Download - Grippers for Space Locomotion
ICRA ’08 Space Robotics Workshop:Orbital Robotics
Grippers for Space Locomotion
Rick WagnerNorthrop Grumman Corporation
May 20, 2008
1
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Contents
• Introduction• Related Work• General Locomotion
Requirements• Gripper Types
• Mechanical• Sticky• Electrostatic
• Algorithmic Considerations
• Conclusion
2
ICRA ’08 Space Robotics Workshop: Orbital Robotics
IntroductionLocomotion in microgravity
Free flying: cold gas jet propulsion, 3-axis stabilizationRail and wheelsWheels or treadsLegged locomotion: walking, leapingTethers
Manipulation in microgravityGripping (grasping) of tools, work objects, retrievalHandoff
Handoff of tools, materialsAstronaut assistanceAssisting other robots
Reacting forcesTool forcesComponent installationObject stabilizationVehicle ACS/RCS or delta-V propulsion dynamic forces
JSC’s mini AERcam
3
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Related Work
JPL LEMUR I and LEMUR II
CMU Skyworker
Northrop Grumman/JPL AWIMR
JSC Robonaut
4
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Related Work (cont.)Eurobot (ESA)
“Autonomous locomotion on ISS modules: the robot system is capable to walk on sections of real-scale modules of the ISS using handrails
Dexterous manipulation of ISS hardware: an operator, equipped with some human-machine interface, can command the robot to perform highly dexterous handling of objects.”
From http://www.esa.int/TEC/Robotics/SEMUC68LURE_0.html and http://www.esa.int/esaHS/SEMI9TNSP3F_iss_0.html
5
ICRA ’08 Space Robotics Workshop: Orbital Robotics
General Locomotion RequirementsSpace microgravity “Prime Directive”
Maintain a positive grip on the space vehicle at all times!Even on a tether, a drifting robot is not a good thing to have near a space vehicleA free flyer walking robot hybrid is a possibility
Even better might be free flyer and walker docking collaboration
Do not damage the host vehicleSpeed of locomotion is usually importantSticky wheels or tracks are possible, but are not in the scope of this workshop presentation
Rails and wheels are also out of scopeWe assume limbed (walking) locomotion
A general microgravity walking algorithm isStart with all feet gripping the structureUngrip one of more feet and relocate and regrip on the structureChange the pose of the robot body in the direction of locomotionRepeat
6
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Mechanical GripperExtensive literature exists on mechanical grippers for terrestrial useSpace applications issues include thermal extremes, mechanism lubrication, and contact friction materials
LEMUR gripper (JPL)
Skyworker gripper (CMU)
7
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Gripping Astronaut HandrailsStandard astronaut handrails are specified in a JSC document
Width, thickness, height above vehicle surface, and distance between standoffs (see references)
Handrails are installed on the exterior of space vehicles wherever astronaut extra-vehicular activity (EVA) is plannedAstronaut Story Musgrave told me that even though they were designed for a “power grip,” astronauts actually use just their fingertips in traversing the handrails
Eurobot handrail gripper (ESA)
Friction clamp concept
8
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Hybrid Mechanical and Sticky Gripper Robot Concept
Space service robotThe long link is 1 meterRepair or assembly tasks
9
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Dry adhesion (gecko feet)Van der Waals forces of intermolecular attractionGood attachment strength, about 10 N/cm2 for geckosWorks in atmosphere and vacuum, low temperature and low humidityCompliant micro/nano-hairs adaptable to a multitude of smooth and rough surfacesCurrently at NASA technology readiness level (TRL) 3Text and pictures adapted from a CMU document
CMU sticky foot work by Dimi Apostolopoulos and Metin SittiSticky feet are also under development at UC Berkeley
10
