Combinatorial Reliability Risk in Power Electronics- Embedded Assessment in Automated Design
Abstract
With increasing demands in electrified drivetrains and directed energy systems, power electronics experience significant thermal, mechanical, and electrical reliability challenges. Though wide bandgap technologies offer increased power density, their deployment is inhibited by limited reliable packaging demonstrations. Current reliability methods use computational modeling in concert with costly accelerated testing, and standard reliability qualifications simply cannot capture operating risks that arise from combined stress inputs, such as electromigration-driven voiding interacting with thermomechanical stresses. The challenge of quantifying these risks on an application specific level prevents thorough validations, and limited methods exist for examining interacting failure drivers. Military operating environments consist of multiple concurrent stresses on electronic components (e.g. temperatures, humidity, vibrations), electric drive and weapon systems must be robust in extreme-stress operating conditions. This effort specifically seeks to identify and characterize failure risk in power packaging interconnect technology arising from multiple interacting stress inputs, for the purpose of enabling electrical design automation that can detect and optimize module layouts against these intermingled reliability risks. The approach will use a combination of physics-based reliability experimentation for assessing failure interactions, along with computational investigation to understand the contributions of these compounded risks. Emphasis on novel metallurgies using Cu-Mo with high temperature die attach materials will be examined for next-generation, powerdense systems; however, lifetime reliability model formulation will be made with generic interacting stresses in mind. Further, statistical covariance modeling techniques will guide the development of optimization algorithms to be integrated into UA’s PowerSynth platform. This tool has previously enabled electro-thermal co-optimization design automation, though reliability considerations have been implicit to date, and not yet capable of enabling designers to detect potential compounding failure risks. As such, the resultant design algorithms which incorporate reliability optimization protocols can significantly improve development time and costs in enabling next-gen military power electronics.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 07, 2023
- Source ID
- FA95502110205
Entities
People
- David Huitink
Organizations
- Air Force Office of Scientific Research
- Office of the Secretary of Defense
- University of Arkansas System