2nd Conference

This document contains the preliminary pages of the Proceedings for the 2nd International Conference. The Table of Contents gives a good overview of the structure of the conference and the papers presented (and discussion thereon).

The Measurement of Highway Pavement Performance
B. G. Hutchinson

Thickness Determination of Flexible Highway Pavements for Mixed Loads and Traffic Volume
M. Livneh and E. Shklarsky

Design of Subsurface and Surface Against Detrimental Compaction and Shear in the Finished Structure
J. L. McRae

Flexible Airport Pavement Design and Performance
G. Y. Sebastyan

A Method for Strengthening Flexible Pavements
J. Lassalle and S. Langumier

Developments in the Application in Practice of a Fundamental Procedure for the Design of Flexible Pavements
G. M. Dormon and J. M. Edwards

Prediction of Pavement Deflections from Laboratory Tests
C. L. Monismith, H. B. Seed, F. G. Mitry, and C. K. Chan

The Computation of Road Deflections under Impulsive Loads from the Results of Vibration Measurements
M. E. Szendrei and C. R. Freeme

Analysis of Deflection Data from the AASHO Test
J.M. Kirk

Consideration of Calculated Strains at Various Depths in Connection with the Stability of Asphalt Pavements
W. Heukelom and A. J. G. Klomp

The Behavior of Asphalt Pavements under Variable Repeated Loads
P. Ferrari

Evaluation of Applicability of AASHO Road Test Results to Corps of Engineers Flexible Pavement Design Criteria
W. J. Turnbull, R. G. Ahlvin and D. N. Brown

A number of attempts have been made over the years to formulate objective measures of pavement performance. In general all of these attempts measured some aspect of the pavement surface distortion in the form of an index of roughness. Hveem has made an excellent survey of the types of equipment used to measure or estimate pavement roughness and the techniques used to analyze and interpret the data.

While many miles of highway pavement have been surveyed with one or other of the instruments described by Hveem, no rational method of profile interpretation has been developed with the exception of the A.A.S.H.O. Road Test formulation reported by Carey and Irick.

This paper explores the formulation of apavement performance measure from a systems oriented approach. The driver functions are examined and existing knowledge on human response to motion is reviewed in order to establish the critical human parameters of the system. A tracking task is described and the results of a field experimental program are given.

One accepted method is that of the U.S. Corps of Engineers based on evaluation of the strength according to the CBR scale. Since this method of the U.S. Corps of Engineers is particularly suited to the calculation of pavement thickness for runways, there is special interest in the development of a method of extending this method for use in determining the thickness of highway pavements. Obviously such an extension should include highway pavement thickness calculation formulae equivalent to those of the U.S. Corps of Engineers as a special case while also providing an appropriate method of calculation for mixed traffic, which is the actual case on the roads. This paper presents the above mentioned extended formulae after first developing the settlement relationships which enables the required translation of the U.S. Corps of Engineers method to highway pavements. Additionally, the paper presents a method of calculating pavement thickness for mixed traffic loads also based on the above mentioned settlement relationships.

Illustrative examples are given in which the Gyratory Testing Machine (GTM) is used to evaluate the compaction and shear characteristics of the various structural elements, including an illustration of the use of the Gyratory Testing Machine (GTM) shear data in the original Prandtl Formula to predict the bearing resistance versus shear strain relationship for the surface course of a flexible pavement.

Discussion of environmental conditions has been purposely avoided in this paper in the interest of brevity. It is recognized that the test specimens should be subjected to environmental conditioning prior to testing. The nature of such conditioning will, of course, vary depending upon the anticipated conditions for any particular geography and weather.

Developments in flexible pavement design methods during the past few years tend to follow two general directions.

Empirical and semi-empirical flexible pavement thickness design analyses have given place to design methods developed on the basis of the rationalization of empirical knowledge by measuring, correlating and expressing design variables involved using statistical methods and analysis.

The AASHO Road Test and the work of the Canadian Good Roads Association (CGRA) Pavement Design and Evaluation Committee are two excellent examples of this approach.

The other direction in the development of flexible pavement design methods is the rational design approach based on determining and limiting the stresses or deflections generated in the subgrade and the pavement structure under the superimposed load as evaluated by the theory of elasticity.

As an introduction to the subject, some of the problems and difficulties encountered in Department of Transport’s design work which made use of the various empirical, semi-empirical and rational methods will be discussed in the light of the data and experience collected by the Department.

They have not dealt with the following points: Why use a strengthening overlay?
When this must be done?
because these points must take into account criteria which are outside the method itself.

This paper is divided into four parts:
– Principle of the method;
– Practical application of the method;
– Examination of four points of special importance; – Possibilities evolved in the method.

This load associated cracking of the pavement has been attributed to fatigue failure of the surfacing resulting from repeated traffic stresses over a period of time. The California Division of Highways and more recently several other agencies, have measured the deformation and rebound occurring innumerous pavement sections and have found a close correlation between the occurrence of cracking and the magnitude of the transient pavement deflections. However, as indicated by the California data it is necessary to consider not only the absolute value of deflection as a damage determinant but also the number of repetitions of load, since in a number of instances, cracking has only developed after large numbers of load repetitions producing potentially damaging deflection.

The prevalence of this form of pavement distress has created a need for methods of predicting the transient deformations of a pavement in order to assess the magnitude of the stresses and strains developed in the surfacing which in turn appear to determine the fatigue life of the pavement. In a general way, it is to be expected that stresses or strains in the asphalt surfacing will increase with the magnitude of pavement deflection, but they are also dependent on the curvature of the deflected surface. However, methods for predicting deflection would be a valuable first step in solving the problem of preventing fatigue failures in pavements.

In order to interpret field measurements a more practical approach has been followed by postulating a “mechanical model” with a number of frequency independent parameters, suggested by the shape of the experimental curves.

Although vibration tests are carried out using sinusoidal forces only, it is possible to analyse a force impulse, such as applied by a moving wheelload, by Fourier techniques and, by combining the relevant impedance value with each frequency component of the pulse, a method has been found by which the resultant road response to the force impulse may be estimated.

Measurements were taken with the equipment consisting of an inverted electromagnetic vibrator supported by a spring on a 1 ton cylinder. The cylinder makes contact with the road through a steel plate 30 cm in diameter. The applied sinusoidal vertical force is measured by a calibrated geophone fixed to the vibrator, while the resultant velocity is measured by a similar calibrated geophone located centrally near the top of the 1 ton mass. This method was first suggested by Thrower but the present equipment has been modified by the addition of the 1 ton mass to lower the natural resonance of the road-equipment system to more realistic frequencies.

The numerous results from the AASHO Road Test offers a good opportunity to check if this assumption of elastic behaviour of road materials can be used for the design of road structures, and the following pages contain an analysis of some deflection data from the AASHO Road Test.

Some attention is also given to a statistical treatment of wheel loads in Dutch practice and to the thickness design of entirely bituminous road structures intended to carry a given traffic on subgrades of variable CBR value.

However, even when all the above factors have been into account in the design, and the pavement has been properly constructed, it has been observed that, if traffic is heavy and intense, the pavement keeps well for a certain period of time, but subsequently cracks begin to appear on the carpet, at first of limited length and depth, which in time deepen and extend to large areas of the pavement, taking on the characteristic aspect of a net.

With the loss of its two important properties of monolithicity and impermeability, the efficiency of the pavement appears completely impaired.

Without leaving out of consideration the alterations brought about by ageing and weathering on the properties of the asphalt cement, the chief factor in causing the formation of cracks has been traced, in the last few years, to a “fatigue” phenomenon due to repeated loads; consequently, in the design of pavements, this phenomenon has been – in a wholly empirical way – taken into consideration, by increasing the thickness in direct proportion with the anticipated increase in repeated loads. In effect, since the true nature of fatigue phenomena in asphalt mixtures was unknown, it has been thus far impossible to evaluate their qualitative and quantitative importance for the purposes of resistance to repeated loads. This gap in basic knowledge takes on an even greater importance in the so-called “rational” methods for the calculation of pavements, whose basic hypothesis on the physical properties of asphalt mixtures do not take into account the element of material fatigue.

