Datasets:

ai-aerospace

/

ams_data_full_2000-2020

like 0

packaging. It provides a 1:l gear ratio. This gearbox was procured with a flanged input interface and a hollow shaft output interface. The flanged input interface provides a bolt-on interface for the motor with misalignments handled by an integrated flex coupling. The output of the Tandler gearbox drives a Sumitorno Servo-Match Gearbox which is a low-backlash precision cycloid torque multiplier designed for heavy duty industrial robotic applications. The Sumitomo gearbox has a reduction drive ratio of 89:l. The Sumitomo input interface is a hollow shaft of the same diameter of the Tandler gearbox. A precision manufactured shaft with machined keyway provides the drive interface and a precision machined adapter plate provides the housing interface and alignment between the two gearboxes. A machined drive shaft is coupled to the output half of the Sumitomo drive. The drive shaft is supportea by a pair of matched angular-contact ball bearings. The bearing set is sized to react to the cyclic bending, side and axial loads resulting from aerodynamic and oscillatory dynamic forces. The drive shaft with the bearing set effectively decouples the Sumitomo drive from all but rotational loads and provides an interface for sting attachment. A 16-bit encoder measures the rotation of the drive shaft which provides the sting’s angular position. It is used in the closed-loop motion control. All of the mechanical components, except the servo motor are mounted inside a cylindrical, titanium housing. Titanium was chosen for its strength and lightweight properties. The additional weight of the FOS, compared to the weight of the existing model support sting and mounting hardware, decreases the existing model support system model weight capability. Attaching the FOS to the existing model support system required a rigid interface that could tolerate high side and torsional loading. A tapered dovetail with bolts was chosen for this interface. The existing sting adapter was also modified as shown in Figure 9 to match this same dovetail interface so both the FOS and existing sting can be quickly interchanged for various testing needs. A removal bolt is also designed into the interface to force the tapered dovetail joint apart for FOS and sting removal. t Figure 9. Comparison of existing to modified sting interface The expected system performance for this design for a 200-lb model is shown in Figure 10. By generating and plotting the various performance curves along with the desired theoretical curve from Figure 2, it was determined that the first limiting loading factor for roll oscillation testing is the model support system; and the next factor is due to the load limits of the balance used for measuring aerodynamic loads. The FOS 395

Document

AMS_2006.pdf

can conceivably exceed the load capability of the test infrastructure. It should also be observed that the drive system is not limited by its speed but by the maximum motor torque. I h 5 E.- t?: E: N- c 4- a. u- LL c- a- 0- .e 3 - c - -. .- 2: 1- Similarly, the expected system performance for a 90-lb model is shown in Figure 11. However, for the much smaller model class, the entire theoretical frequency versus amplitude curve (Figure 2) is achievable. The limiting factor for the system is due to motor speed except at frequency greater than 3.4 Hz, where the limiting factor is the motor torque. \ -Desired Theoretical Curve *Limit Due to Motor Speed --)(-Limit Due to Motor Torque *Balance Limit in Roll Axis - 200 Ib Model - FF-IO Balance (141 N-m in Roll) - FA25, FS 00 Gearboxes - 82048 Kollmorgen Motor 0 5 8 10 12 14 16 18 20 22 25 30 35 40 45 50 Oscillation Angle (degrees) Figure 10. Predicted performance curve for a 200-lb model System Integration and Testing System integration testing was done outside the tunnel in a controlled environment by installing inertia bars on the FOS to verify performance under simulated loading conditions. Various runs were made with varying amplitude and frequencies gradually approaching the limitations of the hardware. 396

Document

AMS_2006.pdf

7 6- 5- h. v 2- 5 4- h- 0- m LL s- m- 2; .I! 3 CI - - .- 0- 2- s- 1- 0 Figure 11. Predicted performance curve for a 90-lb model +Limit Due to Motor Speed *Limit Due to Motor Torque +Balance Limit in Roll Axis +Model Support Limit in Roll - 90 Ib model - FF-IO Balance - FA25, FS 00 Gearboxes - 82048 Kollrnorgen Motor - The hardware ran well at low frequencies and low amplitudes. However, degradation in performance occurred at higher amplitude profiles. As one of the trouble-shooting methods, a torque wrench was used to record the required moment at the motor input interface. The test showed that the torque varied significantly per revolution of motor. A probable cause for the variation is due to an internal alignment problem at one of the drive component interfaces. To isolate and locate the problem, the mechanical subassembly was disassembled and all the individual components were inspected. No anomalies were discovered and all parts were machined within tolerances as specified. The mechanical drive system was carefully reassembled with emphasis on even and opposite torqueing of all bolt patterns and with careful inspection to assure that no contamination or irregularities were present on the critical interfaces. The reassembled unit was then tested again to determine torque consistency per motor revolution. The subsequent test showed a slightly increased overall torque as compared to the previously fowest measured torque but with almost no variation per revolution. Once reassembled, verification tests of the FOS demonstrated the expected performance and system safety. Figure 12 shows a response to a sinusoidal command input of 0.5 Hz at 5-degree amplitude. As shown, the system tracked the roll command to within 3 percent of amplitude. Figure 13 shows the encoder output from an excerpt of a Schroeder sweep command input. This demonstrates the new capabilities as envisioned for the new FOS. Integration and testing have verified system operation and safety. The system still needs to be evaluated to characterize the performance over the full operational envelope. Validation testing with known inertias 397

