Note: Descriptions are shown in the official language in which they were submitted.
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
Dual Position Linear Displacement Midromechanism
1. FIELD OF THE INVENTION
The present method and apparatus relates to dual position mechanisms. More
particularly, the invention relates to compliant bistable mechanisms, and the
invention also
relates to an apparatus and method for a dual position latching mechanism.
2. TECHNICAL BACKGROUND
The term "compliant mechanisms" relates to a family of devices in which
integrally
formed flexural members provide motion through deflection. Such flexural
members may
therefore be used to replace conventional mufti-part elements such as pin
joints. Compliant
mechanisms provide several benefits, including backlash-free, wear-free, and
friction-free
operation. Moreover, compliant mechanisms significantly reduce manufacturing
time and
cost. Compliant mechanisms can replace many conventional devices to improve
functional
characteristics and decrease manufacturing costs. Assembly may, in some cases,
be obviated
entirely because compliant structures often consist of a single piece of
material.
In microelectromechanical systems (MEMS), compliant technology allows each
mechanism of a MEMS system to be an integrally formed, single piece mechanism.
Because
MEMS devices are typically made by a layering and etching process, elements in
different
layers must normally be etched and formed separately from each other.
Additionally,
elements with complex shapes, such as pin joints, require multiple steps and
layers to create
the pin, the head, the pin-mounting joint, and the gap between the pin and the
surrounding
ring used to form the joint. While pin joints do have difftculties in
manufacturing, these
complex shapes do have advantages of allowing large displacements and low
stresses
compared to fully compliant mechanisms.
An integrally formed compliant mechanism, on the other hand, may be
constructed as
a single piece, and may even be constructed in unitary fashion with other
elements of the
micromechanism. Substantially all elements of many compliant devices may be
made from a
single layer. Reducing the number of layers, in many cases, simpliftes the
manufacturing and
design of MEMS devices. Compliant technology also has unique advantages in
MEMS
applications because compliant mechanisms can be manufactured unitarily, i.e.,
from a single
continuous piece .of material, using masking and etching procedures similar to
those used to
form semiconductors.
-1-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
In MEMS as well as in other applications, there exists a large need for
"bistable
devices," or devices that can be selectively disposed in either of two
different, stable
configurations. Bistable devices can be used in a number of different
mechanisms, including
switches, valves, clasps, and closures. Switches, for example, often have two
separate states:
on and off. However, most conventional switches are constructed of rigid
elements that are
connected by hinges, and therefore do not obtain the benefits of compliant
technology.
Compliant bistable mechanisms have particular utility in a MEMS environment,
in which
electrical and/or mechanical switching at a microscopic level is desirable,
and in which
conventional methods used to assemble rigid body structures are ineffective.
Bistable mechanisms present a unique challenge because the compliant elements
must
be properly balanced so that two fully stable positions exist. Even if a
bistable design is
obtained by fortunate guesswork or extensive testing, conventional
optimization techniques
are often ineffective because the design space is so complex, i.e., highly
nonlinear and
discontinuous, with such a small feasible space that gradient-driven methods
are unable to
reach a workable solution. The likelihood that a stochastic method will
stumble onto a
solution is extremely small in fully compliant designs. Hence, it is difficult
to enhance the
fully compliant bistable designs, except through additional experimentation.
However, implementation of designs that allow for large displacements of
bistable
mechanisms can provide for mechanisms that are more predictable and require
less
experimentation to obtain two stable configurations. Adding pin joints to
compliant
mechanisms can allow for these large displacements to enable bistability
without undue
experimentation and analysis. Unfortunately, previous MEMS bistable designs
have
encountered difficulties with applying pin joints to non-stationary members.
Additionally,
various attempts of using non-stationary pin joints have encountered motion
problems as the
result of suction, the bonding of moveable members to the microchip substrate.
Consequently, it would be an advancement in the art to provide a bistable
compliant
mechanism, and particularly a bistable compliant mechanism that would have a
large
displacement to ease design considerations as well as a bistable mechanism
that can employ
pin joints on moveable structures.
Another problem in the art of bistable and dual positional mechanisms is the
inability
of these mechanisms to sustain a force that is greater than the force that
actuates the
mechanism into its stable configurations. To obtain a highly stable mechanism,
the
mechanism should have the ability to lock to one of two positions. One such
stable device is
-2-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
a latching mechanism. Most latching mechanisms provide for interference
engagements
between two members that will maintain a force larger than the engagement
force. However
in creating a highly secure and stable configuration, these latching
mechanisms often require
multiple actuators to both latch and unlatch the mechanism. Further, complex
processes and
operations are needed to operate these mechanisms that limit the number of
applications for
latching mechanisms, especially MEMS applications. This subsequently increases
the cost
and size of the mechanism.
Consequently, it would be an advancement in the art to provide a latching
mechanism
that could lock and unlock through the use of a single actuator. The latch
should also be able
to sustain a force larger than the force used to loclc the latching mechanism
as well as be easy
to unlatch.
BRIEF SUMMARY OF THE INVENTION
The present invention provides a micromechanism having a first stable
configuration
and a second stable configuration. The micromechanism comprises a complaint
leg, a base
member, and a shuttle. The complaint leg has a base end and a shuttle end,
wherein the base
end is coupled to the base member via a pin joint and the shuttle end is
coupled to the shuttle
via a pin joint. In one embodiment, the pin joint that couples the compliant
leg to the base
member is a fixed pin joint and the pin joint that couples the compliant leg
to the shuttle is a
floating pin joint.
The bistable micromechanism may be located on a microchip substrate where the
base member is fixed to the microchip substrate. The bistable mechanism may
further have a
plurality of dimples that extend from the bottom of the mechanism. The dimples
are
configured to elevate the mechanism above the substrate to prevent stiction.
The dimples
may be located on the floating pin joint, the shuttle, or the leg.
The leg may have any number of shapes to allow the mechanism to be bistable.
The
leg may be arch, straight, or "V" shaped. Other shapes are possible so long as
the shape
functions similar to a spring by storing potential energy when compressed. The
mechanism
operates by increasing potential energy in the leg as it is deflected toward
the maximum
potential energy position. From the maximum potential energy position, the
mechanism may
toggle in one of two directions toward one of two low potential energy
positions. Thus, the
leg may toggle between low potential energy positions, where these two low
potential energy
positions are separated by the maximum potential energy position.
-3-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
Another embodiment of the present invention provides for a latching mechanism
comprising a grasping member, a lock slider, and a detent slider. The lock
slider is
configured to induce a locking deflection with respect to the grasping member
to reach an
engaged position. The lock slider is also capable of inducing an unlocking
deflection with
respect to the grasping member, where the unlocking deflection is greater than
the loclcing
deflection. The decent slider is coupled to the lock slider to selectively
maintain the
unlocking deflection with respect to the grasping member. This unlocking
deflection allows
the lock slider to disengage the grasping member and retract to its initial
position.
The grasping member may be a compliant mechanism that has a fixed end and a
free
end that are separated by a generally flexible arm. The free end comprises an
engagement
member and a disengagement member. The engagement member is configured to
engage the
lock slider in a locked configuration and the disengagement member is
configured to engage
the lock slider in an unlocked configuration. The detent slider is operably
connected to the
lock slider. The lock slider is substantially situated with the perimeter of
the detent slider and
the detent slider has a raised member that engages a stop that is situated on
top of the lock
slider. Thus act the lock slider moves, the detent slider will react to its
movements as the
stops abut the raised member. The mechanism is capable of being actuated by a
thermal in-
plane microacW ator in a MEMS application. The actuator can both lock and
unlock the
latching mechanism by driving the lock slider in a single direction.
This latching mechanism follows the process of displacing the lock slider a
first
distance in an actuation direction sufficient to engage a grasping member in a
locked
configuration. Then the step of displacing the lock slider a second distance
in the actuation
direction sufficient to deflect the grasping member in an unlocked
configuration is
performed. Once the grasping members are deflected the step of displacing the
loch slider a
third distance in the actuation direction sufficient to maintain the grasping
member in the
unlocked configuration occurs. Then finally, the step of displacing the lock
slider a fourth
distance in a direction opposite the actuation direction, wherein the third
distance is sufficient
to disengage the grasping member finishes the processes.
Numerous other embodiments of the bistable mechanism and the latching
mechanism
may be created through varying parameters of the mechanism. For example,
additional leg
sets may be added to the bistable mechanism and the type of joints in the
mechanism can be
varied for different applications. In a latching mechanism the members that
induced the
-4-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
locked and unlocked configurations may be varied to create a design that is
capable of being
placed in a wide variety of locations.
DESCRIPTION OF DRAWINGS
Figure 1 is a plan view of one embodiment of a compliant bistable mechanism
according to the invention;
Figure 2 is a plan view of an analytical model to approximate the operation of
the
bistable mechanism of Figure 1;
Figure 3 is a flowchart diagram showing one embodiment of an optimization
process
for compliant mechanisms;
Figure 4 is a graph showing a local opfimurn and a global optimum for an
objective
characteristic to be optimized over a single analytical model characteristic;
Figure 5 is a flowchart diagram showing one embodiment of a first recursive
optimization algorithm suitable for the optimization process of Figure 3;
Figure 6 is a flowchart diagram showing one embodiment of a second recursive
optimization algoritlnn suitable for the optimization process of Figure 3;
Figure 7 is a plan view of an alternative embodiment of a compliant bistable
mechanism according to the invention;
Figure 8 is a plan view of another alternative embodiment of a compliant
bistable
mechanism according to the invention;
Figure 9 is a plan view of yet another alternative embodiment of a compliant
bistable
mechanism according to the invention; and
Figure 10 is a plan view of still another alternative embodiment of a
compliant
bistable mechanism according to the invention.
Figure 11 is a plan view of one embodiment of a compliant pined-pinned
bistable
mechanism according to the invention;
Figure 12 is a plan view of an alternative embodiment of a non-actuated
compliant
pined-pinned bistable mechanism according to the invention;
Figure 13 is a plan view of an alternative embodiment of an actuated compliant
pined-
pinned bistable mechanism according to the invention;
Figure 14 is a plan view of an alternative embodiment of a non-actuated
compliant
pined-pinned bistable mechanism attached to a gear actuator;
-5-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
Figure 15 is a plan of one embodiment of a latching mechanism according to the
invention;
Figure 16 is an operational plan view of an alternative embodiment of the
latching
mechanism according to the invention;
Figure 17 is an operational plan view of an alternative embodiment of the
latching
mechanism attached to an amplified thermal in-plane microactuator; and
Figure 18 is an isometric view of an alternative embodiment of the latching
mechanism according to the invention.
DETAILED DESCRIPTION OF THE INVENTION
The present apparatus and method will be best understood by reference to the
drawings, Wherein like parts and steps are designated by like numerals
throughout. It will be
readily understood that the components of the present mechanism, as generally
described and
illustrated in the Figures herein, could be arranged and designed in a wide
variety of different
configurations. Thus, the following detailed description of the embodiments of
the apparatus
and method, as represented in the Figures, are not intended to limit the scope
of the claim, but
are merely representative of present embodiments of the apparatus and method.
Referring to Figure 1, one example of an integrally formed compliant mechanism
10,
or mechanism 10, is shown. The mechanism 10 may be sized for MEMS
applications.
Although many of the examples presented herein apply to MEMS applications, the
apparaW s
and method of the present invention are not limited to MEMS, but are rather
applicable to
compliant mechanisms in general.
The exemplary mechanism 10 of Figure 1 is a bistable micromechanism, or a MEMS
device that can be actuated between two stable configurations through the
application of an
input force 11. Such a mechanism may be used to perform microswitching
functions or the
like. The mechanism 10 may have a longitudinal direction 12, a lateral
direction 14, and a
transverse or out-of plane direction 16. The mechanism 10 may have a
substantially planar
configuration, e.g., all parts of the mechanism 10 may have substantially the
same thickness
and positioning in the transverse direction 16. The mechanism 10 may have a
shuttle 20
configured to receive the input force 11 and, if desired, exert an output
force on some other
obj ect.
-6-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
The shuttle 20 may be comlected to a first base member 22, a second base
member
24, a third base member 26, and a fourth base member 28. The base members 22,
24, 26, 28
may be stationary, or may be affixed to other stationary or mobile MEMS
devices.
Consequently, the base members 22, 24, 26, 28 are depicted simply as
attachment surfaces.
S The first and third base members 22, 26 may be separated from the second and
fourth base
members 24, 28 by an offset distance 29. The offset distance 29 may be
substantially fixed.
The shuttle 20 may be connected to the base members 22, 24, 26, 28 through the
use of a first
leg 30 and a second leg 32. The legs 30, 32 may be thick enough to be
substantially rigid.
The exemplary mechanism 10 of Figure 1 implements two general types of
compliant
segments, or segments that are designed to provide motion and/or biasing force
through
deformation.
More specifically, the mechanism 10 may have a first shuttle pivot coupling 40
and a
second shuttle pivot coupling 42 that pivotally attach the shuttle 20 to the
first and second
legs 30, 32. Similarly, the mechanism may have a first base pivot coupling 44
and a second
1 S base pivot coupling 46 that pivotally attach the first and second legs 30,
32 with the base
members 22, 24, 26, 28. As shown, the pivot couplings 40, 42, 44, 46 take the
form of small-
length flexural pivots 40, 42, 44, 46, or flexural pivots 40, 42, 44, 46;
however, in other
embodiments, pin joints or other types of compliant members may be used. The
flexural
pivots 40, 42, 44, 46 allow the shuttle 20 to travel from a first position SO
to a second position
S2 by flexing to provide pivotal motion similar to that of a pin joint. Thus,
the mechanism 10
has a first stable configuration 54, or first configuration S4 corresponding
to the first position
SO and a second stable configuration S6, or second configuration 56
corresponding to the
second position S2.
