Gear Bearing Technology

jadesoreAI and Robotics

Nov 13, 2013 (3 years and 9 months ago)


Gear Bearing Technology

Midterm Report

Group 14

Pamela Carabetta

Marisa Jenkins

Sarah Kovach

Anuja Mahashabde

Qi Yan



Mr. John Vranish

Dr. Mircea Badescu

Ms. Kathryn DeLaurentis

Mr. Brian Weinberg

Prof. C M

Prof Mavroidis

Design of Mechanical Systems


Table of Contents

Executive Summary……………………………………………………..2

Problem Description…………………………………………………….4

Mission Statement………………………………………………………5



Gears and Transmissions………………………………………..7

Gear Bearings…………………………………………………...12


Robotic Arms……………………………………………………17

Design Specifications……………………………………………………23


Existing Robotic Arm……………………………………………24


Multidimensional Gear Bearing…………………………………27

Robotic Arm……………………………………………………..28

Robotic Wrist and Hand…………………………………………29


election Matrix…………………………………………………32



Project Management…………………………………………………….34


Works Cited……………………………………………………………..37


Executive Summary

Currently, pl
anetary gears are considered standard for a wide range of machining
applications. While planetary gears are traditionally the status quo, there are applications for
which they are not ideal. For space exploration devices, they are too heavy and expensive
. For
this reason, NASA has been searching for a replacement for the planetary gear.

John Vranish of NASA’s Goddard Space Flight Center developed the concept of the
“Gear Bearing.” Gear bearings improve upon current planetary gear technology because they

lightweight and allow for greater precision. They are also capable of astounding reduction ratios.
The benefits of the application of this technology to NASA are numerous.

The aim of this project is to apply gear bearing technology to a device to be

used for
space exploration. Five concepts have been proposed. The first is to improve an existing robot
arm. This idea would entail replacing the existing gear system with a gear bearing system. The
next concept is a rover that can travel over rough t
errain in outer space. The gear bearings would
transmit improved torque to the axles, thus providing greater exploration capability. The third
concept is a multi
freedom gear bearing. This concept involves taking Vranish’s gear
bearing technol
ogy to another level by creating more degrees of freedom within the gear itself.
The multi
freedom gear bearing would enable an increase in the overall ability of
potential applications. The fourth concept is a robot arm that would contain a hi
gh precision
system. Gear bearings would be used to control the motion of a sensor inside a specified radius.

The final concept is a robotic wrist and hand. This idea focuses on a specific piece of the
robotic arm. The robotic wrist and hand is the be
st concept due to group member interest,
feasibility, and potential application of the multi
freedom gear bearing. Also, the


relevance of a high precision wrist and hand to NASA’s mission makes the concept the most
advantageous choice.


Problem Description

Planetary gears have been desirable for a broad range of applications due to their speed
reduction and torque magnification properties. These gears serve several functions in space
missions, including telescopes, rovers, a
nd robotic arms. Limitations such as weight and cost
often stand in the way of the use of planetary gears for these purposes. For this reason, Mr. John
Vranish from Goddard Space Flight Center designed the gear bearing. Gear bearings combine
the benefit
s of both bearings and gears into a lighter and cheaper version that functions at a
higher level of performance. In addition to these developments, gear bearings create a much
higher speed reduction than typical planetary gears. They also provide a smoot
her roll due to
their bearing properties, making them ideal for application in telescopes. Currently gear bearings
have been used in telescopes; they have not yet been applied to other space applications, such as
robot arms and rovers. For this reason, w
e will investigate the use of gearing bearing technology
in other devices for NASA’s use.


Mission Statement

Mission Statement: Gearing Bearing Technology

Product Description

An application of a Gear Bearing into a device to be
used for space exploration

Key Business Goals

Establish a functional application for which Gear
Bearings will offer significant advantages over other
types of planetary gears and bearings

Design a model of this application by December 2003

Create a working model of the devices by
May 2004

Primary Market


Secondary Market

Robotics Industry


Precision oriented application

Low vibration necessary for application

Heavy Load requiring support in application


Group members

Rutgers Robotics Lab




The evolution of NASA’s space missions calls for a higher level of robotic technology.
Robots offer significant advantages over their human counterparts. While the need for robots in
space is quite clear, there are still many limitations. The ma
in drawback to using robots is that
they need additional power that is already limited by being isolated from the Earth. Their jobs
often require immense precision that is difficult to achieve using the conventional gears and
motors available. Their weig
ht and cost of parts also factor into their disadvantages. Finding a
solution to even one of these problems would make a substantial breakthrough in the ability of
robots to take over some of the more dangerous and tedious tasks in outer space.

