This prospective study was designed to implement the use of Digital Image Correlation to investigate the behaviour of a cylindrical silicone specimen subjected to uniaxial compression, through a custom-designed support system. Concurrently, the load applied on the cylinder was monitored by means of a piezometric load cell located inside the support structure and used to perform both static and dynamic measurements, thus providing a complete picture of the strain of the specimen. The analysis of the sample deformation was conceived to be carried out thanks to an internal patterning of the specimen itself, realized by mixing steel powder directly into the silicone rubber. This approach to DIC represents the most noteworthy aspect of the present research and allows to achieve an accurate measurement of the sample deformation under stress. The support system, specifically manufactured for this study, represents the second focus point. It is devised to achieve two main functions: holding both the specimen and the load cell while maintaining their alignment and allowing a controlled compression force to be applied through a screw-driven mechanism. The main challenge faced during its design process was guaranteeing structural stiffness while providing a clear view of the sample, which is crucial for DIC implementation. The experimental procedure followed a loading sequence, during which the specimen was subjected to two different loads. The increasing force was recorded via the load cell and pictures for the DIC were simultaneously captured by two cameras. Image acquisition and processing was performed thanks to VIC-3D system, a turnkey equipment including the necessary hardware and software. The most challenging task during the DIC analysis was enabling the software to perform the correlation correctly and optimizing its accuracy. To achieve that, two steps have been fundamental: firstly, the acquired images were processed by enhancing contrast and saturation; secondly, a thorough trial-and-error process was carried out in order to find the optimal computational parameters (i.e. subset size and step size). It has to be mentioned that the actual speckle pattern used for the correlation was not the one originally intended, i.e. the steel powder: as a matter of fact, the accidental air bubbles trapped in the silicone turned out to provide a much better correlation than the powder particles, that were neither large nor numerous enough to ensure a successful measurement. Lastly, the resulting data were analysed and compared with the expected mechanical response of the specimen. The examination of the data retrieved from the load cell shows some discrepancies with the expected outcome, especially during the static measurements: the inconsistencies are likely due to the piezometric nature of the load cell. In fact, the dynamic measurements provided more accurate results. However, the most noteworthy data are the ones obtained through DIC: the resulting deformation fields closely match the predicted ones, both in distribution and in magnitude. It is worth mentioning that DIC was also able to detect a slight asymmetry of the load, invisible to the naked eye. Unfortunately, the resolution of the DIC measurement is not remarkable and the deformation maps show some evident discontinuities: these are conceivably caused by lack of contrast and consistency of the accidental speckle pattern provided by the air bubbles. Overall, this study can be considered successful: the custom support structure correctly transmitted the compression load to the specimen and the deformation fields derived from DIC, even though not exceptionally accurate, complied with the expected outcome. In spite of its limitations, this paper offers some insight into DIC technology and its implementation, as well as into the use of internal patterning and its possible shortcomings.
Questa tesi sperimentale si propone di implementare la Digital Image Correlation (DIC) al fine di studiare il comportamento di un campione cilindrico in silicone sottoposto a compressione uniassiale tramite un sistema di supporto appositamente progettato. Inoltre, il quadro completo della deformazione viene fornito dall’utilizzo parallelo di una cella di carico piezometrica collocata all’interno della struttura di supporto; impiegata per eseguire misurazioni sia statiche sia dinamiche, essa monitora costantemente il carico applicato sul cilindro. L’analisi della deformazione del campione viene computata grazie a uno speckle pattern interno al campione stesso, ottenuto mescolando polvere di acciaio direttamente nella miscela bicomponente del silicone. Questo approccio alla DIC rappresenta l’aspetto più rilevante della presente ricerca, in quanto consente di analizzare la deformazione da un punto di vista diverso e ottenerne una misura accurata. Il sistema di supporto, realizzato specificamente per questo studio, costituisce il secondo punto focale. Il design della struttura è pensato per svolgere due funzioni principali: da un lato sostenere sia il campione sia la cella di carico, mantenendone l’allineamento, e dall’altro permettere l’applicazione della forza di compressione tramite un meccanismo a vite. La procedura sperimentale consiste in una precisa sequenza di carico durante la quale il campione è stato sottoposto a due diverse forze deformanti: l’aumento della forza è stato registrato mediante la cella di carico, mentre due telecamere sincronizzate acquisivano le immagini per la DIC. Al fine di ottimizzare i risultati della DIC sono state fondamentali due operazioni: prima le immagini acquisite sono state processate per migliorarne contrasto e saturazione; poi è stato condotto un accurato processo per trial-and-error per determinare i parametri computazionali ottimali. Tuttavia, lo speckle pattern effettivamente utilizzato non è stato quello previsto inizialmente, ovvero la polvere di acciaio: le particelle di