Generally, small-length flexural pivots are thin cross-sectioned segments that
replace
2S traditional pivotal joints while still allowing the joint to be
mathematically modeled as a
traditional pivotal joint. Small-length flexural pivots bend along their
length to allow other
elements of a micromechanism to move relative to each other. Small-length
flexural pivots
can have a wide variety of lengths and shapes to suit multiple designs. Longer
small-length
flexural pivots allow for a large range of motion, while shorter pivots are
easier to model with
techniques, such as pseudo-rigid body modeling.
To facilitate mathematical modeling of the small-length flexural pivots
through
methods such as pseudo-rigid body modeling, it is advantageous to form each of
the flexural
pivots 40, 42, 44, 46 with a length of less than 10% of the length of the legs
30, 32 to which
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
they are attached. However, the length of the flexural pivots 40, 42, 44, 46
must also be
sufficiently long to allow the legs 30, 32 to rotate through the necessary
angle as the shuttle
20 travels between the first and second positions 50, 52. Therefore, depending
on the desired
displacement between the first and second positions 50, 52 of the shuttle 20,
a 10% ratio of
flexural pivot length to leg length may be advantageous. Longer flexural
pivots 40, 42, 44,
46 may also be used, but may require the use of more advanced modeling
techniques.
The first and third base members 22, 26 may be connected together by a first
mounting beam 58, and the second and fourth base members 24, 28 may be
connected
together to form a second mounting beam 59. The mounting beams 58, 59 may have
a length
and thickness selected to permit bending in the fixed-fixed configuration.
Thus, the first and
third base members 22, 26 may operate in conjunction with the first mounting
beam 58 to
form a first deformable mount 60, and the second and fourth base members 24,
28 may
similarly operate in conjunction with the second mounting beam 59 to forni a
second
deformable mount 62. The first and second deformable mounts 54, 56 may be
coupled to the
legs 30, 32 by the first and second base flexural pivots 44, 46, xespectively.
The deformable
mounts 60, 62 may function similar to springs, in that they elastically
deflect when the legs
30, 32 press outward against them.
The motion of the elements of a compliant mechanism is determined by the
geometry
of the elements. Thin or necked-down members, such as the deformable mounts
60, 62 and
the flexural pivots 40, 42, 44, 46, will flex as they receive a load
sufficient to cause a
deformation or a deflection. Conversely, thick members, such as the legs 30,
32 and the
shuttle 20, will remain substantially undeflected under loading of the
mechanism 10.
Typically, some minor bending will occur in thicker members such as the legs
30, 32 and the
shuttle 20; however, such bending is small in comparison to the deflections of
the thimier
members. The combination of the flexural pivots 40, 42, 44, 46 and deformable
mounts 60,
62 with the legs 30, 32 provides the range of motion necessary for the
mechanism 10 of
Figure 1 to move between the first configuration 54 and the second
configuration 56.
The deformable mounts 60, 62 may be undeflected, or only slightly deflected,
when
the shuttle 20 is in the first position 50 or the second position 52 so that
the first and second
configurations 54, 56 are both low potential energy states. However, when the
shuttle 20 is
disposed between or outside the first and second positions 50, 52, the
deformable mounts 60,
62 may be deformed to a larger extent to store a larger amount of potential
energy, and
_g_
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
thereby provide the impetus for the mechanism 10 to return to the first
configuration 54 or
the second configuration 56.
Thus, the mechanism 10 obtains its stability in two configurations from the
existence
of two low potential energy positions. A mechanism implementing low potential
energy
positions provides better control and a larger range of design possibilities
than do bistable
mechanisms implementing residual stress or buckled beamed methods. The
mechanism 10 of
Figure 1 uses a combination of the flexural pivots 40, 42, 44, 46, the
deformable mounts 60,
62, and the legs 30, 32 to create at least two low potential energy positions.
When the shuttle 20 is disposed between the first and second positions 50, 52,
the
deformable mounts 60, 62 may be bent outward. When the shuttle 20 is disposed
outward of
the first position 50 or the second position 52, the deformable mounts 60, G2
may be bent
inward. Deformable mounts need not be configured in the same manner as the
deformable
mounts 60, 62 of Figure 1, but may involve the use of fixed-free cantilever
mounting beams,
zigzagging flexural segments, or the like, as will be shown and described
hereafter.
The mechanism 10 may also be described as a compliant bridge 70 coupled to and
integrally formed with the base members 22, 24, 26, 28. The braclcet labeled
70 indicates
that, in the mechanism 10 of Figure l, the compliant bridge 70 includes the
shuttle 20, the
flexural pivots 40, 42, 44, 46, and the legs 30, 32.
The compliant bridge 70 may be selectively disposable along a first path 72,
which
corresponds to the first position 50 and the first configuration 54, and a
second path 74,
which corresponds to the second position 52 and the second configuration 56.
The paths 72,
74 may each be longer than the offset distance 29, so that the compliant
bridge 70 is curved
or kinked in each of the first and second configvmations 54, 56. Conversion
between the first
and second configurations 54, 56 may simply entail reversing the curvature or
lcinlcing of the
compliant bridge 70.
The compliant bridge 70 is curved or kinked in both configurations 54, 56
because the
tendency of the deformable mounts 60, 62 to stay straight is of a higher
magnitude than the
tendency for the compliant bridge 70 to return to its original state. For
example, the
mechanism 10 may be manufactured in the first configuration 54, so that there
is substantially
no deformation of the mechanism 10 in the first configuration 54. Thus, in the
second
configuration 56, the compliant bridge 70 may be generally bent in a direction
opposite to its
original, undeflected curvature. The compliant bridge 70 may thus exert
outward force
against the deformable mounts 60, 62 in the second configuration 56 due to its
tendency to
-9-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
return to its undeflected state. The deformable mounts 60, 62 must therefore
provide a
resilient force sufficient to counteract the outward pressure of the compliant
bridge 70 in the
second configuration 56, so that the compliant bridge 70 will remain deflected
in the absence
of any external force.
Such tendencies or strengths are a function of the geometry and positioning of
the
elements of the mechanism 10. In Figure 1, the deformable mounts 60, 62 have a
thicker
width than the flexural pivots 40, 42, 44, 46, so that the defonnable mounts
60, 62 have a
comparatively strong tendency to remain straight. The resilient forces
produced by the
deformable mounts 60, 62 and the flexural pivots 40, 42, 44, 46 can be
compared to springs
with different spring constants. Thiclcer elements are analogous to springs
with large spring
constants, and therefore resist deformation more than elements that would be
modeled with a
smaller spring constant.
The compliant bridge 70 may have a first beam section 80 that generally
encompasses
the first leg 30, the first shuttle flexural pivot 40, and the first base
flexural pivot 44.
Similarly, the compliant bridge 70 may have a second beam section 82 that
generally
encompasses the second leg 32, the second shuttle flexural pivot 42, and the
second base
flexural pivot 46. The compliant bridge 70 may also have a central portion 84
that includes
the shuttle 20.
The curvature of the paths 72, 74 may be somewhat exaggerated in the view of
Figure
1 to distinctly show the first and second configurations 54, 56; the paths 72,
74 may, in
reality, be nearly straight. The curvature of the paths 72, 74 may be adjusted
to suit the
desired force and displacement characteristics of the mechanism 10.
Additionally, multiple
other factors may also be adjusted to modify the operation of the mechanism
10.
When the input force 11 is applied to the shuttle 20, the shuttle 20 is moved
toward
the second position 52, thereby pressing the deformable mounts 60, 62 outward.
The
deformable mounts 60, 62 flex outward to accommodate the increased length of
the
compliant bridge 70 as the compliant bridge 70 deflects toward a straight
configuration. The
potential energy present in the mechanism 10 increases as the legs 30, 32
approach a linear
alignment; maximum potential energy may be reached at or near the point at
which the
compliant bridge 70 becomes completely straight.
The maximum potential energy position is the toggle point. At the toggle
point, the
compliant bridge 70 is in an unstable equilibrium, in which the compliant
bxidge 70 is equally
biased between the first and second configurations 54, 56, each of which has a
low potential
-10-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
energy. A slight displacement toward either of the configurations 54, 56 will
snap the
compliant bridge 70 into one of the configurations 54, 56. If continued force
is applied in the
direction indicated by the input force 11, the compliant bridge 70 will be
urged to bend by the
deformable mounts 60, 62 as the defonnable mounts 60, 62 relax into a
comparatively
straight, undeformed, configuration to provide the second stable configuration
56. The
deformable mounts 60, 62 may not fully relax in the second configuration 56
due to the
continued outward force exerted by the compliant bridge 70.
To actuate the mechanism 10 back to the first configuration 54, a return
force, or a
load sufficient to deflect the compliant bridge 70 beyond the toggle point,
towards the first
stable position 54, need only be applied in a direction opposite to that of
the input force 11.
The magnitude of the return force required to return the mechanism 10 to the
first
configuration 54 need not be equal to the magnitude of the input force 11
required to actuate
the mechanism 10 into the second configuration 56. Indeed, the input force 11
and the return
force may be specifically designed to suit the application in which the
mechanism 10 is used.
Through the use of compliant technology, a bistable mechanism may be produced
without conventional, separate members and joints. The functions of such
members and
joints are instead carried out by integrally formed elements that allow for
motion similar to
that of conventional mechanism. While a compliant structure can provide motion
and
displacement similar to that of designs involving conventional rigid elements,
the design
process for a compliant mechanism has obstacles not present in the case of
conventional
mechanisms utilizing separate, rigid parts. For example, the deformable mounts
60, 62 not
only serve as spring-type elements to bias the shuttle 20 toward the first
position 50 or the
second position 52, but they also function as attachments for the first and
second base
flexural pivots 44, 46. Thus, multiple forces and torques simultaneously act
on the
deformable mounts 60, 62. Additionally, comparatively large deflections may
occur.
As a result, determining how the material flexes requires complex non-liner
equations. The complexity of these equations makes it difficult to obtain
closed form
mathematical relationships between geometric characteristics of the mechanism
10 and the
resulting operational characteristics. "Geometric characteristics" refers to
physical properties
of a compliant structure, including member dimensions, member positions, and
material
properties. Hence, with traditional methods, it is difficult to create a
compliant structure that
functions in the desired manner, let alone optimize the structure for a
desired function. As
mentioned previously, the present invention presents a system and method
whereby
-11-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
compliant structures in general can be designed and optimized without the
computationally
intensive procedures that have been necessary in the past.
More specifically, the process of designing and optimizing a compliant
structure can
be simplified by modeling a compliant structure as a mechanism with rigid
members
connected with conventional joints and springs. Such a model may be called a
"pseudo-rigid
body model," which will be referred to as an analytical model in the following
discussion.
Refernng to Figure 2, an analytical model 110 of the mechanism 10 of Figure 1
is
depicted. Each elements of the mechanism 10 is present in the analytical model
110. More
specifically, the input force 11 may be modeled as a longitudinal spring 111
with a linear
configuration. Such a force may be applied by an actuator affixed to the
shuttle 20, which
may resist longitudinal motion of the shuttle 20 in either direction until the
actuator is
triggered. However, if the input force 11 does not vary in linear fashion with
the
displacement of the shuttle 20, some other model may be more appropriate. For
example, if
the shuttle 20 is part of an accelerometer and is actuated only by its own
weight, the
longitudinal spring 111 may be removed entirely from the analytical model 110,
or negated
by setting its spring constant to zero. The shuttle 20 may be represented by a
longitudinal
slider 120 that moves only in the longitudinal direction 12 by virtue of the
symmetry of the
model 110.
The legs 30, 32 may be modeled as legs 130, 132, which are connected by pin
joints
and torsional springs. More precisely, the first and second shuttle flexural
pivots 40, 42 may
be modeled with a single pin joint/torsional spring combination 140 on the
slider 120. The
torsional spring of the pin joint/torsional spring combination 140 represents
the resistance of
the shuttle flexural pivots 40, 42 to bending. The first and second base
flexural pivots 44, 46
may be modeled as first and second pin joint/torsional spring combinations
144, 146 attached
to the legs 130, 132. The torsional springs of the pin joint/torsional spring
combinations 144,
146 represent the resistance of the base flexural pivots 44, 46 to bending.
The resilient force of the first mounting beam 58 may be represented by a
first lateral
spring 155, and the resilient force of the second mounting beam 59 may be
represented by a
second lateral spring 157. The pin joint/torsional spring combinations 144,
146 may be
attached to first and second lateral sliders 158, 159, respectively, which
represent the physical
attachment provided by the first and second mounting beams 58, 59. The lateral
sliders 158,
159 may be constrained to move in the lateral direction 14 by the longitudinal
symmetry of
the deformable mounts 60, 62. The combination 160 of the first lateral spring
155 with the
-12-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
first lateral slider 158 may be analogous to the first deformable mount 60,
and the
combination 162 of the second lateral spring 157 with the second lateral
slider 159 may be
analogous to the second deformable mount 162.
The analytical model 110 has many "analytical model characteristics," each of
which
can be changed to alter the operational characteristics of the model 110. The
analytical
model characteristics are analogous to the geometric characteristics of the
compliant
structure. For example, the legs 130, 132 each have a length and a rest angle
with respect to
the lateral direction 14. Each of the pin joint/torsional spring combinations
140, 144, 146
may have a spring constant that indicates the strength of the torsional
spring. Each of the
linear springs 111, 155, 157 may also have spring constants. Mathematical
relationships
between the analytical model characteristics and operational characteristics
of the analytical
model 110 may be determined using traditional tools of kinematic analysis. The
analytical
model characteristics may be altered to simulate different configurations of
the mechanism
10 in its compliant form.
