Nickel–Titanium (Ni–Ti) shape memory alloy (SMA) is recognized as the top choice for realizing self-expanding cardiovascular devices enabling minimally invasive surgery, thanks to its optimal combination of mechanical, functional, and biocompatibility properties. Its primary applications include peripheral stents to treat occlusive pathologies affecting lower limb vessels and heart valve frames to replace stenotic or insufficient native valves. Ni–Ti implants throughout their lifespan undergo millions of fatigue loading cycles related to body and physiological activities, leading to multi-axial deformation. Preclinical assessment for an equivalent of ten years of working life is required to verify the structural and functional integrity of new devices and prevent fracture phenomena. Recently, the increasing role of computational simulations in the regulatory procedure for medical devices has established a total-life assessment approach. Experimental characterization of the fatigue limit of the material on surrogate samples is combined with high-fidelity structural simulations of devices to predict the failure risk through a fatigue criterion. However, this workflow presents several open issues related to the complexities of Ni–Ti behavior, preventing the use of standard and well-established practices. This Ph.D. thesis aims to investigate complementary predictive approaches to enhance the fatigue assessment of Ni–Ti cardiovascular devices beyond the state of the art, combining advanced experimental and computational techniques. Stent samples with commercial and non-commercial designs were exploited to refine the total-life approach and fill current gaps in the literature regarding the modeling procedure and the fatigue criteria. The impact of modeling choices that are generally overlooked was assessed through a comprehensive sensitivity analysis. Experimental tests on uniaxial multi-wire samples were carried out to characterize the fatigue behavior of the material, while fatigue tests on stents were performed to verify the total-life pipeline. Among several state-of-the-art approaches, ad hoc fatigue indicators accounting for SMA mechanical hysteresis were identified as the most reliable option, predicting the severity of different loading conditions and capturing the most critical device locations. A complementary damage-tolerant approach was investigated, considering crack propagation from manufacturing-induced defects. Crack growth was characterized on notched samples with a suitable size and microstructure, using the cyclic J-integral as crack driving force parameter to capture energy dissipation associated with phase transformation. The approach was preliminarily validated on multi-wire samples, enabling the identification of safe conditions and providing fatigue life predictions in agreement with experimental data. The application to non-commercial stents offered additional insights compared to the total-life approach, addressing the role of material flaws. Finally, a novel one-dimensional phase-field model of fracture and fatigue in SMAs was developed. The model was built upon the Auricchio–Petrini constitutive model following a variational formulation and coupling phase transformation with damage evolution. The model was validated against multi-wire sample uniaxial fatigue data, enabling a distinction of safe and critical conditions and providing encouraging fatigue life predictions. The enhanced predictive capability enabled by the methods proposed in this thesis has the potential to significantly improve device reliability and foster innovative Ni–Ti applications. This thesis offers a pathway to advance current design paradigms by establishing a comprehensive predictive framework that synergistically combines experimental, analytical, and computational approaches.
La lega a memoria di forma Nichel–Titanio (Ni–Ti) rappresenta la scelta principale per la realizzazione di dispositivi cardiovascolari autoespandibili destinati alla chirurgia mini-invasiva, grazie alla sua combinazione ottimale di proprietà meccaniche, funzionali e di biocompatibilità. Le principali applicazioni comprendono stent periferici per il trattamento di patologie occlusive dei vasi degli arti inferiori e strutture di supporto per valvole cardiache impiegate nella sostituzione di valvole native stenotiche o insufficienti. Durante l’esercizio, gli impianti in lega Ni–Ti sono sottoposti a milioni di cicli di carico a fatica associati alle attività corporee e fisiologiche, che inducono stati di deformazione multiassiali. Al fine di garantire l’integrità strutturale e funzionale dei nuovi dispositivi e prevenire fenomeni di frattura, è richiesta una valutazione preclinica equivalente a dieci anni di impianto. Il crescente impiego delle simulazioni computazionali nelle procedure regolatorie dei dispositivi medici ha favorito l’affermazione di un approccio di valutazione total-life. Tale approccio combina la caratterizzazione sperimentale del comportamento a fatica del materiale su campioni surrogati con simulazioni strutturali dei dispositivi, allo scopo di prevedere il rischio di cedimento mediante opportuni criteri di fatica. Tuttavia, la complessità del comportamento meccanico della lega Ni–Ti introduce diverse criticità ancora irrisolte, che limitano l’applicabilità di pratiche standard e consolidate. Questa tesi di dottorato si propone di investigare approcci predittivi complementari per migliorare la valutazione a fatica dei dispositivi cardiovascolari in lega Ni–Ti, attraverso l’integrazione di tecniche sperimentali e computazionali avanzate. Stent con diversi design sono stati impiegati per affinare l’approccio total-life e colmare alcune lacune presenti in letteratura relative alle procedure di modellazione e ai criteri di fatica. L’impatto di scelte di modellazione comunemente trascurate è stato analizzato mediante un’accurata analisi di sensitività. Sono state condotte prove sperimentali su campioni uniassiali multi-filo per caratterizzare il comportamento a fatica del materiale, mentre prove a fatica su stent hanno permesso di verificare l’affidabilità dell’approccio total-life. Tra i diversi criteri proposti in letteratura, indicatori basati sull’isteresi