Ferretti, Jacopo
(2026)
New semiconductor technology for automotive power applications, [Dissertation thesis], Alma Mater Studiorum Università di Bologna.
Dottorato di ricerca in
Automotive engineering for intelligent mobility, 38 Ciclo.
Documenti full-text disponibili:
Abstract
The rapid electrification of the automotive sector requires highly reliable and efficient power electronics, especially traction inverters that use Wide Bandgap (WBG) semiconductors, such as silicon carbide (SiC) and gallium nitride (GaN). However, evaluating the reliability of these emerging technologies presents significant challenges as they are prone to severe degradation under continuous, hard-switching dynamic stress. This thesis therefore proposes a comprehensive framework combining advanced virtual prototyping and experimental setups to enhance simulation accuracy and perform reliability assessment of power electronic devices in electric vehicles. The research is structured around two pillars: modeling and simulation for degradation-aware converter design, and hardware platforms for reliability testing and condition monitoring. Within the modeling domain, a high-fidelity electro-thermal virtual prototype of a three-phase SiC automotive inverter has been developed. This datasheet-driven approach incorporates aging effects and advanced thermal CFD models to evaluate performance limits and system degradation under realistic operating conditions, independently of confidential vendor data. Alongside this, a physics-guided digital twin architecture has been introduced for passive components, bridging the gap between data-driven accuracy and circuit-level simulation efficiency. Both approaches were rigorously validated against manufacturers' virtual models and real-world test bench measurements. To address reliability testing and condition monitoring, two custom hardware solutions were designed and validated. The Loadless GaN (LoGaN) setup enables high-frequency, high-voltage dynamic accelerated lifetime testing of GaN HEMTs, extracting vital degradation data under realistic switching conditions. Meanwhile, the Source-Drain Acquisitor (SoDA) board was developed for condition monitoring, enabling highly accurate, in-circuit, offline on-state resistance (Rds,on) measurements. This research ultimately provides a comprehensive toolset spanning from AI-enhanced digital twins to custom stress-testing hardware that facilitates predictive maintenance, and reduces prototyping costs and improves the overall reliability of next-generation electric vehicle powertrains.
Abstract
The rapid electrification of the automotive sector requires highly reliable and efficient power electronics, especially traction inverters that use Wide Bandgap (WBG) semiconductors, such as silicon carbide (SiC) and gallium nitride (GaN). However, evaluating the reliability of these emerging technologies presents significant challenges as they are prone to severe degradation under continuous, hard-switching dynamic stress. This thesis therefore proposes a comprehensive framework combining advanced virtual prototyping and experimental setups to enhance simulation accuracy and perform reliability assessment of power electronic devices in electric vehicles. The research is structured around two pillars: modeling and simulation for degradation-aware converter design, and hardware platforms for reliability testing and condition monitoring. Within the modeling domain, a high-fidelity electro-thermal virtual prototype of a three-phase SiC automotive inverter has been developed. This datasheet-driven approach incorporates aging effects and advanced thermal CFD models to evaluate performance limits and system degradation under realistic operating conditions, independently of confidential vendor data. Alongside this, a physics-guided digital twin architecture has been introduced for passive components, bridging the gap between data-driven accuracy and circuit-level simulation efficiency. Both approaches were rigorously validated against manufacturers' virtual models and real-world test bench measurements. To address reliability testing and condition monitoring, two custom hardware solutions were designed and validated. The Loadless GaN (LoGaN) setup enables high-frequency, high-voltage dynamic accelerated lifetime testing of GaN HEMTs, extracting vital degradation data under realistic switching conditions. Meanwhile, the Source-Drain Acquisitor (SoDA) board was developed for condition monitoring, enabling highly accurate, in-circuit, offline on-state resistance (Rds,on) measurements. This research ultimately provides a comprehensive toolset spanning from AI-enhanced digital twins to custom stress-testing hardware that facilitates predictive maintenance, and reduces prototyping costs and improves the overall reliability of next-generation electric vehicle powertrains.
Tipologia del documento
Tesi di dottorato
Autore
Ferretti, Jacopo
Supervisore
Co-supervisore
Dottorato di ricerca
Ciclo
38
Coordinatore
Settore disciplinare
Settore concorsuale
Parole chiave
Wide Bandgap
WBG
Silicon Carbide
SiC
Gallium Nitride
GaN
Traction Inverters
Reliability
Virtual Prototyping
Power Electronics Simulation
Digital Twin
Accelerated Lifetime Testing
Condition Monitoring
Predictive Maintenance
Data di discussione
21 Luglio 2026
URI
Altri metadati
Tipologia del documento
Tesi di dottorato
Autore
Ferretti, Jacopo
Supervisore
Co-supervisore
Dottorato di ricerca
Ciclo
38
Coordinatore
Settore disciplinare
Settore concorsuale
Parole chiave
Wide Bandgap
WBG
Silicon Carbide
SiC
Gallium Nitride
GaN
Traction Inverters
Reliability
Virtual Prototyping
Power Electronics Simulation
Digital Twin
Accelerated Lifetime Testing
Condition Monitoring
Predictive Maintenance
Data di discussione
21 Luglio 2026
URI
Gestione del documento: