U.S. patent application number 15/170355 was filed with the patent office on 2016-09-22 for silicon force sensor and method of using the same.
The applicant listed for this patent is Sandia Corporation. Invention is credited to Damon J. Burnett, Thomas B. Crenshaw, Paul C. Galambos, Jeffrey W. Lantz, Erik E. Nishida.
Application Number | 20160273978 15/170355 |
Document ID | / |
Family ID | 56234855 |
Filed Date | 2016-09-22 |
United States Patent
Application |
20160273978 |
Kind Code |
A1 |
Galambos; Paul C. ; et
al. |
September 22, 2016 |
SILICON FORCE SENSOR AND METHOD OF USING THE SAME
Abstract
The various technologies presented herein relate to a sensor for
measurement of high forces and/or high load shock rate(s), whereby
the sensor utilizes silicon as the sensing element. A plate of Si
can have a thinned region formed therein on which can be formed a
number of traces operating as a Wheatstone bridge. The brittle Si
can be incorporated into a layered structure comprising ductile
and/or compliant materials. The sensor can have a washer-like
configuration which can be incorporated into a nut and bolt
configuration, whereby tightening of the nut and bolt can
facilitate application of a compressive preload upon the sensor.
Upon application of an impact load on the bolt, the compressive
load on the sensor can be reduced (e.g., moves towards zero-load),
however the magnitude of the preload can be such that the load on
the sensor does not translate to tensile stress being applied to
the sensor.
Inventors: |
Galambos; Paul C.;
(Albuquerque, NM) ; Crenshaw; Thomas B.;
(Albuquerque, NM) ; Nishida; Erik E.;
(Albuquerque, NM) ; Burnett; Damon J.; (Sandia
Park, NM) ; Lantz; Jeffrey W.; (Albuquerque,
NM) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
Sandia Corporation |
Albuquerque |
NM |
US |
|
|
Family ID: |
56234855 |
Appl. No.: |
15/170355 |
Filed: |
June 1, 2016 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
14064573 |
Oct 28, 2013 |
9383270 |
|
|
15170355 |
|
|
|
|
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
G01L 1/18 20130101; G01L
5/0028 20130101; G01L 1/26 20130101 |
International
Class: |
G01L 1/18 20060101
G01L001/18; G01L 5/00 20060101 G01L005/00 |
Goverment Interests
STATEMENT OF GOVERNMENTAL INTEREST
[0002] This invention was developed under contract
DE-AC04-94AL85000 between Sandia Corporation and the U.S.
Department of Energy. The U.S. Government has certain rights in
this invention.
Claims
1. A method, comprising: applying a compressive preload to a
piezoresistive device, wherein the compressive preload is greater
in magnitude than an operating range of operational loads;
transmitting an operational load within the operating range to the
piezoresistive device so as to produce a reduction in the
compressive preload; and measuring the reduction in the compressive
preload to facilitate determining a magnitude of the operational
load.
2. The method of claim 1, wherein the compressive preload is
applied to the piezoresistive device by tightening of a nut on a
boltshaft of a bolt, wherein the bolt is located on a supporting
structure to which the operational load is applied.
3. The method of claim 2, wherein a bolthead of the bolt is located
in a hole extending from an outer surface of the supporting
structure to an inner surface of the supporting structure, and the
piezoresistive device is located on the boltshaft between the nut
and the inner surface of the supporting structure, the tightening
of the nut applying the compressive preload to the piezoresistive
device.
4. The method of claim 1, wherein the piezoresistive device is a
plate fabricated with a thinned inner region and a thicker outer
region, the thinned inner region facilitates inducing strain into a
conductive trace element located at the thinned inner region
facilitating the measuring of the reduction in compressive
preload.
5. The method of claim 1, wherein the piezoresistive device
comprising a plate of silicon.
6. A method, comprising: applying a compressive preload to a
piezoresistive device, wherein the compressive preload is greater
in magnitude than an operating range of operational loads;
transmitting an operational load within the operating range to the
piezoresistive device; and measuring the reduction in the
compressive preload.
7. The method of claim 6, wherein the compressive preload is
applied to the piezoresistive device by tightening of a nut on a
boltshaft of a bolt, wherein the bolt is located on a supporting
structure to which the operational load is applied.
8. The method of claim 7, wherein a bolthead of the bolt is located
in a hole extending from an outer surface of the supporting
structure to an inner surface of the supporting structure, and the
piezoresistive device is located on the boltshaft between the nut
and the inner surface of the supporting structure, the tightening
of the nut applying the compressive preload to the piezoresistive
device.
9. The method of claim 6, wherein the piezoresistive device is a
plate fabricated with a thinned inner region and a thicker outer
region, the thinned inner region facilitates inducing strain into a
conductive trace element located at the thinned inner region
facilitating the measuring of the reduction in compressive
preload.
10. The method of claim 6, wherein the piezoresistive device
comprising a plate of silicon.
11. A method, comprising: applying a compressive preload to a
piezoresistive device, wherein the compressive preload is greater
in magnitude than an operating range of operational loads;
transmitting an operational load within the operating range to the
piezoresistive device; and measuring the reduction in the
compressive preload to facilitate determining a magnitude of the
operational load.
12. The method of claim 11, wherein the compressive preload is
applied to the piezoresistive device by tightening of a nut on a
boltshaft of a bolt, wherein the bolt is located on a supporting
structure to which the operational load is applied.
13. The method of claim 12, wherein a bolthead of the bolt is
located in a hole extending from an outer surface of the supporting
structure to an inner surface of the supporting structure, and the
piezoresistive device is located on the boltshaft between the nut
and the inner surface of the supporting structure, the tightening
of the nut applying the compressive preload to the piezoresistive
device.
14. The method of claim 11, wherein the piezoresistive device is a
plate fabricated with a thinned inner region and a thicker outer
region, the thinned inner region facilitates inducing strain into a
conductive trace element located at the thinned inner region
facilitating the measuring of the reduction in compressive
preload.
15. The method of claim 11, wherein the piezoresistive device
comprising a plate of silicon.
Description
CROSS-REFERENCE TO RELATED APPLICATION
[0001] This application is a divisional of prior U.S. application
Ser. No. 14/064,573, filed Oct. 28, 2013, which is incorporated
herein by reference in its entirety.
BACKGROUND
[0003] Designing sensors that can measure and survive shock loads
in extremely harsh real-world environments, while still providing a
high signal to noise (S/N) ratio and significant bandwidth, can be
difficult. Further, a small sensor form factor (e.g. a
microelectromechanical systems (MEMS) sensor) can be necessary to
prevent the sensor deleteriously affecting the structure under
load, while facilitating arrayed and embedded applications.
Possible applications for a very small sensor that can survive high
shock loads and provide a high S/N ratio with a high bandwidth can
include crash-test dummy instrumentation, impact forensics,
embedded sensors in smart surfaces for lifetime structural health
monitoring, etc. Such sensors can indicate when a structural
element (e.g., an aircraft wing, a bridge pylon, etc.) has
experienced an out-of-bounds load due to an extreme event (e.g., an
explosion, an earthquake, etc.). A small sensor can also find
application measuring a system response to impact loads, as well as
for validating modeling and simulation tools.
[0004] There are three main types of normal load (pressure)
sensors: capacitive, piezoelectric, and piezoresistive. Each type
has performance and manufacturing advantages and shortcomings
associated therewith. Capacitive sensors typically have a very
small signal (e.g., change in capacitance) relative to the overall
sensor capacitance associated with the complete circuit.
Accordingly, a parasitic capacitance associated with signal lines
on a chip separated from a chip ground by a thin oxide layer can be
difficult to avoid as the ground is intimately connected to the
sensing element. Such factors tend to reduce the signal to noise
ratio for a capacitive sensor. Further, tight tolerances on the
small sensing capacitor gap can also be problematic as a very small
gap is desirable to achieve higher sensitivity, however a very
small gap can limit a range of gap deformation and further lead to
an increase in sensor response signal variability as a function of
dimensional variability of the gap.
[0005] Piezoelectric sensing elements are typically quartz crystals
which require poling and assembly, including bonding of the crystal
into an electronic package. Recent advances in the fabrication of
piezoelectric thin films are facilitating monolithic fabrication of
piezoelectric sensors with a reduction in the reliance on adhesives
at the critical sensing element level. An advantage of
piezoelectric sensors is that they can generate an output voltage
with no input voltage required, although an input voltage is
required for any in-package signal amplification or processing
electronics. However, piezoelectric sensors lack a robustness
required to survive dynamic shock loads of a high magnitude.
[0006] Typically a DC signal (e.g., a low frequency response) is
not sensed in a piezoelectric sensor, while it can be for a
piezoresistive sensor. Additionally, monolithic fabrication
techniques for piezoresistive silicon (Si) sensing elements are
well developed. Si has a large piezoresistive gage factor, can be
mass-fabricated, has a high elastic modulus and possesses a high
ultimate strength in uniaxial compression. However, Si is a brittle
material with tensile and flexural strengths being much lower in
comparison with its compressive strength performance. Such material
properties can render measurement of highly dynamic loads
problematic with a Si sensor material. An application which can be
particularly problematic is where dynamic loading can induce
tensile stress waves into structural components which can lead to
the generation of microcracking in the Si sensor material.
