5th Conference

code ISAP 5th Conference – Titles & Abstracts
05000 5th International Conference on the Design of Asphalt Pavements – Volume I, preliminary pages and Table of Contents
05001 The Development and Implementation of Analytical Pavement Design for British Conditions
S.F. Brown, J.M. Brunton, P.S. Pell

A description is given of the computer program ANPAD which provides three basic facilities for using analytical procedures in pavement design calculations. These include the ability to compute the required layer thickness, the design life or asphalt mix proportions for a balanced design, the particular option being selected on the basis of the available data and design constraints.

Examples are given to illustrate some applications of the program. Prediction methods for dynamic stiffness and fatigue characteristics of asphalt mixes are used in ANPAD. However, the resistance of a particular mix to permanent deformation requires realistic laboratory testing and the techniques available are reviewed in the light of recent experiments. The implementation of analytical design methods in Great Britain is illustrated by some examples of full-scale trials in which unconventional asphalt mixes have been used. The paper also provides a description of the computer program CUDAM which includes the consideration of cumulative damage in fatigue when determining the required layer thickness of an asphalt base.
05002 Thickness Design of Asphalt Pavements – The Asphalt Institute Methods
J.F. Shook, F.N. Finn, M.W. Witczak, C.L. Monismith

Unlike previous editions, the ninth (1981) edition of The Asphalt Institute Thickness Design Manual (MS-1) is based on mechanistic design methodology. Elastic theory and concepts of limiting subgrade strain, to control permanent deformation, and limiting tensile strain in the asphalt layers, to control fatigue cracking, were adopted. Typical material Properties, modulus of elasticity and Poisson’s ratio, were selected from available test data.

A comprehensive computer program was developed to produce design charts for use in the manual. Design charts were prepared for structural sections consisting of asphalt concrete surface and base placed directly on subgrade, asphalt concrete with emulsified asphalt base and asphalt concrete with untreated granular base. Three environmental conditions, each consisting of a separate 12-month temperature distribution, were used to represent temperature effects on asphalt mixture properties. Untreated aggregate bases were considered to have stress-dependent properties. In addition, subgrade and untreated base properties were made to vary on a monthly basis to account for freezing and thawing effects. Loads were represented by a dual tire, equivalent to the load produced by a standard 80 kN single axle load.

An extensive verification study was made using the design charts prepared for the manual. Comparisons were made to several existing design methods, including previous editions of The Asphalt Institute manual. In general, the comparisons produced similar design thicknesses, except at very high traffic and high subgrade strength levels, particularly, when untreated bases are used.

Comparisons also were made to performance data collected on full-scale test road projects and existing highways. Six separate sources of data, representing 402 individual design comparisons, were used in the study. Results indicated that the ninth edition of The Asphalt Institute Thickness Design Manual (MS-1) produced design thicknesses that on the average were 41 mm (1.6 in) greater than the observed thicknesses, but also produced unconservative designs 12.7 percent of the time.
05003 Rational and Practical Designs of Asphalt Pavements to Avoid Cracking and Rutting
J. Verstraeten, V. Veverka, L. Francken

The structural design approaches to asphalt pavements, presented at the 4th International Conference (Ann Arbor 1977) are improved.

For the rational approach. the improvements concern the design against cracking by fatigue of the bituminous layers and are based on a new resistance criterion taking into consideration the beneficial influence of the rest periods between the loads. They also concern the design against permanent deformation (limitation of the rut depth) and this on the basis of new results obtained in the field of the mechanical properties which precise, for bituminous mixes, the law of permanent deformation to be considered in the practical applications.

For the engineering approach, and with a view to including the above improvements. new design charts are set up. Moreover,a simplified method to determine the resistance to rutting is presented; its application complements the information that can be obtained from the design charts.

Finally, experimental verifications are presented.
05004 The AGIP Viscoelastic Method for Asphalt Pavement Design
G. Battiato, C. Verga

In this paper a simplified viscoelastic method for pavement design based on the analysis of a three-layer system subject to moving loads is considered.

The first layer is representative of all the asphalt layers, whose mechanical characteristics are described by a very simple analytical expression which can be obtained by uniaxial creep laboratory tests. The second and the third (semi-infinite) layers are considered elastic. All the layers are supposed to be rigidly bounded and incompressible (Poisson’s ratio = 0.5).

A computer program called PAVIS 3 (Pavement Viscoelasticity – Three Layers) has been developed to calculate strains at all points of the three-layer system, caused by the passage of moving loads. The load is supposed to be uniformly distributed on a round track. The following notable viscoelastic effects have been foreseen:
– the lack of symmetry between strains calculated for positive and negative distances along the direction of load motion;
– the influence of speed;
– the difference between the maximum amount of longitudinal and transverse strains calculated at the base of the asphalt layers.

An experimental validation of the above theoretical results has been performed by measurements carried out on a large test track in southern Italy. The new viscoelastic design method allows the determination of the critical tensile strains in the bituminous layers, under different temperature situations and speeds of commercial vehicles, thus furnishing the necessary parameters for the calculation of the fatigue life of flexible and semi-rigid structures.

A complete subsystem for the prediction of permanent deformations in the asphalt layers is included.

A computer program (MOREL) has been developed for the calculation of permanent deformations. Temperature and traffic conditions are taken into account by means of a nomograph.
05005 Mechanistic Design for Timber Haul Pavement
F.T. Hsia, H.H. Richter, J.W. Padgett

This study was undertaken to establish a thickness design procedure for the design of timber haul pavements using the mechanistic approach. The procedure is based on a model in which the pavement structure is considered as a linear elastic multi-layered system, and the layered materials are characterized by their dynamic mechanical properties. The primary design criteria are fatigue life of the asphalt concrete and permanent deformation of the subgrade. A method of conversion was introduced to convert the timber sale volume to the form of 18-Kip equivalent axle axles. Possible alternative seasonal variations in timber traffic patterns were incorporated into the design system to ensure the optimum use of the pavement and to assess seasonal haul limitations during the spring thaw period. Alternative pavement design strategies were also included to cope with the involuntary underthickness situation. The evaluation of both mechanical properties and economic feasibility of locally-available construction materials from different sources is included in the material characterization subsystem of the design process to dictate the selection of the asphalt and the aggregate base. The material characterization subsystem is made possible through an environmental rectification procedure derived from an in-depth analysis of field measurements, meteorologic data and general trends of environmental patterns. The design was programmed onto a computer of modest size, but an example problem is provided to illustrate the use of the procedure.
05006 A Design for Pavement to Carry Very Heavy Traffic
N.W. Lister, W.D. Powell, R.T.N. Goddard

In the United Kingdom standards for the structural design of road pavements are based largely on results obtained from full-scale experiments. Over the 30 years that many of these experiments have been in existence, comprehensive date have been obtained on their performance. However, there remain serious shortcomings in their use to develop a comprehensive set of design standards. In particular, it is difficult to design roads to carry traffic much greater than the traffic levels accumulated by the full-scale experiments themselves and, because of the recent marked increase in the damaging effect of traffic on the most heavily trafficked roads, there is now an urgent requirement for designs well beyond the proven range of present design recommendations. This paper describes how this need has been met by the development of an interim design in fully flexible construction from the best current information from mechanistic models combined with that from full-scale road experiments and performance testing.

Fatigue cracking, although not usually the first cause of deterioration of fully flexible pavements in the U.K., is important, particularly on the more heavily trafficked roads, because of the consequent need to reconstruct both surfacing and roadbase of a cracked pavement. A computer model of the pavement structure has been developed to predict fatigue performance taking into account variations in pavement temperature and wheel-load spectra. The model has been calibrated by testing a pavement under controlled conditions in a pilot scale test facility and then validated by comparison with the fatigue behaviour of full-scale pavements of known performance.

To ensure adequate resistance to deformation, the deformation behaviour of pavements to carry heavy traffic has been examined in relation to that determined from full-scale and laboratory experiments. Linear elastic analysis was used to demonstrate that vertical stresses in the soil under traffic were not excessive.

This approach has been used to formulate novel designs for roads carrying very heavy traffic, cumulative traffic in excess of 80 million equivalent standard 80 kN axles over a 20 year design life. The design consists of a thick course of well compacted coated macadam placed between layers of rolled asphalt wearing course and lower roadbase. This construction combines the deformation resistance of dense coated macadam with the good fatigue properties of rolled asphalt. Such a design was first adopted for the reconstruction in 1979 of a section of the M6 motorway, where the new 20 year design life was 160 msa. Structural properties of pavement components and of the subgrade were checked during reconstruction and the performance of the pavement has shown the design to be soundly based.
05007 Design Procedure for Premium Flexible Pavements
F.L. Roberts, N. von Quintus, W.R. Hudson

In certain situations, particularly major urban freeways, it is practically impossible to close highways for maintenance due to extremely heavy traffic conditions. In order to overcome this problem, a comprehensive study for the Federal Highway Administration has been carried out to identify flexible and composite pavement structures which will perform as premium pavements with an absolute minimum of maintenance for a period of 20 years. This paper reports the results of a detailed study of the characteristics and requirements for such pavement structures. To accomplish this goal, the study identifies the major distress types which have seriously limited the maintenance-free life of flexible pavements, selects analytical models for use in a structural analysis for the pavement, and develops a design criteria for each distress type. A complete review of material properties was completed to identify those materials capable of satisfying the premium pavement criteria. As a result of this project, a detailed design procedure was developed in which fatigue and transverse cracking, rutting, roughness, and environmental factors such as temperature and moisture, were used to develop design charts for thickness and material selection to meet the requirements for the design and construction of premium pavements. This paper presents a portion of the basic design charts that were developed in that study and includes an example problem to illustrate the procedure.
05008 Pavement Design Optimization by an Analytical Method
I. Gschwendt, I. Poliacek

The paper deals with the description of an analytical design method’s application for flexible pavement at the optimization of pavement structures. Whereby the permissible stresses in the pavement layers, the compression stress on the subgrade as well as the protection of the pavement against subgrade frost penetration effects are taken into account. In assessing the pavement one works with the fatigue strength of materials and the use efficiency of this strength is evaluated by the aid of the "utilization coefficient”. In assessing the temperature regime of the pavement – the temperature change of bituminous layers and hereby the change of their strength and deformation characteristics – are considered also. The changes are expressed by the duration of certain periods and in the case of necessity the changes in traffic load in the course of the year are considered too.

With regard to the significance of several assumptions in the analytical method, a laboratory research of the materials and soils fatigue strength was carried out, further the temperature regime of pavements was measured, as well as the subgrade’s moisture regime. The interface friction of the pavement layers was investigated theoretically, by measurements on pavement models and during the construction of experimental sections. Several results of the research pointed to the necessity to make corrections in the hitherto considered assumptions.
05009 The Texas Flexible Pavement Design System
.L. Lytton, C.H. Michalak, T. Scullion

The Texas Flexible Pavement Design System incorporates all seven components of a complete pavement design system: (1) the "Russian deflection equation method" which uses elastic moduli and thickness of pavement layers to calculate pavement deflections; (2) a non-linear search technique to compute elastic moduli for multiple layers (2 or more) using surface deflection measurements made by non-destructive testing devices; (3) performance equations for all significant types of pavement; (4) distress prediction equations for all significant types of distress on each pavement type; (5) a computerized system for selecting the optimum pavement design for a specific project; (6) a computerized system for optimally selecting projects within a pavement network ( the subject of another paper in the proceedings of this conference); and (7) a consistent method of observing and rating pavement distress and performance.

The development of this system has encompassed a period of years requiring the construction of a flexible pavement data base which incorporates pavement construction histories, material properties, and layer thicknesses on approximately 400 sections of pavement along with climatic , deflection, ride, and distress data from several annual surveys.

This paper summarizes the developments in items (1) through (4 ) mentioned above, notes how they are used in item (5) and discusses the development of survivor curves which are used in item (6). Item (7) was developed several years ago, has been reported previously, and is not covered in this paper.

The entire system is now in a position to be of substantial assistance to the Texas State Department of Highways and Public Transport.ation in its continuing efforts to maximize cost-effectiveness in planning, designing, construction, and rehabilitating its entire state pavement network.
05010 Design Criteria of Asphalt Pavement Structures at High Service Temperatures
A.F. Bissada, S.K. Hamdanl, H.R. Guirguis

This paper includes a study of the behavior of two asphalt pavement structures under traffic and environmental conditions in Kuwait. The objective being to indicate those criteria that should be incorporated in the pavement design procedure with due consideration being given to the performance under the prevailing excessive pavement temperatures, which range between +10C and +75C.

The approach is based on monitoring the performance of each pavement during seven years of service at four different locations. This included an analysis of the major traffic, environmental, and material parameters affecting the performance of the pavement sections. The measured permanent deformation at each section was compared with the calculated reduction in the asphalt layer thickness based on the recent Shell procedure, and a correlation has been found in this respect.

