sensors
Article
Precision Low-Cost Compact Micro-Displacement Sensors
Based on a New Arrangement of Cascaded Levers with
Flexural Joints
Che-Chih Tsao *, Yi-Chun Tseng, Yu-Sheng Chen, Wei-Hsuan Chang and Li-Ting Huo
Citation: Tsao, C.-C.; Tseng, Y.-C.;
Chen, Y.-S.; Chang, W.-H.; Huo, L.-T.
Precision Low-Cost Compact
Micro-Displacement Sensors Based
on a New Arrangement of Cascaded
Levers with Flexural Joints. Sensors
2023, 23, 326. https://doi.org/
10.3390/s23010326
Academic Editor: Grigoris Kaltsas
Received: 20 October 2022
Revised: 28 November 2022
Accepted: 6 December 2022
Published: 28 December 2022
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Sensors 2023, 23, 326. https://doi.org/10.3390/s23010326
Department of Power Mechanical Engineering, National Tsing Hua University, Hsinchu 30013, Taiwan
* Correspondence: cctsao@pme.nthu.edu.tw
Abstract: Existing displacement sensors of micrometers to sub-micron precision are expensive and
have various limitations. This paper reports the design and development of a new contact type
compact micro-displacement sensor of sub-micron precision for a fraction of the cost of commercial
devices. The basic concept of the new sensor system applies a mechanical magnifying mechanism to
magnify a displacement at sub-micron to micron level and uses a low-cost Hall sensor to measure
the magnified displacement. Various conceptual designs for the mechanical magnifying mechanism
based on cascaded levers with flexural joints were studied and a final design, featuring side-by-side
placement of lever structures in a multi-planar layout with adjacent levers coupled by L-shaped
coupling foils, was devised. Prototypes of two different sizes and constructions with mechanical
magnification ratios over 100 were made and tested. Measurement repeatability and accuracy to
sub-micrometer level and a resolution down to hundredths of a micrometer were demonstrated
by a compact Alpha Model prototype. Design modification of parts and a corresponding small
lot size production procedure were devised to provide an estimated bill of material cost per unit
under US$100.
Keywords: displacement sensor; flexure; micro-displacement; Hall sensor
1. Introduction
1.1. Problems and Motivations
Precision displacement sensors or gauges are useful in measuring and checking di-
mensions of manufactured parts, positions of machine stages, assembly precisions of
components in a machine system and in other areas. However, for in-process, in-machine
precision measurements at micron to submicron level, there is currently no low-cost and
compact choice. Traditional mechanical displacement gauges, such as dial indicators or
meters based on gears and spiral spring, have a precision of about 2 µm (micron, or microm-
eter) and are not suitable for sub-micron level measurements. Highly precise displacement
measurement systems, such as interferometers, can easily make sub-micron measurements,
but they are extremely expensive and bulky and generally not suitable for in-process or
in-machine applications. For in-process, in-machine precision measurements, several com-
mercial devices are available, as shown in Table 1. The list includes sensors of non-contact
type and contact type. Capacitive sensor, eddy current sensor, laser triangulation and color
confocal sensor are non-contact type. LVDT (including its probe part) and digital contact
displacement sensor are contact type. For micrometer level measurement precision, capaci-
tive sensor, eddy current sensor, digital contact sensor and color confocal sensor can reach
one micron accuracy and sub-micron resolution. However, they are still expensive. Table 1
shows prices of single sensor heads in USD, not including their companion electronic boxes.
The Hall sensor is very cheap, but its precision is over several micrometers.
https://www.mdpi.com/journal/sensors
Sensors 2023, 23, 326 2 of 28

Table 1. Comparison of specifications and prices of typical commercially available displacement sensors.

Range Accuracy Resolution Dealer Price in USD a

Category
(mm) (µm) (µm) (Brand/Manufacturer)
Capacitive sensor 0.05 0.15 0.001 $2000 (µε) b
Eddy current sensor 1 1 0.1 $2100 (Kaman) c
Inductive sensor (LVDT) 1 6 –i $650 (µε) d
Laser triangulation 10 12 1 $1300 (µε) e
Digital contact
12 1 0.1 $1600 (Keyence) f
displacement sensor
Color confocal laser
7 0.94 0.25 >$8000 (Keyence) g
displacement sensor
0.1~1% of
Hall sensor 0.25–2.5 5 $5 h
full scale
a From quotations by dealers. b Micro Epsilon product specifications [1]. c Kaman product specifications [2]. d Mi-
cro Epsilon product specifications [3]. e Micro Epsilon product specifications [4]. f Typical model GT2-PK12A [5].
g Typical model CL-L030 [6]. h General specification, see Chap. 3 of [7]. i Depending on amplifying electronics.

In addition to costs, commercial devices also have various limitations. While non-
contact measurement is appealing, capacitive sensors have a very small measurement
range. Eddy current sensors can only be used on good conductors [7]. (Chap. 3) Sensors
based on laser triangulation have good resolution but limited linearity [4]. Additionally, the
reflective property of a surface always affects the sensing of any optical sensor. Contact type
sensors can be applied, in general, to any rigid surface. Existing contact type displacement
probes, such as the Keyence example listed in Table 1 and many other commercial models,
generally use an axially moving long rod as the probe, together with an opto-electronic
sensing unit to “read out” the displacement of the rod. This design requires size and carries
intrinsic cost.
Accordingly, if a displacement sensor can be made to have a compact size, to have a
precision comparable to existing precision sensors, and to cost only a fraction of or even
an order of magnitude less than those commercial sensors, then it can be applied in many
situations in large quantities in the field. For example, a machine tool or a positioning stage
can have several sensors installed at strategic locations and combined measurement results
can be used to track machine deformations or position errors in multiple directions due
to thermal or load effects in real time. Such information can lead to further improvement
of process accuracy. These small devices can also be applied to factory machines for
maintenance monitoring and in production lines for process monitoring, all at costs lower
than existing choices.
This paper reports the design and development of a new contact type precision
micro-displacement sensor for measuring a micro-displacement of a target surface with
a measurement range from a few micrometers to hundreds of micrometers and a mea-
surement resolution on the order of 0.01 to 1 micrometer and a measurement accuracy
on sub-micron to micrometer level. The basic concept is to apply a mechanical mecha-
nism to magnify a sub-micron displacement to be measured by about 100 times so that
the magnified displacement can be measured by using an existing low-cost displacement
sensor, such as a Hall sensor. That is, with a reliable mechanical magnifying mechanism of
consistent and repeatable magnification ratio, a low-cost displacement sensor can measure
a micro-displacement to about 1/100th of its own resolution and accuracy.

1.2. Technical Background in Displacement Ampilification Mechanisms

Displacement amplification mechanisms can be found in industrial applications and
in academic research. In the broadest sense, displacement amplification or magnification
mechanism can be separated into two categories: Integral structures and non-integral
structures. By integral structures we mean structures fabricated from a monolithic piece
of solid material or an assembly of parts joined into an integral solid structure. Between
any parts of an integral structure, there is no mechanical contact with relative motion. On
Sensors 2023, 23, 326 3 of 28

the other hand, non-integral structures include discrete parts that are in contact, but not
joined into one solid, and there can be relative motion between parts in mechanical contact.
Kelemenová [8] summarized amplifying mechanisms of various types, including both
integral and non-integral structures.
Flexure-based mechanisms belong to the category of integral structures. In general,
because there is no mechanical contact with relative motion and thereby no need of pro-
cesses of hardening or polishing of contact surfaces, flexure-based mechanisms should have
better precision and repeatability and potentially a lower cost compared to mechanisms of
non-integral structure. However, most implementations seen in industry or in academic
publications are case by case and have very application-specific designs, not suitable for
development into low-cost off-the-shelf products. Some examples are briefly reviewed
below. Chen et al. [9] provided an extensive review of flexure-based displacement am-
plification mechanisms. One frequently applied flexure-based mechanism is the bridge
structure, such as examples shown in Juuti et al. [10]. Bridge structures can be modified
and stacked into a honeycomb link, as described in Muraoka et al. [11]. Another frequently
applied mechanism is the lever structure, from single to multiple stages. For example,
Henmi et al. [12] describes a two-stage lever amplifier used in an actuator. Most exam-
ples described in the literature referenced above are for applications to amplify motions
of actuators, rather than for displacement measurement, although the principle of the
mechanisms can be applied in both ways. In Yu et al. [13], a two-stage lever structure is
applied to reduce a displacement from an actuator to obtain micro-scale displacement for
thin film measurement applications. Yamauchi [14] described a device for measuring rock
deformation in a drilling tube applying a single stage lever mechanism. Bae et al. [15] de-
scribed a piezoelectric actuator driven micro-stage with a build-in magnifying mechanism
for displacement measurement. The magnifying mechanism was a two-stage lever system
with monolithic hinges all made in a piece of metal slab and had a magnification ratio of 30.
Such planar integral structures are also seen in MEMS structures, such as a micro-actuator
structure reported by Nada et al. [16].

2. Basic Concept and Studies of Conceptual Designs

2.1. Basic Concept of the Low-Cost Precision Sensor
The basic concept of the contact type precision micro-displacement sensor of this
research is summarized in Figure 1, using a magnification ratio of 100 as an example. From
Table 1, we see that the Hall sensor has a significantly low cost but a measurement resolution
of about 5 µm. It has an accuracy of about 0.1–1% of full scale and a range of 0.25–2.5 mm.
That is, for a measurement range on the order of 1000 µm and a resolution/accuracy of
1–5 µm, the Hall sensor is a very cost effective choice. Now, for measuring a displacement
down to 0.1 µm, if we can magnify this tiny displacement by a factor of, say, 100, then we
will be measuring 10 µm, which falls well within the Hall sensor’s capability. That is, if a
magnification mechanism can be devised to reliably magnify sub-micron displacements
with good repeatability, then a Hall sensor can be used to resolve a micro-displacement of
0.1 µm with a full range of 10 µm, while the Hall sensor itself is measuring displacements
in a full range of 1 mm with a resolution of 10 µm.
Further, if such a mechanical micro-displacement magnifying mechanism (MDMM)
can be made small with low cost, then a precision Low-cost Compact Micro-Displacement
Sensor (LCMDS) can be constructed by combining it with a Hall sensor, which is also cheap
and of small size.
By applying different mechanical magnification ratios, the micro-displacement sen-
sor can have different measurement ranges, as shown in Table 2. For a specific sensor
measurement range, the corresponding best measurement accuracy and resolution can be
estimated by shrinking the general resolution and accuracy of a typical Hall sensor by the
magnification ratio.
Sensors 2023, 23,
Sensors 2023, 23, 326
x FOR PEER REVIEW 44 of
of 28

Figure
Figure 1. Basicconcept
1. Basic conceptofofthe
the Low-cost
Low-cost Compact
Compact Micro-Displacement
Micro-Displacement Sensor
Sensor (LCMDS)
(LCMDS) using
using a
a me-
mechanical magnification
chanical magnification ratio
ratio of 100
of 100 as an
as an example.
example.

2. Relation
TableFurther, if between mechanical magnification
such a mechanical ratio and magnifying
micro-displacement measurementmechanism
range, with correspond-
(MDMM)
ing accuracy and resolution, based on a Hall sensor of a measurement resolution of
can be made small with low cost, then a precision Low-cost Compact Micro-Displacement 5 µm and an
accuracy of 0.1–1% of the full measurement range.
Sensor (LCMDS) can be constructed by combining it with a Hall sensor, which is also
cheap and of small size.
Measurement Magnified
By Magnification
applying different mechanicalMeasurement
magnification ratios, theAccuracy Resolution
micro-displacement sen-
Range Range
sor can(µm) Ratio
have different measurement as shown in Table(µm)
ranges,(mm) (µm)
2. For a specific sensor
measurement
10 range, the corresponding
250 best measurement
2.5 accuracy and resolution
0.01–0.1 can be
0.002–0.02
estimated
10 by shrinking the
100general resolution and
1 accuracy of a typical
0.01–0.1 Hall sensor by the
0.005–0.05
magnification
25 ratio. 100 2.5 0.025–0.25 0.005–0.05
50 50 2.5 0.05–0.5 0.01–0.1
Table 2.100 25
Relation between mechanical 2.5 ratio and measurement
magnification 0.1–1 range, 0.02–0.2
with corre-
250
sponding 10
accuracy and resolution, 2.5
based on a Hall sensor 0.25–2.5
of a measurement resolution 0.05–0.5
of 5 μm and
500 of 0.1–1% of the full
an accuracy 5 measurement range. 2.5 0.5–5 0.1–1