Gear bea
rings pose a solution to several of these problems. They are lighter than
traditional planetary gear assemblies and also increase precision significantly. They will cost
less than standard planetary gears due to manufacturing differences. A further adva
ncement in
this technology would be a way to make these gears able to transfer work in more than one
direction. The traditional gear bearings and planetary gear assemblies have only one axis. The
group will to attempt to make a two or even three
axial ve
rsion of this gear bearing assembly that
would have applications in any device that has more than one degree of freedom. Simply
considering a robotic arm, it would reduce the amount of motors needed to only one, thus
significantly decreasing the amount of

power needed as well as reducing the weight and
complexity. The saved power would increase proportional to the number of degrees of freedom
added to the gear bearing.



Background research was conducted in order to fully understand Gear Bear
ings in relation to
current technology. Topics include gears and transmissions, gear bearings, motors, and robotic
arms. Initial group meetings indicated a preference for working on the robotic arm. While
concepts were developed for other applications o
f the Gear Bearing, the main focus was on the
robotic arm.

Gears and Transmissions

Virtually any machine involving rotating components uses gears. Gears serve many
mechanical functions, and among these functions are reducing speed, reversing direction of
rotation and changing the axis of rotation. Of the many abilities of gears, the most important is
speed reduction. Often a mechanical device may be powered by means of a small motor. A
small motor may be able to produce large RPMs, however, motors can o
nly transmit so much
torque to a device. Gears can solve this problem rather easily.

While the geometry of a basic gear can be quite complex (Figure1), one important value
is the pitch. When gears turn, the common normal to the surfaces at the contact

point always
intersects the line of centers at the same point. This point is called the pitch point. A circle
traveling through this point on each tooth is called the pitch circle [2].


. Gear Geometry [1]

The most s
ignificant gear concept is that of the gear ratio. When two gears mesh, it is the
ratio of one gear’s diameter in comparison to the other. There is an inverse relationship between
the ratio of diameters (gear ratio) and the ratio of angular speeds of eac
h gear. Along with
rotating in opposite directions, the larger gear, having a greater circumference, will rotate more
slowly than the smaller gear. If the ratio is 2:1, then the smaller gear rotates two times faster
than the larger gear.

Another (and pe
rhaps more definite) way to define a gear ratio is to take the ratio of the
amount of teeth on each gear. This provides a more perfect gear ratio, since diameters can be
subject to machine tolerances. The teeth of a gear also prevent slippage, ensuring t
synchronization of two axles.

Gears come in many forms for various applications. The most common type is the spur
gear (Figure 2). Spur gears have straight teeth, and are used to turn parallel axles. Helical gears


(Figure 3) are similar to spur ge
ars, but instead they have teeth cut at and angle. Contact
between the teeth is gradual, so less impact is delivered to each tooth; helical gears are much less
noisy than spur gears. Helical gears can be used to turn parallel or perpendicular shafts. Be
gears (Figure 4) serve the same function as helical gears, and can have straight, spiral, or hypoid
teeth. The hypoid bevel gear has the added advantage that it can turn perpendicular shafts that
need not lie in the same plane.

Figure 2. Spur Gear [

Figure 3. Helical Gear [3]


Figure 4. Bevel Gear [3]

One type of transmission system that combines these types of gears is the harmonic drive.
A wave generator operates harmonic drives. This rotates and displaces a flexible spline

When the spline is flexed, its teeth press on the teeth of an output ring, causing it to
move. Typically the output ring has one less tooth than the spline so that one revolution moves
the ring by one tooth. There are two types of harmonic drives, the p
ancake and the cup. The
pancake version is not efficient, and the cup can be too massive. The spline of this structure is
not very strong, and can easily fail or strip under certain loading. The mechanical advantage is
questionable, and harmonic drives
can be expensive.

