Europe’s ageing transport infrastructure needs effective and proactive maintenance in order to continue its safe operation during the entire life cycle; European countries have to allocate huge resources for maintaining their service-ability level. This give rise to the necessity of an urgent need to adopt faster and more reliable monitoring inspection approaches to help tackling these issues. The deterioration of structures is most often foreseen by the formation of cracks on concrete surface. The presence of cracks can be a symptom of various problems like expansion and shrinks due to temperature differences, settlement of the structure, improper cover provided during concreting, corrosion of reinforcement steel, heavy load applied, insufficient vibration at the time of laying the concrete or loss of water from concrete surface shrinkage, therefore the identification, measurement and monitoring of cracks on the concrete surface becomes of primary importance. The main currently adopted inspection methods rely on visual marking and rulers, long and cumbersome activity, prone to errors and poorly objective on quantitative analysis because it strongly depends on operator experience. According to UNI EN 1992-1-1:2005 standard , the maximum admitted concrete crack width is 0.3 mm. For this reason, to accurately and reliably measure the target dimension, it is necessary to employ measurement instruments with suitable metrological characteristics (e.g. precision and accuracy at least one order lower than the value to be measured). Otherwise, the crack severity could be misclassified. This thesis work proposes a novel automatic image-based approach able to locate and measure cracks on concrete surfaces respecting the metrological constraint imposed by UNI EN 1992-1-1:2005 standard. Using only one image, the developed method is able to automatically and rapidly locate and measure the average width and length of a crack in an existing concrete structure. The measurement system developed exploits a single camera working in the visible range to acquire a digitized image of the structure being inspected. The software component of the system receives as input the single image framing the crack and gives as output an augmented image where the crack is highlighted as well as its average/max width and length. The measure of the crack width is performed perpendicularly to the crack central line with sub-pixel accuracy. The measurement system has been deployed on a smartphone for operator-based manual inspections as well on embedded systems for remote inspection with robots or Unmanned Aerial Vehicles (UAVs). The strategies developed can be easily extended from concrete inspection applications to any other context where a surface quality control targeted to the identification of eventual damages/defects is required. The activity was triggered by an explicit need within the EnDurCrete project.
In Europa, le infrastrutture civili e di trasporto necessitano di una manutenzione efficace e proattiva per garantire il continuo funzionamento in sicurezza durante l'intero loro ciclo di vita. I paesi europei devono ogni anno stanziare enormi risorse per mantenere il loro livello di funzionalità. Ciò fa sorgere la necessità urgente di adottare approcci di ispezione di monitoraggio più rapidi e affidabili per aiutare ad affrontare questi problemi. Il deterioramento delle strutture è più spesso anticipato dalla formazione di fessure sulla superficie del calcestruzzo. La presenza di fessurazioni può essere sintomo di diverse problematiche quali dilatazioni e ritiri dovuti a sbalzi di temperatura, assestamenti della struttura, copertura impropria fornita in fase di getto, corrosione delle armature in acciaio, carichi pesanti applicati, vibrazioni insufficienti al momento della posa del calcestruzzo o perdite d'acqua per ritiro superficiale del calcestruzzo. Diventa quindi di primaria importanza l'identificazione, la misurazione e il monitoraggio delle fessurazioni sulla superficie del calcestruzzo. I principali metodi di ispezione attualmente adottati si basano su strumenti manuali e righelli: un’attività lunga e ingombrante, soggetta a errori e scarsamente oggettiva sull'analisi quantitativa perché fortemente dipendente dall'esperienza dell'operatore. Secondo la norma UNI EN 1992-1-1:2005, la larghezza massima delle fessure del calcestruzzo ammessa per una generica classe di rischio è di 0,3 mm. Per questo motivo, per misurare in modo accurato e affidabile la dimensione della fessura, è necessario l’impiego di strumenti di misura con caratteristiche metrologiche adeguate (es. precisione e accuratezza almeno un ordine inferiore al valore da misurare). In caso contrario, la severità della fessura potrebbe essere classificata erroneamente. Questo