On the basis of some experiments conducted at the Experimental Laboratory of the Institute of Road, Railway, and Airport Constructions of the University of Naples on asphalt concrete samples submitted to repeated alternating bending stresses, this paper deals with a theory which should make possible to identify the factors which cause fatigue failure in the asphalt mixture, through the study of alterations in the stress distribution and structure of the material brought by repeated loads. It is thus possible to determine the pavement features which have a greater weight on said phenomenon, and thereby to influence same for the purpose of improving pavement behavior.

The present serviceability index method of evaluating pavement performance, developed by the AASHO Road Test staff, appears to have considerable merit in quantitatively assessing pavement condition. However, the specific variables for which objective measurements are taken for use in this method of evaluation are not those normally considered in pavement design.

The various approaches followed in attempting to relate the pattern of behavior represented by the AASHO Road Test results to the pattern inherent in the CE design procedures are explained and comparisons are shown. Mathematical patterns seem to be strongly parallel, but the specific field measurements needed to draw a direct comparison are lacking.

Analysis of Stresses and Displacements in a Three-Layered Viscoelastic System
J, E. Ashton and F. Moavenzadeh, Massachusetts Institute of Technology, Cambridge, Massachusetts

Dynamic Phenomena in Pavements Considered as Elastic Layered Structures
A. Avramesco, Laboratoire Central Des Ponts et Chaussees, Paris, France

Stresses and Displacements in Viscoelastic Layered Systems under Circular Loaded Areas
Y. H. Huang, Council of Highway Investigation and Research, Charlottesville, Virginia

The Theory of Viscoelastic Two-Layer Systems and the Conception of its Application to the Pavement Design
K. Ishihara, University of Tokyo, Tokyo, Japan
T. Kimura, Tokyo Institute of Technology, Tokyo, Japan

Stress-Strain Law for Viscoelastic Flexible Pavement under Temperature Variations
A. B. KU, University of Detroit, Detroit, Michigan

The Analysis of Pavements under Arbitrary Loading
A. L. Yettram, C. A. O’Flaherty and M. E. Fleming, The University of Leeds, Leeds, England

Deflection of Viscoelastic Medium Due to a Moving Load
W. H. Perloff, Purdue University, Lafayette, Indiana
F. Moavenzadeh, Massachusetts Institute of Technology, Cambridge, Massachusetts

Stresses and Displacements in Elastic Layered Systems
J. Verstraeten, Centre De Recherches Routieres, Brussels, Belgium

Stresses in Layered Systems under Static and Dynamic Loading
A. Waterhouse, Bradford Institute of Technology, Yorkshire, England

Stress and Displacement in an Elastic Mass under Semi-Ellipsoidal Loads
J. L. Sanborn, University of New Hampshire, Durham, New Hampshire
E. J. Yoder, Purdue University, Lafayette, Indiana

Predicting Performance of Bituminous Surfaced Pavements
R. D. Barksdale, Georgia Institute of Technology, Atlanta, Georgia
G. A. Leonards, Purdue Universiv, Lafayette, Indiana

The idea of calculations will be recalled, but the reader is strongly advised to refer to the noted work*, where are given the procedures of integration, and almost every bibliographic reference; also, the case of “torsion” may appear as of purely theoretical interest: but it allows to underline the essential features of the phenomena so simply that it has been thought necessary to include it in this paper and it is already studied experimentally in France for the testing of roads.

Formulas or equations have been written only either when they allow to conceive how the mathematical method works, or to precise a physical result. The reader interested in complete developments is again asked to refer to the noted paper* for calculations and references.

—————–

* A. Avramesco: Dynamique des structures elastiques stratifiees Annales des Ponts et Chaussees Janvier-Fevrier,1966.

Two methods, one based on a direct method of Laplace inversion and the other on an approximate method of collocation, were developed to determine numerically the stresses and displacements in two and four-layer systems made of linear viscoelastic materials.

Numerical solutions involving both simple and complicated cases are presented, and the significance of these solutions is discussed.

In this paper, we shall endeavor to propose more realistic stress-strain relationships and concern ourselves with the effect of temperature and viscoelasticity on these relationships.

Analytically, the method involves developing the variational approach to problems in elasticity to deal with the complete slab and foundation system.

Computationally, the use of polynomial expansions is proposed to describe the independent variable (the distribution of the loading) and also the dependent variables (the distributions of transverse shearing forces, bending moments and torque, rotations and deflections in the slab). This technique results in a set of linear simultaneous equations in the polynomial coefficients, and the solution of this set then enables the dependent variables to be calculated at any location, as required.

Recently it has been shown, however, that, for highway pavement materials, the assumption of time independent behavior is not even approximately correct (PISTER and MONISMITH, 1960: PAPAZIAN, 1961; HOUSEL, 1959: SCHIFFMAN, 1959). In order to more appropriately characterize the response of such materials, various time-dependent constitutive equations have been formulated. One of the simpler types of such equations describes the behavior of linear viscoelastic materials. This type of equation can be written in several forms, one of which is (LEE, 1960). In his equation (a sub r) and (b sub r) may be functions of time, space and even temperature. However, linearity requires that (a sub r) and (b sub r) be independent of stress and strain. Assumption of homogeneity and isotropy eliminates the dependence of (a sub r) and (b sub r) on location and orientation of stress within the medium, and greatly simplifies the analysis. It is also often desirable to express Lee’s equation as two equations in which dilational (volumetric) and deviatoric (shearing) effects are separated.

The derivations are made under the assumption that the superposition of the three above-mentioned stresses represents the action of a vehicle wheel on the upper surface of the surfacing course.

Starting from the derived theoretical formulas of the stresses, numerical calculations are performed in the particular case of four-layered systems with continuous interfaces. . ” These calculations permit to draw general conclusions relevant to pavement design.

1. Briefly to outline existing methods of the design of flexible pavements leading up to a brief historical introduction to the mathematical analysis of elastic layered systems. It presents a finite element approach to this problem and treats salient points involved in detail. Certain matrices are quoted in sufficient detail to enable the stiffness matrices for the elements to be calculated. No results are presented at this stage but work is proceeding to compare the finite element analysis first with the Boussinesq theory and second with the results presented by Peattie and Jones. The effect of rigid inclusions in the mass is being studied and it is hoped that figures will be available by August 1967.

2. To describe experimental work involved in the determination of axial stresses in two-layer and three-layer systems, vertically below the central axis of a circular load. The measurements also involve a continuous record of surface displacement and applied load and the determination of relative ‘E’ values. Loads are applied statically and dynamically.

The study was composed of three parts:
1) to write a program for a digital computer, which could be used directly or with minor modifications for any given load distribution,
2) to use that program to develop charts of vertical stress and displacement for a reasonable approximation of usual tire loads, and
3) to compare effects of uniform circular and non-uniform, non-circular loads.

Stresses were considered to be distributed according to Boussinesq theory which assumes a homogeneous isotropic elastic mass. No attempt is made to account for the effects of the layered structure of flexible pavements. Several investigators have described solutions to two and three layer problems. However, since a primary function of this program is to provide a basis for comparison of effects of non-uniform loads with the conventionally assumed uniform circular loads, the work was restricted to the simpler homogeneous case which has been the assumption adopted by other investigators in a large portion of the earlier work.

Thus, progress has been made toward the development of a suitable tool to predict pavement performance. However, changes in material properties under the action of repeated loadings accompanied by permanent deformations and a redistribution of stresses with time cannot be evaluated using elastic theory.

The purpose of this paper is to extend the basis for predicting pavement performance to include repeated loads and time dependent material properties that can be evaluated from laboratory tests. The analysis of pavements using elastic, multi-layered theory is also extended to include numerical solutions of four-layer pavement systems.