Document

AMS_2006.pdf

is also required, and is underway to confirm system implementation and test techniques. Certification and approval for wind-tunnel operation is being processed currently. I -Roll Command Time (Seconds) 2.00 1.50 1 .oo 2 0.50 h 8 CT ; 0.00 s c I .- 5. -0.50 E -1 .oo -1.50 -2.00 a Time (Seconds) I - - - - Encoder Feedback I Figure 12. Typical encode: output signal Figure 13. An excerpt from a Schroeder sweep 398

Document

AMS_2006.pdf

Conclusion A new Forced Oscillation System (FOS) has been desigfied and built that will provide new capabilities and flexibility for conducting dynamic derivative studies. The new design is based on a tracking principle where a desired motion profile is achieved via a fast closed-loop positional controller. The motion profile for the tracking system is numerically generated and thus not limited to sinusoidal motion. This approach permits non-traditional profiles such as constant velocity and Schroeder sweeps. Also, the new system permits changes in motion parameters including nominal offset angle, waveform and its associated parameters such as amplitude and frequency. Most importantly, the changes may be made remotely without halting the FOS and the tunnel. System integration and testing has verified design intent and safe operation. Currently the FOS is being validated for wind-tunnel operations and aerodynamic tests. Once complete, the system is a major enhancement to forced oscillation studies. The productivity gain from the motion profile automation will shorten the testing cycles needed for control surface and aircraft control algorithm development. The new motion capabilities also will serve as a test bed for researchers to study and to potentially improve and/or alter future forced oscillation testing techniques. References 1. Murphy, Patrick C. "Estimation of Aircraft Unsteady Aerodynamic Parameters from dynamic wind Tunnel Testing", AlAA 2007-4016. 2. Etkin, Bernard and Reid, Lloyd D. "Dynamics of flight - Stability and Control", 3rd Edition, John Wiley & Son, Inc. 1996. 3. Armstrong, Richard W. "Load to Motor Inertia Mismatch: Unveiling the Truth", Drives and Controls Conference, Telford England, 1998. 399

Document

AMS_2006.pdf

I REPORT DOCUMENTATION PAGE 6. AUTHOR(S) Form Approved I OM6 NO. 0704-0188 5b. GRANT NUMBER 5c. PROGRAM ELEMENT NUMBER 5d. PROJECT NUMBER 01- 05 - 2006 I Conference Publication I 4. TITLE AND SUBTITLE I 5a. CONTRACT NUMBER 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) NASA Langley Research Center Hampton, VA 2368 1-2 199 9. SPONSORlNGIMONlTORlNG AGENCY NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANEATION REPORT NUMBER L- 19245 IO. SPONSORIMONITORS ACRONYM(S) 17. LIMITATION OF 18. NUMBER OF PAGES ABSTRACT 16. SECURITY CLASSIFICATION OF: a. REPORT b. ABSTRACT c. THIS PAGE U U U uu 418 Boesiger, Edward A. (Compiler) 19a. NAME OF RESPONSIBLE PERSON 19b. TELEPHONE NUMBER (Incbde area code) STI Help Desk (email: help@sti.nasa.gov) (301) 621-0390 5e. TASK NUMBER National Aeronautics and Space Administration Washington, DC 20546-0001 NUMBER(S) NAS4lCP-2006-2 14290 12. DlSTRlBUTlONIAVAlLABlLlTY STATEMENT Unclassified - Unlimited Subject Category 37 Availability: NASA CAS1 (301) 621-0390 13. SUPPLEMENTARY NOTES Edward A. Boesiger: Lockheed Martin S ace Systems Company, Sunnyvale, California NASA Langley point of contact: James E! Wells An electronic version can be found at http:l/ntrs.nasa.gov 14. ABSTRACT The Aerospace Mechanisms Symposium (AMS) provides a unique forum for those active in the design, production and use of aerospace mechanisms. A major focus is the reporting of problems and solutions associated with the development and flight certification of new mechanisms. Organized by the Mechanisms Education Association, the National Aeronautics and Space Administration and Lockheed Martin Space Systems Company (LMSSC) share the responsibility for hosting the AMs. Now in its 38th symposium, the AMS continues to be well attended, attracting participants from both the U.S. and abroad. The 38th AMs, hosted by the NASA Langley Research Center in Williamsburg, Virginia, was held May 17- 19, 2006. During these three days, 34 papers were prcsented. Topics included gimbals, tribology, actuators, aircraft mechanisms, dcployrnent mcchanisrns, releasc mechanisms, and test cquipment. Hardware displays during the supplier exhibit gave attendees an opportunity to meet with devclopers of current and future mechanism components. 15. SUBJECT TERMS Actuators; Bearings; Deployment; Design; Gimbals; Mechanisms; Release; Test; Tribology

Document

AMS_2006.pdf

NASA/CP-2012-217653 41st Aerospace Mechanisms Symposium Compiled/Edited by: Edward A. Boesiger Proceedings of a sy mposium held at Hilton Pasadena, CA Hosted by the Jet Propulsion Laboratory and Lockheed Martin Space Systems Company Organized by the Mechanisms Education Association May 16-18, 2012 May 2012