The stress-strain values in these two pavement structures were determined for a wide range of moduli appropriate to the temperatures and time of loading at different hours of the day and for three given air temperatures: 35C, 27C. and 18C, representing the average daily temperature of the hottest, moderate, and coldest season of the year in this area, respectively. Weighted values of repetitions to failure were derived from three design criteria: subgrade strain, tensile strain in the asphalt layer and shear stress within the asphalt layer controlling its shear failure.

Using Miner’s hypothesis of damage accumulation, the relative damage due to each of these criteria was calculated during the day hours as well as for the whole seasons of the year. The results of this analysis show that the introduction of a criterion for limiting shear stresses within the asphalt pavement layers, in addition to the other two primary design criteria, is necessary for the design and control of permanent deformation failures at elevated service temperatures.
05011 Mechanistic Design of Asphalt Pavements and Verification Using the Heavy Vehicle Simulator
C.R. Freeme, J.H. Maree, A.W. Viljoen

The first approach to the mechanistic design of asphalt pavements in South Africa was presented at the Fourth International Conference on the structural Design of Asphalt Pavements (1) in 1977. This approach was discussed in some detail and a number of examples were given as verification. Several improvements have since been made, mainly as a result of differences found in practice and from extensive testing carried out with the Heavy Vehicle Simulator (HVS) on pavements throughout South Africa. These tests covered a wide variety of designs, materials, traffic and environmental conditions.

The current mechanistic design method is described in the paper. The method takes account of fatigue and deformation in bituminous materials (different approaches being used for thin asphalt surfacings and thick bituminous bases), crushed-stone materials, cementitious materials of various strengths, natural gravels and soil subgrades. The degree of sophistication in the characterization of these materials may range from the selection of data from tables, up to repeated load triaxial and beam bending tests.

The models used are based on simple linear elastic computer programs such as CHEVRON, and a purpose-modified program MECDE, up to the multi-loading program ELSYM.

The criteria for distress used for different materials were developed and improved over a period of time from both laboratory and field tests. Rut deformation limits are based on laboratory repeated-load measurements and the results of tests with the HVS. The limits for cementitious materials are treated in two phases – the pre-cracked and past-cracked phases.

A wide variety of different base types have been tested, using the fleet of four HVSs. These include pavements with crushed-stone bases, bituminous bases, cementitious bases and natural gravel bases, and concrete pavements.

The HVS test consists of applying a variety of wheel loads and measuring the elastic deflection and permanent deformation of the individual pavement layers. A large number of repetitions are then applied at a selected wheel load. During the test numerous indicators are used: these include deflection and deformation with depth; surface deflection and curvature. and rutting; state of cracking; surface profile; temperature; and density and moisture conditions in the pavement.

The large volume of data on the behavior of different pavement types has led to a high degree of confidence in the use of mechanistic design in South Africa. It has also been possible to modify designs in practice and to reduce pavement costs without a loss of confidence that the pavement will carry the expected traffic. In this way many millions of rands have been saved in South Africa, thus justifying many years of research into mechanistic design.
05012 ALIZE III Practice
P. Autret, A.B DeBoissoudy, J.P Marchand

The French method of the structural design of overlays and new pavements is based on several themes many of which have already formed the subject of L.C.P.C. communications presented at Ann Arbor over the last few years : the catalogue of new pavements by Mr Chantereau and Mr Leger (1977), non-destructive testing of pavements by Mr Leger and Mr Autret (1972) and Mr Guillemin and Mr Gramsammer (1972). diagnosis of existing pavement by Mr Bonnot, Mr Autret and Mr De Boissoudy (1977), test sections by Mr Sauterey and Mr Siffert (1972), laboratory tests by Mr, Bonnot (1972).

Today, now that the L.C.P.C. ALIZE III computation program is used extensively in laboratories and companies or in design offices requesting this program, it appears to be worthwhile completing the presentation of the method underscoring, in particular, the numerical values used in France, taking into account the experience gained over fifteen years of laboratory testing, in situ checks, site observations analyses, test section follow-up reports etc….

This article includes V chapters :
– theory of ALIZE program use;
– traffic : Traffic is represented by a typical unit load and application intensity. The French numerical values are indicated together with a key allowing a change of reference axle loads;
– materials : The main standard French materials are presented in the form of sheets summing up their mechanical and physical properties as well as the experimental numerical dues relating to them:
– practical application : This chapter details the process of calculation and indicates how to take account of the dispersion of thicknesses in place, the variation of the mechanical properties of a material and the offset coefficient to be applied to laboratory tests to take account of in situ behaviour in order to result in the design of the structure with some probability of meeting requirements;
– application example : To illustrate each of the three chapters, and also to demonstrate how to transpose the method to suit other conditions than the French environment, the application of a concrete example is given.
05013 Design of Pavements in Tropical Regions with the Finite-Element Method
F. Wolff

Bituminous pavements in tropical regions have – apart from the lacking anti-frost layer – an essentially thinner pavement construction than e.g. German roads with a respectively comparable traffic volume.

On introducing constant modules of elasticity in the layers into the calculation of these pavements with non-bound subbase (e.g. gravel) – only those are investigated here – tensile stress values occur at the bottom of the subbase which cannot be sustained by such materials as experience shows. The conclusion from this is that the stress-strain behavior of road construction materials can only be described with reservation with constant modules of elasticity. In the following stress-strain relationships, taken from latest literature, of asphalts (temperature-dependent modules of elasticity) and of subbase and subgrade (stress-dependent modules of elasticity) are presented.

As the calculation method used so far in road construction engineering – the multi-layer-theory – allows a variation of the stiffness of the materials only in the depth, the finite-element-method is applied here. The usability of this method is investigated by means of examples for which analytical solutions are at hand; based on this an appropriate FE-grid and a limitation of the axially symmetric continuum is developed for the later non-linear calculations. The non-linear stress-strain behavior is taken into account by the incremental method in the FE-program. The material-specific stress-strain behavior of road construction materials is shown by means of the developed FE-program with simple impressive examples.

Furthermore the differential stress-strain behavior due to the differential stress-strain relationships is presented for a given three-layer pavement. Completely different stress distributions in the pavement result from this compared with the calculation with constant modules of elasticity.

Furthermore characteristic values for the estimation of the bearing capacity of a pavement are developed, and the sensitiveness of the values to alteration of the system parameters is investigated. Then pavements which are given by a design curve developed empirically (Road Note 29, 31) are compared with the calculation method and the characteristic values developed herein.

It proves that the calculation method and the characteristic values are suitable to make statements on the stresses and strains of pavements.
05014 Use of Pavement Test Facility for the Validation of Analytical Design Methods
S.F. Brown, J.W. Pappin

The results from an experimental study of asphalt pavements with granular bases are used to illustrate the value of a pavement test facility for validating analytical design methods. The particular aim of the project was to continue development of analytical procedures for the prediction of permanent deformation in pavements but this also involved the measurement and computation of transient stresses and strains. Seven pavements were tested of which three are studied in detail . Each structure had about 40 mm of asphalt, about 140 mm of crushed limestone base and a silty clay subgrade of about 6% CBR. A new finite element computer program called SENOL was used to calculate stresses, strains and permanent deformations. The latter involved a novel two-stage procedure using a plane strain configuration for the final calculation of permanent deformation. The computations of stress and resilient strain were adjusted in the light of the observed behaviour and this exercise illustrated the importance of having adequate non-linear stress-strain relationships for all materials. Although reasonably accurate values of both resilient and permanent strain were calculated, the stresses were low.
05015 Behavior of Flexible Road Pavements Under Tropical Climate
F.J. Gichaga

Most road pavement designs in tropical Africa are empirical and are derivatives of design methods developed in Europe. Experience on some recently constructed roads shows that road pavements have tended to fail sooner than expected thereby leading to unplanned expenditure in the exercise of reinstating them. This paper describes field studies carried out to establish long-term structural behaviour of typical high standard roads in Kenya. The studies involved measurements of elastic deflections using Benkleman Beams, traffic loading patterns, cracking and rut depths for selected test pavement sections along six typical high standard bitumenized trunk roads of varying design procedures. Results of the studies show that pavements are weakened by repetitions of traffic loads but tend to develop strength with age. For a pavement approaching failure cracking is accompanied with high deflections. The older designs, which required mainly surface dressing as a form of surfacing, gave pavements which performed exceptionally well with service lives exceeding fifteen years. The pavement models which were based on newer design procedures, which specified thick asphaltic concrete surfacings, showed cracks and rutting in less than five years of service. Results also indicate that higher pavement deflections are obtained during the months of high rainfall and high temperatures.
05016 Development of an Asphalt Concrete Distress Criteria Using In-Service Planning
R.F. Carmichael III, W.R. Hudson

Data are presented from 63 hot mix asphalt concrete pavements which were selected to fill an experimental factorial of three levels of traffic, three levels of existing pavement surface conditions, and the different environmental regions of the Republic of Argentina. While laboratory fatigue testing and permanent deformation testing of hot mix asphalt concrete specimens offers valuable insight concerning the fatigue and permanent deformation resistance of various mixes, the in-service fatigue cracking and rutting performance of pavements is somewhat different. The objective of the reported study was to develop hot mix asphalt concrete fatigue cracking and rutting relationships using data gathered from in-service experimental test sections. The criteria were specifically developed for the Direction Nacional de Vialidad (DRV), National Highway Department, of Argentina as part of a project to develop a new pavement overlay design procedure. Recognizing the need for a criteria based on Argentine materials, mix designs, environmental loads, and considering that hot mix asphalt concrete reacts differently to in situ loads as compared with laboratory fatigue and deformation tests, the experimental study was designed and performed.

The experiment design and the types of data required are discussed. The primary data are Benkelman Beam deflections, thicknesses, common material’s strength tests, such as the California Bearing Ratio test, historical traffic records, and detailed condition survey information. In addition, the procedure used to establish final materials properties is out lined. Elastic layered theory was the analytical tool used with the field data to predict asphalt tensile strains and subgrade vertical strains. The developed fatigue and permanent deformation models for hot mix asphalt concrete pavements are presented. The final models show excellent agreement with other models in the literature and are being incorporated into a new overlay design procedure for Argentina.

The major steps required for the development and verification of distress criteria using data from operating field sections are presented. The 63 sections used for this development were selected from all parts of Argentina to include the effects of different climates, typical soil types, and construction materials used.
05017 Verification of the Design Method for Asphalt Pavements in Switzerland
I. Scazziga

A structural design method for asphalt pavements based on the results of the AASHO-Road-Test is currently used in Switzerland since now 10 years and a discussion about a possible revision is actually taking place. Informations from three different sources are analyzed in the phase of preparation of this revision. The study of the performance of different pavements under traffic, which have been subjected to a special observation program since their construction and now have up to more than 15 years of age shows, compared to the results of theoretical performance values, that with few exceptions under special conditions the pavements are behaving far more better than could be expected. The size and the degree of precision of some elements contributing to the study is not sufficient to give exact figures of the amount of over-designed thickness. However it can be said that pavements designed for 7 to 15 millions standard axles on a subgrade with a CBR of 10% could easily afford a thickness reduction of the equivalent of 50 mm asphalt concrete when their structure has unbound or to a certain extent cement stabilized materials in the subbase. These findings are already confirmed in the case of semi-rigid pavements by the intermediate results of full scale tests in a circular test track under moving wheel loads. In the comparison of different major design methods pavements designed by the Swiss or AASHTO method are distinctly the most conservative ones, those determined with the Shell pavement design method are the thinner ones at the opposite end. The truth, as in many cases, may well be in the middle of both extremes and this would also be confirmed by the observation program.
05018 Four Asphalt Pavement Case Studies Using a Mechanistic Approach
F. Hugo, P.J. Strauss, G.P. Marais, T.W. Kennedy

Several sections of road pavement in Southern Africa were overlayed in 1973 for different reasons. Three typical sites were selected from these for this study, one which had very little deformation but with some ageing, another section of similar design to the first but with greater deformation and cracking in places and a third section in good structural condition but with a serious problem as far as skid resistance was concerned. All overlays were designed using standard deflection reduction techniques and checked against the latest available information on mechanistic design methodology (1972). The three sections have performed well within the predicted distress criteria as suggested by mechanistic methods and showed no classical fatigue cracking and little deformation. Two of the three sections, however, displayed surface cracking within six years. Analysis of the materials indicated acceptable fatigue performance as well as relatively low stiffnesses as is generally required for thin asphalt. A detailed investigation of the deflection bowl using a modified Benkelman beam to monitor movement under a passing wheel showed a reverse curve right in front of the wheel which indicated considerable visco-elastic behaviour. This lead to the conclusion that in addition to the accepted modes of distress namely tensile strain at the bottom of the asphalt layer and vertical strain at the top of the subgrade, tensile strain at the top of the asphalt due to secondary and horizontal shear stress in the layers as a result of moving loads had also to be considered. This was especially important where the pavement material was susceptible to shear stresses or exhibited visco elastic behaviour. Finally the methodology developed was utilized for the design and construction of a relatively thin sand/calcrete asphalt layer directly on top of an in situ sand to obtain an economical solution using the available sub-standard material.
05019 The Performance of Highway Pavements in the Netherlands and the Application of Linear Elasticity Theory to Pavement Design
J. Ros, A.C. Pronk, J. Eikelboom

In 1971 the Rijkswaterstaat (National Public Works Department) constructed 9 trial sections, each 200 metres in length, on Highway A15 in the Netherlands. Each section was of a different structure and the range included structures in current use, variants on these and even experimental structures. The strains and pressures were measured at different levels in the structures when subject to the load of a standard lorry at varying speeds and temperatures. This paper presents results of these measurements.