This type of pseudo-rigid body modeling can be applied to many different
compliant
structures besides bistable micromechanisms. After creating an analytical
model of the
compliant structure, conventional techniques can be applied to the analytical
model to obtain
closed form equations that relate operational characteristics to analytical
model
characteristics, such as the dimension, orientation, stress, force, thickness,
or width. When
the desired analytical model characteristics have been determined, a compliant
design may be
produced so that comparable operational characteristics can be obtained. This
process of
creating an analytical model of a compliant structure and then characterizing
the structure
with conventional rigid body equations allows for a large number of candidate
designs to be
considered for any given application. Additionally, once a feasible analytical
model is
determined, the model may be optimized to make the structure perform a
specific function in
a predetermined manner.
The initial step in optimizing a compliant structure is to determine the
desired
operational characteristics of the mechanism. For example, the maximum stress
on the
structure elements, such as the small-length flexural pivot, must be lower
than the yield
strength of the material to prevent failure. As another example, it may be
desirable to
minimize an input force, maximize an output force, maximize an output
displacement, or
obtain some other objective characteristic.
-13-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
The mechanism 10 in Figure 1 may be designed through the use of such an
optimization process. The mechanism 10 has several geometric characteristics
that must be
considered when performing an optimization process. First, the mechanism 10 is
to be stable
in two different configurations 54, 56. Bistable design can be difficult to
achieve because of
the complexity inherent in compliant mechanisms; the interactions of flexible
members,
analysis of large deformations, and the like make it difficult to determine
which sets of
geometric characteristics will provide a bistable design. However, through the
use of an
analytical model like that of Figure 2, multiple bistable configurations may
be found.
More specifically, certain ranges of analytical model characteristics may be
determined to yield bistable operation; such ranges may then form the
parameters of further
analysis to select the best specific configuration. Which bistable
configuration is the best
depends on the desired operational characteristics. One operational
characteristic may be
selected and designated an "objective characteristic," or the operational
characteristic that is
to be obtained through the optimization process. The objective characteristic
may be a target
value, such as a desired threshold input force to move the mechanism 10 from
the first
configuration 54 to the second configuration 56. Tn the alternative, the
objective
characteristic may be a value that is to be maximized or minimized; for
example, the
displacement of the shuttle 20 between the first and second positions 50, 52
rnay be
minimized.
Referring to Figure 3, a flowchart depicts one example of an optimization
process 210
according to the present invention. As mentioned previously, the optimization
process 210 is
applicable not only to bistable mechanisms or MEMS, but can be used for
compliant
mechanisms in general. As described previously, the first step may be to
select 212 a
compliant structure, such as the mechanism 10 of Figure 1, and an objective
characteristic to
be optimized. For the mechanism 10, the objective characteristic may be the
displacement
between the first and second positions 50, 52 of the shuttle 20.
Alternatively, a plurality of
objective characteristics may be selected and simultaneously optimized.
After the compliant structure and the objective characteristic have been
chosen, an
analytical model, such as the model 110 of Figure 2, may be created 214.
Traditional
analysis of rigid body members may be used to obtain mathematical
relationships that relate
the objective characteristic to characteristics of the analytical model, such
as member lengths
and angles, spring constants, and the like. Such mathematical relationships
may consist of
closed form equations that can be readily applied by a computer in an
iterative process.
-14-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
Alternatively, a compliant structure may be optimized without using an
analytical
model if a different form of analysis is used. For example, the compliant
structure may be
modeled by computer through the use of finite element analysis or a similar
method. Finite
element analysis creates a geometric model of the compliant structuxe by
dividing the
compliant structure into a large number of small geometric shapes, such as
tetrahedrons. A
computer then utilizes an iterative process involving simultaneous sets of
equations to
determine how the structure will respond to loads.
Thus, the finite element method may be used to evaluate multiple
configurations of
the selected compliant structure without creating closed form mathematical
relationships.
However, finite element analysis is computationally intensive, and would have
to be
performed with each iteration of an optimization algorithm. Consequently,
optimization
through finite element analysis may be too time consuming to be practical.
Therefore, the
following discussion assumes the use of an analytical model.
Returning to Figure 3, once the analytical model has been created 214, initial
values
of the characteristics of the analytical model should be selected 216. The
initial values may
simply be a guess as to what may be close to the optimal values of the
analytical model
characteristics; such a guess may be made through analysis of the model,
experimentation, or
experience.
A first recursive optimization algorithm may then be applied 222 to the
analytical
model. In general, the first recursive optimization algoxitlnn takes the
initial set of values for
the analytical model characteristics and performs an iterative process to
approach a global
optimum, or a set of analytical model characteristics that will produce the
most desirable
value of the objective characteristic. In the exemplary process 210 of Figure
4, the main
purpose of applying 222 the first recursive optimization algorithm is not
necessarily to find
the global optimum itself, but to find the general region in which the global
optimum lies
among local optima. This concept will be illustrated with greater clarity in
Figure 4.
The first recursive optimization algorithm may return a first set of values of
the
analytical model characteristics; the first sat of values bring the objective
characteristic near a
global optimum. A second recursive optimization algorithm may then be applied
224. The
second recursive optimization algorithm may receive the first set of values of
the analytical
model characteristics and may perform iterative steps to obtain a second set
of values of the
analytical model characteristics that provide a value of the objective
characteristic within a
tighter threshold of the global optimum. The threshold is necessary because
the second
-15-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
recursive optimization algorithm may otherwise come infinitely close to the
global optimum
without actually reaching it.
Once the second set of analytical model characteristics has been obtained, the
analytical model characteristics are utilized to derive 226 the geometric
characteristics of the
selected compliant structure that will obtain a similar result. Derivation of
the geometric
characteristics from the analytical model characteristics may be accomplished
by utilizing
known relationships between compliant structures and their rigid body
approximations. For
example, a torsional spring constant from the analytical model 110 may be used
to determine
a thickness for a small-length flexural pivot so that the small-length
flexural pivot simulates
the torsional spring constant in operation. Other geometric characteristics,
such as member
lengths, materials, member angles, and the like may be determined in a similar
manner, i.e.,
by comparison with the second set of values of the analytical model
characteristics.
The optimization process 210 is only one example of an optimization method
according to the present invention; many other processes may also be used. As
mentioned
above, a different method of analysis, such as the finite element method, may
be used in
place of creating 214 the analytical model. Additionally, rather than applying
222 a first
recursive optimization algorithm and applying 224 a second recursive
optimization
algorithm, a single optimization algorithm may be used to approach the global
optimum and
come within the threshold value.
Hence, steps of the optimization process 210 may generally be omitted,
modified,
and/or added to create new embodiments. However, the following discussion is
based on the
optimization process 210. Figure 4 is a graphical illustration of local and
global optima;
Figure 5 provides additional detail regarding the operation of one embodiment
first recursive
optimization algorithm, and Figure 6 provides additional detail regarding the
operation of one
embodiment of the second recursive optimization algorithm.
Referring to Figure 4, a graph 230 shows the value of the objective
characteristic,
mapped against the value of one analytical model characteristic. Although
optimization will
most likely be carned out by varying multiple analytical model
characteristics, the two-
dimensional representation of Figure 4 is somewhat easier to describe and
understand. Those
of skill in the art will recognize that the principles described below are
equally applicable to
optimization processes in which multiple characteristics are simultaneously
changed.
The vertical axis represents the value of the objective characteristic, such
as the
actuation force required to move the shuttle 20 of the mechanism 10 of Figure
1, between the
-16-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
first and second positions 50, 52. In this embodiment, it is desirable for the
actuation force to
be as low as possible. The horizontal axis represents the value of the
analytical model
characteristic that is to be varied to optimize the objective characteristic.
The domain
depicted by the horizontal axis represents the feasible design area for the
analytical model
characteristic that is to be varied.
As shown in Figure 4, there are multiple valleys 232 in which feasible designs
come
to relatively low (i.e., desirable) values of the objective characteristic.
One of the valleys 232
contains a local optimum 234, which is a lowest value of the objective
characteristic within
the valley 232 in which the local optimum 234 is present. The other valley 232
contains a
global optimum 236, which is a Iowest value of the objective characteristic
within the entire
feasible range of the analytical model characteristic. If a high value of the
objective
characteristic were desirable, local and global optima may be peaks instead of
valleys.
If the optimization process is not robust enough, the optimization algorithm
may only
obtain a local optimum, which would produce a less desirable design. Many
optimization
algorithms tend to return the optimum closest to the initial value provided to
the algorithm.
Consequently, if the initial value of the analytical model characteristic
falls within the valley
232 in which the local optimum 234 resides, such an optimization process would
likely return
only the local optimum 234, without finding the global optimum 236 that
provides the best
solution.
In order to avoid the local optimum problem, the optimization process 210 of
Figure
3, applies 222 the first recursive optimization algorithm primarily for the
purpose of escaping
from local optima, such as the optimum 234. The optimization process 210 then
applies 224
the second recursive optimization algorithm to find the value of the
analytical model
characteristic that will be within a smaller threshold of the global optimum
236.
Referring to Figure 5, a more detailed flowchart depicts steps that may be
followed to
apply 222 the first recursive optimization algorithm. The first recursive
optimization
algorithm may be any algorithm known in the art of mechanical design. As
mentioned
previously, the first recursive optimization algorithm may be used alone, if
desired. The
following discussion simply provides an example of one first recursive
optimization
algorithm that may be used in conjunction with a second recursive optimization
algorithm to
obtain robust and accurate results.
More precisely, the first recursive optimization algorithm may comprise a
process
known as simulated annealing. Other stochastic methods, such as genetic
algoritlnns, neural
-17-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
networks, and the like may also be used; however, the following discussion
assumes that
simulated annealing is the selected method. The simulated annealing process
utilizes random
decision-making to ensure that the process is able to move between the valleys
232. A first
ending criterion may first be selected 240. The first ending criterion may be
a required value
of the obj ective characteristic that must be obtained or passed in the
positive or negative
direction, a maximum number of iterations, a maximum change in the objective
characteristic
per iteration, or the lilce. An initial configuration may then be designated
242 as the
configuration with the initial set of values that were selected 216
previously.
The initial con figuration may be set 244 as the interim configuration. The
value of
the objective characteristic 246 may then be determined for the interim
configuration. The
value of the objective characteristic may be obtained through the use of
closed form
equations fiom an analytical model of the selected compliant structure, from a
computerized
process such as finite element analysis, or even through solving the complex
equations
involved with direct analysis of the selected compliant structure.
Once the value of the objective characteristic has been determined, the first
recursive
optimization algorithm may then determine 248 whether the first ending
criterion has been
achieved. As mentioned above, the first ending criterion may take a wide
variety of forms,
and may be tied to the value of the objective characteristic. If the first
ending criterion has
been achieved, the first recursive optimization algorithm may then set 250 the
first values for
the analytical model characteristics as the values of the interim
configuration. The first
values may then be returned by the first recursive optimization algorithm.
If the first ending criterion has not been achieved, the first recursive
optimization
algorithm may then randomly select 262 one of the analytical model
characteristics and
change the selected analytical model characteristic to form a changed interim
configuration.
Not all of the analytical model characteristics need be subject to change;
some may be
designated as characteristics to be held constant. The magnitude of the change
may also be
randomly determined if desired, or may be based in whole or in part on factors
such as the
results of previous iterations, the type of analytical model characteristic
that has been selected
at random, or the like. The first recursive optimization algorithm may then
determine 264
whether the changed interim configuration provides a better value of the
objective
characteristic than the original interim configuration.
In order to determine 264 which configuration is preferable, the value of the
objective
characteristic may be calculated or retrieved, if it has already been
determined and stored, for
-18-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
the interim configuration and for the changed interim configuration. For
example, if the
objective characteristic is minimization of the actuation force, whichever
configuration
provides the lowest actuation force provides the best value of the objective
characteristic.
If the changed interim configuration provides the better result, the changed
interim
configuration rnay be unconditionally set 266 as the new interim
configuration. The old
interim configuration may be discarded. The value of the objective
characteristic may then
be determined 246 with the new interim configuration to begin the cycle again.
If the changed interim configuration does not provide the better result, a
probabilistic
algoritlnn may be applied 268 to determine whether the changed interim
configuration should
be kept. The probabilistic algorithm may be as simple as the computer
equivalent of a coin
toss, or may be based in whole or in part upon othex factors such as the
difference between
the values provided by the two configurations, the history of the recursive
pxocess, or the
like. The first recursive optimization algorithm may determine 270 the results
of the
algorithm, and if the probabilistic algorithm dictates that the changed
interim configuration is
to be kept, the changed interim configuration may be set 266 as the new
interim
configuration.
The probabilistic algoritlun may also dictate that the changed interim
configuration
should be discarded 272, in which case the old interim configuration is kept,
and the cycle
may begin again. Reiteration with the same interim configuration will not
necessarily yield
the same result with each iteration because the first recursive optimization
algorithm may
randomly select 262 the analytical model characteristic to change, and
possibly even the
magnitude of the change. In all likelihood, the interim configuration will be
redefined
numerous times throughout the optimization process. Once the first ending
criterion has been
met, the first values 250 are set as the values of the interim configuration.
The random acceptance or rejection of an inferior interim configuration allows
the
first recursive optimization processes to escape local optima. As can been
seen in Figure 4,
were a configuration to be within the valley 232 of a local optimum 234, an
optimization
process may need to select a number of inferior designs in order to find the
global optimum'
236. In the case of the simulated annealing method, the algoritlnn that
selects or rejects an
inferior design may be based upon a cooling schedule of metal. The annealing
processes
would select fewer and fewer inferior designs as the process cycled. The
probabilistic
algorithm applied by the first recursive optimization algorithm may operate in
a similar
manner.