Document

AMS_2012.pdf

NASA STI Program ... in Profile Since its founding, NASA has been dedicated to the advancement of aeronautics and space science. The NASA scientific and technical information (STI) program plays a key part in helping NASA maintain this important role. The NASA STI program operates under the auspices of the Agency Chief Information Officer. It collects, organizes, provides for archiving, and disseminates NASA’s STI. The NASA STI program provides access to the NASA Aeronautics and Space Database and its public interface, the NASA Tech nical Report Server, thus providing one of the largest collections of aeronautical and space science STI in the world. Results are published in both non-NASA channels and by NASA in the NASA STI Report Series, which includes the following report types:  TECHNICAL PUBLICATION. Reports of completed research or a major significant phase of research that present the results of NASA Programs and include extensive data or theoretical analysis. Includes compila- tions of significant scientific and technical data and information deemed to be of continuing reference value. NASA counterpart of peer-reviewed formal professional papers but has less stringent limitations on manuscript length and extent of graphic presentations.  TECHNICAL MEMORANDUM. Scientific and technical findings that are preliminary or of specialized interest, e.g., quick release reports, working papers, and bibliographies that contain minimal annotation. Does not contain extensive analysis.  CONTRACTOR REPORT. Scientific and technical findings by NASA-sponsored contractors and grantees.  CONFERENCE PUBLICATION. Collected papers from scientific and technical conferences, symposia, seminars, or other meetings sponsored or co-sponsored by NASA.  SPECIAL PUBLICATION. Scientific, technical, or historical information from NASA programs, proj ects, and missions, often concerned with subjects having substantial public interest.  TECHNICAL TRANSLATION. English-language translations of foreign scientific and technical material pertinent to NASA’s mission. Specialized services also include organizing and publishing research results, distributing specialized research announcements and feeds, providing help desk and personal search support, and enabling data exchange services. For more information about the NASA STI program, see the following:  Access the NASA STI program home page at http://www.sti.nasa.gov  E-mail your question via the Internet to help@sti.nasa.gov  Fax your question to the NASA STI Help Desk at 443-757-5803  Phone the NASA STI Help Desk at 443-757-5802  Write to: NASA STI Help Desk NASA Center for AeroSpace Information 7115 Standard Drive Hanover, MD 21076-1320 This page is required and contains approved text that cannot be changed.

Document

AMS_2012.pdf

NASA/CP-2012-217653 41st Aerospace Mechanisms Symposium Compiled/Edited by: Edward A. Boesiger Proceedings of a symposium held at Hilton Pasadena, CA Hosted by the Jet Propulsion Laboratory and Lockheed Martin Space Systems Company Organized by the Mechanisms Education Association May 16-18, 2012 May 2012

Document

AMS_2012.pdf

Acknowledgement The high quality of this symposium is the result of the work of many people, and their efforts are gratefully acknowledged. This extends to vol untary members of the Aerospace Mechanisms Organizing Committee representing the eight par ticipating NASA Centers, co-sponsor Lockheed Martin Space Systems Corporation, and the European Space Agency. Appreciation is also extended to the session chairs, the authors, and particularly the personnel at the Jet Propulsion Laboratory responsible for the symposium arrang ements and publication of these proceedings. A sincere thank you is well deserved by the Mechanisms Education Association officers, particularly Ed Boesiger, who are responsible for the year-to-year management of the AMS including editing of the papers and production of the program. Special thanks to: Donald Sevilla, JPL Host Chairman Louise Jandura, JPL Co-Host Chairman Stuart Lowenthal, MEA General Chairman, Lockheed Martin SS Edward Boesiger, MEA Operations Chairman, Lockheed Martin SS Monica King, JPL Symposium Organization Support Kris Katagiri, JPL Symposium Organization Support Ken Gowey, JPL Graphics Services , Technical Information Section Available from: NASA Center for AeroSpace Information 7115 Standard Drive Hanover, MD 21076-1320 443-757-5802 This report is also available in electronic form at http://www.sti.nasa.gov/ and http://ntrs.nasa.gov/

Document

AMS_2012.pdf

iii PREFACE The Aerospace Mechanisms Symposium (AMS) provides a unique forum for those active in the design, production and use of aerospace mechanisms. A major focus is the reporting of problems and solutions associated with the development and flight certification of new mechanisms. Organized by the Mechanisms Education Association, responsibility for hosting the AMS is shared by the National Aeronautics and Space Administration and Lockheed Martin Space Systems Company (LMSSC). Now in its 41 st symposium, the AMS continues to be well attended, attracting participants from both the U.S. and abroad. The 41 st AMS, hosted by the Jet Propulsion Laboratory (JPL) in Pasadena, California, was held May 16, 17 and 18, 2012. During these three days, 38 papers were presented. Topics included gimbals and pos itioning mechanisms, comp onents such as hinges and motors, CubeSats, tribology, and Mars Science Laboratory mechanisms. Hardware displays during the supplier exhibit gave attendees an opportunity to meet with developers of current and future mechanism components. This publication was prepared by the Jet Propulsion Laboratory, California Institute of Technology, under a contract with the National Aeronautics and Space Administration. A portion of this work was prepared outside NASA, JPL, and California Institute of Technology. Any views and opinions expressed in this outside work do not necessarily state or reflect those of NASA, JPL, or the California Institute of Technology. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not constitute or imply its endorsement by the United States Government or the Jet Propulsion Laboratory, California Institute of Technology. © 2012. All rights reserved.

