Advancing the Use of the Damage Rating Index (DRI) for Evaluation of Concrete Affected by Internal Swelling Reactions (ISR)

Abstract

The damage rating index (DRI) is a microscopy tool and methodology used to identify deterioration caused by internal swelling reactions (ISR) in concrete. The DRI was initially developed to evaluate concrete affected by alkali-aggregate reaction (AAR) and has since begun to be employed for other types of ISR. It utilizes a point-count method to count distinct petrographic and distress features across a surface, providing insight into the type, extent, and pattern of deterioration. This study addresses research gaps related to the DRI’s ability to distinguish between ISR mechanisms, the connection between microscopy and mechanical responses, operator variability, and the use of technology for more efficient data collection. This thesis is paper-based and includes five papers aimed at addressing the identified research gaps. The first topic is divided into two papers, with the first exploring the tool's adaptability by modifying the counting method to better describe deterioration due to delayed ettringite formation (DEF). The study illustrates the enhanced capability of the DRI to characterize DEF-induced deterioration more effectively. The second paper assesses the counting method in relation to mechanical responses. The second topic and third paper investigates the relationship between distress features and mechanical responses. The research employs linear regression through detailed analysis, establishing significant correlations that provide valuable insights into the engineering implications of distress features observed under microscopy. The third topic is variability, and the fourth paper presents an approach for autonomous operator training aimed at improving skill acquisition and reducing variability. This approach proposes a robust training framework, demonstrating promising results in standardizing operator performance. The final topic and fifth paper evaluate a technique for assessing the outcomes of automating the collection of distress feature counts via images. The study compares the accuracy and efficiency of operator-based versus machine-based data collection, highlighting the advantages of automation in terms of consistency and speed. The combined findings from these studies offer a comprehensive understanding of the DRI’s adaptability to ISR mechanisms, distress feature analysis, autonomous training methodologies, and automated evaluation techniques. Future research can build upon these findings and approaches to further innovate and refine the use of the DRI in engineering practices.