-19-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
After the first recursive optimization algorithm has been applied 222, a first
set of
values of the analytical model characteristics may be returned by the first
recursive
optimization algorithm. The first set of values may simply be one set of
values that positions
the objective characteristic somewhere within the valley 232 in which the
global optimum
236 resides. The second recursive optimization algorithm may then be applied
224 to obtain
a second set of values that are within a specified threshold of the global
optimum 236.
Referring to Figure 6, a more detailed view of the operation of the second
recursive
optimization algoritlnn is depicted. The second recursive optimization
algorithm need not be
designed to avoid local optima, but may simply operate to remain within the
valley 232 in
which the first values of the analytical model characteristics lie.
Consequently, gradient
analysis may be effectively used in conjunction with the second recursive
optimization
algorithm. Consequently, the second recursive optimization algorithm rnay
comprise the
generalized reduced gradient method or a similar optimization process. Other
gradient-
driven methods, such as sequential quadratic programming, linear programming,
and the like,
may also be used, but the following discussion assumes the use of the
generalized reduced
gradient method.
The second recursive optimization algorithm may first receive 280 the first
values of
the analytical model characteristics, and designate a second interim
configuration as the
configuration having analytical model characteristics with the first values. A
second ending
criterion may then be selected 282. Like the first ending criterion, the
second ending
criterion may take a variety of forms, and may be based upon the value of the
objective
characteristic.
The second ending criterion may be somewhat more restrictive than the first
ending
criterion; for example, the second ending criterion may require a
comparatively small
threshold distance between the global optimum 236 and the value of the
objective
characteristic. Such a criterion may be measured, for example, by measuring
the differences
between the values of the objective characteristics in successive iterations.
Very small
differences may indicate that the value of the objective characteristic that
is obtained is near
the global optimum 236.
The second recursive optimization algorithm may then determine 284 which
analytical model characteristics are to be kept constant during iteration.
Changing a larger
number of the analytical model chaxacteristics may provide a better result
because each
additional dimension of the analysis provides the possibility for a lower
global optimum. For
-20-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
example, if the graph 230 of figure 4 were expanded to three dimensions, the
new dimension
may well provide a global optimum more desirable than the global optimum 236
that is
available through variation of only one of the analytical model
characteristics. However,
changing a smaller number of analytical model characteristics may expedite the
procedure,
and may avoid introducing geometric characteristics that are not viable from a
manufacturing
standpoint.
The value of the objective characteristic may then be determiiled 286 with the
second
interim configuration. The second recursive optimization algoritlnn may then
determine 288
whether the second ending criterion has been achieved. If so, the second
values may be set
290 as the values of the analytical model characteristics of the second
interim configuration.
If the second ending criterion has not been met, all of the values of the non-
constant
analytical model characteristics may be changed to provide a second changed
interim
configuration. The changes may have a random component, or may be determined
entirely
by factors such as the effects of changes in past iterations on the objective
characteristic.
A vector may then be created 298 to describe the changes made to the
analytical
model characteristics. The vector may take the form of an ordered number set
including all
changes made. Properties of the vector may be measured 300 to determine which
changes
are improvements and which are not, e.g., which changes bring the second
changed interim
configuration comparatively closer to the global optimum 236. Such properties
may be
measured through the use of partial derivatives; the partial derivative of
each analytical
model characteristic with respect to the objective characteristic indicates
the effect that
changes to that analytical model characteristic have on the objective
characteristic.
After the vector has been analyzed, the changes that have not had a positive
influence
on the second interim configuration may be discarded 302. The remaining
changes may be
incorporated 304 into the second interim configuration to form a new second
intexim
configuration. Thereafter, the process may continue cyclically until the
second ending
criterion has been achieved.
The result of the second recursive optimization algorithm is to provide the
second set
of values of the analytical model characteristics, which will provide the
globally optimum
value for the objective characteristic within a comparatively tight threshold.
If desired, such
a result could be obtained solely with the first recursive optimization
algorithm. However,
the randomized nature of the simulated annealing algoritlnn may make accurate
determination of the global optimum a comparatively time-consuming process.
Thus, the use
-21-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
of the second recursive optimization algorithm may make the optimization
process faster and
more accurate.
As mentioned previously, design and optimization methods presented herein are
applicable to a wide variety of compliant mechanisms besides the mechanism 10
of Figure 1.
Additionally, bistable MEMS devices according to the invention may be designed
in other
ways besides those outlined in Figures 3, 5, and 6. Many bistable co~gurations
are possible
aside from the mechanism 10 of Figure 1. Some such configurations will be
shown and
described in comlection with Figures 7 through 10.
Referring to Figure 7, an alternative embodiment of a mechanism 310 with
bistable
properties is depicted; the mechanism 310 may also be sized for MEMS
applications. The
mechanism 310 of Figure 7 may have elements similar to those of the mechanism
10 of
Figure l; however, the mechanism 310 may not have dual, laterally symmetrical
portions.
The mechanism 310 may, for example, have a shuttle 320 disposed to move in the
longitudinal direction 12, much like the shuttle 20 of Figure 1. The first and
third base
members 22, 26, the first leg 30, the first shuttle flexural pivot 40, the
first base flexural pivot
44, and the first mounting beam 58 may be substantially as shown and described
in
connection with Figure 1. As with figure 1, the first base member 22, the
third base member
26, and the first mounting beam 58 may form the first defonnable mount 60.
The shuttle 320 may be integrally formed with and coupled to the first shuttle
flexural
pivot 40. However, in place of the second shuttle flexural pivot 42, the
shuttle 320 may
simply abut a surface 321 against which the shuttle 320 is able to slide.
Thus, the lateral
distance between the first deformable mount 60 and the shuttle 320 is fixed.
The shuttle 320
is constrained to move longitudinally, Iilce the shuttle 20 of Figure 1;
consequently, the
shuttle 320 may move in a manner similar to the shuttle 20.
The first deformable mount 60 may flex outward as the shuttle 320 moves toward
the
toggle point in order to accommodate the aligned lengths of the shuttle 320,
the first Ieg 30,
and the flexural pivots 40, 44. The shuttle 320 may be then snap into one of
two low
potential energy positions as the first deformable mount 60 pushes inward,
toward its
undeformed configuration. The mechanism 310 is simple in design, and may be
actuated
between stable configurations with about half the input force required to
actuate the
mechanism 10, assuming that the displacement of the shuttle 320 between low
potential
energy positions is the same as that of the shuttle 20, and that the surface
321 provides
-22-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
negligible friction. A low input force may be beneficial for some
applications, but may also
cause instability in other applications.
Stability is generally proportional to the force required to toggle a bistable
mechanism
from one low potential energy configuration to another low potential energy
configuration.
An embodiment requiring a small force to toggle the mechanism is not as stable
as a
mechanism that xequires a large force. Nevertheless, a highly stable mechanism
will require
a comparatively larger force to actuate the mechanism between different stable
configurations. A high input force may place excessive energy burdens on a
system,
especially in MEMS applications, in which energy consumption is limited. In
such
applications, minimizing the toggle force may be desirable.
However, stability may be compromised if the toggle force is too low. For
example,
vibrations that occur from outside the system may toggle a bistable mechanism
with a small
input force requirement. Therefore, the application and energy requirements
must be
considered when selecting a bistable mechanism design. One method of
increasing stability
is to increase the number of legs in the bistable mechanism. Increasing the
number of legs
also increases the input force required to toggle the mechanism.
Referring to Figure 8, another embodiment of a mechanism 410 according to the
invention is depicted; the mechanism 410 may also be sized for MEMS
applications, and may
also be bistable. As depicted in Figure 8, the mechanism 410 may receive an
input force 411,
and may have a shuttle 420 disposable between two stable positions. The
shuttle 420 may be
connected to a first base member 422, a second base member 424, a third base
member 426, a
fourth base member 428, a fifth base member 430, and a sixth base member 432.
The shuttle 420 may be coupled to the first and third base members 422, 426 by
a first
leg 30 and to the second and fourth base members 424, 428 by a second leg 32.
Additionally,
the shuttle 420 may be coupled to the third and fifth base members 426, 430 by
a third leg
434 and to the fourth and sixth base members 428, 432 by a fourth leg 436. The
first and
second legs 30, 32 may be integrally formed with and coupled to fixst and
second shuttle
flexural pivots 40, 42, and to first and second base flexural pivots 44, 46.
Similarly, the third
and fourth legs 434, 436 may be integrally foamed with and coupled to third
and fourth
shuttle flexural pivots 450, 452, and to third and fourth base flexural pivots
454, 456.
The first, third, and fifth base members 422, 426, 430 may be connected in
series, and
the second, fourth, and sixth base members 424, 428, 432 may be similarly
connected in
series. Thus, a first deformable member 460, a second deformable member 462, a
third
-23-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
deformable member 464, and a fourth deformable member 466 are formed; each of
the
defonnable members 460, 462, 464, 466 is coupled to one of the base flexural
pivots 44, 46,
454, 456, respectively. The deformable members 460, 462, 464, 466 may be like
those 60,
62 of Figure 1, with the addition of stubs 468 to which the base flexural
pivots 44, 46, 454,
456 are attached. The stubs 468 may help to avoid plastic deformation in the
junctures
between the defornable members 460, 462, 464, 466, and the base flexural
pivots 44, 46,
454, 456. Without the stubs 468, stresses may concentrate in the junctures
between the
defonnable members 460, 462, 464, 466 and the base flexural pivots 44, 46,
454, 456.
The use of multiple sets of legs 30, 32, 434, 436 may prevent the shuttle 420
from
rotating or twisting in a non-linear path. Such rotating or twisting may cause
the shuttle 420
to toggle in an unpredictable manner. Also, the additional legs 434, 436
increase the amount
of force required to actuate the mechanism 420. As discussed above, the
actuation force
generally corresponds to the amount of force that is required to deflect the
deformable
mounts 460, 462, 464, 466 enough to allow the shuttle 420, legs 30, 32, 434,
436, and
flexural pivots 40, 42, 44, 46, 450, 452, 454, 456 to linearly align at the
toggle point.
The addition of two more deformable members 464, 466 constitutes additional
potential energy that must be added to the mechanism 410 by the actuating
force to toggle the
shuttle 420. Thus, by adding third and fourth legs 434, 436 and third and
fourth defonnable
mounts 464, 466, the amount of force required to move the mechanism 410 from
the first
stable configuration to the second stable configuration is roughly double that
of the
mechanism 10 of Figure 1. As a result of the greater required actuating force,
greater
stability is similarly obtained. The mechanism 410 effectively has two
compliant bridges,
one of which extends from the first deformable mount 460 to the second
deformable mount
462, and another of which extends from the third defonnable mount 464 to the
fourth
defonnable mount 466.
As shown in figure 8, the base members 422, 424, 426, 428, 430, 432 may talce
the
form of enclosed anchors 422, 424, 426, 428, 430, 432. The anchors 422, 424,
426, 428, 430,
432 may be affixed to a silicon substrate in stationary fashion. Figure 8 also
depicts the third
and fourth anchors 426, 428 as common elements of the first and third
defonnable mounts
460, 464 and the second and fourth deformable mounts 462, 466, respectively.
This allows
for material and size reduction, and also creates separately deflecting
deformable mounts
460, 462, 464, 466 for each of the legs 30, 32, 434, 436. Attaching each of
the legs 30, 32,
434, 436 to a single defonnable mount 460, 462, 464, 466 simplifies the design
of the
-24-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
mechanism 410 and the calculation of the input force 411 required to toggle
the mechanism
410.
However, the mechaiusm 410 may also be designed without the third and fourth
base
members 426, 428, so that the first and third legs 30, 434 are attached to a
common
deformable mount extending between the first and fifth base members 422, 430,
and the
second and fourth legs 32, 436 are also attached to a common deformable mount
(not shown)
extending between the second and sixth base members 424, 432. The deformable
mounts
would operate in a manner similar to that described above, but with a slightly
different type
of deformation. Such an elongated deformable mount may beneficially allow for
further
material and size reduction. Another alternative embodiment may have a single
central
anchor (not shown) with two cantilevered beams extending from opposite sides
of the central
anchor. The legs 30, 32, 434, 436 may then be coupled to the ends of the
cantilever beams to
provide the required inward force to produce bistable operation.
Another way of controlling the stability of a bistable mechanism such as the
mechanism 410 of Figure 8 is by adjusting the orientations of the mechanism
elements with
respect to each other. The relative angles of the elements control the
function of the
mechanism. Various angles of the mechanism 410 may be varied through a wide
range of
values to create an infinite number of operational variations. This allows a
mechanism like
the mechanism 410 to be custom designed to suit individual applications and
constraints.
More specifically, the mechanism 410 may be manufactured with certain angles
470,
472, 474, 480, 482, 484 in an undeformed configuration, which will be assumed
to be the
first stable configuration of the mechanism 410. A first base pivot angle 470
may be the
angle at which the first and second base flexural pivots 44, 46 extend from
the first and
second defoimable mounts 460, 462. A first leg angle 472 may be the angle at
which the first
and second legs 30, 32 extend from the first and second base flexural pivots
44, 46. A first
shuttle pivot angle 474 may be the angle at which the first and second shuttle
flexural pivots
40, 42 extend from the first and second legs 30, 32. A second base pivot angle
480, a second
leg angle 482, and a second shuttle pivot angle 484 may be analogous to the
first angles 470,
472, 474, but with reference to the third and fourth base flexural pivots 454,
456, the third
and fourth legs 434, 436, and the third and fourth shuttle flexural pivots
450, 452.