The stiffness properties of the subgrade, sand cement, slag and the various types of asphalt were determined as far as possible both in situ and in the laboratory. The results were then used in the BISAR computer program and the computed strains and soil pressures were compared with the measured values. In general, the correlation was reasonable, particularly at the higher asphalt stiffnesses. Fatigue properties were also determined in the laboratory.

The results were also used to compare the computed relative life and the ‘actual’ life based on a 90% cut-off value for the measured strains; the former proved to be higher than the latter.

The performance of the trial sections under normal traffic conditions was monitored for an 8-year period.

Periodic measurements of skid resistance, evenness, rut depth and deflection as well as visual inspections have been carried out. The evenness of all sections was still good after 8 years. There were differences in the degree of rutting and cracking but these were still acceptable.

An overall assessment revealed some of the design methods to be conservative. The results and findings have been partly responsible for thinner structures being used on highways.
05020 Prediction of Rutting in Asphalt Pavements on the Basis of the Creep Test
H.J.N.A. Bolk

In this paper the results are given of further investigations into the practical applicability and accuracy of some existing methods for the prediction of rutting, based on the static creep test. Furthermore, a first step is taken in the development of a new prediction method (behavioural model) based on static as wellas semi-dynamic creep tests.

The measured rut depths on eight test sections were analysed in order to determine the rutting behaviour of asphalt pavements in practice. Using core samples from these sections, creep tests have been carried out which allowed the input material parameters for the various rutting behavioural models to be determined. This was followed by a comparison of the computed values with the measured rut depths.

It appears that of the existing prediction methods considered here, the Shell method produces the relatively best results. By using corrected temperatures an improvement can be achieved. The principal problem, however, remains the fact that the gradient of the calculated deformation curve is too small.

The main differences between the new method and the Shell method concern the determination of the air and asphalt temperature, the correction for the missing lateral confinement in the creep test and the correction for the dynamic effect. For the determination of the absolute magnitude of the rut depth the new method generally produces (apart from one test section) an equivalent or an even distinctly better result. It is especially important that the gradient of the calculated deformation curve agrees better with the behaviour in practice after the initial stage. It appears that a (slight) improvement in the ranking is obtained with respect to the Shell method, this especially applies to the overall picture.

This investigation is based on a retrospective prediction. Consequently, it appears possible to make a simple correction for the influence of an extremely hot summer period. In a final behavioural model such effects have to be taken into account in an analogous way.
05021 Verification and Calibration of PDMAP and COLD Computer Programs
C. Saraf

This paper presents the results of a study which was conducted to verify and calibrate the distress prediction models included in the two computer programs PDMAP and COLD. The specific forms of distress considered were fatigue cracking and permanent deformation, which were covered by PDMAP and low-temperature cracking covered by COLD.

The required data for this study were obtained from the Florida Department of Transportation and were used in the verification and calibration of PDMAP models. Utah Department of Transportation supplied the data for the verification and calibration of the model used in COLD program.

A brief description of the steps involved in this study is included in this paper. The method of verification and calibration of all the three models is also briefly described and the calibrated models along with their outputs are presented for comparison with the field observations.

The results of this limited study indicated that pavements built with two different surface thicknesses (2" and 3") will require two fatigue distress prediction models. However, only one rate of rutting prediction model was needed to predict rut depths in most of the test sections (88% of total) within +/-50% of the observed rut depths.

The COLD program was calibrated and used in comparing the predicted probability of cracking with the observed frequency of cracks in the field. The results of this comparison indicated that the model can predict the frequency of cracking qualitatively in pavements using the same grade of asphalt binder. Pavements using different grades of asphalts would probably need their own calibration of the prediction model.
05022 Verification and Application of the VESYS Structural Subsystem
W.J. Kenis, J.A. Sherwood, T.F. McMahon

At the Fourth International Conference, the general modular framework of the VESYS structural subsystem was presented. Since then, major improvements to the models were made. These improvements allow for the input of variation in material properties with seasonal environmental changes, discrete axle load groups, up to eight layers, prediction of fatigue in any layer, low temperature cracking damage and the permanent deformation occurring in each layer. One of three optional modes of the Primary Response Model uses the principle of quasi elasticity to compute time step solutions to the Chevron N-layer elastic theory BVP, resulting in the time varying response at any point in the pavement due to a stationary load applied to the pavement’s surface. This response is treated in a similar manner as in VESYS-IIM through Boltzman’s superposition principle to calculate the pavement’s viscoelastic response when the load is moving across the pavement’s surface at any speed. A second option allows for input of a single modulus for each layer and season, and the elastic layer theory BVP calculates a single time-independent response. A third optional mode computes the layer moduli indirectly from the layer creep curves. The terms "viscoelastic-plastic" and direct and indirect "elastic-plastic" are used to distinguish between the three different options available.

A comprehensive study of the predicted versus measured behaviors of selected AASHO Road Test sections is made using the new VESYS subsystem. In addition. the results of two in-house research studies conducted at FHWA are summarized as examples of the use of VESYS in a research capacity. One of these is an effort to evaluate the potential of "Sulphlex" paving mixtures in light of the recent energy crisis. Another is that dealing with the development of load equivalencies which realistically represent load induced distress in various environments. The use of these functions are a critical component in studies dealing with life cycle costing analyses such as cost allocation and selection of optimum maintenance strategies.
05023 Evaluation of In-Situ Moduli and Pavement Life from Deflection Basins
W.P. Kilareski, B.A. Anani

The determination of the in-situ pavement material moduli is an essential step in the non-destructive structural evaluation of flexible pavements. If the modulus values of the various layers can be determined, then the remaining pavement service life can also be predicted. This mechanistic evaluation can then serve as a rational basis for pavement management systems. This paper describes the results of a research project designed to determine the in-situ modulus values and remaining pavement life based upon the deflection basin obtained from a Road Rater device.

Elastic layer theory and the Gaussian method of elimination were used to formulate surface deflections in terms of elastic modulus, load intensity, and layer thickness. The resulting equations were used to analyze the effect of changing layer modulus on the surface deflection. On the basis of this analysis, a computer program for calculating in-situ modulus values from Road Rater deflection basins was developed.

The calculated in-situ modulus values were then used to calculate the maximum tensile strain in the bituminous concrete base laver of a four laver pavement system. The Road Rater deflection basin was used to calculate the surface curvature index, SCI; the base curvature index, BCI, and the base damage index, BDI. These criteria were correlated with the maximum tensile strain, the fatigue cracking of the pavement sections and the traffic loading. A relationship was established between the SCI and the traffic loading resulting in the prediction of remaining pavement life. Results are presented graphically in terms of the structural number (SN) of the pavement section.
05024 Distress/Performance Relationships and Overall Performance Concepts for Flexible Pavements
R.D. Pedigo, F.L. Roberts, W.R. Hudson, F.N. Finn

As traffic loads, environment, and other forces act upon the pavement system, the pavement responds with stress, strain, deformation, fracture, or wear. These responses are termed "behavior". Most of the analysis models currently used to design flexible pavement structures are predictors of behavior. When the predicted behavior reaches a limiting value, distress is initiated. With time, various types of distress develop and interact, and the resulting accumulation of distress acts to counter the ability of the pavement to serve traffic. It has long been recognized that the ability to predict such serviceability losses before they occur could be of great value to agencies that are charged with providing and maintaining roadways. In this paper flexible pavement distress/performance relationships are examined, the development of performance production models is discussed, and the combination of various pavement condition and performance variables into an overall performance measure is illustrated.

Pavement condition and serviceability data from several sources were examined, and distress/ performance relationships were investigated. It was found that rut depths less than 0.2 inches (0.5 cm.) have no effect on serviceability. The variable which was most highly correlated with serviceability was (root)C+P where C+P = cracking plus patching in sq. ft. per 1,000 sq. ft. Relatively, low values of R^2 were obtained from regression analyses, and the addition of other distress variables did not significantly improve the fits obtained with C+P alone.

Some significant problems encountered with the data examined in this research effort are discussed. The authors suggest that in the light of these problems, indirect or interactive performance modeling techniques are more suitable than standard approaches to performance modeling.

The greatest current demand for pavement performance models comes from agencies that have begun to develop pavement management systems. Systematic pavement management requires that comparison be made among alternative plans of action in order to determine the most appropriate strategy for designing, maintaining and rehabilitating a pavement section or a pavement network. If several factors are to be involved in the comparison of alternative strategies, then a method of combining these factors into a single overall performance indicator is required in order to carry out such comparisons efficiently and economically. Three methodologies for achieving such a combined index are discussed: (1) deduct values, (2) utility theory and (3) unique sums. Some advantages and disadvantages of each technique are noted, and examples of the use of these techniques are provided. These methodologies provide highway agencies with a means to efficiently construct a more relevant measure of pavement performance for improved pavement management.
05025 Development of a Road Profile Statistic for Compatible Pavement Evaluation
D.W. McKenzie, W.R. Hudson

One of the most important applications of the Surface Dynamic Profilometer is to provide a stable calibration reference for Response-Type Road Roughness Measuring instruments (RTRRM). The latter devices, of which the Maysmeter is typical, are relatively inexpensive and are used by many agencies for routine pavement monitoring.

Two approaches are possible for developing profile statistics suitable for both roughness measuring device calibration and for general roughness evaluation. The first is dynamic modelling of a hypothetical device, with certain physical constants pre-defined and with a sequence of profile elevations taken as system input. The alternative statistical approach which is described in this paper is to obtain data from RTRRM instruments on representative road sections whose profiles have been measured and then use regression techniques to select a profile statistic which the instruments are capable of measuring precisely.

A special class of profile statistics, termed Root-Mean-Square Vertical Acceleration (RMSVA), has been shown to reveal many of the road surface properties normally associated with roughness. An RMSVA-based roughness index, which was tailored to describe the behavior of eight Maysmeters run on 29 pavement test sections, is now the basis of a large-scale calibration program in Texas.

Although the Maysmeter calibration problem motivated the development of RMSVA roughness indices, careful monitoring of a set of calibration test sections and other pavements have revealed interesting surface properties that could never be detected by Maysmeters or by other RTRRM devices which reduce roughness evaluations to a single number. The RMSVA indices computed from a road profile can provide a "signature" that reflects roughness over a broad range of profile Wavelengths. Already, distinctive signatures corresponding to certain pavement classes, or types of deterioration, have been tentatively identified and are presented herein. Their interpretation remains a promising subject for future research.
05026 A Method for the Evaluation of the Structural Condition of Pavements with Thick Bituminous Road Bases
M. Dauzats, R. Linder

The first disorders found in the bituminous pavements built in France since 1979 have led to accelerated development of a method for the analysis and follow-up of this type of pavement.

After a summary statement of the behavior of bituminous pavements, based on 1400 km, this document describes a method based in part on the general principles of organization common to all stabilized structures and in part on specific tests; the originality of the method lies in the interpretation of these tests. Following a short description of the means of investigation, a large share of this document is devoted to analysis of the mechanical performance of the pavement based on such standard tests as deflection, the product R x d, and recent tests like the ovalization test, the value of which is that it measures elastic strains in the pavement. Tests of materials taken in situ, in particular the direct tensile test whose foundations are given, serve to determine the moduli of the materials and the characteristics at failure, and, via correlations, the allowable fatigue strain at 10^6 cycles.

The method makes use of calculation of the probability of failure of the pavement based on the probabilistic hypothesis of the appearance of structural deterioration according to the dispersion of thicknesses, the nature of the material, and the fatigue law of the material.

This method, used in the follow-up and maintenance of pavements, helps to judge the structural quality of the pavement and provides the information needed in deciding what maintenance work should be done, or in some cases for overlay design.
05027 Lacroix Overlay Design by Three Layer Analysis
W.Th. Hoyinck, R. van den Ban, W. Gerritsen

A practical method is presented for evaluating existing pavements by three layer analysis of Lacroix deflection bowls. To this and the two upper layers are first considered to act as one compound layer, behaving like a plate, resting on an elastic foundation of infinite depth, according to Hogg’s model.

