder content selection for the I-710 freeway reconstruction near Long Beach, California and evaluation of these materials for rutting with the Caltrans Heavy Vehicle Simulator

– Comparison of RSST-CH results with rutting performance under truck trafficking at the Westrack accelerated pavement testing experiment.

The performance prediction methods used for these projects are reviewed. The use of other tests performed using the shear tester, such as the shear frequency sweep, and shear stiffness from the RSST-CH are also compared with observed performance on field and HVS sections. Based on the case histories, this paper summarizes use of the RSST-CH for mix evaluation and design, and the design criteria used for mix design and pavement evaluation.


Evaluating Tensile Strength of Asphaltic Paving Mixtures Using a Hollow-Cylinder Tensile Tester
G.G. Al-Khateeb and W.G. Buttlar

A Hollow-Cylinder Tensile Tester (HCT) was recently developed to obtain fundamental tensile properties of asphalt paving mixtures at low and intermediate temperatures. The HCT was developed as a surrogate test device for the Superpave Indirect Tensile Test (IDT), and can measure creep compliance, tensile strength, dynamic modulus, and resilient modulus. Because the HCT is a simple, portable, and rugged test device, it is a very useful and powerful tool for mixture design, control and forensic evaluation. This paper investigates the use of the HCT and IDT to measure the first failure and ultimate tensile strength of asphalt paving mixtures. As expected, ultimate tensile strength was found to be highly test-dependent. In a carefully controlled study involving four different polymer modification levels, the HCT ultimate tensile strength was found to be strongly dependent upon modification level, as opposed to the IDT ultimate tensile strength, which was weakly correlated to modification level.


Evaluation of Stability and Sensitivity of Hot-Mix Asphalt Mixtures Using Gyratory Shear Strength
B. Birgisson, D.D. Darku, R. Roque and B.E. Ruth

At the time of development, the SuperPaveTM gyratory compactor was developed only to densify mixtures. Since the advent of the SuperPaveTM gyratory compactor, a number of SuperPaveTM compatible gyratory compactors have been developed that measure the gyratory shear strength during compaction. This paper explores the use of the gyratory shear strength parameter obtained in SuperPaveTM compatible compactors as an index of mixture sensitivity and stability, while maintaining the SuperPaveTM mixture design framework. Eight different mixtures were prepared and evaluated, including four fine-graded and four coarse graded mixtures, as defined by whether the gradation curve passes above or below the restricted zone, respectively. The aggregates used all had known rutting performance histories in asphalt mixtures in Florida. The results show that the gyratory shear strength can be used as an index to evaluate both the stability of hot-mix asphalt mixtures, as well as the sensitivity of mixtures to asphalt content.


Development and Validation of a Pavement Response Evaluation Model
R.V. Siddharthan, P.E. Sebaaly, M.M. El-Desouky and M. El-Mously

The paper provides the following relative to a recently developed finite-layer based pavement response evaluation model:
(1) verification of its predictive capability and
(2) demonstration of its applicability.

The model incorporates important pavement response factors such as the noncircular contact area, complex contact stress distributions, vehicle speed, and viscoelastic material characterization. The verification of the proposed approach and the ensuing computer program was undertaken using:
(1) results from ELSYM5, a widely used pavement response model and
(2) field measured strain responses from a well instrumented road test carried out under a variety of loading conditions (load level and vehicle speed).

A very good agreement was observed in all cases. This paper also documents many important pavement responses computed by the proposed model under a variety of loading conditions. Pavement responses for a conventional single axle with dual tires traversing two types of pavements with different contact stress distributions have been reported.


Non-uniform Tire Pressure Effects on Surface Initiated Cracks with Everflex: a Three Dimensional FEA Tool for Flexible Pavements
H. Wu, J.P. Mahoney, G. Turkiyyah, L.M. Pierce, J. Uhlmeyer, R. Mulvaney

This paper presents a preliminary examination of the causes of surface initiated cracking. Of special interest are non-uniform tire pressure effects. These effects are examined via EverFlex–a three-dimensional finite element analysis tool for flexible pavement systems. The interactive and computational features of EverFlex are also presented. A case study is presented based on Minnesota Road Research Project sections.


Evaluation of Non-uniform Tyre Contact Stresses on Thin Asphalt Pavements
M. De Beer, C. Fisher and F. J. Jooste

Improved quantification of the shape and distribution of actual tyre-pavement contact stresses resulted in enhanced definitions of 3D-tyre-pavement contact stresses for the design and analysis of flexible pavements. It is now possible to describe the 3D-load/stress regimes with a series of discrete load values that were measured using Stress-In-Motion (SIM) technology, as well as being predicted from trained Artificial Neural Networks (ANNs). This paper illustrates the importance of these new load/stress inputs in pavement design on pavements with relatively thin asphalt surfacings, typically used in southern Africa. This investigation concentrates on the quantification of several pavement response parameters as a result of non-uniform and non-circular shaped contact stresses at different asphalt moduli, compared to the traditional circular uniformly loaded tyre patch. The Finite Element Method (FEM) was used for the detailed analyses. In addition, the responses of an un-cracked and cracked pavement structure were also compared, under circular and rectangular loading shapes of varying levels of loads and contact stresses using a customised semi-analytical FEM code. The response data were benchmarked with multi-layered linear elastic theory, and also indicated the importance of both the load shape and level of contact stress on pavement performance. The importance of using and managing actual tyre-pavement contact stress data for more rational design and analysis of pavements incorporating thin asphalt layers is highlighted.


Prediction of the Behaviour of a Flexible Pavement Using Finite Element Analysis with Non-Linear Elastic and Visco-Elastic Models
P. Hornych, J -P. Kerzreho and S. Salasca

In the past few years, LCPC has been implementing in its finite element code CESAR-LCPC a module for pavement modelling, including non-linear models for asphalt materials and unbound granular materials. This program has been used to model results of a full scale experiment on a flexible pavement, with a granular base, performed on the LCPC accelerated pavement testing facility. The experimental results indicated that the response of the pavement depends strongly on the level of load and on the water content of the unbound granular layers. The modelling of the pavement was performed in 3D, and several modelling hypotheses were successively tested : linear elasticity, a non-linear elastic model for the unbound layers (Boyce model) and a visco-elastic model for the asphalt concrete (Huet-Sayegh model). The most complete model, coupling non-linear elasticity for the unbound materials and visco-elasticity for the asphalt concrete, led to realistic predictions of the pavement response for different levels of load and different loading speeds.


Effects of Frost Heave on the Longitudinal Profile of Asphalt Pavements in Cold Regions
S.Kameyama, M. Kato, A. Kawamura, K. Himeno and A.Kasahara

Longitudinal profile of an Expressway in a cold region of Japan was measured weekly during winter using inertial profiler. The International Roughness Index (IRI) calculated from the measured profiles was analyzed to determine its seasonal changes during the coldest period. While the IRI of segments including embankments and bridges changed little during winter, the IRI of some of the segments, including cuts, was approximately double what it was in fall. A tendency of increasing IRI with increasing freezing index was observed in these segments. This indicated that the IRI’s increase in winter was affected by frost heave. Multiresolution decomposition using a wavelet was applied to the longitudinal profile to analyze the wave characteristics of the profile affected by frost heave. It was found that the component of the decomposed profile whose wavelength was 4 through 16 m greatly increased as the freezing index increased.


Development of a Rational Design Procedure for Pavements Subjected to Frost Action
G. Dore, N. Rioux and P. Pierre

In northern climates, frost action is a major cause of pavement deterioration. In current pavement design approaches, modern analytical tools are commonly used to assess the adequacy of pavement structures with respect to the effect of traffic. However, in most design methods, procedures to take environmental factors into consideration are still essentially empirical. At the most, some design methods include procedures by which material properties vary as a function of moisture and temperature. A new method for the verification of pavement structures with respect to the effects of freezing and thawing is proposed. The method is based on a four steps iterative procedure. In the first step, the mechanical and thermodynamic response of the pavement structure to traffic and climatic conditions are calculated. The calculated responses are then used to predict performance, through distress specific models, in the second step of the procedure. The results are compared to the specified project objectives in the third step. If the objectives are not met, the fourth step of the procedure includes a decision three to help enhance the pavement structure. A preliminary version of the method is being implemented in the province of Quebec. It has been validated for typical Quebec road network conditions through the collection of specific performance data from typical road sections. The validation project has also included the development of operational procedures to measure the new parameters required by the method.


Seasonal Variation of Moisture and Bearing Capacity in Roads with a Thin Surface Dressing Wearing Course
S. Erlingsson, G. Bjarnason and V. Thorisson

Road structures in Iceland undergo annual freeze-thaw cycles, giving rise to environmental fatigue of the structure. In order to better understand the effect of environmental factors on the response and durability of the road structure, seasonal variability of moisture and temperature have been monitored for three years in the base and the subbase layers of three old test road sections in SW Iceland. Structural stiffness has further been measured regularly using the Falling Weight Deflectometer (FWD). The volumetric moisture content varies significantly, depending on seasonal variance (freeze-thaw) as well as the precipitation. Both a long-term seasonal variation and a short-term variation can be seen. The amount of moisture increases during short thawing periods during the winter and during the spring thaw period. During rainy periods the moisture in the layers increases, especially in the upper base course layers of the road, but is reduced quickly after the rainfall stops. From the FWD results the deflections and back-calculated moduli show good correlation with the measured moisture content in the pavement structure.


Top-Down Cracking, Damage and Hardening in Practical Flexible Pavement Design
N.H. Thom, Y. Choi and A.C. Collop

This paper presents an approach to design against traffic-induced fatigue cracking, which allows topdown cracking, damage and hardening to be taken into account. The resulting computer-based prediction method makes use of recent laboratory research data on crack propagation and damage growth, both modelled in terms of calculated strain under load. Predictions relate first to damage growth, making the assumption that this is in the form of micro-cracks, which accumulate and grow in every part of the pavement experiencing tension. When cracks can no longer be considered small, strain in the region of the crack tip is related to propagation rate. The analysis methods used to obtain strain under load are sufficiently accurate for sensible usage, but incorporate engineering simplifications, avoiding the need for finite elements computations. Calculation of strain near the surface allows prediction of top-down cracking. Computation is incremental, first as damage accumulates and then as individual cracks grow.


An Improved Tool for Structural Design of Flexible, Composite and Rigid Structures
D. Leonard, L. Francken

New structural design software is presented that deals with flexible, composite and rigid structures. The pavement design is based on a trial-and-error process involving the use of the linear elastic multilayer model for flexible and composite pavements and the strength of materials (Westergaard’s theory) for the rigid ones. Evidence of illogical results in terms of fatigue performance using current pavement design philosophy for composite pavements are shown. Therefore, a new way of design is proposed to clear this issue.


