Nowadays, aluminothermic welding has become the most diffused technique for joining two rails. This process has almost totally substituted the previous bolted joints, since it requires less maintenance and it provides also higher comfortability, due to absence of discontinuities between rails. On the other side, one of the most important criticism of welded rails stands in the failure mode, which, most of times, is represented by semi-elliptical fatigue crack propagation, terminated with brittle fracture. Therefore, the present thesis aims to measure this kind of defects, by means of Non-Destructive Testing; in particular, Conventional Ultrasonic Testing (UT) and Phased Array (PA) will be adopted. They are able to store the response amplitude of the defect related to its depth, according to a certain number of decibels. It will be considered the worst case for this type of NDT, which occurs when defects are located at rail’s foot. Three main results will be provided. First of all, we will show how, according to defect positions, UT and PA performances change; thus, we will provide an indication regarding the best method for each case. Secondly, after comparing experimental results for both smooth rail and long-welded rail, we will demonstrate how the microstructure of the molten zone is able to modify the response amplitude. Finally, a reliability study in terms of Probability of Detection (PoD) will be provided for each defect, and it will demonstrate how low depth defects are hard to be detected, independently from the chosen method.
Oggigiorno, la saldatura alluminotermica è diventata la tecnica più diffusa per collegare due rotaie. Questo processo ha quasi del tutto sostituito le precedenti giunzioni bullonate, perché richiede meno manutenzione e fornisce anche maggior comfort, dovuto all’assenza di discontinuità tra le rotaie. Dall’altra parte, una delle criticità più importanti delle rotaie saldate risiede nel tipo di rottura che, molte volte, è rappresentato da una propagazione di cricche semiellittiche di fatica, che termina con una frattura fragile. Per questo motivo, il presente lavoro si pone di misurare questo tipo di difetti, per mezzo di Controlli Non Distruttivi (CND): in particolare, Ultrasonic Testing convenzionale (UT) e Phased Array (PA) saranno utilizzati. Questi riescono a registrare l’ampiezza di risposta del difetto in relazione alla sua profondità, fissato un certo numero di decibel. Sarà considerato il peggior caso per questo tipo di CND, che avviene quando i difetti sono localizzati sul piede della rotaia. Tre principali risultati verranno provati. Prima di tutto, mostreremo come, in base alla posizione del difetto, le performance di UT e PA cambiano; per questo, forniremo un’indicazione riguardo al miglior metodo da scegliere per ogni caso. In secondo luogo, dopo aver paragonato i risultati sperimentali di rotaia liscia e rotaia saldata, dimostreremo come la microstruttura della zona fusa può modificare l’ampiezza di risposta. Per finire, uno studio di affidabilità in termini di Probability of Detection (PoD) sarà fornito per ogni difetto, e sarà dimostrato come sia difficile individuare difetti di bassa profondità, indipendentemente dal metodo scelto.
Reliability of phased array ultrasonic inspections applied to welded railway rails
SERAFINI, ITALO
2018/2019
Abstract
Nowadays, aluminothermic welding has become the most diffused technique for joining two rails. This process has almost totally substituted the previous bolted joints, since it requires less maintenance and it provides also higher comfortability, due to absence of discontinuities between rails. On the other side, one of the most important criticism of welded rails stands in the failure mode, which, most of times, is represented by semi-elliptical fatigue crack propagation, terminated with brittle fracture. Therefore, the present thesis aims to measure this kind of defects, by means of Non-Destructive Testing; in particular, Conventional Ultrasonic Testing (UT) and Phased Array (PA) will be adopted. They are able to store the response amplitude of the defect related to its depth, according to a certain number of decibels. It will be considered the worst case for this type of NDT, which occurs when defects are located at rail’s foot. Three main results will be provided. First of all, we will show how, according to defect positions, UT and PA performances change; thus, we will provide an indication regarding the best method for each case. Secondly, after comparing experimental results for both smooth rail and long-welded rail, we will demonstrate how the microstructure of the molten zone is able to modify the response amplitude. Finally, a reliability study in terms of Probability of Detection (PoD) will be provided for each defect, and it will demonstrate how low depth defects are hard to be detected, independently from the chosen method.File | Dimensione | Formato | |
---|---|---|---|
TesiDefinitiva.pdf
non accessibile
Dimensione
5.87 MB
Formato
Adobe PDF
|
5.87 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/146601