Recent Advances in Materials Characterization and Modeling of Pavement Systems

This paper presents a new method using x-ray tomography imaging to quantify the specific surface area of aggregates. The method reconstructs the three-dimensional (3D) representation of individual particles and computes the specific surface area and the sphericity (a shape factor) of individual particles. The quantified specific surface area of a limestone aggregate in the sieve size range (3/8"-No.4) indicates that the specific surface area and the sphericity have a large variation among individual particles. However, the overall specific surface area of the aggregate particles within the same sieve size range can be accurately obtained using about 25–30 particles. It has been found that the specific surface area of the aggregates is much larger (84%) than that of the spheres of equivalent size.

A new methodology is presented in this paper for determining cross-anisotropic aggregate properties, i.e., directional dependency of moduli and Poisson's ratios, from stress path testing. The proposed laboratory characterization procedure requires conducting large stress excursion tests performed under slightly different stress path slopes at similar stress states that are representative of pavement wheel loading conditions. In accordance, cross-anisotropic aggregate properties are determined by varying slightly the stress path slopes during testing and then by employing an error minimization approach to interpret the test results. The solution methodology adopted also satisfies positive strain energy requirements. Such an anisotropic material property set obtained from following the proposed methodology would essentially be useful in the finite element analysis of the anisotropic and stress dependent granular material behavior often observed in unbound pavement layers.

Base and subgrade soil stiffness are important parameters for material characterization in highway construction. During and after pavement construction, it is very important and cost-effective to have a reliable technique that can measure the stiffness of in-situ base and subgrade layers accurately and quickly. The new highway design guide proposed by AASHTO requires that such measurement be conducted. A new field-testing technique has been developed to measure small-strain shear modulus of soils (which can be converted into elastic modulus) using the bender element technique. The device used in this method includes a pair of cone penetrometers, each fitted with two bender elements, which can be pushed into the base, subbase and subgrade layers. A pulse generated by a function generator is used to activate the transmitters. Vibration of the transmitters produces shear waves that propagate through the soil in the horizontal and inclined directions and are captured by the receivers. Then from the measured shear wave velocities, soil stiffness can be determined. The technique is compared with the current practice of soil stiffness determination based on the CBR test.

An understanding of the cracking behavior of hot mix asphalt mixtures provides an important basis for the development of mixtures with improved cracking resistance and pavement cracking performance models. This paper describes the use of random assemblies of displacement discontinuity boundary elements to model the behavior of asphalt mixtures. These random assemblies of boundary elements form particles of varying sizes, which are arranged to simulate the discrete nature of mixtures. To account for the aggregate and mastic properties, each particle is connected to other particles by specifying the cohesion, friction, and tensile properties of the material between particles. The crack initiation and crack growth are simulated using two distinct crack growth laws, based on sequential and parallel crack growth rules. Cracks are allowed to grow along particle boundaries. A series of simulations were performed of the load-deformation and cracking behavior of a typical 19.0-mm nominal maximum aggregate size mixture commonly used by the Florida Department of Transportation. The effects of crack growth rules, particle size, and localization to geometric effects were studied. In summary, the method presented appears to provide a valuable tool for studying the mechanistic behavior of asphalt mixtures.

The minimization of deformation in hot mix asphalt is an important aspect of pavement design. The performance of asphalt mixtures is strongly influenced by both the behavior of aggregate particles and the rheology of the binder. The resulting rheology of the composite material — hot mix asphalt — is significantly shifted from that of the asphalt binder. Rheology of binders and mixtures has been evaluated by testing, which includes Dynamic Shear Rheometer (DSR) and Uniaxial Cyclic Modulus (UCM) measurements, respectively. Master curves of stiffness modulus are constructed using principles of time-temperature superposition that consider both the elastic and viscous components of the modulus. The Federal Highway Administration (FHWA) Accelerated Load Facility (ALF) study has provided an important source of data since materials that were evaluated represented wide ranges in performance. Data from the rheology measurements have been compared to deformation performance of pavements tested at a constant temperature. The determination of the high temperature viscosity properties versus plastic deformation and straining are discussed. A practical framework for the extraction of material parameters from cyclic frequency sweep testing for use in specifications for quantifying the quality of asphalt binders and mixtures is presented.

This paper presents a new simplified method of constructing a master curve of asphalt mix using test data covering a large range of temperatures from –18°C to 55°C. It utilizes sigmoidal fitting function and compressive cyclic (complex) modulus test data obtained at matrix combination of different frequencies and test temperatures. In the master curve construction, the time temperature superposition was modeled two different ways. First, using known time-temperature superposition equations, and second, shilling test data experimentally, i.e., not assuming any functional form for the time-temperature relationship. The master curve construction was done using an Excel spreadsheet with the solver function, which is a tool for performing optimization with non-linear least squares regression technique. The analysis of over sixty mixtures indicated that the experimental approach agreed the best with the Arrhenius shifting equation, correlation coefficient R² being 0.922. Also, the experimental shifting was most flexible, producing the best fit among the studied shifting equations due to the fact that it has the most degree of freedom.