quest’ultima, infatti, sono risultate troppo piccole e troppo poco numerose per garantire una misura affidabile. Al contrario, le bolle d’aria accidentalmente intrappolate nel silicone si sono rivelate efficaci nel fornire una buona correlazione e hanno perciò assunto il ruolo di speckle pattern. I dati risultanti sono stati poi analizzati e confrontati con la deformazione attesa. L’analisi dei dati ottenuti dalla cella di carico mostra alcune discrepanze rispetto ai risultati previsti, probabilmente dovute alla natura piezometrica della cella di carico; a prova di ciò, le misure dinamiche hanno fornito risultati più accurati rispetto a quelle statiche. Tuttavia, i dati più significativi sono quelli ottenuti tramite la DIC: i campi di deformazione risultanti corrispondono in buona misura a quelli attesi, sia per distribuzione che per entità. È interessante notare come la DIC sia stata in grado di rilevare anche una leggera asimmetria del carico rispetto all’asse del cilindro, invisibile a occhio nudo. Purtroppo, la risoluzione della DIC non è stata particolarmente elevata e le mappe di deformazione mostrano alcuni evidenti discontinuità: tali difformità sono imputabili alla scarsa omogeneità e al basso contrasto dello speckle pattern formato dalle bolle d’aria, caratteristiche giustificabili data la sua natura fortuita. Nel complesso, questo studio può essere considerato riuscito: la struttura di supporto ha trasmesso con successo la forza di compressione al campione e i campi delle deformazioni derivati dalla DIC, anche se non troppo accurati, risultano coerenti con quelli previsti. Nonostante le sue limitazioni, questo lavoro offre un’interessante introduzione alla tecnologia della DIC e alla sua applicazione, oltre che un contributo alla comprensione dell’uso di uno speckle pattern interno e delle sue possibili criticità.
SVILUPPO DI UN SISTEMA DI CARICO PER MISURE DI COMPRESSIONE SU MATERIALI SOFFICI TRAMITE 3D-DIC
GIGLIONI, ALICE
2024/2025
Abstract
This prospective study was designed to implement the use of Digital Image Correlation to investigate the behaviour of a cylindrical silicone specimen subjected to uniaxial compression, through a custom-designed support system. Concurrently, the load applied on the cylinder was monitored by means of a piezometric load cell located inside the support structure and used to perform both static and dynamic measurements, thus providing a complete picture of the strain of the specimen. The analysis of the sample deformation was conceived to be carried out thanks to an internal patterning of the specimen itself, realized by mixing steel powder directly into the silicone rubber. This approach to DIC represents the most noteworthy aspect of the present research and allows to achieve an accurate measurement of the sample deformation under stress. The support system, specifically manufactured for this study, represents the second focus point. It is devised to achieve two main functions: holding both the specimen and the load cell while maintaining their alignment and allowing a controlled compression force to be applied through a screw-driven mechanism. The main challenge faced during its design process was guaranteeing structural stiffness while providing a clear view of the sample, which is crucial for DIC implementation. The experimental procedure followed a loading sequence, during which the specimen was subjected to two different loads. The increasing force was recorded via the load cell and pictures for the DIC were simultaneously captured by two cameras. Image acquisition and processing was performed thanks to VIC-3D system, a turnkey equipment including the necessary hardware and software. The most challenging task during the DIC analysis was enabling the software to perform the correlation correctly and optimizing its accuracy. To achieve that, two steps have been fundamental: firstly, the acquired images were processed by enhancing contrast and saturation; secondly, a thorough trial-and-error process was carried out in order to find the optimal computational parameters (i.e. subset size and step size). It has to be mentioned that the actual speckle pattern used for the correlation was not the one originally intended, i.e. the steel powder: as a matter of fact, the accidental air bubbles trapped in the silicone turned out to provide a much better correlation than the powder particles, that were neither large nor numerous enough to ensure a successful measurement. Lastly, the resulting data were analysed and compared with the expected mechanical response of the specimen. The examination of the data retrieved from the load cell shows some discrepancies with the expected outcome, especially during the static measurements: the inconsistencies are likely due to the piezometric nature of the load cell. In fact, the dynamic measurements provided more accurate results. However, the most noteworthy data are the ones obtained through DIC: the resulting deformation fields closely match the predicted ones, both in distribution and in magnitude. It is worth mentioning that DIC was also able to detect a slight asymmetry of the load, invisible to the naked eye. Unfortunately, the resolution of the DIC measurement is not remarkable and the deformation maps show some evident discontinuities: these are conceivably caused by lack of contrast and consistency of the accidental speckle pattern provided by the air bubbles. Overall, this study can be considered successful: the custom support structure correctly transmitted the compression load to the specimen and the deformation fields derived from DIC, even though not exceptionally accurate, complied with the expected outcome. In spite of its limitations, this paper offers some insight into DIC technology and its implementation, as well as into the use of internal patterning and its possible shortcomings.I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.12075/23585