Document

AMS_2012.pdf

v CONTENTS Symposium Schedule ............................................................................................................ .................. viii Symposium Organizing and Advisory Committees .................................................................................. xii Development of Brushed and Brushless DC Motors for use in the ExoMars Drilling and Sampling Mechanism ..................................................................................................................... ............................ 1 Robin Phillips, Massimo Palladino & Camille Courtois Rotary Percussive Sample Acquisition Tool (SAT): Hardware Development and Testing ......................17 Kerry Klein, Mircea Badescu, Nicolas Haddad, Lori Shiraishi & Phillip Walkemeyer Lock & Release Mechanism for the CHOMIK Penetrator Device and its Tribological Properties ...........29 Marcin Dobrowolski & Jerzy Grygorczuk Harmonic Drive™ Gear Material Selection and Life Testing....................................................................39 Jeffrey Mobley & Jonathan Parker New Supplier – Hardware Duplication – Some Pitfalls ........................................................................... .53 Edwin Joscelyn Development of the Vibration Isolation System for the Advanced Resistive Exercise Device ................67 Jason Niebuhr & Richard Hagen Passive Thrust Oscillation Mitigation for the CEV Crew Pa llet System ...................................................81 Matthew Sammons, Cory Powell, Joe Pellicciotti, Ralph Buehrle & Keith Johnson The Damper Spring Unit of the Sentinel 1 Solar Array .......................................................................... ..97 Frans Doejaaren & Marcel Ellenbroek Ultra-low-weight Rotary Actuator for Operation on Mars and Pin Puller Mechanism Based on a Novel Shape Memory Alloy Technology ...........................................................................................................111 Nestor Nava, Marcelo Collado, Francisco Alvarez, Ramiro Cabás, Jose San Juan, Sandro Patti & Jean-Michel Lautier Design and Performance of the Telescopic Tubular Mast .....................................................................127 Mehran Mobrem & Chris Spier Development of Variable Reluctance Resolver for Position Feedback ..................................................141 Gregory Leibovich & Sara Senanian FeF 3 Catalytic Influence on PFPE Lubricants Lifetime under Loaded Conditions .................................147 Lionel Gaillard, Antoine Mariot, Catalin Fotea & Roland Holzbauer Trade Studies for a High Torque Density Planetary Gearbox ................................................................155 Jeffrey Mobley Single Motion Actuated Shape Memory Alloy Coupling ...... ...................................................................161 Alberto Perez, J. Newman & M. Romano Development and Testing of a High Compact Stepper Motor Mechanism ............................................167 Jörg Schmidt & Greg Wright

Document

AMS_2012.pdf

vi Cryogenic Temperature Testing of NEA Fuse Wire Mechanism ...........................................................175 Edwin Vega & Geoff Kaczynski Developmental Testing of Electric Thrust Vector Control Systems for Manned Launch Vehicle Applications .................................................................................................................. ..........................181 Lisa Bates & David Young Development of High Temperature High Current Contact Technology in Slipring Assemblies for the BepiColombo MPO & MTM Spacecrafts ............................................................................................. ...195 Fabrice Rottmeier, Mikaël Krummen & Mickaël Miler Lessons Learned to Avoid Coax Cable Failure in Moving Mechanical Mechanisms .............................211 Sheah Pirnack DLR’s Dynamic Actuator Modules for Robotic Space Applications .......................................................223 Armin Wedler, M. Chalon, K. Landzettel, M. Görner, E. Krämer, R. Gruber, A. Beyer, H-J. Sedlmayr, B. Willberg, W. Bertleff, J. Reill, M. Grebenstein, M. Schedl, A. Albu-Schäffer & G. Hirzinger Mars Science Laboratory Rover Integrated Pump Assembly Bellows Jamming Failure .......................239 Michael Johnson, Joel Johnson, Gajanana Birur, Pradeep Bhandari & Paul Karlmann Development of the Descent Brake Mechanism for the Mars Science Laboratory ...............................253 David Dowen, Jeff Moser & Jeff Mobley Mars Science Laboratory Differential Restraint: The Devil is in the Details ..........................................263 Elizabeth Jordan Mars Science Laboratory’s Dust Removal Tool ................................................................................... ..279 Kiel Davis, Jason Herman, Mike Maksymuk, Jack Wilson, Philip Chu, Kevin Burke, Louise Jandura & Kyle Brown A Zoom Lens for the MSL Mast Cameras: Mechanical Design and Development ................................293 Daniel DiBiase, Jason Bardis & Rius Billing Wet Chemistry Automated Sample Processing System (WASP) ..........................................................311 Juancarlos Soto, James Lasnik & Shane Roark & Luther Beegle Refinement of a Low-Shock Separation System ................................................................................... .329 Chuck Lazansky Ares I Linear Mate Umbilical Plate and Collet ................................................................................. .......345 William Manley, Gabor Tamasy & Patrick Maloney GMI Spin Mechanism Assembly Design, Development, and Test Results ............................................359 Scott Woolaway, Mike Kubitscheck, Barry Berdanier, David Newell, Chris Dayton & Joseph Pellicciotti Lessons Learned from the TIRS Instrument Mechanisms Development ...............................................373 Jason Budinoff, Richard Barclay, James Basl, Konrad Bergandy, Thomas Capon, Bart Drake, Michael Hersh, Chris Hormann, Edwin Lee, Adam Matuszeski, Armani Nerses, Kenneth Pellak, Kermit Pope, Joseph Schepis & Ted Sholar

Document

AMS_2012.pdf

vii Resolution for Fretting Wear Contamination on Cryogenic Mechanism ................................................399 Charles Clark Design and Manufacturing Considerations for Shockproof and Corrosion-Immune Superelastic Nickel-Titanium Bearings for a Space Statio n Application ................................................407 Christopher DellaCorte & Walter Wozniak Wear of Steel Ti6Al4V in Vacuum ............................................................................................... ...........424 Timothy Krantz & Iqbal Shareef Angular Runout Test Setup for High-Precision Ball Bearings ................................................................439 Scott Miller, Jonathan Wood & Stuart Loewenthal LightSail-1 Solar Sail Design and Qualification ............................................................................... .......451 Chris Biddy & Tomas Svitek A Novel Release Mechanism Employing the Principle of Differential Coefficients of Thermal Expansion .........................................................................................................465 Clint Apland, David Persons, David Weir & Michael Marley A Nichrome Burn Wire Release Mechanism for CubeSats ....................................................................479 Adam Thurn, Steve Huynh, Steve Koss, Paul Oppenheimer & Sam Butcher, Jordan Schlater, Peter Hagan Antenna Deployment Mechanism for the Cubesat Xatcobeo. Lessons, Evolution and Final Design ....489 Jose Antonio Vilán Vilán, Miguel López Estévez & Fernando Aguado Agelet