If the angles 470, 472, 474, 480, 482, 484 are not selected correctly, the
mechanism
410 may not be bistable. For example, the angles 470, 472, 474, 480, 482, 484
should all be
positive, i.e., in the direction shown, so that the legs 30, 32, 434, 436 do
not compress into a
-25-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
zigzag configuration with the flexural pivots 40, 42, 44, 46, 450, 452, 454,
456 when the
input force 411 is applied. Similarly, the angles 470, 474, 480, 484 should
preferably not be
too extreme to avoid zigzagging andlor the buildup of concentrated bending
stresses at the
locations where the flexural pivots 40, 42, 44, 46, 450, 452, 454, 456 attach
to the shuttle
420, the legs 30, 32, 434, 436, and/or the deformable mounts 460, 462, 464,
466.
The first angles 470, 472, 474 need not be the same as the second angles 480,
482,
484. If desired, different angles may be used for analogous parts of the
mechanism 410 to
obtain a specific input force-to-displacement curve, additional stable
configurations, or the
like.
Figure 8 further demonstrates the use of the mechanism 410 as a compliant,
bistable
electrical switch. In this embodiment, the shuttle 420 may be positioned to
close an electric
circuit when actuated to the second stable configuration. More specifically,
the shuttle 420
may positioned to abut an output force receiver 490, in the form of an
electrical connection
490, in the second stable configuration. The electrical connection 490 may
take the form of a
first contact 492 and a second contact 494 separated from the first contact
492 by an air gap
or an evacuated space. The shuttle 420 may be made of, coated with, or simply
attached to
an electrically conductive material capable of closing and completing the
circuit. When the
mechanism 410 is acW ated, the shuttle 420 simultaneously engages the ends of
the first and
second electrical contacts 492, 494 to close the circuit. The circuit may then
be opened again
by actuating the mechanism 410 back to the first stable configuration.
Although Figures l, 7, and 8 have shown embodiments in which comparatively
rigid
legs are connected by small length flexural pivots, other configurations may
accomplish
bistable operation without requiring rigid legs. For example, a single arched
beam may
perform the functions of the legs as well as the flexural pivots. Such a
configuration will be
shown and described in connection with Figure 9.
Referring to Figure 9, another embodiment of a mechanism 510 according to the
invention is depicted. The mechanism 510 may be bistable, and may be sized fox
MEMS
applications. Furthermore, the mechanism SIO, as shown, may be utilized as an
accelerometer. The mechanism 510 may be designed to receive an input force S 1
l, which
may come from acceleration. The mechanism 510 may have a shuttle 520
configured to
move in the longitudinal direction 12; the shuttle 520 may have a weight 521
affixed to or
integrally formed with the shuttle 520.
-26-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
The mechanism 510 may also have base members 422, 424, 426, 428, 430, 432 and
defonnable mounts 260, 262, 264, 266 like those of Figure 8. However, in place
of the legs
and flexural pivots of Figure 8, the mechanism 510 of Figure 9 may have a
first arched beam
530 with a first end 532 attached to the first deformable mount 260 and a
second end 533
attached to the second deformable mount 262, and a second arched beam 534 with
a first end
536 attached to the third deformable mount 264 and a second end 537 attached
to the fourth
deformable mount 266. The arched beams 530, 534 may each have a substantially
uniform
cross section.
The configuration depicted in Figure 9 may be the first stable configuration,
and may
also be the manufacturing configuration in which no significant deformation is
present. The
arched beams 530, 534 may be advantageous in that their structure is very
simple, and no
sharp discontinuities are present to cause stress concentrations. Furthermore,
they are
comparatively easy to manufacture. However, the arched beams 530, 534 may make
the
mechanism 510 somewhat more difficult to model due to their length and the
irregularity of
their deformation.
As the input force 511 is applied, the arched beams 530, 534 may be compressed
into
an "S" shape. When the shuttle 520 has passed the toggle point, the arched
beams 530, 534
may move into a second stable configuration, in which the "S" shape is
maintained by the
inward pressure of the deformable mounts 260, 262, 264, 266. In the embodiment
of Figure
9, the input force 511 may be applied by acceleration. More specifically, as
the base
members 422, 424, 426, 428, 430, 432 move in a direction opposite to that of
the input force
51 l, inertia will act on the weight 521 to draw the shuttle 520 in the
direction of the input
force 51 l, with respect to the base members 422, 424, 426, 428, 430, 432.
The mass of the weight 521 may be carefully selected so that the mechanism 510
snaps into the second stable configuration when the acceleration on the
mechanism 510
reaches a certain threshold. For example, the mechanism 510 may be used to
signal a vehicle
impact for subsequent diagnosis, or to trigger safety features such as airbags
in real time.
The mechanism 510 may simply be combined with the electrical connection 490 of
Figure 8
to close a circuit upon detection of impact. Such a bistable accelerometer may
be useful in
many other applications as well. The mechanism 510 may even be reconfigured to
provide
more than tvvo stable configurations, so that different configurations can be
obtained from
different accelerations.
-27-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
Like the mechanisms 10, 410 of Figures 1 and 8, the mechanism 510 may also be
described in terms of compliant bridges. More specifically, each of the arched
beams 530,
534 operates in conjunction with the shuttle 520 to form a compliant bridge.
A bistable mechanism according to the invention may be triggered in a wide
variety
of ways besides acceleration. With reference to MEMS applications, a bistable
mechanism
may be triggered by mechanical linear or rotary devices, electrostatic
actuators such as comb
drives, or actuators driven by thermal expansion. Such actuation will be shown
and
described in greater detail in connection with Figure 10.
Referring to Figure 10, yet another alternative embodiment of a mechanism 610
with
bistable MEMS application is shown. The mechanism 610 may be configured
somewhat
similax to tile mechanism 410 of Figure 8, with a shuttle 420 connected to
first, second, third,
and fourth legs 30, 32, 434, 436 by first, second, third, and fourth shuttle
flexural pivots 40,
42, 450, 452, respectively. However, only two base members 622, 624 are
present, and the
deformable mounts 460, 462, 464, 466 of Figure 8 have been replaced by first,
second, third,
I S and fourth deformable mounts 660, 662, 664, 666, which are provided as
examples of
alternative types of deformable mounts
A "defonnable mount" is not limited to the deformable mounts 60 62, 460, 462,
464,
466 depicted in Figures 1, 7, 8, and 9. Rather, any stricture that is capable
as being modeled
or functioning as a spring can be used to create multiple low potential energy
positions.
While the embodiments of the previous figures utilize a fixed-fixed beam in
each deformable
mount, other types of deformable mounts may be implemented.
For example, the deformable mount may be a beam that cantilevers from a rigid
surface at one end and is fixed to a small-length flexural pivot at the other
end. The bending
of the beam at the non-fixed end will provide the biasing force that creates
the two low
potential energy positions. Alternatively, the defonnable mount may be a fixed-
fixed arched
beam or a beam that is initially deflected. In yet another variation of the
bistable mechanism,
the defoimable mount may be a structure similar to the bistable mechanism
itself, but with
only one stable position; the deformable mount may simply resist any
deflection from the one
stable configuration to provide biasing force. One of ordinary skill in the
art will recognize
that there are many methods of creating a force to urge a compliant bridge
into multiple low
potential energy positions.
In the mechanism 610 of Figure 10, the first deformable mount 660 has a fixst
compressible member 670 that extends from the first base member 622 in
cantilevered
-28-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
fashion to attach to the first leg 30. Similarly, the second deformable mount
662 has a second
compressible member 672 that extends from the second base member 624 in
cantilevered
fashion to attach to the second leg 32. Each of the compressible members 670,
672 may take
the form of a relatively thin beam with a looped portion; the looped portion
may be
compressed to provide the biasing force to store potential energy, thereby
performing
essentially the same function as the mounting beams 58, 59 of Figure 1.
Similarly, the third and fourth deformable mounts 664, 666 may have third and
fourth
compressible members 674, 676, respectively. The third and fourth compressible
members
674, 676 may extend from the first and second base members 622, 624 to attach
to the third
and fourth legs 434, 436, respectively. Each of the compressible members 674,
676 may take
the form of a relatively thin beam with a bend or kink to ensure that the beam
will buckle
under compression. The compressible members 674, 676 may have such a shape
that the
buckling is elastic over the feasible range of travel of the shuttle 420.
The mechanism 610 may receive an input force from an actuator 680 disposed in
close proximity to the mechanism 610. As mentioned previously, a wide variety
of actuator
types may be used, including micromechanical devices such as worm gears, rack
and pinion
assemblies and the Iike, electrostatic actuators such as comb drives, and
thermal actuators.
The exemplary acW ator 680 of Figure 10 is a thermal microactuator. The
actuator 680 may
have a shuttle 682 positioned close to the shuttle 420 of the mechanism 410
when the
mechanism 410 is in the first stable configuration. The shuttle 682 of the
actuator 680 may
be coupled to a first base member 684 and a second base member 686 by a
plurality of
expansion members 690, each of which is comparatively long and thin.
When electric current passes from the first base member 684 to the second base
member 686, the current travels through the expansion members 690. The
expansion
members 690 heat up in response to the current, and expand to press inward
against the
shuttle 682. The shuttle 682 is thereby driven to move toward the shuttle 420
of the
mechanism 610. The shuttle 682 may contact the shuttle 420, and exert pressure
on the
shuttle 420 until the mechanism 610 has passed the toggle point.
After the mechanism 610 passes the toggle point, the potential energy stored
by the
mechanism 610, and more specifically, stored by the compressible members 670,
672, 674,
676, may act to move the mechanism 610 into the second stable configuration,
thereby
pulling the shuttle 420 out of contact with the shuttle 682. If desired, the
shuttle 420 may be
-29-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
moved back to the first stable configuration through the use of a second
actuator (not shown),
or through the application of electrical or thermal energy directly to the
mechanism 6I0.
In the alternative, the shuttle 682, and indeed, the entire actuator 680, may
be
integrally formed with the shuttle 420 of the mechanism 410. Thus, the
actuator 680 may be
used to push the mechanism 610 into the second stable configuration, and to
pull the
mechanism 610 back into the first stable configuration. Once the current is
disconnected, the
expansion members 690 will cool, thereby retracting the shuttle 682 to pull
the shuttle 420
back into its original position.
According to another embodiment, a bistable mechanism may incorporate the
functionality of an actuator. For example, such a mechanism may have members
that
thermally expand under the application of electric current to push a shuttle
into a second
stable position. The shuttle may return to the first stable position after the
current is
disconnected. In the alternative, the shuttle may remain at the second stable
position after the
current is disconnected, but application of current elsewhere on the bistable
mechanism may
serve to return the shuttle to the first stable position.
Figure 11 is an alternative embodiment of a bistable mechanism 710 that
employs pin
joints in MEMS application. The bistable mechanism 710 is similar to the
previously
discussed mechanism, in that it has a base member 712 and a shuttle 716 that
are coupled
together by a flexible leg 720. However, in the present embodiment the base
member 712
and the base end 724 of the leg 720 as well as the shuttle end 728 of the leg
720 and the
shuttle 716 are coupled together by pin joints 730, 732. The bistable
mechanism 710 has a
first stable configuration and a second stable configuration that are obtained
by applying an
actuation force 740 to the shuttle 716.
In the embodiment depicted in Figure 11, the bistable mechanism 710 actuates
when a
determined actuation force 740 is placed upon the shuttle 716 in a
longitudinal direction 12.
Similar to other embodiments previously discussed, the bistable nature of the
mechanism,
710 is in part produced by a fixed lateral 14 distance between the base member
712 and the
shuttle 716. In a MEMS application, the base member 712 is typically fixed to
the microchip
substrate 742 or is an integral part of the substrate 742. Alternatively, the
base member 712
may be fixed to another micromechanism which may or may not be fixed to the
substrate
742. Generally, the base member 712 needs only to substantially maintain the
lateral 14
distance bet~.veen itself 712 and the shuttle 716. In Figure 11, the shuttle
716 has a fixed
lateral 14 distance from the base member 712. This fixed lateral 14 distance
is produced by
-30-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
the shuttle being configured to slide on a surface 748 that maintains a
longitudinal 12 travel.
Thus, the lateral 14 distance between the base end 724 and the shuttle end 728
of the leg 720
is fixed.
This fixed lateral 14 distance allows the micromechanism 710 to function in a
bistable
manner. The bistability is obtained by the linear length of the leg 720 being
longer than the
lateral 14 distance between the base member 712 and the shuttle 716. The
linear length of
the leg 720 is measured as the distance from the axed pin joint 730 to the
floating pin joint,
730 add to the length from the center of the floating pin joint 732 to the
shuttle's 732 contact
with the sliding wall 748. In the configuration depicted in Figure 1 l, the
linear length of the
leg 720 is equal to the lateral distance between the base member 712 and the
shuttle 716
divided by the cosine of the offset angle 752 between the two pin joints 730,
732. Thus when
an offset angle 752 between the pin joints 730, 732 exists, the linear length
of the leg 720 will
be longer than the lateral 14 distance between the base member 712 and the
shuttle 716. This
requires the linear length of the leg 720 to compress as the shuttle 716 is
biased from the
stable configuration depicted in Figure 11 to the other stable configuration.
As the linear length of the leg 720 compresses, potential energy builds up
within the
leg 720. The potential energy is stored within the deformed and flexed
material of the leg
720. In the embodiment of Figure 11, the energy is stored within the curvature
of the arched
leg 720. The leg 720 is depicted in Figure 11 in the first stable
configuration that is at a
relative zero potential energy position. The potential energy is zero because
the curvature of
the leg 720 is at its rest state. If the leg is bent or deflected, the
potential energy will increase
in the leg, causing the leg 720 to bias baclc to its rest configuration. The
configuration of the
micromechanism 710 depicted has two rest configurations, the first stable
configuration and
the second stable configuration.