By inserting a Young’s modulus for the top layer applicable to the Lacroix test conditions, the compound upper layer can be split up in two layers, having the same flexural rigidity. Testing of both uncracked and cracked sections of the same road is sufficient to express the structural properties in terms of affective layer thickness of the first layer and Young’s moduli of the second and third layer. These data can then be entered into the Shell 1978 Design Manual to establish the required overlay thickness. Taking fully into account the settlements of the supports of the reference T-frame of the Lacroix Deflectograph as well as the supports of the Benkelman Beam resulted in a relationship between Lacroix and Benkelman deflections. This enables one to use Lacroix data for evaluation of pavements by methods requiring the input of Benkelman Beam deflections. Verification of the adopted model is demonstrated for different types of construction and temperature conditions by comparing computed deflection bowls with recorded ones. A practical example of the evaluation and overlay design method is given.
05028 Structural Capacity of In-Place Asphaltic Concrete Pavements from Dynamic Deflection
H.F. Southgate, G.W. Sharpe, R.C. Deen, J.H. Havens

The proper design of asphaltic overlay thicknesses involves four major factors: the in-place modulus of the subgrade, an estimate of the structural capacity of the existing pavement, estimates of the future traffic expressed as equivalent axle loads and required or desired design levels, and a thickness design procedure. This paper deals with estimating the in-place subgrade modulus and the remaining load-carrying capacity of the existing pavement. The method presented herein is valid for any Road Rater or other dynamic tester such as the Dynaflect. This procedure was based upon a 600-pound (272.4-kg) peak- to-peak dynamic load applied at a rate of 25 Hz. The steady state deflections have to be adjusted for load, dynamic frequency, and location of sensors. This method should be applied only to those testers that use a constant vibratory load.
05029 Pavement Survey as a Mean to Define a Road Network – Preventive Maintenance Policy
M. Poulet, J.C. Gramsammer

In 1965, France initiated a vast program for the strengthening of its national roads (about 30,000 km) and the development of its motorway network (over 5,000 km in 1981). In order to preserve the capital thus invested and to avoid the reoccurrence of the situation which, in the winter of 1962-1963, led to the serious deterioration of part of its highway facilities, a maintenance policy, as preventive as possible, was set up as of 1972.

This maintenance policy is based in particular on systematic and periodic surveying of the condition of pavements. To accomplish this, the French Road Research Laboratories (Laboratoires des Ponts et Chaussees) were required to adapt and to develop nondestructive surveying systems, initially designed to study the behaviour of traditional flexible pavements, to enable the evaluation of pavements having thick layers of treated materials, the investigation of surface properties (skid resistance, evenness) of pavements, and the surveying of road alignment characteristics and installations.

After having described these new investigation facilities and illustrated briefly their applications within the area of continuous pavement surveying and more detailed nondestructive testing, the article describes the methods used for processing the measurement data, and in particular the creation of a highway data bank.
05030 Aspects of the Interpretation and Evaluation of Falling Weight Deflection (FWD) Measurements
A.C. Pronk, R. Buiter

In the Netherlands the Falling Weight Deflectometer (FWD) is used more and more for routine evaluation of existing pavements, design of new pavements and investigation of mechanical properties of (new) pavement materials (recycled asphalt, crushed concrete, cement bound crushed asphalt etc.).

The paper deals with some aspects and problems that may arise with the interpretation of FWD measurements. Subjects like the apparent velocity of the deflection disturbance, the effective layer thickness of the asphalt, the determination of the subsoil modulus, non-linearity and the effect of loose asphalt layers are involved.

For this purpose in 1978 a test program on a number of full-depth flexible pavement structures was started. Afterwards pavements with a bound road base (sand cement, slag) and a few with unbound bases (crushed concrete, crushed asphaltic concrete) were included in the test program.

The result obtained on the basis of a linear-elastic multi-layer model were compared with results from surface wave propagation methods and laboratory tests on in situ taken cores. Although further investigations into inertia effects are desirable it can be concluded that falling weight deflection measurements using a (non) linear-elastic multi-layer model provide reliable results for the structural evaluation of pavements.
05031 Simplified Methods for Evaluation of Asphalt Pavements
J.M.M. van der Loo

In this report a procedure is described to evaluate flexible pavements in a quick and simple way by means of deflection measurements with the falling weight deflectometer. Full depth asphalt pavements and pavements having an unbound base layer are considered as a linear elastic multi layered system. With aid of a computer program developed by Chevron deflections and strains at the interfaces of the layers are calculated. At a standard asphalt temperature of 18C the design life of a pavement, expressed in a number of standard axle loads, is derived from a permissible strain level in the asphalt layer and the subgrade. Due to a test load of 50 kN the characteristics of a deflection bowl are related to the design life of the pavement. Within a range of operating temperatures of 5 to 30C the results of the deflection measurements can be related to the standard temperature. Related to the deflection characteristics the residual life of a pavement can be estimated from available data about passed and expected number of axle loads. The simplified evaluation method has to be an aid for determining road sections having a high priority for maintenance planning. Measuring a large number of roads of a road system in a more general way these sections can be selected. The method is not dealt with determining overlay thickness. Detailed investigations of selected road section together with economical considerations will give the answer how the pavement has to be improved.
05032 A Pavement Management System for Provincial Roads in the Netherlands
A.A.A. Molenaar, Ch.A.P.M. van Gurp

This paper describes a pavement management system developed by the Laboratory for Road and Railroad Research of the Delft University of Technology. Since the described models and techniques are developed, using data gathered on provincial roads, the system is intended for this type of secondary roads. For reasons of efficiency and simplicity distinction is made between evaluation techniques on two levels, i.e. on a network level and on a project level. Special attention has been paid to the development of performance models to predict the deterioration of the pavement condition in terms of strength, ride quality and skid resistance. Overlay design graphs are given for the rehabilitation of the pavement strength and ride quality of the pavement surface. At the same time it will be shown how the evaluation techniques are used to estimate the budget required to maintain a pavement network. Finally a method for the optimization of the overlay design on project level will be described.
05033 Improved Pavement Performance Relationships in Brazil
C.A.V. Queiroz, W.R. Hudson

A considerable amount of public money is spent on pavement maintenance every year, as pavements continuously deteriorate over time due to traffic loadings and climatic factors. For efficient use of maintenance resources, it is necessary to estimate the future condition or deterioration level of the different pavement sections in a specific network. This estimate is only possible if the pavement engineer or planner has available reliable predictive models.

Predictive models of this type can be used to identify those road projects most in need of preventive and corrective maintenance or rehabilitation. Pavement predictive models also find their application in Pavement Management Systems, in which all the activities involved in providing pavements are considered in an integrated, coordinated manner.

A major highway research project has been conducted in Brazil since 1975, and has the objective of yielding parameters, methods, and models to permit the optimization of investments in highway construction and maintenance so as to minimize the total road transportation cost. To accomplish this, considerable data has been collected in Brazil on pavement behavior and performance. Pavement attributes have been monitored for up to three years and include measurements of roughness, cracking and patching, rutting, and Benkelman Beam deflections. Test pits were dug to measure in-situ density, CBR, and moisture while laboratory testing produced material gradings, Atterberg limits, laboratory CBR and density of the road materials, as well as asphalt surfacings resilient moduli. These data, predictive models, and test results are analyzed in this paper.
05034 An Integrated Maintenance System for the Assessment, Diagnosis and Treatment of Flexible Pavements
M.S. Snaith, H.J.H. Bailie, E.T. Stewart, D.M. Orr

The Department of the Environment for Northern Ireland, Roads Service and the Department of Civil Engineering of the Queen’s University of Belfast established a Joint Project to effect improvements in the planning of road maintenance in Northern Ireland. The problem has been tackled by considering maintenance under three headings: relatively rapid road condition surveys, priority assessment for remedial treatment, and objective quantitative testing of the road pavements with various selected road machines.

The MARCH program suite was available from the City Engineers’ Group in England and was adopted to give an initial framework for maintenance management. The screening and priority assessment are carried out by the program following data collection by relatively unskilled staff, for discrete maintenance lengths of road. In parallel with this, considerable work has been carried out in the development and proving of various road machines to ensure an accurate assessment of serviceability and to assist in the determination of the cause of any decline where this is observed.

To obtain the maximum benefit from the above, a data bank has been established to act as an interface between the MARCH suite and the information gathered by the road machines. Furthermore it is suggested that within this data bank a "diagnostic" subroutine may be used to determine both rapidly and accurately the state of a road pavement.

The paper demonstrates the manner in which the various individual developments forma coherent maintenance planning system, which in turn seeks to standardise pavement assessment, enhance remedial treatment selection and maintain a reasonable level of pavement serviceability in a time of financial restraint.
05035 Rational Pavement Management in the City of Amsterdam
Ir. G.H. Kellersmann, Ir. J.v.d. Klooster

Within the scope of increasing interest for rational pavement management the Road Department started to build up an automated Road Databank (comprising about 6,300 ‘street-sections’) in 1975 and started to build up a new Pavement Management System in 1978. This new system can be fitted in a model (see figure 1) and comprises a procedure which will be executed continuously with a cycle time of 2 years.
1. Execution of a broad Periodical Visual inspection (P.V.I.) and Overall Measurements on all street-sections to separate a group of ‘good’ and ‘suspect’ street-sections.
2. Execution of an Extensive Visual Inspection (E.V.I.) and Supplementary Investigation only for the suspect street-sections to separate a (second) group of ‘good’ street-sections and ‘bad’ street-sections.
3. Determination of road technical structural maintenance measures only for the bad street- sections.
4. Determination of final structural measures also defined by other boundary conditions like renewal or installation of pipes and cables, tramrails, modifications of road profiles, only for the bad street-sections.
5. Preparation of structural maintenance programs on long and on short terms.
6. Updating these programs by means of results of the latest E.V.I. and Supplementary Investigation.
7. Relating budget shares to maintenance levels by means of data in Road Databank (using Road Databank as a management instrument).

– All data of street-sections are stored in the automated Road Databank, so that ranges, tests and selections of groups of good, suspect and bad street-sections can be executed by computer.
– The Overall Measurements concern evenness, road friction and deflection.
– The P.V.I. and E.V.I. are based on the well known visual inspection system developed by the Texas Transportation Institute, whereby the E.V.I. registrates the location of the damages on maps.
– The Supplementary Investigation concerns measurements of evenness, road friction and deflection; estimations of axle loads; plate bearing tests, creep tests, etc.
05036 Washington’s State’s Pavement Management System
R.V. LeClerc, T.L. Nelson

This paper reviews the problem facing transportation agencies with their present method of financing pavement upkeep and rehabilitation, and describes a solution based on pavement management practices. The goals of the pavement management system developed for the Washington State Department of Transportation are presented, together with a description of the four broad areas of data processing which combine to constitute the foundation of the system.

Basic information file data on pavements which were available or generated to form the background for the system are listed, and the incorporation of these data into a master file is described. Also shown are the methods used to analyze and convert the master file data from pavement condition ratings to pavement performance curves for each project. The performance curves are then used, together with appropriate cost data, to optimize rehabilitation at the project level, producing the most cost-effective type and time of pavement fix. How the network level programming follows from project level optimizing to give a six-year (or other) rehabilitation schedule is set forth. Included is the monitoring feature which provides a summary of pavement condition throughout the pavement system before and after the rehabilitation action.

Means for adjusting the program to fit budget or manpower constraints are reviewed. The use of the PMS to determine what the rehabilitation dollars will give in terms of pavement serviceability over the six-year period is discussed, with illustrations to show how the program can be adjusted to achieve a particular level of serviceability. Also shown are the means for utilizing the program to demonstrate the consequences of delayed rehabilitation.

Other possible applications and uses of the WSDOT/PMS data analysis procedures in areas of pavement design, construction, and maintenance are outlined.

It is concluded that the system, operating on biennial pavement condition ratings, provides a good solid framework for orderly analyses to estimate the economic benefits of the type, timing, and sequence of rehabilitation activities applied to a pavement. As such it is expected to be a great aid to WSDOT in providing the citizens of the state with the best pavements for their tax dollars. The system features procedures for easy updating and simple modification which should make it particularly suitable and attractive for adoption by other transportation agencies utilizing their own pavement data information files.
05037 Development of an Improved Pavement Management System
D.R. Luhr, B.R. McCullough, A. Pelzner

The computer program, Pavement Design and Management, System (PDMS), was developed at The University of Texas at Austin through a cooperative effort with the U.S. Forest Service. This pavement management system calculates and optimizes pavement design and rehabilitation strategies on a project level after being given information about available construction materials, material characteristics, expected traffic volume and loads, various costs, and required pavement performance. An improved version of PDMS was developed that greatly increases the capabilities of the system, while at the same time improves the rational basis for predicting pavement performance.