Asphalt concretes are composite materials consisting.of interspersed aggregates, asphalt binder and air voids. Their constitutive behavior depends largely on the interaction of the aggregates and the binder, which have drastically different stiffnesses. This behavior is further complicated by the non-linear viscoelastic behavior of the binder. The paper at hand presents a fundamental methodology for describing the constitutive behavior of asphalt concretes. It involves three main steps. The first step involves the development of mechanistic binder models from rheological measurements and their implementation into stress-strain relationships suitable for FEM modeling. Binder non-linearity is treated in a piece-wise linear fashion, whereby the binder model constants are assigned as a function of strain level. The second step involves capturing the microstructure of asphalt concretes through imaging techniques and processing the images to translate them into FEM grids. The third step involves verification of these asphalt concrete models by comparing predictions of shear modulus and phase angle to the properties measured in the laboratory using a Simple Shear Test (SST).

Permanent deformation is one of the most significant distresses that causes severe damage in asphalt concrete (AC) pavements. It is caused by high traffic loads associated with high field temperatures. It is believed that permanent deformation develops at a small rate and accelerates with the initiation of microcracks in asphalt pavements. An anisotropic viscoplastic continuum damage model is developed to describe the permanent deformation of asphalt pavements. The model is based on Perzyna's formulation with Drucker-Prager yield function modified to account for the material inherent anisotropy. The material anisotropy parameter is measured through microstructural analysis of two-dimensional sections of asphalt mixes. Furthermore, a damage parameter is included in the model in order to quantify the nucleation of microcracks that develop to macrocracks at later stages. The model's parameters are all determined from strength and static creep tests.

Estimation of field compaction requirements to achieve a target density is important for predicting construction costs of asphalt concrete pavements. A model to predict the effects of mixture volumetric properties on compaction efficiency is desirable to optimize the design of asphalt mixtures. A compaction model is developed to predict laboratory compaction by the Superpave gyratory compactor. The presented work represents only a preliminary evaluation of the model to extend it to field compaction conditions. A pressure dependent porous material with elastoplastic matrix is assumed to contract under a prescribed compaction pressure induced by the Superpave gyratory compactor to a 6 in. cylindrical sample. Plastic strains are integrated from an incremental elastoplastic constitutive equation by forward difference method. Three model constants, q₁, q₂ and σ_y, are calibrated to determine measured deflections from the gyratory compactor by means of Levenberg-Marquardt nonlinear parameter estimation algorithm. A statistical correlation between estimated parameters and volumetric components are constructed to evaluate the effects of mixture properties on the estimated model constants. The proposed constitutive model accurately predicts the volume change behavior of mixture specimens undergoing gyratory compaction. Mixture volumetric properties also show correlations with the estimated model constants.

A constitutive model based on an extended form of the Schapery continuum damage formulation is currently being evaluated and developed as a comprehensive material model for asphalt concrete. This model considers the viscoelastic, damage, and viscoplastic components of asphalt concrete behavior over the full range conditions of interest for the mechanistic prediction of flexible pavement distresses. The focus of the present paper is limited to the viscoplastic response component at intermediate and high temperatures. The results confirm earlier findings that asphalt concrete in compression is a thermorheologically simple material well into the large strain viscoplastic regime at elevated temperature. The study demonstrates that the proposed viscoplastic model component provides a good representation of the viscoplastic response of asphalt concrete in uniaxial unconfined compression. Given the validity of time-temperature superposition for viscoplastic response, the viscoplastic material parameters can be calibrated from a limited number of uniform time and uniform load creep and recovery tests. Typical viscoplastic material parameters are derived for a representative asphalt concrete mixture.

Some results of an extensive experimental and numerical research project on the response of Asphalt Concrete (AC) are presented. In the project, a material model for AC is being developed and implemented in the Finite Elements package CAPA-3D. Also, the test set-ups, instrumentation, and test procedures necessary to characterize the material and determine the model parameters were developed, In an early stage of the project, it was recognized that failure in tension and compression followed a different mechanism and needed to be described separately in order to capture the response to alternating loads. From the results of uniaxial tension tests it became apparent that also in tension pronounced non-linearity occurred prior to the peak load. The adaptations necessary to incorporate this in the model are presented in this paper. The response predicted by the model relations is compared to that observed in laboratory tests to validate the model relations and it can be seen that the observed behavior is described quite well by the model.