SUMMARY
[0007] The various, exemplary, non-limiting embodiments presented
herein relate to a sensor which can measure and survive shock
loading while still providing a high signal/noise ratio with
significant bandwidth are presented. In an exemplary, non-limiting
embodiment, a load sensing system can comprise a load sensor
comprising a piezoresistive sensor. The load sensing system can
further comprise a load application component configured to locate
the load sensor on a supporting structure. In a further embodiment,
the load application component can apply a compressive preload to
the load sensor, wherein the compressive preload results in
application of a compressive load to the piezoresistive sensor. And
in a further embodiment, while under an operational load, the load
application component can effect a relaxation in the compressive
load on the piezoresistive sensor.
[0008] A further exemplary, non-limiting embodiment comprises a
method for measuring a load while providing a high signal/noise
ratio with significant bandwidth. The method can comprise applying
a compressive preload to a piezoresistive device, wherein the
compressive preload is greater in magnitude than an operating range
of operational loads. The method can further comprise transmitting
an operational load within the operating range to the
piezoresistive device so as to produce a reduction in the
compressive preload. The method can further comprise measuring the
reduction in the compressive preload to facilitate determining a
magnitude of the operational load.
[0009] A further exemplary, non-limiting embodiment comprises a
load sensing system, wherein the system can comprise a load sensor
comprising at least one piezoresistive sensor, a first washer, and
a load application component. In an embodiment, the load
application component can comprise a nut and bolt, the bolt
comprising a bolthead, a boltshaft and a threaded portion of the
boltshaft. The load application component can be configured to
locate the load sensor and the first washer on a supporting
structure, wherein the bolt is located in a hole extending from an
outer surface of the supporting structure to an inner surface of
the supporting structure. The bolt can be located such that the
bolthead is situated proximal to the outer surface and the
boltshaft extends through the hole beyond the inner surface. In an
embodiment, the first washer can be situated between the bolthead
and the outer surface. The load application component can be
further configured to apply a compressive preload to the load
sensor by tightening the nut on the boltshaft, resulting in
application of a compressive load to the at least one
piezoresistive sensor. In an embodiment, the compressive load can
generate a first voltage state at the at least one piezoresistive
sensor. The load application component can be further configured to
receive an operational load on the bolthead. The operational load
can cause a compression of the washer and the operational load can
be transferred along the boltshaft causing a relaxation in the
compressive preload on the load sensor. In an embodiment, the
relaxed compressive load can generate a second voltage state at the
at least one piezoresistive sensor.
[0010] The above summary presents a simplified summary in order to
provide a basic understanding of some aspects of the systems and/or
methods discussed herein. This summary is not an extensive overview
of the systems and/or methods discussed herein. It is not intended
to identify key/critical elements or to delineate the scope of such
systems and/or methods. Its sole purpose is to present some
concepts in a simplified form as a prelude to the more detailed
description that is presented later.
BRIEF DESCRIPTION OF THE DRAWINGS
[0011] FIG. 1 is a sectional view of a sensor system, according to
an embodiment.
[0012] FIG. 2 is three dimensional view of a sensor system
according to an embodiment.
[0013] FIG. 3 is a schematic of a sensing element, according to an
embodiment.
[0014] FIG. 4 is a sectional view of a sensing element, according
to an embodiment.
[0015] FIG. 5 is a sectional view of a sensor assembly which
includes a sensing element, according to an embodiment.
[0016] FIG. 6 a photograph illustrating a sensor assembly which
includes a plurality of sensing elements, according to an
embodiment.
[0017] FIG. 7 is a photograph illustrating a side view of a sensing
element with an etched cavity, according to an embodiment.
[0018] FIG. 8 is a photograph illustrating a sensing element, with
a trace located thereon, according to an embodiment.
[0019] FIG. 9 illustrates a die mount to locate a sensor assembly,
according to an embodiment.
[0020] FIG. 10 is a block diagram illustrating a sensor assembly
with a portion of the sensor assembly protruding out of a die mount
opening, according to an embodiment.
[0021] FIG. 11 is a photograph illustrating a load sensor assembly,
according to an embodiment.
[0022] FIG. 12 is a FEM study illustrating strain distribution in a
sensing element, according to an embodiment.
[0023] FIG. 13 is a FEM study illustrating strain distribution in a
sensing element, according to an embodiment.
[0024] FIG. 14 presents a chart depicting the effect of fillet
radii on fracture load, according to an embodiment.
[0025] FIG. 15 presents a chart representing testing data for
quasi-static compressive loads for a number of sensing elements,
according to an embodiment.
[0026] FIG. 16 presents a chart representing testing data for
quasi-static compressive loads for a number of sensing elements,
according to an embodiment.
[0027] FIG. 17 presents a calibration curve for a sensor, according
to an embodiment.
[0028] FIG. 18 presents a calibration curve for a sensor, according
to an embodiment.
[0029] FIG. 19 presents a chart illustrating measured forces for a
square wave pulse input, according to an embodiment.
[0030] FIG. 20 presents a chart illustrating measured forces for a
triangle wave pulse input, according to an embodiment.
[0031] FIG. 21 presents a chart for a triangle shaped pulse with an
applied preload, according to an embodiment.
[0032] FIG. 22 is a sectional view of a sensor system, according to
an embodiment.
[0033] FIG. 23 is a flow diagram illustrating an exemplary,
non-limiting embodiment to facilitate measurement of an impact load
based upon a relaxation of a compressive load.
[0034] FIG. 24 is a flow diagram illustrating an exemplary,
non-limiting embodiment to fabricate a load sensing component.
DETAILED DESCRIPTION
[0035] Various technologies pertaining to a sensor which can
measure and survive shock loading while still providing a high S/N
ratio with a significant bandwidth are presented, wherein like
reference numerals are used to refer to like elements throughout.
In the following description, for purposes of explanation, numerous
specific details are set forth in order to provide a thorough
understanding of one or more aspects. It may be evident, however,
that such aspect(s) may be practiced without these specific
details. In other instances, well-known structures and devices are
shown in block diagram form in order to facilitate describing one
or more aspects.
[0036] Further, the term "or" is intended to mean an inclusive "or"
rather than an exclusive "or". That is, unless specified otherwise,
or clear from the context, the phrase "X employs A or B" is
intended to mean any of the natural inclusive permutations. That
is, the phrase "X employs A or B" is satisfied by any of the
following instances: X employs A; X employs B; or X employs both A
and B. In addition, the articles "a" and "an" as used in this
application and the appended claims should generally be construed
to mean "one or more" unless specified otherwise or clear from the
context to be directed to a singular form. Additionally, as used
herein, the term "exemplary" is intended to mean serving as an
illustration or example of something, and is not intended to
indicate a preference.
[0037] The various embodiments presented herein relate to a sensor
configured to measure high magnitude forces over a wide bandwidth,
where such forces can range from static loading through repetitive
loading extending to frequencies of several hundred kilohertz (kHz)
or more. In an embodiment, the sensor can be configured as a
washer-like device and can include at least one embedded
piezoresistive sensing element (e.g., a MEMS based sensor). In a
further embodiment, the sensing element can be formed from Si. A Si
sensing element can be fabricated to include a
photolithographically patterned strain gage(s) on a deformable
sense plate to provide a high sensitivity linear response (e.g.,
about 1-10 mV/1000 lbf, i.e., about 0.2-2 mV/kN) to input loads of
up to about 10,000 lbf (about 45000 N). The patterned strain gages,
or `wires` facilitate operation of the sensor as a bridge circuit
to generate a measurable voltage difference. As described herein, a
range of sensors can be fabricated to operate over a range of
applied loads by implementing relatively simple design changes, for
example, changing the diameter of a sense plate region, changing
the thickness of a sense plate region, etc. Small Si sensing
elements (e.g., of about 10 .mu.m to about 100 .mu.m in diameter)
enable multiple sensors to be embedded in a washer-like device to
facilitate redundant load measurement and further enable a
determination of loading direction (e.g., off-axis, angled or side
loads) by comparing outputs from sensors at different locations on
the washer.
[0038] Readily available non-brittle engineering materials having
ductile and/or compliant properties can be utilized to package the
Si sensor die(s) in a layered material stack arrangement that can
protect the brittle Si to facilitate survival of impact loads of a
high magnitude(s) For example, the enclosing materials can act to
dampen undesirable high frequency signals that may cause damage to
the Si and associated components. Such materials can include
glass-reinforced epoxy laminate sheets (e.g., a printed circuit
board, PCB), copper (Cu), polyimide, aluminum (Al), steel, etc. A
low-level (micro-) compressive preload can be applied to the sensor
to ensure that any gaps in the material stack are closed. Prior to
operation a high level (macro-) compressive preload can be applied
the sensor system including the Si sensing element. The magnitude
of the macro-preload may be selected, e.g., such that irrespective
of the operational loading on the sensor system, the Si sensing
element will always experience compressive loading.
