1. IC Integration Technologies for Flexible Hybrid Electronics (FHE)
- Funded by NSF SBIR and Contracted through Uniqarta for one year
- Performed accelerated testing of FHE considering Temperature and Humidity together
- Conducted failure analysis on parts failed during testing (Digital Microscope)
- Working on developing degradation model and reliability prediction
2. Creation of an Adaptive Remaining Useful Life (RUL) Time prediction Model of Power Electronic Modules
- Funded by NREL DOE for two years, ending period: September 30, 2018
- Developing degradation models and probabilistic reliability prediction models
- Developing a Model to predict RUL on real time basis as new data become available
3. Implementing QMS and Process Improvement Tools at D&M Industries
- Funded by D&M Industries continuously for last three years
- Overview: Helping D&M Industries for establishing QMS and Continuous Improvement Process
Funded by D&M Industries continuously for last three years
4. Reliability Oriented Design for DC-Link Capacitors in Power Electronic Converters
- Partially Funded by CQRME for last three year
- Task 1: Review of SiC MOSFET Based Three-Phase Inverter Lifetime Prediction
- Overview: This project presents a review on SiC MOSFET based 3-phase inverter lifetime prediction. The inverter-level lifetime prediction flowchart is first illustrated with the system failure rate and system Mean Time to Failure derived which integrate six in-series stresses. An overview of SiC MOSFET power module Physics-of-Failure is then presented. Following it are the step-by-step lifetime modeling comparison and summarization, including power loss analysis, thermal modeling, thermal-mechanical modeling, and damage accumulation modeling. Furthermore, cycle counting algorithms and active power cycling tests specific to SiC MOSFET are discussed. Finally, a detailed system-level thermal profile extraction from SiC MOSFET based 3-phased inverter in the hybrid electrical vehicle application is conducted. The simulation result is consistent with theoretical power loss and thermal equivalent circuit based junction temperature evaluation. Thus it can be used for further system lifetime modeling and the reference for 3-phase inverter based power cycling tests in SPWM mode.
- Task 2: Comparison and Analysis of SiC MOSFET Gate Oxide Degradation Indicators
- Overview: SiC MOSFETs have been applied in multiple high-power, high-voltage applications such as PV systems, electric vehicles, and solid-state transformers. It is critical to ensure their long-time reliable operation considering the high maintenance cost and potential economic impact. SiC MOSFET gate oxide has been widely studied due to its inherent reliability issues compared with Si devices, such as lower inversion channel mobility, higher interface state density, and smaller conduction band offset. To evaluate the gate oxide reliability and estimate its lifetime, offline accelerated tests and online condition monitoring can be conducted. This project presents a comprehensive comparison and analysis for SiC MOSFET among these six indicators and another indicator, i.e. gate driver turn-on energy. Four accelerated tests are designed and conducted. The indicator verification tests are carried out. Based on the test results, the indicator shifts are analyzed. Finally, the comparison among these seven indicators is made and the conclusion is drawn.
5. Prognostics and Maintenance Planning of Complex Systems in Dynamic Environmental/Operational Conditions
- Partially Funded by CQRME for last three year
- Overview: This project aims at developing efficient methods for prediction of remaining useful life (RUL) and maintenance planning of complex systems working in dynamic environmental/operational conditions. We propose a novel data-driven approach addressing the effects of dynamic conditions for online prediction of RUL. Moreover, we extend our approach for the complex systems consisted of stochastically dependent components, which the failure or the degradation state of a component will influence the degradation process of other components of the system. Considering dynamic conditions and stochastic dependency between components add to the complexity of the problem and increase the computational cost. Therefore, our goal is to get an accurate prediction of the RUL at the reasonable computational time developing statistical and machine learning methods for learning from historical data and condition monitoring signals. The proposed research will provide an accurate and online prediction of the RUL of the system that can be used to find the optimal maintenance strategy and warranty plans as a next step of our work.
6. Reliability of Cyber Physical Systems
- Partially Funded by CQRME for last one year
- Overview: Cyber-Physical Systems (CPS) encompass physical processes controlled and monitored by computational efforts. The range of applications includes medical systems, assisted living, traffic control, automotive, process control, critical infrastructure systems, etc. Nowadays, it becomes crucial to estimate the reliability profiles of such systems. This project presents a review on reliability and degradation models for Cyber Physical Systems. As a starting point, a CPS can be decomposed in physical components and cyber components. Degradation models have been applied for physical components, and Non-homogeneous Poisson Processes (NHPP) for cyber components reliability models. A model is proposed to assess the reliability profile of a Cyber-Physical System (CPS) based on the mixed-degradation profiles of its critical physical components and the reliability profile of its critical cyber component.