Public defence: Kültigin Demirlioğlu

Background

Bridges play a vital role in facilitating safe and efficient transportation networks. Over their lifespan, they are subjected to various factors such as diverse environmental conditions, natural hazards, and heavy loads, all of which pose significant risks to their structural integrity, as evidenced by recent bridge failures. Consequently, there has been a growing interest in adopting Structural Health Monitoring (SHM) systems to continuously assess the condition of bridges. Among the various SHM methods, there is a clear trend towards embracing drive-by monitoring methods due to their cost-effectiveness and efficiency. These methods utilize sensors mounted on vehicles traveling across the bridge to capture its dynamic response. This eliminates the need for directly installing sensors on the bridge structure, as required in conventional SHM approaches.

Objectives

This PhD thesis has two primary objectives. The first objective is to explore the limitations of existing drive-by monitoring methods, particularly in identifying the natural frequencies and mode shapes of bridges. 

The second objective is to develop innovative drive-by monitoring techniques that can construct mode shapes and detect bridge damage more effectively. These advancements aim to address the identified limitations and to bridge the existing gap in the literature. 

Method

The research begins by investigating the limitations of current drive-by monitoring methods in accurately estimating the natural frequencies and mode shapes of simply supported bridges. It evaluates the precision of the methods using operational modal analysis and bridge modal components, emphasizing the importance of addressing road roughness effects for accurate mode shape recovery. 

The impact of road roughness on vehicle bridge interaction and the extraction of bridge-related dynamic information is further explored, leading to recommendations for optimal road roughness discretization to balance computational efficiency with the accuracy of numerical simulations.

Building on these outcomes, the study focuses on the efficacy of drive-by monitoring methods in estimating mode shapes for bridges supported by elastic bearings. By comparing various scenarios, including different vehicle speeds, road roughness effects, and traffic, the research highlights the challenges of estimating modal displacements near elastic supports. 

This study enhances our understanding of the limitations of the current methods and sets the foundation for the development of improved techniques. Subsequently, a novel method called Reference-based Component Scaling (RCS) is introduced for estimating bridge mode shapes. Utilizing its innovative framework, the RCS method stands alone as the only vehicle scanning method capable of accurately estimating the mode shapes of bridges, regardless of their symmetry, elastic supports, realistic damping ratios, and road roughness. 

Finally, a novel framework is introduced for identifying damage in bridges through continuous wavelet analysis of accelerations collected by two sensors installed on a moving vehicle. This method relies on changes in the static response of the bridge, demonstrating higher sensitivity to damage compared to its dynamic response. By focusing on the static response, the method eliminates the need for modal parameters, reduces the influence of road roughness, and can detect and locate damage even without data from an undamaged bridge.

Conclusion

As a result, this PhD thesis significantly enhances our understanding of utilizing drive-by monitoring approaches, with a particular focus on mode shape identification and damage detection. The contributions elevate the current state-of-the-art in drive-by monitoring and establish a solid foundation for future research and practical applications.