Thin Asphalt Pavements on Soft Soil
C.A.P.M. van Gurp and A.J. van Leest

This paper presents results of a study on the development of an integrated design and structural evaluation procedure for thin asphalt pavements. The procedure incorporates geotechnical design features, pavement design and evaluation, and the assessment of the reinforcing effect of geosynthetics in unbound road bases. The methodology uses the strain at the bottom and surface of the asphalt layer, the compressive strain at the top of the bound road base, the stress state at middepth of the unbound road base, the strain at the bottom of the bound road base, and the vertical compressive strains at the top sub-base and subgrade as design criteria. For the design of structural overlays, the load-carrying capacity of the pavement structure is expressed by the so-called Modified Structural Number, SNC, to circumvent the tedious procedure of backcalculation of layer stiffness moduli. An especially developed software tool called ‘QUASAR’ enables easy and very quick use of the multitude of equations to assess the structural condition of asphalt roads and to determine the required strengthening.


In-Place Densification of Hot Mix Asphalt and Verification of Superpave
B.D. Prowell, E.R. Brown and M.H. Huner

In order to determine the optimum asphalt content for hot mix asphalt, the correct laboratory compaction effort needs to be applied during the design phase. The laboratory compaction effort for the Superpave mixture design system has been subject to refinement since first being introduced in 1994. The objective of this study was to verify the accuracy of the existing gyration table for the Superpave gyratory compactor. The verification includes data from two major research efforts, an evaluation of 40 field projects and the evaluation of 46 test sections of an accelerated loading test track at the National Center for Asphalt Technology. The results indicate that modified binders reduce the rate of pavement densification. The data appears to validate the currently specified gyration levels for low design traffic levels. Significant differences were found between the compactive efforts of different brands of Superpave gyratory compactors. These differences may be attributed to different internal angles applied to the samples.


Measurement of Edge Effects on Pavements with Thin Asphalt Surfacing
J. Aksnes, I. Hoff and H. Mork

A full-scale pavement test has been performed at Sandmoen near Trondheim. Two different pavements were constructed, one of them were instrumented for the measurement of load responses at different depths and offsets from the pavement edge in the base and subbase layers. Results from controlled traffic and plate loading are reported in this paper. Vertical stress has been successfully measured both under the base and the subbase layers. The results show an increase in vertical stress towards the pavement edge. This effect is evident at the bottom of the base layer for load positions closer than one meter from the edge. Increasing horizontal stress at the bottom of the base layer is recorded when a load is applied on the pavement surface. Deformation measurements show large plastic strains in granular materials, especially near the pavement edge. Parts of these strains recover when the load position is changed.


Reflective Cracking in Asphalt Overlay on Existing PCC
F. Zhou and L.Sun

Reflective cracking is the main premature pavement distress of the asphalt overlay on existing PCC in China. A new anti-reflective cracking measure-Special Steel Grid (SSG) was developed to retard the reflective cracking. The preliminary test results showed that the SSG1 strongly reinforced the asphalt concrete and performed the best. In order to investigate the process of initiation and propagation of reflective cracking and validate further the effect of SSG1 on retarding the reflective cracking under repeated loads, both laboratory scale test pavements and field test roads were constructed. Both results indicated that SSG1 could effectively retard the occurrence of reflective cracking. And the double reflective cracks were observed in both laboratory and field. Furthermore, 3-D FEM was used to analyze the 3-D stress field in composite pavement structure (AC/PCC) in order to explain the observed phenomena. It was found that the interface debonding and associated horizontal tensile stress made the double reflective cracking initiate and propagate vertically in asphalt overlay. Finally, the optimal thickness of asphalt overlay for old PCC with different foundation was recommended based on the results of both 3-D FEM and field test roads.


Laboratory and Field Experimental Investigations About the Properties of a Bituminous Thin Layer with Porous Aggregate
A. Bucchi, G. Dondi and A. Simon

A presentation and a critical analysis are made about the physical and functional characteristics of an ultra-thin porous layer. This work in particular deals with experimental tests carried out on a bituminous mix having part of the aggregate constituted of expanded clay. This mixture represents an improvement of the usual antiskid layer with discontinuous grading, concurring to obtain a lightening of the mixture and an increase of the skid resistance. The experimental study allowed to analyse the methodologies of employment of the gyratory compactor and the procedures of volumetric design. The analysis has concurred to verify also the possibilities of production and the performances of this bituminous mix. The comparison between the preliminary laboratory tests and in-situ tests for skid resistance, permeability and void index, allows to verify the data obtained during the optimisation study. The results obtained during construction also provided valuable indications on the parameters for design and quality requirements.


Temperature and Density Differentials in Asphalt Concrete Pavement
K.A. Willoughby, J.P. Mahoney, L.M. Pierce, J.S. Uhlmeyer and K.W. Anderson

The Washington State Department of Transportation (WSDOT) has conducted an examination of open-textured areas in hot- mix asphalt paving over four construction seasons, 1995, and 1998 through 2000. These study programs examined over 60 production paving projects to determine the cause of these open-textured areas in the finished pavement. In 1995, examination of these areas revealed that a potential cause may be temperature differentials in the mat. The 1998 study program eliminated aggregate segregation as the only cause. WSDOT focused on the types of equipment being used (material transfer devices/vehicles, pavers, rollers, etc.) to investigate any patterns of temperature differential occurrence in 1999. In 2000, nuclear density testing through the open-textured areas was performed to determine the severity of density differentials in the mat. The bottom line is that pavements that experienced large temperature differentials during placement produced substantial density differentials in the finished mat.


Factory Production Control According to the New PREN 13108-21
E. Eustacchio and G. Klimisch-Ibler

The European Standard prEN 13108-21 "Bituminous mixtures – Quality – Part 21: Factory Production control" describes how mixing plants shall operate their production control. In an Austrian research project six mixing plants of different age and design were investigated to show the effects of a quality management system as described in prEN 13108-21 on the results of mixture control. It is intended that the results of this project help to estimate future quality costs and their effects on the price of bituminous mixtures. On the other hand the results should assure that the frequencies set in prEN 13108-21 are realistic and that mixing plants are able to fulfil the requirements of this FPC system.


Performance Requirements on Asphalt Mixtures/Layers in Asphalt Contracts
N. Ulmgren

New forms of contracts, where requirements are expressed in performance characteristics on the asphalt mixtures or on the built in asphalt layers instead of the traditional recipe approach have been introduced in Sweden since a couple of years. The performance parameters that are at first hand used in these contracts are abrasion, stability, bearing capacity and water sensitivity all measured on the mixtures or on cores from the built in layers. All these functional methods are at hand as prEN-standards. Both parties are new to these new forms of contracts and the client has to learn how to express his needs in functional terms and the contractor has to learn to produce a suitable mixture and pavement and take full responsibility for the results for a longer period of time. In the paper the development of criteria for contracts stipulating performance characteristics on the asphalt mixtures or on the built in asphalt layers are described. The paper also gives an account of the experience so far as of how the different laboratory methods may give a rightful forecast of the performance of the asphalt materials in the road.


Alternative Contracting Models for Maintenance and Rehabilitation of Pavement Networks
R. Haas, C. Raymond, J. Yeaman and L.C. Falls, I. Rickards

The privatization of maintenance and rehabilitation for road networks can be in the form of several types of alternative contracting models. These range from privatized or outsourced maintenance, to public-private sector partnerships, to competitive or legislated highway maintenance to privatized highway asset management and maintenance, all generally involving terms of five years or less. Another type of model involves long term performance based contracts with terms of ten years or more. The allocation of risk between the public sector owner and the contractor varies with type of model. Any public sector owner contemplating privatization should carefully assess the pros and cons of each type of model as well as a number of influencing factors. The paper contends that despite a number of issues involved, the long term performance based contracting model offers the best potential for innovation and cost savings. Finally, it presents recommendations toward maximizing these benefits.


User Delay Impacts of Alternative Traffic Plans for Maintenance and Rehabilitation Interventions
S.L. Tighe and R. Al Assar

Whether or not user delay costs due to maintenance and rehabilitation (M&R) interventions should be included in life cycle analysis is a controversial subject. However, there is a growing awareness or acceptance by public agencies that they should be considered either directly, in life cycle analysis, or at least indirectly in terms of the "down time" represented by delays. In fact, there is some contention that this down time is no different than that occurring when workers are not on the job due to illness, etc.A number of quite sophisticated user delay models have been developed which calculate slowing plus queuing delays, and the associated costs. A comprehensive model was developed in Ontario’s, Ontario Pavement Analysis of Costs (OPAC) 2000 pavement design system, as reported at the 8th Conference in Seattle in 1997. This paper first identifies the available models, their principal features and their limitations. Among the limitations for most of the sophisticated models is the extensive requirement for input data, which in turn makes them cumbersome to use. The real issue suggested in the paper is twofold: (a) obtaining quantitative numbers on delay times and costs of sufficient reliability for the life cycle analysis, and (b) being able to evaluate the alternative traffic plans, including detours. To address these issues, the paper shows how user delay calculations can be simplified to yield approximate but still sufficiently reliable numbers for life cycle comparison of alternative M&R strategies as well as comparisons of a wide range of alternative traffic plans.


Selection of Pavement Maintenance By Use of Accelerated Load Testing
C.B. Nielsen, J. Rosenberg and J. Raaberg

A three-year research project was commenced in 1998 at the Danish Road Institute to establish a basis for evaluating the flow rutting and weathering resistance of specific types of Danish asphalt pavement materials. The objective of the present paper is to combine the definition of road classes, with respect to flow rutting and weathering resistance, with an empirical model to estimate the development of rutting in the asphalt pavement. The empirical model is based on Accelerated Load Testing in the Danish Asphalt Rut Tester (DART). Different maintenance strategies are evaluated from an economic and technical point of view and the benefit from accelerated load testing is assessed.


Towards a Performance Related Seal Design Method: New Empirical Test Method Using Scaled Down APT and Theoretical Performance Model
T. Milne, M.F.C. van de Ven and K.J. Jenkins

Practice has shown that bituminous road surfacing seals, with "straight" or modified bituminous binders, do not always behave as expected or predicted. An investigation towards a performance related seal design method was thus initiated at the Department of Civil Engineering, University of Stellenbosch during 1998. Part of this investigation focuses on the development of a new empirical test method using the scaled down accelerated pavement tester (model mobile load simulator) for seal performance prediction. The development of the new empirical test method is provided, with the findings of the first series of test results, where the performance of straight and modified single seals is compared. Current research is examining comparative performance of the different seal binders at road surface temperature at or approaching softening point (50ºC). The range of seal binders tested includes bitumen roller EVA, SBS and SBR modified binders, and the 80/100 penetration grade bitumen control. The development of a theoretical model aimed at assisting with the prediction of road surfacing seal serviceable life is discussed. The empirical test results will be used to assist in the evolution of seal design improvements where applicable, and in the calibration of the theoretical model. The comparison of empirical and theoretical model parameters relating to applicable performance aspects of the road surfacing seal is made. Aspects examined through literature search and original research include the seal components (aggregate, binder), pavement, environment and imposed load, and seal design and performance.