[0039] The sensor design(s) facilitates control of a stress level
at the Si sensor die and thus the effective load range of the
sensor, while maintaining a high S/N ratio (e.g., of greater than
about 100:1). Further, the various sensor designs presented herein
are amenable to scalable manufacturing, take advantage of the high
sensitivity and high compressive strength of Si, and they
facilitate integration of the sensor with other sensors and
actuators as part of a complete impact response system. Further,
owing to application of a piezoresistive sensor, sensor response is
typically linear (e.g., as compared with a quadratic response for a
capacitive sensor), whereby the linear response can ease
calibration and electronic signal conditioning.
[0040] FIGS. 1 and 2 illustrate a load sensing apparatus 100 in
accordance with various non-limiting embodiments, where FIG. 1
illustrates a vertical section through A-A and FIG. 2 presents a
three dimensional (3D) representation. As shown in FIGS. 1 and 2, a
supporting structure such as a bolt 130 and nut 140 can be utilized
to locate a load sensor 120 on a housing 110. In an aspect, the
housing 110 can be a wall, a panel, a skin, a frame, or other
structure of a device into which the load sensor 120 can be
incorporated.
[0041] As also shown in FIG. 1, at least one sensing element 122
can be located inside the load sensor 120, with design and
operation of the sensing element(s) 122 described further herein,
particularly with reference to FIGS. 3, 4, and 5. The bolt 130
comprises a bolthead 131, a boltshaft 133 and a threaded portion
135 of the boltshaft 133 onto which the nut 140 can be attached
(e.g., nut 140 can be threaded onto the boltshaft 133). Bolt 130 is
located in a hole 116 formed (e.g., drilled) in the housing 110,
with the boltshaft 133 and the threaded portion 135 located in the
hole 116. The bolthead 131 is located in a recess 115 formed in the
housing 110 about the hole 116 (e.g., recess 115 is a countersunk
hole). In an embodiment, the bolthead 131 can be situated proximal
to an outer surface (or sidewall) 118 such that a load (e.g.,
P.sub.applied as further described herein) applied to the outer
surface 118 also acts, in part, on the bolthead 131. The base of
the recess 115 acts to retain the bolthead 131 while the boltshaft
133 extends through the hole 116 and protrudes from an inner
sidewall (or surface) 117 of the housing 110 opposite to the recess
115. Load sensor 120 having a washer-like configuration can be
located on boltshaft 132. Nut 140 can be utilized to locate the
sensor 120 against the inner sidewall 117. Such locating of sensor
120 against the inner sidewall 117 by nut 140 may be performed,
e.g., by tightening nut 140. The tightening of nut 140 can further
act to apply tension T on bolt 130 which locates bolthead 131 in
recess 115. Electrical connection to the sensing elements 122 can
be via connector 180, with signals (e.g., a change in voltage
applied to the at least one sensing element 122) being transmitted
and/or received by load measuring system 182. Based on electrical
changes at one or more sensing elements such as element 122, a
state of the at least one sensing element 122 can be determined,
such as the difference between a first state (e.g., a preloaded
state) and a second state (e.g., under impact loading). For
example, an impact load can cause a relaxation in the preloaded
state of a load sensor 120 (as further described below) and the one
or more sensing elements, such as element 122, can include a
Wheatstone bridge, whereby the change from the first loading state
to the second loading state causes a change in voltage across the
Wheatstone bridge. The Wheatstone bridge facilitates measurement of
the load on a sensing element 122 by balancing the resistance in
two legs of the bridge circuit.
[0042] It is to be appreciated that the term `locate` is utilized
herein to convey positioning and/or attachment of a component,
e.g., a first component relative to a second component. For
example, the first component can be located adjacent to the second
component. In another example, the first component can be located
with reference to the second component, hence, while the first
component and second component may not be adjacent or touching, the
second component (e.g., load sensor 120) may be located with
reference to a first component (e.g., any of an inner sidewall 117,
recess 115, hole 116, etc.), where the second component is
separated from the first component by a third component (e.g., any
of washers 150, 160, or 170 as further described below).
[0043] In an embodiment, rather than placing the load sensor 120
under direct compressive loading (e.g., the load sensor is located
between the nut 140 and the housing 110, as shown, e.g., in FIG.
22), a plurality of gapping components, e.g., washers 150, 160 and
170 can be utilized to apply a compressive preload, P.sub.preload,
to the load sensor 120. Application of a preload can facilitate
linearization of the response of load sensor 120 as a function of a
reduction in the compressive loading of the sensing element(s) 122
located inside the load sensor 120, as further described herein. In
an embodiment, washer 150 can be located, i.e., situated, in the
recess 115 between the bolthead 131 and the bottom of the recess
115. Washers 160 and 170 can be located on boltshaft 133, e.g.,
washer 160 can be located between the load sensor 120 and sidewall
117, while washer 170 can be located between the load sensor 120
and the nut 140, as illustrated in FIG. 1.
[0044] Washers 150, 160 and 170 can be formed from a material
having a ductile and/or compliant nature such that during
application of either of P.sub.preload or P.sub.applied, a load
(e.g., L.sub.1 at washer 160 and L.sub.2 at washer 170) is
maintained on load sensor 120, where such materials include, in a
non-exhaustive list, Cu, Al, polyimide, steel, PCB, etc. Washers
150, 160 and 170 can be formed from a resilient material, per the
above, which can have a degree of elastic recovery such that as a
load is intermittently placed on any of washers 150, 160 or 170,
the washer can be compressed and upon removal of the intermittent
load the washer can expand back to, or near to, an original size,
thereby compensating for the various dimensional changes in
structure 100 as the load is applied and removed.
[0045] A high degree of damping can increase a load survivable by
load sensor 120, but typically (e.g., with a conventional load
sensor) such damping can lead to a reduction in the bandwidth,
where bandwidth herein relates to the frequency range and/or
highest frequency being measured during loading. Accordingly, the
configuration illustrated in FIGS. 1 and 2 facilitates protection
of the load sensor 120, whereby any of the washers 150, 160 and/or
170, in conjunction with the bolt 130 and nut 140, can act to
absorb a damaging portion of an impact load P.sub.applied while
imparting a known portion of the load P.sub.applied through the
load sensor 120 (and accordingly any sensing element(s) 122
included therein).
[0046] In an embodiment, a macroscopic preload can be applied to
the load sensor 120 (and accordingly any sensing element(s) 122
included therein) based upon tightening of nut 140 along bolt
thread 135. In another embodiment, a microscopic preload can be
applied to a sensing element 122 based upon fabrication of the load
sensor 120, as further described herein.
[0047] During preloading, P.sub.preload, of the load sensor 120,
tightening of nut 140 along bolt thread 135 can place the load
sensor 120 and washers 150-170 under compression loading C, and
accordingly place the bolt 130 in a state of preload tension T,
with the compressive and tensile loading being respectively applied
by the shoulder 132 of bolt 130 and working face 142 of nut 140.
During operation, application of an impact load, P.sub.applied, on
the exposed upper surface 134 of bolthead 131, e.g., in direction
F, can act to reduce the preload tension T in the bolt 130 and
thereby reduce the P.sub.preload compression C on the load sensor
120. In an embodiment, if a transmitted portion of the impact load
P.sub.applied does not exceed the P.sub.preload compression C
applied to the load sensor 120, the load sensor 120 (and
accordingly any sensing element(s) 122 included therein) can
survive application of impact load P.sub.applied as the load sensor
120 has already withstood the more controlled (e.g., low loading
rate) application of P.sub.preload.
[0048] For example, sensing element(s) 122 can always be in a state
of compression, or near compression, where in an example if
P.sub.preload=10000N, and P.sub.applied=4700N, the sensing element
122 is still experiencing a compressive load of 5300N. Furthermore,
in another embodiment, the application of P.sub.preload can also
ensure that the load sensor 120 is in a state of compression of
sufficient magnitude to facilitate linear response of the load
sensor 120 during application of the impact load P.sub.applied.
Effectively, application of the P.sub.preload can remove any gaps
or looseness in the hardware stack (e.g., components 110, 120, 130,
140, 150, 160, and 170, and also components included in sensor 120
as described below) resulting in a tight component assembly that
can smoothly transmit a portion of the impact load P.sub.applied to
P the load sensor 120 (and accordingly any sensing element(s) 122
included therein). In an aspect, it can be considered that there
are two levels of protection, the microscopic compressive preload
applied to the sensor 120 and sensing element(s) 122 located in
load sensor 120, and the macroscopic compressive P.sub.preload
applied to the load sensor 120 and the portion of the P.sub.preload
acting to maintain the hardware stack tightly coupled.