Measurement Magnified
As pointed outMagnification
already, because there is no mechanical contact Accuracy Resolution
with relative motion,
Range Measurement Range
flexure-based mechanisms Ratio should have better repeatability and (μm)lower cost (μm) compared
(μm) (mm)
to mechanisms of non-integral structure. There are two frequently seen flexure-based
10 250
mechanisms: Bridge structure and lever structure. 2.5 The latter is0.01–0.1
comparatively 0.002–0.02
simple and
easy to construct. Accordingly, lever structure was selected for the design and0.005–0.05
10 100 1 0.01–0.1 construction
25
of the mechanical 100
displacement 2.5
magnifying mechanism 0.025–0.25
of this research. To0.005–0.05
obtain large
50 50 2.5
magnification ratios, larger than 25 or so, multiple lever structures need to be0.05–0.5 0.01–0.1
used and
connected 100in cascaded stages 25 into an integral structure
2.5 of levers. 0.1–1 0.02–0.2
A magnified displacement
in a lever250structure in one10stage drives another 2.5lever structure in the next stage
0.25–2.5 and the
0.05–0.5
original500displacement is magnified
5 further. 2.5 0.5–5 0.1–1
By geometry, the cascaded lever structures can take two kinds of layout arrangements.
The lever structures
As pointed can be disposed
out already, because with
thereone lever
is no structurecontact
mechanical on topwithof another
relativeand with
motion,
the planes of motion of the arms all on the same plane. This can
flexure-based mechanisms should have better repeatability and lower cost compared to be called the co-planar
layout. The lever
mechanisms structures can
of non-integral also beThere
structure. disposed side frequently
are two by side with all flexure-based
seen arms in parallel and
mech-
with the Bridge
anisms: plane ofstructure
motion of andeach armstructure.
lever different The
but parallel
latter is to each other. This
comparatively can be
simple andcalled
easy
the multi-planar
to construct. layout. lever structure was selected for the design and construction of
Accordingly,
Flexural joints are used asmagnifying
the mechanical displacement hinges of themechanism
levers and of forthis
coupling between
research. lever struc-
To obtain large
tures of successive stages. There are two categories of flexural joints,
magnification ratios, larger than 25 or so, multiple lever structures need to be used and namely monolithic
structures,
connected which can be stages
in cascaded made as integral
into parts ofstructure
an integral a monolithic mechanism,
of levers. and flat-springs,
A magnified displace-
which need to be assembled to a mechanism [12] (Section 8.6).
ment in a lever structure in one stage drives another lever structure in the next stage and
Most existing
the original flexure-based
displacement displacement
is magnified further.amplifying mechanisms, as referenced in the
literature review above, use co-planar layout,
By geometry, the cascaded lever structures can takeespecially whentwo the kinds
mechanism is made
of layout from
arrange-
aments.
monolithic
The lever structures can be disposed with one lever structure on top of another due
piece of solid metal with build-in monolithic flexural joints. This could be and
to ease of manufacturing because flexural joints in integral forms can be made by drilling
with the planes of motion of the arms all on the same plane. This can be called the co-
holes on a metal slab followed by making lever arms by wire EDM (Electric Discharge
planar layout. The lever structures can also be disposed side by side with all arms in par-
Machining) of the slab.
allel and with the plane of motion of each arm different but parallel to each other. This
can be called the multi-planar layout.
 Most existing flexure-based displacement amplifying mechanisms, as referenced in
the literature review above, use co-planar layout, especially when the mechanism is made
from a monolithic piece of solid metal with build-in monolithic flexural joints. This could
be due to ease of manufacturing because flexural joints in integral forms can be made by
drilling holes on a metal slab followed by making lever arms by wire EDM (Electric Dis-
Sensors 2023, 23, 326 5 of 28
charge Machining) of the slab.
In this research, MDMM conceptual designs of both co-planar layout and multi-pla-
nar layout using monolithic and flat-spring flexural joints were explored in order to find
In this
a reliable andresearch,
compactMDMM
design conceptual
of cascadeddesigns of both co-planar
lever structures layout
with flexural andofmulti-
joints a magni-
planar layout using monolithic
fication ratio of about 100. and flat-spring flexural joints were explored in order to
find a reliable and compact design of cascaded lever structures with flexural joints of a
magnification ratio of about 100.
2.2. Conceptual Design A: Monolithic Structure of Co-Planar Layout with Monolithic Flexural
Joints—Exploratory Study
2.2. Conceptual Design A: Monolithic Structure of Co-Planar Layout with Monolithic Flexural
Joints—Exploratory
A monolithicStudy three-stage cascaded lever structures was first designed, as shown in
FigureA monolithic
2, and studied.three-stage cascaded
Each lever leverwas
structure structures
designedwastofirst
havedesigned, as shownratio
a magnification in of
Figure 2, and studied. Each lever structure was designed to have a magnification
five, with the goal of a total magnification ratio over 100. The input contact place is locatedratio of
five,
at A with
in thethefirst
goalstage
of a total
levermagnification ratio over
L1 and the output 100. Theat
is located input
B oncontact place
the third is located
stage lever L3.
at A in the first stage lever L1 and the output is located at B on the third stage
The monolithic flexural hinge structures (LH1, LH2 and LH3) were in the form of circular lever L3.
The monolithic flexural hinge structures (LH1, LH2 and LH3) were in the form
notch and were designed to avoid fatigue, assuming the application of aluminum, by us- of circular
notch and were designed to avoid fatigue, assuming the application of aluminum, by using
ing the stress formula from Merken [17] with a rough estimate of bending angles. The
the stress formula from Merken [17] with a rough estimate of bending angles. The output
output end of lever L1 is coupled to the input place of lever L2 by a coupling joint CF12,
end of lever L1 is coupled to the input place of lever L2 by a coupling joint CF12, which is
which is a smaller monolithic flexural hinge. The coupling between stage 2 and stage 3 at
a smaller monolithic flexural hinge. The coupling between stage 2 and stage 3 at CF23 is
CF23 is the same.
the same.

2. Conceptual
Figure 2.
Figure Conceptualdesign A,A,
design lever armarm
lever height 7 mm,
height widthwidth
7 mm, 5 mm,5flexural hinge radius
mm, flexural hinge3 radius
mm, flex-
3 mm,
ural section thickness 1 mm, coupling joint radius 1 mm, flexural section thickness 1 mm; mechanical
flexural section thickness 1 mm, coupling joint radius 1 mm, flexural section thickness 1 mm; me-
advantage
chanical of each lever
advantage 5; MFlever
of each indicating
5; MF holes for mounting
indicating holes foronto a back plate.
mounting onto a back plate.

Deformation
Deformationanalysisanalysisusing
usingAnsys
Ansyssoftware
software was
was applied
appliedto the structure
to the to observe
structure to observe
the displacement of the levers and deformation of the joints and the magnification ratio
the displacement of the levers and deformation of the joints and the magnification ratio
was found to be very small. Figure 3a is a total deformation graph of the structure when
was found to be very small. Figure 3a is a total deformation graph of the structure when
a force of 1 N is applied downward at contact place A, showing that the displacement at
aoutput
force Bofis1only
N isslightly
appliedlarger
downward at contact place A, showing that the displacement at
than that of input A.
output TheBmain
is only slightly
issue largertothan
was found be atthat
the of input A.
coupling joints, CF12 and CF23. As shown in a
close up view of joint CF23 in Figure 3b, shearcoupling
The main issue was found to be at the strains arejoints, CF12This
obvious. andisCF23. As when
because, shown in
atwo
close up view
adjacent lever ofarms
jointrotate
CF23 about
in Figure
their3b, sheareven
hinges, strains are obvious.
at small This istranslational
angles, relative because, when
two adjacent also
displacement lever armsdue
occurs rotate about
to the their hinges,
rotations, even at relative
thereby causing small angles, relative
shearing. transla-
The small
tional
couplingdisplacement
joints under also occurs
shearing due the
prevent to the rotations,
lever arms from thereby
furthercausing
rotationsrelative shearing.
and therefore
The
limitsmall coupling joints
the magnification under shearing prevent the lever arms from further rotations
effect.
To resolvelimit
and therefore the issue, additional flexure
the magnification effect.must be built into the coupling joints to ac-
commodate both relative rotational movements and relative translational movements. To
achieve this in a monolithic co-planar layout, design of the coupling joints could become
complicated and the overall size of the structure could increase. Another issue of the
co-planar layout is that the height of the structure, measured along the direction of dis-
placement at position A, can become too large and no longer “compact”. For these issues,
the design approach was switched to multi-planar layouts.
 Sensors2023,
Sensors 23,x326
2023,23, FOR PEER REVIEW 66ofof2828

(a)

(b)
Figure
Figure 3. (a) is a total
total deformation
deformationgraph
graphofofthe
thestructure
structure when
when a force
a force of 1ofN1isNapplied
is applied downward
downward at
at contact
contact place
place A;A;
(b)(b)
is is a close-up
a close-up view
view ofof joint
joint CF23
CF23 andexaggerated
and exaggeratedmotion
motionanalysis
analysisbetween
betweenL2L2
and
and L3.
L3.

2.3. Conceptual
To resolveDesign B: Assembled
the issue, Structure
additional flexureofmust
Multi-Planar
be builtLayout with
into the Monolithic
coupling Hinges
joints to ac-
but Flat-Spring Couplers—Exploratory Study
commodate both relative rotational movements and relative translational movements. To
In this
achieve thisdesign, levers areco-planar
in a monolithic disposedlayout,
side bydesign
side with all arms
of the in parallel
coupling joints and
couldwith the
become
Sensors 2023, 23, x FOR PEER REVIEWplane of motion of each arm different but parallel to each other, as the multi-planar7layout of 28
complicated and the overall size of the structure could increase. Another issue of the co-
shown layout
planar in Figure 4. Bythe
is that this multi-planar
height layout, the
of the structure, height ofalong
measured the assembly can be
the direction ofreduced.
displace-
ment at position A, can become too large and no longer “compact”. For these issues, the
design approach was switched to multi-planar layouts.

2.3. Conceptual Design B: Assembled Structure of Multi-Planar Layout with Monolithic Hinges
but Flat-Spring Couplers—Exploratory Study
In this design, levers are disposed side by side with all arms in parallel and with the
plane of motion of each arm different but parallel to each other, as the multi-planar layout
shown in Figure 4. By this multi-planar layout, the height of the assembly can be reduced.
A challenge associated with this approach is how to couple levers placed side by side
to effectively transfer the displacement from one stage to the next stage. An L-shaped flat
(a)
spring foil was devised for this purpose. As depicted (b) in Figure 4, the coupling foil CF23 is
mounted
Figure with
Figure4.4.Conceptual
a horizontal
Conceptualdesigndesign B,
bare section
B, assembled
D 2 on top of lever 2 and a vertical bare section D1
structureofofmulti-planar
assembled structure multi-planarlayout
layout with
with monolithic
monolithic hinges,
hinges, but
to the
but sideflatofspring
“L-shape”
“L-shape”
lever
flat L3. The
spring
couplers
samebetween
couplers
between
islevers;
for the coupling
levers;
(a) (a) CF23
coupler
foilisCF12.
coupler CF23As
“pull”
already
is “pull”
type;
described,
type;
(b) CF23 is (b) CF23
“push”
there
is
type.
are both
“push” rotational relative motions and linear relative motions when two adjacent lever
type.
armsArotate aboutassociated
challenge their hinges.
withFigure 5 showsisan
this approach howexaggerated depiction
to couple levers of this
placed sidephenom-
by side
enon. The horizontal
to effectively baredisplacement
transfer the section of the coupling
from one stagefoiltois the
devised for accommodating
next stage. An L-shaped flat the
spring foilrelative
rotational was devised for this
motions andpurpose.
the verticalAs bare
depicted in Figure
section 4, the
for linear coupling
relative foil CF23 is
motions.
mounted with a horizontal
Key parameters bare
of this section
design D2 on top
approach wereof lever
tested 2 and
andastudied
vertical by
bare section
using D1 ele-
finite to
the side of lever L3. The same is for the coupling foil CF12. As already
ment models, as one example shown in Figure 6a, with the goal to achieve a total magni- described, there are
both rotational
fication ratio over relative motions
100. The and linear
following relative motions
observations when two adjacent lever arms
were obtained.
rotate about their hinges. Figure 5 shows an exaggerated depiction of this phenomenon.
Sensors 2023, 23, 326 7 of 28

(a) (b)
(a) section of the coupling foil is devised
(b) for accommodating the rotational
The horizontal bare
Figure 4.motions
relative4.
Figure Conceptual design
anddesign
Conceptual B, assembled
assembled
the vertical
B, structure
bare section of multi-planar
multi-planar
for linear
structure of layout with
relative motions.
layout with monolithic
monolithic hinges,
hinges,
but “L-shape”
but “L-shape” flat spring
Key parameters
flat spring couplers
of this between
design
couplers levers;
approach
between (a) coupler
were
levers; (a) couplerand
tested CF23
CF23 is “pull”
“pull”
studied
is type;
by type; (b)
using(b) CF23 is
finite
CF23 is
“push”
element type.
models,
“push” type. as one example shown in Figure 6a, with the goal to achieve a total
magnification ratio over 100. The following observations were obtained.

Figure 5. Exaggerated
Figure 5.
Figure Exaggerated depictionofof
Exaggerated depiction
depiction ofrelative
relativetranslational
relative translational
translational displacement
displacement
displacement dd and
d and and rotational
rotational displacement
displacement
rotational displacement
δ
δ between
between two
two adjacent
adjacent levers
levers at
at the
the location
location of of
thethe coupler,
coupler, in in
the the conceptual
conceptual design
design
δ between two adjacent levers at the location of the coupler, in the conceptual design B. B. B.

(a)
(a)

(b)
(b)
Figure
Figure 6. (a)
Figure 6. (a) A
A finite
A finiteelement
finite elementmodel
element modelofof
model ofthe
the
the conceptual
conceptual
conceptual design
design
design B; (b)
B;
B; (b) (b) undesired
undesired
undesired shear
shear
shear deformation
deformation
deformation
found
found in one
found in one ofof the
of the monolithic
themonolithic hinges.
monolithichinges.
hinges.