Compaction of HMA Using a High Frequency, Double Drum Vibratory Roller
J.A. Scherocman, Y. Nohse and K. Hokari

The purpose of this research is to determine the relationship between roller speed, vibratory frequency, and number of roller impacts per meter and the effect of this relationship on the density obtained for a hot mix asphalt (HMA) pavement layer. Roller test sections were constructed in Japan, California, and Kentucky using various combinations of roller speed (two or three levels), roller vibratory frequency (two or three levels) and number of roller passes (two levels). It was determined that for a given vibratory frequency, as roller speed increased (number of impacts per meter decreased), density decreased. It was also determined that, in general, for a given roller speed, density increased as the vibratory frequency increased (impacts per meter increased).


Very Thin Asphalt Concrete as Runway Wearing Course. Quality Control by Mean of Microscopy Techniques
H.C. Korsgaard and H.H. Bunner

In 1998 the first part of the main runway, 04L-22R, at Copenhagen Airport was successfully tendered for the resurfacing of the approximately 30.000 m2 wearing course. The works included the milling off of the top 20 mm of the old wearing course and the construction of a new 20 mm thick wearing course type where the asphalt paving machine is spreading 1,2 – 1,4 kg/m2 of emulsion in front of the open graded asphalt concrete being paved. The rest of the runway has successively been resurfaced during the period from 1999 to 2002 using the same method. The tender documents were based on the new type of specifications, which are based purely on functional requirements in this case including end-performance requirements after an eight-year warranty period. Amongst other topics the requirements used in the specifications are described below. During construction of the new wearing course, the client controlled the work intensively in order to be able to evaluate the actual quality of the new pavement and in order to be able to predict the development of the performance of the pavement using the first test results as a platform for the comparison with future measurements. For these purposes traditional laboratory test methods as well as new test methods like thin section microscopy were used. This has placed Copenhagen Airport A/S in a position making it possible for Copenhagen Airport to assess whether the maintenance methods and the maintenance efforts of the contractor during the eight-year warranty period are satisfactory or not in relation to the functional requirements of the new wearing course. The output of the traditional laboratory analysis as well as the output of the analysis of thin section microscopy during the four-year period is described below. The use of this pavement type has resulted in unexpected advantages. One of the unexpected advantages is that compared to conventional pavements for runways almost no rubber from the wheels of the aircraft fastens to the pavement surface and therefore no derubberisation work has been necessary after four year’s use. Another unexpected advantage is that much less de-icer material is needed for the de-icing of the pavement compared to traditional runway pavements.


The Rehabilitation of the Ricchieri Highway in Argentina with SMA and Thin-SMA Technologies
P.E. Bolzan

The present paper describes the application of Stone-Mastic Asphalt (SMA) and thin-SMA technologies that were used for the rehabilitation of the Ricchieri highway near the international airport of Buenos Aires, Argentina. This dual carriageway connects Buenos Aires city with the international airport and carries more than 120,000 vehicles per day. Due to the presence of reflective cracking (composite pavement sections), rutting, and lack of sufficient skid resistance, the road authorities conducted an investigation into the functional and structural conditions. From the rehabilitation techniques available, a variable thickness flexible overlay of SMA application was selected. The main goal was to cover both the functional and the structural needs with a single layer application together with some localised repairs. As a result a new heavy duty, high skid resistance wearing course surface has been obtained with a single application on a heavily trafficked highway. At present the performance of both SMA and thin-SMA layers are excellent.


The Effect of Reconstruction on the Stiffness, Balance and Stress Dependency of Pavement Layers
P.W. de Bruin, A.T. Visser and G.J. Jordaan

Pavement reconstruction / rehabilitation usually involves the re-use and retension of existing pavement layers. Often, these layers are retained as sub-layers with the addition of strength in terms of new layers on the top of the pavement. Hence, in order to accurately analyse the behaviour of the rehabilitation effect, both the old as well as the new pavement layers need to be characterised accurately. Currently, little or no information or guidelines are available to assist the researcher or design engineer to accurately predict or simulate the elastic properties of the rehabilitated pavement. The objective of the paper is to demonstrate the effect of rehabilitation actions on pavement properties such as balance, pavement structure behaviour (stress-stiffening and / or stress-softening etc) and stress dependency. An experimental section of road was evaluated before and after rehabilitation and the stiffness and associated stresses were determined in situ for the different uniform sections for modelling purposes along the road. It was found that the elastic modulus changed dramatically with the addition of new layers, even if the layer was not disturbed.


Cold Mix Recycling of Milled Pavement Material for the Construction of Low-Volume Roads in Saudi Arabia
A. Al-Dughaim

Due to the vast network of low-volume roads in the Kingdom and the important social and economic benefit provided by them, the Ministry of Communications (MOC) is keen on the enhancement of low volume road sector through the application of appropriate technology. Cold mix recycling is one of the techniques to economically upgrade the existing low-volume tracks and to construct good quality new low-volume pavements. A substantial quantity of milled material ( asphalt ) is generated from the rehabilitation and maintenance projects of the existing pavements. Several hundreds kilometers of low-volume roads have been constructed by performing cold mix recycling of the milled material generated from these projects. Both emulsion and cutback asphalt have been used as recycling agents in preparing cold mixes. The roads constructed in Riyadh and Hail districts are evaluated through visual condition survey and laboratory characterization. The paper concludes that most of these cold mix recycled roads have shown good performance. They are also economical and provide environmental benefits. It is recommended that more detailed evaluation of the milled material should be performed.


Development of Asphalt Overlay Performance Models From the C-LTPP Experiment
S. Tighe, R. Haas and N. Li

The Canadian Long Term Pavement Performance (C-LTPP) study, initiated in 1989, involves 65 sections in the 24 provincial sites that received rehabilitation comprising various thicknesses of asphalt overlays. This paper describes the impacts of the various alternative rehabilitation treatments on pavement performance in terms of roughness progression under comparative traffic loading, climate, and subgrade soil conditions. Factor effects, including climatic zone, subgrade type and traffic level were also evaluated. Some findings are that: (a) in wet, high freeze zones, thinner overlays show a higher rate of roughness progression than thicker overlays, regardless of subgrade type; (b) in dry, high freeze zones, roughness progression for medium and thick overlays is relatively small; (c) in wet, low-freeze zones, thinner overlays combined with a fine subgrade show the highest rate of roughness progression. The methodology developed in this study for pavement roughness evaluation can be applied to performance trends analysis of other LTPP data.


Pavement Evaluation and Strengthening Design: Sixteen Years Experience
B.A. Hakim, S.F. Brown and R.J. Armitage

Practical implementation of analytical pavement evaluation and strengthening design using the Falling Weight Deflectometer (FWD) and associated techniques has resulted in many developments in the past sixteen years and represents an excellent example of "paving the gap" between research and practice. This paper presents case studies to illustrate the techniques in various applications. Use of the Nottingham Asphalt Tester (NAT) to assess the mechanical properties of bituminous layers from tests on cores is commonly carried out to support FWD findings and for use in pavement strengthening design. Non-linearity of subgrade material crucially affects FWD measured deflections and, hence, the back analysed results. Therefore, determination of subgrade stress levels and the consequent stiffness variations with depth result in more realistic designs. Other case studies are described to illustrate:

– Whole life cost and pavement design optimisation for clients such as Design Build Finance and Operate (DBFO) consortia, using mechanistic techniques.

– Practical implementation of end-product performance testing such as FWD, Dynamic Cone Penetrometer (DCP) and Dynamic Plate Bearing Tester (DPBT) on foundations, and NAT on bituminous samples, to support applications of mechanistic design techniques.

– Use of the FWD to improve the design of rehabilitated concrete pavements using the ‘crack and seat’ technique with bituminous overlay.

– Use of a new analysis technique for prediction of bond stiffness between bituminous layers from FWD data, to assist in detailed structural evaluation.


Implementation of Falling Weight Deflectometer Technology and Development of Analytical Pavement Design in Thailand
S. Lekso, T. Ruenkrairergsa and C. Phromsorn

Thailand has experienced an exceptionally rapid economic development over more than 10 years, leading to major infrastructure problems and costly rehabilitation of the road network at short intervals. The Thai Department of Highways aims to reduce the maintenance costs through more appropriate pavement evaluation and design. A contract agreement for the "Implementation of Falling Weight Deflectometer Technology and Development of Analytical Pavement Design Project" was signed between the Department of Highways, Thailand and the Danish Road Directorate in October 1998. Funding for the project was provided through a Danish Mixed Credits Loan for the total costs of the project. The duration of the project was three years ending in December 2001. The project has included intensive testing of 130 road sections by the Falling Weight Deflectometer and by Benkelman Beams. In addition, the pavement structures were checked and samples brought to the laboratory for further tests. The activities were summarised in ten research tasks. The present paper describes the activities used in the successful implementation of the new technology in the highway organisation and examples of the results. The technical content of the project is published in fourteen Technical Reports and Manuals available at the Department of Highways, Thailand.


In-Situ Assessment of Stiffness Modulus for Highway Foundations During Construction
P.R. Fleming, J.P. Lambert, M.W. Frost and C.D.F. Rogers

Several portable field devices that measure stiffness modulus are reviewed in detail in this paper including the German Dynamic Plate Test (also known as the Lightweight Drop Tester), the TRL Foundation Tester (UK), the Prima (Denmark) and the Humboldt Soil Stiffness Gauge (USA, also known as the GeoGauge). Laboratory and field data are presented which explain the many important influences on the measured data and demonstrate comparative performance with respect to the Falling Weight Deflectometer. These field data show significant scatter and site specific correlation. A strategy for compliance testing during construction, as part of a performancebased specification approach for the UK, is suggested. Conclusions are made regarding the devices’ relative merits and limitations, and considerations for their introduction into contractual use for routine assessment during construction.


Back-Calculation of Pavement Layer Moduli and Forward-Calculation of Stresses and Strains
W. Zhang and P. Ullidtz

The "analytical-empirical" (or "mechanistic-empirical") method is widely used in pavement design or evaluation. The method has three steps:
1. Determine the moduli of the pavement layers
2. Calculate the critical stresses or strains under the design load(s)
3. Compare critical to permissible values (or use stresses or strains in a deterioration model)

If the pavement layer moduli are determined from an inverse analysis (backcalculation) of Falling Weight Deflectometer (FWD) data, then the first two steps in this procedure constitute the "analytical" part. This paper deals with these two steps in relation to deflections, stresses and strains measured under a FWD. In step 1 and 2 above, the pavement "response" is calculated using a theoretical model, in terms of deflections in step 1 and of stresses and strains in step 2. All theoretical models involve simplifications with respect to reality and even the most "sophisticated" theoretical model must be verified against real pavement structures. To verify a theoretical model, the stresses or strains calculated in step 2 may be compared to measured stresses or strains. During the past ten years this has been done on three instrumented pavements in the Danish Road Testing Machine and on three in situ pavements in Sweden. Strains at the bottom of the asphalt layer and stresses and strains at the top of the subgrade (which are frequently used in design methods) were measured and compared to theoretical values. Three different theoretical models were used to first backcalculate the layer moduli and then forward calculate the stresses and strains. The models were: Layered Elastic Theory with linear elastic materials, the Method of Equivalent Thicknesses with a nonlinear subgrade and the Finite Element Method where any layer may be non-linear elastic. For the strains at the bottom of the asphalt, reasonably good agreement was mostly found with all theoretical models, although some comparisons gave very large differences. For the strains at the top of the subgrade, the assumption of linear elastic materials gave very poor agreement. A reasonably good agreement could be obtained with a non-linear elastic subgrade. In most cases the best agreement (and the most reasonable values for layer moduli) was found using the Method of Equivalent Thicknesses.