meccanica tipica delle leghe a memoria di forma si sono dimostrati i più affidabili, consentendo di valutare la severità delle diverse condizioni di carico e di individuare le regioni più critiche del dispositivo. In parallelo, è stato investigato un approccio complementare di tipo damage-tolerant, che considera la propagazione di cricche a partire da difetti introdotti durante il processo di fabbricazione. Il processo di propagazione di cricca è stato caratterizzato su campioni con dimensioni e microstruttura adeguate, adottando il J-integral ciclico per tenere conto della dissipazione di energia associata alla trasformazione di fase. L’approccio è stato preliminarmente validato su campioni multi-filo, consentendo l’identificazione di condizioni di sicurezza e fornendo previsioni di vita a fatica in accordo con i dati sperimentali. L’applicazione a stent non commerciali ha fornito informazioni complementari rispetto all’approccio total-life, mettendo in luce il ruolo dei difetti del materiale. Infine, è stato sviluppato un modello monodimensionale basato sull’approccio phase-field per la descrizione dei fenomeni di frattura e fatica nelle leghe a memoria di forma. Il modello è stato formulato a partire dal modello costitutivo Auricchio–Petrini, seguendo una formulazione variazionale e accoppiando la trasformazione di fase con l’evoluzione del danno. La validazione del modello, condotta mediante dati sperimentali di fatica uniassiale su campioni multi-filo, ha permesso di distinguere tra condizioni sicure e critiche, fornendo previsioni promettenti della vita a fatica. La maggiore capacità predittiva resa possibile dai metodi proposti in questa tesi ha il potenziale di migliorare significativamente l’affidabilità dei dispositivi cardiovascolari e di favorire applicazioni innovative della lega Ni–Ti. Nel complesso, questo studio offre un’evoluzione degli attuali metodi di progettazione, proponendo un approccio predittivo integrato che combina in modo sinergico tecniche sperimentali, analitiche e computazionali.
Predictive approaches for the fatigue assessment of Nickel–Titanium cardiovascular devices
Brambilla, Alma
2025/2026
Abstract
Nickel–Titanium (Ni–Ti) shape memory alloy (SMA) is recognized as the top choice for realizing self-expanding cardiovascular devices enabling minimally invasive surgery, thanks to its optimal combination of mechanical, functional, and biocompatibility properties. Its primary applications include peripheral stents to treat occlusive pathologies affecting lower limb vessels and heart valve frames to replace stenotic or insufficient native valves. Ni–Ti implants throughout their lifespan undergo millions of fatigue loading cycles related to body and physiological activities, leading to multi-axial deformation. Preclinical assessment for an equivalent of ten years of working life is required to verify the structural and functional integrity of new devices and prevent fracture phenomena. Recently, the increasing role of computational simulations in the regulatory procedure for medical devices has established a total-life assessment approach. Experimental characterization of the fatigue limit of the material on surrogate samples is combined with high-fidelity structural simulations of devices to predict the failure risk through a fatigue criterion. However, this workflow presents several open issues related to the complexities of Ni–Ti behavior, preventing the use of standard and well-established practices. This Ph.D. thesis aims to investigate complementary predictive approaches to enhance the fatigue assessment of Ni–Ti cardiovascular devices beyond the state of the art, combining advanced experimental and computational techniques. Stent samples with commercial and non-commercial designs were exploited to refine the total-life approach and fill current gaps in the literature regarding the modeling procedure and the fatigue criteria. The impact of modeling choices that are generally overlooked was assessed through a comprehensive sensitivity analysis. Experimental tests on uniaxial multi-wire samples were carried out to characterize the fatigue behavior of the material, while fatigue tests on stents were performed to verify the total-life pipeline. Among several state-of-the-art approaches, ad hoc fatigue indicators accounting for SMA mechanical hysteresis were identified as the most reliable option, predicting the severity of different loading conditions and capturing the most critical device locations. A complementary damage-tolerant approach was investigated, considering crack propagation from manufacturing-induced defects. Crack growth was characterized on notched samples with a suitable size and microstructure, using the cyclic J-integral as crack driving force parameter to capture energy dissipation associated with phase transformation. The approach was preliminarily validated on multi-wire samples, enabling the identification of safe conditions and providing fatigue life predictions in agreement with experimental data. The application to non-commercial stents offered additional insights compared to the total-life approach, addressing the role of material flaws. Finally, a novel one-dimensional phase-field model of fracture and fatigue in SMAs was developed. The model was built upon the Auricchio–Petrini constitutive model following a variational formulation and coupling phase transformation with damage evolution. The model was validated against multi-wire sample uniaxial fatigue data, enabling a distinction of safe and critical conditions and providing encouraging fatigue life predictions. The enhanced predictive capability enabled by the methods proposed in this thesis has the potential to significantly improve device reliability and foster innovative Ni–Ti applications. This thesis offers a pathway to advance current design paradigms by establishing a comprehensive predictive framework that synergistically combines experimental, analytical, and computational approaches.| File | Dimensione | Formato | |
|---|---|---|---|
|
Brambilla_PhD_Thesis.pdf
non accessibile
Descrizione: PhD Thesis
Dimensione
30.28 MB
Formato
Adobe PDF
|
30.28 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/10589/248277