Electron Beam Melting (EBM) is an additive manufacturing technology that has gained a relevant momentum for industrial use, especially for the production of titanium components in the aerospace sector and biomedical implants. EBM allows one to produce complex parts, from highly customized components to series production with some advantages against competitor methods, namely the highest productivity and the capability to produce challenging materials like brittle intermetallic alloys. Up to date, process instability is still a crucial issue in EBM, which could generate defective products or early termination of the process, with consequent losses of time and wastes of resources. Moreover, the lack of repeatability represents a major problem, given the stringent quality requirements in above in the aerospace and biomedical sectors. In this framework, the goal of this work is to investigate novel methodologies to detect the onset of defects during the EBM process, by combining in-situ sensing via machine vision with image-based statistical modelling and monitoring tools. In particular, two different methodologies for the in-line detection of geometrical distortions were developed, based on the reconstruction of the contour of the printed slice in each layer. Both of the methodologies rely on the acquisition of high resolution videos during the process, and their performances have been tested and compared by means of a build composed by Ti6Al4V specimens. Moreover, through the design of a custom camera mount device, a high-speed camera was installed on the EBM system, leading to the first attempt to keep under control the process stability by means of high-speed video imaging in EBM. A preliminary study on the detection of hot-spots, i.e., regions of the slice that stay hot for a long time with anomalous cooling patterns are presented and discussed.
L’Electron Beam Melting (EBM) è un processo additivo che in questi ultimi decenni ha avuto uno slancio significativo nell’impiego industriale, specialmente per quanto riguarda la lavorazione di componenti in titanio nel settore aerospaziale e biomedico. L’EBM consente di realizzare sia componenti complessi con ottima flessibilità, sia produzioni in serie sfruttando alcuni vantaggi propri della tecnologia stessa come l’alta produttività e la capacità di processare materiali difficili da lavorare (fragili). Attualmente l’instabilità del processo rappresenta un’importante criticità che porta alla generazione di difetti o interruzioni del processo stesso con il conseguente spreco di tempo e risorse. Inoltre, la mancanza di ripetibilità rappresenta un grande ostacolo dati i rigorosi requisiti di qualità propri del campo aerospaziale e medico. In tale contesto, l’obbiettivo di questa tesi è quello di approfondire nuovi metodi atti all’identificazione dell’insorgenza di difetti durante il processo EBM, combinando al monitoraggio in-situ della macchina una modellazione statistica basata sull’analisi delle immagini. Nello specifico, sono state sviluppate due metodologie differenti per il riconoscimento in linea di difetti geometrici basati sulla ricostruzione dei contorni di ogni strato processato. Entrambi gli approcci si basano sull’acquisizione in alta risoluzione di video durante il processo e i risultati sono stati comparati e validati attraverso una stampa di provini in Ti6Al4V. Successivamente, grazie alla progettazione di un supporto personalizzato, è stata resa possibile l’installazione di una camera ad alta velocità che ha portato al primo tentativo di controllo della stabilità di un processo EBM attraverso video ad alta velocità, unico nel suo genere. Infatti, a seguire, è presentato e discusso uno studio preliminare sull’identificazione degli hot-spot, porzioni di superficie che subiscono un processo di raffreddamento anomalo.
Optical in-situ process monitoring for electron beam melting additive manufacturing
VALSECCHI, GIORGIO;CASOLO, LORENZO
2018/2019
Abstract
Electron Beam Melting (EBM) is an additive manufacturing technology that has gained a relevant momentum for industrial use, especially for the production of titanium components in the aerospace sector and biomedical implants. EBM allows one to produce complex parts, from highly customized components to series production with some advantages against competitor methods, namely the highest productivity and the capability to produce challenging materials like brittle intermetallic alloys. Up to date, process instability is still a crucial issue in EBM, which could generate defective products or early termination of the process, with consequent losses of time and wastes of resources. Moreover, the lack of repeatability represents a major problem, given the stringent quality requirements in above in the aerospace and biomedical sectors. In this framework, the goal of this work is to investigate novel methodologies to detect the onset of defects during the EBM process, by combining in-situ sensing via machine vision with image-based statistical modelling and monitoring tools. In particular, two different methodologies for the in-line detection of geometrical distortions were developed, based on the reconstruction of the contour of the printed slice in each layer. Both of the methodologies rely on the acquisition of high resolution videos during the process, and their performances have been tested and compared by means of a build composed by Ti6Al4V specimens. Moreover, through the design of a custom camera mount device, a high-speed camera was installed on the EBM system, leading to the first attempt to keep under control the process stability by means of high-speed video imaging in EBM. A preliminary study on the detection of hot-spots, i.e., regions of the slice that stay hot for a long time with anomalous cooling patterns are presented and discussed.| File | Dimensione | Formato | |
|---|---|---|---|
|
2019_07_Casolo_Valsecchi.pdf
non accessibile
Descrizione: Testo della Tesi
Dimensione
32.86 MB
Formato
Adobe PDF
|
32.86 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/10589/148767