(1) The coupling foils should be pulled rather than pushed: The first question is in what
orientation the L-shaped flat spring foil should be mounted between two adjacent
lever arms. In the configuration of Figure 4a, when an input force pushes on position
A the output displacement of lever L2 “pulls” lever L3 through the coupling foil
CF23; while in the configuration of Figure 4b, lever L2 “pushes” lever L3. Considering
mechanics of material, pulling should be preferred because it results in mainly tension,
which is easier to take by the coupling foil than pushing, which has a potential of
 (1) The coupling foils should be pulled rather than pushed: The first question is in what
orientation the L-shaped flat spring foil should be mounted between two adjacent
lever arms. In the configuration of Figure 4a, when an input force pushes on position
A the output displacement of lever L2 “pulls” lever L3 through the coupling foil
Sensors 2023, 23, 326 CF23; while in the configuration of Figure 4b, lever L2 “pushes” lever L3. Consider- 8 of 28
ing mechanics of material, pulling should be preferred because it results in mainly
tension, which is easier to take by the coupling foil than pushing, which has a poten-
tial of causing bulking in the foil. Finite element simulations using Ansys software
causing bulking in the foil. Finite element simulations using Ansys software confirms
confirms this consideration, as a model configuration of Figure 4b gives a smaller
this consideration, as a model configuration of Figure 4b gives a smaller magnification
magnification ratio than the model configuration of Figure 4a does.
ratio than the model configuration of Figure 4a does.
(2) For an assembly of lever structures of a fixed size, the thickness of the coupling foils
(2) For an assembly of lever structures of a fixed size, the thickness of the coupling foils
has a minimal effect on the magnification ratio.
has a minimal effect on the magnification ratio.
(3) The coupling foils should be in contact with the lever arms with widths as large as
(3) The coupling foils should be in contact with the lever arms with widths as large as
possible, as depicted as W1 and W2 in Figure 4a, so that there is sufficient stiffness in
possible, as depicted as W 1 and W 2 in Figure 4a, so that there is sufficient stiffness in
the coupling, which helps to improve the magnification ratio.
the coupling, which helps to improve the magnification ratio.
(4) A larger width (as W2 in Figure 4a) of each lever structure also increases the structure
(4) A larger width (as W 2 in Figure 4a) of each lever structure also increases the structure
stiffness
stiffness and
and thereby
thereby helps
helps toto improve
improve the magnification ratio.
the magnification ratio. However,
However, activation
activation
force
force also increases with a thicker lever structure. For an input force not exceeding1
also increases with a thicker lever structure. For an input force not exceeding
N,
1 N,the width
the width ofofthe levers
the leversininthe
thedesign
designofof
Figure
Figure 6a6awaswassetsettotobebe8 8mm.
mm.
(5) The sizes of the bare sections of the coupling foil affect achievable
(5) The sizes of the bare sections of the coupling foil affect achievable magnification
magnification ratio
ratio
significantly, especially the horizontal bare section D 2. The bare sections D1 and D2
significantly, especially the horizontal bare section D2 . The bare sections D1 and D2
on
on the
the coupling
coupling foil foil are
are devised
devised to to accommodate
accommodate the relative linear
the relative linear movements
movements and and
relative rotational movements between adjacent, coupled lever
relative rotational movements between adjacent, coupled lever arms. Sizes too small arms. Sizes too small
make
make thethe cascaded
cascaded leverlever structures
structures tootoo stiff
stiff to
to deform,
deform, but but sizes
sizes too large make
too large make thethe
coupling too soft to effectively transfer displacement, both hurting
coupling too soft to effectively transfer displacement, both hurting the magnification the magnification
ratio.
ratio. Therefore,
Therefore,for for aa cascaded
cascaded lever lever structure
structure of of aa given
given size,
size, there should be
there should be anan
optimal
optimal range
rangeof ofbare
baresection
sectionsizes.
sizes.Different
Different values
valuesof ofD1Dand
1
D
and were
2 D
2
tested
were to
tested ob-
to
serve
observetheir effects
their effectsononthethe
total magnification
total magnificationratioratioand
andthe theresults
resultsare are summarized
summarized in in
Figure 7. D = 2 mm and D = 1.6 mm were selected as close to
Figure D1 = 2 mm and D2 = 1.6 mm were selected as close to optimal for this design.
1 2 optimal for this design.

Figure 7. Effect
Effect of
of the
thesizes
sizesofofbare
baresections
sectionsofofthe coupling
the foils
coupling onon
foils thethe
total magnification
total ratioratio
magnification of the
of
the conceptual
conceptual design
design B ofBFigure
of Figure
6a. 6a.

After
After the
the above
above adjustments
adjustments andand optimization,
optimization, the
the best
best total
total magnification
magnification waswas 57,
57,
which was still far short of the targeted 100. Close examinations of the
which was still far short of the targeted 100. Close examinations of the deformation atdeformation
at different
different stages
stages andand locations
locations in the
in the model
model of the
of the cascaded
cascaded leverlever structures
structures revealed
revealed that
that
therethere
werewere significant
significant shearshear
strainsstrains
in thein the horizontally
horizontally placedplaced monolithic
monolithic flexuralflexural
hinge,
hinge,
as shownas shown
in Figurein Figure
6b, and6b,these
and shear
these strains
shear strains
ate upate up a significant
a significant portion
portion of theofideal
the
ideal magnification ratio. Thus, the hinges needed improvement, which led
magnification ratio. Thus, the hinges needed improvement, which led to the next concep- to the next
conceptual
tual design.design.
2.4. Conceptual Design C: Assembled Structure of Multi-Planar Layout with Flat-Spring Couplers
but Re-Oriented Monolithic Hinges—Exploratory Study
In order to avoid shear strains in the hinges, the horizontally placed hinges should
be reoriented to become aligned along the direction of input force, that is, in the vertical
direction. The vertically placed monolithic hinges can be arranged to respond to the input
force by tension or by compression. From observations obtained in Section 2.3, flexural
joints in tension may be a better choice. However, a tensile arrangement of monolithic
 In order to avoid shear strains in the hinges, the horizontally placed hinges should
be reoriented to become aligned along the direction of input force, that is, in the vertical
direction. The vertically placed monolithic hinges can be arranged to respond to the input
force by tension or by compression. From observations obtained in Section 2.3, flexural
Sensors 2023, 23, 326 9 of 28
joints in tension may be a better choice. However, a tensile arrangement of monolithic
hinges in the cascaded lever structures could result in a somewhat complicated design.
This is because, in order to load the hinge in tension, the hinge needs to be placed above
hinges in the cascaded lever structures could result in a somewhat complicated design. This
the lever arm, and since a monolithic hinge takes up certain space, the distance between
is because, in order to load the hinge in tension, the hinge needs to be placed above the lever
the axis of the hinge and the input place (effort point) cannot be too small. This will not
arm, and since a monolithic hinge takes up certain space, the distance between the axis of
be helpful for designing a “compact” sensor device. As a result, we decided to test a com-
the hinge and the input place (effort point) cannot be too small. This will not be helpful
pressive arrangement
for designing of monolithic
a “compact” sensor device.hingesAsfirst.
a result, we decided to test a compressive
Finite element
arrangement modeling
of monolithic wasfirst.
hinges applied to help determine the optimal thickness of the
flexural section
Finite elementof the hinge in
modeling waseach stage to
applied that minimizes
help determine deformation
the optimal in that single
thickness stage.
of the
Other
flexural section of the hinge in each stage that minimizes deformation in that single stage. of
parameters were kept the same as those of the conceptual design B. Simulations
total
Otherdeformations
parameters were indicated
kept thethat a total
same magnification
as those over 100
of the conceptual is achievable.
design A detailed
B. Simulations of
design was made, indicated
total deformations as shownthatin Figure
a total 8. Finite element
magnification oversimulation showedAthat
100 is achievable. a down-
detailed
designforce
ward was made,
of 5 Nasatshown in Figure
position 8. Finite
A resulted in aelement simulation
displacement showed
of 0.6 μm at that a downward
position A and 60
force of 5 N at position A resulted in a displacement
μm at output position B, indicating a magnification of 100. of 0.6 µm at position A and 60 µm at
output position B, indicating a magnification of 100.

Figure 8. The conceptual Design C.

Figure 8. The conceptual Design C.
However, 5 N appeared too large for a mere 0.6 µm displacement. Further analysis
found However,
that even5larger
N appeared
forces aretoo large to
needed forproduce
a mere larger
0.6 μm displacement.
displacement Further
while analysis
the magni-
found that even larger forces are needed to produce larger displacement
fication ratio decreases with increasing displacement. To obtain a displacement of 1 µm while the mag-
nification ratio decreases with increasing displacement. To obtain a displacement
needs 10 N and the magnification ratio shrinks to 68. To obtain 10 µm displacement will of 1 μm
needs 10 N
need over 100and
N ofthe magnification
force. ratio shrinks
This is undesirable. Input to 68. To
forces obtain
could 10 μm by
be reduced displacement
reducing thewill
need over 100
thicknesses N of
of the force.sections
flexure This is but
undesirable. Input forces
then unfavorable could be
deformation reduced
due by reducing
to compression
would
the increase. It
thicknesses ofappeared
the flexurethatsections
this dilemma of using
but then compressive
unfavorable monolithic
deformation duehinges was
to compres-
difficult to resolve.
sion would increase. It appeared that this dilemma of using compressive monolithic
hingesTherefore, it was
was difficult to decided
resolve. to move to conceptual design D to use flat springs for
the hinges.
Therefore, it was decided to move to conceptual design D to use flat springs for the
hinges.
2.5. Conceptual Design D: Assembled Structure of Multi-Planar Layout with Flat-Spring Couplers
and Hinges
2.5. Conceptual Design D: Assembled Structure of Multi-Planar Layout with Flat-Spring
A. System layout
Couplers and Hinges
Through the studies of the previous conceptual designs, it was concluded that a better
A. System layout
approach for the hinge would be flat-spring in tension, which could minimize unfavorable
Throughasthe
deformation studies
well as hingeof the previous conceptual designs, it was concluded that a bet-
sizes.
ter approach for the the
Figure 9 depicts hinge would be
conceptual flat-spring
design in tension,
of a single which could
lever structure, which minimize
includes unfa-
a
base frame
vorable LB2, an arm
deformation (beam)
as well LA2 and
as hinge a hinge LH2 that connects the arm to the base
sizes.
frame. The flat-spring
Figure 9 depicts foilthehinge is attached
conceptual to a of
design base mounting
a single face
lever P1L2 on the
structure, baseincludes
which frame a
and to one end of the arm. The amounting can be by screw or by joining
base frame LB2, an arm (beam) LA2 and a hinge LH2 that connects the arm to the base such as brazing
or soldering
frame. or glue. A small
The flat-spring bare section
foil hinge of thetofoil
is attached spring
a base (Ls2 ) is toface
mounting facilitate
P1L2theonhinge
the base
function. The arm includes a protruded feature with a second mounting face P2L2, which
is parallel to the first mounting surface P1L2. When the protruded feature on the arm
is pushed downward (i.e., toward −z direction), downward displacement at the second
mounting face P2L2 is amplified at the end of the arm at a third mounting face P3L2 by a
factor of L2 /l2 , the ratio of the distances from the third mounting face P3L2 and the second
mounting face P2L2 to the base mounting face P1L2.
 hinge
hingefunction.
function.TheThearm armincludes
includes a protruded
a protruded feature with
feature a second
with mounting
a second face P2L2,
mounting face P2L2,
which is parallel to the first mounting surface P1L2. When the protruded
which is parallel to the first mounting surface P1L2. When the protruded feature feature on the on the
arm
armisispushed
pusheddownward
downward (i.e., toward
(i.e., toward−z −z
direction), downward
direction), downward displacement at theatsec-
displacement the sec-
ond
ondmounting
mountingface faceP2L2
P2L2 is is
amplified
amplified at the endend
at the of the armarm
of the at aat
third mounting
a third face P3L2
mounting face P3L2
Sensors 2023, 23, 326 by a factor of L 2/l2, the ratio of the distances from the third mounting face P3L2 and
by a factor of L2/l2, the ratio of the distances from the third mounting face P3L2 10 the
of 28 the
and
second mounting face P2L2 to the base mounting
second mounting face P2L2 to the base mounting face P1L2. face P1L2.

Figure 9. Single lever design of the conceptual design D.

9. Single
Figure 9.
Figure Single lever
leverdesign
designofofthe
theconceptual
conceptualdesign D. D.
design
Figure 10 explains the assembly of a magnification mechanism of three stages of the
Figure 10 explains the assembly of a magnification mechanism of three stages of
lever structures in exploded
Figure 10 explains theview. The three
assembly lever structures
of a magnification are of basically
mechanism similar
of three con-of the
stages
the lever structures in exploded view. The three lever structures are of basically similar
figuration as the one depicted in Figure 9. The lever structures are
lever structures in exploded view. The three lever structures are of basically similar disposed side by side con-
configuration as the one depicted in Figure 9. The lever structures are disposed side by side
with all arms
figuration in
as theparallel and
one depicted with the
in theplane
Figure of motion
9. of
The of
lever ofeach arm
structures different but parallel to side
with all arms in parallel and with plane motion each armare disposed
different butside by
parallel
each
with other. Theinfirst
allother.
arms stageand
parallel lever L1 includes
with planeaof
theincludes base frameofLB1, an arm LA1 and a hinge
to each The first stage lever L1 a motion
base frame each
LB1,arm different
an arm LA1 and but aparallel
hinge to
LH1.
each The arm LA1 also has a protruded feature as the contact place CP, onto which a ball,
LH1. other.
The arm TheLA1first
alsostage
has alever L1 includes
protruded featureaasbase frame LB1,
the contact placean CP,arm
ontoLA1which and a hinge
a ball,
for example,
LH1. The armmade
LA1 of Zircon,
also has ais mounted
protruded as the
feature contact
as the artifact
contact 5 as the
place input
CP, onto point.
which Thea ball,
for example, made of Zircon, is mounted as the contact artifact 5 as the input point. The
magnification
for example, ratio
made from the contact
of Zircon, place CPas
is mounted to the contact
end of the arm P3L1 is L1input
/l1. Lever L1 The
magnification ratio from the contact place CP tothe the end of theartifact 5 as the
arm P3L1 point.
is L1 /l1 . Lever L1
is oriented with the output end P3L1 of the arm LA1 pointing to the −x direction.
magnification
is oriented withratio from the
the output endcontact
P3L1 ofplace CP to
the arm LA1 thepointing
end of the
to thearm−xP3L1 is L1/l1. Lever L1
direction.
is oriented with the output end P3L1 of the arm LA1 pointing to the −x direction.

Figure10.
Figure Theconceptual
10.The conceptualdesign
designDDininexploded
explodedview.
view.

