Optimal design of a micromachined force gauge under uncertainty

Abstract

Micromachined force gauges, like many microdevices, are subject to inevitable and inherent uncertainty in dimensional and material parameters that lead to variability in the gauge accuracy and reliability. This paper presents a methodology for the design of a micromachined force gauge under uncertainty, in which the measured force and the maximum stress in the beams in the gauge are related to five uncertain design parameters representing the dimensions of the beam, the material property and the deflection of the beam under the force being measured. A deterministic large-deflection beam model is combined with a sampling-based stochastic model to obtain the distributions of the measured force and the maximum stress corresponding to the interactive influence of the uncertainty in the five design parameters. An optimization problem is formulated to maximize the force measurement accuracy subject to a constraint on the maximum allowable stress. On the basis of the optimization results, design windows for the deflection and the force that maximize the accuracy are derived as functions of the uncertainties in the design parameters. The results provide directly usable information on the optimum geometric design of the force gauge for given uncertainties in the manufacturing and operation of the device.

Statistical Investigation of the Mechanical and Geometrical Properties of Polysilicon Films through On-Chip Tests

“…For example, for silicon tuning fork commercial gyroscopes, it was shown in [ 6 ] that, in order to obtain a sense-drive frequency separation accuracy on the order of 2%, fabrication tolerances for critical suspension springs had to be close to the limit of the deep reactive-ion etching (DRIE) technology. Similarly, in [ 7 , 8 ], it was shown that, in order to make monolithically integrated microtensile testers, the role of uncertainties in the stiffness of critical MEMS components can become considerable.…”

Silicon cantilever sensor for micro-/nanoscale dimension and force metrology

“…However, owing to their small size of typically a few hundreds of microns in length and a few tens of microns in width, AFM probes are not convenient for force calibration of commercial tactile probing tools. Furthermore, by increasing the cantilever length with respect to typical AFM probes both reliability and accuracy can be improved (Mawardi and Pitchumani 2005). Therefore, transfer standards based on silicon microcantilevers with optimized dimensions have been developed as well as calibration procedures operating the transfer standard both as stiffness artifact and as load cell (Behrens et al 2003).…”