Variation-aware modelling of micro-scale piezoelectric energy harvesters

The inherent tolerances in MEMS fabrication pro­cesses introduce variability that causes the device performance to deviate systematically from its designed values. In this work, we have used Monte-Carlo (MC) simulation to propagate process variability through a piezoelectric energy harvester (PEH) finite element model (FEM) and its extracted equivalent circuit model (ECM) to discover the uncertainty models of its performance metrics, such as optimal power and resonance frequency and its ECM parameters. Characterization results of a set of fabricated PEH devices were compared against the MC-derived uncertainty models to explain the statistical spread.