Purdue Home | Purdue Search | Visit Purdue | Giving  
Administration |  People   | Contact  
Photo Gallery
Annual Reports

Target System
The focus of PRISM research is a contacting radio frequency (RF) capacitative MEMS, such as that shown. This type of device is widely used for contact actuators and capacitative switches, and involves metal-dielectric contact. A fixed-fixed aluminum membrane makes repeated contact with a dielectric contact pad. In the up state the switch presents a very small capacitance, which allows the signal to travel through the transmission line with very low loss. On the other hand, in the down state, the suspended bridge contacts the dielectric layer forming a large capacitance, which is an effective short circuit at high frequencies. Consequently, the incident signal is reflected and the switch presents a high isolation.

Schematic of radio-frequency MEMS showing membrane (yellow) over contact pad (gold). Application of a voltage difference between the membrane and the pad causes the membrane to make contact with the pad, reflecting the signal in the transmission line (blue).
Failure Modes
Though the technology has much promise, RF MEMS switches are prone to a variety of failure mechanisms. Dielectric charging is caused by trapped charges inside the thin dielectric layer and leads to uncontrollable changes of the actuation voltage versus time, and eventually, complete device failure.
Contact area damage and wear leads to contact resistance degradation
even for relatively low power levels. This effect is probably due to local
strain hardening and change of chemistry due to strong mechanical stresses
at the contact area and is further exacerbated by switched RF currents. It
is often manifested as contact resistance increase versus time that renders
the switch unusable. MEMS switches are very sensitive to environmental
contaminants and humidity. For example, while high humidity levels favor
high adhesion forces that may lead to failure due to the stiction,
nearly-zero humidity is also not ideal since it significantly increases the
friction coefficients between the contacting surfaces, leading to premature
failure. Mechanical failure mechanisms are tied to reductions in the
physical size of the device, resulting in an interaction of microstructure
with interfaces and free surfaces. Mechanical stress and creep development
are important at very high or very low temperatures.Under conditions of shock loading, mechanisms related to materials microstructure, such as spalling, ductile and brittle fracture and delamination, become important.

Research Thrusts
The work to be undertaken by PRISM is divided into five thrusts and three cross-cutting technology groups.
Thrust 1: Contact Physics. The focus of Thrust 1 is failure mechanisms directly associated with metal-dielectric contact, including dielectric charging, current channeling, contact-area evolution and damage, and the influence of environmental factors such as humidity on contact. A significant focus of the center is the atomistic resolution of adhesion and stiction phenomena, and their influence on macroscale physics.
Multiresolution simulations of contact physics

Thrust 2: Multiscale Modeling of MEMS response.
  This thrust addresses the coupled electro-thermo-mechanical response of the MEMS structure. A critical element is the development of micromechanical materials models based on atomistic simulations for the description of single and polycrystalline plasticity, dislocation and defect evolution and the inclusion of size effects during plastic deformation.
Size effects during plastic deformation (Hunter and Koslowski, 2007)

Thrust 3: Multiscale Models for Aerodynamic Damping.
This thrust addresses the external fluid domain, and targets the aerodynamics of fluid damping. An important feature is the accurate characterization of squeeze film damping, particularly during the time-periodic transition between the continuum and the rarefied gas regimes as the gas gap opens and closes. Simulations spanning the continuum regime, described by the Navier-Stokes equations, to rarefied regimes, described by the Ellipsoidal-Statistical Bhatnagar Gross-Krook (ES-BGK) model, are proposed.
Streamlines and pressure contours for moving  micron-sized plate at Kn=0.1 using ES-BGK model (Alexeenko, et al, 2006)

Thrust 4: Uncertainty Quantification.
MEMS fabrication processes create uncertainties in critical physical dimensions. Electrical, thermal and mechanical properties depend on specific processing techniques and initial stress, micro-structural, and surface states, among others. Since these quantities form inputs to our simulations, the effect of their uncertainty on predictions must be quantified. Thrust 4 seeks to quantify this through the use of the generalized polynomial chaos (gPC) formalism. A particular challenge is the extension of this methodology to multiscale multiphsyics systems.

Polynomial chaos in a transverse comb drive simulation. The effect of uncertain Young's modulus on tip deflection is investigated using polynomial chaos of different orders and compared to Monte Carlo (MC) simulation(Agarwal and Aluru, 2007).

Thrust 5: Integration of Models and Numerics.
The focus of this task is to develop a seamless numerical framework to ensure that the models and numerics developed in the Center thrusts mesh well together without compromising accuracy or convergence. Three main groups of techniques are to be integrated. At the macroscale, the backbone for computation will be anunstructured finite volume method (FVM), superimposed with the material point method (MPM), a Lagrangian particle technique. Micro-scale contact will be resolved by molecular dynamics methods, and coupled to the macroscale through reduced-order models.

Verification and Validation: Experimental Program

     A significant experimental effort is planned to provide high-quality validation data. Our experiments start with a careful characterization of material and interface phenomena  to provide initial conditions on the microstructure and to validate important micro-scale models. The experiments reflect a three-tier structure, starting with single physics experiments focusing independently on the electrical, mechanical and thermal domains and graduating to multiphysics experiments that couple two domains - electro-mechanical, thermo-mechanical and electro-thermal, and culminating finally in system-level validation experiments. 
     Another unique aspect of our program is the development of validation data for uncertainty quantification in microsystems. Leveraging experiments from the IMPACT Center at the University of Illinois, a small group of test structures are to be used to extract over 30 material and geometric properties including overetch, Young’s modulus, residual stress and material density. Unlike traditional characterization techniques, the methodology allows measurement of these parameters over the entire device lifetime. In addition, all the extracted properties rely on only three measured electrical parameters: capacitance, frequency, and voltage.   Large numbers of MEMS structures are to be characterized in this way, and their statistical geometric and property distributions related to outputs of single and multiphysics experiments, as well as system-level measurements of MEMS performance. The resulting database will be used to validate our uncertainty quantification effort, and will provide a first-of-its-kind view of how uncertainty is propagated in these microsystems.

Purdue Homepage | Webmaster | Purdue Search | Campus Map | Purdue Directories
Copyright © 2006, Purdue University, all rights reserved.
An equal access/equal opportunity university.