Document

AMS_2012.pdf

viii SYMPOSIUM SCHEDULE WEDNESDAY, 16 MAY 2012 7:30 Wednesday Presenters' Breakfast - San Marino Room, Hilton Pasadena 8:00 CHECK-IN AND REFRESHMENTS - Skylight Arcade, Hilton Pasadena 8:30 INTRODUCTORY REMARKS - International Ballroom, Hilton Pasadena Don Sevilla, Host Chairman, Jet Propulsion Laboratory, Pasadena, CA Stuart Loewenthal, General Chairman, Lockheed Martin Space Systems, Sunnyvale, CA Dr. Charles Elachi, Director, Jet Propulsion Laboratory, Pasadena, CA 9:00 SESSION I – ACTUATOR COMPONENTS AND THEIR USE IN DRILLS Lionel Gaillard, Session Chair ESA/ESTeC, Noordwijk, The Netherlands  Development of Brushed and Brushless DC Motors for use in the ExoMars Drilling and Sampling Mechanism Robin Phillips, Maxon Motor AG, Sachseln, Switzerland, et al  Rotary Percussive Sample Acquisition Tool (SAT): Hardware Development and Testing Kerry Klein, Jet Propulsion Laboratory, Pasadena, CA, et al  Lock & Release Mechanism for the CHOMIK Penetrator Device and its Tribological Properties Marcin Dobrowolski, Space Research Centre of the Polish Academy of Sciences, Warsaw, Poland, et al  Harmonic Drive™ Gear Material Selection and Life Testing Jeffrey Mobley, Sierra Nevada Corporation, Durham, NC, et al  New Supplier – Hardware Duplication – Some Pitfalls Edwin Joscelyn, Aeroflex, Hauppauge, NY 11:30 LUNCH Lunch for AMS Attendees in the California Ballroom, Hilton Pasadena 12:30 SESSION II – HINGES & DAMPERS Michael Kubitschek, Session Chair Ball Aerospace & Technologies Corp., Boulder, CO  Development of the Vibration Isolation System for the Advanced Resistive Exercise Device Jason Niebuhr, Apogee Engineering, Colorado Springs, CO; et al  Passive Thrust Oscillation Mitigation for the CEV Crew Pallet System Matthew Sammons, ATK Aerospace Systems, Beltsville, MD; et al  The Damper Spring Unit of the Sentinel 1 Solar Array Frans Doejaaren, Dutch Space B.V., Leiden, The Netherlands; et al 2:00 BREAK 2:15 SESSION III – DEPLOY, THEN A POTPOURRI OF POSTERS Colin Francis, Session Chair Space Systems/Loral, Palo Alto, CA  Ultra-low-weight Rotary Actuator for Operation on Mars and Pin Puller Mechanism Based on a Novel Shape Memory Alloy Technology Nestor Nava, Arquimea Ingenieria, S.L., Leganés, Spain; et al  Design and Performance of the Telescopic Tubular Mast Mehran Mobrem, Astro Aerospace - Northrop Grumman Aerospace Systems, Carpinteria, CA; et al

Document

AMS_2012.pdf

ix  Development of Variable Reluctance Resolver for Position Feedback Gregory Leibovich, Ducommun LaBarge Technologies, Carson, CA; et al  FeF 3 Catalytic Influence on PFPE Lubricants Lifetime under Loaded Conditions Lionel Gaillard, European Space Agency, Noordwijk, The Netherlands; et al  Trade Studies for a High Torque Density Planetary Gearbox Jeffrey Mobley, Sierra Nevada Corporation, Durham, NC  Single Motion Actuated Shape Memory Alloy Coupling Alberto Perez, Naval Postgraduate School, Monterey, CA; et al  Development and Testing of a High Compact Stepper Motor Mechanism Jörg Schmidt, Phytron Elektronik GmbH, Grabenzell, Germany; et al  Cryogenic Temperature Testing of NEA Fuse Wire Mechanism Edwin Vega, NEA Electronics, Inc., Moorpark, CA; et al 6:00 -10:00 RECEPTION - California Ballroom, Hilton Pasadena Invited component suppliers display current products and provide tutorials. Local high school FIRST Robotics Team demonstrations, and a light buffet meal. THURSDAY, 17 MAY 2012 7:00 Thursday Presenters' Breakfast - San Marino Room, Hilton Pasadena 8:00 SESSION IV – ROCKETS TO ROBOTS Brett Kennedy, Session Chair Jet Propulsion Laboratory, Pasadena, CA  Developmental Testing of Electric Thrust Vector Control Systems for Manned Launch Vehicle Applications Lisa Bates, NASA Marshall Space Flight Center, Huntsville, AL; David Young, Raytheon – Jacobs ESTS Group / NASA MSFC, Huntsville, AL  Development of High Temperature High Current Contact Technology in Slipring Assemblies for the BepiColombo MPO & MTM Spacecrafts Fabrice Rottmeier, RUAG Space Switzerland, Nyon, Switzerland, et al  Lessons Learned to Avoid Coax Cable Failure in Moving Mechanical Mechanisms Sheah Pirnack, Lockheed Martin Space Systems, Denver, CO  DLR’s Dynamic Actuator Modules for Robotic Space Applications Armin Wedler, German Aerospace Center (DLR), Wessling, Germany; et al 10:00 BREAK 10:15 SESSION V – MARS SCIENCE LABORATORY Ruben Nalbandian, Session Chair Moog, Inc., Chatsworth, CA  Mars Science Laboratory Rover Integrated Pump Assembly Bellows Jamming Failure Michael Johnson, Jet Propulsion Laboratory, Pasadena, CA, et al  Development of the Descent Brake Mechanism for the Mars Science Laboratory David Dowen, Sierra Nevada Corp., Louisville, CO; et al  Mars Science Laboratory Differential Restraint: The Devil is in the Details Elizabeth Jordan, Jet Propulsion Laboratory, Pasadena, CA  Mars Science Laboratory’s Dust Removal Tool Kiel Davis, Honeybee Robotics Spacecraft Mechanisms Corp., New York, NY; et al 12:15 LUNCH Lunch for AMS Attendees in the California Ballroom, Hilton Pasadena