Thesecond
The secondstage
stagelever,
lever,L2,
L2,already
alreadydescribed
describedin inFigure
Figure9,9,isisplaced
placedininparallel
parallelrelative
relative
Figure
to lever10.
L1,The
butconceptual design
oriented with itsDoutput
in exploded
end ofview.
arm P3L2 pointing toward the +x direction.
to lever L1, but oriented with its output end of arm P3L2 pointing toward the +x direction.
That is, in order to have a compact form factor and small characteristic size, adjacent lever
That is, in order to have a compact form factor and small characteristic size, adjacent lever
The second
structures stage with
are aligned lever,the
L2,output
already described
ends in Figure
of the arms pointing9, is to
placed in parallel
opposite relative
directions.
structures are aligned with the output ends of the arms pointing to opposite directions.
to lever L1,
Further, thebut oriented
second with its
mounting output
face P2L2 end
of theof arm
arm LA2
P3L2ispointing
aligned toward
to the end theface
+x direction.
P3L1
Further, the second mounting face P2L2 of the arm LA2 is aligned to the end face P3L1 of
That
of theis,arm
in order
LA1 suchto have
thataan
compact
L-shapedform
foilfactor andfoil,
flexural small characteristic
as described size, adjacent
in Section 2.3, CF12lever
structures
couples leverareL1aligned with
and lever L2the output
at the ends The
two faces. of the arms
lower pointing
portion of thetofoil
opposite directions.
is mounted to
the arm LA1
Further, at its end
the second face P3L1face
mounting andP2L2
part of
of the arm
upper portion
LA2 is mounted
is aligned to the toendtheface
second
P3L1 of
mounting face P2L2 of the lever L2. Thus, the displacement at the output end of the arm
LA1 can be transmitted to the input end of the arm LA2 of lever L2.
The construction of the third stage lever, L3, and its coupling with the second stage
are basically similar. The output end of the third stage lever is the measurement point MP.
The 3 lever structures are assembled side by side together, with a spacer stripe (SP12
and SP23) between adjacent lever structures to prevent unwanted sliding contact between
 ples lever L1 and lever L2 at the two faces. The lower portion of the foil is mounted to the
arm LA1 at its end face P3L1 and part of the upper portion is mounted to the second
mounting face P2L2 of the lever L2. Thus, the displacement at the output end of the arm
LA1 can be transmitted to the input end of the arm LA2 of lever L2.
The construction of the third stage lever, L3, and its coupling with the second stage
Sensors 2023, 23, 326 11 of 28
are basically similar. The output end of the third stage lever is the measurement point MP.
The 3 lever structures are assembled side by side together, with a spacer stripe (SP12
and SP23) between adjacent lever structures to prevent unwanted sliding contact between
thearms,
the arms,and
andbecome
becomean anintegral
integralstructure,
structure,as
asdepicted
depictedininFigure
Figure11.
11.AAtiny
tinydisplacement
displacement
at the contact place CP is magnified at the measurement place MP by a cascaded
at the contact place CP is magnified at the measurement place MP by a cascaded ratio ratioof
of

Total theoretical
Total magnification
theoretical = (L
magnification /l11/l)1)(L(L
=1(L 2 /l
2/l22) (L3 /l3)3 )
3/l (1)
(1)

Figure11.
Figure TheMDMM
11.The MDMMconceptual
conceptualdesign
designD.
D.

Arrows 10 and 20 indicate directions of the input micro-displacement and of the

Arrows 10 and 20 indicate directions of the input micro-displacement and of the out-
output magnified displacement of the assembly, respectively. Assuming the magnification
put magnified displacement of the assembly, respectively. Assuming the magnification of
of each lever structure is five, the total theoretical magnification is then 125, two orders of
each lever structure is five, the total theoretical magnification is then 125, two orders of
magnitude amplification.
magnitude amplification.
B. The Design of contactofforce
construction and the
the third flatlever,
stage springL3,hinges
and its coupling with the second stage
are basically similar. The output end of the third
A simplified analytical model was developed for stage lever is the measurement
an initial approximationpoint MP.of
design
B.the flat-spring
Design hinges.
of contact The
force model
and assumes
flat spring that, in an assembled stages of cascaded lever
hinge
structure, output displacement and force
A simplified analytical model was developed can be transferred from approximation
for an initial one stage to thedesign
next stage
of
by “perfect” couplings with no loss and all forces in the vertical direction
the flat-spring hinges. The model assumes that, in an assembled stages of cascaded lever are balanced. It is
also assumed
structure, outputthat, as a rough approximation,
displacement and force can be thetransferred
bending offrom the hinge flat spring
one stage to thecan
nextbe
Sensors 2023, 23, x FOR PEER REVIEWviewed as a simple cantilever beam. Accordingly, this simplified cascaded
stage by “perfect” couplings with no loss and all forces in the vertical direction are bal-lever model can
12 of 28
be analyzed in mechanics as three free bodies with each body supported by a small elastic
anced. It is also assumed that, as a rough approximation, the bending of the hinge flat
simple beam, as illustrated in Figure 12.
spring can be viewed as a simple cantilever beam. Accordingly, this simplified cascaded
lever model can be analyzed in mechanics as three free bodies with each body supported
by a small elastic simple beam, as illustrated in Figure 12.

Figure
Figure12.
12. A
A simplified analyticalmodel
simplified analytical modelofof
thethe cascaded
cascaded lever
lever structures.
structures.

By force balancing, relations among input force F1, inter-lever forces F2 and F3, lever
arm weights, w1, w2 and w3, and geometric parameters can be formulated as follows:
⋅ ⋅
𝐹 = (2)
Sensors 2023, 23, 326 12 of 28

By force balancing, relations among input force F1 , inter-lever forces F2 and F3 , lever
arm weights, w1 , w2 and w3 , and geometric parameters can be formulated as follows:

M1 − w1 · L21 + F2 · L1
F1 = (2)
l1

M2 − w2 · L22 + F3 · L2
F2 = (3)
l2
L3
M3 − w3 · 2
F3 = (4)
l3
where the bending moments of the hinges can be approximated by the simple beam relation:

θ3 EI
M3 = (5)
Ls3

θ2 EI
M2 = (6)
Ls2
θ1 EI
M1 = (7)
Ls1
with θ 3 , θ 2 and θ 1 as the angles of the rotation of the lever arms, Ls3 , Ls2 and Ls1 are the
sizes of bare sections of the hinge springs, E is the Young’s modulus of the material of the
hinge spring, and area moment of inertia of the cross-section of the hinge spring expressed
in terms of its thickness t and width b as:

bt3
I= (8)
12
The input displacement, corresponding to the acting position of input force F1 , is:

din = d1 = l1 θ1 (9)

Additionally, the displacements at position 2 and 3 are:

L1
d 2 = l2 θ 2 = d (10)
l1 1

L2 L2 L1
d 3 = l3 θ 3 = d2 = d (11)
l2 l2 l1 in
Rearranging Equations (3)–(11) and inserting them into Equation (2) gives the expres-
sion of input force F1 as a function of input displacement din and system parameters:
     
din EI w1 L1 1 din L1 EI w2 L2 din L1 L2 EI w3 L3 L2 L
F1 = − + − + − · · 1 (12)
Ls1 l1 2 l1 Ls2 l1 l2 2 Ls3 l1 l2 l3 2 l3 l1 l2

By applying Equation (12), the dimension of the hinge flat-spring can be designed
to match the suitable range of contact force. Conceptual design D uses aluminum as the
lever structures, spring steel foils as hinges and takes key parameters set or learned from
previous designs:
l1 = l2 = l3 = 10 mm
L1 = L2 = L3 = 50 mm
Lever arm cross-section 8 mm × 8 mm:

Ls1 = Ls2 = Ls3 = 1 mm.

 ll11 == ll22 == ll33 == 10
10 mm
mm

LL11 == LL22 == LL33 == 50

50 mm
mm
Lever arm cross-section
Lever arm cross-section 88 mm
mm ×× 88 mm:
mm:
Sensors 2023, 23, 326 13 of 28
Ls11 == Ls
Ls Ls22 == Ls
Ls33 == 11 mm.
mm.
With these material and geometric
geometric information,
information,the
therelations
relationsbetween
betweeninput
inputforce
forceand
and
With
input these material and geometric
displacement information,were
hinge dimensions
dimensions the relations
computedbetween input force
andshown
shown inFigures
Figures
at different hinge were computed and in
and input displacement at different hinge dimensions were computed and shown in
13 and 14. By selecting flat-springs
flat-springs of
of thickness
thickness tt == 0.1
0.1 mm
mm andand width
widthbb==88mm,
mm,thetheinput
input
Figures 13 and 14. By selecting flat-springs of thickness t = 0.1 mm and width b = 8 mm,
contact force will be from zero to slightly
slightly over
over 88 N
N inin the
the full
full10
10μm
μmrange
rangeofofthe
themeasure-
measure-
the input contact force will be from zero to slightly over 8 N in the full 10 µm range of the
ment. This suggests that this design
design appeared
appeared viable.
viable.
measurement. This suggests that this design appeared viable.

Figure
Figure13. Relationsbetween
13.Relations between input
input
input force
force
force andand input
input
and displacement
displacement
input displacement of a of aa three-stage
three-stage
of cascaded
cascaded
three-stage lever lever
cascaded lever
structure different
structureatatdifferent hinge
hinge thickness
thickness
thickness t,t, when
t, when hinge
when hinge width
width
hinge width bb == 88as
b = 8 mm, mm, as
ascomputed
computed
mm, by
byEquation
by Equation
computed (12) of (12)
Equation (12)
of the
the simplified
simplified analytical
analytical model.
model.

Figure 14. Relations between input force and input displacement of a three-stage cascaded lever
structure at different hinge width b, when hinge thickness t = 0.1 mm, as computed by Equation (12)
of the simplified analytical model.

3. Prototypes and Tests

Prototypes were developed in three phases. In the first phase a model of three stages
of lever of comparatively large size, 82 mm in length, was built and tested to verify the
basic design concept and metrology parameters. In the second phase, compact models of
25 mm long with two stages of levers were built and tested. Finally, the compact model
was modified and packaged into an Alpha model with features of a pre-commercial model.

Sensors 2023, 23, 326
of lever of comparatively large size, 82 mm in length, was built and tested to verify the
basic design concept and metrology parameters. In the second phase, compact models of
25 mm long with two stages of levers were built and tested. Finally, the compact model
was modified and packaged into an Alpha model with features of a pre-commercial
model. 14 of 28

3.1. 82 mm Proof of Concept Prototype

3.1. 82Amm Proof
first of Concept
prototype Prototype
built was a proof-of-concept prototype using the design parameters
A first prototype
concluded in Section built
2.5 was a proof-of-concept
B. Figure 15 shows a prototype
photograph using the design
of the parameters
prototype displacement
concluded in Section 2.5 B. Figure 15 shows a photograph of the prototype
magnifying mechanism. This 82 mm long mechanism looks basically identical to the de- displacement
magnifying mechanism.
piction of Figure Thisthat
11, except 82 mm long mechanism
flat-spring hinges and looks basically
couplers wereidentical
mounted to the
by screws
depiction of Figure 11, except that flat-spring hinges and couplers were mounted by screws
and washers. The lever arms and bases were made from aluminum by wire EDM. A stain-
and washers. The lever arms and bases were made from aluminum by wire EDM. A
less-steel ball was used as the contact ball. A first surface reflector was attached to the
stainless-steel ball was used as the contact ball. A first surface reflector was attached to the
uppersurface
upper surfaceofof thethe output
output endend B,be
B, to toused
be used for optical
for optical lever lever measurement.
measurement. A hole A hole was
was
openedononthe
opened the base
base frame
frame underneath
underneath the output
the output end Bend
for B for mounting
mounting sensors.
sensors.

Figure15.
Figure 15.The
The proof-of-concept
proof-of-concept prototype,
prototype, with with an overall
an overall dimension
dimension of 82
of 82 mm mm
× 28 ××
mm 2824
mm × 24 mm.
mm.

InInorder
order tototest and
test calibrate
and the prototype,
calibrate a testing/calibrating
the prototype, setup was
a testing/calibrating setup built.
wasAs built. As
shown in Figure 16, the setup was basically a big lever with a metal push
shown in Figure 16, the setup was basically a big lever with a metal push beam supported beam supported
at
atthe
theleftleftend
endasas fulcrum
fulcrum (F)(F)
andandat right end end
at right by a by
vertically erected
a vertically linear linear
erected stage with
stagea with a
digital micrometer head as the effort end (X). The prototype displacement magnifying
digital micrometer head as the effort end (X). The prototype displacement magnifying
mechanism was placed close to the fulcrum underneath the push beam with the contact
mechanism was placed close to the fulcrum underneath the push beam with the contact
ball in contact with the lower surface of the push beam at a load point (Y). The lever ratio
ball
of FXin tocontact
FY waswith the 10:1.
roughly lowerThus,
surface of the
when the push beam at
micrometer a load
head pushedpoint
up(Y). The lever ratio
or lowered
of FXthe
down to push
FY wasbeam roughly 10:1. end
at the effort Thus,by 1when
µm, antheinput
micrometer headofpushed
displacement about 0.1 upµmortolowered
down the push beam at the effort end by 1 μm, an input displacement
the prototype could be generated. An eddy current sensor (sensor 1), Karman KD2306 1U1, of about 0.1 μm to
Sensors 2023, 23, x FOR PEER REVIEW 15 of 28
the placed
was prototype to thecould
side ofbethegenerated.
contact ball Anofeddy current to
the prototype sensor (sensor
precisely 1), Karman
measure the inputKD2306
displacement
1U1, was placed at loadtopoint Y as of
the side thethe
reference
contactforball
theofcalibration.
the prototype to precisely measure the
input displacement at load point Y as the reference for the calibration.

Figure16.
Figure 16.Setup
Setupforfor testing/calibrating
testing/calibrating thethe LCMDS
LCMDS prototypes.
prototypes.

Three different methods were applied to test/calibrate the prototype displacement

magnifying mechanism, with the aim to verify the feasibility of the design, and most im-
portantly, to test the repeatability of the hardware implementation.
A. Measuring prototype output displacement by a second eddy current sensor
A second eddy current sensor (sensor 2 in Figure 16), Micro-Epsilon eddyNCDT 3100
Sensors 2023, 23, 326 15 of 28

Three different methods were applied to test/calibrate the prototype displacement

magnifying mechanism, with the aim to verify the feasibility of the design, and most
importantly, to test the repeatability of the hardware implementation.
A. Measuring prototype output displacement by a second eddy current sensor
A second eddy current sensor (sensor 2 in Figure 16), Micro-Epsilon eddyNCDT
3100 EPS08, was placed at the output sensor mounting seat under the output end B of
the prototype mechanism (Figure 15) to measure output displacement and to check the
magnification ratio and measurement repeatability.
Initially, it was observed that the output displacement at B measured by the second
eddy current senor exhibited vibrations with amplitudes of 40–50 µm even when the input
end at A was stationary. It was suspected that the nearly overhanging arm of the L3 lever
was picking up the floor vibrations caused by frequent movements of members of our
laboratory, which is on the fifth floor of the engineering building. A simple damper with
a small acrylic piston in a small open oil cup was added underneath the lever arm (D, in
Figure 15) and basically eliminated the excess vibration.
Three measurements were performed to correlate the input displacement din , as
measured by the first eddy current sensor under the push beam of the calibration setup,
to the output displacement dout , measured by the second eddy current sensor. In each
measurement, the input was set relative to an arbitrary starting position and changed
by adjusting the micrometer head of the test setup to lower down the push beam and
thereby pushed down the contact ball. The output was also recorded as change relative to
a starting position corresponding to the staring position of the measurement. For repeated
measurements, the test setup and the prototype were returned to the original starting
position as measured by the readout signal of the first eddy current sensor, to ensure that
all measured curves have the same origin.
The results are shown in Figure 17. As the second eddy current sensor used for
measuring the prototype output has a maximal range of only 0.8 mm, the calibration could
only cover 6 µm of the input displacement. The slopes of the measured curves indicate
that the magnification ratio exceeds 150:1 in the first 2 µm and then approaches 100:1. The
unidirectional repeatability Ri of the prototype were estimated according to ISO 230-2 [18]
by applying the following relations:
r
1 n
∑

2
si ↑= j = 1
xij ↑ − x ↑ (13)
n−1
Ri ↑= 4si ↑ (14)
Take note that, although the “actual measurement” was performed on the magni-
fied displacements, the “effective measurement” of the prototype device was for the pre-
magnified input micro-displacement. In applications, measured output displacement
will be converted into input micro-displacement using the measured curves. Therefore,
the quadruple standard deviation 4si was estimated from the output displacements, but
the effective repeatability should be mapped to the input displacement according to the
measured curves, as highlighted in Figure 17. Thus, the repeatability of the prototype
mechanism, based on the measurement by the second eddy current sensor, was estimated
to be 0.005–0.2 µm over the 6 µm range, which is reasonably good, even considering the
small range around 5.5 µm that gave the 0.2 µm value.
 effective repeatability should be mapped to the input displacement according to the meas-
effective repeatability should be mapped to the input displacement according to the meas-
ured curves, as highlighted in Figure 17. Thus, the repeatability of the prototype mecha-
ured curves, as highlighted in Figure 17. Thus, the repeatability of the prototype mecha-
nism, based on the measurement by the second eddy current sensor, was estimated to be
Sensors 2023, 23, 326 nism, based on the measurement by the second eddy current sensor, was estimated16toof be28
0.005–0.2 μm over the 6 μm range, which is reasonably good, even considering the small
0.005–0.2 μm over the 6 μm range, which is reasonably good, even considering the small
range around 5.5 μm that gave the 0.2 μm value.
range around 5.5 μm that gave the 0.2 μm value.