A structural analysis of AASHO Road Test pavement sections was conducted using a non-linear elastic layer procedure. The materials were characterized for four seasonal periods in this analysis, and the results compared favorably with deflection measurements made at the road test.

A regression analysis was performed to develop a performance prediction equation. The dependent variable used was the change in Present Serviceability Index (PSI), divided by the change in vehicle applications, for each seasonal period. The use of this performance variable allowed the consideration of seasonal characteristics of pavement materials and seasonal traffic volumes. Pavement deterioration is calculated by adding the change in PSI due to each vehicle type, instead of using Miner’s rule to estimate cumulative damage. Since each vehicle type is considered separately, there is no need to consider equivalency factors. This is an important advantage since AASHTO equivalency factors are found to have serious limitations.

The performance variable is used to predict the PSI-Traffic curve for the pavement structure, thereby allowing the evaluation of the performance area under this curve. This performance area can be used to calculate the cost/area ratio, which assists in evaluating the pavement performance, especially when vehicle user costs are not known.

It is concluded that the additional capabilities provided through these developments are important to a pavement management system. The mechanistic approach used in developing the performance model provides a good foundation for the inevitable extrapolation that occurs in design practice. However, it is no substitute for a comprehensive data base.
05038 The Development and Use of a Pavement Management System in the United Kingdom
P.D. Thompson, L.W. Hatherly

In 1970, the Committee on Highway Maintenance, set up by the UK Ministry of Transport reported on all aspects of highway maintenance including standards, engineering, management, training and many others. For the management of maintenance it recommended the development of a maintenance assessment system which led to the system produced by the Transport and Road Research Laboratory (TRRL), Computerised Highway Assessment of Ratings and Treatments (CHART).

The CHART system embraces inspection, rating, computer processing and the output of information which can be in several different forms. In addition to the collection of CHART visual information, the overall management system encompasses mechanical measurements for the assessment of structural condition (deflection), of skidding resistance (SCRIM) and texture measurement and of road profile. The use of the Bump Integrator and of the high-speed profilometer (HSP) for profile measurement is described.

The paper deals with the development of the CHART system and its various components as well as illustrating the way in which the rating is carried out by the computer. The various items of mechanical equipment are also described and illustrated.

The rating system is complicated and varies depending on the defect being rated. For interpretation of the priorities for structural treatment, a number of ratings are combined. Although it is not possible to comment on the range of ratings computed, an illustration is given to show how the results are interpreted and the priorities assessed.

A variation of the assessment system output is used by the Greater London Council in the management of the Capital’s major road network. For the mechanical measurements and accident records, transparent overlays are produced which can be placed on a map of the network. In this way, the engineer can quickly assess those sites in need of his attention.

Finally. some innovations are referred to which are being developed in the course of the continual enhancement of the system.
05039 Development of a Pavement Management System for the Arizona Department of Transportation
R. Kulkarni, K. Golabi, F. Finn, E. Alviti, L. Nazareth, G. Way

A comprehensive Pavement Management System (PMS) was developed to assist the highway engineers and managers of the Arizona Department of Transportation in making consistent and the most cost-effective decisions related to maintenance, design, and rehabilitation of pavements. The major elements of the PMS are: a Data Base Management System, a Network Optimization System (NOS), and a Project Optimization System (POS). In this paper, the development, testing, verification, and implementation of the NOS are described.

The NOS can be used to determine which rehabilitation policies will achieve prescribed performance standards at a minimum cost. This information can be used to prepare 1-year, 5-year, and 10-year pavement rehabilitation budgets to maintain prescribed performance standards. With an iterative procedure, the NOS can also be used to determine the highest performance standards that can be maintained with a fixed budget. These results can be used to allocate a given budget among different roads in the state, in a manner that provides the best possible value for the public dollar.

The NOS is fully operational on the Arizona DOT’S computer facilities. The system is currently being used in assisting the management to formulate budgetary and engineering policies related to maintenance and rehabilitation of pavements. Because of its modular nature, it has been possible to make several improvements in the NOS with only input data restructuring, but no programming changes.
05040 Implementation and Verification Examples of Successful Pavement Management
M.A. Karan, R. Haas, D.A. Kobi, A. Cheetham

The development of pavement management technology has undergone considerable progress during the past decade. Techniques now exist for all components of the pavement management process, although they are still far from perfect.

An important development in pavement management has been at the network level. However, this has been relatively recent in that pavement management started at the project level. As a consequence, most pavement management implementation and verification to date has been at the project level and has been fairly widely reported in the literature. In the late 1970s, however, the first successful cases of implementation and verification of network level pavement management began to appear. These have involved a comprehensive set of procedures and techniques that can be applied in a common framework and are applicable to a range of jurisdictions from small cities to large State or Provincial Highway Departments.

The foundation of pavement management is good data and its evaluation. In order to be used effectively, however, the data and evaluation should itself be "manageable". A means for achieving such manageability, not only in evaluation but also in using the information in pavement management, lies in automation. Savings in time and costs, through increased efficiency, can be realized. Such automation can range from the initial data acquisition, and its processing to the priority programming for networks through the actual design and economic calculations for individual projects to the use of computer graphics and tabulations for communicating the results.

This paper describes a comprehensive set of pavement management procedures which have been developed and successfully applied to a variety of rural and urban networks. It outlines a general framework for these procedures, involving both the network and project levels of pavement management, and then provides examples of the implementation at both levels. These examples begin with the acquisition, analysis and presentation of inventory data. Use of the data in network level implementation is then described as well as the development of a priority program of work and the evaluation of alternative budget levels. Automation in data acquisition and processing, data analyses, priority analyses and optimization including the economic evaluation, and computer graphics presentation of results is illustrated in the examples.

Brief examples of project level implementation are also provided. These similarly illustrate the use of automation from initial data acquisition to design analysis to final presentation of results,
05041 Optimization of Highway Maintenance and Structural Rehabilitation
D.W. Potter, W.R. Hudson

Funds allocated for maintenance and rehabilitation of road pavements are becoming an increasingly important component of highway authority budgets. This paper describes an investigation into the effect of the maintenance/rehabilitation policy on the cost to the highway authority and the cost to the road user over a specified analysis period. Two-lane rural roads with a range of pavement structures, pavement conditions and traffic volumes are considered. The Highway Design Model is used to estimate both authority and user costs under a range of maintenance/rehabilitation policies. Policies which minimize overall cost (authority cost plus user cost) are determined, together with the sensitivity of overall cost to maintenance/ rehabilitation policies. Policies which minimize overall cost (authority cost plus user cost) are determined, together with the sensitivity of overall cost to maintenance/rehabilitation policy. A methodology for incorporating future uncertainties (traffic growth, monetary inflation and available funds) is included.

The results of the study show that maintenance activities can significantly affect overall costs. It is realized that this study is very restrictive in terms of the road traffic situations considered. However, coupled with studies of a similar nature (Refs 18 and 19) a reasonable basis is established for this conclusion. Detailed conclusions concerning recommended maintenance practice for specific cases and the associated cost saving are presented. The point is made that the recommended maintenance practice can vary considerably depending on the aspects of the case being considered. Also, while cost increases associated with departures from the recommended strategy may be small when expressed on a percentage basis, the monetary value is by no means insignificant.

The study indicates’that quantities which depend on future events can be incorporated into a present-time evaluation in a realistic manner. Although decision analysis methods have been used in the areas of business management for several years, their adoption by road construction and maintenance authorities has been somewhat slower.

Because discount rate has a major influence on costs when the analysis period is-medium to long term, the need for a careful interpretation of the effect of inflation in evaluating future costs is emphasized. Further, considerable effort is warranted in order to obtain a satisfactory estimate valid for the duration of the analysis period.
05042 The Texas Rehabilitation and Maintenance System
R.L. Lytton, D.T. Philllps, C.Y. Shanmugham

The Texas Rehabilitation and Maintenance System (RAMS) is a series of computer programs which have been written to assist in the management of pavement networks. In the management process, specific projects and maintenance strategies must be selected and the best sequence of rehabilitation and maintenance activities over a planning horizon must be planned so as to remain within the available resources of budget, men, materials, and equipment. The selected projects should maximize the expected benefit from the use of these resources in terms of reliability and traffic served.

This paper describes the computer programs in some detail, gives the mathematical formulation of two of them, discusses how all of them are used in sequence in achieving an optimal set of project selections, and gives some simple examples of the use of the programs. Reference is made to more detailed reports which document the programs and describe the algorithms used in solving the optimization problems that are involved. The algorithms are described as state-of-the-art developments in operations research.

The example problems demonstrate that computer-assisted decision making in pavement rehabilitation and maintenance management can be expected to realize between 2 and 27 percent more beneficial project selections.
05043 A Method of Integrated Priority Programming and Budget Level Analysis for Pavement Maintenance and Rehabilitation
R. Haas, A. Cheetham, M.A. Karan

Total expenditures for maintenance and rehabilitation of a pavement network should represent an optimum combination of the two types of activities. Under the usual situation of budget constraints, this requires priorities to be established.

This paper describes an integrated method for programming maintenance and rehabilitation for paved road networks for any chosen program period. It begins with a common inventory of condition, serviceability, structural adequacy, traffic, unit costs and other information. The maintenance programming subsystem evaluates alternative treatments for different types, density and severity of distresses and produces a demand-based budget using a maximization of cost effectiveness. The rehabilitation programming subsystem similarly evaluates alternatives and a priority list of year-by year projects, over the program period, based on benefit maximization, is produced. The total of maintenance and rehabilitation costs for any given year does not exceed the total budget limit.

A case application is provided to illustrate the method. It uses the arterial street network of a small city, subdivided into 100 sections. The outputs include section-by-section, year-by-year, recommended programs of maintenance and rehabilitation work.

An additional feature of the method is a capability for evaluating the long-term effect of various budget options of average network serviceability. The rehabilitation budget levels, one representing the expected funding and one being zero budget, have been tested for the case application. As expected, average network serviceability was estimated to decrease significantly over the 10-year programming period for the zero budget case.

Finally, it is recommended that year-by-year updates be carried out on the inventory and the maintenance and rehabilitation programs.
05044 Airfield Pavement Performance Prediction and Determination of Rehabilitation Needs
M.Y. Shahin, S.D. Kohn

Since many in-service airfield pavements are approaching their design lives, there is a need for a comprehensive pavement management system. Such a system has been developed for the U.S. Air Force by the U.S. Army Construction Engineering Research Laboratory (CERL). The system provides the user with a pavement condition rating procedure and methodologies to select a timely maintenance strategy. This rational and systematic approval should result in decreased maintenance costs.
05045 Development of a System for Efficient Pavement Management for Municipalities
P.C. Koning

A Working Group set up by the Studie Centrum Wegenbouw (Study Centre for Road Construction) is developing a system for the application of efficient road management by municipal road managers in the Netherlands. Starting from a provisional system, a suitably adapted system has been developed by testing and verification in actual practice. This has more particularly been done in a "modal" municipality chosen for the purpose in this country. To this end, a comprehensive data acquisition survey of the whole road network was carried out, and visual inspections and measurements were performed over a period of some years. Also, standards for maintenance criteria were established, prediction methods for the behaviour with the passage of time were investigated, and planning techniques were developed.

The result of this testing and verification procedure comprises recommendations for setting up a scheme of data acquisition and visual inspection and for the exemtion of various measurements and their value assessment. Standards and behaviour predictions should be used with the necessary caution.

The amount of detail, and the number of factors, determining the execution of maintenance measures increase according as the planning period is shorter. The content of the data acquisition survey, and the measuring and inspection methods employed, are more particularly adapted to these circumstances. It is explained how the collected data are used in planning the maintenance for the various terms envisaged.

In the first phase of the Working Group’s activities which have yielded the results reported here the emphasis has been on the development of a systematic procedure which is clearcut and suitable for practical application.

This paper therefore has a descriptive character. The results of visual inspections and of measurements are dealt with only in so far as they play a part in connection with the development of the above-mentioned systematic procedure.
05046 Programmable Calculators in the Assessment of Overlays and Maintenance Strategies
P. Ullidtz, K.R. Peattie

A technique for assessing the serviceability of an existing pavement and predicting its future life under specified traffic loading and seasonal conditions has been developed. It is based on a simplified form of multi-layer elastic theory which can be incorporated in programs for the larger programmable pocket calculators. This program will handle structures having up to four layers consisting of one bituminous layer, two unbound granular layers and a subgrade.

The properties of the material in the existing pavement are determined from non-destructive measurements of surface deflection using a Falling Weight Deflectometer or some similar instrument which measures deflection at the centre of the loaded area and at various distances from it. A method is given for converting the moduli of bituminous materials determined at one temperature and time of loading to values appropriate to other temperatures and times of loading.

The overlay program gives the residual life, in years, of each of the granular layers in a pavement and of the subgrade. It will also provide the residual life of each of these layers after the application of a bituminous overlay of stated thickness and modulus to the pavement structure. Different values of moduli can be assigned to the various layers according to the seasons of the year being considered.