[0049] In an embodiment, one or more sensing elements 122 can be
incorporated into a load sensor 120 to facilitate load measurement
as previously described with reference to FIGS. 1 and 2. FIGS. 3
and 4 illustrate a design for a sensing element 300 (which can be
considered to be equivalent to sensing element 122), according to
an embodiment, where FIG. 4 is a sectional view along B-B. The
sensing element 300 can be formed from a plate 310 of material,
which may, e.g., be single crystal Si.
[0050] In an example, the thickness of plate 310 can be on the
order of about 100 .mu.m thick to facilitate high load measurement.
Further, the thickness of the plate 310, or a portion thereof, can
be thinned (e.g., to a thickness of about 10 .mu.m) by any suitable
technique such as an etch process (e.g., a Bosch etch), where the
broken line 313 on FIG. 3 indicates the location of the sidewalls
313 of a thinned region 312 on FIG. 4.
[0051] Alternatively, plate 310 can be constructed using silicon on
insulator (SOI) device technology, whereby a wafer can be obtained
of a required thickness with and/or without the thinned region 312.
A surface (e.g., a top, or upper, surface in the plane of the
illustration of FIG. 3) of plate 310 can be patterned with at least
one trace, e.g., traces 320 and/or 330, as shown in FIG. 4. As
presented further herein, traces 320 and 330 can be formed from
four separate traces 320A and 320B, and 330A and 330B, where traces
320A and 320B are located relative to the periphery of thinned
region 312, and traces 330A and 330B can be located more centrally
relative to the thinned region 312.
[0052] As explained further herein, when a load (e.g.,
P.sub.preload or P.sub.applied) is applied to and/or distributed
across the top surface of the plate 310, plate 310 can become
deformed (e.g., by bending) about the thinned region 312.
Accordingly, deformation of plate 310 can lead to strain
transduction occurring in the traces 320 and 330 on the top surface
of the plate 310, where in where in typical configurations the
highest strain tends to occur. While the traces 320 and 330 can be
of any dimension and form, in an embodiment the traces 320 and 330
can be formed from any suitable conductor such as doped (e.g.,
greater than about 1.times.10.sup.26 cm.sup.-3) n-type polysilicon
and can be of any suitable dimension, e.g., about 0.3 .mu.m, in
width and height. In a further embodiment, traces 320 and/or 330
can be isotropic. In another embodiment, traces 320 and/or 330 can
have a piezoresistive gage factor coefficient of about -18.
[0053] As shown in FIG. 3, the traces 320A, 320B, 330A and 330B can
form a serpentine pattern repeatedly crossing back and forth over
the thinned region 312 of the plate 310 that can experience the
highest radial strain during deformation of the plate 310.
[0054] In an embodiment, the traces 320A, 320B, 330A and 330B can
be arranged as a full Wheatstone bridge, in which a plurality of
bridge resistance connections 340-370 are collectively patterned on
the surface of plate 310. Patterning the plurality of bridge
resistance connections 340-370 onto the surface of plate 310 can
mitigate any temperature variation effects as all the resistive
legs (e.g., traces 320A, 320B, 330A and 330B) of the Wheatstone
bridge can be subject to the same temperature and thermal boundary
conditions. Further, patterning the plurality of bridge resistances
340-370 onto the surface of plate 310 can also maximize sensitivity
as the opposing resistor pairs 340 & 350 and 360 & 370 in
the Wheatstone bridge can be subject to either tension (outer
resistors 340 and 370 located at the edge of plate 310) or
compression (inner resistors 350 and 360 located at the center of
plate 310). Patterning the plurality of bridge resistances can
cause differences in resistance due to strain to increase the
change in voltage, rather than offset it. In embodiments, the
Wheatstone bridge does need to be balanced. Obviating such a need
can facilitate monitoring the offset (e.g., at outputs 340 and 370)
for problems associated with electrical design, circuit
connections, input/output electrical connections, and the like. It
is to be appreciated that any suitable resistance can be utilized
to facilitate operation and sensing by the Wheatstone bridge (e.g.,
traces 320A, 320B, 330A and 330B), in an embodiment, a resistance
of each leg of the bridge can typically be in the order of tens of
kilohms.
[0055] While not shown in FIGS. 3 and 4, a layer of insulating film
(e.g., silicon dioxide (SiO.sub.2), and/or silicon nitride (SiN)
layer 570 shown in FIG. 5) can be deposited between the traces 320
and 330 and the surface of plate 310 to prevent shorting. Further,
while not shown in FIGS. 3 and 4, a layer of insulating film (e.g.,
layer 580 in FIG. 5) can also be applied to cover the surface of
the traces 320 and 330 to provide a measure of wear protection.
[0056] As previously mentioned, the brittle sensing element 300 can
be packaged to protect the sensing element from the most extreme
loads and can further transmit a percentage of the applied load to
the die 310. Protection afforded by sensor packaging enables the
sensor element 300 survive the high loads and loading rates. FIGS.
5-11 provide illustration of various packaging embodiments and a
testing configuration, and are presented to facilitate
understanding of calibration of an applied load to a measured
sensor response. Based thereon, various predictions can be made
regarding a transfer function through a sensor package for an
anticipated load level(s) for the various materials, stiffness, and
thicknesses chosen in the package design. The various calibration
curves, applied loads, FEM analyses, sensor dimensioning, material
selection, etc., are presented to facilitate understanding of the
various concepts presented herein and are not to be taken as
definitive values, with any parameter magnitude, material,
dimension, etc., being replaceable by any other value suitable in
accordance with the embodiments presented herein. Hence, the
parameter magnitude(s), material(s), dimension(s), etc., can be
selected to facilitate design of a sensor, and adjusted for
different anticipated loads in different sensor applications.
[0057] FIG. 5 illustrates a sensor assembly 500 which includes a
sensing element 300, according to an embodiment. As illustrated, a
plate 310 has located thereon traces 320 and 330, as previously
described in FIGS. 3 and 4. Thinned region 312 indicates where
material has been removed (e.g., by Bosch etching) to facilitate
thinning of plate 310 to a thickness R which is less than the
original thickness P of plate 310.
[0058] In an embodiment, the thinned plate region 312 can increase
the degree of deformation which can occur at that region relative
to a plate with a constant thickness P throughout.
[0059] Electrical connection to traces 320 and/or 330 is
facilitated by electrical contacts 510 which can be connected to a
printed circuit board (PCB) 530 via electrical contacts 520 and
590, bond pads 550 and trace elements 560. In an embodiment, the
bond pads 550 can be formed from any suitable material, e.g.,
aluminum. In an embodiment, trace elements 560 can be formed on the
PCB 530, and further can be formed from any suitable conductive
material, e.g., gold. In an embodiment, electrical contacts 590 can
be solder bumps and in a further embodiment, electrical contacts
520 can be a die attach, formed, for example, from indium (In). As
previously mentioned, traces 320 and/or 330 can be optionally
located between a lower insulating layer 570 and/or an upper
insulating layer 580. Further, an adhesive layer 540 can be located
between the upper layer 580 and the PCB 530, where adhesive layer
540 can be formed from any suitable material such as an epoxy
resin.
[0060] FIG. 6 is a photograph illustrating a sensor assembly 600
which includes a plurality of sensor assemblies 500 (and according
sensing elements 300), according to an embodiment. As previously
mentioned, the example configuration presented herein relates to a
load sensor 120 being located in a compressive loading system
comprising a bolt 130 and a nut 140.
[0061] Accordingly, as shown in the example configuration, a
plurality of sensing elements 300 (e.g., comparable to sensing
elements 122 of FIG. 1) can be located on a PCB 530, where in the
example embodiment, the PCB 530 is in the form of a washer-like
configuration to facilitate location of the load sensor 120 on to
the boltshaft 133. In the example configuration presented in FIG.
6, four sensing elements 300 (each located in a sensor assembly
500) are located on PCB 530, whereby readings from each of the four
sensing elements 300 can be analyzed and, based thereon, a
determination can be made regarding a direction of loading relative
to the operating plane of the four sensing elements 300. PCB 530
has further formed thereon a plurality of traces 610 which are
connected to respective inputs and/or outputs 340-370.
[0062] FIG. 7 is a photograph illustrating a side view of a sensing
element 300 with the etched cavity 312 readily discernible,
according to an embodiment. As shown, each sensing element 300 can
be located on PCB 530 by an adhesive layer 540, where during
location of a sensing element 300 on to PCB 530 the adhesive layer
540 can wick into a gap formed between the electrical contacts 520
on PBC 530 and the upper oxide layer 580. FIG. 8 is a photograph
illustrating a plate 310, with a trace 320 and/or 330 located
thereon, according to an embodiment. Further indicated on FIG. 8
are the bridge resistance connections 340-370. Electrical contacts
520 can be utilized to supply and/or measure voltage, etc., as
traces 320 and/or 330 undergo loading deformation. Further,
additional indium bumps 810 can also be utilized as symmetrical
support for electrical contacts 520.