In many applications a large gear ratio is very desirable. One way to achieve this is to
utilize a gear train. Combing several gears in series greatly increases the speed of the final gear.
Reversing this configuration can greatly red
uce the speed of the final gear in the train. Worm
gears produce a large gear ratio. In this configuration, a gear meshes with the threads on a shaft,
such that the shaft can turn the gear, but the gear cannot turn the shaft. One rotation of the shaft
dvances the gear by one tooth. This is a very effective method of speed reduction, providing
that a worm gear can be housed in whatever the particular application.


Figure 5. Worm Gear [3]

An important specific type of gear train is the planetary gear
. These are common in
many types of transmissions. Planetary gears allow for larger gear ratios, as well as turning the
input and output shafts in the same direction. A center gear (the “sun”) can mesh with all the
outer gears (the “planets”) at the sam
e time. These planets are connected to a carrier, while they
engage the inside of the outer ring. A shaft can then be attached to this outer ring. One
advantage of planetary gears is that they transmit higher torque. There are three meshes in the
ge planetary gear, as opposed to a one
gear mesh of the average parallel shaft gear (spur or
helical). This can make the transmitted torque three times as much. Planetary gears are low
backlash, which means there is less motion lost due to improperly siz
ed teeth and tooth spaces.
All three planets share the load on a planetary gear equally. This distribution of the load is
conducive to a longer life for the part. Planetary gears also readily accept high
speed inputs.
Since motors perform more efficien
tly at higher speeds, they work best in conjunction with a
planetary gear. The planetary gear effectively provides a low speed and high torque to a system
from a high speed/low torque motor [3].


Figure 6. Planetary Gear [3]

While very useful, planetary

gears do have limitations. Gear ratios larger than 10:1
typically require more stages of planetary gears, which may not be possible in certain systems,
due to factors such as size and weight. The carrier on a typical planetary gear may also require a
cond stage, and forfeits some of the gear’s efficiency. A separate ball
bearing system is used
for stability, as well as locating gear components and input or output drives. This complicates
and enlarges the system. Planetary gears also may be expensive

Gear Bearings

To solve many of the problems encountered with transmissions, John Vranish of NASA’s
Goddard Flight Center developed a machine part he calls “gear bearings.” The goal of gear
bearings was to create an “over
achieving planetary speed red
ucer, something that could step
down a motor output dramatically in a simple and lightweight form (
Mechanical Engineering
2002).” Although gear bearings may look like planetary gears, there is a distinct difference in
that the gears are not supported by
bearings. The first gear bearing created achieved a 70:1 ratio.


Vranish has also used a technique known as “phase tuning” to achieve an impressive ratio of
2000:1 [4].

Figure 7. Gear Bearing [4]

Figure 8. Exploded Gear Bearing [4]

Gear bearings can

be classified as either Spur Roller Gear Bearings or Phase
Gear Bearings. Spur roller gear bearings feature a spur gear with a coaxially mounted roller on
top. The roller diameter equals the pitch diameter of the spur gear, and the spur gear tee
th are
crowned where they border with the roller. The same idea is used to add a roll race on top of a
ring gear. These components, along with a sun gear, planets and a ring are assembled in a very


similar fashion to that of a planetary gear. The fixtur
e is stable, and does not require any
additional support. Thrust bearing contact occurs at the highest point on the crowned teeth,
which allows for speed matching. The planetary roller gear has high strength and efficiency.
The rollers improve the accur
acy when the gears mesh, by setting locations for the gears with
respect to each other. Also, the gears efficiently carry the rollers.

For phase
shifted gear bearings, there is a spur gear with its upper half rotated so as to be
exactly out of phase with
the lower half. The teeth in the upper half are positioned over the gaps
in the teeth of the lower half. The gears are in continuous contact as one drives the other,
however, they are out of phase. The teeth can be beveled and the motion can be optimize
d for
maximum strength and efficiency. The system works in a similar way to that of a 4
way thrust
bearing. At least one top half and one bottom half will always be engaged. This makes the gear
bearing very strong and the motion very smooth.