Document

AMS_2012.pdf

x 1:15 SESSION VI – MORE MSL, INSTRUMENTS & SEPARATION Doug Packard, Session Chair NEA Electronics, Inc / Rocketstar Robotics, Inc, Moorpark, CA  A Zoom Lens for the MSL Mast Cameras: Mechanical Design and Development Daniel DiBiase, MDA Information Systems, Inc. – Space Division, Pasadena, CA, et al  Wet Chemistry Automated Sample Processing System (WASP) Juancarlos Soto, Ball Aerospace & Technologies Corp., Boulder, CO; et al  Refinement of a Low-Shock Separation System Chuck Lazansky, Sierra Nevada Corporation, Louisville, CO  Ares I Linear Mate Umbilical Plate and Collet William Manley, NASA Kennedy Space Center, FL; et al 3:15 BREAK 3:30 SESSION VII – INSTRUMENTS Charlie Hodges, Session Chair Sierra Nevada Corporation, Durham, NC  GMI Spin Mechanism Assembly Design, Development, and Test Results Scott Woolaway, Ball Aerospace & Technologies Corp., Boulder, CO; et al  Lessons Learned from the TIRS Instrument Mechanisms Development Jason Budinoff, NASA Goddard Space Flight Center, Greenbelt, MD; et al  Resolution for Fretting Wear Contamination on Cryogenic Mechanism Charles Clark, Lockheed Martin Space Systems, Palo Alto, CA 6:30-10:00 BANQUET – Pasadena Conference Center 6:30 – 8:30 Social Hour, Dinner and Live Jazz Band 8:30 – 10:00 Entertainment and Show FRIDAY, 18 MAY 2012 7:00 Friday Presenters’ Breakfast - San Marino Room, Hilton Pasadena 8:00 SESSION VIII – TRIBOLOGY/BEARINGS Terri Taylor, Session Chair Honeywell International, Glendale, AZ  Design and Manufacturing Considerations for Shockproof and Corrosion-Immune Superelastic Nickel-Titanium Bearings for a Space Station Application Christopher DellaCorte, NASA Glenn Research Center, Cleveland, OH; et al  Wear of Steel Ti6Al4V in Vacuum Timothy Krantz, NASA Glenn Research Center, Cleveland, OH; et al  Angular Runout Test Setup for High-Precision Ball Bearings Scott Miller, Lockheed Martin Space Systems, Palo Alto, CA; et al 9:30 BREAK

Document

AMS_2012.pdf

xi 9:45 SESSION IX – SMALL SESSION ON SMALL SATELLITES Scotty Allen, Session Chair ATK Space Systems, Goleta, CA  LightSail-1 Solar Sail Design and Qualification Chris Biddy, Stellar Exploration, San Luis Obispo, CA; et al  A Novel Release Mechanism Employing the Principle of Differential Coefficients of Thermal Expansion Clint Apland, Johns Hopkins University / Applied Physics Laboratory, Laurel, MD; et al  A Nichrome Burn Wire Release Mechanism for CubeSats Adam Thurn, Naval Research Laboratory, Washington, D.C.; et al  Antenna Deployment Mechanism for the Cubesat Xatcobeo. Lessons, Evolution and Final Design Jose Antonio Vilán Vilán, Universidad de Vigo, Vigo, Spain; et al 11:45 SPECIAL PRESENTATION Development of the Mars Science Laboratory Rover “Curiosity”, from Concept to Launch 12:15 TECHNICAL SESSIONS CONCLUSION Edward Boesiger, Operations Chairman, Lockheed Martin Space Systems, Sunnyvale, CA • Herzl Award Presentation • Closing Remarks 12:30 LUNCH Lunch for AMS Attendees in the California Ballroom, Hilton Pasadena 1:30 – 4:30 JPL TOUR 1:30 Buses depart hotel for JPL 2:00-4:00 Facility tour at JPL 4:30 Buses return to hotel

Document

AMS_2012.pdf

xii SYMPOSIUM ORGANIZING COMMITTEE Donald R. Sevilla, Host Chairman, JPL Louise Jandura, Co-Host Chair, JPL Stuart H. Loewenthal, General Chairman, Lockheed Martin Edward A. Boesiger, Operations Chairman, Lockheed Martin Steven W. Bauman, NASA GRC William Caldwell, NASA Ames Jared Dervan, NASA MSFC Carlton F. Foster, NASA MSFC (retired) Claef F. Hakun, NASA GSFC Christopher P. Hansen, NASA JSC Wayne Jermstad, NASA JSC Alan C. Littlefield, NASA KSC Ronald E. Mancini, NASA Ames (retired) Fred G. Martwick, NASA ARC Donald H. McQueen, Jr., NASA MSFC Gérard Migliorero, ESA/ESTeC Robert P. Mueller, NASA KSC Fred B. Oswald, NASA GRC Minh Phan, NASA GSFC Joseph P. Schepis, NASA GSFC Mark F. Turner, NASA ARC Robin Tutterow, NASA LaRC James E. Wells, NASA LaRC