Field Application of the Resilience Design Procedure for Flexible Pavement
J. L. Beaton, E. Zube, and R. Forsyth, State of California Highway Transportation Agency, Sacramento, California

The Design and Evahration of Asphalt Concrete Pavements Using Continuous Waves
W. H. Cogill, University of New South Wales, Kensington, Australia

The Influence of Compaction Methods and Condition on the Structural Behavior of Compacted Subgrades
H. E. Wahls and L. J, Langfelder, North Carolina State University, Raleigh, North Carolina

Structural Effect of Restraint Layer on Subgrade of Low Bearing Capacity in Flexible Pavements
T. Yamanouchi, Kyushu University, Fukuoka, Japan

Experiences With the Use of Wedges Against Frost Heaving
0. A. Taivainen, University of Oulu, Oulu, Finland

Considerations on the Structural Number
H. Takeshita, Chuo University, Bunkyo-Ku, Tokyo, Japan

Measurement and Interpretation of the Elastic Compressibility of Subgrades and Its Relation to the Behavior of Asphalt Pavements
C. Ruiz, Graduate School, Highway Engineering, Buenos Aires, Argentina

Soil-Cement Properties Determined by Repeated Loading in Relation to Bases for Flexible Pavements
J. K. Mitchell, University of California, Berkeley, California
Chih-Kang Shen, Loyola University, Los Angeles, California

Structural Design of Asphalt Pavements With Local Materials for Heavy Loads
L. M. Zalazar, Graduate School, Highway Engineering, Buenos Aries, Argentina

Since 1960 the California Division of Highways has been utilizing pavement deflection measurements as a guide for the determination of the structural adequacy of existing pavements and the magnitude of necessary reconstruction. In order to introduce the deflection factor into flexible pavement design at the preliminary stage, a device known as the resiliometer has been developed by the Materials and Research Department. This instrument subjects a 4″ diameter by 4″ high soil specimen to cyclic dynamic loads with intensities varying from 0 to 50 lbs. per sq. in. In 1960 a program aimed at correlation of laboratory resilience measurements with field deflection data was initiated as the first step toward incorporation of the test soil resilience value into the flexible design procedure. Twenty-five sampling locations from California highways and the AASHO test road were included in this study, The results of What is considered to be a successful field correlation program were presented at the 1962 International Conference on Flexible Pavement Design.

This paper reports on subsequent studies initiated for the purpose of utilizing the pavement deflection-resilience correlation in practical design situations. It includes a description of the development of a method for selecting design moisture content and density of preliminary samples and establishment of a curing period between fabrication and testing of sensitive” soils. Upon establishment of these test criteria, 40 preliminary samples from 21 different projects were tested and analyzed utilizing design moisture content and density criteria. After completion of construction, the roadways subject to analyses were tested for transient deflection in order to determine the relationship between predicted and actual field deflection utilizing established design criteria. A coefficient of corelation of 0.86 with a confidence band ranging from 0.73 to 0.92 was obtained. Averaging the results of two or more samples from a single roadway increased the correlation coefficient to 0.93. Field trial data are presented, along with examples of the design procedure.

Asphaltic concrete pavements are layered structures, in which the highly stressed zone near the surface transmits and spreads the vehicle load to the lower layers. The wearing surface must offer resistance to abrasion. Also, it must possess elastic stiffness in order to enable the lower courses to be stressed less highly: this permits the use of material having poorer bearing qualities for the lower layers.

It has been shown by Burmister, and by Cumming & Gerrard that, if the elastic constants at various depths in such a structure are known, the stresses can be calculated in the layers under any given configuration of surface load. This paper is concerned with the determination and meaning of elastic constants in layered media of this kind.

It is the intent of this paper to review, on the basis of current knowledge, the validity of the preceding assumptions. Because the bulk of the available information regarding strength or strength indices for compacted solids is based on laboratory data rather than field data, the initial phase of paper will review the factors that have been shown to affect the shearing resistance of laboratory-compacted samples. Next the available data correlating laboratory and field behavior will be considered. Also, limited data on the influence of field environmental conditions and traffic on subgrade performance will be presented. Finally the implications of the data that is being reviewed to current laboratory testing procedures will be discussed.

As the materials for restraint layer in this study, the following items were adopted:
(1) Soil-cement method,
(2) Method by low pressure polyethylene net.

The investigated spots have been in frost-susceptible ground, the road-parts in question being situated on both sides of culverts. The following cases have been investigated:
– culverts without any transition wedges,
– culverts with short irregular wedges, filled with sand,
– – culverts with long regular sand wedges, and
– a wedge filled with insulating lightweight aggregate.

Observations have been made regarding the changes in the roads, and particularly with respect to the grades in the winter as well as in the summer time. Some special observations regarding the depth of freezing and the temperatures in the structures of the road have also been made, when sand and lightweight aggregate have been used as insulating material in the wedges.

The investigation has produced as a result a proposal for the shape and the dimensions of the frost displacement wedge for culverts, and as a result of this proposal certain viewpoints regarding its use have also been presented.

This paper attempts to show the considerations concerning flexible roads, as the following items:

1. The tentative relations between structural number, wheel load and CBR were obtained from the field investigations.

2. The coefficient of relative strength “a” seems to be correlative with the modulus of deformation of the material.

3. By the use of the results obtained, the cause of cracking of the asphalt pavement on the National Highway Route No. 17 was analysed.

Modern design methods directly or indirectly based on the elastic layers theory and on experience, tend to achieve that the stresses and strains due to the transient loads, anywhere in the structure, be compatible with the characteristic of the materials under any service condition, somehow or other bearing in mind the effect of fatigue. Whatever the point of view, we arrive at the conclusion that the elastic behaviour of a structure under loads, essentially depends upon the harmony between the critical tensile stress in the lower zone of the upper asphaltic layers and the other critical vertical stress at the level of the inferior bearing layers.

The elastic behaviour of the structure requires that in the inferior noncemented layers, and especially in the subgrade, the vertical stress that reaches them does not cause shearing stresses, which surpass the strength of the material controlled by the lateral support. It should be remembered that in dealing with compressible materials, the vertical reaction related to the recoverable reduction of the volume, can fundamentally control the elastic deflections of a structure, and, consequently, control the development of the fatigue of the asphaltic layers responsible for its cracking.

Flexible pavement sections containing soil-cement base courses are analyzed using allowable stresses and strains and modulus values determined from the repeated loading tests. Combinations of asphalt concrete and soil-cement base thickness are found which, for a range of subgrade strengths, limit critical stresses and strains in the pavement sections to acceptable values. The conditions examined are 1) tensile strain at the bottom of the asphalt layer, 2) vertical compressive strain in the subgrade, 3) tensile stress in the soil-cement, and 4) tensile strain at the bottom of the soil-cement layer, Surface deflections are also examined.

It is shown that if a “weak” soil-cement base (E = 100,000 psi) is used over a weak subgrade, design may be controlled by either subgrade compressive strain or tensile stress in the base. On stronger subgrades the subgrade compressive strain controls. For a “strong” soil-cement base (E = 1,000,000 psi) the design is controlled by tensile stress in the base for all subgrade strengths if surface deflections are not critical.

It is shown that no single value of equivalency, in terms of inches of granular base required to provide the same subgrade protection as one inch of soil-cement, exists, but that the equivalency depends on the modulus of the soil-cement, modulus of the subgrade, and thickness of the asphalt layer. Previously reported equivalency value of 1.5 to 3.3 fall well within the range predicted by the analysis.

The analyses presented in this paper are intended primarily to show how soil-cement bases can perform a valuable load carrying and distributing function in a pavement section. Before the results of such studies can be used for design purposes, however, more information is needed concerning fatigue characteristics of soil-cement, cracking and other weathering effects that may develop in the field, and the suitability of the laboratory repeated loading tests for prediction of stresses and deflections in the field.

The zone belongs to a subtropical region with warm climatic conditions and heavy rains over the whole year, but reaching peak figures during spring. Rainfall on the past Spring-Summer (1965-1966) from October 1, 1965, up to April 30, 1966, made one of the highest figure of the century with a total of 1.726 mm (68 inches) for the mentioned period of seven months. Also the Parana River which flows at the north of the route marked on March 2nd the second great flood of the century, being the maximum on 1905, producing many troubles in the ground affecting the drainage of the zone which normally run off to the river.

Drainage is always a problem due to the extreme flatness of the natural ground which contains an excellent black “top soil” for agricultural purposes, used most in the cultivation of rice. This fact means an important factor against the good behaviour of pavements, because of the moisture accumulation on the top of the soil profile and moisture accumulated within.