In compari
son to existing transmission technology, the gear bearing is superior. With
respect to the very popular state
art planetary gear, the gear bearing has many advantages.
It does not need a carrier, and therefore is already stronger and more efficien
t. The geometry of
the gear bearing compensates for the function performed by regular bearings on a planetary gear.
Eliminating supporting bearings also strengthens the structure. Gear bearings have high intrinsic
load bearing capacities. The other pro
minent existing transmission system is the harmonic drive.
Gear bearings offer much more strength and stability than the harmonic drive, and can be equally
as efficient if not more. They are also smaller and simpler, as well as cheaper than both of the
forementioned systems. Gear bearings appear to be the next step in creating the best possible
transmission system. They outclass the alternatives in terms of efficiency, strength, size,
mechanical advantage, simplicity and cost [5].



The basic type
s of linear and rotary motors are AC/DC, stepper, servo, hydraulic,
pneumatic and ultrasonic. DC and stepper motors are commonly used in robotic applications
because of their easy control, replacement, lower costs and high precision motion control.

The ba
sic principle behind the operation of any type of electromagnetic motor is the
interaction of magnetic fields between two components: the rotor or armature and the stator. This
interaction results in linear motion as in a linear motor or provides a torque
causing the armature
component to rotate as in a rotary motor. The stator is the stationary part of the motor while the
rotor moves. The differences in motors arise from how this interaction is brought about and how
the magnetic fields are set up.

The DC
motor contains a rotor called an armature and a stator. The stator in this case has
a fixed magnetic field as a result of either permanent magnets or electromagnetic windings. DC
voltage is provided to the armature through a commutator, which is mechanical

or electrical in
nature. In brushed DC motors, the commutator has carbon brushes that provide DC voltage to the
armature windings. A force acts on the current carrying rotor in the magnetic field and causes it
to rotate.

The stepper motor comes in three t
ypes: permanent magnet, variable reluctance and
hybrid, where the types refer to the differences in the rotor. The Unlike DC motors, the stepper
motors causes motion in incremental steps or is a digital device. A step drive processes digital
pulses and cau
ses incremental changes in the current in the stator windings. The time lag
between these incremental changes determines the speed of the motor. This in turn causes
incremental motion of the rotor and leads to rotation.


Motors are commonly combined with ge
arboxes to vary the output speeds and torque as
needed in the application. In this case, the motor will be coupled with the Gear
Bearing to
achieve the desired output.

The motion of the motor can be controlled though a feedback system such as a resolver o
encoder. Just as with the motors, there is an extensive range of feedback devices available, but
with respect to high precision optical encoders are most commonly used. Encoders are also
classified in terms of the position description they provide. Absol
ute encoders give the exact
angular position of the motor shaft without any reference position involved, each location has its
own unique code. Incremental encoders describe position with respect to a reference position.

The optical encoder consists of a l
ight source, a LED or a laser that is projected through
thin slits in a rotary disc for rotary encoders. A receiver on the opposite side of the disc
transforms the light projected into an electric signal from which position can be obtained. The
optical enc
oder is mounted on the motor and detects the motion of the motor output shaft [6].

Figure 9. Motor and Housing [7]


Robotic Arm

As NASA evolves the goals and capabilities of their missions, robots and robotic arms,
will become more common on unmanned ex
cursions and also on the space shuttles and space
station themselves. Their outstanding ability to service the station and shuttle without the help of
an astronaut allows the crew to remain inside of the craft. This procedure is notably safer than
the al
ternative. The ability of robotic arms to take samples of dirt and rocks allows information
to be sent back to earth from places in the solar system that are many years away from being
accessible to human visitors. Talks have even begun on creating a rob
otic colony on Mars to
prepare the planet for human existence [8].

Aside from the fact that using robotic arms increases the safety of the crew, it may also
increase the effectiveness of the repairs. Maintenance may be accomplished exactly to
ions by implementing a computer program and eliminating human error.