Document

AMS_2012.pdf

1 Development of Brushed and Br ushless DC Motors for use in the ExoMars Drilling and Sampling Mechanism Robin Phillips*, Massimo Palladino** and Camille Courtois+ Abstract This paper presents a summary of work performed to qualify two COTS (Commercial Off The Shelf) motor types (one Ø13 mm brushed and one Ø22 mm brushless) for operation in a Martian atmosphere for the ExoMars Drilling and Sampling Mechanism. We present all the major steps in this process, which included an analysis of features that needed changing from the standard industrial motor design, a development program that was undertaken to select an appropriate design, and a qualification campaign that was then applied to the modified motors. Introduction Maxon motor is well known for having built all the drive and steering motors for JPL’s Mars Pathfinder (Sojourner) and MER (Spirit & Opportunity) rovers where the soundness of the basic design has been demonstrated over the 7+ years that Opportunity has now spent roving on the Martian surface. Maxon was also selected by ESA to develop motors for its ExoMars rover mission. However, the significant differences in requirements for the Drilling and Sampling mechanism did not allow the selection of the same type of motors as MER used. Specifically, the specified lifetime and power output for the main drill drive motor required the use of a brushless motor since the expected lifetime of a brushed motor was not sufficient for this application. Additionally, space restrictions inside the drill required the use of a smaller brushed motor (Ø13 mm) than had been used in either of the previous JPL missions. As with the JPL RE25 development, the aim was to adapt industrial standard motors (i.e., a so called “Commercial Off The Shelf” or COTS design) rather than to develop a custom solution in order to lower development costs. The process followed in the development program included three main steps: 1. A detailed analysis of the standard industrial motor design to identify features that were unlikely to work correctly when exposed to the environmental conditions and duty cycles requirements of the ExoMars mission. 2. Research into possible solutions to the identified design problems and then an appropriate redesign of the motors. 3. Manufacture and qualification testing to simulate all relevant aspects of the ExoMars mission on the new design. Mars Rover Electric Motors Background Successful operation of electric motors on the Martian surface was initiated by the Viking project which put two landers on Mars in 1976. The Viking landers used numerous motors in the twin cameras (e.g., for rotating the camera housing and moving the scanning mirror) as well as for the sampling boom and communications antenna pointing. That electric motors have a history of causing technical difficulties for * maxon motor ag, Sachseln, Switzerland ** ESA, ESTEC, Noordwijk, The Netherlands + RUAG Space, Nyon, Switzerland Proceedings of the 41st Aerospace Mechanisms Symposium, Jet Propulsion Laboratory, May 16-18, 2012

Document

AMS_2012.pdf

2 Martian applications is revealed by the appearance of the “Surface-sampler boom motor” in the list of “Top Ten Problems” for the Viking development program between Feb 1973 and Sept 1974 (Ezell and Ezell 1984, p. 253). After Viking there was a long pause until the Mars Pathfinder landed with the small Sojourner rover in 1997 which contained 11 maxon RE16 brushed motors. The Sojourner rover covered about 100 m before the lander communication link failed, thereby ending the mission after nearly 3 months. In covering the 100 m in just under 3 hours of driving time the drive motors rotated approximately 0.5 million times (with another 0.5 million during Earth based testing) with no indication of motor problems (Braun 1998). Following the success of the Mars Pathfinder mission, two larger and much more capable rovers were flown on the Mars Exploration Rover (MER) missions that landed the rovers Spirit and Opportunity on Mars in 2003. Each rover contained 39 maxon motors for drive and steering functions, solar array deployment, camera mast actuators and the actuators for the science arm joints. Two different motor types were used, a modified RE25 as can be found in maxon’s standard program and a RE20 that was specially developed for the MER program. The success of these missions is well known and as of January 2012 Opportunity was still active and had driven nearly 35 km over the previous 8 years. For the follow on MSL mission (launched in 2011), due to the increased distances the rover was expected to cover, brushless motors were selected. The development of these motors and associated gearboxes proved much more complex than expected and, as has been well documented in the press, ultimately led to a two year delay in the launch, emphasizing again the complexity involved in specialist motor design. Selection of Motor Types for the ExoMars Application The ExoMars program (as originally conceived by ESA) envisaged many motors of numerous types being used for various different functions on the rover, a summary of which is shown in Table 1. For the purposes of the development program being reported on here, the requirements for the Drilling and Sampling mechanism, hereafter “Drill”, were considered. Where possible the RE20 and RE25 brushed motors that were used on the MER missions were used in the original design. However, three applications in the drill required new motor sizes or types. These were the Drill mandrel clamp which has to fit within the diameter of the drill bit (see Figure 1) and hence is limited to Ø13 mm, the main drill drive Table 1: List of motor applications on the ExoMars rover Function Motor type Quantity on rover Wheel Drive RE 20-25 6 Steering Drive RE 20-25 6 Deployment Drive RE 20-25 6 Camera Mast Pan Axis RE 20-25 1 Camera Mast Tilt Axis RE 20-25 1 Drill Positioner Translation RE 20-25 1 Drill Positioner Rotation RE 20-25 1 Drill Positioner Jettison RE 20-25 1 Drill Translation EC22long 1 Drill Rod Mag Rotation RE 20-25 1 Drill Rod Mag Clamp RE 20-25 1 Drill Rod Lower Clamp RE 20-25 1 Drill Mandrel (main drive) EC40 1 Drill Mandrel Clamp RE13 1 Drill Tool RE13 1 SPDS=Sample Preparation & Distribution System Function Motor type Quantity on rover SPDS CSTM RE 20-25 1 SPDS BSD RE 20-25 1 SPDS Jaw actuation EC22 1 SPDS De-block actuation RE 20-25 1 SPDS Position/Rotary Motion RE 20-25 1 SPDS Dosing Mechanism RE 20-25 2 SPDS Carousel RE 20-25 1 Solar Array Deployment RE 20-25 1 Battery Isolation Switch RE13? 2 Life Marker Chip (LMC) Instrument Inlet Valve EC14fl 4 LMC Bellows pump EC8 4 LMC Rotary Valve RE13 4 Mars Organic Molecule Analysis (MOMA) Instrument RE10 2