A pavement management program has been developed to determine the effect of applying an overlay on the total life of the pavement or to compare the benefits of applying one of a number of alternative maintenance procedures such as overlaying and surface dressing. The extension of life resulting from the application of each maintenance procedure considered is determined and, if the costs of each are known, the ratio of functional benefit to cost is evaluated. The most effective maintenance strategy for a given road network within a stated cost budget can be found by applying the pavement management program to the individual road sections making up the complete network.
05047 ESSO Overlay Design System
S. de la Taille, P. Schneck, F. Boudeweel

The method of new pavement design developed at Esso has been adapted to provide an overlay design system. The basic idea is to modelize the old pavement in order to be able to calculate stresses and strains, considering the traffic and climate conditions. It is then possible to define design criteria, corresponding to the requirements of an overlay:
– the old pavement must be relieved.
– the overlay itself must work under acceptable conditions.

The modelization is achieved through the old pavement analysis, while the mechanical properties of the mixes considered are assessed : dynamic modulus, resistances to permanent deformations and to fatigue. The minimum thickness of the overlay is computed for each mix considered in order to match the design criteria. A selection between the various solutions is made from technical and economical considerations, to find out the most suitable one for the work.

An example of application of the system is presented. The problem was to find out a suitable mix and to compute the required thickness for an overlay of a ring road in Northern Ireland. Ruts and cracks in the temporary wearing surface made it compulsory to overlay the structure. The analysis of the old pavement (visual inspection, borings and deflection measurements) showed that it was very weak. Analysis of deflection values permitted a modelization, and the obtention of design criteria: reduction of horizontal tensile strain at the bottom of the old pavement to 1.2 x 10^-5, overlay fatigue index of 120 x 10^-3, and acceptable resistance to permanent deformations of the overlay mix.

Six mixes altogether were envisioned for the study: 2 Hot Rolled Asphalts, 2 Asphalt Concretes and 2 Dense Bitumen Macadams for Wearing Courses. In a first step, a single course overlay was considered, and the minimum thickness for each mix computed. Because of the high values obtained, a two-course overlay was then envisaged: a wearing course made of 40 mn of Hot Rolled Asphalt, and a base course to be determined. The minimum thickness of the base course for each mix was computed to match the design criteria, which led, together with considerations on the resistance to permanent deformations of the mixes, to selecting the Asphalt Concrete with the lower binder content in a 100 mn layer. This solution was retained and the pavement overlaid without any particular problem. A deflection follow-up was carried out: the overlay considerably reduced the deflection values, down to the level computed with the Esso Overlay Design System, pointing out the validity of the procedure.
05048 Pavement Evaluation and Overlay Design – Practical Method of Belgian Road Research Center
V. Veverka, J. Romain

This paper presents a practical method of pavement evaluation and overlay design. The proposed method results from field observations and assessments of a large number of flexible roads, from a study of statistical distributions of the parameters influencing the life expectancy of a bituminous pavement (deflection, load characteristics, temperature and thickness of the pavement, composition and mechanical properties of bituminous materials) and from an application of layered elastic theory. The paper presents the description of empirical and theoretical predictive models developed in the context of this study and aiming at the evaluation of the life expectancy and the distress evolution of the existing pavement. The proposed method includes:
a) pavement evaluation procedure
b) overlay design.

The pavement evaluation using empirical or theoretical predictive models is based on four parameters: cumulative number of heavy commercial vehicles (during the past service period), deflection, degree of cracking and repairs and rut depth.

The fatigue condition of the pavement is evaluated from deflection and cumulative number of heavy commercial vehicles. To evaluate the structural condition, the observed and predicted values of degree of cracking and rut depth are compared. The diagnosis is formulated and the remedies (further maintenance, strengthening, reconstruction) are chosen according to structural and fatigue conditions.

The overlay design uses:
a) design charts (current design method)
b) computer programs (full design method).
The overlay design charts have been developed by application of layered elastic theory. Several examples illustrate the practical implementation of the proposed method.
05049 The TRRL Method for Planning and Design of Structural Maintenance
N.W. Lister, C.K. Kennedy, B.W. Ferne

Effective use of the large sums of money now being spent world wide on strengthening roads requires a design system capable of matching spending to needs; it should establish priorities for work and also the nature and extent of strengthening on the roads selected for treatment. Economical solutions can generally be obtained if a relatively thick overlay is applied before the structural integrity of the road pavement is seriously impaired, when the road is in a critical and not a failed condition. The Transport and Road Research Laboratory has therefore developed a method for the planning of structural strengthening that can predict the remaining life of the pavement so that the strengthening can be timed to coincide with the onset of critical conditions; the method can also design the necessary thickness of overlay required to extend the life of a road to carry the predicted future traffic.

Papers to previous Conferences have described earlier stages in the development of the method from a systematic study of the relation between the deflection under a loaded wheel and the performance of pavements in the Laboratory’s many full-scale road experiments; the work has been supported by detailed observations of the deflection behaviour of pilot-scale pavements and by analytical techniques. This paper outlines the method concentrating on the latest work to complete its development, evidence of its validity and examples of its use.

Extensive experimental data is presented that has been used to develop charts relating deflection to pavement temperature so that deflections measured over a range of temperatures on different pavements can be corrected to equivalent design deflections at a standard temperature of 20°C. The development of pavement deterioration and its relation to deflection levels is illustrated, in particular the increasing uncertainty as pavement failure approaches, uncertainty that may be avoided by suitable definition of critical conditions and timely maintenance intervention. Typical evidence is presented defining experimental relations between deflection and critical life. The information has been consolidated into design charts for the prediction of remaining pavement life to the onset of critical conditions; separate design charts have been produced for the four main types of road base used in the United Kingdom. Measurements on a number of in-service pavements demonstrate the validity of the charts over a wide range of traffic. Observations of the reduction in deflection achieved by different types of overlay material are presented. These results have been combined with information on the performance of overlaid pavements to produce overlay design charts. Levels of probability of achieving a particular residual life or, in the case of an overlay, an extension of life are defined on the design charts.

Maintenance decisions based primarily on deflection but supported by other information have been made on many roads in the United Kingdom over a 20 year period. Comparisons are made of the cost-effectiveness of the deflection approach with decisions based on engineering judgement alone. Four major maintenance schemes on Motorways and Trunk Roads, and also 9 smaller projects under the control of one maintenance unit are considered. The use of deflection to determine the depth of deterioration in a Motorway pavement and the design of its reconstruction is also briefly described. Finally the paper indicates the present use of the deflection method in the United Kingdom for planning of maintenance priorities and for detailed design of strengthening measures.
05050 A Study of Resiliency for Pavement Design in Brazil
J. Medina, E.S. Preussler, S. Pinto, L. Motta

Resiliency of pavement materials and subgrade soils as determined through repeated load tests is studied for pavement design of Brazilian Highways. A research work conducted by the pavement mechanics group at the Federal University of Rio de Janeiro with the support of the Brazil Federal Department of Highways is underway since 1978. Satisfactory comparison has been obtained between deflections measured and those computed using the FEPAVE program and data from repeated load tests. The large deflections producing premature cracking and rapid deterioration of asphaltic surface layers are due usually to the low resilient modulus of granular layers submitted to low confining pressures and not to resiliency of subgrade soils. Moisture contents of subgrade soils in the tropical region analyzed are quite low: at optimum of standard Proctor test or below it. Also, many soils of lateritic nature exhibit good behavior as subgrade and pavement materials with respect to resiliency.

Fatigue characteristics of asphaltic mixtures determined at different temperatures are used to establish an approach to overlay design.
05051 Cracking in Wearing Courses
J.P. Marchand, H. Goacolou

The use of fracture mechanics in the study of pavements is a further step towards a better knowledge of the mechanisms, and thus of the influences of the parameters, governing the behaviour of a pavement.

The model presented in this article forms a part of this approach, since it takes into account the vertical discontinuities created by the transverse shrinkage cracks of courses treated with hydraulic binders. Its two-dimensional character is well suited to the study of pavements and the finite element method provides a good approximation to the modelling of the crack. The examples given show that it is possible to determine the directions of propagation of an existing crack for each type of loading encountered in pavement mechanics. It may be matched to a multi-layer two-dimensional model for use in interpreting the results of testing, to detect the presence of a vertical crack in the treated road base before it has begun its propagation into the coated material.

The presence of such a crack is reflected by a substantial increase in deflection and a decrease in the radius of curvature by comparison with measurements made on a crack-free pavement.

The wearing course design method proposed takes into account the influence of cracks ; it is presented in a branching form that schematically represents the various stages of propagation of the crack. The model may be used to give an approximate determination of the time taken by cracks to spread upward into the wearing courses.

This design method yields a better qualitative understanding of cracking phenomena in coated materials.
05052 Probabilistic Fatigue Design for Flexible Pavements
K.R. George, S.K. Nair

A general framework for the probabilistic prediction of load associated fatigue distress mode is presented. Using this framework, a prediction model is developed for fatigue life treating-traffic, material properties, and environmental effects as stochastic variables. The main steps in developing the model are:
(1) solution for structural response (using CHEV5L program) in terms of tensile strain at the-bottom fibers of base,
(2) prediction of fatigue life from structural response using an empirical relationship proposed by Hwang and Witczak [1979],
(3) development of regression model relating fatigue life to such field variables as load, ambient air temperature, subgrade support value, and pavement geometry,
(4) prediction of cumulative fatigue damage employing Miner’s rule, and
(5) development of design equation specifying reliability.

By virtue of the fact that the parameters affecting fatigue distress mode are random variables, cumulative distress, delta, is a random variable. The critical damage at failure, denoted as D, is not always close to one, as implied in Miner’s law, but in fact assumes a wide distribution: therefore, D is a random variable also. As both delta and D are random variables, fatigue life can be estimated with only a certain reliability (confidence level). Reliability, R, is the probability that serviceability will be maintained at adequate levels, from a user’s point of view, throughout the design life of the facility. An expression relating mean value and coefficient of variation of delta and R is proposed, which, by virtue of its deterministic format, can be used in design.

Pavement design using the above expression will require distribution functions (or frequency distributions) for wheel load, ambient air temperature and subgrade support value of the region under consideration; typical distributions of these random variables are presented. Respectively, load, air temperature, and subgrade support value conform to shifted exponential, Weibull, and lognormal distributions.

To assess the reasonableness of the proposed procedure, structural number selected for a specific traffic loading, environmental, and subgrade conditions has been compared with that of the conventional AASHTO Interim Guide design procedure and also with that of the zero-maintenance design procedure.
05053 The Interpretation of Repeared Load Parameters for a Glacial Subgrade from Soil Properties
R.W. Kirwan, E.R. Farrell, D.N.D. Hartford, T.L.L. Orr

This paper presents the results of an investigation to establish simple methods for estimating the resilient and permanent deformation parameters of a till subgrade from standard laboratory tests.

The subgrade deformation parameters necessary for the satisfactory design of a flexible pavement depend on the water content and dry density of the subgrade soil and on the applied stresses and other factors. A programme of repeated load triaxial tests has been carried out to investigate the influence of these variables on the required parameters such as resilient modulus, Poisson’s ratio and creep compliance. The experimental data from the repeated loading tests has been used to prepare charts relating the resilient modulus and creep compliance to the dry density and water content. These charts show that the creep compliance is very sensitive to change in water content and, except for samples with high dry densities, the resilient modulus is more sensitive to changes in water content than dry density and decreases as the water content increases.

Repeated load tests were also carried out in a laboratory pavement simulator. The deformation parameters interpreted from the pavement deflections measured in these tests are in close agreement with the values predicted by the charts from the repeated load triaxial testing programme for the in-situ dry density and water content.

The effect of variations in the modulus of resilience and creep compliance of the subgrade on the performance and design life of a pavement was investigated. This was carried out by examining the experimental results and the results of a parametric analysis using a finite element program. The results of the parametric analysis show that changes in the subgrade properties have a greater effect on rutting of the pavement than on the fatigue life of the asphalt layer.
05054 Field Validation of an Overlay Design Procedure to Prevent Reflection Cracking
B.F. McCullough, S.B. Seeds

A computer model was developed in 1977 under the sponsorship of the FHWA for analysis of reflection cracking in overlays on rigid pavements. This paper provides a description of the program (RFLCR) and a brief discussion on the theory behind its development. More importantly, the paper discusses the sensitivity of the program to certain input parameters and provides some basis (in the performance of overlays in Texas) for verification and acceptance of the procedure. The model will, however, continue to be examined, improved and calibrated for use in Arkansas as part of a joint project with the University of Arkansas, the Arkansas State Highway and Transportation Department.
05055 Structural Finite Element Design of Unbound Material Pavements from Cyclic Loading Triaxial Tests
J.L. Paute, J. Martinez

In this paper we intend to expose the application of a structural design method for unbound granular layered pavements founded on the direct determination of materials and soils behaviour parameters by cyclic triaxial loading, and on the utilization of obtained experimental results in structural design by means of a finite element computer code.

