Characteristics of damaged asphalt mixtures in tension and compression

Abstract

This paper addresses the measurement and modelling of the damaged properties of asphalt mixtures including the fracture, healing and viscoplastic deformation of the asphalt mixtures in both tensile and compressive loading as being affected by their composition and conditioning with ageing and exposure to temperature and moisture. An energy-based mechanics is applied to obtain the material fundamental properties such as surface energies, bond energies, anisotropy, yield functions and plastic potential functions that are valid for actual asphalt mixtures, viscoelastic crack growth criteria under both tensile and compressive loading, a simple mechanics-based method of determining the fatigue endurance limit, and the measurement and prediction of healing in restoring the damage done by fracture. Healing is anti-fracture and cracking is the net result of the interplay of these two complimentary mechanisms. Because fracture in asphalt mixtures is not the growth of a single crack but the simultaneous growth of multiple cracks that start out as air voids, this fact leads to the use of the growth of damage density to characterise fracture in an asphalt mixture. It was discovered that the form of Paris’ law applies to the growth of damage density of asphalt mixtures in both tensile and compressive loadings. The importance of this fact lies in many developments from this discovery, e.g. compressive monotonic loading of cylindrical test samples permits a direct determination of the Paris’ Law coefficient and exponent. In all cases, measured material properties are presented as they vary with mixture composition and with conditioning such as moisture and ageing, both in the lab and in the field. The measurements of these properties are made simply, quickly and accurately by the use of mechanics so that an entire characterisation of the properties of an asphalt mixture in tension and compression can be completed in the space of one day. The net effect is to reduce the efforts expended in the lab and the systematic error due to the assumptions made by the existing models and simultaneously to increase the efficiency and cost-effectiveness of materials testing and raise the reliability of the design of mixtures, pavement structures and specifications and the prediction of the life cycles in as-built pavements.