CONSIDERATION OF STRAIN AT FAILURE AND STRENGTH IN PREDICTION OF PAVEMENT THERMAL CRACKING

Abstract

Theoretical principles of viscoelasticity that are needed for the prediction of thermal cracking of asphalt pavements are reviewed in this study. The assumptions of simplification made in the existing literature devoted to the prediction of pavement cracking are examined. It is found that, although in the studies the complexity of this phenomenon is well realized, the proposed prediction methods include several simplifications. These studies also assume that strength of the asphalt mixture or binder is the only failure criterion needed to predict cracking. These methods do not consider the possibility that strain at failure may control the cracking. It is also fount that in the majority of these methods the affect of cooling rate on strength of asphalt binders and mixtures is not considered. Furthermore, the glass transition behavior know for asphalt materials has not been integrated in the calculations of thermal strains and the resulting stresses in pavements. In this study attempts were made to introduce improved procedures to account for the characteristics of failure strain and strength, for the cooling rate dependency of strength and failure strain, and for the glass transition behavior of asphalt binders and mixtures. The study is focused on bringing the attention to the possible role of failure strain in the prediction using the proposed procedures are presented for one neat binder and one polymer-modified binder. The predicted cracking temperatures based on failure strain are compared with based on the conventional concept of thermal stress and strength. Results of the study indicate that strain tolerance characteristics of asphalt binders, dependency of failure properties (strain and strength) on cooling rate, and glass transition behavior could be effectively integrated in prediction models for a more reliable method. The results also indicate that failure strain is an important criterion that needs to be considered in pavement cracking prediction and can result in more critical cracking temperatures than strength. These results, although admittedly based on theoretical analysis and limited laboratory data, merit significant validations in the field.