The first and second stable configurations correspond to two low potential
energy
positions along the travel of the shuttle 716. The mechanism also has a toggle
point that
corresponds to a high potential energy position and is positioned between the
two low
potential energy positions. Therefore, either direction the micromechanism is
biased from
the toggle point will cause the leg 720 to snap to one of the two low
potential energy or rest
positions. Because an object will tend to stay at the lowest potential energy
configuration
until acted upon by a force, the mechanism 710 remains stable in both
configurations. While
the embodiment depicted in Figure 11 shows an arched leg 720, any shape of leg
that is
capable of compressively or deflectively storing energy may be applied to the
current bistable
-31-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
mechanism. For example, a leg that is straight at the stable or rest
configurations may be
employed. The leg would plastically defect or buckle as the Ieg approaches the
toggle point.
The energy created by the deflecting or buckling would force the
micromechanism into one
of the stable configurations.
In another embodiment, the legs may be generally "V" shaped where two straight
legs
intersect at an angle. Potential energy is stored in the legs as the angle
between the two legs
increases or decreases. This embodiment, however, does produce high stress
levels at the
intersection point. This may cause problems at the micro level. In yet another
embodiment,
different raised and lowered contours may be applied to the sliding surface to
create high
potential energy positions. This may be accomplished by causing the leg 720 to
compress as
it slides over a raised contour in the sliding surface 748. The leg 720 will
tend toward a low
potential energy position on either side of the high potential energy raised
contour of the
sliding surface 748.
To facilitate actuation between stable configurations and to maintain the
curvature of
the leg 710, the leg 710 is coupled to the base member 712 and the shuttle 716
through the
use of pin joints. The micromechanism in Figure 11 implements two different
types of pin
joints; a fixed pin joint 730 and a floating pin joint 732. Pin joints 730,
732 differ in design
and manufacture fiom macro sized pin joints. Because of the material
properties of MEMS,
special designs of pin joints must be implemented.
The fixed pin joint 730 of the micromechanism 710 is made in the layering
process
used in MEMS, where different layer form different members of the joint. The
fixed pin
joint 730 has a central pin that is situated within a hole of a round or
otherwise shaped soclcet
752. The top of the pin has an enlarged head 756 to prevent the socket 752
from slipping off
of the pin. Typically, the socket 752 is manufactured on a layer above the
substrate 742 of
the chip to prevent suction. Stiction is caused by Van Der Waals and
electrostatic forces
between the members and the substrate 742. Once suction occurs, it is
difficult or impossible
to free the MEMS member. To prevent this, the mechanism members may be
maintained
above the substrate 742 and other members through the use of dimples. A dimple
is a small
section of a mechanism that extends from the bottom of the mechanism to engage
the
substrate 742. These dimples provide a relatively small surface area contact
with the
substrate 742, such that the Van Der Waals and electrostatic forces may be
overcome to
allow movement.
-32-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
The floating pin joint 732 functions in a slightly different manner than does
the axed
pin join 730. The floating pin joint 732 must not only be able to hinge two
members of a
mechanism 710, but it must also be capable of sliding over the substrate 742
without stiction
problems. In one embodiment, the floating pin joint 732 is comprised of a ring
760 rigidly
attached to a pin mount 764. The ring 760 is also rigidly attached to the
shuttle 716. The pin
mount 764 extends into the center of the ring 760 and a pin extends from the
surface of the
pin mount 764 and toward the substrate 742. The pin mates with a floating
socket 768 that is
also situated within the center of the ring 760. The floating socket 768 has a
bridging
member 772 that rigidly attaches the leg 720 to the floating socket 768. The
bridging
member 772 may also be configured to ride along the top of the ring 760 as the
pin, pin
mount 764, and ring 760 rotate relative to the floating socket 768, bridging
member 772, and
the leg 720. In one embodiment the ring 760, the floating socket 768, and the
leg 720 are
manufactured from a first layer and the pin mount 764 and the bridging member
772 are
manufactured from a second layer.
Another embodiment of a floating pin joint 732 is comprised of an inner disk
and an
outer ring. The inner disk is situated within the center of the outer ring.
The imler ring is
attached to a first arm that bridges over the outer ring and the outer ring is
fixed to a second
ann that extends in the opposite direction of the first arm. The first arm
further comprises a
lip that cantilevers over the inner disk to maintain their relative
engagement. The outer ring
and the inner disk rotate relative to each other where the first and second
arms maintain
engagement and prevent slipping.
Pin joints provide advantages in certain applications over fully compliant
joints. For
example, pin joints allow for larger travel with smaller mechanism members,
when compared
to fully compliant mechanisms. Additionally, pin joints have the advantage of
providing
large ranges of rotational motion with minimal effect on surrounding
components. In the
present embodiment of the bistable mechanism, the use of pin joints allows the
shape and
curvature of the leg to be relatively unaffected in the first and second
stable configurations.
On the other hand, a compliant joint may require a significant section of the
leg to bend and
deflect to allow for the full range of motion needed. While the bending and
deflecting of a
leg in certain bistable application is acceptable, the present embodiment
relies on the shape
and curvature of the leg to control the potential energy storage. It is the
control of this
potential energy that allows the mechanism to function.
-33-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
Potential energy storage is controlled by changing the curvature of the leg as
it moves
through the toggle point. The potential energy stored in the leg 720 increases
as the leg
approaches the toggle point and then correspondingly decreases as the leg 720
moves away
from the toggle point. The potential energy is stored by changing the
curvature of the leg
away from its rest position. However if in order to facilitate motion, the
curvature of the leg
continues to deflect after the toggle point, potential energy may increase in
the leg 720 after
the toggle point. If the potential energy continues to increase after the
toggle point, then the
mechanism may only have one low energy position and thus no longer be
bistable. Thus for
the present embodiment, it is advantageous to maintain the curvature of the
leg through the
entire range of motion. This curvature is most easily maintained through the
use of pin
j oints.
Figure 12 demonstrates an alternative configuration of a bistable mechanism
810 that
uses two sets of legs that are coupled to pin joints. The mechanism 810 is
depicted in the
first stable configuration and functions similarly to the single leg
embodiment depicted in
Figure 11. This embodiment comprises a first set of legs 812 and a second set
of legs 816.
Each leg in the sets 812, 816 has a base end 820 and a shuttle end 824,
wherein the base ends
820 is coupled to a fixed pin joint 826 and the shuttle end 824 is coupled to
a floating pin
joint 827. Each base end 820 is connected to the fixed pin joint 826 by a
socket 828 that is
attached to the base end 820 of the legs 812, 816. The socket 828 has a hole
through which a
pin protrudes (not visible in Figure 12) to rotatably mount the leg 812, 816
to the substrate
832. The socket 828 is capable of rotating about the pin.
In the embodiment depicted in Figure 12 the base member to which the fixed pin
joints 826 are attached may be a microchip substrate 832. However, the base
member can
take various embodiments such as a raised structure above the substrate 832 or
a member of
another mechanism. The base member need only maintain the relative lateral 14
distance
between the base end 820 and the shuttle end 824. Other embodiments of the
fixed pin joints
are possible and are within the scope of this application. Additionally, the
fixed pint joint
may be replaced with a fully compliant mechanism, such as a small length
flexural pivot.
While this type of joint may not be a preferred embodiment for the given
bistable mechanism
810, one of ordinary skill in the art will recognize compliant joint designs
that will not
interfere with the bistable nature of the mechanism 810.
While the base ends 820 of the legs 812, 816 are couple to fixed pin joints
826, the
shuttle ends 824 are couple to floating pin joints 827. The floating pin
joints 827 provide a
-34-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
rotational coupling between the legs 812, 816 while also having the ability to
translate with
the displacement of the shuttle 836. The shuttle end 824 is coupled to the
bridging member
838 that is attached to the floating socket 840. The shuttle 836 end is
conversely attached to
the ring 842 and the pin mount 844. The pin extends from the surface of the
pin mount 844
into a hole in the floating socket 840. This pin engagement allows the legs
812, 816 to rotate
relative to the shuttle 836. The floating pin joint 827 may also translate
along the surface of
the substrate 832 because no element of the floating pin joint 827 is fixed to
the substrate
832. Additionally, the minimal surface engagement of the floating socket 840
and the ring
842 helps to limit stiction between the floating pin joint 827 and the
substrate 832. Thus, as
the mechanism 810 toggles from the first stable configuration to the second
stable
configuration, the floating pin joints 827 can translate along the surface of
the substrate 832.
Figure 13 depicts the micromechanism 810 of Figure 12 in the second stable or
actuated configuration. As discussed above, this figure demonstrates that the
position of the
floating pin joints 827 moves with the shuttle 836 when the mechanism 810
actuates. This
can be best viewed by comparing the longitudinal 12 position of the floating
pin joint 827 to
the longitudinal position of fixed pin joint 826. This comparison between the
mechanism's
810 position in Figure 12 and the position in Figure 13 demonstrates that the
floating pin
joint 827 travels along the substrate with shuttle 836.
While Figures 12 and 13 depict the floating pin joints 827 as an external
member of
the shuttle 836, the floating pin joints 827 may be an internal st1-ucture on
the body of the
shuttle 836. For example, a type of fixed pin joint that sits on top of the
shuttle 836 may be
implemented. However, current MEMS manufacturing processes requires the
connecting
members to be manufactured of the same silicon or other material. Therefore in
order to
place a type of fixed pin joint on the shuttle 836, the fixed pin joint 826
that is attached to the
substrate 832 should be raised to the same layer as the pin joint on top of
the shuttle. In yet
another embodiment, the pin joint attached to the shuttle end 824 of a leg
from the first leg
set 812 can be the same pin joint that connects a leg from the second leg set
816. Thus, the
attachment location of the two pin joints would be the same position on the
shuttle. In this
embodiment there is only one set of pin joints down the center of the shuttle
836. The pin
joint is coupled to two different legs that move relative to each other. In a
variation of this
embodiment, the two legs may share a common floating pin joint where no
shuttle is present
and the floating pin joint may be the actuation member. In yet another
embodiment, the
-35-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
floating pin joints may be replaced with a fully compliant joint, such as a
small length
flexural pivot.
Figure 13 further demonstrates that the curvature of the arched legs 812, 816
in the
second stable configuration is substantially similar to the curvature of the
arched legs 812,
816 in first stable configuration. As discussed above, the curvature of the
arched Legs 812,
8I6 correspond to the linear distance between the fixed pin joint 826 and the
floating pin
joint 827. Because the legs 812, 816 have only one fixed end (the base end
820), the rotation
about the fixed pin joint 826 is in an arched or circular path. This arched
path allows for two
locations of the shuttle end 824 to exist that have the same linear distance
between the fixed
pin joint 826 and the floating pin joint 827. These two locations allow the
arched legs 812,
816 to be in their non-deflected ox rest states which correspond to a common
curvature.
A close comparison of Figures I2 and I3 will reveal that there is a slight
variance in
the leg curvatures between the first arid second configurations in the two
drawings. This may
be attributed to the terminals 848 that the shuttle 836 abuts. To ensure
proper contact
between the shuttle 836 and the terminals 848, it is prefez~red that the
shuttle 836 continue to
bias against the terminals 848 when substantially at the second stable
configuration. In other
words, the shuttle 836 abuts the terminals 848 before the shuttle 836 reaches
the second low
potential energy position. As a result, in this embodiment, the two curvatures
may vary
slightly.
The terminals 848, depicted in Figure I3, are one embodiment of many possible
applications for a bistable micromechazlism. The present embodiment functions
as an
electrical switch. The micromechanism 810 in Figure 12 demonstrates an open
circuit
confzguratian and the micromechanism in Figure 13 demonstrates a closed
circuit
configuration. The shuttle 836 has a contact 852 that, when actuated, closes
the circuits as
shown in Figure 13. The terminals 848 have guiding protrusions that
facilitates a proper
engagement and alignment with the corresponding contact 852,
The bistable mechanism znay also incorporate other structures to help ensure
proper
engagemezxt. For example, Figure 13 depicts a slot 860 running in the
longitudinal 12 length
of the shuttle 836. Within the slot 860 is at least one guiding mount 864 that
is fixed to the
-36-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
substrate 832. The guiding mount 864 helps to maintain a linear path of travel
for the shuttle
836. Furthermore, the mechanism 810 can be assisted in traveling in a linear
path by
maintaining the first set of legs 812 substantially symmetric about the
shuttle 836 to the
second set of legs 816. This symmetry helps to keep the mechanism balanced
during
actuation. In the present embodiment, the guiding mounts 864 are "T" shaped to
prevent the
shuttle 836 from lifting from the substrate 832. One of ordinary skill in the
art will recognize
that there are multiple methods of maintaining a linear travel of the shuttle
836 by
implementing various structures and shapes. Additionally, the bistable
mechanism 810 has a
wide variety of applications that would be known to one ordinarily skilled in
the art,
including the applications of the fully compliant bistable mechanism 410
discussed above.
Figure 14 demonstrates a bistable micromechanism 810 coupled to an actuator
system
870. The actuator system 870 depicted in Figure 14 is capable of toggling the
shuttle 836,
fromm the ftrst stable configuration to the second stable configuration and
then back again to
the first stable configuration. The system 870 has a toothed extension 874
that is attached to
the shuttle 836. The toothed extension 874 is capable of driving the shuttle
836 in both
directions along the longitudinal 12 axis. The toothed extension 874 is
coupled to at least one
gear 878, the gear having a center hub 882. The gear 878 is capable of driving
the toothed
extension 874, which in tuna actuates the shuttle 836. In the embodiment shown
in Figure 14,
the gears 878 are driven by thermal in-plane micro actuators 886. The micro
actuators 886
are capable of a displacing an actuator shuttle 890 when a current runs
through them. By
tangentially connecting the shuttles 890 to the gear hubs 882, the actuator
886 can drive the
gear 878 by driving the gear hub 882 for a half rotation and the letting the
rotational
momentum drive the gear 878 slightly over the top of the gear hub 882. Then
when the
acW ator 886 retracts, it pulls the gear 878 back through the other half of
the rotational cycle.