Asphalt Pavement Performance Research Project in Lithuania
A. Braga, V. Puodziukas

Currently there are three PMS systems used in Lithuania- HDM-III, HDM-4 and DAVASEMA (Lithuanian PMS). HDM pavement performance models are used in all of them. With the purpose of calibration and adaptation of those models in 1997 Lithuanian Pavement Performance Research Project was executed. The research data gathered in four years of the Project gives opportunity to do some conclusions on asphalt pavement performance in Lithuania. The first stage attempts of the adaptation of input data and calibration of HDM models to local conditions are described in this article.


Modelling Flexible Pavement Performance Using Canadian Historic Data
T. Kazmierowski and L. Ningyuan

The Ministry of Transportation of Ontario (MTO) has recently completed the development of a Second- Generation Pavement Management System (PMS2). One of the advanced features built in the system is a number of performance prediction models used for analysing short-medium term pavement needs at the network level for maintenance and rehabilitation treatments planning purposes. Data samples recording more than 25 years of field pavement performance history were used in modelling deterioration characteristics of flexible pavements for each functional class of road. These models range from simple polynomial curves to sophisticated sigmoidal performance models. Substantial efforts were put into this phase of model analyses, including data source and pre- process, model analysis, and using engineering judgement to assess the impacts of various pavement rehabilitation treatments on overall pavement performance. Some major findings and recommendations from this study may be useful for highway engineers and practitioners in the areas of pavement life-cycle analysis, multi-year rehabilitation programming and economic analysis of thin asphalt pavements versus thick asphalt pavements.


Implementation of Ground Penetrating Radar Technology in Asphalt Pavement Testing
T. Scullion and T. Saarenketo

Over the past decade Ground Penetrating Radar (GPR) applications have been under development within both the Texas Department of Transportation (TxDOT) and the Finnish National Roads Administration (FNRA). In the mid 1990’s GPR systems moved into mainstream usage, improved software was developed and training schools were conducted. TxDOT now has 4 complete air launched GPR systems for pavement evaluation. GPR technology is now used on a routine basis as the first step in planning the rehabilitation of flexible pavements. In Finland GPR has been implemented in the area of quality assurance testing of new asphalt overlays. Using a limited number of calibration cores the GPR data are used to compute the inplace air void content profile of new overlays. In the 2000 construction season a penalty system was implemented based on these GPR results, and the feedback from both FNRA and the contracting community has been positive. This paper will provide an overview of these development and implementation efforts. The basics of GPR will be presented together with a discussion of the methods used to convert GPR waveforms into information useful to pavement engineers. A number of successful case studies will be described. GPR was initially promoted as a rapid layer thickness measuring tool. It certainly provides useful thickness information for the upper layers of flexible pavements. However as will be described in the case studies GPR can provide substantially more information than simply layer thickness.


Step-Frequency Radar Applied on Thin Pavements
X. Derobert, J-M. Simonin, L. Laguerre and C. Pichot

In the field of road construction and maintenance, the need for information on the thickness of very thin road layers is not being satisfied by means of commercial impulse ground penetrating radar (GPR), due to the inability of such devices to operate over ranges of several gigahertz. As a result, research has focused on the design of a step-frequency radar technique, capable of working with very high-frequency synthetic pulses. The principle of the step-frequency technique is presented herein. An ultrawide band antenna (within the family of Vivaldi antennas) has been developed for road applications; it was created using "stripline" technology and yields a bandwidth greater than one decade, over a severalgigahertz range. GPR dynamic measurements were obtained from selected road construction and maintenance test sites (e.g. the Circular Pavement Fatigue Test Track, composed of a number of known structures). Results have shown improved resolution in comparison with a commercial impulse GPR system. Indeed, thin asphalt surfacing thicknesses can be measured from raw stepfrequency radar data in instances when very thin asphalt surfacing requires signal processing for thickness measurements.


Performance-Based Pay Factors for Asphalt-Concrete Construction Reflecting Fatigue and Rutting Effects
J.A. Deacon, C.L. Monismith, J.T. Harvey and L. Popescu

This paper describes a rational and feasible method to quantitatively establish pay factors for new asphalt concrete pavements. The approach uses performance models for fatigue and rutting based on the analysis of accelerated pavement tests from the Caltrans Heavy Vehicle Simulator (HVS) and the Federal Highway Administration Project WesTrack. For rutting, the influence of asphalt content, air-void content, and aggregate gradation are considered. For fatigue, air-void content, asphalt content, and asphalt concrete thickness are included. Costs are established using a cost model considering only agency cost consequences of delaying or accelerating the time to the next rehabilitation. For the as-constructed mix the relative performance RP (ratio of off-target ESALs to target ESALs) is determined for both fatigue and rutting. The shortest RP for the combined RP’s for mix and pavement characteristics considered for a specific distress mode permits determination of the pay factor from the cost model.


The Benefits of User-Friendly Suspensions
A.C. Collop and D. Cebon

This paper examines the effects of ‘road friendly’ heavy goods vehicle suspensions on long-term flexible pavement performance. A deterministic Long Term Pavement Performance Model (LTPPM) is used to calculate pavement damage due to realistic traffic and environmental loading. The traffic is modelled first as a fleet of steel sprung heavy goods vehicles and second as a fleet of ‘road friendly’ air suspended vehicles. Two generic types of flexible pavement construction are considered representing a major road and a minor road. Pavement life predictions are compared for the two cases and with results from a simple road damage analysis based on the ‘fourth power law’. It is concluded that changing to a fleet of ‘road friendly’ vehicles would not significantly affect the life or maintenance costs of thicker asphalt pavements (motorways and trunk roads) whereas the life of thinner pavements (minor roads) would be increased significantly if the vehicle fleet changed to road friendly suspensions.


The Combined Effects of Tire Contact Stresses and Environment on Surface Rutting and Cracking Performance
L.A. Myers, C. Drakos and R. Roque

This paper will describe how the combined effects of tire contact stresses and environmental conditions may control the surface cracking and rutting performance of pavements. Measured tire contact stress data will be used in conjunction with a range of environmental conditions to evaluate their effects on near-surface stresses that are conducive to rutting and cracking. Findings to date have indicated that temperature gradients during cooling periods are most conducive to surface cracking. The work has also shown that an overall shear stability evaluation during critical high temperature periods may need to be conducted to better assess the near-surface rutting potential of a particular mixture in given environment.


Asphalt Pavement Deterioration Models for Mild Climatic Conditions
A. Loizos, J. Roberts and S. Crank

There is currently significant international activity in model development, including the EU PARIS and RIMES projects, and the models within HDM-4. However, due to climatic and material variations across Europe, more models are needed. The Paper describes the ongoing adaptation of a suite of innovative performance models, based on HDM-4, for asphalt pavements for use in mild (Mediterranean) climates and operating conditions. The models have significant changes and additions to improve their realism, consistency, explanatory power and sensitivity, and are being implemented within a Road Infrastructure Management System (RIMS) being developed for the Greek Government. It is hoped that the work will be a positive contribution to HDM-4 development. The Paper provides an overview of the requirements of a satisfactory modelling system, a description of the modelling of structural degradation, and aspects of the models for cracking, rutting and roughness.The Paper compares the models with the official HDM-4 models in terms of a study for crack modelling using Greek SHRP LTPP data.


Performance Models for Deep in Situ Recycled, Bitumen Stabilised Pavements Under Accelerated Traffic
F. Long, H.L. Theyse, S. Robroch and J. Liebenberg

This paper describes research to assess the use of foamed bitumen and bituminous emulsion treated materials with deep in situ recycling (DISR) technology. The paper discusses Heavy Vehicle Simulator (HVS) and laboratory test results on these materials. The results showed that the treated materials have higher resistance to permanent deformation than fatigue. The HVS sections failed after the addition of water, exhibiting erosion and pumping. The laboratory test results show that treating the material with bituminous binders and cement increases the permanent deformation resistance, and if enough binder is added, the flexibility of the material improves. The HVS data are used to develop preliminary structural design models for effective fatigue and permanent deformation, and to determine damage factors. The materials are more load sensitive in fatigue than permanent deformation.


Asphalt Concrete Response: Experimental Determination and Finite Element Implementation
S.M.J.G. Erkens, X. Liu, A. Scarpas and A.A.A.Molenaar, J. Blaauwendraad

At the previous conference the beginning of an extensive experimental and analytical investigation into the mechanisms leading to the initiation and propagation of damage in asphalt concrete pavements was reported. The objectives of this Asphalt Concrete Response (ACRe) project were twofold, firstly a 3-dimensional, strain rate sensitive, temperature and loading history dependent constitutive model would be formulated and implemented in the finite element package CAPA-3D. Secondly, the necessary experimental set-ups, testing procedures and dataanalysis methods for determination of the model parameters would be developed. At the time the prototype formulation of the model was reported, along with preliminary test results and the way in which the model parameters were determined on the basis of the experimental results. A simulation of the dynamic non-linear response of a pavement was included to demonstrate the possibilities of the approach. In the past years, the project progressed rapidly, with continuous interaction between the two objectives: on the one hand the model dictated what should be measured in a test, while on the other hand, the response observed in the tests set the requirements for the model. This led to regular adaptations in the numerical formulation as well as the development of highly accurate test set-ups for parameter determination. Throughout the project, test results were compared to model predictions to verify applicability. In this contribution both, the numerical and the experimental aspects of the project will be presented in detail.


ILLI-PAVE Based Flexible Pavement Design Concepts for Multiple Wheel-Heavy Gear Load Aircraft
M. Thompson, F. Gomez-Ramirez and M. Bejarano

ILLI-PAVE, a structural axi-symmetric non-linear finite-element pavement analysis software, has been enhanced to characterize flexible pavement response and performance under Multiple Wheel-Heavy Gear Load aircraft using Mechanistic-Empirical concepts. Several analysis techniques using the Principle of Superposition (PS) are applied to account for response interaction (stresses, strains and deflections) from multiple wheel loads. It has been shown that, for engineering purposes, the PS is valid and can be used to determine load-induced stresses and deflections of MW assemblies provided the SW responses are accurately computed or measured. To characterize Flexible Pavement performance using Mechanistic-Empirical concepts, fatigue-cracking and subgrade permanent deformation models have been adopted. The horizontal strain at the bottom of the asphalt layer is used to assess fatigue-cracking damage. For subgrade rutting, a Subgrade Stress Ratio criterion (SSR = applied stress ¡A soil strength) is used.


