The shift from internal combustion to electric vehicles necessitates addressing new challenges posed by novel components, materials, manufacturing processes and different operating regimes. This includes assessing new types of excitations and damages from a reliability perspective. This paper investigates a solution to enhance the reliability of the Printed Circuit Board (PCB) within a Power Electronic Unit (PEU). PCBs are critical electronic and mechanical supports for functional components (e.g. diodes, transistors, resistors) in numerous applications. Controlling critical vibration levels is crucial to prevent electronic component breakage and subsequent PEU functionality loss.
Existing solutions often focus on improving sensor-PCB bonding, surface treatments, or stiffening components’ support structures. The proposed approach, instead, exploits Locally Resonant Metamaterials (LRMs) to reduce the vibrational load experienced by the PCB. LRMs are created by adding resonating elements to the host component. They provide excellent Noise, Vibration, and Harshness (NVH) performance within specific frequency ranges, while being lightweight and come with a high design freedom, making them adaptable even to mass-sensitive or design-constrained components.
Since direct LRM implementation on the PCB is not feasible in this specific application, an alternative is explored, namely treating the spider frame which is an aluminum die-casted part securing the PCB. Previous simulation studies by the authors have demonstrated that this treatment significantly reduces the vibrational load on the PCB. In this study, the LRM solution is fabricated, and experimental validation is performed using a modal test, followed by a shaker test mimicking operating conditions.
Multiple LRM configurations are explored: initially two concepts specifically tuned to individual PEU resonance frequencies, and later a hybrid configuration combining resonant additions to simultaneously reduce vibrations at both targeted peaks. Finally, damage is calculated using both experimental and simulated data, comparing configurations with and without the LRM solution, demonstrating how LRMs can significantly enhance component reliability.