The first robotic arms were carried by missions such as Luna
13, Venera
13 and

14, and
the Mars
3, and
6. These basically used a spring system with no other control systems.
implemented on later Luna missions as well as the Surveyor probes used complex and
actuation controls [9].

The Surveyor probes in 1967 and 1968 became increasingly more complicated. The arm
had a varying length and three degrees of freedom with a fixed e
effector that sampled soils.
The azimuth had 112 degrees of angular movement; 42 degrees were in elevation, and a length
variation from 58 to 150 cm. Some later soviet probes achieved an elevation freedom of 100
degrees angular and a wrist rotation of

180 degrees. The Viking probes from 1976 expanded to
four degrees of freedom: 120 azimuth, along with wrist rotation, elevation and length [9].


The MPL had the most complex design to date in 1999, with 4 degrees of freedom [10].
MVACS was made of a gra
phite and epoxy resin and measured 2.2 meters long with a mass of
3.5 kg (5 kg with electronics). Although its small power requirements (10
20 W) benefited the
long journey, it did not allow the arm to move quickly. In fact, it barely moved more than a f
tenths of a degree a second with an accuracy of 1 cm and a repeatability of .5 cm [11]. This was
also a result of the need for small actuators. At max power, it was able to move at about 7 cm/s
with a force of 80N [9]. The power supply consisted of D
C motors with a max power of 42 W
and a 2 stage reduction made up of a harmonic drive and planetary gear [11]. The actuators are
able to produce anywhere from around 10 to 90 N
m of torque depending on the joint. Each
joint also contains a heater and a t
emperature sensor to keep the operating conditions needed for
the motors at above 208 K [11].

Table 1 shows the information on all four of the joints. Joints 1
4 are as follows: shoulder
azimuth, shoulder, elbow and wrist.

Table 1: Joint capabilities fo
r MVACS [11]


Joint 1

Joint 2

Joint 3

Joint 4


Actuator output torque






Gear ratios





Min angle






Max angle






Max speed (no load)













The software used to move the arm allows the positions to be programmed by relative or
absolute position. It is also able to find recover glitches in the programming, which is a must
when functioning alone on a planet far f
rom its inventors [11].

Figure 10: MPL [10]

More massive arms are found on the space shuttles and the new international space
station. The Space Shuttle Remote Manipulator System (RMS) is present on the space shuttles.
15 meters lo
ng, this arm has seven degrees of freedom: two degree shoulder, one degree elbow,
three degree wrist, and a 1 degree end effector for grabbing. It is controlled by a crew
inside of the space shuttle, and assists in docking, maintenance, and even he
lps in moving the
astronauts around outside of the shuttle [12].


The International space station will have several arms to assist the crew in repairs and
other tasks. The most multifaceted of them is called Canadarm 2. Its main responsibilities
helping in the assembly of the main components of the station, handling large payloads,
replacing parts and general maintenance, assist the astronauts in space walks. It also is used as a
transportation device around the station.

Perhaps its most ingeni
ous feature is its ability to move over the entire station. It is
designed to be able to move in inchworm fashion from grapple fixture to grapple fixture to reach
every corner of the station. The power supply is routed through each fixture to the arm
ending on its location. Figure 11 shows a simple representation of the Canadarm2. The two
large arms are 3.5 meters each, and contain 7 joints [12].

Figure 11. Canadarm2 [12]


There are many differences between these two arms. The RMS arm is 17.6 m
eters long
and has a mass of approximately 1,800 kg. It has a peak power of 2,000W and a mass handling
capacity of 116,000 kg. The average power to keep it functioning is 435 W and it has a stopping
distance under maximum load of 0.6 meters [12].

Canadarm2, Canadarm returns to earth every shuttle mission and is fixed to a
particular part of the shuttle. It has six degrees of freedom, and its joint rotation is limited to
approximately 160 degrees and has no sense of touch. Its length is 15 meters
long and weighs
410.5 kg. Its diameter is 33cm and has a mass handling capacity of 266,000 kg. An important
aspect is its speed. Unloaded it has the capacity to move at 60 cm/s, however loaded, this
reduces to about 6 cm/s. It is composed of 16 plies o
f a high modulus carbon fiber epoxy and,
like Canadarm2, is controlled autonomously or by astronaut [13].

