order=0 cellpadding=0 cellspacing=0 width=100%>
Yahoo! is not affiliated with the authors of this page or responsible for its content.
Caramel: contamination and reliability analysis of microelectromechanical layout - Microelectromechanical Systems, Journal of JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 8, NO. 3, SEPTEMBER 1999
309
CARAMEL: Contamination And Reliability
Analysis of MicroElectromechanical Layout
Abhijeet Kolpekwar, Tao Jiang, and R. D. (Shawn) Blanton
AbstractCARAMEL (Contamination And Reliability Analy-
sis of MicroElectromechanical Layout) is a CAD tool for MEMS
fault model generation. It is based on the integrated circuit
contamination analysis tool CODEF [1] and is capable of ana-
lyzing the impact of contamination particles on the behavior of
microelectromechanical systems. CARAMELs simulation output
indicates that a wide range of defective structures are possible
due to the presence of particulate contaminations. Moreover,
electromechanical simulations of CARAMELs mesh represen-
tations of defective layout has revealed that a wide variety of
misbehaviors are associated with these defects. Several thousand
contamination simulations were performed using CARAMEL on
the surface micromachined comb-drive resonator. The results
generated by CARAMEL identies the comb drive as the most
defect prone region of the microresonator and the deposition of
the rst structural layer as the most vulnerable processing step.
[383]
Index TermsDefects, faults, MEMS tests, resonator.
I. I
NTRODUCTION
T
HE development and deployment of new MEMS products
will not occur due to advances in design, packaging and
processing alone. Testing methodologies must be developed
in concert that are capable of assessing faulty behavior (in the
form of fault simulation and automatic test pattern generation)
along with design for testability (DFT) structures that improve
and ensure the end quality of MEMS-based products. It should
also be noted that many new and existing MEMS applications
are integral parts of safety-critical systems. Consequently, the
need to address reliability of MEMS makes the testability
problem that much more important. To ensure high quality and
reliability of MEMS, a comprehensive testing methodology
must be developed that allows devices to be tested econom-
ically with a very high level of condence. Success of any
testing methodology is highly dependent on the fault models
employed. Fault models that do not cover real defective
behavior can reduce defect coverage and degrade test quality.
The work presented here addresses this need.
MEMS fault models, unlike their digital and analog coun-
terparts, must explicitly consider the impact of defects on
the micromechanical structures. Our rst step toward devel-
Manuscript received August 31, 1998; revised March 19, 1999 and June
25, 1999. This work was supported by the National Science Foundation under
Grant MIP-9702678 and the Defense Research Projects Agency under Rome
Laboratory, Air Force Materiel Command, USAF, under Grant F30602-97-2-
0323. Subject Editor, W. N. Sharpe, Jr.
The authors are with the Electrical and Computer Engineering Department,
Center for Electronic Design Automation, Carnegie Mellon University, Pitts-
burgh, PA 15213-3890 USA.
Publisher Item Identier S 1057-7157(99)07194-2.
oping effective MEMS fault models centers on the inductive
generation of faulty behaviors from particle contamination
simulations. We have chosen the folded-exure comb-drive
microresonator
1
as our research vehicle because it possesses
many of the basic structures (beams, joints, springs, etc.) that
can form the core primitives of a MEMS design library [2][4].
We believe this analysis of the resonator will provide the basis
for developing generic fault models that are applicable to a
wide class of surface-micromachined MEMS.
MEMS failures can result from stiction, manufacturing
variations, undesired residual stress, and particulate contam-
inations. Our experience is that a major cause of the hard-
to-detect faulty behavior in MEMS is due to particulate
contaminations that occur during and after various process
steps of fabrication [5]. Particulates can cause a signicant
perturbation in the structural and material properties of the
microstructure [6]. Thus, a formal assessment of both the
possible defective structures and the corresponding faulty be-
haviors of MEMS design primitives will lead to the formation
of effective MEMS fault models. Such fault models will
undoubtedly lead to methods for fault grading, test generation,
design-for-testability, and design for fault avoidance. The fault
modeling process can also be used to form links between
defects and faulty behaviors. Such links would aid in diagnosis
by helping identify the process steps that are likely to produce
the observed faulty behavior [7].