New Evaluation of Pavement Performance Through Mechanical Fatigue Testing
P. Des Croix

Mix mechanical fatigue testing is almost thirty years old in our laboratory. The classical fatigue approach is to record the number of cycles for failure versus initial stress or strain. The standard fatigue law has the following form: delta = A Nsuper-B, with delta = (strain) or (stress). A new approach based on damage theory approach is presented in this paper leading to a new law: fatigue damage rate asubF versus strain amplitude level eta. Typical results are shown with a standard mix formulation with pure and modified bitumen. With this new approach, the procedure appears as independent of the type of loading (as far as an "homogeneous" testing procedure is considered). The testing time (as it is no more necessary to go up to the failure of the sample) can also be reduced.


Crack Growth Behavior of Asphalt Mixtures and Its Relation to Laboratory and Field Performance
R. Roque, B. Birgisson, B. Sangpetgnam and Z. Zhang

Laboratory and field studies have resulted in a more complete understanding of the cracking behavior of asphalt mixtures and of the cracking mechanisms associated with asphalt pavements. This work has also led to the development of a viscoelastic fracture mechanics-based crack growth law that is capable of fully describing both initiation and propagation of cracks in asphalt mixtures for any combination of loading and temperature conditions. The model requires the determination of only four fundamental mixture parameters that can be obtained from less than one hour of testing using the SuperPaveTM Indirect Tension Test (IDT). These parameters can account for micro-damage, crack propagation, and healing for stated loading conditions, temperatures, and rest periods. This paper summarizes the developments leading to the model, and descri-bes the model and its potential use in the areas of mixture design and optimization, performance-based specifications, and pavement design.


RILEM – Interlaboratory Tests on Performance Prediction of Pavements
M.N. Partl and H. Piber

First conclusive results of an ongoing international RILEM test programme are presented where performance predictions of different laboratories based on their own test methods and models are compared. Slabs and technical data of two motorway test sections were provided to 16 laboratories participating on a voluntary basis. This paper focuses on rutting and fatigue predictions for a period of 10 years. It was found that the predicted rut depths varied over a wide range. Compared to rutting, the fatigue predictions were in better agreement. However, the conclusion that fatigue models are more accurate could not be drawn. For both rutting and fatigue prediction, none of the laboratories followed a procedure and methodology which was directly comparable. This made clear that further exchange and co-ordination of research efforts is extremely necessary.


Evaluation of Pavement Performance Performance of Bituminous Roadbase Layers
S.F. Said, R. Fredriksson, H. Carlsson and H. Jansson

Four different bituminous roadbase mixes were used in a new segment of a highway. The bituminous mixes were modified stone mastic asphalt called Viacobase D, which is used as a large aggregate roadbase mix, grouted macadam, which is coated with open graded asphalt concrete called Runbase, conventional bituminous gravel mix, type AG25/B180, which was used in the reference sections, and a modified AG25/B180, which has a high content of stone particles. Pavement performance was studied 1995 – 2001 through measurements of deflection, rut depth and unevenness and manual distress identification, as well as traffic counts. A number of cores were taken for studying performance-related (functional/mechanical) properties of mixes in the laboratory. The evaluation is based on pavement performance. Furthermore, similar structures were tested in a Circular Test Track. In addition to evaluation of the roadbase layers, the purpose was to validate laboratory and field methodologies for predicting distress in flexible pavements, in particular roadbase layers.


Study on the Performance-Based Design Method of Flexible Pavement
S. Lijun, Z. Hongchao, L. Fangyan, L. Liping and L. Hui

Along with the increasing traffic volume, vehicle’s speed and axle load, serious initial failures occurred on the heavy-duty asphalt pavements in recent years. The rapid deterioration attracted more and more attention on the conventional road technique. To improve the pavement design methods, field investigation and mechanistic analysis have been conducted, and six new types of distress were addressed herein. The research indicates that the conventional pavement technique only be suitable for the ordinery pavement design but not the heavy-duty pavements. On the freeway, the dynamic hydraulic pressure under the moving-vehicles’ tires became an unneglectable factor causing the initial failures. The non-uniformly distributed load is another important factor which will cause the local damage, and the other important factor is the overload. According to the pavement structure behavior equation covering the whole life cycle, which is eatablished in this paper, a new bitumen pavement design method based on performance was developed.


Highway Pavement Performance Models
L. Gaspar

Since 1991, 62 Hungarian pavement sections have been yearly monitored to develop pavement performance models for various combinations of traffic-pavement type-subsoil strength. The condition evaluation technique and its application areas are presented. The investigation of resurfaced and redressed trial sections allowed to evaluate the actual condition improving effect and the cost-effectiveness of various intervention types. The time-series condition data of the 5-year monitoring of several main and secondary road sections of altogether 3000 km length will be used to validate the performance models based on the measuring results of 62 trial sections.


Bitumen Rubber Asphalt: Report on the Long Term Performance in South Africa
C.J. Potgieter, D.E. Sadler and E.M. de Villiers

The three authors were involved with bitumen-rubber (B-R) binders and asphalts since its introduction into the Republic of South Africa (RSA) in the early 1980’s. Recently they revisited most of these old projects to evaluate the performances of (B-R) asphalts over the last 20 years. This paper contains the result of this survey, it also details the B-R asphalt and binder properties at the time of manufacture. The outcome culminated in the present RSA specification for B-R binders and asphalts. Their specifications are also furnished in this paper. B-R products are normally at a disadvantage in that they are placed where the conventional asphalts have already failed and on top of this must now prevent the old asphalt’s deficiencies from reflecting through the new surfacing again. B-R binder and asphalt performed this role admirably.

The crack dampening and waterproofing provided by the bitumen rubber stress absorbing membrane interlayers (SAMI’s) and B-R asphalts, as well as the enhancement of structural capacity provided by the latter, were found to be far superior when compared with conventional binders and asphalts under similar conditions. For heavily trafficked and distressed roads B-R asphalts and SAMI provide the more economic overlay in the RSA.


Engineering Properties of Asphalt Concrete Mixtures Utilised in Thailand
T. Ruenkrairergsa, S. Lekso and P. Silarom

The Department of Highways, Thailand, is considering improving its method of evaluating and designing flexible pavements from the conventional empirical method to the semianalytical method. Prior to doing so, the engineering properties of pavement materials of each layer need to be known. Since the engineering properties of Asphalt Concrete (AC) mixtures are the parameters of most concern in the pavement design process, these need to be considered in detail. Therefore, a research study into the engineering properties of AC mixtures used in Thailand was initiated. In this study, more than 300 Asphalt Concrete core samples were taken from more than 70 selected test sections around Thailand. The cores were tested by various laboratory-testing methods, such as the Indirect Tensile Strength test, the Resilient Modulus test and the Fatigue test to obtain the engineering properties of Asphalt Concrete in Thailand. The results indicate no major difference between Asphalt Concrete Mixtures used in different regions of Thailand. The average value of the Indirect Tensile Strength was found to be 1.08 MPa at 25°C. The Resilient Modulus was found to be quite high, the average values were found to be 2,468 / 4,410 / 7,794 / 13,183 MPa at 45 / 35 / 25 / 15°C, respectively, corresponding to nearly a twofold increase of the Resilient Moduli for every 10°C decrease in the testing temperatures. In addition to the laboratory tests, non-destructive testing of the test sections were performed using the Falling Weight Deflectometer (FWD) to obtain the backcalculated AC moduli. Subsequently, a comparison between the AC Resilient Moduli determined from laboratory testing and backcalculated AC moduli from FWD testing of selected road test sections was made. The AC Resilient Moduli determined in the laboratory are found to be higher than the AC moduli obtained from backcalculation by a factor of between 1.45 and 1.91. The differences were found to decrease with increasing temperature.


Using Field Measured Stresses and Strains to Quantify Flexible Pavement Responses to Loading
A. Loulizi, I.L. Al-Qadi, G.W. Flintsch and T.E. Freeman

One of the 12 instrumented sections of the Virginia Smart Road was used to evaluate the effect of several loading parameters on the measured stresses at different layers and the measured horizontal transversal strain in the bottom of the hot-mix asphalt (HMA) layer. It was found that speed does not affect the measured vertical compressive stress in all layers, but does significantly affect the measured horizontal transversal strain under the HMA layer (190.5mm below the pavement surface). Strain was found to decrease by a factor of 2.7 when the speed increases from 8km/h to 72.4km/h at 25°C. Variation in tire inflation pressure from 551.6kPa to 724kPa was found not to affect the measured vertical compressive stress in all the layers and the measured horizontal transversal strain in the bottom of the HMA layer. Vertical stresses under the HMA layer were found to vary linearly with the applied tires’ load at 25°C.


Estimation of the Plastic Strain in the Pavement Subgrade and the Pavement Functional Condition
W. Zhang and P. Ullidtz

At the present time, most subgrade failure criteria use the permissible stress or strain at the top of the subgrade. The permissible stress or strain is related to the volume of traffic, and is mostly independent of soil type, moisture content and mechanical properties of the soil. Over the past six years, three test pavements in the Danish Road Testing Machine have been constructed and instrumented. Accelerated loading tests and freeze/thaw experiments were conducted. Based on the measured data, models to estimate the plastic strain in the subgrade and the pavement functional condition were established. In the models it is assumed that the plastic strain in the subgrade and the pavement functional condition are related to the number of load repetitions, resilient strain and stress. The parameters used in the models were obtained from the measured data using an incremental-recursive procedure. For both the measured plastic strain in the subgrade and the pavement profiles, reasonable predictions were produced with the models.