Multiple other methods for toggling the mechanism 810 exist and are within the
scope
of the present application. For example, a single thermal in-plane
microactuator may be
coupled to the shuttle 836 for a single locking operation. Additional
actuators may be added
to displace the shuttle 836 between stable configurations. Further, actuators
may be attached
to the fixed pin joints 826 to induce a rotational force on the pin joint 826
sufficient to toggle
to actuator. One of ordinary slcill in the art will recognize that a wide
variety of embodiments
exist for the bistable mechanism and are within the scope of the present
application.
While the compliant bistable micromechanism provides the ability to toggle
between
two stable configurations, it is limited in the amount of force that can be
placed upon the
-37-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
mechanism without toggling it to the previous configuration. In order provide
a mechanism
that is capable of receiving a force larger than the actuation force, a
latching or locking
mechanism needs to be implemented. A latch mechanism is disclosed that is
configured to
induce a locked configuration and an unlocked configuration by actuating a
slider in a single
direction. A locking member actuates a first distance in an actuation
direction to induce the
locked configuration and the locking member actuates a second distance induce
the unlocked
configuration. The second distance is a displacement that is in addition to
the first distance
the loclc slider displaces. The latching mechanism may also have a detent
slider that allows
the lock slider to travel in a direction opposite the actuation direction.
An embodiment of the latching mechanism is depicted in Figure 15. Referring to
Figure 15, a latching mechanism 910 having a latched and an unlatched position
is depicted.
The latching mechanism 910 has lock slider 912 that is slidably disposed
substantially within
the perimeter of a decent slider 916. The lock slider 912 is maintained within
the perimeter of
the decent slider 916 by a raised member 920 that bridges over the lock slider
912. The
height of the raised member 920 is sized to engage a first stop 922 and a
second stop 923 that
are located on the lock slider 912.
The detent slider 916 is slidably disposed withal a base member 924. The base
member 924 may be a housing or a surface which provides a structure from which
the lock
slider 912 and the detent slider 916 may travel in relation thereto. In a MEMS
application the
base member would be attached to the-substrate to allow only in-plane travel.
In Figure 15,
the detent slider 916 is slidably maintained on the surface by a bracket type
base member
924. The lock slider 912 is similarly coupled to the surface by a guiding
mount 926 fixed
within a guiding slot 927. The guiding mount 926 has an overhanging head that
extends over
the lock slider 912 to confine the travel of the lock slider 912 to the length
of the slot 927.
The lock slider 912 also has a locking mount 928 that is conf°igured to
engage a
grasping member 932. In the embodiment shown in Figure 15, the grasping member
932 is a
mechanism that has a fixed end and a free end that are separated by a
generally flexible arm.
The grasping member 932 comprises an engagement member 936 and a disengagement
member 940 that are located at the free end. The arm is also a compliant
mechanism, in that
it pivots about a single piece integrally formed member. Additionally, the
grasping member
932 may be a compliant member because it 932 resist deflection in a spring-
Like manner
without multiple parts or additional force absorbing members.
-38-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
In the embodiment of Figure 15 the grasping member 932 pivots about a mounting
member 944 that is in a fixed location. The engagement member 936 is
configured to engage
the locking mount 928 in a locked configuration and the disengagement member
940 is
configured to engage the locking mount 928 in an unlocked configuration. The
locked
configuration and the unlocked configuration can be induced by an actuation
force on the
lock slider 912 in a single actuation direction. The detent slider 916 is also
configured to
engage the engagement member 936 of the grasping member 932 to maintain the
unlocked
coWguration. Once the detent slider 916 engages the engagement member 936, the
lock
slider 912 can be retracted from the grasping member 932 by a biasing force
that may or may
not be induced by an actuator.
The relative configurations and motion of the members of the latching
mechanism
910 may be best illustrated by referring to Figure 16. Figure 16 is an
operational view of the
latching mechanism represented by sub-figures 16A-E. Figure 16A is the initial
configuration of a lock slider 910 that is symmetric about the lateral 14
length of the lock
slider 912. This symmetry provides two grasping members 932 that are capable
of engaging
the locking mount 928. In this configuration, the lock slider 912 and the
associated locking
mount 928 are not engaging the grasping members 932. This is the initial
unlocked
configuration.
The lock slider 912 engages the grasping members 932 in Figuxe 16B upon
receiving
a locking force in the lateral 14 direction. The locking force is sufficient
to displace the lock
slider 9I2 a first distance. The Locking force need only be of a magnitude
capable of sliding
the lock slider a first distance and deflecting the grasping members 932 is a
locking
deflection which induces a locked configuration. However, the actuation
distance of the
locking force is prefexably the minimal length required to engage the locking
mount 928 to
the grasping members 932. In the embodiment depicted in Figure 16, the first
distance is
preferably identified as the distance required to engage the locking mount 928
to the grasping
members 932 or of a distance equal to the distance between the first stop 922
and the raised
member 920.
In this Ioclced configuration, as depicted in Figure 16B, the latching
mechanism 9I0 is
substantially more stable that a typical bistable mechanism. Because the
locking mount 928
has an interference engagement with the engagement members 936, the engagement
is
capable of sustaining a force opposite the actuation direction that is larger
than the actuation
force. This is in contrast to a typical bistable mechanism where the force
required to actuate
-39-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
the mechanism back to its first stable configuration is equal to or less than
the force that
originally actuated the mechanism to the second stable configuration. Thus,
the amount of
force that can be applied to the lock slider 912 in a direction opposite the
actuation direction
is only limited by the structural and engagement properties of the grasping
member 932 and
the locking mount 928.
In a preferred embodiment, the engagement of the grasping members 932 and the
locking mount 928 is accomplished by two members locking in the in the same
plane. If one
member slips above or below that plane, then the lock slider 912 and the
grasping member
932 will disengage. The disengagement force would thus be the force required
to induce an
out-of plane 16 slipping between the two members. However, various geometric
features can
be implemented to the latching mechanism 910 to prevent slipping. For example,
an
overhanging lip may be added to prevent slipping or the engagement member 928
may have a
groove that prohibits out of plane slipping.
Alternatively, the force that the lock slider 912 is capable of sustaining may
be
limited by the weakest failure mode of the grasping member 932. In the
embodiment
depicted in Figure 16, the weakest failure mode may be the tensile strength of
the thin
flexible arms 952. If a force is placed upon the lock slider 912 that induces
a tensile force in
the thin flexible arms greater than the maximum tensile load, then the
locl~ing mount 928 will
be released from the grasping member 932, and will return to the original
position of Figure
16A. Additionally, stress concentrations in the engagement members 936 or in
the
intersection of the thin flexible arms 952 with the mounting members 944 may
cause the
members to fail and allow the lock slider to return to the original position
of Figure 16A.
Such stress concentrations may be caused by the geometry of the grasping
members 932 or
the locking mount 928. Further, fatigue stresses in the members may also
create failure
location in the latching mechanism 9I0.
While, the latching mechanism 910 is capable of maintaining a large force in a
direction opposite the actuation direction, the latching mechanism is capable
disengaging the
grasping member 932 by a force generally equal to the actuation force. To
disengage the
lock slider 912 from the grasping members 932, a force is place upon lock
slider 912 in the
lateral 14 actuation direction that displaces the lock slider 912 a second
distance. As depicted
in Figure 16C, the second distance is a distance capable displacing the lock
slider 912 a
distance sufficient to engage the disengagement members 940 of grasping
members 932. The
detent slider 916 is configured to react to the second displacement of the
lock slider 912,
-40-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
wherein the reaction displaces the detent 916 slider from the position
depicted in Figures 16A
and 16B to the position depicted in Figure 16C.
In the embodiment shown, the detent slider 916 is displaced the second
distance when
the first stop 922 abuts the raised member 920. A discussed previously, the
lock slider 912 is
configured to slide within the detent slider 916. The raised member 920 is the
section of the
detent slider 916 connecting the two sides of the detent slider 916 and
bridging over the lock
slider 912. The raised member 920 is situated above the lock slider 912 at a
height sufficient
allow the lock slider 912 to slide relative to the detent slider 916. However,
the raised
member 920 is also of a height sufficient to abut the stops 922, 923 in
response to a
determined movement of the lock slider. In Figure 16C, the first stop 922
abuts the raised
member 920 when the lock slider 912 displaces the second distance. The second
displacement distance drives the detent slider 916 to engage the engagement
members 936 of
the grasping members 932. A shown in Figure 16C, the detent slider 916 lodges
between the
two grasping members 932, such that when the lock slider 912 retracts from the
disengagement members 940 the grasping members 932 are maintained in an
unlocked
configuration by the detent slider 916.
The locked and unlocked configurations, in the embodiment shown, relate to the
relative position of the grasping members 932 to the lock slider 912. The
locked
configuration, Figure 16B, occurs when the locl~ing mount 928 engages the
engagement
mounts 936 of the grasping members 932. In order to engage the engagement
mounts 936
the locleing mount 928 must first spread apart the grasping members 932 that
are spaced more
narrowly than the width of the locking mount 928. As the locking mount 928
first abuts the
grasping member 932, the angled edged of the locking mount 928 spread apart
grasping
members 932 until the flat edge of the locking mount 928 reaches the flat edge
of the
engaging mounts 936, as depicted in Figure 16B. This flexing and spreading of
the grasping
members 932 to engage the locking mount 928 and the resulting engagement is
refereed to
herein as the locking deflection. The locking deflection of the grasping
members 932 allows
the lock slider 912 and the grasping members 932 to create a locked
configuration.
The unloclced configuration, Figure 16C, occurs as the lock slider is
displaced in the
actuation direction and the locking mount 928 abuts the disengagement members
940 of the
grasping members 932. In the embodiment shown, the disengagement member 940 is
an
angled structure where the angle is substantially parallel to the angled edges
of the locking
mount, as shown. The disengagement members 940 of the two grasping members 932
are
-41-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
spaced at a distance narrower than the width of the locking mount 928. The
difference in
width between the locking mount 928 and the grasping members 932 causes the
angled edges
of the locking mount 928 to spread apart the disengagement members 940 and
resultantly
spread apart the grasping members 932 in an unlocking deflection. This
operates in a
ramping manner, such that as the disengagement member 940 is driven further
into the
angled edge or ramp, the grasping members 932 move further away from their
rest or locked
configuration.
In the embodiment shown, the spacing apart of the two grasping members 932 in
the
unlocking deflection is larger than the spacing of the two grasping members
932 in the
IO locking deflection. This difference in spacing between the grasping members
932 in the
locking deflection and grasping members 932 in the unlocking deflection is of
sufficient
spacing to allow the locking mount 928 to retract in a direction opposite the
actuation
direction without re-engaging the locking mount 928. As depicted in Figure
16D, the
grasping members 932 are maintained in the unlocking deflection by the detent
slider 916 at a
spacing larger than the width of the locking mount 928.
With the spacing of the grasping members 932 maintained by the detent slider
9I6,
the lock slider 912 can retract a distance such that when the detent slider
916 disengages the
engagement members 936, the locking mount will not re-engage the engagement
member
936. To disengage the decent slider 916 from the grasping members 936, the
lock slider is
equipped with a second stop 923. The stop 923 is raised above the surface of
the lock slider
912 at a height sufficient to abut the raised member 920 of the detent slider
916. Thus, as the
lock slider 912 retracts from the grasping members 932, the second stop 923
abuts the raised
member 920 and retracts the detents slider 916 fiom the engagement members 936
of the
grasping members 932. Once the decent slider 916 retracts, the grasping member
returns to
its initial unengaged configuration, Figure 16E.
One advantage of the present latching mechanism is that the mechanism may lock
and
unlock by a single directional force. Figure I 6 demonstrates this single
directional force by
the lateral 14 force arrows depicted in sub-figure 16B and 16C. The first
force 956 induces a
translation of the lock slider 912 the first distance, as shown in Figure 16B.
The force need
only be of sufficient magnitude and distance to induce an unlocking deflection
in the
grasping member 932 to engage the locking mount 928 in the locked
configuration. This
same force creating source that induced the loclced configuration may be used
to induce the
unlocked configuration. This second force 966, in the embodiment shown, may
also act upon
-42-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
the lock slider 912 in the same direction as the fzrst force 956. The second
force induces a
displacement in the lock slider 912 to translate a second distance, where the
second distances
is in addition to or longer than the first distance from the initial
configuration in Figure 16A.
The difference in displacement length of the lock slider 912 between these two
forces
956, 960 is demonstrated by the relative size of the arrows in Figure 16B and
C. The first
force 956 arrow is shorter than the second force 960 arrow. This indicates
that second
displacement is further from an initial reference point than is the first
displacement. Thus,
the lock slider 912 is further from its initial configuration, Figure 16A, at
the second
displacement, Figure 16C, than it is at the first displacement, Figure I6B.
While the net
distance of the second displacement is larger than the first displacement, the
actual distance
travel by the lock slider to induce the unlocked configuration need not be
longer than the
actual distance traveled to induce the locked configuration. The distance the
lock slider 912
travels from the initial configuration to the loclced configuration may be
longer or shorter
than the distance the lock slider 912 must travel from the locked
configuration to the
unlocked configuration, depending upon embodiment.
