A number of pavement structures experience deterioration due to high traffic volume and growing weights. Recently, the Texas Legislatures passed bills allowing trucks of gross vehicle weight (GVW) up to 556 kN routinely to use a route in south Texas along the Mexican border. Thus, there is a need to model pavement responses due to various types of overweight truck traffic (OTT) by taking into account axle loads, configuration, and pavement layer material characterizations in order to provide a guideline to assess the existing pavement performance and expected service life. It is for this purpose that the nonlinear cross-anisotropic pavement analysis finite element program (NCPA) has been developed. Stress dependent and directionally different resilient modulus and Poisson's ratios are incorporated into the finite element formulation to model the pavement response. As a tool to assess the performance of the pavement, the procedure to calculate the overall rutting and the cracked area was included in the formulation Intensive nondestructive testing has been performed to identify the existing pavement test section geometry and layer properties. In addition, a fiber optic based Weigh-in Motion (WIM) sensor was developed and tested. It is expected to be a promising device to monitor traffic by showing a reliable response. Sampled materials from the test section were tested to characterize their stress-dependent, cross-anisotropic and permanent deformation properties. Constitutive models are verified by comparing the predicted displacements with field displacements measured with the Multi-Depth Deflectometer (MDD). The result was that the least error between predicted and measured displacements is generated by the nonlinear cross-anisotropic model. In addition, the cross-anisotropic characteristic of the asphalt concrete material is introduced and evaluated based on the relationship between the backcalculated static and dynamic modulus. This addition improves the accuracy of the assessment of pavement performance with respect to both rutting and fatigue cracking. Charts to evaluate the service life of the existing pavement subjected to OTTs are established in terms of the unit service life consumed due to the rutting and fatigue cracking with the various observed combinations of pavement geometry, traffic load, and material properties.

Work done by several researchers reveals that unbound aggregate materials show nonlinear cross-anisotropic behavior. The incorporation of cross-anisotropic properties significantly improves the predictions of stress distribution by reducing tensile stresses computed within granular layers. Existing pavement analysis and design approaches, however, generally assume the pavement structure to be linear isotropic layered system. This assumption is motivated by the difficulties in determining cross-anisotropic resilient material properties from laboratory experiments and lack of pavement anisotropic analysis programs. Recently, the International Center for Aggregates Research (ICAR) developed a methodology to characterize unbound aggregate layers by considering stress-sensitivity and nonlinear cross-anisotropy. The ICAR model requires nine coefficients to account for stress-sensitivity and anisotropy of vertical, horizontal, and shear moduli. Unfortunately, ICAR testing protocol is time-consuming and expensive to perform and certainly do not lend themselves to routine testing. Since it is important to be able to consider the stress-sensitive and anisotropic nature of unbound granular materials, a simple procedure was proposed by accounting for the effects of aggregate gradation and shape properties in predicting the cross-anisotropic modular ratio of unbound granular materials. Variable confining pressure type repeated load triaxial tests were performed on six aggregate sources with three different gradations and three different moisture contents. The experimental results were analyzed within the framework of nonlinear cross-anisotropic elastic model in order to determine the model coefficients. Image analysis techniques were utilized to measure aggregate shape properties. The gradation and shape properties were fitted using a cumulative distribution function and nonlinear regression analysis, which is capable of capturing the complete distribution of these properties. The experimental and analytical results indicate that the vertical resilient modulus is greater than the horizontal resilient modulus and that aggregate physical properties significantly affect the anisotropic resilient behavior. Based on finite element analysis, the anisotropic resilient behavior has substantial effect on the critical pavement responses. Thus, it is extremely valuable to approximate the degree of cross-anisotropy in unbound aggregates and to use it as input in the pavement analysis programs to adequately model unbound aggregate bases for pavement design and analysis.