RheologicaI Behavior of Asphalt Pavings Under Traffic
0. Andersson, National Swedish Road Research Institute, Stockholm, Sweden

An Experimental Investigation of the Stresses, Strains, and Deflections in a Layered Pavement Structure Subjected to Dynamic Loads
S. F. Brown and P. S. Pell, University of Nottingham, England

The Surface Wave Method
R. Jones, E. N. THrower, and E, N. Gatfield, Road Research Laboratory, Berks, England

Field Measurement of Dynamic Elastic Moduli of Materials in Flexible Pavement Structures
C, T. Metcalf, Shell Oil Company, San Francisco, California

New Developments in Vibrations Techniques
L. W. Nijboer, Rijkswaterstaat Rijkswegenbouwlaboratorium, Delft, Netherlands

Vibratory Study of Stabilized Layers of Pavement in Runway at Randolph Air Force Base
A. A. Maxwell and A. H. Joseph, Waterways Experiment Station, Vicksburg, Mississippi

Deformability, Fatigue, and Healing Properties of Asphalt Mixes
P. Bazin and J. Saunier, Compagnie de Raffiage Shell Berre, Paris, France

Results of Fatigue Tests on Different Types of Bituminous Mixtures
J. M. Kirk, Asfaltindustriens, Oplysningskontor, Copenhagen V, Denmark

Fatigue of Asphalt Pavement Mixes
P. S. Pell, University of Nottingham, England

Effect of Asphalt Aging on the Fatigue Properties of Asphalt Concrete
B. A. Vallerga, F. N. Finn, and R. G. Hicks, Materials Research and Development, Oakland, California

The Behavior and Performance of AsphaIt Pavements with Lime-Fly Ash-Aggregate Bases
E. J. Barenberg, University of Illinois, Urbana, Illinois

The present paper deals with measurements of the compression of wearing courses under a wheel moving at different speeds and at different axle loads and also similar measurements in a road under normal traffic. The properties of asphalt concrete have also been studied by analysis of stress relaxation and creep data of laboratory specimens recorded after different rates of compression and loading.

The approach has been to build up a three layer system investigating first a single layer, a clay subgrade, then a two layer system incorporating a granular base layer and finally the complete structure after the addition of a bituminous surface.

This paper is concerned with results from the first two of these systems, the two layer results being less complete and generally of a more preliminary nature than those for the single layer.

The main aim of the work has been to check the validity of linear elastic theory as applied to, firstly a homogeneous semi-infinite mass (Ahlvin and Ulery), and secondly a two layer system (Jones). In order to do this, stresses and strains have been measured within the pavement at various depths and radii and on sufficient planes to determine, by superposition, stress and strain conditions at a particular point. In addition, values of in-situ modulus and Poisson’s ratio have been calculated from the results.

The present paper describes work in progress at the Road Research Laboratory, England, on some aspects of the method. The experimental technique employs frequencies within the range of 25 c/s to 30,000 c/s, and emphasis is placed upon manual and computer aids used to derive the experimental dispersion curves from the initial data. Recent work on the basic theory indicates how theoretical dispersion curves can be calculated for most types of road structure. Inversion of the dispersion curve to give unique solutions for elastic properties and layer thicknesses is possible only in rare cases, notably single layers of concrete or cement-bound materials. Inversion of the dispersion curves for other cases is possible if some of the parameters, layer thicknesses for example, are known.

To provide an adequate check on the theory, models of roads are being tested. Experimental dispersion curves obtained in two-dimensional models with one or two superficial layers give excellent agreement with theoretical dispersion curves calculated from independent measurements of the elastic properties and thicknesses of the uniform materials i.e. aluminum alloy, Perspex and linoleum, used in the models.

The paper also gives a brief account of the practical uses which have been made of the Surface Wave Method at the Road Research Laboratory. The thickness, elastic properties and strength properties of concrete and cement-bound layers have been derived, and a modified technique using one frequency only has been developed for studying variations of quality of cemented road layers. Possible extension of these applications to other materials is discussed. Other applications have included investigations to follow changes in the properties of road bases of granular, cement-bound and bituminous materials with time and traffic.

Tests at Colorado’s experimental base project were made to investigate the properties of pavements containing relatively thick layers of asphalt stabilized sands in the base course. Results here showed that it was necessary to adopt a separate method of analysis – other than the past conventional approach – to evaluate those pavements containing thick layers of asphalt-treated base. Temperature was also found to exert a significant influence on results.

In other tests at Minnesota’s extensive field design project, velocities we obtained on a wide variety of materials contained in pavements in the regular highway system throughout the state. Again it was found that analysis of the data was dependent on the particular type of pavement involved: pavements with thick requiring different techniques than those with thin surfaces.

The paper deals with vibration phenomena in a road construction. The theories, developed by Jones (R(oad) R(esearch) L(aboratory), England, are discussed, both for the case of a single slab of bound material, and for a compound slab, consisting of two layers. His theories cover the case of “no friction” between the two layers and the case of “no slip.”

The measuring technique is elucidated and the mathematics applied in the analysis of vibration phenomena are discussed. Experiments carried out on a single slab (c.q. cement concrete and asphalt on a granular base) and on compound slabs (asphalt on lean concrete, asphalt on sand-cement, asphalt on blast furnace slag) and asphaltic concrete on an asphaltic base are discussed. The latter covers one case of “no friction” (asphalt on lean concrete), the other experiments represent cases of “no slip.”

Means to control the correctness of the analysis of the vibration measurements are discussed. Two classes of materials are distinguished, viz. hydraulically bound materials (cement bound material, blast furnace slag) and bitumen bound materials. It is indicated that the value of the Young’s modulus of the first group of materials is independent of frequency, whilst for the second group the rules by which the frequency influences the E-value is elucidated.

For the material of the bituminous layer the concentration of the mineral aggregate in the layer (Cv) is used as a criterion. The value of Cv can be determined from vibration measurements and from an analysis of the material. A second means of control by comparing measured and calculated R-values (Spring constant of the structure) is discussed.

The paper gives a survey of values of the Young’s modulus obtained in the field on hydraulically bound materials and the values of Cv on bituminous bound materials and points to some discrepancies for the cement bound material. Calculations show that for some existing roads in the Netherlands the radial stress at the bottom of the second layer will be 80% higher in the case of “no friction”, than in the case of “no slip,” demonstrating the constructional advantages of a “no slip” construction over “no friction.”

The basic discussion in this paper centers about the last-named item, primarily because it was felt that there would be a distinct change in strength with time as a result of the addition of lime. In order to record possible changes in strength, it was decided to utilize the method of nondestructive testing described herein in an effort to evaluate the strength changes with time, particularly with respect to lime stabilization. It is of course recognized that changes in strength might be brought about by other factors, and certainly these will not be ignored in yearly evaluations. It is planned to carry on this study for several years longer.

The ever-increasing traffic necessitates a constant improvement in road quality. The bituminous mix, the main purpose of which was to impermeabilize and improve the surface quality, is becoming more and more important as a load carrier in the road structure.
Under these circumstances, it appears extremely necessary to obtain a method, which can rationally determine the thickness of the mixes to be used.

At present, numerous methods have been proposed: they are usually based on the elastic characteristics of the utilized material. In the case of flexible pavements, this is only an easy approximation; in fact, the mechanical properties of bituminous mixes depend on the temperature and the time of loading, which deforms them.

In order to broaden our knowledge in this field, we have designed and manufactured a laboratory apparatus, which enables us to study the deformability and behaviour under repeated load of the bituminous mixes, under conditions which are equivalent to those of their utilization on the road.

In this article are contained the description of the utilized apparatus, in addition to the results concerning the determination of the complex modulus and the fatigue caused by bending. This article additionally contains an attempt to correlate the fatigue provoked by this apparatus and that caused by a loaded wheel rolling on a bituminous carpet.

Finally, after having studied the relative influence of several parameters (nature of bitumen, void-content, etc. . .) on the bending and fatigue, the first results are given of a study on the healing of mixes after rupture and on their recovery ability, after a fatigue period, which has not yet caused noticeable cracking.

In the last ten years, a number of rational design methods have been developed, based upon the usual engineering principle that the calculated stress in the different materials used, must not exceed a certain permissible stress, which depends upon the properties of the material.