We have enhanced the process simulator CODEF [1] into
a tool called CARAMEL (Contamination And Reliability
Analysis of MicroElectromechanical Layout) for analyzing
the impact of contamination particulates on the properties
of microelectromechanical structures. CARAMEL is an inte-
gral component of our MEMS fault model generation (see
Fig. 1). CARAMEL performs process simulation and creates
a three-dimensional (3-D) representation of the defective mi-
croelectromechanical structure. It then extracts a mesh netlist
representation from the defective structure whose form is com-
pletely compatible with the mechanical simulator ABAQUS
[8]. Mechanical simulation of the mesh then allows us to link
the contamination of concern to a defective structure and a
faulty behavior. Observed faulty behaviors are then classied
and used to form fault models at the next level of abstraction.
Monte Carlo iteration around the ow of Fig. 1 provides a
mechanism for creating realistic fault models for MEMS.
Here, we describe CARAMEL and illustrate its use in
MEMS fault model generation. The resonator structure (see
1
In the remaining parts of this paper, we will refer to the folded-exure
comb-drive microresonator simply as the resonator.
10577157/99$10.00
©
1999 IEEE 310
JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 8, NO. 3, SEPTEMBER 1999
Fig. 1.
MEMS particulate contamination analysis using CARAMEL.
Fig. 2) under consideration belongs to a class of MEMS known
as surface-micromachined MEMS. Prototype surface micro-
machining processes are available from MCNC (Multi-user
MEMS Processes service (MUMPs) [9]), Analog Devices
iMEMS process [10], and from Sandia National Labs [11].
We have selected the MUMPs process for the contaminations
simulations of the resonator due to its open availability. The
seminal paper on the resonator is given in [12]. Analytic
models of the resonators pertinent characteristics can be
found in [13][15]. The resonator structure is a mature case
study in the design of suspended MEMS which are now
used in commercial accelerometers [10], [16], gyroscopes and
micromirror optical beam steering [17]. Future commercial
applications are in resonator-based oscillators [18], IF mixers,
high-Q IF lters for communications and microstages for
probe-based data storage [19].
The rest of this paper is organized as follows. Section II
describes prior work in the area of MEMS testing. Section III
describes MEMS contamination analysis using our tool
CARAMEL. In Section IV, simulation results obtained using
CARAMEL on the surface-micromachined microresonator are
presented. Finally in Section V, we present conclusions.
II. R
ELATED
W
ORK
Most research in MEMS centers on design, technology,
and packaging problems and not testing. However, a growing
number of researchers have been concerned with MEMS
Fig. 2.
Top view of a comb-drive microresonator.
testability. Here, we discuss representative work performed
in this area.
Fault Modeling and Simulation of MEMS
Fault models abstract the behavior of physical defects to a
high-level representation of the unit under test. In the case of
MEMS, where physical failure mechanisms are much more
complex due to presence of mixed domains, developing fault
models becomes a difcult challenge.
In [21], an extensive overview of the issues and possible
solutions for the problems related to MEMS fault modeling,
simulation, test generation, design for testability, and built-in
self test (BIST) is presented. MEMS defects are modeled using
the concept of mutants and saboteurs. Mutants use analog-
like faults to model nonelectrical (mechanical, optical, etc.)
defects. Mutants represent defects that cause value changes in
microstructure parameters. Saboteurs, on the other hand, model
defects that cause components to be removed or added to the
microstructure. Fault simulation is performed by modifying the
fault-free schematic of the microsystem through the instanti-
ation of mutants or saboteurs. A still-open question involves
characterization and selection of the mutants and saboteurs
that represent realistic MEMS defects.
In [22], the importance of the MEMS model and its relation-
ship to defects of the MEMS structure is stressed. An electrical
model is also used to represent mechanical and electrical
components of MEMS. However, the model is constructed
in a such way so as to allow the accurate modeling of a
wide variety of MEMS defects. They have focused on the
co