Sticking contact (adhesion) between moving components in Micro-ElectroMechanical Systems (MEMS) is a major problem limiting reliability. Sticking contact, referred to as ?stiction failure,? is fatal to the component; stiction prevents the individual component from operating properly and, hence, jeopardizes the reliability of the overall device. Preliminary experiments, performed by the investigators, indicate that electrically induced structural vibrations can lead to the initiation of stick-release. The purpose of this research endeavor is to create a theoretical foundation, validated by experiments on real MEMS devices, for using electrical excitation to drive MEMS components in order (i) to prevent adhesion (through mechanical dithering) and (ii) to repair stiction failures. This non-contacting (noninvasive) approach will enable stick-prevention and stick-release of a component without causing damage to the component or its neighbors. Moreover, the electrical actuation may be built into the MEMS chip, using the existing functionality of the chip to overcome adhesion. This is a cost effective, easy-to-implement approach that may be used in-situ. Also note that the framework developed here will also be applicable on nano-scale devices (NEMS).
This work is broken into an experimental and theoretical component. The experimental portion of this project involves using simplified geometries (micro-cantilevers) and real MEMS components (gear systems, mirrors, etc.) to validate the model results and to demonstrate the utility of electrically induced vibrations in both stiction prevention and stiction repair. Theoretical models will be created to develop a more complete understanding of the fundamental mechanics involved in the dithering and stick-release processes.
Currently there are no commercially available MEMS or Nano-Electro Mechanical Systems (NEMS) with contacting/sliding parts due to the reliability issue caused by adhesion failures. To enable their commercial introduction, practical techniques for the prevention of stiction and repair of failed devices under normal service conditions are required. The proposed methodology is a viable approach, enabling commercialization. The general framework developed here may be extended to nano-devices.
The educational portion of this project integrates MEMS vibrations testing into a required undergraduate lab course. This will give students hands-on experience with MEMS and will demonstrate that near field forces (negligible in macro-scale tests) may not be ignored in all micro-scale tests. Outstanding undergraduates and particularly those from under-represented groups (note that the UNM is a Hispanic Serving Institution (HSI)), will be involved in the research.
Sticking contact (adhesion) between moving components in small electromechanical systems, also referred to as MEMS, is a major problem limiting the manufacturing and reliability of miniaturized components. This type of adhesion is so common that it has a special name – stiction. Although present at all length scales, stiction is a more considerable problem for smaller length scales (dimensions less than 1 mm). The purpose of this research was to study stiction at these smaller length scales with an experimental effort, mainly, and also a theoretical effort. The experimental effort characterized the adhesion energy between small-compliant components (microcantilevered beams of silicon) to a rigid substrate (silicon as well). In a controlled manner, the microcantilevered beams of silicon were the peeled off of the substrate. Using a Michelson type (Figure 1) interferometer the shape of the beam as it was peeled was found, see Figure 2. From its shape the adhesion energy was found using a fracture mechanics model that was developed as a result of this work. Results, Figure 3, show that the previous studies did not capture all the effects of mechanical loading and that the adhesion energies previously reported were artificially low for larger values of h. These data were then used to determine parameters needed to structurally vibrate the cantilevers in order to cause them to un-adhere from the substrate, see Figure 4. This type of behavior, Mathieu-type behavior, was predicted from the developed theory. These data were then applied to the more practical mechanical system of gears. Gears at these smaller length scales stiction-fail to one another easily. In this work, a gear train was designed wherein the gears themselves could be vibrated out of their plane of rotation with a dithering motion. The dithering motion is hypothesized to lower the incidences of stiction failure of the gears to one another, see Figures 5 and 6.