[0063] FIGS. 9, 10 and 11 illustrate example configurations for
packaging of a sensor, according to an embodiment. As previously
mentioned, the various embodiments presented herein can be directed
towards incorporation of one or more sensing elements (e.g.,
sensing elements 300) into a washer-like configuration. As shown on
FIG. 9, a sensor assembly (e.g., sensor assembly 500) can be
incorporated into a die mount 910. Die mount 910 can include a
plurality of openings 920 (e.g., well cutouts) whereby each opening
can be utilized to locate a sensor assembly 500. As shown in FIG.
10, each sensor assembly 500 can be located in the die mount
opening 920 such that a portion of the sensor assembly 500 can
protrude, by distance M, above the surface H of the die mount 910.
Hence, during assembly of a washer assembly (e.g., assembly 1110
shown in FIG. 11), a washer or disc of material 1010 can be placed
on the exposed surface of the sensor assembly 500, such that a
micro-level preload can be applied to the sensing elements 300 in
the sensor assembly 500 which can further cause the washer 1010 to
deform over the surface of sensing element 300 by the distance M.
In an embodiment, washer 1010 can be formed of a compliant material
such as nylon. A second washer 1020 can be placed under the PCB 530
as required to facilitate protection of the sensing element
310.
[0064] It is to be appreciated that any number of layers can be
utilized during assembly and packaging of a sensing element 300.
Hence, once the final washer assembly 1110 is assembled with the
sensor assembly 500 and washers 1010 and 1020 located in an outer
shell 1120, the sensor assembly 500 and the respective sensing
elements 500 can have a micro-level preload applied thereto. During
subsequent assembly of the washer assembly 1110 (e.g., forming load
sensor 120) into the load sensing device apparatus 100, a further
compressive load (e.g., a macro-level preload) can be applied to
the sensor assembly 500, whereby the macro-level preload and the
micro-level preload can combine to form P.sub.preload.
[0065] To facilitate understanding, the following presents
calculations and test data relating to a structure (e.g., sensor
assembly 500) fabricated in accordance with the various embodiments
presented herein. In an embodiment, a primary change in resistance
across the Wheatstone bridge (e.g., traces 320A, 320B, 330A and
330B) can be due to strain being induced at the upper surface of
the plate 310. The strain can be induced as a function of the plate
310 being deformed, e.g., bent. In an embodiment, the strain of the
upper surface of plate 310 can be maximized at edge region E (e.g.,
as shown in FIG. 3), where the thinning of region 312 from a
thickness P (e.g., thickness of plate 310) to a thickness R, can
concentrate strain at region E.
[0066] Finite element (FEM) studies showing the strain distribution
in plate 310, and traces 320 and 330, are presented in FIGS. 12 and
13.
[0067] Optical inspection of the thinned region 312 formed by the
etching process indicate that the fillet radius (FR) region where
the etched plate meets the remainder of the plate 310 was not
sharp, but rather had a significant fillet radius estimated at 100
to 200 .mu.m. The effect of fillet radius is further address in
FIG. 21.
[0068] It is to be appreciated that the solutions can over-predict
edge stresses. A load P.sub.applied was applied to the sensor in
these calculations rather than the fraction of the load that
actually is felt by the sensor due to load division between the
sensor 310 and the package 910. It is to be noted that the
calculation identifies regions of high stress near the edges of the
plate in FIG. 12, however such a concentration can be mitigated
with a larger fillet. In this solution a distributed load is
applied across a quadrant of a circular sensor plate region 312
surrounded by a Si support structure 310. A nylon washer (soft
material) covered by a steel washer (stiff material) can be modeled
on top of the Si plate (neither the nylon nor steel elements are
shown) to incorporate a deforming element and maintain load element
contact with the plate 310 as it deforms.
[0069] As previously mentioned, the washers (e.g., washers 1010 and
1020) can be considered as a first level of packaging protection
for the sensor element 310. The sensor plate 310 can fail (fracture
at the edge of the plate) at the loads modeled in FIGS. 12 and 13
due to the relatively low strength of Si in bending, further
presented in the design calculations hereinafter. It is to be noted
that forming a sensor plate with SOI could result in a sharper FR
and hence, may be more closely modeled in accord with FIGS. 12 and
13.
[0070] FIG. 14 presents chart 1400 illustrating the effect of
fillet radii on fracture load, e.g., in the corner of thinned
region 312, according to an embodiment. FEM analysis of fracture
load based on maximum stress at membrane radius (plate edge) for
different fillet radii are shown, plot 1410 is the determined
values for a 50 .mu.m radius, plot 1420 is the determined values
for a 100 .mu.m radius, and plot 1430 is the determined values for
a 200 .mu.m radius. A calculated fracture load from the example
design calculations presented herein is shown in line 1440, while
the measured fracture load is shown in line 1450 (note: Si sensor
element fractured at load just beyond that shown in FIG. 20).
[0071] Further, an approximate fracture strength for Si in bending
is shown in 1460. The fracture strength of Si is reached at a low
load (.about.10000 N) for the small radius etch fillet of 50 .mu.m
(which is in approximate agreement with the design calculation
herein) and at a much higher load (.about.30000 N) for the large
radius etch fillet of 200 .mu.m (which is in approximate agreement
with an experimental fracture load, as presented herein). Thus, the
results shown in FIG. 12 support utilization of a large etch fillet
radius (estimated at about 200 .mu.m) to reduce the stress at the
plate edge, increasing the maximum survivable load and decreasing
the sensor signal by an amount consistent with the observed
experimental results.
[0072] To facilitate understanding of the various embodiments
presented herein, the following equations are presented. As an
approximation, the plate 310 can be considered, in aspect, as a
clamped circular plate having a fixed/fixed boundary at the edge E.
An analytical expression for maximum stress (.sigma..sub.R) at the
edge of a clamped circular plate is presented in Eqn. 1:
.sigma. R = 3 .pi. r 2 w 4 .pi. t 2 Eqn . 1 ##EQU00001##
[0073] Further, the associated strain (.epsilon..sub.R) can be
calculated using Young's modulus, according to Eqn. 2:
.epsilon. R = .sigma. R E = 3 r 2 w 4 t 2 E Eqn . 2
##EQU00002##
[0074] where w is the distributed load (N/m.sup.2), E is Young's
modulus (N/m.sup.2), t is the thickness of plate 312, and r is the
plate radius (m). Owing to radial strain having a larger magnitude
than a circumferential strain at the edge of the circular plate
(e.g., edge E of thinned region 312), Eqns. 1 and 2 can be
considered to be directed towards radial strain.
[0075] Returning to Eqns. 1 and 2 and incorporating a
piezoresistive gage factor for highly doped n-type polysilicon
(e.g., G=-18), a change in resistance (.DELTA.R) for a given strain
(.epsilon..sub.R) can be calculated, according to Eqn. 3:
.DELTA.R=GR.epsilon..sub.R Eqn. 3
[0076] Further, an approximate bridge circuit output voltage
(V.sub.Out) as a function of input voltage (V.sub.in) can be
derived for a half bridge design according to Eqn. 4:
V out = 2 V in [ R 1 - R 2 R 1 + R 2 ] Eqn . 4 ##EQU00003##
[0077] where R.sub.1 is the resistance of the outer trace 320
(e.g., traces 320A and 320B) and R.sub.2 is the resistance of the
inner trace 330 (e.g., traces 330A and 330B), according to FIG.
3.
[0078] With reference to FIG. 3, for the sensor design the
resistances of the two outer traces 320A and 320B can be made equal
to each other and the resistances of the two inner traces 330A and
330B can be made equal to each other. Accordingly, the sensor
design can be approximated as Wheatstone halfbridge circuits owing
to the shorter center of the trace 320 resistors, while under
compressive loading, and hence can experience less strain and
contribute less change in resistance than the longer edge of plate
resistance traces 330 may do.
[0079] For a simple circuit design, the input ground 350 and
V.sub.output 340 and/or 370 can be connected to minimize noise,
according to FIG. 3, which can effectively create a voltage divider
circuit. For a voltage divider circuit the change in output sense
voltage for a given change in resistance (in this case due to
strain at the edge of the plate (according to Eqn. 3)) can be
derived according to Eqn. 5:
.DELTA. V out = R 2 ( R 1 + R 2 ) 2 V in .DELTA. R 1 Eqn . 5
##EQU00004##
[0080] As previously mentioned, the materials in package 1110
surrounding the sensor assembly 500 can protect the sensing
element(s) 300 from the most extreme loads and can transmit a
percentage of the load P.sub.applied to the sensing element(s) 300.
All of this protection enables a sensor assembly 500 (and
associated components) to survive high loading of a high
frequency.
[0081] However, the relationship between applied load to a measured
sensor response is required to be known, e.g., for calibration
purposes. Accordingly, knowledge is required to be able to predict
the transfer function of the anticipated load levels (e.g.,
P.sub.applied and/or P.sub.preload) through the materials,
stiffness, and thicknesses chosen in the design of any of sensor
assembly 500, sensing element(s) 300, washer assembly 1110, etc.
Based on such knowledge, various design parameters can be
established and/or adjusted in accordance with the various
anticipated loads to be potentially encountered in a variety of
applications.
[0082] With reference to the package configurations shown in FIGS.
