Back to search

Democritus University of Thrace (DUTH)

Έλεγχος και επεμβάσεις δομικών στοιχείων από ωπλισμένο σκυρόδεμα με ειδικά τσιμεντοκονιάματα και ευφυή υλικά

Abstract

dc:description

Civil infrastructures are essential long-term investments of a nation, and the proper operations of these are fundamental for the well-being of society. Therefore, extending the lifetime of existing infrastructure and structures without compromising their safety levels is paramount. Thus, developing innovative inspection and refurbishment methods for reinforced concrete structural elements is at the forefront of research and practical interest. The present doctoral thesis investigates the utilization of smart piezoelectric materials for real-time inspection of reinforced concrete structural elements and monitoring their structural integrity. In addition, the efficacy of smart piezoelectric patches for assessing cement-based mortars and Fiber Reinforced Polymers (FRP), commonly used to refurbish reinforced concrete structural members, and their performance as reinforcement materials was investigated. In this scope, the thesis introduces new innovative methodologies for acquiring and evaluating the electromechanical responses of piezoelectric sensors. The Electromechanical Impedance method captures the variation of the electromechanical response of a piezoelectric patch attached to the structural element due to the degradation of its structural integrity. Therefore, the Electromechanical Impedance method is accompanied by a technique for identifying changes in the electromechanical responses of the piezoelectric patch and correlating them with the degradation of the structural integrity of the structural element. In the international literature, the changes in the electromechanical responses were recognized using statistical indices, with RMSD and MAPD being the most widespread. In this doctoral thesis, two new methodologies for inspecting the structural integrity of reinforced concrete structural elements, which utilize machine learning techniques, were developed and investigated. The first methodology introduced in this thesis employs the hierarchical clustering method to identify the degradation of the structural integrity of the structural element from the changes in the responses of the piezoelectric sensors. The hierarchical clustering method is classified as an unsupervised machine learning method, so its application requires a smaller amount of data. Contrary, the second methodology introduced in this thesis was based on the development of a one-dimensional model Convolutional Neural Network (CNN), which belongs to the category of supervised machine learning methods. Although it requires a larger amount of data to train the CNN model, it exhibits an impressively high level of accuracy and reliability. In order to evaluate the proposed methodologies, an extensive experimental program was carried out, in which 3.142 electromechanical responses from 62 piezoelectric sensors were acquired. The Electromechanical Impedance method captures the variation of the electromechanical response of a piezoelectric patch attached to the structural element due to the degradation of its structural integrity. Therefore, the Electromechanical Impedance method is accompanied by a technique for identifying changes in the electromechanical responses of the piezoelectric patch and correlating them with the degradation of the structural integrity of the structural element. In the international literature, the changes in the electromechanical responses were recognized using statistical indices, with RMSD and MAPD being the most widespread. In this doctoral thesis, two new methodologies for inspecting the structural integrity of reinforced concrete structural elements, which utilize machine learning techniques, were developed and investigated. The first methodology introduced in this thesis employs the hierarchical clustering method to identify the degradation of the structural integrity of the structural element from the changes in the responses of the piezoelectric sensors. The hierarchical clustering method is classified as an unsupervised machine learning method, so its application requires a smaller amount of data. On the contrary, the second methodology introduced in this thesis was based on the development of a one-dimensional model Convolutional Neural Network (CNN), which belongs to the category of supervised machine learning methods. Although it requires a larger amount of data to train the CNN model, it exhibits an impressively high level of accuracy and reliability. In order to evaluate the proposed methodologies, an extensive experimental program was carried out, in which 3.142 electromechanical responses from 62 piezoelectric sensors were acquired. The experimental program is divided into two parts. The first part includes four cylindrical fiber-reinforced concrete mortar specimens under monotonic and repeatable compressive loading. Three piezoelectric sensors were attached to the external specimens to inspect their structural integrity. The results highlight the ability of piezoelectric sensors to promptly recognize the degradation of the structural integrity of fiber-reinforced concrete mortar specimens. In addition, the induced stress influences the electromechanical responses of the piezoelectric patches, which is strongly reflected in the statistical indicators of damage. Therefore, under fluctuating stress conditions, the statistical indices present a high probability of presenting false alarms, reducing their reliability to some extent. On the contrary, the proposed methodologies, which utilize machine learning techniques in structural integrity inspection, exhibit elevated reliability even under variant loading levels. In addition, three prismatic specimens of standard dimensions were constructed and subjected to four-point bending under repeated loading. Piezoelectric sensors were attached to the test specimens in several network configurations to optimize their performance in the inspection of their structural integrity. Additionally, two FRC beams were constructed and tested under repeated four-point bending loading. The experimental results show the beneficial influence of adding discrete fibers on concrete flexural strength and post-peak behavior. The Electromechanical Impedance method demonstrates encouraging results in the structural integrity inspection of fiber-reinforced specimens under four-point bending. It can be concluded that there is a correlation between the distance of the position of the piezoelectric sensor and the position of the formation of the critical crack in its successful and valid detection. Hence, the application of a dense network of piezoelectric sensors increases the probability of successful and prompt detection of the degradation of the structural integrity of the specimen. The second part includes the investigation of a new reinforcement technique in real-scale reinforced concrete structural elements. Firstly, two reinforced concrete beam-column joint sub-assemblages were constructed with similar geometric characteristics and reinforcement details. One of the beam-column joints was strengthened by applying C-FRP ropes to specially shaped grooves on the surface of the joint body and column. Furthermore, piezoelectric sensors were attached to the external surface of the joint body as well as to the C-FRP ropes of the reinforced specimen to check its structural integrity. The proposed strengthening technique shows encouraging results, increasing the load-bearing capacity, energy absorption and hysteretic behavior in comparison to the reference specimen. The results of the application of piezoelectric sensors demonstrate their efficacy for the inspection of the structural integrity of beam-column joints and the proposed refurbishment technique. In addition, two reinforced concrete deep beams were constructed and reinforced with C-FRP ropes. One of the deep beams had a rectangular cross-section, while the second one had a slab cross-section. Piezoelectric sensors were placed on the deep beams to inspect their structural integrity and monitor the refurbishment performance. The proposed strengthening technique significantly increased the load-carrying and displacement capacity of the deep beams and changed the failure mode of the deep beams to flexural.

Degree

thesis:*
Grantor dc:publisher
Democritus University of Thrace (DUTH)
Year dc:date
2024

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Sapidis, Georgios
  • Σαπίδης, Γεώργιος

Subjects

dc:subject × 39

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/56003

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Sapidis, Georgios; Σαπίδης, Γεώργιος. Έλεγχος και επεμβάσεις δομικών στοιχείων από ωπλισμένο σκυρόδεμα με ειδικά τσιμεντοκονιάματα και ευφυή υλικά. Democritus University of Thrace (DUTH), 2024. http://hdl.handle.net/10442/hedi/56003