Figure17.
Figure 17.The
Theresults
resultsof
ofthe
thecalbiration
calbirationof
ofthe
theproof-of-concept
proof-of-conceptprototype,
prototype,by
byusing
usingan
aneddy
eddycurrent
current
Figure 17. The results of the calbiration of the proof-of-concept prototype, by using an eddy current
sensor
sensor to measure output displacements from three separate measurments.
sensor to
to measure
measure output
output displacements
displacements from
from three
threeseparate
separatemeasurments.
measurments.

B. Measuring
B. Measuring prototype
prototype output
output displacement
displacement bybyananoptical
opticallever
lever
B. Measuring prototype output displacement by an optical lever
As the
As the
above
the above
measurement
abovemeasurement
measurementbyby
by the second eddy current sensor was onlyonlyable to cover
thethe second
second eddyeddy current
current sensor
sensor was was
only able to able to
cover
a 6
cover μm a range,
6 µm a laser
range, aoptical
laser lever
optical method
lever was
method further
was conducted
further to verify
conducted the
to measure-
verify the
a 6 μm range, a laser optical lever method was further conducted to verify the measure-
ment repeatability
measurement of the prototype
repeatability mechanism
of the prototype covering the targeted
mechanism full 10 μm measure-
ment repeatability of the prototype mechanism covering covering
the targetedthefull
targeted
10 μmfull 10 µm
measure-
ment range.
measurement range.
ment range.
Figure 18illustrates
Figure illustrates thesetup
setup ofthethe laseroptical
optical lever.AAlaserlaser wasmounted
mounted above
Figure 1818 illustrates the
the setup of of the laser
laser optical lever.
lever. A laser was was mounted above above
the
the testing/calibrating
testing/calibratingsetup setup with
setupwith the
withthethelaser
laserbeam
beam pointing
pointingdownward.
downward. A wide
A widenotch was
notch
the testing/calibrating laser beam pointing downward. A wide notch was
opened
was on
opened the push
onpush beam
the push so that, when the prototype mechanism was positioned under-
opened on the beambeam so when
so that, that, when the prototype
the prototype mechanism
mechanism was positioned
was positioned under-
neath the push
underneath the beambeam
push for calibration,
for the the
calibration, L3 lever
L3 armarm
lever andand
thethe
first surface
first reflector
surface reflectoron it
neath the push beam for calibration, the L3 lever arm and the first surface reflector ononit
(Figure
it 15)
(Figure15) was
15)was exposed
wasexposed to
exposedtotothethe laser
thelaser above.
laserabove.
above.TheThe laser
Thelaser beam
laserbeam
beamhithit the
hitthe first
thefirst surface
first surface reflector,
surface reflector,
reflector,
(Figure
reflected upward,hit
reflected hit anotherreflector
reflector fixedabove
above thetest test setupandand thenwas was redirected
reflected upward,
upward, hit another
another reflector fixedfixed above the the test setup
setup and then then was redirected
redirected
toaadistant
to distantscreen
screenofofgraph
graphpaper,
paper,as asshown
shownin inaa photograph
photographof of the
the laser
laser spot
spot on
on the
the screen.
screen.
to a distant screen of graph paper, as shown in a photograph of the laser spot on the screen.

Figure 18. Optical lever setup illustration and a photograph of the laser spot on a distant screen.
Figure 18.
Figure 18. Optical
Optical lever
lever setup
setup illustration
illustration and
and aa photograph
photograph of
of the
the laser
laser spot
spot on
on aa distant
distant screen.
screen.

Three measurements were conducted following a procedure similar to the one de-
Three measurements were conducted followingfollowing a procedure similar
similar to
to the one de-
scribed in the previous sub section, except that the output values were distances of move-
scribed in the previous sub section, except that the output values were distances of move-
ment of the laser spot on the screen. Figure 19 shows the results of measurements. The
ment of the
the laser
laser spot
spot on
on the
the screen.
screen. Figure
Figure 19
19 shows
shows the
the results
results of
of measurements.
measurements. TheThe
unidirectional repeatability Ri over the 10 μm measurement range was estimated, by the
unidirectional repeatability Rii over the 10 µm μm measurement range was estimated, by the
similar method used in Figure 17, to be 0.05–0.2 μm, which is in good consistency with
similar method used in Figure 17, to be 0.05–0.2 μm, which
0.05–0.2 µm, which is is in
in good
good consistency
consistency with
with
the results from the eddy current sensor measurement. The accuracy A of the prototype
the results from
from the
the eddy
eddy current
current sensor
sensor measurement.
measurement. The accuracy A of the the prototype
prototype
mechanism was also estimated, according to the same ISO 230-2 standard by applying the
following relation:
A = max[ xi + 2si ] − min[ xi − 2si ] (15)
and was calculated to be 0.4 µm.
 mechanism
mechanism was
was also
also estimated,
estimated, according
according to
to the
the same
same ISO
ISO 230-2
230-2 standard
standard by
by applying
applying the
the
following
following relation:
relation:
Sensors 2023, 23, 326 𝐴
𝐴== max
max 𝑥̅𝑥̅ + 2𝑠 −
+ 2𝑠 − min
min 𝑥̅𝑥̅ −
− 2𝑠
2𝑠 17 (15)
of 28
(15)
and
and was
was calculated
calculated to
to be
be 0.4
0.4 μm.
μm.

Figure
Figure
Figure 19. Results of
19. Results of three
three separate measurements of
separate measurements of the proof-of-concept prototype
the proof-of-concept prototype by
by using
using the
the
optical
optical lever.
lever.

C. Measuring
C. Measuring
Measuring prototype
prototype
prototype output
output
output displacement
displacement
displacement byby
by a Hall
aa Hall
Hall sensor
sensor
sensor
AA Honeywell
A Honeywell SS495A
Honeywell SS495A Hall
SS495A Hall effect
Hall effect sensor
effect sensor was
sensor was mounted
was mounted to
mounted to the
to the output
the output sensor
output sensor mounting
sensor mounting
mounting
seat
seat of
of the
the 82 mm
82 mm
82 prototype
mm prototype mechanism
mechanism (Figure
prototype mechanism (Figure 15).
(Figure 15). A
15). A tape
tape magnet,
magnet, cutcut from
from magnetic
magnetic
tape
tape obtained
obtained from
from the
the Misumi
Misumi Corporation,
Corporation, was was fixed
fixed toto the
the lower
lower surface
surface ofof the
the output
output
end
end facing
facing the
the Hall
Hall sensor
sensor (C,
(C, in
in Figure
Figure 15).
15). An
An Arduino
Arduino Uno Uno board
board was
was used
used to
to read
read the
the
output voltage from the Hall sensor and send
output voltage from the Hall sensor and send it to a laptop it to a laptop
laptop PCPC for
PC for records.
for records. A prototype
records. A prototype
sensor
sensor device
device was
was thus
thus constructed.
constructed.
Before
Before testing
testing the
the prototype
prototype sensor
sensor device,
device, aa separate
separate testtest on
on the
the Hall
Hall sensor
sensor waswas
conducted
conducted by by moving
moving aa tape
tape magnet
magnet relative
relative to
to the
the sensor.
sensor. TheThe output
output voltage
voltage toto input
input
displacement
displacement relation
relation shows
shows very
shows verygood
very goodlinearity
good linearityand
linearity andconsistency,
and consistency,as
consistency, asshown
as shownin
shown inFigure
in Figure20.
Figure 20.
20.
The Arduino Uno had an analog to digital conversion resolution
The Arduino Uno had an analog to digital conversion resolution of 10 bit and the Hall of 10 bit and the Hall
sensor
sensor had
had aa voltage
voltage output
output scale
scale of
of 555V.
V. Therefore,
V. Therefore, the the resolution
resolution ofof voltage
voltage measurement
measurement
was
was 0.005
0.005 V V (0.1%).
(0.1%). The
The standard
standard deviation
deviation of of the
the measurements
measurements was was 0.006
0.006 V,
V, which
which waswas
likely
likely to
to reflect
reflect the
the resolution
resolution of of the
the voltage measurement of
voltage measurement of the
the Arduino
Arduino Uno.Uno. 0.006
0.006 V
V
maps
maps toto aa repeatability
repeatability ofof 12.5
12.5 μm.
µm.
μm.

Figure
Figure 20.
20. Output
Output voltage
voltage of
of aa Honeywell
Honeywell SS495A
SS495A Hall
Hall senor
senor in
in relation
relation to
to the
the displacement
displacement of of aa
tape magnet
tape magnet
tape with
magnet with respect
with respect to
respect to the
to the sensor,
the sensor, three measurements
three measurements
sensor, three relative
measurements relative to
relative to one
to one arbitrarily
one arbitrarily set
arbitrarily set positon
positon as
set positon as
as
the
the origin.
origin.
the origin.

However,
However, itit should
shouldbe
should benoted
be notedthat
noted thataaaseparate
that separatecalibration,
separate calibration,
calibration,like
like in Figure
in in
like Figure 20,
Figure cannot
20,20,
cannot be
be
cannot
applied
applied
be to
to calibrate
applied calibrate the
to calibrate prototype
the the
prototype
prototype displacement
displacement
displacement magnifying
magnifying
magnifying mechanism
mechanism
mechanismthat used
that that aa dif-
useduseddif-
a
ferent
ferent piece
piece
different of tape
of of
piece tape magnet
tapemagnet
magnet in aa slightly
inin slightlydifferent
aslightly differentgeometric
different geometricenvironment,
geometric environment,because
environment, because itit was
because was
found
found that neither
that neither the
neitherthe magnetic
themagnetic polarity
magneticpolarity
polaritynornor
nor the
the
the strength
strength
strength is
is uniform
uniform
is uniform over
thethe
over
over the surface
surface
surface of
of
of the
commercial magnet tape. Therefore, there is almost no way to obtain two pieces of small
tape magnets with the same magnetic properties.
In testing the sensor device prototype, again, three measurements were conducted
following a procedure similar to the one described in sub section A, except that the output
values were the output voltage of the Hall sensor. Figure 21 shows the results of the
 the commercial magnet tape. Therefore, there is almost no way to obtain two pieces of
small tape magnets with the same magnetic properties.
Sensors 2023, 23, 326
In testing the sensor device prototype, again, three measurements were conducted18 of 28
following a procedure similar to the one described in sub section A, except that the output
values were the output voltage of the Hall sensor. Figure 21 shows the results of the meas-
urements. Again, the quadruple standard deviation was estimated from the Hall sensor
measurements. Again, the quadruple standard deviation was estimated from the Hall
output voltage, but the effective repeatability and accuracy mapped to the input displace-
sensor output voltage, but the effective repeatability and accuracy mapped to the input
ment according
displacement to the measured
according curves,curves,
to the measured as highlighted in Figure
as highlighted 21. Thus,
in Figure 21. the repeatabil-
Thus, the
ity of the prototype mechanism was estimated to be 0–0.4 μm over the 10 μm
repeatability of the prototype mechanism was estimated to be 0–0.4 µm over the 10 range
µm and
the accuracy
range and the was 0.7 μm.
accuracy was 0.7 µm.

Figure 21.
Figure 21. Results of three
three separate
separate measurements
measurements of the proof-of-concept prototype by
proof-of-concept prototype by using
using a Hall
sensor
Hall and aand
sensor tape
a magent to measure
tape magent output
to measure displacement,
output increasing
displacement, inputinput
increasing displacement meaning
displacement
decreasing
meaning the gap distance
decreasing between
the gap distance the Hall
between thesensor and the
Hall sensor tape
and themagnet at position
tape magnet B onBthe
at position onthird
arm.
the third arm.