The inchworm
like Canadarm2 on the other hand is made of 19 plies of high strength
carbon fiber. It is permanently in space and has a range of motion

that is only limited by the
number of Power Data Grapple Fixtures on the space station. Its seven degrees of freedom are
supplemented by the fact that it is able to change its configuration without moving its hands.
The rotation in all seven joints can
rotate up to 540 degrees, which is larger than the rotation of a
human arm. It is equipped with automatic collision avoidance and even force moment sensors
that provide limited sense of touch. Weighing 1800 kg and being 17.6 meters long with a
diameter o
f 35 cm, this arm is not only larger, but also heavier than Canadarm. Unloaded, it can
move around 37 cm/s. Loaded, it can move anywhere form 15 to 1.5 cm/s, depending on the
task [13,14].


The differences in the tasks require many different types of robo
ts and robotic arms.
These tasks may specify size, power, weight, resilience to environment, speed and precision.
With the continual interest and strive for more missions outside of Earth’s atmosphere, the need
to continually come up with easier methods

of building these robots increases.


Design Specifications

As a technology
push product, there are several potential applications to which a gear
bearing can be applied. Design specifications vary greatly depending on the project. In general,

aspects to the design will be consistent despite variations in the initial concept. The top
priorities for characteristics of a device for a space mission include being lightweight, size
efficient, and durable. John Vranish’s gear bearing necessarily ex
hibits all of these
characteristics. The application of the gear bearing must extend these principle qualities to the
rest of the design. For the most part, these specifications will be met through researching the
proper materials and designing the model

around those materials.


Brainstorming Ideas

Existing Robotic Arm

The idea of using an existing robotic arm revolves around the idea of testing the
effectiveness of the gear bearing in an application. The group would purchase a robotic arm
(Figure 12
) with given specification and programming. The arm would then be partially taken
apart to replace a planetary gear with a gear bearing. Tests would be set up to compare the
existing capabilities of the robotic arm to the new capabilities with the gear b
earing installed.

This concept offers several advantages in that it is well within the group’s abilities.
Because of this fact, tangible results will be produced. Its limitations include a large and risky
initial investment in a robotic arm that may or
may not effectively respond to the use of gear
bearing technology. On a more intellectual level, it may fail to offer the group a sufficient
challenge and a sense of accomplishment.

Figure 12. Lynx Robot Arm [15]



The concept of building a rover

deals with the idea that the extra torque created by the
Gear Bearing may be used to effectively propel its wheels over surfacing typically not travelable
by current designs. Initial ideas involve the use of either one or two motors per axel (Figures 13,

14). This number would vary based on the specific applications. The concept also includes the
option of adding a solar panel partially adjustable by the use of a gear bearing (Figure 15).

The rover has much to offer NASA in terms of a relevant technolo
gy. Additionally, the
project is within the group’s capabilities. The Robotics Lab is also knowledgeable on the topic
of rovers. One disadvantage is the lack of a clear reason for using gear bearing technology on a
rover. The smooth motion characterist
ic of gear bearings would offer little benefit to this
specific application. Also, gear reduction is not always necessary in rover devices. The group
additionally lacks a strong interest in pursuing this concept.

Figure 13. Rover with one motor per ax


Figure 14. Rover with two motors per axle

Figure 15. Solar Panel gear bearing application to Rover


Multidimensional Gear Bearing

Professor Mavroidis developed this concept during the meeting with the group members,
their advisors, and John

Vranish. Professor Mavroidis suggested the benefit to making a design
such as the gear bearing spherical. The group is currently examining at least a two degree
freedom gear bearing.

Schematics are shown Figures 16 and 17. The gear bearing design wo
uld be used as the
mechanism for the x
direction. As the motor input comes in from the left, the x
axial gear
bearing outputs a reduction on that same axis. Due to the mechanics of the gear bearing, it is
proposed to put another gear bearing around it, o
ffset at an angle

from the x
axis. The rotation
in the outer shell of this one can be made to act as the center “sun” for the second, causing it to
rotate. The resulting output in this new direction can be taken off of the planets or the outer shell
of the outer gear be
aring. It should be noted that this is very much theoretical, and requires more
time to consider its practicality.