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Sticky Foot for AWIMRCMU (Dimi Apostolopoulos and Metin Sitti) provided material samples and various prototypes of sticky feet for the AWIMR (Automated Walking Inspection and Maintenance Robot) project in 2005JPL (Brett Kennedy) installed a prototype sticky foot on LEMUR II (at right)
11
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Stick Foot Material TestsIn the summer of 2005 we received samples of sticky material for test from CMU
White siliconeClear polydimethylsiloxane (PDMS)PDMS with circular grooves
We used various sizes of circular punches to create test specimensWe used a weight and pulley apparatus to apply preload and pulloff forces
12
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Sticky Silicone on Aluminum TestsSilicone on bare aluminum results
Better than 2:1 pull-off to preload force
Aluminum Creep Test Results
We performed two trials with 2 kiloPascal preload and 0.5 kiloPascal pull load and measured the time for the aluminum plate to pull off:
75 seconds50 seconds
Storage test showed no degradation after two weeks
5/8 Silicone on Bare Aluminum
0.0
500.0
1000.0
1500.0
2000.0
2500.0
3000.0
3500.0
4000.0
0.0 2000.0 4000.0 6000.0 8000.0 10000.0
Preload pressure (Pascals)
Pull-
off p
ress
ure
(Pas
cals
)
Run OneRun TwoRun ThreeRun FourAverage
5/8 inch white silicone on bare aluminum. September 29, 2005
1.0 Inch White Silicone on Bare Aluminum
0
500
1000
1500
2000
2500
0 500 1000 1500 2000 2500 3000 3500 4000
Preload Pressure (Pascals)
Pul
l-off
Pre
ssur
e (P
asca
ls)
Run OneRun TwoRun ThreeRun FourRun FiveAverage
13
ICRA ’08 Space Robotics Workshop: Orbital Robotics
JSC-Supplied Space Shuttle TileMuch higher pull-off forces were measured for the shuttle tile than for bare aluminumIn general, smoother surfaces stick better than rough surfaces
The shuttle tile had a shiny surface finish
No visible residue was evident with any of the CMU-supplied sticky materials we tested
5/8 Inch White Silicone on Shuttle Tile
0.0
2000.0
4000.0
6000.0
8000.0
10000.0
12000.0
0.0 2000.0 4000.0 6000.0 8000.0 10000.0
Preload Pressure (Pascals)
Pul
loff
Pre
ssur
e (P
asca
ls)
Run OneRun TwoRun ThreeAverage
14
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Polydimethylsiloxane (PDMS)The PDMS does not seem to have as much absolute adhesion pressure as the white silicone, but the preload sensitivity is highPull-off pressure on Kapton for the PDMS sample was much higher than for bare aluminum
7/8 Inch PDMS on Bare Aluminum
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0 1000 2000 3000 4000 5000 6000
Preload Pressure (Pascals)
Pul
l-off
Pre
ssur
e (P
asca
ls)
Run OneRun TwoRun ThreeRun FourRun FiveAverage
7/8 PDMS on Kapton
0.0
500.0
1000.0
1500.0
2000.0
2500.0
3000.0
3500.0
0.0 1000.0 2000.0 3000.0 4000.0 5000.0 6000.0
Preload pressure (Pascals)
Pul
l-off
pres
sure
(Pas
cals
)
Run OneRun TwoRun ThreeRun FourAverage
1.0 PDMS on Shuttle Tile
0.0
200.0
400.0
600.0
800.0
1000.0
1200.0
1400.0
1600.0
0.0 500.0 1000.0 1500.0 2000.0 2500.0 3000.0 3500.0 4000.0
Preload pressure (Pascals)
Pul
l-off
pres
sure
(Pas
cals
)
Run OneRun TwoRun ThreeRun FourAverage
15
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Sticky Foot Material Test ConclusionGeneral microgravity locomotion is feasible with sticky foot grippersMay need to avoid areas of thermal insulation to avoid damaging
Pull-off-to-Preload Ratios for Shuttle Tile
0
1
2
3
4
5
6
7
8
9
10
0 2000 4000 6000 8000 10000
Preload Pressure (Pa)
Ratio
1.0 PDMS1.0 Silicone5/8 Silicone-ST5/8 Silicone (9/29)
Pull-off-to-Preload Ratio for Bare Al
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Preload Pressure (Pa)
Ratio
1.0 PDMS7/8 PDMS1.0 Silicone5/8 Silicone (8/16)5/8 Silicone (9/29)1.0 Sticky Stuff3/4 Silicone
Pull-off-to-Preload Ratios for Kapton
0
1
2
3
4
5
6
7
8
9
10
0 1000 2000 3000 4000 5000 6000 7000 8000
Preload Pressure (Pa)
Ratio
1.0 Grooved PDMS1.0 PDMS7/8 PDMS3/4 Silicone1.0 Sticky Stuff
16
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Electrostatic gripper for space use
Conceptual model of an electrostatic attachment device. Left: lower and upper views of an assembled foot. Right: Exploded view.