To
Tosum
sumup upininother
otherwords,
words, when
whenapplying
applying thethe
prototype sensor
prototype device
sensor to measure
device a
to measure a
micro-displacement, a best-fit curve will first be generated from the curves in Figure
micro-displacement, a best-fit curve will first be generated from the curves in Figure 21. 21.
Then
Then aameasured
measuredvoltage
voltageoutput
output will bebe
will converted
convertedto atomicro-displacement
a micro-displacementaccording to to
according
this best-fit curve. The converted micro-displacement will have a repeatability
this best-fit curve. The converted micro-displacement will have a repeatability between between 0 0
and 0.4 µm and an accuracy of 0.7 µm.
and 0.4 μm and an accuracy of 0.7 μm.
Thus, the feasibility of the design and the ability of the prototype device to mea-
Thus, the feasibility of the design and the ability of the prototype device to measure
sure micron level displacements at sub-micron repeatability and accuracy are verified
micron level displacements at sub-micron repeatability and accuracy are verified and con-
and confirmed.
firmed.
3.2. 25 mm Compact Model
3.2. 25 mm Compact
Following Modelprocedure and based on the design features of the 82 mm proof
the design
Following
of concept the design
prototype, procedure
a compact 25 mm andmodelbased
was on the design
designed features of The
and developed. the 82 mm proof
compact
model alsoprototype,
of concept has a measurement
a compactrange25 mm ofmodel
10 µmwas anddesigned
repeatability and accuracy
and developed. Theincompact
the
sub-micron
model also level
has abut uses only two
measurement leverofstructures.
range 10 μm and repeatability and accuracy in the sub-
micronFigure
level22but
shows
usesthe
onlyfinal
twodesign and key dimensions of the compact model. The
lever structures.
magnification ratio was designed
Figure 22 shows the final design andto be aboutkey127. The thickness
dimensions of theand featuremodel.
compact dimensions
The mag-
of the flexural foils were determined from finite element simulations
nification ratio was designed to be about 127. The thickness and feature dimensions with three majorof the
goals. The first design goal was to keep the contact force below 0.3 N, close to commercial
flexural foils were determined from finite element simulations with three major goals. The
contact type displacement probes. The second goal was to make sure the flexural foils to
first design goal was to keep the contact force below 0.3 N, close to commercial contact
work within endurance limit. Flat spring foils were strip steel products from Sandvik [19].
type displacement probes. The second goal was to make sure the flexural foils to work
Analysis and finite element simulation of the model were conducted to check that stresses
within
in endurance
the hinges limit. Flatwere
and couplers spring
way foils were
below strip
the steel products
endurance limit offrom Sandvik [19].
the material. Anal-
Addi-
tionally, the third goal was to maintain the magnification ratio of the mechanism under in
ysis and finite element simulation of the model were conducted to check that stresses
the hinges
different and couplers
model werethat
orientations, wayis,below
when the
the endurance
direction oflimit of the
gravity material.
acting on theAdditionally,
model is
the third Simulation
different. goal was to maintain
showed the magnification
a magnification ratio
ratio close to of
125the
andmechanism
direction ofunder
gravitydifferent
has
amodel orientations,
negligible that is, when the direction of gravity acting on the model is different.
effect on it.
 Sensors 2023, 23, x FOR PEER REVIEW 19 of 2
Sensors 2023, 23, x FOR PEER REVIEW 19 of 28

Sensors 2023, 23, 326
Simulation showed a magnification ratio close to 125 and direction of gravity has a negl
model is different. Simulation showed a magnification ratio close to 125 and19direction
of 28 of
gible effect on it.
gravity has a negligible effect on it.

Figure 22. Layout and key dimensions of the MDMM compact model.
Figure 22. Layout
Figure 22. Layoutand
andkey
key dimensions
dimensions of the
of the MDMMMDMM compact
compact model.
model.

Thealuminum
The aluminum bodies of
thethe lever structures were made by wire EDM and CNC mil
The aluminumbodies bodiesof of lever
the structures
lever were
structures made
were by wire
made by EDM
wire and
EDM CNCand CNC
ing. TheThe
milling. flexural foils
flexural were
foils werecut
cutbybylaser.
laser.Due
Dueto tothe
the small
small size ofof the
the model,
model,aaset
setofof fixture
milling. The flexural foils were cut by laser. Due to the small size of the model, a set of
were made
fixtures and aand
were made step-by-step procedure
a step-by-step procedurewas developedtoto
was developed ease
ease assembly.
assembly. Figure
Figure 23 23 is
fixtures
isCAD
a CAD
were
model
made and aofstep-by-step
depiction the use of the
procedure
fixtures in
was
the
developed
assembly of the
tomodel.
ease assembly.
Figure 24
Figure
model depiction of the use of the fixtures in the assembly of the model. Figure 2
23 is a CAD model
shows depiction of the use
withofawith
the
Hallfixtures in the assembly of the model. Figure
showsa picture of the
a picture compact
of the model
compact model asensor and
Hall sensor a tape
andmagnet
a tape mounted.
magnet mounted.
24 shows a picture of the compact model with a Hall sensor and a tape magnet mounted.

Figure 23. Fixtures for the assembly of the compact model.

23. Fixtures
Figure 23.
Figure Fixturesfor
forthe assembly
the of the
assembly compact
of the model.
compact model.
Sensors 2023, 23, x FOR PEER REVIEW 20 of 28

Sensors2023,
Sensors 2023, 23,326
x FOR PEER REVIEW 20 of 28
28
Sensors 2023,23,
23, x FOR PEER REVIEW 20 of
20 of 28

Figure 24. A prototype of the compact model, with an overall dimension of 25 mm × 11 mm × 12

mm.
Figure 24. A prototype of the compact model, with an overall dimension of 25 mm × 11 mm × 12
Figure 24. A prototype of the compact model, with an overall dimension of 25 mm × 11 mm × 12
mm. 24. A prototype of the compact model, with an overall dimension of 25 mm × 11 mm × 12 mm.
Figure
mm.
The relationship between input force and input displacement of the compact model
The relationship between input force and input displacement of the compact model
was The relationship
measured between
byplacing
placing input force
thecompact
compact and input
model on displacement
a precision scaleofand
the applying
compact model
a vertical
The relationship
was measured by between
the input force
model andoninput displacement
a precision scale of applying
and the compact model
a vertical
was measured
micro-stage to by placing
push on the
the compact
contact ballmodel
of theoncompact
a precision scaleFigure
model. and applying
25 showsa vertical
the meas-
was measured
micro-stage by placing
to push the compact
on the contact ball of model on a precision
the compact scale 25
model. Figure and applying
shows a vertical
the measured
micro-stage
ured results to push
onpush on the contact
threeondifferent ball
compact of the compact model. Figure 25 shows the meas-
micro-stage
results to
on three differentthe contact
compact ball of models,
models, the but model.
butcompact
all made alltomade to the
theFigure
same 25same
showsspecifications.
the meas-
specifications. The
uredinput
The results on three
force is different
below Ncompact models, butrange.
all made to the same specifications.
ured force
input results
is on three
below N0.15
different
0.15 within within
the 10 the
compact 10
models,
µm μm
range.but all made to the same specifications.
The input force is below 0.15 N within the 10 μm range.
The input force is below 0.15 N within the 10 μm range.

Figure
Figure 25.
Figure25. Inputforce
25.Input forcein
force inrelation
in relationto
relation toto input
input
input displacement
displacement
displacementof of
thethe
ofthe compact
compact
compact model
model
model measured
measured
measuredfromfrom
fromthreethree
three
Figure 25. Input force in relation to input displacement of the compact model measured from three
different prototypes
differentprototypes
different ofthe
prototypesofof thecompact
the compact
compact model.
model.
model.
different prototypes of the compact model.
Tests
Tests and
Tests and calibrations
calibrations of of the
ofthe compactmodel
thecompact modelwere were conducted
conducted using
using the
thethesetupsetup of Fig-
of Fig-
Figure Tests
16 and calibrations
calibrations
following the same of the compact
compact
procedure and
model
model
the
were
were
same
conducted
conducted
electronic
using
using
readout the setup
setup
setup used
of
of Fig-
on
ure 16
ure 16 following
following the
the same
sameprocedure
procedureand andthethesame
same electronic
electronic readout
readoutsetup
setupused on the
used on the
ure82
the 16mmfollowing the
prototype same
with procedure
the Hall and
sensor. the same
Figure electronic
26 shows readout
the setup
compact used
model on
set the
in
82
82 mm prototype with the Hall sensor. Figure 26 shows the compact model set in meas-
82mm
mm prototype
prototype
measurement
with the
with
position
the Hall
Hallsensor.
insetup.
the setup.
sensor.Figure
Figure2626shows
Itfound
was found
shows
that thethe
the
compact
compact
floor
model
model
vibration set set
effect inonin meas-
meas-
the
urement
urement position
position in the
in the
the setup. It was
setup.ItItwaswasfound that
foundthat the floor
that thefloor vibration
floor vibrationeffect on
effect the
on output
urement
output position
lever, in
encountered in thetests
initial tests ofmmthethe vibration
82 mm prototype, effect
was not thethe
onseen output
output
in the
lever,
lever, encountered in the initial of the 82 prototype, was not seen in the compact
lever,encountered
compact encountered in in the
theinitial
initialtests
testsofofthe
the8282mmmm prototype,
prototype, was
was notnot
seenseen in the
in the compact
compact
model. model.
model.
model.

Figure26.
26. A prototypeof
of thecompact
compact modelininthe
the measurementposition
position in thesetup
setup of Figure16.
Figure 26.AAprototype
Figure prototype ofthe
the compactmodel
model in themeasurement
measurement positionininthe
the setupofofFigure
Figure
Figure
16. 26. A prototype of the compact model in the measurement position in the setup of Figure
16.
16. Figure 27 shows the results of the measurements. Note that the trend of decreasing
Figure
voltage w.r.t.27 shows the
increasing results of the
displacement in measurements. Note from
Figure 27 is different that the trend
Figure 21,of decreasing
Figure 27 shows the results of the measurements. Note that the trend ofbecause the
decreasing
voltage
tape w.r.t.
Figure
magnets 27increasing
shows
used in thethedisplacement
tworesults
cases of
had in Figure
the
different 27 is different
measurements.
polarities. Note
The from
thatFigure
the
repeatability 21,the
trend
of because
of the
decreasing
compact
voltage w.r.t. increasing displacement in Figure 27 is different from Figure 21, because the
tape magnets
voltage w.r.t. used in
increasingthe two cases
displacementhad different
in Figure polarities.
27 is The
different repeatability
from Figure of
21, the com-
model was estimated
tape magnets to be
used in the two0–0.7
casesµm
had over the 10polarities.
different µm range, The with the poorest
repeatability the com- the
because
of number
pact
tape model
occurring was estimated to be 0–0.7 μm over the 10 μm range, with the poorest number
pact magnets
modelin was used
a narrow inrange
the two
estimated becases
toaround 3 had
0–0.7 µm different polarities.
μm displacement.
over the 10 μmThe
range, The
accuracy repeatability
with the 0.5 µm.of
waspoorest the
As forcom-
number
pact model was estimated to be 0–0.7 μm over the 10 μm range, with the poorest number
Sensors 2023, 23, x FOR PEER REVIEW
occurringin a narrow range around 3 μm displacement. The accuracy was 0.5 μm.21As of for
28

the measurement resolution, the maximal voltage difference over the full measurement
Sensors 2023, 23, 326
range was 0.53 V, while the Arduino Uno readout board has a resolution of 10 21 of 28
bits over a
full voltage
occurring in range of 5range
a narrow V, which corresponds
around to a voltage
3 μm displacement. resolution
The accuracyof 0.005
was 0.5 V,
μm.which
As for is
about
the 1/1000 of 0.53
measurement V, therebythe
resolution, themaximal
displacement resolution
voltage is also
difference overabout 1/1000
the full or 0.1 μm.
measurement
the measurement
range was 0.53 V,resolution, the maximal
while the Arduino Unovoltage
readoutdifference
board hasover the full measurement
a resolution of 10 bits over a
range
full voltage range of 5 V, which corresponds to a voltage resolution ofof0.005
was 0.53 V, while the Arduino Uno readout board has a resolution 10 bits
V, over
which is
a full voltage range of 5 V, which corresponds to a voltage resolution of 0.005 V, which is
about 1/1000 of 0.53 V, thereby the displacement resolution is also about 1/1000 or 0.1 μm.
about 1/1000 of 0.53 V, thereby the displacement resolution is also about 1/1000 or 0.1 µm.

Figure 27. Results of three separate measurements of a prototype of the 25 mm compact model by
using a Hall sensor and a tape magent to measure output displacement, increasing input
displacement meaning decreasing gap distance between the Hall sensor and the tape magnet on the
output arm.
Figure 27.Results
Figure27. Resultsofofthree
threeseparate measurements
separate measurementsof aofprototype of the
a prototype 25 mm
of the compact
25 mm model
compact by by
model
using
usinga Hall sensor
a Hall and aand
sensor tape magent
a tape tomagent
measureto
output displacement,
measure output increasing input displacement
displacement, increasing input
3.3. LCMDS
meaning
displacement Alphagap
decreasing
meaning Model
distance between
decreasing the Hallbetween
gap distance sensor and
thethe tape
Hall magnet
sensor andon the
the output
tape arm.on the
magnet
outputBased
arm. on the 25 mm compact model specifications, an Alpha Model of the LCMDS
3.3. LCMDS Alpha Model
was designed and developed. The Alpha Model design includes a few modifications over
3.3. Based onAlpha
the 25Model
mm compact model specifications, an Alpha Model of the LCMDS was
andLCMDS
additions to the compact model in order to be as close to a commercial model as pos-
designed and developed. The Alpha Model design includes a few modifications over and
sible.Based on the 25 mm compact model specifications, an Alpha Model of the LCMDS
additions to the compact model in order to be as close to a commercial model as possible.
First, a package
was First,
designed is added to the
and developed. MDMM compact model. The package basically in-
a package is added The Alpha
to the MDMM Model designmodel.
compact includes a few
The modifications
package basicallyover
cludes
and
includes an
analuminum
additions base, to
to the compact
aluminum base, tomodel
whichthe
which inthe MDMM
order
MDMM to be is close
is as mounted,
mounted, toand and an acrylic
a commercial
an acrylic modelprotective
as pos-
protective
casing, consisting of a cover screwed to walls, as shown in a side view of the design in in
casing,
sible. consisting of a cover screwed to walls, as shown in a side view of the design
Figure
Figure 28. a package is added to the MDMM compact model. The package basically in-
First,
28.
cludes an aluminum base, to which the MDMM is mounted, and an acrylic protective
casing, consisting of a cover screwed to walls, as shown in a side view of the design in
Figure 28.

Figure28.
Figure 28.Design
Designofofthe
the packaging
packaging and
and thethe Contact
Contact Relay
Relay Mechanism.
Mechanism.