Instrumented Field Test on Light Expanded Clay Aggregate Used in Pavements
I. Hoff, A. Watn, E. Oiseth and K.O. Amundsgard

Damages due to frost penetration into frost susceptible sub soil are a severe problem in the Nordic countries. Traditionally road construction materials like gravel and crushed rock have very low insulation capacity and quite thick layers are needed to avoid frost penetration. Light weight expanded clay aggregate (LWA) is a promising alternative because of its good insulation properties. To investigate the mechanical and thermal properties of the LWA-material in pavement structures a research project was initiated by a producer. The material has been studied in the laboratory by several tests including repeated load triaxial testing. A full scale laboratory test was also performed with a real pavement structure built in the laboratory and subjected to cyclic loading. These tests showed that the LWA-material behaves almost as traditional gravel or crushed rock material, provided that the stresses are small enough to avoid crushing of the grains. To validate the findings from the laboratory, a full scale test field was built just outside Trondheim. The test was constructed with 10 cm of asphalt and 30 cm crushed rock over 40 cm of LWA. The test field was instrumented with sensors for measuring horizontal and vertical stresses and deformations in the LWA-layer. In addition temperature sensors were installed in all layers of the structure. The test field was trafficked by heavy vehicles for two years, and in addition controlled loading with trucks, FWD and plate load tests was performed. In addition to the instrumented section a wedge-out section with gradually less material covering the LWA-material was constructed. At the weak end of this section only 4 cm of asphalt and 16 cm of crushed rock covered the LWA-material. Some rutting was observed initially after the field was constructed. This is believed to be caused by the use improper compaction equipment. The development of rutting stabilised after a few months except for the weakest part of the wedge-out section where a total of 80 mm of rutting accumulated after two years of service. The test field was excavated in August 2001 and the LWA-material was examined to check for crushing of grains. Some crushing was observed, but the crushing was not severe and is not considered to be a major problem.


The Design and Performance of Thin Surfacing Layers
J. Carswell

Thin surfacing layers, originally developed in the 1980s, have grown dramatically in their application on the strategic highway networks. This paper reviews the design considerations of a range of thin surfacing systems within Europe and elsewhere and outlines the advantages over more traditional surface treatments. The scope for innovation is high and both wide grading ranges and different layer thicknesses have been successfully used. In particular, the role of the binder in thin layer applications is considered in some detail. The performance of a number of road contracts carried out under a range of climatic conditions is reported, including performance data from Sweden, United Kingdom and the United States. The road sites are still under assessment and all of them are performing satisfactorily. Given the wide application range for thin layer technologies, the paper also discusses the possible failure criteria for the various thin surfacing systems.


Skid Resistance Evolution of Different Wearing Course Techniques According to Traffic
M. Gothie

This article presents how the skid resistance measurements carried out in France with the ADHERA device are grouped together in a computerized file named CARAT. This file allows specific studies to be carried out that show the relationship between skid resistance and various elements such as accumulated heavy goods vehicles traffic (AHGVT), the techniques used for wearing courses and their main characteristics. The AHGVT is divided into several different classes, according to quantity , and for 11 different techniques this article presents the evolution of skid resistance values under the effect of traffic and measured by the ADHERA device. The analysis that was carried out highlighted several wearing course techniques presenting and maintaining good skid resistance characteristics when exposed to traffic.


Relating Road Roughness to Human Discomfort and Health Impact
C. A. Lenngren and J. Granlund

The International Roughness Index is the choice parameter of most road engineers for describing pavement surface roughness. It has been proven to satisfactorily explain phenomena such as pavement performance and pavement deterioration. However, vertical motion frequencies and amplitudes affecting humans the most are different from those used in the model. Humans are also affected by side motion, which is not accounted for at all. At a road test with instrumented vehicles it was found that the index did not indicate rough road sections being the most adverse for humans. From a previous study of performance based road maintenance it was also found that IRI was less suitable for this type of roads. A laser profiler commonly used for road surface surveying used for assessing the IRI was used for trying othe r measures of the roughness. Raw data were stored so that other models could be tried including horizontal acceleration. A comparison was then made to the instrumented vehicle test data. A discussion follows about vehicle simulation model versus generic measures for pavement management use. Finally, some suggestions are made for a ride quality index adjusted to human aspects.


Developments and Experience with the Testing of the Acoustic Performance of Asphalt Surfaces
S. Phillips

Type approval for proprietary thin asphaltic surfacings has recently been introduced in the UK through the Highway Authorities Products Approval Scheme (HAPAS). This scheme includes a noise test that gives a noise classification for comparing different surfacings or even for the prediction of traffic noise. However, the Statistical Pass-by test used has only limited application when used alone and therefore additional test procedures were needed. One of these is the Close-proximity (CPX) method test that measures the noise close to dedicated test tyres. The CPX method developed by the Transport Research Laboratory has the ability to enable the homogeneity of the noise of road surfacings to be determined arbitrarily along their length. When combined with other systems being developed for assessing acoustic properties indirectly, the current methods give a fuller description of road surface performance.


Noise Reduction vs Wearing Properties
J. Valtonen, I. Hyyppa and P. Sainio

One of the most promising way to protect citizens from noise is to lay low-noise asphalt on urban and sub-urban roads. Two test sections were laid in October 1999, but their wearing properties have not been satisfying. During the winter 1999-2000 the wearing properties of 17 different mixes were examined in the laboratory with PWR-equipment. The main result was, that SMA 5 was wearing more than twice as much as SMA 16. In the summer 2000 the low-noise SMA 5 was laid on four different roads near Helsinki. The ruts were measured in the spring 2001 to find out how much SMA 5 has worn compared to coarser pavements. These results show, that the ruts of SMA 5 were six times as deep as the ruts of SMA 11 and ten times as deep as the ruts of SMA 16. Noise measurements were done before and after the paving in the neighbourhood of the roads. When the speed limit was 50 km/h, the noise reduction was only 3 dB(A) both 10 and 30 m from the road, but when the speed limit was 80 km/h, the noise reduction was even 7 dB(A) even though only two of four lanes were paved with SMA 5. The opposite side of the road still had the old SMA 18-pavement. Also tyre-noise measurements have been carried out and one of the main result was, that after one winter the noise from a SMA 5 has increased significantly because of studded tyres.


Two-Layer Porous Pavements and Noise Reductions in Denmark
H. Bendtsen and L.E. Larsen

Noise from the tire/road interaction is one of the two dominant sources to road traffic noise. The other source is engine noise. A research project was started in Denmark in 1999, with the purpose of developing noise reducing road surfaces for urban roads with speeds around 50 km/h. 3 test sections with two-layer fine grade porous drainage pavements and a reference section have been constructed on an urban road with 7000 vehicles per day. The pavements are cleaned twice a year by high-pressure water spraying and sucking. Some hypotheses on noise reduction, durability and traffic safety have been defined. A multi-disciplinary research group has established a comprehensive measurement program to test these hypotheses. This program includes acoustics (pass-by noise, road surface sound absorption, and noise inside vehicles), permeability, surface texture using laser, study of plane and thin sections from drill cores, friction, traffic safety, social surveys of annoyance and economic evaluation. The measurements are repeated every year, the goal is to continue in the whole lifetime of the pavements. The latest results on noise reduction, porosity and surface characteristics after 3 years of measurements are presented and discussed.


Development of Low-Noise Pavement Function-Recovery-Machine
T. Abe and Y. Kishi

Research into ways to recover the higher drainage and noise- lowering function of porous low-noise asphalt was carried out. Tests in the laboratory indicated that the use of cavitation is very effective for this purpose. A recovery machine based on this was developed. The machine’s water ejection system includes the utility of cavitation, induced by submerged high-speed water jets, held in place using a low pressure chamber, along with water jets in air. Field tests have been carried out on normal roads and on expressways and the machine has been confirmed to be very effective. This paper introduces the research into function-recovery technology of low-noise pavement and the machine consequently developed.


Asset Valuation Methodologies and Performance Measurement in Life-Cycle Analysis
L.C. Falls and R.Haas

Pavement management systems, and now more broadly asset management systems, have been accepted and implemented by many agencies worldwide. Life cycle analysis is a key component of these systems at both the project and network level. However, the incorporation of asset value in life cycle analysis has received little attention. Rather, current and future costs are the prime elements. The time has come though where owners or operators of the assets, the latter particularly in the case of privatization, are starting to require the explicit incorporation of asset value in the life cycle analysis. In other words, the issue is what was the asset worth when built, today, and what is it estimated to be in future years under various alternative strategies and funding scenarios. This paper is based on a highway asset valuation and performance indicators study carried out for the Transportation Association of Canada. It describes the role of asset valuation in asset management, the available methodologies and their applicability and the direct incorporation and reporting of asset value in the life cycle analysis. As well, the paper describes the associated performance indicators related to the general, macro level, service quality to users, functional effectiveness and preservation effectiveness. Finally, the paper identifies the major issues and requirements involved in the proper application or use of asset valuation in life cycle analysis.


Preventive Maintenance Versus Reconstruction: Life Cycle Cost Analysis of Various Options
T.V. Scholz, R.G. Hicks and J. Moulthrop

Life cycle cost analysis is increasingly being recognized by public agencies as an effective tool to assist in the selection of construction, maintenance, and rehabilitation treatments. The Federal Highway Administration has developed a life cycle cost analysis methodology that will likely become the standard in the industry in the USA. The methodology can be used to evaluate the life cycle costs strategies including preventative maintenance, rehabilitation, and reconstruction. For preventative maintenance treatments to be more widely accepted, they must not only extend the life of the existing pavement, but also be shown to be cost effective, i.e., lower life cycle cost than the alternates.

This paper presents:
1) an identification of preventative maintenance treatments and their benefits;
2) a description of the life cycle cost analysis process utilized in this study; and
3) comparative results that evaluate the life cycle cost for pavements that receive preventative maintenance treatments applied early on in the life of the pavement compared with those that receive only major rehabilitation/reconstruction strategies.

The findings indicate that preventative maintenance at early stages in a pavements life is cost effective in all of the scenarios studied. However, the reader should be aware that the estimated lives and costs used in this study are based on presently available information and engineering judgment. Deviations from these values could affect the final conclusions.


Life Cycle Cost Analysis Gives the Possibility to Pay for the Value of the Pavement
A. Huvstig

Performance requirements on road surfaces are sometimes being used as an alternative to materials specifications for achieving pavement quality. Pavement performance affects user costs, safety costs, environmental costs, etc. Life cycle analyses and new deterioration models make it possible to put a value on the performance of a pavement both today and in the future, which in turn means being able to set a monetary value on the performance of a pavement handed over by a contractor.This paper describes a method for assigning a value to specific surface characteristics. This enables the client to pay for a performance characteristic rather than contractor costs.Life cycle cost analysis (LCC/WLC) and PMS models are based on specific economic models used to calculate a net present value (NPV). A normally used simplification is the net present cost model (NPC). There are some major flaws in these simplifications, which cause unnecessary costs for the maintenance. This paper also describes the background of the economic models, and recommends a better way to use economic models as basic input in LCC/WLC models and PMS models.


Environmental Improvement of Binder Content and Aggregate Gradation Determination
M. Jegsen, A. Kargo and E. Nielsen

One of the threats to both working and external environment in analysing asphalt material has been the use of hazardous solvents for determining the aggregate gradation, the binder content and the recovery of binder. The objective of this paper is to provide the technical basis for a large reduction of solvents used through introducing a binder content by ignition method, AM-1. The paper verifies and documents through round robin testing that quality control of binder content and aggregate gradation with this technique has the same (or perhaps slightly improved) precision compared to the "traditional" solvent based methods for the benefit of the environment. It is estimated that this new method fully implemented will reduce the amount of solvents used in asphalt laboratories in Denmark by 90%.