The background and application of the stiffness matrix approach in the dynamic analysis of pavements and other layered systems is presented. The stiffness matrix approach involves a triple transformation of the problem. The first transformation is transformation of loading from the time to the frequency domain. The second and third are spatial transformations of loading, leading to the solution of the problem in the frequency-wave number domain. Finally, inverse transformations are used to return the pavement response to the original spatial-time domain. The most important advantage of the stiffness matrix approach is that the infinity (radiation condition) of the half-space and lateral boundaries is perfectly satisfied. The other advantage is that a single 4 x 4 (2 x 2) complex stiffness matrix provides a theoretically exact description of a single pavement layer (or a half-space). The method can be used for numerous purposes: in modeling and analysis of the response of a pavement to dynamic loads, simulation of nondestructive tests like falling weight deflectometer (FWD) and wave propagation based tests (e.g. Spectral Analysis of Surface Waves — SASW), evaluation of dynamic characteristics of pavement systems (natural frequencies and damping characteristics), etc. The results presented include the dynamic response of pavements to wheel loads, for the purpose of prediction of vibration (deformation) levels, and to small radii circular sources, for simulation of nondestructive pavement testing.

With the ageing of road networks all over the world, more and more emphasis is placed on the cost-effective rehabilitation of pavements. A meaningful assessment methodology is therefore required, for the evaluation of the existing structural condition of pavements, and for the effect thereof on possible alternative rehabilitation options. The mechanistic analysis of the bearing capacity of pavements is a methodology that can accurately take into account the effect of rehabilitation measures. This methodology requires that material properties (elastic moduli or stiffness, etc) of existing pavement layers be determined and modelled accurately in order to calculate the structural capacity of the pavement to be rehabilitated. However, one of the most important aspects of such a methodology is to accurately identify uniform pavement sections for modelling purposes. Currently, the cumulative sum of deviations technique is used, during the rehabilitation design phase, to identify uniform sections for modelling purposes. The objective of this paper is to give guidelines on an advanced application of the cumulative sum of deviations method / technique to identify uniform pavement sections more accurately. This method / technique takes into account the variability of the pavement structural layers along the length of the road. It is based on a holistic approach, which also includes visual assessment data, rut depth measurements and deflection bowl parameters etc. This technique will assist the designer to get more accurate and reliable uniform pavement sections, which in effect will minimize errors in the data that will be used during the rehabilitation design process.

Joint Deterioration of PCC pavement is a process that is controlled by a variety of quantifiable and other non-easily quantifiable parameters. In this study, artificial neural network (ANN) approach was used to correlate the time-dependent joint deterioration behavior with quantifiable experimentally-based aggregate durability parameters such as durability factor, percent expansion and freeze-thaw index. To achieve this objective, the historical Kansas pavement management system database, along with corresponding aggregate and core durability reports for a number of PCC sections, were combined to produce the needed time-dependent joint deterioration database. The resulting database was split into training, testing and validation sub-bases. Upon a number of ANN training, testing and validation processes, a time-dependent ANN-based joint deterioration model was developed. The developed dynamic model utilizes a number of aggregate durability factors to project the time-dependent joint deterioration behavior for up to 20 years after construction. Accordingly, model predictions for year n are sequentially used as inputs for predictions for year n+1. Overall, model predictions are noted to be logical and in good agreement with field observations.

This study focuses on the development and performance of a comprehensive artificial neural network (ANN) model for the analysis of jointed concrete slabs under simultaneous aircraft and temperature loading. Using the results from the ILLI-SLAB finite element program, a comprehensive artificial neural network model was trained for the different loading conditions of gear load only, temperature load only, and simultaneous aircraft and temperature loading cases. Special consideration has been given to the loading of a typical jointed slab assembly under the tri-tandem type gear of the Boeing 777 aircraft. In addition to various slab load locations (interior, corners, and two mid-span slab edges) and joint load transfer efficiencies, a wide range of realistic airfield slab thicknesses and subgrade supports have been considered as ANN input conditions. Comparing the ANN predictions to the ILLI-SLAB solutions validated the performance of the ANN model. The trained ANN model gave maximum bending stresses and maximum vertical deflections within an average error of 1.4 percent of those obtained directly from ILLI-SLAB analyses. It is also shown that the principle of superposition is not valid for simply adding the critical pavement responses (slab deflections and bending stresses) obtained under gear loading only and temperature loading only cases. The summation does not usually produce the same pavement response as obtained from the simultaneous gear and temperature loading. The typical ANN prediction time is about 0.3 million times faster than the average ILLI-SLAB finite element solution. The use of an ANN-based design tool is deemed to be very effective for studying hundreds or thousands of "what if" scenarios for including the temperature effects in pavement analysis and design.