5-11, the various components comprise a plurality of materials of
differing properties. The percentage of load transmitted to the
sensing element(s) 300 can depend on the stiffness of the materials
forming the two parallel load paths through the hardware stack: one
through the sensor assembly 500 and one around the sensor assembly
500, as previously described with reference to FIG. 10. A primary
deforming element is the epoxy washer 530, and the stiffness of
each component in components 500, 300, etc., can depend on the
Young's modulus for each material (epoxy, steel, Si, etc.), the
thickness of each component, and the area over which the load acts
at each component.
[0083] At the interface(s) between the sensor assembly 500, the
washer 1010 and and the support 910, a portion of the load is
carried by the sensor assembly 500 and a portion by the surrounding
support 910. The same deformation occurs in the Si sensing element
300 as the steel in the surrounding support 910, but a greater load
is carried by the steel, primarily because it has much more area
under load. The fraction of load carried by the sensing element 300
and that carried by the surrounding steel structure 910 can be
calculated according to:
P.sub.App=P.sub.SI+P.sub.ST Eqn. 6
[0084] where P.sub.App is the total load applied to the sensing
element 300 and the support 910, P.sub.SI is the portion of the
load passing through the sensing element 300, and P.sub.ST is the
portion of the load passing through the support 910. Ignoring any
difference in height between the sensor assembly 500 (e.g.,
comprising sensing element 300) and the support 910, which would be
eliminated by a preload procedure as previously described, it can
be considered that the die and substrate are at the same height and
subject to the same total applied load. The sensor assembly 500 and
the support 910 can deform the same amount (.delta.), but the
stiffer structure will carry a larger proportion of the applied
load.
F = k .delta. Eqn . 7 k = EA / L Eqn . 8 .delta. = F k = F L E A
Eqn . 9 ##EQU00005##
[0085] In the preceding set of equations, F=P is the applied force
or load, k is spring stiffness, E is Young's modulus, A is area,
and L is length. Therefore, since .delta. is the same for the steel
support 910 and the Si parts of the sensor assembly 500:
.delta. ST = .delta. SI = [ P ST L E ST A ST ] = [ P SI L E SI A SI
] Eqn . 10 ##EQU00006##
[0086] L is the same for the steel and Si and is equal to the
thickness of the sensor assembly 500. Therefore:
P ST = P SI [ E ST A ST E SI A SI ] Eqn . 11 P App = P SI + P SI [
E ST A ST E SI A SI ] Eqn . 12 P SI = P App { E SI A SI E SI A SI +
E ST A ST } Eqn . 13 P ST = P App { E ST A SI E SI A SI + E ST A ST
} Eqn . 14 ##EQU00007##
[0087] The fractional load P.sub.Si applied to the Si, can be used
to calculate a sensor response change in output voltage,
.DELTA.V.sub.out for a given applied load. First, the distributed
load on the sensor assembly 500 is calculated, which is then
utilized to calculate the strain at the edge E of the sensor plate
310:
w = P SI A SI Eqn . 15 .epsilon. R = 3 r 2 w 4 t 2 E SI = 3 r 2 P
SI 4 t 2 E SI A SI = 3 r 2 P App 4 t 2 [ 1 E SI A SI + E ST A ST ]
Eqn . 16 .DELTA. V out = R 2 ( R 1 + R 2 ) 2 V in .DELTA. R 1 = R 2
( R 1 + R 2 ) 2 V in GR 1 .epsilon. R = [ R 1 R 2 ( R 1 + R 2 ) 2 ]
[ V in G ] [ 3 r 2 P App 4 t 2 ] [ 1 E SI A SI + E ST A ST ] Eqn .
17 .DELTA. V out = { [ R 1 R 2 ( R 1 + R 2 ) 2 ] [ V in G ] [ 3 r 2
4 t 2 ] [ 1 E SI A SI + E ST A ST ] } P APP Eqn . 18
##EQU00008##
[0088] Based upon the previous equations and exemplary embodiments,
the following presents an exemplary determination for a maximum
load that can be applied without sensor failure due to excessive
stress at the edge E of the plate 310. The following further
presents an exemplary method for measuring the sensitivity of
sensor assembly 500 in terms of voltage change per unit input
load.
[0089] The failure stress for the plate 310 in bending is
approximately 500 MPa and the Young's modulus is 170 GPa. In an
exemplary embodiment, the plate 310 can be 100 microns thick and 1
mm in diameter. With reference to Eqns. 1 and 2 the distributed
load that can cause the failure stress to be reached at the edge E
of plate 310 can be determined to be:
w=2.7.times.10.sup.7 N/m.sup.2 (27 MPa, 3970 lbf/in.sup.2).
[0090] It is to be appreciated that w is the distributed load on
the plate 310 transmitted through the sensor assembly 500, not the
load on the support 910 which is to be determined. The associated
maximum radial strain at the edge E of the plate 310 can be
calculated from Eqns. 1 and 2, and the maximum change in resistance
from Eqn. 3 for R.sub.1=40 kOhm.
.epsilon..sub.R=0.0029, .DELTA.R.sub.1=2 kOhm
[0091] It is to be noted that either a thicker plate region 312 or
a smaller diameter plate can be utilized to increase the determined
maximum load. However, this could result in a loss in sensor
response signal, hence, a design compromise can be made to
determine the desired combination of high maximum load and high
sensitivity.
[0092] For the exemplary design presented herein (e.g., FIGS.
3-10), assuming four sensing elements 300 included in sensor
assembly 500 of size 2 mm by 1.4 mm:
A.sub.SI=4(2.times.10.sup.-3)(1.4.times.10.sup.-3)=1.1.times.10.sup.-5
m.sup.2
[0093] The steel area of support 910 is:
A.sub.ST=2.5.times.10.sup.-4-1.2.times.10.sup.-5=2.4.times.10.sup.-4
m.sup.2
[0094] for a 19.1 mm (0.75 inch) OD steel support washer 910 with a
6.35 mm (0.25 inch) diameter ID, and four cutouts 920 slightly
larger than 2 by 1.4 mm to accommodate the sensor assemblies
500:
E.sub.SI=170 GPa and E.sub.ST=200 GPa
[0095] Per the Eqns. 13 and 14 above:
P.sub.ST=0.96 P.sub.App and P.sub.SI=0.04 P.sub.App
[0096] For an evenly distributed load over the support washer 910
one fourth of this load is on each sensor assembly 500 (e.g., each
of the four sensing elements 300), 0.01 P.sub.App per die. It was
previously calculated that a distributed load of 2.7.times.107 N/m2
over the die will lead to sensor plate fracture for a 100 .mu.m
thick, 1 mm diameter Si plate 310, accordingly:
P .sub.SI-failure=(2.7.times.10.sup.7)(1.1.times.10.sup.-5)=300N(68
lbf), across all four sensing elements 300.
P.sub.SI-failure,per die=75N (18 lbf), per each sensing element
300.
P.sub.App=300/0.04=7500N (1700 lbf)
[0097] applied to the support washer 910 for sensor die failure. In
terms of the normal stress on the full area of support washer 910
this corresponds to:
.sigma. washer = 7500 2.5 .times. 10 - 4 30 MPa ( 4400 lbf / in 2 )
##EQU00009##
[0098] The sensitivity of the output voltage to this applied load
can be calculated from Eqn. 18:
.DELTA. V out = 0.1 V , for 7500 N applied , 1.33 .times. 10 - 5 V
N , 13.3 mV kN , ( 59 mV / klbf ) ##EQU00010##
[0099] where V.sub.in=9V, G=18, R.sub.1=43.8 kOhm, R.sub.2=33.6
kOhm, r=500 .mu.m, t=100 .mu.m, E.sub.SI=170 GPa, E.sub.ST=200 GPa,
A.sub.SI=1.1.times.10.sup.-5 m.sup.2, A.sub.ST=2.4.times.10.sup.-4
m.sup.2 for the calculations above.
[0100] Sensors, as shown in FIGS. 5-11, were loaded
quasi-statically in compression and sensor response curves as a
function of applied load were generated. The sensor response
calibration from the test data was used in dynamic tests using a
Split-Hopkinson bar test apparatus to determine measured load
during high load rate testing. In addition, maximum load at failure
was determined experimentally.
[0101] FIG. 15 presents chart 1500, representing testing data for
quasi-static compressive loads for a number of sensing elements.
FIG. 15, plots 1510-1550 represent time vs. V.sub.out, for a
plurality of load inputs, 1000 lbf (4445 N) as shown in plot 1510,
2000 lbf (8890 N) as shown in plot 1520, 3000 lbf (13335 N) as
shown in plot 1530, 4000 lbf (17780 N) as shown in plot 1540, and
5000 lbf (22230 N) as shown in plot 1550.