17
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Electrostatic FootCapacitor plate force equation
We measured somewhat lower forces than the equation suggests
Surface charge buildup effects were significantHumidity level was controlled at 50% RHKapton punch-through voltage is about 3,000 volts per milli-inchTesting in vacuum is indicated for future work
F = Aε0V2/d2 F is force of attraction in NewtonsA is area of the plates in square metersV is voltaged is the distance between the plates in metersε0 is free-space permittivity (8.55 x 10-12)
18
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Electrostatic Foot TestingTest Objectives
Evaluate the electrostatic model (area, voltage, force, and dielectric thickness relationship) for real materialsEstablish the foot’s sensitivity to surface irregularitiesEvaluate the suitability of the foot to microgravity locomotion in a space environment
Test SetupFixed electrostatic foot with suspended test surface at ground potentialWe used two different design volt meters for verification
PSC8 DC 13.8 VPower Supply
HV350 HV Power Supply
AC
Red Red
Black Black
Green
Line in
Foot
100 MegOhm
100 MegOhm
Load
Ground Bus
19
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Electrostatic Foot Test (four inch foot)
Four inch diameter octagonal foot wrapped with half mil Kapton Black kapton ground plane
Solar cells on ground plane
20
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Electrostatic Foot Tests (2.5 inch foot)
2.5 inch diameter circular foot wrapped with 2 mil Kapton
2.5 inch diameter circular foot test with solar cells Aluminized Kapton ground plane
21
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Electrostatic Foot Results
Summary of Data for Two Mil Kapton
0
100
200
300
400
500
600
700
800
0 500 1000 1500 2000 2500 3000
Voltage
Pres
sure
(Pas
cals
) Bare GroundPlane
Kapton GroundPlane
Summary of Test DataFoot with two mil KaptonIncludes both 4-inch and 2.5 inch foot data
22
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Electrostatic Foot ConclusionShear capability was found to be an order of magnitude greater than normal pull force
This is a known effect as a result of polarization of surface molecules
We tested the analytical electrostatic model with real materials and found that an empirical model is more suitable for use as a guide to designWe established the foot’s sensitivity to surface irregularities andevaluated the suitability of the foot to microgravity locomotion and other uses in a space environmentFuture work
Vacuum TestingPolarity Reversal TestingBi-Polar Device Testing(at right, by Hobson Lane,Northrop Grumman)
23
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Algorithmic ConsiderationsSticky feet require some preload which is reacted in tension by the other feet
With the tripod gait in a hexapod robot, preload reactions are roughly equal to the preload: pull-off to preload ratio must be greater than 1.0Pairwise opposed gait gives twice the pull-off margin during foot preloading (but with 2/3 the walking speed)
Sticky foot preload can be refreshed by a force pumping action of the legs in a hexapod robot
Push down triples alternately to overcome relaxation creepForce sensing at the feet is probably necessary to detect marginal adhesion during locomotion
Applies to both sticky and electrostatic feet
24
ICRA ’08 Space Robotics Workshop: Orbital Robotics
Space Locomotion Gripper ConclusionFor reacting high tool forces, mechanical grippers are indicated
Requires astronaut handrails or other vehicle design features for gripping
For minimal complexity for locomotion, sticky foot grippers are indicated
May need to avoid thermal insulation and solar cells to avoid damage
For locomotion on thermal insulation and solar cells, electrostatic grippers are a possibility
Much work remains to be done, e.g., time dependent surface charge phenomena need investigating (especially in vacuum)Robustness of dielectric materials needs to be demonstratedVacuum may improve performance (but may lead to unanticipated problems, e.g. corona)Bipolar and/or polarity reversing devices are interesting
25
ICRA ’08 Space Robotics Workshop: Orbital Robotics
ReferencesJSC 26626A, Extravehicular Activity (EVA) Hardware Generic Design Requirements Document, EVA and Crew Equipment Projects Office, May 1995The Challenges of Extra-Vehicular Robotic Locomotion Aboard Orbiting Spacecraft, Fredrik Rehnmark, Robert O. Ambrose, and Michael Goza, proceedings of the 2004 IEEE International Conference on Robotics and Automation (ICRA), New Orleans, LA, April 2004, http://ieeexplore.ieee.org/iel5/9126/29025/01308135.pdfEurobot End-Effectors, S. Michaud et al., Proceedings of the 8th
ESA Workshop on Advanced Space Technologies for Robotics and Automation, November 2004, http://robotics.estec.esa.int/ASTRA/Astra2004/Papers/astra2004_C-03.pdfBrett Kennedy, et al., “Limbed Excursion Mechanical Utility Rover (LEMUR),” JPL Technical Report, Jet Propulsion Laboratory, California Institute of Technology, 2003Mark Showalter, et al., “Development and Comparison of Gait Generation Algorithms for Hexapedal Robots Based on Kinematics with Considerations for Workspace,” ASME IDET/CIE, August 2008 (to appear)