Figure 16. Top view of 2D gear bearing


Figure 17. Side view of 2D gear bearing

While extremely challenging, this idea has much to offe
r in terms of developing a novel
technology that would be state of the art. The success of this project would be meaningful for
NASA and also other members of the robotics industry. The cost would also be low since the
only real expense would result from

the rapid prototyping machine. However, there is a high
level of risk associated with this concept because it pushes the capabilities of the group.

Robotic Arm

The initial project description from John Vranish suggests the use of his device in a
c arm. While many of the specifics of the device would be left up to the group members,
the basic concept of applying the Gear Bearing to a robotic arm holds potential. The
characteristics of producing large reduction ratios with high precision could off
er much in terms
of creating a technology for use by NASA.

This concept (Figure 18) offers the advantage of creating a usable technology. While
regular gears can put the hand in a general region, gear bearings can be used to operate the
sensor to cover a
ll the ground within a certain radius. The gear bearing appears to have ideal


characteristics for this type of application. The actual functions of this device could include an
arm that can drill, screw, and grab objects with a high amount of torque. A
disadvantage to this
concept is that it would require an extensive knowledge about robotic arms to create a state of
the art device. In this way, the concept may be slightly out of the capabilities of the group.

Figure 18. Robot Arm with high precision


Robotic Wrist and Hand

This concept was formed based on the idea that building an entire arm might be too
difficult to produce with a high level of quality. By focusing on a specific part of the arm, the
group would increase the likelihood of cr
eating a useable technology. Like the arm, the wrist
and hand would be capable of drilling, screwing, and grabbing. The group would focus on a
wrist designed to work with fine motor control. The development of an arm to produce the gross
motor control w
ould be left to other sources. Additionally, if possible, the multidimensional
Gear Bearing would be used in the wrist so that only one motor would produce the desired
motion of the wrist. Figure 19 shows the schematic for this design.


The basic principl
e of the robotic arm end
effector assembly involves a motor with an
optical encoder for the purpose of precision motion control that provides the input to the gear
bearing. The gear bearing causes the desired speed reduction and transfers the rotary motio
n to a
worm gear setup. The worm gear then causes rotation in the connecting gear. This connecting
gear has a linkage that causes translational or linear motion in the linkage plate of the grabber.

The linkage plate has five fingers placed symmetrically
around its circumference. The
schematic of the finger shows that there are two joints involved with two degrees of rotation, one
in each joint. A wire runs along the entire length of the fingers. If the wire is stretched by
moving the linkage plate back
, the fingers are stretched. When the wire is relaxed by moving the
linkage plate forward, the fingers curl or grab an object.

The forward/backward motion of the linkage plate is controlled by the direction of
rotation of the gear, which can be reversed v
ia a clutch mechanism. The use of a worm gear
assembly serves as a locking mechanism, ensuring that the motion of the fingers is controlled by
the motor only. The gear bearing leads to smooth precise motion of the grabber. This is useful
for high precis
ion and handling of delicate objects. Sensors can be added to allow for a sense of
“touch” to the fingers.

This concept offers the advantage of creating a usable device that is state of the art. It
allows the group to work on the concept of a multidimen
sional gear bearing without assuming
the risk of working on it alone. Its disadvantage is that the project does not have a unified goal.
By focusing some of the group’s attention on the multidimensional gear bearing, energy will be
diverted away from the

hand application.


Figure 19. Robot Hand/Wrist Assembly



Selection Matrix

(0 = worst, 5 = best)