[0102] As shown in respective plots 1510-1550, the load inputs ramp
up linearly to the respective maximum level (e.g., 1000 to 5000
lbf) in 1 to 5 seconds. Further, as shown, after reaching a defined
maximum, the loads return back to zero load in a symmetrical
manner. The measured sensor response(s) was repeatable. At loading
and unloading rates of between 10 to 100,000 lbf/sec (44.45 to
4.445.times.105 N/sec) almost identical response curves result. The
loading rates presented in FIG. 15 correspond to a minimum signal
period of 10 msec and bandwidth of .about.100 Hz, which can be
considered to be equivalent to a quasi-static test condition.
[0103] The presented data was measured for two of the four sensors
(e.g., two of four sensor assemblies 300) located on an epoxy
washer 630, where each upper line in the load pairing is the load
measured on a first sensor assembly 300, and where each lower line
in the load pairing is the load measured on a second sensor
assembly 300. For the arrangement shown in FIG. 6, all four sensor
assemblies 300 showed similar responses to the two presented in
FIG. 15, but with different offset voltages. As shown in FIG. 15,
the resulting signals 1510-1550 have a low noise with a S/N ratio
of 100:1, facilitating measurement resolution of approximately 100
N. As depicted at 1560 (and similarly for all the plots 1510-1550)
there is a no-response section of the curve(s) during which the
slack in the testing equipment is taken up, and followed by a
steeper rise in response at the low load levels as gaps in the
sensor package 1110 are closed and the full washer area is engaged.
Initially at low load levels a larger fraction of the load is borne
by the Si plate 310, however, once the gaps in the sensor package
1110 are compressed out the sensor response is approximately
linear.
[0104] FIG. 16 shows the same data as that presented in FIG. 15 but
plotted as load versus sensor output (V), where curves in group
1670 are for the first sensor 300 and the curves in group 1680 are
for the second sensor 300. As shown, hysteresis is small. If the
initial low load parts of the curves are ignored the response is
linear with a slope of approximately 7 mV/1000 lbf mV (1.57 mV/kN)
for both sensing elements 300. The sensing elements 300 depicted
were identical designs with a plate diameter of 1 mm and a
thickness of approximately 100 .mu.m. Calibration curves from this
data (sensors 1 and 4) are shown in FIGS. 17 and 18 based on a
quadratic curve fit (e.g., 2.sup.nd order least squares polynomial
fit) and were used to measure loads during dynamic testing at much
higher load rates. FIG. 17, plot 1710 presents the data while plot
1720 is the calibration curve (quadratic curve fit). FIG. 18, plot
1810 presents the data while plot 1820 is the calibration curve
(quadratic curve fit).
[0105] The test data from the dynamic tests was generated using a
Split-Hopkinson bar apparatus. Briefly, the Split-Hopkinson bar
utilizes a gas-gun to drive a projectile into a long steel rod and
impart a shaped shock pulse along the rod. A pulse shaper
(typically a copper disk) at the impact point of the projectile on
the incident bar is utilized to shape the pulse, imparting, for
example, a square wave or triangle wave pulse. The test sample
(e.g., a sensor assembly 600 containing the sensing elements 300)
is placed at the end of the incident bar opposite the projectile
impact. Two configurations were tested, a direct load configuration
and a preloaded protected sensor configuration in which the sensor
assembly 600 is inside the protected housing of sensor washer
1110.
[0106] The calibration curves from the quasi-static testing were
utilized to convert V.sub.out recorded (e.g., on a high speed
digital oscilloscope) into force for the data shown in FIG. 19 (a
square wave pulse input) and FIG. 20 (triangle wave pulse input).
FIG. 19, plot 1910 is the data measured for the first sensor 300,
plot 1920 is the data measured for the second sensor 300, and plot
1930 is the data measured from a quartz-based calibration load
cell. FIG. 20, plot 2010 is the data measured for the first sensor
300, plot 2020 is the data measured for the second sensor 300, plot
2030 is the data measured from a quartz-based calibration load
cell. As shown in FIGS. 19 and 20, the first and second sensors 300
matched the calibration data from the load cell, plots 1930 and
2030.
[0107] A degree of noise is present for the sensor element 300
signals as well as the quartz load cell. The noise can be due to
both electronic and mechanical sources. Ringing and overshoot at
the front of the measured pulse are evident for both the quartz
gage and the sensor elements 300. Overshoot is more pronounced for
the sensor elements 300. An approach to clean up the sensor
elements 300 signals was to apply a low level preload
(approximately 2 kN) to the sensor washer 1110. Such an approach
can remove any slack internal to the sensor washer 1110 and
further, move the sensor elements 300 signals from the quadratic to
the linear region on the calibration curve, as well as reduce
mechanical noise sources.
[0108] In some of the samples tested, the sensor washer 1110 was
still in a state of compression after preload. FIG. 21 presents a
chart 2100 for a triangle shaped pulse with the preload. FIG. 21,
plot 2110 is the data measured for the first sensor element 300,
plot 2120 is the data measured for the second sensor element 300,
and plot 2130 is the data measured from the quartz-based
calibration load cell.
[0109] The noise on signals from the sensor elements 300 was
significantly reduced and the overshoot at the front of the pulse
measurement was eliminated. The square wave pulse (as shown in FIG.
20) was similarly cleaned up by preloading. Further, electronic
noise sources were not cleaned up, specifically 60 Hz noise, as is
present in FIG. 19 or 20. If electronic noise sources were reduced
(e.g. with a 60 Hz filter) a measurement resolution similar to that
shown for the quasi-static data (e.g., per FIG. 16) can be
achievable. A peak load of 35,000 N (7900 lbf) was measured, as
shown in FIG. 21, which corresponds to a stress at the sensor
washer of about 140 MPa (20 ksi). Further, elimination of the
squared term in the calibration curve (not shown) brought the
preloaded MEMS sensor data to closer agreement with the quartz
calibration sensor. A signal rise time of about 10 .mu.sec was
tracked by the sensors (FIGS. 18 and 19) corresponding to a maximum
demonstrated frequency of about 100,000 Hz.
[0110] The test results indicate the failure load was about 5 times
higher than expected (>35000 N measured vs 7500 N predicted
(P.sub.APP)) and the sensor response was 8 times lower than
expected (1.57 mV/kN measured vs 13.3 mV/kN predicted). The
predicted load failure was based on an oversimplification of the
physical situation and hence was purposely conservative. It was
anticipated that the sensor response curve would approximately
track the load since the strain and output voltage track the stress
at the edge E of the plate 310. A key oversimplification for the
previously presented design calculation was the application of a
fixed/fixed boundary condition. In the sensor package the thick
steel in die mount 910 in contact with a sensing element 500 would
not allow much deformation and would tend to produce a fixed
boundary. It is to be noted that in the data shown, e.g., FIGS.
16-21, a thin (0.001 inch, 0.0254 mm) layer of polyimide tape (a
relatively soft material) was placed in between the PCB 530 and the
die mount 910 to provide some measure of wear protection. This soft
material would tend to make the boundary more compliant (less
fixed). Also the PCB 530 at the edge of the sensor plate contacting
the thicker section of the sensing element 300 is a softer material
that would allow some deformation. Hence, the boundary condition at
the edge of the sensor plate may be improved by modeling as
somewhere between a fixed boundary and a simply supported boundary
in which the slope at the edge of the plate is not assumed zero.
For a simply supported circular plate the edge stress is zero, so
any flexibility at the edge E of the plate 310 would tend to reduce
the stress, increase the maximum achievable load, and reduce the
signal output. For a simply supported plate it would be better to
have the piezoresistive strain sensing elements at the center of
the plate rather than the edges.
[0111] The maximum survivable load may be under-predicted and the
sensor response may be overpredicted may be due to the effect of
the etch fillet radius FR at the edge E of the sense plate 310 is
not modeled in the design calculation previously described. As
shown in FIG. 13 a sharp corner at the edge FR of the sense plate
310 causes a stress concentration resulting in plate stresses
beyond the fracture limit for Si in bending. As previously
mentioned, a radius at this location will reduce this stress
concentration leading to a higher load required for fracture, less
strain, and therefore less sensor response voltage.
[0112] As previously mentioned, a micro-preload can be applied to
the sensor assembly 1110, for example to compensate for the various
components comprising assembly 1110 are not tightly packed when
assembly 1110 is first assembled. Specifically parts of the load
path may be in contact while other parts are not. An amount of
preload can be required to compress the initially contacted parts
of the washer assembly 1110 enough to bring the full washer area
into contact in order to achieve full load division. In particular
the die stickup (FIG. 5, distance M) may be such that the assembly
510 is in contact with the die mount 910 while the surrounding PCB
530 is not. The sensor plate 310 can initially carry the entire
load as the sensor washer 1010 is compressed enough until the PCB
530 comes into contact with the die mount 910. A function of the
sensor plate 310 initially carrying the entire applied load can
account for the initially higher slope of the load response curves
(quadratic term in the calibration equations). In an embodiment, a
small amount of preload can be applied to cause the entire areas of
the respective components 910, 530 and 310 to contact and the
sensor response to enter the linear range. Owing to the ability to
monitor sensor response voltage during the preload process (per
FIGS. 17 and 18) it is possible to determine how much preload is
required to compress the various components comprising sensor
assembly 500 and thus determine when enough micro-preload has been
applied for complete contact.