While a single force actuation is capable of inducing a loclced and unlocked
configuration in the mechanism, the force in a direction opposite the
actuation direction is
required to retract the lock slider 912 which in turn retracts the detent
slider 916. Various
embodiments may be applied to the present mechanism to provide this retracting
force
without the necessity of additional actuators. In one embodiment, the latching
mechanism is
microelectromechanical mechanism that is located on a microchip. In this
embodiment,
limiting actuators is important to size and power restrictions.
In a MEMS application, the first and second forces 956, 960 may be induced by
single directional actuator. One such actuator is the amplified thermal in-
plane microactuator
970 as depicted in Figure 17. The actuator, as shown, comprises three non-
amplified thermal
in-plane microactuators 972, 974 that are connected in an amplified
configuration. The
amplified configuration is created by coupling the base members 976 of a
central actuator
974 to the shuttles 980 of two side actuators 974. When the shuttles of the
side actuators 974
actuate, they amplify the displacement of a central shuttle 982 that is
coupled to the lock
slider 912.
The thermal in-place microactuators 972, 974 actuate their shuttles 980, 982
when a
current is passes through opposing base members 976, 984. The shuttles 980,
982 will
actuate a distance that corresponds to the amount to current within the each
actuator. Thus,
-43-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
simultaneously applying a current through all three actuators 972, 974 will
cause the central
shuttle 982 of the amplified actuator 970 to displace a first distance. If
that current is
increased, then the actuation distance will correspondingly increase.
Once the current is removed, the central shuttle 982 will return back to its
unactuated
position. The actuator 970 requires no additional foxce or energy to return to
the initial
configuration. Potential energy is generated when the shuttle 982 is
displaced. The potential
energy is then xeleased, driving back the shuttle with a determined foxce to
the initial
configurations. The foxce is sufficient to retract any member, such as a lock
slider 912, that
is coupled to the central shuttle when the current is removed.
When the amplified thermal in-plane microactuator 970 is attached to the lock
slider
912, a current of a determined magnitude within the actuator will displace the
lock slider 912
a first distance until it engages the grasping member 932. The current source
may then be
removed from the amplified actuator 970. Because the loclc slider 912 is
locked to the
engagement members 940 of the grasping members 932, the microactuator cannot
return to
the unactuated position. Thus, the potential energy generated by actuating the
central shuttle
982 is stored by the lock slider 912 and grasping member 932 engagement.
A second current, that is large than the fixst current, may next be applied to
the
microactuator to actuate the shuttle a second distance that is further from
the unactuated
distance than is the first distance. Because the central shuttle 982 is
already locked at the 1-irst
distance, energy is only need to actuate the central shuttle 982 the
difference between the first
distance and the second distance. This second distance drives the lock slider
912 into the
unlocked configuration which in fizm drives the decent slider 916 into an
unlocked
engagement with the grasping member 932. Once the current grasping members 932
are in
their unlocking deflection the current may be removed. Because the loclc
slider 912 is no
longer engaged with the engagement members 936, the potential energy in the
amplified
actuator 970 retracts the central shuttle 982 and the attached lock slidex 912
to the initial
position. While retracting, the second stop 923 of the lock slider 912 abuts
the raised
member 920 of the detent slider 9I6 to retract the decent slider 916 from the
grasping
member 932.
Other methods of creating a reactive force on the latching mechanism 9I0,
without
the necessity of additional actuation membexs, are possible. For example, the
lock slider 912
may be attached to a spring member that elongates as the lock slider travels
the first and
second distances. Once the actuation force is removed, the spring will retract
the lock slider
-44-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
912. Further, a spring member may be place in a configuration where the spring
compresses
as the Iock slider 912 travels the first and second distances. Once the
actuation force is
removed, the spring member will bias the lock slider 912 to its initial
position. The spring
member may be any number of members known to one of ordinary skill in the art.
For
example, a simple spring or an elastic member could be used to create the
retracting force. In
MEMS applications, a system of small length flexural pivots could be used to
create a spring
type force on the lock slider 912. These spring member applications would
allow the
actuation member that locks and unlocks the mechanism to not be rigidly
attached.
While it is preferably in one embodiment for only a single directional
actuator to be
used in the present latching mechanism, multiple actuators or force creating
devices can be
used to induce the locked and unlocked configuration as well as retracting the
lock slider 912
and the detent slider 916. For example, in a macro embodiment an operator's
hand could be
used to loclc, unlock, and retract the mechanism. In another embodiment, the
reaction force
of an impulse force acting upon the loclc slider 912 or other members could
retract the lock
slider 912 and the detent slider 916. In yet another embodiment, the lock
slider 912 may be
mounted in a manner so that gravity retracts the lock slider 912 and the
detent slider 916. In
this configuration, the detent slider 916 would be configured such that it
would not disengage
the grasping member 932 because of gravity, but would disengage when abutted
by the lock
slider 912. This may be accomplished by lower mass of detent slider 916 when
compared to
the lock slider 912 or by a high coefficient of friction material located on
the engagement
members 936.
Figure 18 demonstrates and isometric view of a preferred embodiment of the
latching
mechanism, wherein the mechanism has a lock slider 912, a detent slider 916,
and at lest one
grasping member 932. However, a significant number of variations of this
latching
mechanism can be created by modifying various elements of the mechanism. Many
of these
embodiments can be captured in a discussion of the method of latching a
mechanism
disclosed herein. First, a lock slider 912 or similar member is displaced a
first distance in an
actuation direction to engage a grasping member 932. The lock slider 912 and
grasping
members 932 may be of any structure or configuration that allows the two
members to
engage. This engagement may be within a plane defined by the longitudinal 12
and lateral 14
datum lines or may be in an out-of plane 16 direction, perpendicular to the
plane.
Next the lock slider is displaced a second distance in the same direction as
the first
displacement to create an unlocked configuration. Similar to the locked
configuration, the
-45-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
unlocked configuration can occur in any manner that allows the lock slider 912
to disengage
from the grasping member 932, within the longitudinal 12 and lateral 14 datum
line plane or
out of this plane 16. Then, the detent slider 916 is displaced a third
distance in the same
direction as the first and second directions to maintain the unlocked
configuration of grasping
member 932. This may be accomplished in many ways beyond simply wedging the
two
grasping members apart as depicted in Figure 16C. Additionally, the step of
displacing the
lock slider 912 the second distance and the step of displacing the decent
slider 916 the third
distance may occur simultaneously. Further, the second distance may equal the
third distance
in some embodiments. Having these two distances equal may be desired in
applications
where the distance an actuator is allowed to travel is limited.
Finally, the lock slider 912 is displaced a fourth distance that is in the
opposite
direction of the first, second, and third directions. This displacement of a
fourth distance is
sufficient to disengage the grasping member 932. Additional steps may also be
present in the
disclosed process, such as, the detent slider 916 may be displaced or
retracted in a direction
opposite the first, second, and third directions. When the detent slider is
retracted, it
disengages the grasping member 932 in the unlocked configuration. This process
may also
occur simultaneously with the step of displacing the lock slider 912 the
fourth distance. The
process described above is not intended to be limited by the structure
described above. The
process need only provide for a method of locking and unlocking a latching
mechanism by
applying a force in a single direction.
Other embodiments, beyond what are depicted in the figures may be used in the
latching mechanism. Some embodiments may not include all members as depicted
in the
figures. For example, the latching mechanism may only include a lock slider
912 and a
grasping member 932. This mechanism would operate in a similar manner as the
embodiment depicted in Figure 18, except that detent slider 916 would not be
required to
maintain the grasping members 932 in an unlocked configuration. The grasping
members
932 need only be maintained in an unloclced configuration for an amount of
time sufficient to
allow the lock slider 912 to retract from the grasping members 932.
To maintain the grasping members 932 in an unlocked configuration for a
duration of
time, a dampening member could be attached to the grasping members 932 to
prevent them
from returning to the locked configuration when the lock slider 912 is
removed. The
grasping member 932 would be induced into the unlocked configuration, as
discussed above.
However, once the lock slider 912 induces the unlocking deflection in the
grasping members,
-46-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
in need only be displaced in the direction opposite the actuation direction
before the
dampened grasping members 932 return to the locked configuration. This
dampening may be
obtained by attaching fasteners to the grasping member 932 than dampening
their motion
from the unlocking deflection to the locking deflection in the longitudinal
direction 12.
Alternatively, the material of which the grasping members 932 are comprised
may have
dampening properties that causes the grasping members 932 to return slowly to
the locking
deflection. In some embodiments the latching mechanism may be made of
plastics, metallic,
or silicon material. In yet another embodiment, the grasping members 932 may
be
submerged in a fluid that dampens the movement of the grasping members 932.
Various
structural members could be added to the latching mechanism to assist in the
process.
In other variations of the latching mechanism, the unlocking deflection need
not be
induced by the lock slider 912 abutting the disengagement members 940 of the
grasping
members 932. The present mechanism encompasses any method of opening grasping
members 932 or disengaging the lock slider 912 that may be accomplished when
moving the
Ioclc slider 912 or the detent slider 916 relative the grasping members 932.
For example, the
detent slider 916 may have a deflecting member, such as a protruding arm, that
engages a
disengagement member on the grasping member 932 to cause an unloclcing
deflection. In
another example, the disengagement members 932 may have a deflecting member
that
engages a disengagement member on the lock slider 912 of the detent slider
916. This could
be in the form of a paix of rigid ams extending from the grasping member 932
that engage a
ramped section to the detent slider 916 unlock the grasping member 932.
In the embodiment depicted in Figure 18, the grasping members 932 are induced
into
the unlocking deflection by the angled edge of the locking mount 928. However,
various
other method of inducing the unlocking deflection can be accomplished by
varying the
structure of the loclc slider 912. Various extending structures and ramped
member can be
employed to induce the unlocking deflection in grasping members 932. One
ordinarily
skilled in the art will recognize that there are many variations to the
invention that may be
accomplished without departing from the latching mechanism disclosed.
In yet another variation of the mechanism, the Loclcing mount 928 may deflect
to
engage rigid grasping members 932. The locking mount 928 and the grasping
members 932
need only deflect relative to each other, such that they interlock one with
the other. The
locking mount 928 could simply be a part of a flexible member that is
deflected into a
locking deflection with the grasping members 932. The Locking mount 928 could
have a
-47-
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
dampening member attached that allows the locking mount 928 to remain in the
unlocked
configuration fox a time sufficient to retract the lock slider 912 from the
grasping member
932. Alternatively, the detent slider 916 may engage the locking mount 928 to
maintain the
unlocked configuration until the lock slider 912 is retracted.
In another embodiment, the locking mount 928 need not be on the front section
of the
lock slider 912. The loclcing mount may be a number of notches or channels in
the side of
the lock slider. Additionally, the grasping members 932 may have a notch on
top of the lock
slider 912 for an out-of plane 16 engagement with the lock slider 912.
Multiple variations of
locking mount 928 or any section of the lock slider 912 to the grasping
members 932 are
possible and are within the scope of this application. In yet another
variation of the latching
mechanism, the position of the lock slider 912 relative to the detent slider
916 may be
reversed. The detent slider 916 may be slidably situated within the locking
mount 912. The
mechanism would simply operate similar to the embodiment of Figure 16.
The structure and the operation of the grasping members 932 could be easily
varied to
create alternative embodiments of the present mechanism. In the embodiment
shown in
Figure 18, the grasping mechanism 932 is a complaint mechanism where the thin
flexible arm
952 provides a flexible joint. However, other non-complaint structure could be
employed as
well as alternative embodiments of the compliant grasping member 932. For
example, the
grasping member 932 could be a spring loaded pivoting mechanism with two
engagement
members 936. Alternatively, the engagement members 936 could be configured to
face away
from each other to engage to the locking mount 928.
Besides the variations of the locking and unlocking structure of the latching
mechanism, the raised member 920 and the stops 922, 923 may also be varied.
For example,
the stops may be mounted on the sides of the loclc slider 912 in such manner
that it engages a
member of the detent slider 916 to drive the slider in the operation of the
mechanism. In yet
another variation, an actuator coupled to the mechanism may selectively abut
different
members during different displacement, thus eliminating the stops. For
example, an actuator
could be configured to drive the lock slider 912 a first distance, but then
actuate the lock
slider 912 and the detent slider 916 together the second distance. This could
be accomplished
by adding a structure to the detent slider 916 that abuts the actuator during
the displacement
of the second distance. Alternatively, the actuator could only drive the
detent slider the
second distance and a stop on the detent slider 916 could drive the lock
slider 912 the first
_q.8_
CA 02421755 2003-03-07
WO 02/23606 PCT/USO1/28614
distance. Various shaped and oriented protrusions or structures could be
employed to cause
the detent slider 916 to react to the movement of the lock slider 912.
In generating variations of the latching mechanism, varying the orientation of
the
mechanism should be considered as a method of generating a large number of
embodiments.
The mechanism as described in Figure 18 is substantially in-plane, meaning
that most of the
movement occurs in a single plane. However, the engaging of the locking mount
928 to the
grasping members 932 could be accomplished by changing the orientation of the
operation of
the mechanism members. For example the mechanism could be simply rotated by
90° about
the lateral axis 14. Thus, the locking and unlocking of the mechanism could
occur out-of
plane. The detent slider 916 might simply be a lifting ramp that raises the
locking mount 928
or the grasping member 932 above the engagement plane to allow the mechanisms
to retract
from each other.
The mechanisms disclosed herein may be modified in many other ways to suit a
wide
variety of applications. The invention may be embodied in other specific forms
without
departing from its structures, methods, or other essential characteristics as
broadly described
herein and claimed hereinafter. The described embodiments are to be considered
in all
respects only as illustrative, and not restrictive. The scope of intellectual
property rights is,
therefore, indicated by the appended claims, rather than by the foregoing
description. All
changes that come within the meaning and range of equivalency of the claims
are to be
embraced within their scope.
-49-