This increases the need of knowledge of the strength of the materials used in road construction. As far as bituminous mixes are concerned, a number of workers have investigated the fatigue strength by various methods and Pell gives a review of this research. The most important general findings are that for a given mix, the fatigue strength – that is the number of loadings to failure – will depend upon the strain, while temperature and speed of loading is of negligible importance, as found by Saal and Pell, and further that the fatigue strength depends upon the bitumen content of the mix as stated by Pell. A number of tests, which have been carried out in order to obtain a more general picture of factors, which influence the fatigue strength are described in the following.

Irrespective of design methods, however, there can be little doubt that under traffic loading the layers of a flexible pavement structure are subjected to continuous flexing and the possibility of fatigue cracking exists under these conditions. With the advent of larger wheel loads in conjunction with heavier traffic it is likely that fatigue considerations will assume greater importance in the future for the economic design of asphalt pavements.

These developments have led to research into the fatigue behaviour of asphaltic mixtures being carried out in a number of different countries. It is not proposed in this paper to review in detail the present information available in this field as this task has recently been done elsewhere but to report the work carried out at Nottingham University since the first conference.

Before giving details of the tests and their results a few comments will be made on the relevant facts which have emerged from earlier work as it is hoped that this will assist the assessment of the experimental results and an understanding of the importance of the various factors involved.

Possibly the greatest difficulty in interpreting fatigue test results arises from the fact that they are influenced by the method of testing. Generally testing methods fall into one of two types, either “controlled stress” when the loading is in the nature of an alternating stress of constant amplitude, or “controlled strain” when the loading is in the form of an applied alternating strain or deflection of constant amplitude. The influence of the type of test is well illustrated by considering the effect of stiffness on the fatigue life of identical specimens tested by both methods.

The fatigue test results for each asphalt were interpreted with the aid of three-layered theory for three typical roadway structural sections. The variables included in this analysis were surface stiffness (E1), base stiffness (E2), and surface and base thickness.

The results of the interpretation confirmed the hypothesis presented in reference (1) that artificially aged asphalts impart better fatigue properties to asphalt concrete than unaged asphalts at comparable consistencies. In addition, it was shown that for equivalent pavement designs the thickness of the surface layer and the magnitude of the base modulus significantly affect the fatigue life of asphalt concrete.

This presentation is divided into three parts: First is a review of the properties of pozzolanic materials. Included in this review will be the proposed design procedure and the rationale behind the proposed procedure. Next is the review of the field performance of the pavements with pozzolanic base materials. Finally a brief discussion of the significance of the findings and recommendations for changes in the procedure are presented.

The pavements considered in this study were selected from several parts of the country to make the study more representative, Since it was not feasible for the author personally to collect all of the data, specific information was required from various contractors, materials suppliers, and industrial organizations. Much of the data presented herein are in response to these requests.

Stress and Strain Measurements in Experimental Road Sections Under ControlIed Loading Conditions
K. H. Gusfeldt and K. R. Dempwolff, Deutsche Shell A. G., Wilhelmsburg, West Germany

Observed and Calculated Strains at Various Depths in Asphalt Pavements
A. I. G. Klomp and TH. W. Niesman, Koninklijk/Shell Laboratorium, Amsterdam, Netherlands

Testing Flexible Pavements Under Normal Traffic Loadings by Means of Measuring Some PhysicaI Quantities Related to Design Theories
L. W. Nijboer, Studie-Centrum Wegenbouw, Arnhem, Netherlands

Deflection Prediction in Prototype Pavements
J. R. Morgan, University of Melbourne, Victoria, Australia
J. C. Holden, County Roads Board of Victoria, Victoria, Australia

AnaIysis of the Elastic Behavior of Flexible Pavement
H. Y. Fang, Lehigh University, Bethlehem, Pennsylvania
J. H. Schaub, West Virginia University, Morgantown, West Virginia

EvaIuation of El Toro Airfield by Layered Theory
J. P. Nielsen, U. S. Naval Civil Engineering Laboratory, Port Hueneme, California

Viscoelastic Properties of Bituminous Mixtures
G. Sayegh, Laboratoire Central des Ponts et Chaussees, Paris, France

The Effects of the Rheological Properties of Asphalt on Strength Characteristics of Asphalt Concrete
W. L. Hewitt and F. 0. Slate, Cornell University, Ithaca, New York

A Practical Approach to Flexible Pavement Design
F. P. Nichols, Jr., National Crushed Stone Association, Washington, D.C.

L.C.P.C. Studies on Pavement Design
J. Bonitzer and PH. Leger, Laboratoire Central des Ponts et Chaussees, Paris, France

In earlier trials it has been found that some 80% of the forces brought to bear are absorbed by the pavement itself. A method was developed to measure stress and strain through the complete thickness of the structure. Preliminary tests made with strain gauges embedded in carrier blocks revealed that a simple, linear relationship existed between tyre pressure and horizontal strain in the road.

Following these promising preliminary trials we set ourselves the task of ascertaining stress and strain in all three dimensions under controlled loading conditions with a single wheel. In doing so, as many factors as possible were simplified and kept constant, for instance, both acceleration and deceleration on the test section were deliberately avoided. The object of these trials was solely to study the strain and stress pattern itself. Repeated tests served to verify the measurements under varying temperature, speed, and load conditions. Fatigue tests were deliberately segregated from the test track and transferred to the laboratory along with the determination of the strength values.

On the basis of these preliminary tests the experimental road was so dimensioned that the wheel exerts pressure on the centre line of the test area and no edge effects distort the strains and stresses. The experimental road, moreover, readily permitted variations in the road construction, built on a scale of 1 : 1. It was thus possible to avoid any change in the strain and stress conditions which might be caused in a scaled-down model.

Properties of materials as well as differences between the load system used in the mathematical approach and the real wheel load are elucidated.

It is pointed out that as a first stage in this study a simple construction (two layer system) and a simple load system (single wheel) have to be preferred for the experimental investigation. Conditions of testing were chosen in such a way that a single construction thickness sufficed as variation in the properties of the chosen all bituminous construction could be obtained by testing over the four seasons at different temperatures.

The measuring techniques are elucidated as the strain gauges at various levels and the optical deflection technique used present YEW developments. Details of the construction and of the loading system are given. The properties of the materials have been determined both in the field by in situ measurements (vibration technique) and in the laboratory, and the properties during the test runs are discussed.

From existing mathematical methods for the elastic design theory graphs have been prepared for use of handling and with a view to simplify application, both for the Burmister method and for the “slab” method (Jeuffroy).

The results of measurements under the centre of loading are discussed and a comparison is made for all the physical properties measured between measured data and calculated values according to the two methods. From the results conclusions are drawn as to the correctness by which the elastic theories are able to predict actual stresses, strains and deflections in the road construction under traffic. The results of some measurements to determine the influence on the lateral position of the wheel are discussed.

Tests were carried out on an air-dry graded sand contained in a tank 6 ft x 6 ft x 4 ft-8 inches high. The surface was loaded by a 10-inch diameter flexible membrane to a maximum pressure of 30 psi and deflections were recorded under both initial and repeated applications of the load. Stresses in both the vertical and horizontal direction were measured using miniature earth pressure cells. The measured stresses at representative points along the centre-line were applied to 4 inch diameter by 8 inch high samples and the strains measured. These strains were then integrated to compare with the surface deflection measured in the sand tank.

The paper describes the special techniques used to calibrate the pressure cells. These were found necessary to overcome the problem of apparently statically inadmissible stress ratios reported by previous investigators. Methods used to measurs axial and radial strains in the triaxial samples to an accuracy of +/- 0.0002 inches are also described.

The results indicate that surface deflections in the prototype pavement under first-cycle loading can be predicted reasonably well, although the method is not suitable as a design technique because of the refined experimental measurements required. The stress measurements indicate that the simple Boussinesq elastic theory is inadequate, particularly for prediction of radial stresses, It is suggested that consideration of anisotropic material behaviour would lead to better stress prediction in real materials.

A simple equation based on the elastic theory is developed. The modulus of subgrade reaction, K, can be easily estimated by use of the equation if the thickness of the pavement components and pavement surface deflection as measured with the Benkelman beam are known. An equation using nondimensional techniques based on methods of dimensional analysis is developed for analyzing the results of Benkelman beam deflection tests on flexible pavement.

