1 December 2007 in Mechanics of Time-Dependent Materials
1 January 2008 in Road Materials and Pavement Design
This paper presents the use of three dimensional (3D) finite element modeling (FEM) techniques to determine asphalt pavement response to loading. In mechanistic-empirical pavement design, the properties of materials used in pavement layers must be specified before the pavement response to imposed loads can be determined. A numerical method is used to obtain the relaxation modulus of linear viscoelastic materials, such as asphalt concrete, from the dynamic modulus measured on specimens procured from the field. Linear elastic behavior is assumed for granular materials. In addition, regarding finite element modeling, key factors such as model geometry, material properties, load and boundary conditions, element type, and mesh refinement are discussed in detail. The adopted Global-Local (GL) FEM approach is capable of simulating pavement responses to multiple axle loads with different load configurations, speed, and temperatures.
1 April 2010 in International Journal of Pavement Engineering
The objective of this study is to evaluate the creep compliance (D(t)) of asphalt concrete (AC) mixtures for thermal cracking prediction of flexible pavements. Various AC overlay design factors influencing the thermal cracking resistance of flexible pavements, such as mixture properties and pavement structure, were included in the evaluation. Two sources of D(t) data were considered: (1) measurement from indirect tensile test and (2) numerical interconversion of complex modulus E*. Design input levels in the mechanistic–empirical design guide software do significantly impact the predicted thermal cracking distresses. For AC maintenance overlay design purposes, level 1 and 2 analyses yield very similar thermal cracking predictions, whereas level 3 analysis significantly underpredicts the extent of cracking when compared with to level 1 analysis. Some discrepancy exists in the thermal cracking predicted from measured and interconverted D(t) due to the inherent approximation nature of numerical interconversion methods. However, level 1 and 2 analyses using interconverted D(t) values provide results closer to the predictions from measured D(t) values than does level 3. Three mixtures are used in this study. The various analyses indicate that using a more ductile AC mixture for the surface layer significantly reduces the amount of thermal cracking at failure.