Characterization of Asphalt Odors and Emissions
C. Lange and M. Stroup-Gardiner

A total of nine asphalts were selected from five regions of the United States. These asphalts were evaluated using a gas chromatograph; compounds were separated into six odor volatile organic compound (VOC) groups. Changes in these odor VOC groups were evaluated after the addition of one of two commercially available odor reducers used at one of two concentrations (1:5,000 and 1:12,000). Results indicate oxygen reactions are a major contributor to the formation of odor VOCs and that odor reducers appear to act as anti-oxidants. Mass loss, measured every minute over a 90-minute period can be related to the formation of odor VOCs. This frequent measurement of mass loss is better correlated with odor VOCs than the industry standard rolling thin film oven. Concurrent opacity measurements are also correlated with odor VOCs. The addition of odor reducers reduces both mass loss and opacity.


Noise Reduction with Porous Asphalt Costs and Perceived Effect
L.E. Larsen and H. Bendtsen

Three test sections of two-layer porous asphalt and a reference section of dense asphalt concrete were laid on Oster Sogade in Copenhagen in 1999. An extensive measurement program including noise, surface characteristics and traffic safety is part of the project, which also includes a before-after questionnaire survey on noise annoyance among those living along the road. Two-layer porous asphalt pavements are more expensive than ordinary dense pavements, but comparisons done as part of the project of the cost of porous asphalt, noise barriers and sound insulation shows the asphalt to be economically attractive where there is a need for reduction of the traffic noise. The questionnaire survey shows that the reduction in decibel is matched by a reduction in annoyance both in and outside the homes of those living along the road.


Porous Asphalt and Traffic Safety
P. Greibe

The behaviour of porous asphalt is different from that of dense asphalt concrete. The open structure of porous asphalt reduces traffic noise, drains water from the road surface and reduces the thermal conductivity. These differences can have an effect on road safety. This article presents a literature review on porous asphalt and safety and furthermore summarises the preliminary results from a Danish pilot study using porous asphalt on an urban road.


Austrian Experiences with Winter Maintenance on Porous Asphalt
J. Litzka

The paper deals with the specific requirements for winter maintenance on sections with porous asphalt and the experiences in Austria with some additional informations from neighbouring countries. After extensive use of porous asphalt on motorways in Austria until the early nineties the application of this type of surface course has been reduced significantly. Besides other problems like clogging, durability and difficulties with local repairs the main reason for this change is the difficulty of winter maintenance on porous asphalt connected with increased risk for the highway operator. This is of special importance for the Austrian climatic situation with rather cold winters and a large numer of freeze-thaw cycles.


Over Ten Years Experience with Porous Road Surface
C.J. Padmos

The use of porous road surface is attractive because the emission of rolling noise is less compared with situations where Dense Asphalt Concrete is used. This means that provisions for noise abatement can be lower and cheaper. In general the road users are very satisfied with porous road surfaces. They are not interested in the lower noise production, but in the fact that under wet conditions they have less problems with plash and spray. Only in the case that a road will be constructed near living areas de people living nearby insist on the use of quiet road surfaces with the highest reductions. Is the use the use of porous asphalt so advantageous as it seems? What about the other properties of porous asphalt, does they remain on the same good values? After ten years it is good to make up the balance and to formulate the wishes for improvements.


Pavement Performance During Thaw Weakening
V.C. Janoo and L. Barna

Freeze-thaw tests were conducted on full-scale test sections in the Frost Effects Research Facility at the Cold Regions Research and Engineering Laboratory. During the thaw period, the test sections were subjected to a single wheel load of 133 kN. FWD tests during the thaw period showed that the maximum base reduction factor was 67%, and the thaw-weakening period was approximately two weeks. For the subgrade the maximum reduction factor was 56%, and the thaw-weakening period was about three weeks. The critical strains in the base were 300- 500% larger then pre-freeze strains. In the subgrade the critical strain reached over 1000% its pre-freeze values. Permanent deformations at the end of thaw in the base and subgrade were 12 and 55 mm, respectively. The total deformation of the base and subgrade when compared to the surface rut depth was 7 mm. The reduction factors for the base and subgrade were 64% and 3-6%, respectively.


The Model Mobile Load Simulator as a Tool for Evaluating Asphalt Performance Under Wet Trafficking
A. de F. Smit, L. Walubita, K. Jenkins and F. Hugo

Model Mobile Load Simulator (MMLS3) tests have proven capable of evaluating the moisture susceptibility and relative performance of asphalt mixes through the use of accelerated pavement testing (APT) techniques. A number of case studies using the MMLS3 are presented. Wet MMLS3 trafficking has been conducted under varying temperature conditions (ambient up to 50 °C). Non-destructive stiffness measurements with seismic devices were performed intermittently during MMLS3 testing to monitor changes in the asphalt stiffness. It was found that stiffness loss, micro-cracking, and stripping is evident when the mix is susceptible to water damage. Indirect tensile strength and fatigue performance in laboratory testing were used as a measure of the distress due to the microfracturing and water damage. It has been found that the fatigue life expectancy of asphaltic materials susceptible to moisture damage was significantly reduced by wet MMLS3 trafficking. Overall, the MMLS3, used in conjunction with forensic testing, appears to be a valuable tool for supplementing full-scale APT devices for evaluating distress due to wet axle trafficking. It offers the potential for evaluating surrogate tests for quick detection of susceptibility to damage due to wet trafficking.


The Effects of Freeze – Thaw Periods on a Test Pavement in the Danish Road Testing Machine
W. Zhang and R.A. Macdonald

Accelerated load testing (ALT) by dual wheel loading at load levels from 40 kN to 60 kN were performed on the second test pavement, RTM2, constructed and instrumented in the Danish Road Testing Machine between 1995 and 1997. To accelerate the deterioration and failure of RTM2, and the rehabilitated RTM3 test pavement, further freeze-thaw experiments were conducted during 1998 and 1999. The experiments consisted of accelerated load testing at 60 kN applied load during the thaw periods of three freeze-thaw cycles. During these freeze-thaw cycles, the temperatures in the pavement layers were monitored during the thaw periods, and the porewater suctions in the Subgrade, the pavement material responses and the pavement profiles were measured regularly. This paper introduces the RTM2 and RTM3 test pavements and the experimental procedures employed during these freeze-thaw periods. Models for estimating the plastic strains at the top of the Subgrade, the permanent deformation of the Subgrade and the pavement functional deterioration, which were developed during accelerated load testing of the RTM2 and RTM3 test pavements at constant +25°C conditions, are presented. Conclusions are drawn from the three freeze-thaw experiments.


The Fourth Power Law and Thin Surfaced Flexible Pavements
D. Alabaster, J. de Pont and B. Steven

Recent surveys of New Zealand’s Heavily Vehicle industry have indicated a desire to move to higher allowable axle weights. A series of accelerated dynamic loading tests have been conducted at the Canterbury Accelerated Pavement Testing Indoor Facility (CAPTIF) to investigate the proposed changes. The first test examined the appropriateness of using the 4th Power Law, developed on structural asphaltic concrete pavements, to determine increases in pavement wear on the thin surfaced flexible pavements widely used in Australasia. The experiment found power relationships between 3 and 9 for the conventional 4th Power law model. A proposed new twopart power model, the ‘compaction-wear’ model, more accurately reflected the performance of the pavement. The exponent values for the compaction component were between 1 and 3.4, and 1.8 and 3 for the wear component. This model has significant implications, which are not predicted with the 4th Power Law, for roading authorities considering increases in mass limits for heavy vehicles.


Asphaltic Plug Joint Characterization and Performance Evaluation
M.N. Partl, S. Hean and L. Poulikakos

This paper presents the initial results of research on polymer-modified asphaltic plug joints (APJ) on bridges and an overview on the practical aspects and scientific background of the new Swiss guidelines, which were released by the Swiss Federal Roads Office (ASTRA) and developed in close coordination with the German road authorities. It focuses on recent EMPA research for laboratory and field characterization of the APJ materials as well as on the evaluation of the relevant mechanical characteristics for the APJ system such as adhesion, permanent deformation and low temperature cracking. In addition, results from a construction monitoring and long-term field performance research project on eighteen plug joints on bridges are presented. Further development of accelerated full-scale laboratory tests at EMPA are also discussed. The experience with the new Swiss guidelines and the research results show, that the performance of polymer-modified asphalt plug joints can be improved significantly if materials are selected and handled professionally and construction is carried out by well trained teams. It was found that tests and requirements are suited to detect and eliminate inadequate materials and systems, thus reducing the risk of damage and encouraging the development and improvement of those joints. It was also found that accelerated full-scale laboratory testing might be of assistance in this respect.


Crack Geometry Analysis in Asphalt Samples Obtained from a Heavy Vehicle Simulator Test Site
P. Offrell, J.Harvey and C. Scheffy

The purpose of the study was to compare vertical crack propagation in samples obtained from a HVS test site and to know whether cracks propagate through all the asphalt layers, and if not, how deeply they penetrate. Furthermore, the purpose was also to investigate any correlation between compaction grades and crack appearance as well as the effect of bonding between asphalt concrete layers. An enhanced technique of marking and recording hairline cracks in the surface is described. The results show no significant correlation between airvoid content and crack initiation. However, the results do indicate a correlation between crack density and air-void content. It seems as if the lack of bonding between the two AC layers has to some extent prevented the crack from propagating down into the lower AC layer. The crack patterns are similar in the different layers, and indicate that the cracks have initiated somewhere close to the bottom surface of the upper asphalt concrete layer. The enhanced method of marking cracks using UV-light has proven very useful.


Evaluation of Asphalt Pavement Analyzer as a Tool to Predict Rutting
A.Cooley and P.S. Kandhal

A standardized laboratory test to predict the rutting potential of an HMA that is accelerated,relative-ly inexpensive, and able to be used during quality control/quality assurance (QC/QA) testing would be of great benefit. Currently the most common type of standardized laboratory test of this nature is a loaded wheel tester (LWT). The objective of this study was to evaluate the Asphalt Pavement Analyzer (APA), the most common LWT in the U.S., to determine its suitability as a general method of predicting the rut potential of HMA mixes. Ten HMA mixtures of known field performance were selected for testing by the APA. Various APA testing parameters were used in a full factorial experiment. The combination of testing parameters which provided APA rut depths closest to and highly correlated with the actual field rut depths, were recommended to fine tune the APA testing protocol.


Results of Accelerated Tests on Flexible Pavements in Brazil Using the UFRGS-DAER Simulator
F.P. Goncalves, J.A.P. Ceratti and R.M.Rodrigues

Full-scale accelerated tests with traffic simulators have been used to improve methodologies for pavement design. To evaluate the performance of conventional and polymer modified asphalt mixtures a study is being conducted on a full-scale test facility at the Federal University of the Rio Grande do Sul. A linear traffic simulator has been used to carry out accelerated tests on six test sections of flexible pavements in asphalt concrete. The most important variables for asphalt mixtures performance evaluation in field are: evolution of surface cracking, rutting in wheel tracks, deflections, horizontal tensile strains in the bottom of the asphalt layer, measured with electrical resistance strain gauges, and vertical stress measured with total pressure cells in different locations of the pavement structure. This paper presents an evaluation of the several stages that composes the experimental part proposed. Special emphasis is given to results involving residual horizontal stress developed on the granular base layer during the compaction process.