Robotic Arm


Gear Bearing


Robotic Wrist and

Predicted Quality of End

1.)Referring to

2.)Referring to group’s

PKFo敦敲r楮g 瑯 瑩m攠

















NKFqo group memb敲s

OKFqo 乁pA











o楳k of 䙡c汵re






Ch慬汥lg攠瑯 䝲oup







NKFApp汩捡瑩ln 瑯 kApA

OKFApp汩捡瑩ln 瑯
oobo瑩捳 fndus瑲y

























Robotic Wrist and Hand


Multidimensional Gear Bearings


Robotic Arm




Existing Robotic Arm


Based on both the selection matrix and group discussions, the group has selected the
concept of the Robotic Wrist and Hand. The group members believe that the compromise of
multiple ideas will produce a project that holds mutual interest, a high level of
potential, and
quality results within the given time frame. Choosing the robotic hand and wrist allows the
group to pursue their interest in the multidimensional Gear Bearing without the heavy risk
involved if a functional part is not developed. Addition
ally, the project is small enough that the
combined efforts of the group members will result in a meaningful project from NASA’s
perspective. Thus, overall, there is greater interest in pursuing this project over the other four. A


patent search indicated

that no patents for robotic systems currently utilize gear bearing
technology. ( Thus, a robotic hand and wrist will be a unique project. This reason in
addition to feasibility concerns makes creating a robotic wrist and hand a sound decision.


Project Management

This section describes the activities pertaining to our project for each group member up to the
midterm presentation.

Pamela Carabetta:

Researched and compiled information on “gears and transmissions” and “gear bearings”
for the
background section of the report.

Wrote “executive summary” for report

Created “Rover” concept

Responsible for weekly progress reports

Compiled final draft of report

Finalized power point presentation

Marisa Jenkins:

Researched and compiled information o
n “robotic arms” for the background section of
the report.

Wrote “significance” for the report.

Collected information on the “Improving an existing robot” concept

Created “Multidimensional Gear Bearing” concept

Made “timeline”

Sarah Kovach

ed patents on existing technology

Wrote “problem description,” “mission statement,” “discussion,” “design specifications”

Created “Robotic Arm” concept

Created selection matrix


Compiled first draft of report

Created preliminary power point presentation

uja Mahashabde:

Researched and compiled information on “motors” for the background section of the

Assisted on website creation

Collaborated with Qi on “Robotic Wrist and Hand” concept

Organized references for report

Qi Yan:

Created website, respons
ible for website management

Assisted on research of motors for background

Collaborated with Anuja on “Robotic Wrist and Hand” concept

Organized references for report



Although not very specific, this timeline serves as a marker of general

goals per month
for the continuation of the project.

October 2003

Project Selection

Identifying customer needs

Background research

Patent search

Brainstorming and Project specifications

Midterm Project Presentation

November 2003

Design Specifications

Final Design Parameters

Initiate computations

Safety considerations

Costs research

December 2003

Final CAD drawings

Finish Computations

Finite Element Analysis

Final Project Presentation

January 2004

Begin Machining

Continue modifications if necessary


March 2004

Test prototype

Make corrections

May 2004

Make final prototype

Final Project Presentation


Works Cited

[1] Efunda Engineering Fundamental, October 2003, “ Nomenclature of Common Gear,”


[2] Juvinall, R.C., and Marshek, K.M., 2000,
Fundamentals of Machine Component Design
John Wiley and Sons, New York.

[3] Howstuffworks, Octobe
r 2003, “ How gears work,”

[4] Sharke, P., August 2002, “The Start of a New Movement,” Mechanical Engineering

[5] National Aeronautics and S
pace Administration, “Technical Support Package: Gear Bearings
and Gear
Bearing Transmissions,” NASA Tech Briefs GSC


[6] The Product Finder, October 2003, <>.

[7] Siemens, October 2003,

[8] Yim, Mark, “Modular Reconfigurable Robots in Space Applications,”

[9] Ulivi, Paolo, February 2, 2001, “Manipulators for Solar System Exploration,”


[10] National Aer
onautics and Space Administration, December, 1999, “Mars Polar Lander/

Deep Space2,” <

[11] Bonitz, Robert G. et. al., June 1999,
“MVACS Robotic Arm,” Submitted to:
Journal of

Geophysical Research, <

[12] National Aeronautics and Space Administration, 2001, “Educational Brief: Humans and

Robots,” EB


[13] National Aeronautics and Space Administration, 2003, “Space Station Assembly:

Elements: Canadarm2 and the Mobile Servicing System,” Curator: Kim Dismukes,


[14] Boeing, “International Space Station: Canadian Mobile Servicing System,”




[15] HobbyTron, October 2003, <