Materials experimental study is achieved in triaxial cells, with a completely pneumatic loading system, allowing the cyclic application of axial stress and cell pressure, as well as sample axial and radial strains measurement. In this way, it is possible to characterize the behaviour of unbound materials and soils under traffic, as well in terms of resilient response as permanent deformation at large number of cycles. Sample dimensions (160 mm diameter by 320 mm height) authorizes the study of materials with 0/20 mm particle size distribution.

Experimental results, obtained on unbound crushed stone and soils, have been analysed in terns of resilient and permanent deformations moduli. The influence of applied stresses on the elastic moduli values shows a great dependence between shear modulus and spheric pressure and between bulk modulus and deviatoric ratio. Presented in terms of permanent deformation at a given number of cycles versus applied stresses results allow drawing stress-strain equivalent diagrams. They give permanent deformation moduli values directly utilizable in structural irreversible deformations calculus at a given cumulated traffic.

Taking account of different obtained experimental results, structural design is carried out with a finite element computer code. In this way, it is possible to integrate non-linear elastic behaviour and applied stress versus cumulated strain parameters. Calculations have been achieved for various structures types with different geometrical and mechanical properties.

In conclusion to our work and as shown by other researchers, cyclic triaxial testing appears to be an efficient way to characterize unbound granular materials and soils for pavement design. The resulting experimental data can be directly used in calculus for determinating structural resilient response and permanent deformation. Design based on such methods should lead to a better structural optimization and it’s of interest to develop and generalize it.
05056 Characterization of Freeze/Thaw Affected Granular Soils for Pavement Evaluation
T.C. Johnson, D.M. Cole, L.H. Irwin

As part of a continuing investigation of techniques for predicting the effects of frost action on pavements, we have conducted field and laboratory tests to determine the nonlinear resilient modulus of granular soils affected by freezing and thawing. Three sands with varying amounts of silt were subjected to repeated-load triaxial compression tests in the frozen state, immediately after thawing, and at various stages of recovery as moisture tension increased during desaturation. By means of regression analyses the resilient moduli were expressed as functions of the significant parameters. In the frozen state these included temperature or ice content, total moisture content, and either the second stress invariant divided by the octahedral shear stress, J sub 2 / tau sub oct, or the octahedral shear stress alone. In the thawed and recovering states the significant variables were the moisture tension, the dry density and either of two stress parameters, J sub 2 / tau sub oct or the first stress invariant J sub 1.

To verify the laboratory characterizations, in-situ dynamic plate bearing tests were conducted on three field test sections, each consisting of about 50 mm of asphalt concrete and 1500 mm of one of the three soils tested in the laboratory, underlain by a gravelly sand subgrade soil. Repeated plate-bearing (RPB) tests were conducted 13 times between October 1978 and September 1979, and falling weight deflectometer (FWD) tests were performed 9 times from February through April 1980. Both types of tests were conducted at two load levels, in the range from 200 to 600 kPa plate pressure, and the testing dates encompassed changes in seasons and in conditions in the test soils ranging from frozen to thawing, recovering and completely recovered. The surface deflection basins were measured in each test, and prevailing subsurface temperatures and moisture tensions were monitored.

The sequence and thickness of the layers prevailing within the test soil during each test were defined by temperatures, frozen or thawed state, and moisture tension level. The layer thicknesses, plate loads, and nonlinear materials characterizations for each layer as determined from the laboratory tests, were entered into an elastic layered system program termed NELAPAV, to calculate stresses, strains and deflections. By an iterative process the program calculates moduli and stresses that are mutually consistent.

The calculated deflections agree reasonably well with the deflections measured during in situ load testing. The calculated resilient moduli show the expected seasonal variation, with extremely high values in the frozen condition, decreasing dramatically upon thawing, and increasing somewhat during late spring, summer and fall. From these results it was concluded that laboratory repeated-load triaxial tests performed at successive stages of the freeze/ thaw/recovery cycle can accurately characterize the nonlinear, stress-dependent moduli representative of each stage. Soil moisture tension measured in situ through cycles of thawing and recovery can provide the basis for assessment of the timing of the thawing/recovery cycle and of the laboratory determined modulus representative of each stage.

It was also found that NELAPAV could be used to calculate moduli from a measured deflection basin, by a process of successive trials. This method is considered less accurate than the method employing both laboratory and field tests, but its use is appropriate for pavement evaluations in cases where non-destructive testing is to be used exclusively, rather than combined with sampling and laboratory testing.
05057 Analysis and Design of Pavement to Resist Thermal Cracking Using Fracture Mechanics
R.L. Lytton, U. Shanmugham

A fracture mechanics based model has been developed that predicts the occurrence of transverse cracking in asphalt concrete pavements. Cracks are assumed to begin at the surface of the pavement and propagate downward as temperature cycling occurs. The rate of crack growth is based upon the Paris and Erdogan equation from which the number of temperature cycles required to crack a pavement can be calculated. The summation of the reciprocals of the cycles to failure was used to define a cumulative damage function, which is used to calculate observed crack frequency with acceptably small standard errors.

The mechanistic model calculates the change in stress intensity factor during each day due to the temperature cycle. Stress intensity factors are calculated from regression equations generated from a multi-factorial experiment using a finite element model of a multi-layered pavement structure. Fracture parameters are derived from readily available rheological properties of the bitumen.

Data used to validate the model were collected in Michigan. An empirical equation with very strong statistics was developed from the Michigan data and may be used to predict crack spacing. The empirical equation is limited in application to subgrade conditions and mix properties typical of Michigan pavements. The mechanistic model is sensitive to both subgrade and mix variables and thus was used to develop an empirical equation useful in the design of pavements to resist temperature cracking.

The mechanistic model is capable of providing more insight into the causes of low temperature cracking, and suggests other areas where the model can be improved profitably including the healing of asphalt concrete pavements due to residual stress, combining several modes of crack propagation, and determining fracture properties of aged mixes.
05058 Laboratory and Field Fatigue Characterization for Sulphur Extended Asphalt Paving Mixtures
J.P. Mahoney, R.L. Terrel

Data from experimental pavements recently constructed and tested at the Washington State University Test Track are analyzed to determine the fatigue relationships for conventional asphalt mixtures. The laboratory mix designs, test track construction and instrumentation are discussed. Additionally, laboratory fatigue parameters are derived for the same three mixtures by using sawed beams from the experimental pavements and tested in a rolling wheel fatigue apparatus. The fatigue analysis reveals that the conventional dense graded asphalt concrete has a flatter fatigue curve when compared to the two mixtures which contained added sulphur (30/70 and 40/60). This can be a desirable characteristic at high levels of repetitions to failure (low bending strains). The sulphur extended asphalt mixtures are able to accommodate higher bending strains at lower levels of repetitions to failure. The differences in the fatigue relationships for both the conventional asphalt concrete and the sulphur extended asphalt mixtures as determined by the fatigue parameters (K sub 1 and K sub 2) are small. The need for durability evaluation of SEA paving mixtures is discussed.
05059 Recycling Asphalt Concrete: Guidelines and Performance Potential
D.N. Little, J.A. Epps, R.J. Holmgreen

Recycling asphalt concrete pavement materials presents a very attractive alternative to pavement rehabilitation which must be considered. Recycling may conserve not only appreciable quantities of asphalt cement but also aggregates, providing not only economic but also environmental advantages.

The Guidelines for Recycling Pavement Materials (NCHRP Report 224) is summarized in this paper. These guidelines provide the tools to the pavement engineer for equitable consideration of recycling alternatives.

Recycled materials have proved comparable and, at times, even superior to conventional asphalt concrete based on both laboratory and in situ testing. Results of in situ pavement studies using the Dynaflect on twenty-five recycled pavements in eleven states are presented.

Laboratory characterization of recycled mixture properties of creep compliance, time-temperature shift, permanent deformation, thermal cracking and fatigue are presented. These properties were input into the VESYS IIM structural subsystem to predict performance of the recycled materials used in pavement systems in a variety of climates. Predicted performance of the recycled materials compared favorably with that of a dense graded, high quality asphalt concrete used as a control or comparison mixture.

Recycled asphalt concrete mixtures can be designed to provide acceptable structural properties and acceptable resistance to low temperature induced stresses.
05060 A Summary of Sulphur-Asphalt Design Technology
G.J. Kennepohl, D.C. Bean, L.J. Miller, R.C.G. Haas

Sulphur-asphalt pavement design technology has undergone extensive development in North America and abroad since the early 1970’s. These pavements represent a viable alternative to conventional asphalt pavements and they offer potential economic and performance advantages. Extensive laboratory, field and analytical investigations have been conducted to verify the design technology.

This paper reports on the design technology for sulphur-asphalt pavements as developed by Gulf Canada Limited and cooperating organizations and individuals. Several major aspects of the technology are addressed, including materials production, properties and quality assurance, structural design and economic considerations. Examples of each aspect are presented.

Sulphur-asphalt mixes have been found to have a lower temperature susceptibility than conventional mixes and they therefore offer the designer increased flexibility with respect to low-temperature and permanent deformation requirements. In addition, increased fatigue life and/or thickness reductions can be achieved under certain conditions. Production, transport, placement and compaction of sulphur-asphalt mixes is done with conventional methods and equipment. It is essential to produce the sulphur-asphalt binder by preblending the components, rather than by separately introducing the components and "pugmill blending" them, for reasons of improved engineering properties plus quality assurance.

The paper provides a summary list of some of the field trials conducted in North America, Europe and the Mid East. Observations to date have verified the laboratory and theoretically based design results related to permanent deformation and low-temperature cracking, as well as the upgrading possible with low-quality aggregates. Thickness reduction and/or fatigue resistance advantages remain to be verified.

Finally, the paper briefly reviews the potential economic advantages of using sulphur-asphalt. These relate to materials cost savings, conservation of materials and potential performance advantages.
05061 Characterization and Design Analysis of Bituminous Pavements
S.K. Khanna, M.G. Arora, K.R.N.S. Setty

With rapid motorisation and industrialization in India, bituminous pavement surfacings are becoming more and more common in the country. However, the design of bituminous pavement layers is still based on ad hoc experience and empirical strength tests. This paper aims to develop analytical method of design for bituminous pavements based on more realistic strength characterisation tests simulating the stress environment in the pavement layer under actual traffic loads. Three types of bituminous pavement specifications commonly used in India viz. bituminous concrete, bitumen bound macadam and sheet asphalt as recommended by the Indian Roads Congress (IRC) have been included in the study. Water bound macadam comprising crushed stone aggregate/overburnt brick aggregate was employed as a base course. Local Boorkee-soil (I.S. classification –SM) represented the subgrade soil.

Stiffness moduli of pavement component materials were determined by the conventional triaxial tests. In certain cases where horizontal stresses developed at the bottom of the bituminous layers under traffic loads are tensile in nature, the stiffness modulus may be different from the case where the horizontal radial stresses are in compression. In order to simulate the radial stresses in tension, thick hollow cylindrical triaxial specimens were cast and tested under axial compression. Radial tension was simulated by inducing water pressure inside the central hole. Analysis of results of all pavement materials tested indicates nonlinear relationship between strength modulus and applied stress. Since stiffness is also dependant on duration of loading, creep tests in tension as well as compression have been carried out. The creep compliance is also dependant on stress level as well as duration of loading.

Semi-full scale pavement sections were installed in the pavement testing laboratory employing the local soil as subgrade and the WBM and various bituminous specifications as the base and surface courses. Over each combination of base course/subgrade, three different thicknesses of bituminous pavement layers of each specification were installed and the load-deformation and pressure transmission characteristics of the pavement system were observed under static plate loads simulating the traffic loads. The results of semi-full scale testing exhibit supremacy of bituminous concrete layer over bitumen bound macadam and sheet asphalt in terms of its greater resistance to surface deformation and better pressure transmission characteristics.