AAspecial
specialContact
ContactRelay
Relay Mechanism
Mechanism (CRM)
(CRM) is attached
is attached ontoonto
the inside surface
the inside of theof the
surface
cover of
cover of the package casing. The purpose of the CRM is to avoid direct contact between the
Figure 28. the package
Design of the casing.
packaging The andpurpose of the
the Contact CRM
Relay is to avoid direct contact between
Mechanism.
MDMM contact ball and an external target surface to be measured,
the MDMM contact ball and an external target surface to be measured, thereby protecting thereby protecting the
internal mechanism from any possible disturbance or contamination. The CRM includes a
the internal
A specialmechanism
Contact Relay fromMechanism
any possible disturbance
(CRM) is attached or contamination. The CRM
onto the inside surface in-
of the
thin but wide flexural cantilever beam with its fixed end attached to the cover and its free
cludesofa the
cover thinpackage
but wide flexural
casing. The cantilever
purpose beam
of the with its
CRM is fixed
to enddirect
avoid attached to the
contact cover
between
end attached with a seat holding a contact ball. A small opening on the cover allows the
andMDMM
the
CRM its free end
contact attached
contact
ball to ballout
stick with
and anaexternal
above seat holding
the top targetasurface
surface contact ball.
to
of the cover. A small
beThe CRM opening
measured, thereby
contact onprotecting
ball isthe
thecover
allows
the
ball to bethe
internal CRM
in actual contactfrom
mechanism
contact ball ato
with any stick out
to beabove
possible
surface the top
disturbance
measured. The surface
or of the
contamination.
cantilever beamcover. The
Themade
was CRM CRM
in-
contact
cludes aball
thinis the
but ball
wide to be in
flexural actual contact
cantilever with
beam a
withsurface
its to
fixed be
from 0.1 µm thick spring steel and had certain stiffness to limit the displacement of the measured.
end attached The
to cantilever
the cover
and its
CRM free
ball andend attached
its seat with in
to be only a seat holding
a direction a contact
normal to theball.
cover.A small opening
The lower onof
surface thethecover
allows
ball seatthewasCRM
madecontact
to be in ball to stick
contact outcontact
with the above ballthe of
topthesurface
MDMMofwith the some
cover.preload
The CRM
contact ball is the ball to be in actual contact with a surface to be measured. The cantilever
Sensors 2023, 23, x FOR PEER REVIEW 22 of 28

Sensors 2023, 23, 326 beam was made from 0.1 μm thick spring steel and had certain stiffness to limit 22 the
of 28dis-
placement of the CRM ball and its seat to be only in a direction normal to the cover. The
lower surface of the ball seat was made to be in contact with the contact ball of the MDMM
with
force.some preload
Therefore, anyforce. Therefore,
displacement any displacement
is transferred fully to is
thetransferred fully toFigure
internal MDMM. the internal
29
MDMM. Figure 29 depicts the construction
depicts the construction of the CRM. of the CRM.

Figure
Figure 29. Design of the
Design of theContact
ContactRelay
RelayMechanism
Mechanismin in exploded
exploded view.
view.

In addition,
In addition, aa pre-screening
pre-screeningprocedure procedureforfor magnet
magnet tapetapematerial
materialwaswasdeveloped
developed andand
conducted to select areas of strong magnetic field to be used
conducted to select areas of strong magnetic field to be used with the Hall sensor. with the Hall sensor. TheThe
procedure basically scans a Hall sensor over surface of a magnet tape and records voltage
procedure basically scans a Hall sensor over surface of a magnet tape and records voltage
output values to create a magnetic strength map of a tape. As a result, the maximal
output values to create a magnetic strength map of a tape. As a result, the maximal differ-
difference of voltage output over the full measurement range was significantly increased.
ence of voltage output over the full measurement range was significantly increased.
Further, the foil thickness of the flexural hinge of the second lever was increased from
Further,
0.01 mm to 0.02themmfoilinthickness of the flexural
order to increase hingeofofthis
the stiffness theoutput
secondlever
leverforwas increased
improved from
anti-
0.01 mm to 0.02 mm in order to increase the stiffness of this output
vibration capability. The modification increases the first mode natural frequency from 33 Hz lever for improved
anti-vibration
to 102 Hz, above capability.
the 10–55 The modification
Hz range used in some increases the first
commercial mode
contact typenatural frequency
displacement
from 33 Hz
sensors, to 102 Hz,
for example, theabove
Keyence thedigital
10–55 contact
Hz range used
sensor in some commercial contact type
[5].
displacement
Further, the sensors,
MDMM fordesign
example,usedthe Keyence
in the compact digital
model contact sensor [5].
was modified with a mechani-
callyFurther,
balancedthe MDMM
lever 2 arm.design
As canused in the
be seen compact
from model
the picture was modified
in Figure 24, the L2with a mechan-
lever arm
slacksbalanced
ically slightly downward,
lever 2 arm.because
As can it beisseen
supported
from the basically
pictureatinone end 24,
Figure by its
thehinge
L2 leverandarm
couplerslightly
slacks with itsdownward,
center of mass overhanging.
because As a result,
it is supported if the sensor
basically at onedevice
end by is to
itsbe usedand
hinge
in an orientation
coupler different
with its center from overhanging.
of mass the erected orientation,
As a result,such as insensor
if the horizontal
deviceorientation
is to be used
for measuring sideway displacements, the magnet-to-sensor gap
in an orientation different from the erected orientation, such as in horizontal orientation size at zero-input force
state will be different (larger than that of the erected orientation,
for measuring sideway displacements, the magnet-to-sensor gap size at zero-input force since gravity component
is smaller). However, this does not prevent the sensor from functioning, the difference
state will be different (larger than that of the erected orientation, since gravity component
between the maximal and the minimal of output voltage over the full measurement range
is smaller). However, this does not prevent the sensor from functioning, the difference
will become smaller, thereby reducing sensor resolution and sensibility. Figure 30 shows
between the maximal and the minimal of output voltage over the full measurement range
the design to adjust the location of the center of mass of the L2 lever arm. Basically, some
will become
mass was removed smaller, fromthereby reducing
the lever arm and sensor resolution
an extension withanda sensibility. Figure 30was
brass counterweight shows
the design to adjust the location of the center of mass of the L2
added to the opposite end of the arm. Ideally, the center of mass should be located in thelever arm. Basically, some
mass
middle was removed
of the from the
bare section lever
of the armsoand
hinge, thatanarmextension
mass poses with noamoment
brass counterweight
to the rotationwas
added
center ofto the
the arm.
opposite
However,end ofallowable
the arm.spaceIdeally,
underthethe
center of massby
limitations should be located
the casing wall and in the
middle
Sensors 2023, 23, x FOR PEER REVIEW of theon
hinge clamp bare
thesection
lever base of the
didhinge, so that
not permit thearm
idealmass poses
design. A no moment to
compromise wasthemade
rotation
23 of 28
center
to lowerof the center
arm. However,
of mass by allowable
about 1.2spacemm, under the limitations
as indicated by the
by “adjusted casing
center wall and
of mass”
in Figure
hinge clamp30. on the lever base did not permit the ideal design. A compromise was made
to lower the center of mass by about 1.2 mm, as indicated by “adjusted center of mass” in
Figure 30.

Figure 30.
Figure 30. Design
Design and
andconstruction
constructionofofaabalanced
balancedlever
leverarm
arm2.2.

Figure 31 shows a picture of the completed Alpha Model in side view. Figure 32
shows two packaged devices, with the Alpha Model in the foreground.
Sensors 2023, 23, 326 23 of 28
Figure 30.
Figure 30. Design
Design and
and construction
construction of
of aa balanced
balanced lever
lever arm
arm 2.
2.

Figure3131
Figure
Figure 31 shows
shows aa picture
picture
a picture
shows ofcompleted
of theof the completed
the completed AlphainModel
Alpha Model
Alpha Model in side
side view.
in side view.
Figure Figure 32
32 shows
view. Figure 32
shows
two two packaged
packaged
shows two packaged devices,
devices, with with the
the Alpha
devices, with the Alpha
Model Model
in the
Alpha in the
the foreground.
foreground.
Model in foreground.

Figure31.
Figure
Figure 31.The
31. Thecompleted
The completed
completed LCMDS
LCMDS
LCMDS Alpha
Alpha Model,
Model,
Alpha in side
in side
Model, in side view.
view.
view.

Figure 32. Two packaged LCMDS prototypes, the Alpha Model in the foreground.
Figure32.
32.Two
Twopackaged
packaged LCMDS prototypes,
Figure LCMDS prototypes, the the Alpha
Alpha Model
Model inforeground.
in the the foreground.

Testsand
Tests
Tests andcalibrations
and calibrations
calibrations of
of of the
thethe Alpha
Alpha
Alpha Model
Model
Modelwere were conducted
conducted
were conducted usingusing aa second
a second
using second calibration
calibration
calibration
setup,
setup, which
setup,which is capable
whichisiscapable
capableofof of positioning
positioning
positioning andand
and testing
testing a packaged
a packaged
testing a packaged or unpackaged
or unpackaged
or unpackaged prototype
prototype
prototype
inboth
in
in botherected
both erectedand
erected andhorizontal
and horizontalorientations.
horizontal orientations.Figure
orientations. Figure3333
Figure 33shows
showsthe
shows thenew
the new
new setup
setup
setup operating in
operating
operating in
horizontal
in horizontal orientation.
orientation. Similar
Similartotothe
thesetup
setup ofofFigure
Figure 16,
16,a manual
a manual linear
linear
horizontal orientation. Similar to the setup of Figure 16, a manual linear micro-stage actu- micro-stage
micro-stage actu-
ates the
ates thethe
actuates push
push beam
push
beambeamat at
at location
location X.X.The
locationX. The load
Theload point Y
load point sees1/5
Y sees
sees × of
1/5×
1/5× of the displacementofof
the displacement
displacement of the
the
the micro-stage,
micro-stage, which which
which was was precisely
was precisely measured
precisely measured
measured by by a Keyence
by aa Keyence CL-L030
Keyence CL-L030 color
CL-L030 color confocal laser
color confocal
confocal laser
laser
micro-stage,
displacement
Sensors 2023, 23, x FOR PEER REVIEWdisplacement sensor.
sensor. The
The Keyence
Keyence sensor
sensorhashas
an an
accuracy
accuracyof ~1
of µm
~1 and and
μm a resolution
a 24of
resolution
of 28of
displacement sensor. The Keyence sensor has an accuracy of ~1 μm and a resolution of
~0.25 μm.Therefore,
~0.25µm. Therefore, at load point
at load
load Y, setting
point a displacement
Y, setting
setting of 0.2 µm accuracy and 0.05 and
µm 0.05
~0.25 μm.
resolution
Therefore,
is achiveable.
at point Y, aa displacement
displacement of 0.2
of 0.2 μm accuracy
μm accuracy and 0.05
μm resolution
μm resolution is is achiveable.
achiveable.

Figure33.
Figure 33.AAsecond
second setup
setup forfor testing/calibrating
testing/calibrating the LCMDS
the LCMDS prototypes
prototypes at different
at different orientations,
orientations, the
the picture showing the setup operating in horizontal orientation.
picture showing the setup operating in horizontal orientation.

The results of calibration are shown in Figure 34. The curves of the two orientations
appeared quite similar and the maximal difference of voltage output over the full 10 μm
range was increased to over 1.4 V in both cases. Uni-directional repeatability and accu-
racy, estimated from 3 measurements by the same method used previously, and resolu-
 Figure 33. A second setup for testing/calibrating the LCMDS prototypes at different orientations,
Sensors 2023, 23, 326 24 of 28
the picture showing the setup operating in horizontal orientation.

The results of calibration are shown in Figure 34. The curves of the two orientations
The results
appeared quiteof calibration
similar are maximal
and the shown in difference
Figure 34. The curves output
of voltage of the two orientations
over the full 10 μm
appeared quite similar and the maximal difference of voltage output over the
range was increased to over 1.4 V in both cases. Uni-directional repeatability full 10 µmaccu-
and
range was increased to over 1.4 V in both cases. Uni-directional repeatability and accuracy,
racy, estimated from 3 measurements by the same method used previously, and resolu-
estimated from 3 measurements by the same method used previously, and resolution, based
tion, based on using the 10-bit Arduino Uno readout board, are summarized in Table 3.
on using the 10-bit Arduino Uno readout board, are summarized in Table 3.

Figure Results
Figure34.34. of calibrations
Results of the LCMDS
of calibrations of the Alpha
LCMDSModel operating
Alpha Modelin two different in
operating orientaitons.
two different
orientaitons.
Table 3. Parameters of performance of the LCMDS Alpha Model.
Table 3. Parameters of performance of the LCMDS Alpha Model.
Repeatability Accuracy Resolution
Alpha Model
(µm)
Repeatability (µm)
Accuracy (µm)
Resolution
Alpha Model
Erected orientation 0.20(μm) 0.42 (μm) 0.03 (μm)
Horizontal orientation 0.54 0.36 0.04
Erected orientation 0.20 0.42 0.03
Horizontal orientation 0.54 0.36 0.04
4. Discussions of Related Issues
4.1. Effect of Variations of Part Dimensions and Assembly Precisions
4. Discussions of Related Issues
Assembly of parts inevitably introduces relative positioning errors among parts. In
4.1. Effect of Variations of Part Dimensions and Assembly Precisions
practice and for cost reason, no two MDMMs can be manufactured and assembled to
Assembly
identical of parts
dimensions. As inevitably
tape magnets introduces
cannot berelative positioning
identical errors among
either, as already parts. In
mentioned,
practice and for cost reason, no two MDMMs can be manufactured and
future commercial LCMDS devices will need calibration individually before shipping. assembled to iden-
tical dimensions.
Therefore, it will beAsconvenient
tape magnets cannot be identical
if positioning variationseither, as already
in assembly canmentioned,
be toleratedfuture
to a certain extent
commercial LCMDS as long
devicesas the
willmagnification ratio
need calibration does not change
individually beforesignificantly and
shipping. Therefore,
measurement repeatability
it will be convenient is preservedvariations
if positioning in each device.
in assembly can be tolerated to a certain
Test results support such tolerance for assembly
extent as long as the magnification ratio does not change variations. Parts forand
significantly at least three
measurement
prototypes of the 25 mm MDMM compact
repeatability is preserved in each device. model were fabricated, assembled and tested
during the development process and all three prototypes worked with similar repeatability
and accuracy in a measurement range of 10 µm. It is reasonable to assume that small
variations of part dimensions and assembly precisions existed among these prototypes.
However, these small variations did not appear to affect the functionality of the prototypes,
as shown by the test results of two of the prototypes from Figures 27 and 34.