It is concluded that there is a close agreement, theoretically and experimentally, between the Benkelman beam and conventional test methods. It is suggested that the Benkelman beam could be adopted for general use for the evaluation of pavement performance and structural design of flexible pavements at a savings of considerable time and money.

At the Laboratoire Central des Ponts et Chaussees, several methods have been developed over the past five years in order to study these properties in the case of bituminous mixtures. They provide rheological models and corresponding analytical expressions utilizable for the design of bituminous pavements.

The characteristics we chose are complex moduli (Young’s modulus and shear modulus) and Poisson’s ratio, determined from vibration methods. The main reason for this choice is that it allows experimental conditions resembling those prevailing in actual pavements under traffic loads. Creep tests were used too in order to cross-check and extend results from dynamic experimentation to very low frequencies or slow deformations.

In the first part of this paper we shall discuss results obtained with a bituminous concrete of the so-called “Le Mans” type (used for the wearing course of the racing circuit of Le Mans). A rheological model and analytical expressions for the complex moduli and Poisson’s ratio will be derived.

In the second part we shall discuss results obtained with other types of bituminous mixtures and try to draw some general conclusions about the behaviour of bituminous mixtures in the range of linear (small) deformations.

Finally we shall compare results of in situ measurements with those of laboratory experiments. It will be seen that it is possible to predict the performance in situ from laboratory experiments.

For this investigation, asphalt concrete test specimens were prepared in the laboratory using limestone aggregate of one gradation at an asphalt content of 5.5 percent, based on the total weight of the mix. Five asphalts were used: three were from Venezuelan crudes and had penetrations of 70, 94, and 173; and the other two were of penetration grade 85-100, one from Rocky Mountain crude and the other from California crude. The effects of asphalt viscosity on the compressive strength of asphalt concrete at 80 and 140F, and at a rate of deformation of 0.05 inch per minute, were evaluated. Triaxial specimens were tested at zero and at 30 pounds per square inch lateral confinement. Also, the influence of rate of loading on the compressive strength of asphalt concrete at 80F was investigated. The data were analyzed to determine pavement resistance to tire contact pressure.

Modulus of rupture for mixtures using each type of asphalt was determined at 40 and 80 degrees F. In addition, specimens were subjected to repetitive loading in flexure at a deformation of plus and minus 0.005 inch from the neutral axis, and the modulus of rupture was determined after 1,000 and 100,000 load repetitions. The influence of repetitive loading on flexural strength is shown.

The influence of asphalt viscosity on the tensile strength and deformation of asphalt concrete was investigated. For each type of asphalt, specimens were tested in tension at 0, 20, 40, and 60F, and at a rate of loading of 0.002 inch per minute. The effect of rate of loading on the tensile strength of asphalt concrete at 0 and at 40F was investigated.

Instrumentation and test procedures are described. The application of the results of this investigation to aspects of asphalt pavement design is discussed.

This paper points out the fallacy of constant equivalency ratios and cites the fact that, even on the same project, the relative strength ratio between the same two materials varies with the magnitude of load and, most particularly, with depth below the surface.

A more practical approach to a universal design procedure is proposed. Every design would include provisions for:
(1) permanently firm, non-resilient support at the subgrade level, to be assured where necessary by means of chemical stabilization,
(2) a base course affording stability without brittleness, and
(3) a binder-surface combination of high stability with adequate cohesion to resist horizontal shear stresses from acceleration and deceleration.

Thicknesses would be essentially standard for standard materials under standard traffic loadings. Although minor thickness adjustments might be made where only non-standard materials are available, all three components should still be present, and overall thickness should be great enough to prevent overstressing the roadbed soil beneath the lowest improved layer.

The appendix contains a brief summary of the latest deflection and performance data from experimental sections on Virginia highways. The design recommendations are largely based upon these data.

Engineers are concerned about more exact design of new pavements to support heavy traffic, and particularly motor-road pavements. They are concerned too about a detailed evaluation of the destructive effects of traffic according to axle load, speed, and environmental effects. French law allows single axle loads up to 18 t, much heavier than other countries (except Belgium), and one is concerned about the advantages and disadvantages of the situation – and the very important problem of highway strengthenings. French highways are frequently broad enough for the present traffic, but insufficiently thick for the actual loads they must support.

The total length of highways that should be strengthened either by a bituminous surface course, or by a layer of cement-treated or granulated slag-treated material is at least 20,000 km for the national road system (one quarter of the whole system). It is comprehensible that such a problem played a major role in the planning of our research program.

Little information of immediate practical application about strengthening problems can be found in the AASHO-Road Test. Thus, it is necessary to evaluate the quality of the pavement to be strengthened from measurements, essentially deflection measurements (or, more generally speaking, measurements of the mechanical response to a load application). Devices like the Benkelman beam allow one to make relatively few deflection measurements per day, and are no longer adequate, for experience shows that static deflexions are frequently highly scattered about their mean, with standard deviations up to 1/4 of the mean, or even more; therefore it is necessary to make a great number of measurements. This is the purpose of the Lacroix-deflectograph. On the other hand, mean static deflection of a pavement section does not suffice to characterize fully the quality of this section. There may be road sections of poor load-bearing ability with very low deflections. Also, deflection shows seasonal variations, which are different for different structures.

Moreover, in the case of bituminous pavements, it depends on temperature and, as shown by experimental results, on vehicle speed. This makes it necessary to vary measurement conditions, to perform measurements with time-varying loads, and, on the other hand, to try to derive from deflection data certain intrinsic characteristics whose direct measurement would be difficult, such as Young’s moduli of pavement materials. We were led to think that it would be useful to focus attention not only on peak deflections, but on the whole deflection profile, and especially on the radius of curvature in the neighborhood of the peak deflection. This also explains why we have concerned ourselves with all kinds of methods of mechanical testing of pavements, as well as with mathematical models allowing us to compute deflections from intrinsic characteristics of the structures.

Suitable criteria for pavement failure conditions are needed. One of the major teachings of the AASHO-Road Test was confirmation of the fact that pavement failure was due to fatigue by repeated load applications. However this statement poses more questions than it solves. The paper outlines some of the questions.

Banquet Address before the Second International Conference on the
Structural Design of Asphalt Pavements at Ann Arbor, Michigan, August 9, 1967
by
John C. Kohl, Executive Secretary
Division of Engineering, National Research Council, Washington, D.C.

Pavement Deflections From Laboratory Tests and Layer Theory
B. S. Coffman, Ohio State University, Columbus, Ohio

The Properties, Behavior and Design of Bituminous-Stabilized Sand Bases
J. S. Gregg, G. L. Dehlen, and P. J. Rigden, National Institute for Road Research, Pretoria, South Africa

Load Transmission Characteristics of Asphalt Treated Base Courses
C. L. Monismith, R. L. Terrel, and C. K. Chan, University of California, Berkeley, California

A Field Experiement of Asphalt-Treated Bases in Colorado
R. I. Kingham, The Asphalt Institute, College Park, Maryland
T. C. Reseigh, Colorado Department of Highways, Denver, Colorado

Mechanical Properties of Asphalt Pavement Materials
B. F. Kallas and J. C. Riley, The Asphalt Institute, College Park, Maryland

Hot Mixed Sand Asphalt Bases in Oklahoma
B. C. Hartronft, State of Oklahoma Department of Highways, Oklahoma City, Oklahoma

The Use of Asphalt Pavement Structures in the Australian Environment
A. J. Scala and E. J. Dickinson, Australian Road Research Board, Victoria, Australia

(1) Moving loads under which the pavement response is essentially elastic.

(2) Stationary loads under which the pavement response is a complex function of time-dependent stresses and strains.

(3) Acceleration loadings which can impose shear and/or normal forces on a pavement.

A major problem in predicting deflections under such loadings is in defining the response of each of the load-supporting materials using laboratory or field derived properties. Once found, however, such defining properties may be used to describe the stress-strain-time characteristics of the pavement under any wheel loading.

This study was directed at an attempt to interpret the results of in-situ deflection measurements on a theoretical basis for loading class 1, and to some extent, 2. The pavement studied was a short section of U.S. 62 running north and slightly east of Columbus, Ohio. This pavement, a four lane divided highway, was constructed in the summer of 1963 and the materials used in constructing the pavement were sampled at the time of construction and tested in the laboratory. Tests were performed on each material over the range of temperatures or densities and moisture contents expected in the field.

