In the era of Industry 4.0 sensor and measurement technology is constantly developed further. New industrial applications become feasible and former rather black-box areas of interest become more transparent thanks to in-situ data acquisition. Where previously quality and machine health state analysis in running production systems was cumbersome, multivariate sensor based analysis today allows the generation of new process intelligence. Data driven decisions facilitate an efficient and objective production management. However, data becomes only beneficial if it is combined with process expertise, interpreted correctly and reduced to the relevant information. Additionally, in the manufacturing system various data sources need to be combined since certain dynamics on one side can only be explained by knowledge of the other. Hence, the importance of data management, transformation and analysis in the industrial environment rises at the same momentum as new metrology techniques are deployed. In this thesis the process of high pressure die casting is studied with a focus on the dynamics during the three characteristic phases: pre filling, die filling and holding pressure phase. High pressure die casting is a complex manufacturing process involving high pressure, velocities and temperatures. Many parameters and variables influence the casting process. To increase process knowledge one HPDC machine within the BMW Group foundry Landshut is equipped with in-die melt pressure sensors. This extends the process monitoring capabilities by a previously not considered area. The essential framework of this thesis is to acquire data from a high output series production under industrial conditions. Its objective is to analyze process variability as revealed by the in-die melt pressure. The acquired data in form of pressure time series is interpreted and described by the application of domain knowledge. The process variability is analyzed in a multivariate approach by the application of a vectorized principal component analysis. The findings are outlined and brought into relation to other available information, such as maintenance and casting quality. Conclusively, potential developments for future endeavors are compiled.
Nell’era dell’Industria 4.0, le tecnologie legate a sensori e sistemi di misurazione hanno avuto un costante sviluppo. L’acquisizione di dati in loco ha reso possibile l’implementazione di nuove applicazioni industriali, portando alla luce ambiti di interesse non ancora esplorati. Se in precedenza si è sempre fatto affidamento ad analisi complesse, basate sulla qualità e sullo stato di salute delle macchine, ad oggi analisi sensoriali multivariate hanno favorito lo sviluppo di tecniche efficienti per la gestione dei processi, dove le decisioni basate sui dati raccolti sono parte fondante dei risultati ottenuti. Tuttavia, l’elaborazione di tali dati risulta vantaggiosa solo se combinata con una buona conoscenza dei processi produttivi, in modo da raccogliere informazioni rilevanti e interpretarle correttamente. Inoltre, un’analisi che consideri diverse sorgenti di dati è necessaria per fare fronte alle dinamiche dei sistemi di produzione. Pertanto, nel contesto industriale, lo sviluppo della gestione, analisi ed elaborazione dei dati è stata accompagnato dalla nascita di nuove tecniche metrologiche. Il lavoro di tesi ha come obiettivo quello di analizzare un processo di pressofusione ad alte pressioni, ponendo particolare attenzione alla dinamica delle tre fasi principali: precarico, riempimento dello stampo e mantenimento della pressione. La complessità di questo processo produttivo è dovuta alle alte pressioni, alla velocità e alle temperature che caratterizzano la tecnologia. Con lo scopo di monitorare in maniera più efficiente il processo di pressofusione, un sensore di pressione è stato inserito nello stampo di una macchina HPDC all’interno dell’impianto produttivo di BMW, nella località di Landshut. Mirando ad analizzare nello specifico la variabilità del processo, resa nota dal sensore di pressione, il lavoro di tesi si presta a documentare i dati provenienti dalle numerose uscite del processo produttivo. Questi sono stati inoltre interpretati e descritti a valle di una dettagliata conoscenza relativa alla tecnologia considerata, mentre la variabilità del processo è stata studiata con un approccio multivariato, tramite un’analisi delle componenti principali. I risultati sono stati riportati e condotti in relazione ad informazioni legate alla manutenzione e qualità del processo di pressofusione. Infine, sono state avanzate alcune proposte nell’ottica di sviluppi futuri.
Process monitoring of high-pressure die casting with in-die pressure sensors
ZEISER, ALEXANDER
2018/2019
Abstract
In the era of Industry 4.0 sensor and measurement technology is constantly developed further. New industrial applications become feasible and former rather black-box areas of interest become more transparent thanks to in-situ data acquisition. Where previously quality and machine health state analysis in running production systems was cumbersome, multivariate sensor based analysis today allows the generation of new process intelligence. Data driven decisions facilitate an efficient and objective production management. However, data becomes only beneficial if it is combined with process expertise, interpreted correctly and reduced to the relevant information. Additionally, in the manufacturing system various data sources need to be combined since certain dynamics on one side can only be explained by knowledge of the other. Hence, the importance of data management, transformation and analysis in the industrial environment rises at the same momentum as new metrology techniques are deployed. In this thesis the process of high pressure die casting is studied with a focus on the dynamics during the three characteristic phases: pre filling, die filling and holding pressure phase. High pressure die casting is a complex manufacturing process involving high pressure, velocities and temperatures. Many parameters and variables influence the casting process. To increase process knowledge one HPDC machine within the BMW Group foundry Landshut is equipped with in-die melt pressure sensors. This extends the process monitoring capabilities by a previously not considered area. The essential framework of this thesis is to acquire data from a high output series production under industrial conditions. Its objective is to analyze process variability as revealed by the in-die melt pressure. The acquired data in form of pressure time series is interpreted and described by the application of domain knowledge. The process variability is analyzed in a multivariate approach by the application of a vectorized principal component analysis. The findings are outlined and brought into relation to other available information, such as maintenance and casting quality. Conclusively, potential developments for future endeavors are compiled.File | Dimensione | Formato | |
---|---|---|---|
2019_12_Zeiser.pdf
non accessibile
Descrizione: 2019_12_Zeiser TESI
Dimensione
17.11 MB
Formato
Adobe PDF
|
17.11 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/10589/151421