4.2. Effect of Temperature

Thermal expansion and contraction will cause a change of length of a lever arm.
Therefore, the question is: Does temperature variation cause change of magnification ratio?
From Equation (1), we can see that when temperature varies the lengths of load (L1 ~L3 )
and the lengths of effort (l1 ~l3 ) will expand, or contract, by the same proportion. Therefore,
the expansions or contractions will cancel each other out and a constant magnification ratio
will be maintained.
 tio? From Equation (1), we can see that when temperature varies the lengths of load (L1~L3)
and the lengths of effort (l1~l3) will expand, or contract, by the same proportion. Therefore,
the expansions or contractions will cancel each other out and a constant magnification
ratio will be maintained.
Sensors 2023, 23, 326 25 of 28
4.3. Effect of Magnetic Disturbance
The use of Hall sensor and magnet makes the LCMDS device potentially sensitive to
external
4.3. Effectmagnetic
of Magneticdisturbances.
Disturbance To understand how significant the effect can be, varia-
tions of output voltage of the Alpha Model were measured with a neodymium magnet
The use of Hall sensor and magnet makes the LCMDS device potentially sensitive to
placed
externalatmagnetic
differentdisturbances.
distances from and in different
To understand orientations
how significant withcan
the effect respect to the Alpha
be, variations
Model. It was found that at a distance of ~40 mm the neodymium magnet
of output voltage of the Alpha Model were measured with a neodymium magnet placed began to affect
the voltage distances
at different output, and
fromthe
andcorresponding magnetic field
in different orientations with strength
respect toexperienced by the Al-
the Alpha Model.
phas
It wasModel
foundat that
that at distance
a distancewas measured
of ~40 mm theto be about 0.2
neodymium mT (or
magnet beganabout three the
to affect to eight
times
voltagethe magnitude
output, and theofcorresponding
the geomagnetic field)field
magnetic at the Alphaexperienced
strength Model. That by is,
theunless
Alphasan ex-
Modelmagnetic
ternal at that distance was measured
disturbance to bethree
higher than aboutto0.2 mT times
eight (or about
thatthree
of thetoearth
eight is
times
imposed
thethe
on magnitude of the geomagnetic
Alpha Model, field)
the device will notatbe
the Alpha Model. That is, unless an external
affected.
magnetic disturbance higher than three to eight times that of the earth is imposed on the
Alpha Model,
4.4. Design for the device Small
Low-Cost will not
LotbeSize
affected.
Production
With the
4.4. Design performance
for Low-Cost of Size
Small Lot the Production
Alpha Model demonstrated and related issues ad-
dressed,
Withcurrent planningofofthe
the performance theAlpha
next Model
step ofdemonstrated
development and is related
to enable low-cost
issues small lot
addressed,
size production
current planningofofa the
Betanext
Modelstepwith size and features
of development comparable
is to enable to the
low-cost Alpha
small Model.
lot size
The BOM (bill of materials) cost of the Alpha Mode was slightly
production of a Beta Model with size and features comparable to the Alpha Model. over US$350, within
which machining
The BOM (bill costs were a cost
of materials) significant portion.
of the Alpha Mode Thewas design
slightly and manufacturing
over US$350, withinof the
which
Beta machining
Model coststwo
will have were a significant
features to reduceportion. The
costs. The design and manufacturing
first feature of the was
is that the MDMM
Beta Model will have two features to reduce costs. The first feature
redesigned to reduce number of parts and to make all parts into extruded geometriesis that the MDMM was that
redesigned to reduce number of parts and to make all parts into extruded
can be joined by glue rather than by screws. Figure 35 illustrates the idea of part geometries thatshape
can be joined by glue rather than by screws. Figure 35 illustrates the
simplification by extruded geometry. Figure 35a shows the design of the L1 lever arm of idea of part shape
simplification by extruded geometry. Figure 35a shows the design of the L1 lever arm of the
the Alpha Model, which requires precision machining to make a few small features. Fig-
Alpha Model, which requires precision machining to make a few small features. Figure 35b
ure 35b shows the simplified design, which, except for the circular recess at the top, is
shows the simplified design, which, except for the circular recess at the top, is basically an
basically
extrusionanof aextrusion of a 2D geometry.
2D geometry.

(a) (b) (c)

Figure 35. An
Figure 35. An example
exampleshowing
showingthethe idea
idea of of part
part shape
shape simplification
simplification by extruded
by extruded geometry,
geometry, from from
(a) to
(a) to (b) and manufacturing by applying a multi-part intermediate object
and manufacturing by applying a multi-part intermediate object (c). (c).

The second feature is that each part will be made by a process using a multi-part
intermediate object. As all parts are designed as extruded geometries, multiple parts of
the same geometry can be included into one multi-part intermediate object, which is a
single workpiece of the same cross-sectional geometry as the part and this multi-part
intermediate object can be produced by machining at a cost comparable to a small single
part but effectively at a much lower per part cost. For example, Figure 35c shows a 4-part
intermediate object including 4 pieces of the lever arm of Figure 35b. After the multi-part
intermediate object is machined, individual parts can be separated by wire EDM, which
will be a one-time operation of repetitive cut at a low cost.
Figure 36a shows an exploded view of the Alpha Model (excluding the casing and
the CRM), which included 20 parts and 12 screws. After simplification, the designs and
arrangement of parts are almost completely changed as shown in Figure 36b, which includes
12 parts (including a side wall that is part of the casing) and 4 screws. By using an eight-part
intermediate object, the BOM cost of one Beta Model was estimated, based on quotations,
to be less than US$100.
 Figure 36a shows an exploded view of the Alpha Model (excluding the casing and
the CRM), which included 20 parts and 12 screws. After simplification, the designs and
arrangement of parts are almost completely changed as shown in Figure 36b, which in-
cludes 12 parts (including a side wall that is part of the casing) and 4 screws. By using an
Sensors 2023, 23, 326 26 of 28
eight-part intermediate object, the BOM cost of one Beta Model was estimated, based on
quotations, to be less than US$100.

(a)

(b)
Figure 36.Design
Figure 36. Designmodification
modification ofAlpha
of the the Alpha
Model Model (excluding
(excluding the packaging
the packaging and the CRM)andforthe CRM) for
small
small lotproduction;
lot size size production; (a) Alpha
(a) current current Alpha
Model andModel and
(b) after (b) after modification
modification and simplification.
and simplification.

5. Conclusions
5. Conclusions
A new contact type compact displacement sensor device of sub-micron accuracy with
A new
low-cost contact
potential type
was compactand
developed displacement
tested. The sensor devicecombines
basic concept of sub-micron accuracy with
a mechanical
low-cost potential was developed and tested. The basic concept combines
magnifying mechanism of side-by-side placement of lever structures with flexural hinges a mechanical
in
magnifying
a multi-planar mechanism
layout with of side-by-side
a Hall placement
sensor to achieve of lever structures
good repeatability, with and
compactness flexural
low- hinges
in a multi-planar
cost. layout
Prototypes of two withsizes
different a Hall
and sensor to achieve
constructions¸ good
including anrepeatability, compactness
82 mm long proof-of-
concept prototype using three stages of lever structures and 25 mm long
and low-cost. Prototypes of two different sizes and constructions¸ including an 82 mmcompact models
usingproof-of-concept
long two stages of levers, with mechanical
prototype magnification
using three stages ofratios over
lever 100 made and
structures and tested.
25 mm long
A compact Alpha Model, which included a compact model mechanism with a balanced
compact models using two stages of levers, with mechanical magnification ratios over 100
lever arm and a contact relay mechanism in a protective package, was shown to have a
made and tested. A compact Alpha Model, which included a compact model mechanism
repeatability of 0.20–0.54 µm, an accuracy of 0.36–0.42 µm and a resolution of 0.03–0.04 µm
with a balanced range
in a measurement lever of
arm andAn
10 µm. a contact
important relay mechanism
finding inthe
is the use of a protective package, was
L-shaped coupling
foils for coupling levers placed side-by-side and the critical design parameters of the bare
section gaps of the foil, which determine the effectiveness of transferring displacements
and accommodating relative motions between the adjacent lever arms. To the authors’
knowledge, this is the first study and demonstration of using multiple stages of lever
structures in side-by-side multi-planar layout for precise displacement magnification with a
magnification ratio over 100. Part design modifications and corresponding manufacturing
procedure for small lot size production were designed to provide an estimated bill of
material cost per device under US$100. The low-cost devices can potentially be applied
in quantities in measuring errors of machine stages in multiple directions, monitoring
machine status and monitoring process status.
Sensors 2023, 23, 326 27 of 28

6. Patents
Key technical contents of the work reported in this paper are currently patent pending
in Taiwan, Mainland China and the United States [20,21].

Author Contributions: Conceptualization, C.-C.T.; methodology, C.-C.T. and Y.-C.T.; software,

Y.-C.T.; validation, Y.-C.T., W.-H.C. and Y.-S.C.; formal analysis, Y.-C.T., W.-H.C. and Y.-S.C.; investiga-
tion, Y.-C.T., W.-H.C., Y.-S.C. and L.-T.H.; data curation, Y.-C.T., W.-H.C. and Y.-S.C.; writing—original
draft preparation, C.-C.T. and Y.-C.T.; writing—review and editing, C.-C.T.; visualization, C.-C.T.,
Y.-C.T., W.-H.C., Y.-S.C. and L.-T.H.; supervision, C.-C.T. All authors have read and agreed to the
published version of the manuscript.
Funding: This research was funded by the Ministry of Science and Technology of the R.O.C. (Taiwan),
grant number MOST 107-2218-E-007-006.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments: We thank D. Shaw, C.K. Sung and S.C. Lin of the NTHU PME Department for
their valuable discussions with the research team.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Micro Epsilon. capaDTCT 6110 Datasheet. Available online: https://www.micro-epsilon.com/download/products/cat-capa/
dax-capaNCDT-6110--en.html#page=2&zoom=Fit (accessed on 15 October 2021).
2. Kaman. KD-2306 Datasheet. Available online: http://www.kamansensors.com/pdf_files/Kaman_KD-2306_data_sheet_web.pdf
(accessed on 15 September 2022).
3. Micro Epsilon. InduSensor Catalog. Available online: https://www.micro-epsilon.com/download/products/cat-induSENSOR-
en.pdf (accessed on 15 October 2021).
4. Micro Epsilon. optoNCDT 1320 Datasheet. Available online: https://www.micro-epsilon.com/download/products/cat-
optoncdt/dax-optoNCDT-1320-en.html#page=2&zoom=Fit (accessed on 15 October 2021).
5. Keyence. High-Accuracy Digital Contact Sensor GT-2 Series, Product Catalog, Keyence Corporation, 2016.
6. Keyence. Color Confocal Laser Displacement Sensor CL-3000 Series, Product Catalog, Keyence Corporation, 2018.
7. Slocum, A.H. Precision Machine Design; Prentice-Hall International Inc.: Hoboken, NJ, USA, 1992.
8. Kelemenová, T.; Dovica, M.; Sedlačko, P. Displacement Amplifying Systems. Acta Mechatron. Int. Sci. J. Mechatron. 2016, 1, 1–5.
9. Chen, F.; Zhang, Q.; Gao, Y.; Dong, W. A Review on the Flexure-Based Displacement Amplification Mechanisms. IEEE Access
2020, 8, 205919. [CrossRef]
10. Juuti, J.; Kordás, K.; Lonnakko, R.; Moilanen, V.-P.; Leppävuori, S. Mechanically Amplified Large Displacement Piezoelectric
Actuators. Sens. Actuators A 2005, 120, 225–231. [CrossRef]
11. Muraoka, M.; Sanada, S. Displacement Amplifier for Piezoelectric Actuator based on Honeycomb Link Mechanism. Sens.
Actuators A Phys. 2010, 157, 84–90. [CrossRef]
12. Henmi, N.; Sumi, Y.; Tanaka, M. Fast Drive of Displacement Magnification Mechanism with Flexure hinge using Loading Type
Impact Damper. J. Mech. Sci. Technol. 2010, 24, 211–214. [CrossRef]
13. Yu, Y.; Sung, B.; Yang, W.; Ge, Y. Strain Type Micro-Nano-Scale Micro-Nano Displacement Sensor. CN Patent CN 101435690B,
8 December 2010. (In Chinese, English translation available from Google Patent)
14. Yamauchi, T. The Annular Multilayer Rock Deformation Testing Device. CN Patent CN 1185580A, 24 June 1998. (In Chinese,
English translation available from Google Patent)
15. Bae, G.H.; Hong, S.W.; Lee, D.W. A Study on Precision Stage with Displacement Magnification Mechanism Using Flexure-Based
Levers. Res. J. Appl. Sci. Eng. Technol. 2014, 7, 2225–2231. [CrossRef]
16. Nada, Y.; Medhat, M.; Nagi, M.; Marty, F.; Saadany, B.; Bourouina, T. Mechanical Displacement Multiplier: 250 µm Stable
Travel Range MEMS Actuator Using Frictionless Simple Compliant Structures. In Proceedings of the MEMS 2012, Paris, France,
29 January–2 February 2012; p. 1161.
17. Merken, P.; Smal, O.; Debongnie, J.; Raucent, B. Mechanical Based Circular Notch Hinge Design. January 2007. Available
online: https://www.researchgate.net/publication/277065173_Mechanical_based_circular_notch_hinge_design (accessed on 15
September 2022).
18. ISO 230-2: 2014(E); Test Code for Machine Tools—Part 2: Determination of Accuracy and Repeatability of Numerically Controlled
Axes. ISO: Geneva, Switzerland, 2014.
Sensors 2023, 23, 326 28 of 28

19. Anonymous. SANDVIK Precision Strip Steel, SANDVIK, Application Note. 2019. Available online: https://www.materials.
sandvik/globalassets/global/downloads/products_downloads/strip-steel-and-strip-based-products/sandvik-precision-
strip-brochure-19-0863.pdf (accessed on 15 September 2022).
20. Tsao, C.C.; Tseng, Y.C. Low-Cost Compact Micro-Displacement Sensor. U.S. Patent 16/677,174, 7 December 2019. Counterpart in
Taiwan: Taiwan Patent 109138905, 6 December 2020.
21. Tsao, C.C. Low-cost Compact Micro-Displacement Sensor. CN Patent 202210474163.8, 29 April 2022.

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.