[0113] As previously described (e.g., per Eqn. 1) the sensor
survival load is most sensitive to a thickness of thinned region
312 versus plate thickness 310, and further a radius J of the
thinned region 312, per FIG. 3. Reducing the plate radius J or
increasing the thickness of thinned region 312 can result in
significantly higher maximum achievable loads (squared
functionality for both parameters). However, such an approach can
lead to a loss of sensor response (per. Eqn 18). In addition,
reducing the percentage of load carried by the sensor plate 310 and
thinned region 312 by decreasing the size of the die mount 910 and
increasing the stiffness of the surrounding materials (e.g.
Tungsten, E=400 GPa, instead of Steel, E=200 GPa) can increase the
maximum achievable load before sensor plate failure, but with a
linear functionality (per Eqns. 1, 13 and 14).
[0114] An approach to increase sensor response for a given strain
is to increase the number of strain-sensitive `wires` (e.g., traces
320 and 330) on the surface of the thinned sensor region 312 and
position the wires more precisely at the maximum strain locations
on the plate 310. For instance, for a calculation to accurately
determine the plate deformation in terms of fixed vs. simply
supported boundary conditions, an approach may be to place the
highest density of `wires` more at the center of the thinned region
312 than at the edge E. The number of wires can be more densely
patterned than that shown in FIG. 3 owing to the traces being
photolithographically patterned much thinner and closer together in
order to increase R.sub.1 and R.sub.2 (e.g., as defined in Eqn. 18)
and thereby increase sensor output voltage for a given load. Hence,
in an embodiment, to achieve a more sensitive, higher maximum load
sensor can involve increasing the radius J to provide more area for
piezoresistive `wires` (a denser pattern of traces at the optimum
radial location micromachined at the limit of photolithographic
precision), while at the same time increasing the thickness of the
thinned region 312 such that the plate 310 can handle a higher
load.
[0115] FIG. 22 presents a sensor apparatus 2200, in accord with an
embodiment. As shown, the load sensor 120 can be placed where the
washer 150 is located in FIGS. 1 and 2. By locating the load sensor
120 between the bolthead 131 and the support structure 110, a
compressive load applied to the outer surface 134 of bolthead 131
can be measured directly at sensor 120. It is to be appreciated
that while the various embodiments presented herein are directed
towards an arrangement whereby a load sensor 120 is located on a
boltshaft 133 of a bolt 130, the various embodiments are not so
limited and can be directed towards any configuration which can
utilize at least one Si sensing element and associated structure(s)
as presented herein. For example, the sensing elements 300 can be
incorporated into a system whereby the sensing element(s) 300 are
positioned between two plate structures which maintain a
compressive load on the sensing element(s) 300 during
operation.
[0116] FIGS. 23 and 24 illustrate exemplary methodologies relating
to a sensor which can measure and survive shock loading while still
providing a high signal/noise ratio with significant bandwidth are
presented. While the methodologies are shown and described as being
a series of acts that are performed in a sequence, it is to be
understood and appreciated that the methodologies are not limited
by the order of the sequence. For example, some acts can occur in a
different order than what is described herein. In addition, an act
can occur concurrently with another act. Further, in some
instances, not all acts may be required to implement the
methodologies described herein.
[0117] FIG. 23 illustrates a methodology 2300 relating to a sensor
which is subjected to a compressive preload and can measure and
survive shock loading while still providing a high signal/noise
ratio with significant bandwidth are presented, according to an
embodiment. As previously mentioned, a preload can be applied to
the sensor, whereby during operation of the sensor, the magnitude
of the preload can be reduced as a function of the applied
operational load. With knowledge of the preload amount, and the
final preload during sensor operation, the magnitude of the applied
operational load can be determined.
[0118] At 2310, a load sensor can be formed and include at least
one piezoresistive sensing element. As previously described, the
sensing element can be formed from Si. A concern with Si elements
is their inherent brittleness in the presence of a tensile stress,
a concern which can be addressed by application of a compressive
preload on the Si sensing element. Based, at least in part, upon
the application, the sensing element can be placed in a containing
structure which can protect the sensing element(s) as well as also
containing other material components which can minimize impact
stresses that could cause fracture of the sensing element. In an
embodiment, the load sensor can be utilized in a configuration
where the load sensor is supported on a boltshaft, where the load
sensor is in the form of a washer. To facilitate determination of
application of loading in a number of directions (e.g., off-center
loading) relative to the load sensor, the load sensor can be formed
with a plurality of sensing elements, e.g., four sensing elements
to facilitate load determination in four directions (e.g.,
effectively cardinal directions) relative to the position of the
load sensor on the boltshaft. The containing structure can be
formed from steel. Further, the material components can be placed
either side of the sensing element(s) to provide a degree of
compliance to the load sensor. The materials can be formed from any
ductile and/or compliant material, as previously described
herein.
[0119] At 2320, a micro-preload can be applied to the load sensor.
The micro-preload can act to compensate for the various components
in the load sensor not being tightly packed during assembly of the
load sensor, e.g., parts of a load path may be in contact while
other parts are not. As previously described, an amount of preload
can be applied to compress initially contacted parts of the load
sensor to achieve full load division between the plurality of
sensing elements, and hence facilitate a linear sensing
response.
[0120] At 2330, as previously described particularly with reference
to FIGS. 1 and 2, the load sensor can be located on a support
structure which can be utilized to locate the load sensor relative
to a component to be sensed, and also to apply a macro-preload to
the load sensor. A support structure can be a nut and bolt, onto
which the load sensor is located, whereby the bolt is located in a
recessed portion of the component to be sensed, e.g., a countersunk
hole in a wall of the component to be sensed. A plurality of
spacing components, e.g., washers, can be utilized to facilitate
load absorption during operation of the load sensor. The washers
can be formed from ductile and/or compliant material.
[0121] At 2340, a macro-preload can be applied to the sensor,
where, in an embodiment, the macro-preload can be of a magnitude
greater than an anticipated maximum load to be experienced during
operation of the load sensor. The macro-preload can be several
times greater than the micro-preload. For example, with the
exemplary embodiment, a compressive macro-load can be applied to
the load sensor, e.g., by tightening of the nut on the bolt, where
the loading results in tensile stress being applied to the
boltshaft, the tensile stress can be equal but of opposite
magnitude to the compressive macro-load.
[0122] At 2350, during operation a load can be applied to the
bolthead which can act to force the boltshaft in an inward
direction relative to the component wall. Accordingly, owing to the
boltshaft moving in a direction opposite to the loading direction
of the nut and bolt, the compressive load resulting from the
compressive macro-preload is reduced.
[0123] At 2360, the amount by which the compressive preload reduced
can be determined, thereby enabling determination of the magnitude
of the load applied to the bolthead.
[0124] FIG. 24 illustrates a methodology 2400 relating to
fabricating a sensor with a thinned region to facilitate
concentrating a load at the thinned region. At 2410, a plate can be
formed to act as a sensing plate. In an embodiment, the plate can
be formed from Si. In another embodiment, the plate can have a
thickness of about 100 .mu.m.
[0125] At 2420, a portion of the sensing plate can be thinned. The
thinned region can be of any suitable thickness, e.g., about 10
.mu.m thick, to facilitate a higher degree of strain occurring in
the thinned region during loading than occurs in the un-thinned
portion of the plate. The thinned region can be formed by any
suitable operation, such as an etching operation. Further the
thinned region can have a circular configuration.
[0126] At 2340, at least one trace element can be formed over the
thinned region. The trace(s) can form a Wheatstone bridge, or
similar circuit system, where, in an embodiment, the trace(s) can
be of a circular serpentine pattern configured to operate over the
thinned region. Appropriate connections, wiring, etc., can be
attached to the trace(s). The sensing plate with the traces formed
thereon can be located in a load sensor, as previously
described.
[0127] At 2440, the load sensor can be incorporated into a load
sensing system. During application of a load to the load sensor, a
strain can be created at the thinned region owing, in part, to the
difference in thickness between the thinned region and the rest of
the thicker plate. The strain at the thinned region can further
induce strain in the trace(s) which can accordingly affect the
voltage, etc., across the Wheatstone bridge. Based upon the change
in voltage on the Wheatstone bridge the reduction in compressive
load on the plate can be determined and accordingly, the degree of
applied load.
[0128] What has been described above includes examples of one or
more embodiments. It is, of course, not possible to describe every
conceivable modification and alteration of the above structures or
methodologies for purposes of describing the aforementioned
aspects, but one of ordinary skill in the art can recognize that
many further modifications and permutations of various aspects are
possible. Accordingly, the described aspects are intended to
embrace all such alterations, modifications, and variations that
fall within the spirit and scope of the appended claims.
Furthermore, to the extent that the term "includes" is used in
either the details description or the claims, such term is intended
to be inclusive in a manner similar to the term "comprising" as
"comprising" is interpreted when employed as a transitional word in
a claim.
* * * * *