Axisymmetric finite element model has been developed for a multi-layer pavement system which is based on the non-linear elastic stress-strain behaviour of the pavement component materials under simulated traffic loads. The stiffness of pavement materials as determined from laboratory characterization tests are used as design input into the analytical model to predict the load response characteristics of the semi-full scale pavement sections. The developed analytical model exhibits closer agreement with the experimental data than the conventional linear elastic model. It is proposed to apply this model for structural evaluation of existing weak flexible pavements in India and evolve design charts for Benkelman Beam Deflection method of design of bituminous overlays under the Research Scheme R-6 sponsored by the Ministry of Shipping and Transport (Roads Wing), Government of India.
05062 Volume II of 5th Conference Proceedings – preliminary pages and Table of Contents

Sponsoring Organizations
Conference Committees
General Program
Opening Session Addresses
05063 Moderator’s report and discussion of Session I papers
C. L. Monismith, M. W. Witczak

Moderators: C. L. Monismith, Professor of Civil Engineering, University of California, Berkeley, USA
M. W. Witczak, Professor of Civil Engineering, University of Maryland. College Park, USA

In this report, the moderators will attempt to use the papers which have been presented to the conference (and included in this session) to develop some general recommendations about the design process within the framework of Figure 1. We believe that the information contained in the papers provides a basis for any organization to develop its own design procedures for asphalt-type pavements. Moreover, as long as good construction procedures are followed in building the resultant designs, we believe that this methodology should provide a high level of confidence in the performance of the pavements. Should proper construction procedures not be followed, it permits a reasonable estimate to be made of the shortfall in performance, e.g., as measured by traffic transported, and to take appropriate action.
05064 Moderator’s report and discussion of Session II papers
J. Bonnot, J. Remain, J. Verstraeten

Moderators: J. Bonnot, Technical Director, Laboratoire Central Des Ponts et Chaussees, France
J. Remain, Scientific Advisor, Belgian Road Research Center
J. Verstraeten, Deputy Director, Belgian Road Research Center

In this session, the emphasis is placed on the "Verification" of design methods and models; that is, the comparison of predicted and observed performances. Nine papers have been assigned to this Session, coming from six countries: 3 papers from the United States, 2 papers from the Netherlands, and one paper each from Kenya, South Africa, Switzerland and the United Kingdom. Papers have been offered that deal with such comparisons in the case of overall empirical methods and in the case of the elements involved in theoretical or semi-theoretical methods, namely:
– stresses and strains
– deflections
– rutting or permanent deformation
– cracking
– PSI as an overall performance index

Further, comparisons between predictions from different design methods have been made. These items will be reviewed in succession. We shall then attempt to place the different contributions in perspective and to synthesize the state-of-the-art in the matter of verification.
05065 Moderator’s report and discussion of Session III papers
A.A.A. Molenar, R.C. Koole

A.A.A. Molenar, Delft University of Technology,
R.C. Koole, Shell Research, Amsterdam, The Netherlands

This section on pavement evaluation is scheduled between the session on pavement design and the verification of design sessions we had yesterday and this morning, and the sessions on pavement management and pavement rehabilitation tomorrow.

Your moderators feel that indeed this is the right time to discuss pavement evaluation aspects, because after we have built our pavements in Session I and II and before we will discuss how we can get the most out of our maintenance and rehabilitation budget, we should discuss techniques and methodologies by which we are able to determine where, when and which maintenance activities should be applied in order to restore or to keep the condition of the pavement system at a desired level of service. Your moderators like to describe evaluation of pavements as: the process of measuring the structural and functional condition of the pavement, followed by verifying whether the condition is above or below predefined minimum acceptance levels and warning levels.

Furthermore, one should estimate the length of the time period that will elapse between exceeding the warning level and reaching the threshold (i.e. the minimum acceptance level), since this enables scheduling in time of maintenance activities (Figure 1). Also, a first estimate should be made of the nature of the required maintenance activity in order to be able to plan the maintenance budget over a certain time period. In order to be able to make such projections, regular check-ups of the pavement condition are necessary (1, 2, 3). This definition of pavement evaluation given by your moderators is wider than what is meant by evaluation in the opinion of others. They define evaluation as measuring properties of pavement layers, like for instance elastic moduli.

For the time being your moderators like to stick with their cinemascopic definition since we feel these limited definitions do incorporate the danger that taking measurements or calculating elastic moduli are becoming a goal themselves. This of course may never be the case. We will come back to this in detail in the second part of this session. Also, the wide definition of evaluation given by your moderators stresses the axiom that the strength of the chain is determined by its weakest link. By this we mea" that measuring, data processing and overlay design techniques should be equally well developed. For instance, it makes no sense to spend a lot of money to obtain deflection equipment which is capable of characterizing very accurately the deflection bowl if an overlay design technique is used which utilizes only the maximum deflection as input. However since we know different opinions exist on the meaning of the word evaluation, we would encourage the authors to comment on the definition given here by your moderators, and to give their views on what the evaluation process should be.

Having some well known definition makers in our session like Dr. Hudson and Dr. Finn
we are sure we will succeed in this.
05066 Moderator’s report and discussion of Session IV papers
R. Haas, W.R. Hudson

Ralph Haas, Professor of Civil Engineering, University of Waterloo, Canada
W. R. Hudson, Professor of Transportation Engineering, The University of Texas at Austin, U.S.A.

This Session deals with Pavement Management and Pavement Management Systems. It is our honor and our responsibility to summarize the papers selected for this session. But to simply summarize the papers is not enough; rather, it seems necessary to provide a background and an organized framework for comparing and putting into perspective the contributions made by the work described in these papers
05067 Moderator’s report and discussion of Session V papers
Jon A. Epps, R. G. Hicks

Jon A. Epps, Professor of Civil Engineering, Texas A&M University, U.S.A.
R. G. Hicks, Professor of Civil Engineering, Oregon State University, U.S.A.

Expenditures for streets and highways are a large part of today’s transportation budgets. A sizeable portion of these budgets is disbursed for pavements with an increasing proportion of this amount devoted to preserving the existing pavement network. In spite of the increases in expenditures for maintaining the existing systems, estimates indicate that funds will not be sufficient to keep streets and highways at an acceptable level of serviceability. This, together with current inflationary trends and the necessity for energy conservation, requires the efficient management of our local street and highway systems.

The purpose of this report is to summarize the recent developments pertaining to the maintenance and rehabilitation of highways. Included will be a general discussion of maintenance and rehabilitation strategies available to engineers, methods available to optimize the selection of the best strategy for a given situation (pavement condition, traffic volume, etc.), and finally, methods for determining overlay thicknesses and/or predicting remaining life of existing pavements. The discussion is based not only on the review of the ten papers accepted for this session, but also the critical review of other recent publications.

In this report pavement maintenance is defined as those processes, both preventative and corrective, which do not involve major alterations in the existing pavement structure. Rehabilitation includes reconstruction, overlays and recycling, and their combinations to restore or improve the serviceability and/or structural capabilities of the pavement. A general framework for such activities is shown in Figure 1. The sections which follow include: (1) discussion of these alternatives and methods used to optimize their selection, and (2) where overlays are required, descriptions of methods for computing the thickness to ensure adequate service.

Organization of Session
In carrying out their task, your moderators have tried to strike a balance between the information presented in the ten papers and information presented elsewhere. The intention of this approach is to stimulate discussion and to allow the authors to add finishing touches to work completed over one year ago. Professor Epps will summarize the papers dealing with maintenance and rehabilitation strategies. Professor Hicks will summarise those papers which consider overlay design. Authors will then have an opportunity to rebut our remarks. This will be followed by discussion (both prepared and from the floor). Dr. Epps will conclude with the summary of the significant findings and recommendations resulting from this session.
05068 Moderator’s report and discussion of Evening Session
F.N. Finn

Moderator: F.N. Finn, Consulting Engineer, U.S.A.

This session is the evening meeting; it is not quite as well organized as the formal sessions are; it is an informal session in that regard and I want to ask you to bear with US a bit. We have an opportunity for some very lively discussions and we have some questions. We also will have some discussion this evening for all of you to participate in and I want you to feel free to get up and make remarks, ask questions, put people on the spot.

I wanted to make just a few opening remarks to lay a bit of ground work on Pave- ment Management Systems. I know you are being inundated with an awful lot of information and at times that can be and will be confusing to US. Sometimes when we talk to each other about pavement management systems we are very sincere and we try to be as factual about it as possible, convinced that what we have to say is in fact the latest word in technology. But sometimes when we begin to talk to administrators, policy-makers and decision-makers and sometimes when we talk to, what I call, the traditional highway engineers, it looks to them as though it is all fantasy. So then the question that is posed to those who are here this evening: is it fact or fantasy?
05069 Moderator’s report and discussion of Session VI papers
P. S. Pell, S. P. Brown, C. K. Kennedy

P. S. Pell, Professor of Civil Engineering, University of Nottingham, United Kingdom
S. P. Brown, Reader in Civil Engineering, University of Nottingham, United Kingdom
C. K. Kennedy, Head of Department of Civil Engineering, Plymouth Polytechnic, United Kingdom

The properties and characterization of materials relevant to performance is an essential and inherent part of the structural design process for asphalt pavements. This topic, therefore, has been dealt with generally in many of the papers presented in other sessions and this particularly applies to asphaltic materials which, of course, constitute the main structural layers in asphalt pavements. However, the purpose of this session Is to present papers whose primary concern is with certain types of material and their behavior under specific conditions rather than overall pavement design methods. In all cases the material response and performance has to be considered in the context of the role they play in the structure of the complete pavement.

The eleven papers assigned to this session and critically summarized below by the Moderators, cover seven topics. Tne characterization of soils and granular sub-base material, both for analysis and performance, is required input data for all structural pavement design methods and is often treated with considerable simplification. In the view of the Moderators, this is an area which warrants considerably more attention and becomes increasingly more important as the asphalt layers become thinner and a number of papers are relevant to these unbound materials. Improvement in the properties of asphalt mixes by the use of various additives is an ongoing development of growing interest for reasons of economy and the improved performance required under present-day heavy traffic loadings. This is reflected by two papers on sulphur-asphalt. The current concern over the best and most economic use of natural resources has led to the development and use of recycled asphaltic material and this is the subject of one paper.

Four papers are concerned with the mechanisms and design sub-systems related to various types of cracking of asphalt layers, namely, reflection cracking, thermal cracking and a probabilistic approach to fatigue cracking. Each of these deals with a difficult aspect of performance criteria which are still in an early stage of development and not yet incorporated in complete pavement design methods. Tne final paper reports the development of a design procedure for Indian conditions based on characterization testing of materials commonly used in that country.
05070 Moderator’s report and discussion of Session VII papers
J. F. Shook, W. Visser

J. F. Shook, Principal Engineer, The Asphalt Institute, USA
W. Visser, General Manager of Pavement Consultancy Service, The Netherlands

The two moderators of Session VII were asked by the conference Technical Advisory Committee and the International Executive Committee to summarize the results of the proceeding six sessions, and to report, from their respective points of view, the outlook for the near future, particularly with reference to the outlook for implementation and to research needs. As explained earlier by Mr. Visser, we have separated our report into two parts. I will review and report on Session I, Pavement Design Methods, Session II, Verification of Design Methods and Session VI, Material Properties. Mr. Visser will report on the remaining three sessions. We hope this procedure will not be too confusing and that we will not contradict one another to any great extent.

I am pleased, particularly to be able to include verification of design procedures in my portion of this report. Much effort has been put into the development of design procedures, both empirical procedures and procedures based on theoretical or semi-theoretical models, in the past twenty years. Somewhat less attention has been given to material testing I believe, and considerably less to field verification. Field verification is very difficult, and I am pleased to note that 24 of the 33 papers in Session I, II and VI make some reference to field verification, as do, also, many of the papers in Session III and V.

Now, for the rest of my presentation; first, I will summarize the more important points raised by the moderators in their respective reports, and then discuss the issues of implementation and research.
05071 Comments on the Next Conference (the 6th)

The moderators of Session VII asked the audience to make comments about the next (sixth) conference:
1. Should there be a sixth conference?
2. Dn we need a change in format of the conference?
3. Should the scope be changed to include other topics?
4. How can the conference help to encourage Implementation?

The participants contributed to the Discussion with written comments:
05072 Corrections to papers printed in Volume I

This file contains all the errata for the published papers. These corrections have been appended to the papers to which they are relevant.
05073 List of Registrants

The names, affiliations, and addresses of all 5th Conference participants.
05074 Rehabilitation Design Incorporating Past Experience of Pavement Behaviour
M. C. Grant, P. C. Curtayne

It is often found in practice that too narrow an approach is taken in the design of pavement rehabilitation. Selected rehabilitation procedures are used in which only a few of the important factors that should affect the design are considered. Furthermore, much of the information that could be obtained by evaluating the condition of the pavement and its past behaviour is not fully exploited. Although these deficiencies may be recognized it is often difficult to combine qualitative judgements with sophisticated numerical analysis.

This paper presents a broad framework for the evaluation of pavements. Emphasis is laid on the cause of distress in the choice of the type of rehabilitation and the use of various types of information that can be used in the design process. Because obtaining information is expensive, a selective and iterative procedure is used starting from simple observations and progressing to more sophisticated testing, if justified. Irrespective of the amount of testing, uncertainties regarding the best form of rehabilitation will remain and the final choice can be made with the help of Bayesian decision diagrams. Some simple tests are described that can provide information about the cause of distress and the structural capacity that can be used in this approach.
05075 Index of Contributors and Discussers

An alphabetical listing of contributors and discussors.