lavoro di tesi propone un nuovo approccio automatico, basato su immagini, in grado di localizzare e misurare fessure su superfici in calcestruzzo rispettando il vincolo metrologico imposto dalla norma UNI EN 1992-1-1:2005. Utilizzando una sola immagine, il metodo sviluppato è in grado di localizzare e misurare automaticamente e rapidamente la larghezza e la lunghezza di una fessura su una superficie. Il sistema di misura sviluppato sfrutta una singola telecamera operante nel campo del visibile per acquisire un'immagine digitalizzata della superficie da ispezionare. Il componente software del sistema riceve in input la singola immagine che inquadra la crepa e fornisce in output un'immagine aumentata dove viene evidenziata la crepa e la sua larghezza e lunghezza media/max. La misura della larghezza della fessura viene eseguita perpendicolarmente alla linea centrale della fessura con una precisione sub-pixel. Il sistema di misurazione è stato implementato su uno smartphone per eseguire ispezioni manuali da parte dell'operatore e su sistemi integrati per l'ispezione remota con robot o velivoli senza pilota (UAV)). Le strategie sviluppate possono essere facilmente estese a qualsiasi altro contesto in cui sia richiesto un controllo di qualità superficiale mirato all'identificazione e misura di eventuali danni o difettosità.
Crack Analyser: a novel image-based NDT approach for measuring crack severity / Giulietti, Nicola. - (2022 Mar 21).
Crack Analyser: a novel image-based NDT approach for measuring crack severity
GIULIETTI, NICOLA
2022-03-21
Abstract
Europe’s ageing transport infrastructure needs effective and proactive maintenance in order to continue its safe operation during the entire life cycle; European countries have to allocate huge resources for maintaining their service-ability level. This give rise to the necessity of an urgent need to adopt faster and more reliable monitoring inspection approaches to help tackling these issues. The deterioration of structures is most often foreseen by the formation of cracks on concrete surface. The presence of cracks can be a symptom of various problems like expansion and shrinks due to temperature differences, settlement of the structure, improper cover provided during concreting, corrosion of reinforcement steel, heavy load applied, insufficient vibration at the time of laying the concrete or loss of water from concrete surface shrinkage, therefore the identification, measurement and monitoring of cracks on the concrete surface becomes of primary importance. The main currently adopted inspection methods rely on visual marking and rulers, long and cumbersome activity, prone to errors and poorly objective on quantitative analysis because it strongly depends on operator experience. According to UNI EN 1992-1-1:2005 standard , the maximum admitted concrete crack width is 0.3 mm. For this reason, to accurately and reliably measure the target dimension, it is necessary to employ measurement instruments with suitable metrological characteristics (e.g. precision and accuracy at least one order lower than the value to be measured). Otherwise, the crack severity could be misclassified. This thesis work proposes a novel automatic image-based approach able to locate and measure cracks on concrete surfaces respecting the metrological constraint imposed by UNI EN 1992-1-1:2005 standard. Using only one image, the developed method is able to automatically and rapidly locate and measure the average width and length of a crack in an existing concrete structure. The measurement system developed exploits a single camera working in the visible range to acquire a digitized image of the structure being inspected. The software component of the system receives as input the single image framing the crack and gives as output an augmented image where the crack is highlighted as well as its average/max width and length. The measure of the crack width is performed perpendicularly to the crack central line with sub-pixel accuracy. The measurement system has been deployed on a smartphone for operator-based manual inspections as well on embedded systems for remote inspection with robots or Unmanned Aerial Vehicles (UAVs). The strategies developed can be easily extended from concrete inspection applications to any other context where a surface quality control targeted to the identification of eventual damages/defects is required. The activity was triggered by an explicit need within the EnDurCrete project. File | Dimensione | Formato | |
---|---|---|---|
Tesi_Giulietti.pdf
Open Access dal 22/09/2023
Descrizione: Tesi_Giulietti
Tipologia:
Tesi di dottorato
Licenza d'uso:
Creative commons
Dimensione
33.09 MB
Formato
Adobe PDF
|
33.09 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.