Comparative analysis of models for predicting permanent strain in unbound granular materials

Abstract

Unbound granular materials (UGMs) are essential for the durability of pavement structures, especially in the base and subbase layers of flexible pavements. This study aims to enhance our understanding of UGMs by evaluating widely-used models for predicting permanent strain under varying stress levels and load cycles. Models developed by Tseng and Lytton (1989), Gidel et al. (2001), Korkiala-Tantu (2005), and Fan Gu (2015) were assessed for their predictive accuracy. These models primarily focus on variables such as stress levels, the number of load cycles, and materialspecific properties. The analysis revealed that these variables significantly influence model performance. As a result, this study emphasizes the importance of considering a broad range of stress conditions for more accurate predictions of permanent strain in UGMs. In evaluating permanent deformation, the study examined key concepts, such as the relationship between permanent strain and the number of load cycles, the rate of permanent strain increase, and the shakedown phenomenon, in two types of UGMs: granite and limestone. An in-depth examination was conducted over 100,000 loading cycles, providing nuanced insights into UGMs behavior. Various factors, including the magnitude of applied stress, the number of load cycles, material density, and moisture content, among others, were identified as the most influencing permanent deformation. For the durable and resilient design of pavements, a comprehensive understanding of these variables is crucial. In this study, log-log graphs were essential in clarifying complex data relationships. This method enriched the analysis and improved the interpretability of our findings, adding both clarity and rigor to the research. This study provides crucial guidelines for choosing appropriate models for predicting UGMs behavior under expected stress and loading scenarios. The insights gained are essential for improving both the design and assessment of flexible pavement structures. Moreover, the research emphasizes the importance of continued efforts to enhance the accuracy of existing predictive models and to further our grasp of the behavior of UGMs.