Computer vision-based structural health monitoring and condition assessment for small to medium-span bridges

Tools

Obiechefu, C.B., 2022. Computer vision-based structural health monitoring and condition assessment for small to medium-span bridges. PhD, Nottingham Trent University.

Preview

Text
Final thesis.pdf - Published version
Download (5MB) | Preview

Abstract

Many UK bridges are currently nearing the end of their life cycles, and therefore their maintenance and inspection routines assume a higher priority, since they become more likely to fail. Bridge maintenance routines are still predominantly based on visual inspections which are expensive, time and labour consuming and prone to human error. Structural Health Monitoring (SHM) systems, mostly utilising contact sensors have been introduced to some larger structures such as the Humber and Severn bridges to complement visual inspections. Contact sensing, however, requires access to the structure, involves working at height and is expensive. More importantly, while many larger structures have received the aid of SHM, older assets, especially small bridges with small traffic volumes still rely solely on visual inspections due to high costs associated with such SHM systems. These challenges can be circumvented by employing the fast-emerging computer-vision based structural health monitoring systems (CV-SHM), which are much more affordable, can be set up to not require working at height, nor require direct access to the structure, and which do not cause traffic disruptions.

This thesis proposes an affordable and accurate CV-SHM and damage detection system to complement conventional bridge inspection routines for small to medium-size bridges. The framework comprises firstly, a computer vision (CV)-based sensing system which consist of a consumer-grade image acquisition device such as a GoPro or smartphone camera, a computer, and an image processing algorithm. This system obtains structural response without access to the structure by using image processing and feature detection techniques to analyse images of a structure captured during loading and obtain its structural response. The second part of this framework is the response analysis for damage detection and characterisation. Here, a set of data-based techniques are developed based on response information such as displacements, curvatures, inclination angles, and strains. Displacements serve as primary response obtained from the monitoring process. Others (strain, curvatures, and inclination angles) are secondary responses obtained by manipulating the primary. In this approach, response from any section of the structure can be analysed, without the requirement for further structural information, such as flexural rigidity (EI), or distance to support, unlike similar studies in literature. The condition of a structure can then be determined by comparing response measurements collected at a first inspection Others (strain, curvatures, and inclination angles) are secondary responses obtained by manipulating the primary. In this approach, response from any section of the structure can be analysed, without the requirement for further structural information, such as flexural rigidity (EI), or distance to support, unlike similar studies in literature. The condition of a structure can then be determined by comparing response measurements collected at a first inspection (

Item Type:	Thesis
Creators:	Obiechefu, C.B.
Date:	December 2022
Rights:	This thesis has been submitted to Nottingham Trent University (NTU) for the degree of PhD. I declare that the work in this thesis was carried out following the regulations of NTU. I certify that all material in this thesis which is not my own work has been referenced and that no material has previously been submitted and approved for the award of a degree by this or any other University. The research reported in this thesis was conducted at Nottingham Trent University between October 2017 and December 2022. Any views expressed in this thesis are those of the author and in no way represents the university.
Divisions:	Schools > School of Architecture, Design and the Built Environment
Record created by:	Linda Sullivan
Date Added:	13 Jun 2023 10:49
Last Modified:	13 Jun 2023 10:49
URI:	https://irep.ntu.ac.uk/id/eprint/49192