For field deflection data, reference rods were placed in the pavement at nine locations and periodic measurements were made during 1964 and 1965. As a part of these measurements the temperatures of the asphaltic concrete layers were determined and the pavement was trenched to determine the moisture content and density of the subbase and upper subgrade as well as the thicknesses of the pavement layers.

The objective of these field tests was to define the material conditions existing at the time of each deflection measurement. With material conditions defined by the field tests, and using the lab test data, it was possible to calculate theoretical deflections for the different loading conditions with the layer elastic theory.

Thisis a report of comparisons between the resulting calculated and measured deflections. In a larger sense it is a test of the validity of the concept of combining individually determined time-dependent moduli to represent the material and loading conditions to be found in flexible pavements.

An intensive laboratory study was carried out on one representative bitumen-sand mixture. Triaxial tests carried out over a range of temperatures and a wide range of strain rates revealed that the strengths of the mixture could be expressed in terms of the Coulomb relationship. While the angle of friction phi remained fairly constant, the c-intercept varied in an orderly manner with strain rate and temperature, the effects of which could be combined using the time-temperature superimposition concept. Tensile strengths measured in direct extension varied more rapidly with rate of strain and temperature, but could again be combined using the superimposition concept. Repeated load triaxial tests carried out with various stress intensities and durations indicated that the modulus of resilient deformation was inversely related to the ratio between the major and minor principal stresses.

A field study was carried out on eight sections of road in South West Africa with various compositions of 6-inch thick bitumen-sand base. Relationships were found between the deformations which had occurred under traffic and the in-situ CBR, in-situ vane shear strength, and bearing capacity determined from triaxial tests on cores of the stabilized base. Very little deformation was observed beneath stationary trucks with 7000-lb wheel loads, Although more deformation might have occurred during the warmer season. The strengths associated with satisfactory performance were very much less than those specified for conventional granular base-courses. Plate bearing tests indicated that the moduli of resilient deformation of the bitumen-stabilized bases were of the same order as those of conventional unbound granular bases, so that it was unlikely that the use of these bases would permit reductions to be made in the thickness of cover to the subgrade relative to the design cover with an unbound base. Mixtures having dense gradings were generally found to have higher strengths and moduli of resilient deformation than mixtures lacking in fine material. The study is continuing.

Because of the widespread interest generated by the AASHO Road Test in asphalt-treated bases, and because of limited availability of data on the characteristics of certain asphalt-treated materials (primarily those treated with asphalt emulsions and liquid asphalts), this investigation was undertaken. Specifically, the paper describes:
1) laboratory measurements of the properties of asphalt-treated materials,
2) methods whereby the laboratory data can be used to predict pavement deflections within the framework of suitable theory, and
3) comparisons between predicted deflections and measured deflections in prototype pavement structures with base courses consisting of materials treated with asphalt cement, asphalt emulsion, and liquid asphalt.

From these analyses, a method is suggested whereby layer equivalencies can be established for specific situations through suitable laboratory tests on appropriately conditioned materials, and through analysis of systems representative. at least to a degree, of pavement structures. However, it is emphasized that no one equivalency can be applied to a specific material; moreover, for the emulsion and liquid-asphalt-treated materials, the equivalencies would appear to change during the initial period after placement because of curing.

The road was opened to traffic in the spring of 1965. The paper reviews the performance of the test sections through the spring of 1966. The Present Serviceability Index Concept developed at the AASHO Road Test was used to judge performance. Measurements have been made with both the CHLOE Profilometer and the Bureau of Public Roads Roughometer.

The load response characteristics of the various base types were determined using the Benkelman Beam for deflection measurement. In addition, a measure of the curvature of the pavement was made with the Dehlen curvature meter developed in South Africa. Beam deflection data were used with layered system theory to determine elastic moduli for the various asphalt base types. The moduli showed excellent agreement with those determined in the laboratory for similar temperature conditions and a loading frequency of one cycle per second. Theory and inferred moduli were used for asphalt bases to derive thickness-deflection relationships for average surfacing and base temperatures observed in the field. Such relationships provide one means of obtaining equivalent thicknesses of different materials when critical deflection values become available.

Relationships between applied forces acting on the various paving materials and resultant deformational response were developed by several laboratory testing procedures. Sinusoidal stresses varying in magnitude and frequency were applied axially to unconfined cylindrical asphalt paving mixture samples at several temperatures. The resultant axial strains were measured and dynamic complex moduli were determined. Values of Poisson’s ratio for these materials were determined by similar testing procedures. Repetitive triaxial tests were made on untreated granular base and subbase, and on the subgrade soil. Repeated short duration deviator stresses were applied to triaxially confined specimens and the resultant recoverable or resilient axial deformations were measured. From these tests resilient deformation moduli were determined for varying confining pressures at several levels of deviator stress. Repetitive flexure tests were also made on asphalt paving mixture beam samples. Repeated controlled short duration loads were applied to beam specimens and the resultant dynamic deflections of the beam were measured. Stiffness moduli were determined from results of these tests.

In addition to strength moduli, standard test data were obtained and are reported on the materials incorporated In the experimental project.

The strength moduli determined by the laboratory tests are used to calculate pavement deflections according to N-layer elastic theory. A computer solution developed by Chevron Research Company Is used for these computations . Generally, the calculated deflections for all asphalt base course test sections are lower than Benkelman beam deflections measured on the Colorado Experimental Base Project.

As of July, 1966 over 200 miles of two lane pavement have been constructed with hot mixed sand asphalt base as well as about that many miles utilizing this material for overlay and widening.

The remainder of this paper will be concerned with results of various testing on five different projects over a period of approximately four years. An attempt will be made to describe the components of the base, the apparent effect of the surfacing, the variation in the physical properties of the asphaltic cement, and the load carrying capacity of the pavement structures.

At present, these materials are applied as surface courses to pavements at airports and in urban areas and they are relatively expensive. Research is in progress to compare their behaviour with that of natural and crushed granular materials now generally used as the structural element.

Some results of the continuous measurement of temperature at various depths in the layer of asphaltic concrete at a site near Melbourne, are given.

Model experimental sections have been laid at one site to compare the load spreading properties of the two types of base material and, by surface and subgrade deflection combined with subgrade stress measurements, their relative behaviour is demonstrated.

A moderated discussion of the papers presented in Session VII.

Performance Studies of the Mexico City International Airport
L. M. Aguirre, D. Sanchez, and M. Zarate, Secretaria De Obras Publicas, Mexico, D.F.

U.S. Navy Experience with the Performance of Asphalt Pavements Subjected to High Pressure Aircraft Tires
P. P. Brown and C. E. Rhodes, Department of the Navy, Washington, D.C.

Use of Viscoelastic Concepts to Evaluate Laboratory Test Results and Field Performance of Some Minnesota Asphalt Mixtures
L. J. Gardner, Standard Oil Company of California
E. L. skok, Jr., University of Minnesota, Minneapolis, Minnesota

The Behavior of Flexible Pavements Under Moving Wheel Loads
N. W. Lister and R. Jones, Road Research Laboratory, Berks, England

A New Method in Correlation Study of Pavement Deflection and Cracking
Kuang-Yuan Kung Taiwan Highway Bureau, Republic of China, Taiwan, China

Long Term Deflection Study of an Exceptionally Maintenance Free Pavement
A. Lee, Maryland Road Commission, Brooklandville, Maryland
S. Williams, Bureau of Public Roads, Washington, D.C.
W. G. Mullen, North Carolina State University, Raleigh, North Carolina

The Lacroix L.C.P.C. Deflectograph
E. Prandi, Laboratoire Central des Ponts et Chaussees, Paris, France

Recent Full-Scale Flexible Pavemenl: Design Experiments in Britain
G. F. Salt, Road Research Laboratory, Berks, England

Field Performance Studies of Flexible Pavements
Pavement Design and Evaluation Committee, Canadian Good Roads Association, Ottawa, Canada

Since the early 1950s, the U.S. Navy has been operating high performance carrier jet aircraft on its training and fleet support stations. These aircraft utilize tires with inflation pressures which have steadily increased with the gross aircraft weights and restrictions on the size of landing gear. Operational pressures now exceed 400 psi.