Finite Element Simulation of Rutting on Superpave Pavements
M. Hossain and Z. Wu

Rutting is a major distress on asphalt pavements. This paper presents a threedimensional (3-D) nonlinear finite element (FE) model to simulate the initiation and propagation of rutting damage on the Superpave test sections at the Kansas Accelerated Testing Laboratory (KATL). A creep model and a Drucker-Prager model was chosen to characterize the permanent deformation characteristics of the Superpave mixtures and the aggregate base and subgrade materials, respectively. The 3-D FE model was built in several steps. First, a two-dimension (2-D) linear-elastic FE model was constructed. Then, the proposed non-linear material models were incorporated into the 2-D model and finally, into a 3-D model. However, the 3-D simulation was deemed to be very time consuming to implement, and a 2-D model was chosen for full simulation. The rutting simulation using the 2-D FE model was done up to 10,000 repetitions of the K-ATL tandem axle load. The predicted rut depths checked reasonably well with the measured values.


Performance Prediction with the MMLS3 at Westrack
A.E. Martin,T. Ahmed, D.C. Little, F. Hugo, P. Poolman and M.Mikhail

The one-third scale Model Mobile Load Simulator (MMLS3) was used to traffic five pavement sections (including one replicate) at WesTrack to establish and validate its ability to reliably predict rutting performance under full-scale trafficking. Researchers found similar performance rankings and conducted statistical analyses for field and laboratory results to compare performance under both loading conditions of three coarse-graded sections that showed poor performance and one fine-graded section with good performance. A standard rut depth (RD) analysis method and MMLS3 testing conditions were suggested to determine a RD for (1) comparison with acceptable performance criteria or (2) use with a theoretical rutting ratio based on stress analyses to predict performance under full-scale trafficking. A quantitative comparative analysis based on the assumption that rutting is dependent on the pavement structure and loading and environmental conditions was conducted to successfully validate the hypothesis required in the performance prediction methodology.


The Implications of Dynamic Loading and Tyre Type on the UK Road Network
D.B. Merrill, D. Blackman and V. Ramdas

DIVINE, an international project sponsored by OECD, has highlighted the need to consider the interdependence of components of the road system (heavy goods vehicles, tyres, pavement construction, traffic levels etc.) when optimising the economics of road infrastructure. EC COST Action 334 examined more closely the effects of wide-single and dual tyres on pavement wear. This paper describes work carried out by TRL as part of the UK Highways Agency research programme, to interpret the outcomes of these two studies with respect to traffic and pavement conditions in the UK. The work described in this paper focuses on three principal areas: the effect of dual and single tyres; the influence of dynamic loading on pavement wear; and a whole life cost assessment of the significance of these effects. The effect of dual and single tyres was investigated using instrumented pavements in TRL’s Pavement Test Facility. In an accelerated trafficking trial the relative rates of rut formation were measured. Contact stress distributions under six types of tyre were measured and compared with the observed responses and deformations in the test pavements. In a complimentary investigation, the influence of dynamic loading on pavement wear is being examined on a number of test sites on UK roads. The road profile and deflection of these sites was surveyed over a period of time. Any deterioration was correlated to the dynamic load profile recorded by a specially instrumented heavy goods vehicle. Using results from these studies a whole life cost analysis will be carried out to evaluate the financial implications of these factors on road maintenance.


HVS – Testing of Icelandic Low Volume Road Structures
T. Ingason, L.G. Wiman and H. Haraldsson

Two typical Icelandic road structures were tested in the Heavy Vehicle Simulator facilities at VTI in Sweden. The aim of the project was to compare these two structures and to compare them with Swedish and Finnish structures tested with the HVS. Most of the public road system in Iceland is low volume roads (traffic less than 300 AADT on 79 % of the roads), but goods transport is growing and the heavy traffic is increasing. The structures tested were composed of Icelandic aggregates, transported to Sweden. One with unbound base course, and the other with bituminous bound base course, both had thin surface dressing on top. All results indicated more bearing capacity of the structure with bituminous base course. Propagation of rutting was greater than expected and twofold of rutting measured in a Swedish low volume road structure. It should though be considered that the base layer material used, did not fulfil specifications for grain size distribution and the bituminous base course did not get enough time to harden before the test started. Results from theoretical calculations (3D FEM) compare reasonably with results of response measurements. The data collected during these tests will be stored in a common Finnish-Swedish database.


Numerical Validation of Pavement Performance at the Louisiana Accelerated Loading Facility (ALF)
B. Huang, L.N. Mohammad, M. Rasoulian, F.L. Roberts and H. Qin

This paper presents the numerical simulation results of the Louisiana Accelerated Loading Facility test pavements. Three asphalt pavement test lanes: one with conventional hot-mix asphalt mixtures, one with crumbrubber modified (CRM) wearing course, and one with CRM asphalt base, were tested using the ALF. These test lanes were numerically simulated through a three-diminensional finite element analysis using the commercial finite element software, ABAQUS. Fundamental engineering properties of the conventional and the crumb-rubber modified asphalt mixtures were studied. A Drucker-Prager based viscoplastic constitutive model was used to characterize the asphalt concrete. Elastoplastic constitutive models were employed for other paving materials. Dynamic traffic load was applied during the finite element analyses. Pavement distress of rutting was predicted using the stress and strain obtained from numerical simulation and field collected data. These predictions were compared with the ALF test results.

09122 Authors Index

An aplhabetical listing of authors.
09123 Conference Introduction

– Foreword
– Conference Committees / Sponsors
– Daily guide to scientific sessions
– Conference Summary
– Distribution of participants by country
09124 Conference Opening Addresses

Welcome Address
– Thomas Egebo, Permanent Secretary of the Danish Ministry of Transport
– Mads Lebech, Mayor of Frederiksberg Municipality
– Henning Christiansen, Director General, the Road Directorate, Ministry of Transport
– S F Brown, University of Nottingham, Chairman ISAP Board

Conference Keynote Address – Analytical Tools for Design of Flexible Pavements
Per Ullidtz

This paper falls into two parts. In the first, the widely used analytical-empirical method of pavement design and evaluation is discussed and in the second two simulation models are presented. There are important differences between the assumptions on which the theoretical models are based, and the reality of pavement materials and structures, and these differences are important both for the determination of input values (elastic parameters) and for the calculation of pavement response. Linear elastic theory often results in incorrect moduli, when used for backcalculation of layer moduli from deflection testing, and in questionable stresses and strains, when used for forward calculation. Including non-linear materials characteristics may improve the theoretical model, but no theoretical model has yet been conclusively verified with experimental data. The empirical relationships used to predict pavement deterioration from critical stresses or strains, are equally problematic.

In the second part of the paper two theoretical models are presented. The first model simulates the deterioration of a section of pavement over time using an incremental-recursive process. The model is stochastic and considers the spatial variation of pavement materials, layer thickness and traffic loads as well as seasonal variations. The second model deals with the forces on, and the displacements of, the individual grains in a particulate medium. This model does not rely on empirical relationships to predict permanent deformation or failure, but models this in the same process as forces and displacements. The input to this model includes the grain size distribution, the shape of the grains and the degree of compaction, parameters that are similar to those used for specification of materials and quality control. It is concluded that an interaction between development of theoretical models and experimental verification is needed to improve the understanding and predictability of the complex process of pavement deterioration.


Plenary Session I – U.S. Perspective on Design and Construction of Perpetual Asphalt Pavements
David E. Newcomb

The construction of long-lasting hot mix asphalt pavements has been practiced for a number of decades in the United States. Full-depth (asphalt courses used for all layers above subgrade) and deepstrength (asphalt surface and asphalt base over a minimal aggregate base above subgrade) pavements were originally designed for 20-year life expectancies. One of the primary advantages to these designs was that the total pavement sections were thinner when compared to conventional designs of asphalt over thick aggregate bases. As these full-depth and deep-strength pavements performed beyond their design lives, many only required surface restoration such as a thin overlays or mill and overlay. This practice of replacing only the surface offers a number of rehabilitation advantages in terms of speed of construction (user delay costs) and construction costs. The challenge for today is to obtain a longer surface life on a long-lasting asphalt support structure. Recent efforts in materials selection, mixture design, performance testing and pavement design offer a methodology which may be employed to obtain very long term performance from asphalt pavement structures (greater than 50 years) while periodically (approximately every 20 years) replacing the surface (top 25 to 100 mm) of the pavement. This concept has been proposed for use in Europe and it is rapidly gaining acceptance in the U.S. The common theme in these approaches is to combine a rut resistant, impermeable and wear resistant top structural layer with a rut resistant and durable intermediate layer and a fatigue resistant and durable base layer. This paper presents the ongoing efforts to develop methods for Perpetual Pavements in the United States. The pavement is designed to avoid bottom-up fatigue cracking by providing a total asphalt thickness that keeps the tensile strain at the bottom of the layer below a critical level and/or the use of a slightly asphalt-rich bottom layer. Rutting resistance in the intermediate layer can be obtained through the use a large-stone mixture or a dense-graded Superpave mixture. The surface course is comprised of a premium asphalt mixture which may be either a stone matrix asphalt (SMA) or a high-quality Superpave mixture. For areas where the use of an open-graded friction course (OGFC) is desired, this feature can be incorporated and viewed as a renewable surface.


Plenary Session II – Overheads From the Plenum Session on Perpetual Pavements
Brian Ferne

Mr Ferne’s presentation slides.


Plenary Session III – Overheads From the Plenum Session on Noise
Ronny Klaeboe

Mr Klaeboe’s presentation slides.


Plenary Session IIII – Overheads From the Plenum Session on Noise
Hans Bendtsen

Mr Bendtsen’s presentation slides.


Reports on Plenum Sessions
Jorgen Christensen, Johann Litzka, Christian Busch

Report on the Plenum Session on Perpetual Pavements by J. Christensen.

Report on the Plenum Session on Noise by J. Litzka.

Report on the Plenum Session on AASHTO 2002 by C. Busch.


Open-Forum: New Contractual Relationships as a Driver of Innovations. Dissemination of Knowledge.
Mikael Thau, Joe P. Mahoney and David E. Newcomb

Panel discussion and audience comments/questions about the topics addressed.


Closing of Conference
S F Brown

Parting reflections from Stephen F. Brown, Chairman ISAP Board.

09133 Exhibitors

A listing of 9th Conference Exhibitors and their contact details.
09134 Participants

A listing of 9th Conference Participants and their contact details.