Hybrid Model-Based and Data-Driven Solution for Uncertainty Quantification at the Microscale

Abstract

Background: Due to their size, Micro Electromechanical Systems (MEMS) display performance indices affected by uncertainties linked to the mechanical properties and to the geometry of the films constituting their movable parts. Objective: In this perspective, a recently proposed multiscale and hybrid solution for uncertainty quantification is discussed. Methods: The proposed method is based on the (deep) learning of the morphology-affected elasticity of the polycrystalline films, and of the microfabrication-induced defective geometry of the devices. The results at the material and at the device levels are linked through a reduced-order representation of the response of the entire device to the external stimuli, foressen to finally feed a Monte Carlo uncertainty quantification engine. Results: Preliminary results relevant to a single-axis resonant Lorentz force micro-magnetometer have shown a noteworthy capability of the proposed multiscale deep learning method to account for the mentioned uncertainty sources at the microscale. Conclusion: A promising two-scale deep learning approach has been proposed for polysilicon MEMS sensors to account for both materials- and geometry-governed uncertainties, and to properly describe the scale-dependent response of MEMS devices.

On-Chip Tests for the Characterization of the Mechanical Strength of Polysilicon

“…Monte Carlo simulations were then adopted to assess the spreading of results linked to devices, which were all identical if not for the uncertainties of sources due to the microfabrication. To speed up such investigation, deep learning strategies were recently proposed in [14,15]. Even though the stochastic mechanical resistance of ThELMA polycrystalline silicon was already statically explored [16], these studies were mostly focused on in-plane fracture.…”

“…Monte Carlo simulations were then adopted to assess the spreading of results linked to devices, which were all identical if not for the uncertainties of sources due to the microfabrication. To speed up such investigation, deep learning strategies were recently proposed in [14,15].…”

Optimization of the Geometry of a Microelectromechanical System Testing Device for SiO2—Polysilicon Interface Characterization

“…This localized cracking-like failure mode typically occurs at the interface between silicon dioxide and polycrystalline silicon. The fatigue and delamination phenomena both lead to a progressive shift in resonance frequency, structural stiffness (also affected by uncertainties at the microscale; see [18][19][20][21][22]), and electrical resistance, thereby affecting the long-term reliability of these devices [9,[23][24][25].…”