This paper describes typical operations at several Naval Air Stations, the criteria to which flexible pavements have been constructed, and the performance of these pavements. In general, there are many well constructed high quality flexible pavements performing well under these high tire pressures at Naval Air Stations. Limited failures have occurred in areas of concentrated landings and of marginal quality materials and construction.

The paper also describes the procedures used by the Navy for assessing the suitability of flexible pavements for use by aircraft with high pressure tires.

The following testing has been done on these test sections in the field: (1) fractional plate load testing with in-place moisture and density determination of each layer, (2) determination of Benkelman Beam deflections, rut depth, and cracking for each section and (3) determination of the roughness index and present serviceability rating for each section. Items (2) and (3) are being run about once per year on each of the test sections. The roughometer index is used along with the rut depth and cracking to determine a present serviceability index periodically. The trend of the PSI with traffic will define the performance of the test section.

The following laboratory tests were performed on all 50 embankment soils: Atterberg limits, standard moisture-density, gradation, stabilometer R-values, and California Bearing Ratio tests. The granular bases and subbases have had the first four above-mentioned tests run on them plus CBR and triaxial tests on a selected number of them. A special study was made on the granular materials and it was found that the R-value test adequately classified the strength characteristics of the bases and subbases.

Extractions were made of the asphalt surface and base mixtures, and penetration, softening point, and ductility tests are being performed on the recovered asphalt. Four-inch cores of the bituminous materials have also been taken from each test section to determine density, cohesiometer, and modified tensile strength of the mixture. The cohesiometer and modified tension did not give very meaningful results. It was felt that the asphalt mixtures could be evaluated better if a nondestructive test could be devised which used a loading condition close to the magnitude of a wheel load. A repetitive type test was also considered preferable over a static test. This also involved evaluation of the mixture based on the viscoelastic rather than elastic or rupture theory.

In 1964 a study under the sponsorship of the Asphalt Institute was started at the University of Minnesota to try to develop a practical laboratory test for asphalt mixtures using some of the viscoelastic concepts developed at Ohio State University and the University of California. The goal of this work is to run tests on mixtures cored from the 50 test sections and compare the performance of these mixtures in the field with the stress-strain properties of the samples.

Both constant stress tests and repeated load tests have been made on a number laboratory prepared mixtures and, at the writing of this paper, the project at the University of Minnesota is at the stage of initial testing of some of the asphalt cores. Minimal meaningful comparisons with performance are only possible at present because most of the test sections have not as yet failed. The apparatus has been developed and a number of laboratory mixtures have been tested under different magnitudes of vertical and lateral pressures with one temperature.

This paper will cover the theoretical considerations, the apparatus developed for testing the mixtures, the test results and effects noted, and the results of the tests on field cores made up to the present time.

Dynamic loading effects generated by vertical motion of the wheel passing over surface irregularities has been shown to produce up to 35 per cent higher (or lower) surface loads than the nominal (static) wheel load on experimental sections of poor riding quality from an unladen test vehicle travelling at 15 m.p.h. In addition, variations of up to 40 per cent were recorded in the vertical component of stress in the subgrade beneath the same type of construction attributable to variations of nominally the same materials; the largest variations were observed beneath a crushed stone base and the smallest beneath uncracked cement-bound bases. The effects of dynamic loading and variability of materials are usually inter-related, and suggestions are made for further studies under normal traffic conditions.

Difficulties of defining the appropriate elastic moduli of soils, unbound materials and bituminous materials for use in the elastic analysis are discussed. Comparisons are made of the measured values of the surface deflection and the vertical component of stress in the soil with those calculated using moduli deduced from vibrational experiments. Suggestions are made for further experimental work to remove some of the present uncertainties.

The correlation between flexible pavement deflection and pavement performance has been studied for many years, Many experiments were carried out in this field. Most of them were run on the basis of the magnitude of deflection; some were on the basis of “Radius of Curvature.” Still a few others on the basis of “Bending Index.” Each has its specialty. However, the results do not agree either on the basis adopted or on the conclusions stated. . Furthermore, none of them can be free from inconsistencies to a satisfactory degree. Therefore which one is the best still can not be certain.

The writer found a new method based on “Slope of Deflection” which is derived from a typical deflection curve by means of basic principles of the “Strength of Materials.” . This new method is not only sound in theory but also strongly supported by field tests and observations on Virginia test highways. This paper will first discuss the existing methods for comparison purpose, and then the writer’s new method “Slope of Deflection.” Finally two suggestions for securing more accurate data in field measurements are also discussed.

In 1960 the study was expanded to include observations on newly completed flexible pavements in addition to existing pavements selected in 1955. At that time the entire study was included as a project under the HPS-HPR program of the Bureau of Public Roads. Each year all newly completed flexible pavements have been added to the program so that initial deflection observations are available.

In the fall of 1963, a physical sampling investigation of the pavement layers, base and subgrade of one of the pavements in the original study, U.S. 40 between Frederick and Hagerstown, was undertaken. Analysis of the behavior of another of these pavements, I-70S between Frederick and Rockville, has been reported under a separate study.

A number of interesting findings have resulted from the over-all study. Included are observations that spring deflections are higher than fall deflections, outer wheel path deflections are higher than inner wheel path deflections. and, most significant, for the range of pavement thickness encountered in Maryland, the magnitude and dispersion of deflections may be grouped into different categories according to subgrade soil types.

The over-all findings hold true for U.S. 40 West, the pavement that is the subject of this paper. In addition, the continuous good performance of this pavement may be attributed to adequate structural thickness and to an unusually well-protected subgrade. Subgrade protection has been afforded by an impervious bituminous concrete surface course coupled with good longitudinal drainage.

The Benkelman beam was used for the first investigations on pavement deformability. It possesses, it seems to us, two main drawbacks:
– its precision is limited
– the rate of operation is low: in our experience, 50 to 80 measurements per day with a well trained crew.

The use of the so-called JYC optical deflectometer allowed of only limited improvements. In 1956, M. Lacroix attempted to design an apparatus which would be able to perform a large number of deflection measurements per day, Teamwork, including agencies of the Service des Ponts et Chaussees (Dordogne, Maine-et-Loire) and the LCPC resulted in the development of several successive prototypes, the first of which was already able to perform 7,000 to 8,000 measurements per day.

A deflectograph includes:
– A truck travelling at approximately constant slow speed, with a rear axle load of 13 tons;
– A measurement beam, which is analogous to a Benkelman beam. This beam rests on the pavement during measurement and is automatically shifted from one point of measurement to the next.
– A measuring and recording system converting displacement of the tip of the probe arm of the beam into an electrical signal.

The main original feature of the deflectograph is the intermittent displacement of the beam, which allows measurement to be performed during rest periods.

On standard apparatus (about 20 now in operation), the influence lines of the deflection for each measurement point is photographically recorded. A digital recording system, using a magnetic tape, was recently tried.

Since 1960 six further experiments have been constructed, investigating the performance of materials which have recently come into common usage in the country, and examining in more detail points of doubt or anomaly arising from previous work. These experiments were mentioned as proposals for future work in the original paper, and this contribution thus constitutes an extension to it.

The paper first describes the changes which have taken place since 1962 in the procedures adopted at the British Road Research Laboratory for carrying out this work and then goes on to give details of the six new experiments.

Detailed results are given for only one of these as the remainder have not yet been in service long enough for measurements to be significant. The possible implications of some of the early measurements are, however, briefly discussed.

Extensive regression analyses on the Stage 1 pavement inventory indicates the significance of age and strength on pavement performance. The regression equations are most significant at pavement ages of 6 to 11 years. Extensive data is shown for loss of strength during the spring of pavements subject to frost action. Regression analyses were not successful in predicting the loss of strength. Uniform and non-uniform pavements are described together with data showing the improved performance and reduced thicknesses that are achieved with uniform pavements. Small weak areas are associated with non-uniform pavements. A design method for new pavements is outlined together with construction control methods based on Benkelman tests. Construction control methods to attain uniformity are emphasized as being supplementary to design and equally important.

Terminal Remarks by Prof. W.S. Housel, Chairman, Executive Committee.