Document

AMS_2012.pdf

3 motor which due to the required lifetime and power requirements need to be a Ø40-mm brushless motor and the drill translation unit which due to power requirements was best achieved with a Ø22-mm brushless motor. In order to avoid the motor problems that plagued the MSL development, a program was started to design and qualify both of these two new motor types. This program also had the express goal of making sure that the technology needed for making such motors was available outside of the USA so as to be free of ITAR restrictions. The first stage of this program was to analyze the existing commercial motor designs for features that were unlikely to be compatible with the ExoMars environmental and operating specifications. Figure 1: The ExoMars drill box breadboard (left) and drill mandrel EM (right) Specifications Summary Although much of the specification for the motors is similar to that for a terrestrial application, there are a few key areas that a significantly outside of the normal application regime that a COTS motor would be expected to be designed for, these are summarized in Table 2. Table 2: Summary of ExoMars specification that is significantly different to an industrial standard application Non-operating temperature range -120 °C to 125 °C Operating temperature range -55 °C to 30 °C Operating atmospheric conditions 1 bar, Earth standard 5-10 mbar CO 2 (Martian atmosphere) Vibration environment Sine 100 Hz, 33 g Random 20-2000 Hz, 17.2 g rms Shock loading 100 Hz: 25 g , 300 Hz: 400 g 2 kHz – 10 kHz: 1500 g Several features of both motors that were incompatible with these specifications had common solutions, such as modifying the bearing grease to Braycote 601EF for the low temperature pressure environment and changing the wiring to a type compatible with standard ESA wiring specifications. Various materials in the standard motor designs were modified, such as the aluminium housing and front flange of the EC22 which were changed to titanium, in order to avoid problems with differential expansion (aluminium bearing seat to stainless steel bearing race for example) or to reduce the mass. Other features, however, were specific to each motor type and are described in more detail in the next two sections.

Document

AMS_2012.pdf

4 Key Problems for Brushed Motors Derating factor The RE13 standard motor has a specification sheet defining the maximum mechanical output power (i.e., torque and speed) allowed. The extremes of the continuous operating regime are determined by the maximum allowed winding temperatures (125 °C in the case of the RE13). Normally the winding will cool via a mixture of conduction (via the shaft), convection (via the air surrounding the winding) and radiation. A reduction of atmospheric pressure will cause a corresponding reduction in the convection component of this cooling and hence in the available performance from the motor. In order to quantify the size of this effect, a simple test was conducted with an RE13 motor in a vacuum chamber. The resultant measured thermal resistances (Table 3) could then the programmed into maxon’s standard motor simulation software and new maximum continuous torque calculated. As can be seen from Table 3, the presence of even a small atmosphere causes a dramatic improvement in thermal conductivity. Table 3: Measured winding to housing ther mal resistance and corresponding maximum continuous torques for RE13 Atmospheric pressure (all with Earth standard composition) Thermal Resistance (Winding->Housing) [K/W] Maximum continuous torque [mN-m] 1 bar 7 2.39 7±1 mbar 9 1.85 <1x10-3 mbar 50 0.811 Vibration and Shock Considerations for Brushed Motors For the RE13 motor, there was concern that during shock and vibrations (mainly from launch and landing) there is no contact between the motor winding and the housing. This is particularly of concern since the RE13 motor winding is a cantilever (Figure 2) and the gap between the winding and housing can in the worst case tolerances be only 170 µm, as shown in (Figure 3). Figure 2: RE13 rotor showing bell shaped winding attached to shaft Figure 3: Detail of gap between RE13 winding and housing Based on a resonance search performed on the rotor by itself (see Figure 4), the winding fundamental frequency and the associated amplification factor were found to be 529 Hz and 3.5 respectively. Considering the random vibration requirement, by using the Miles equation for displacements, the winding displacement is given by: Winding Housing Magnet

Document

AMS_2012.pdf

5 3 332) ( fresQfres PSDdispRMS where fres is the motor winding fundamental frequency, Q is the amplification factor at resonance, PSD( fres) is the specified power spectral density of acceleration at frequency f and dispRMS is the RMS displacement. The motor winding RMS displacement during random vibration was found to be 25 µm with a 3 sigma peak of 75 µm. Therefore, any contact between the motor winding and the housing is excluded, even at 3 sigma (since the gap is no less than 170 µm). A random vibration test was performed on the fully assembled RE13 as part of the qualification testing and no signs of contact between winding and housing were seen after the test. A shock analysis was performed to assess the behavior of the RE13 motor winding, considering the required shock response spectrum. A half-sine shock of 1500 g amplitude and 0.5 ms was applied. Figure 5 shows that the shock generated (in red) envelopes the required shock response spectrum (in dotted blue). Figure 4: Test setup for the random vibration test conducted on the RE13 winding Figure 5: Shock response spectrum applied on the motors The displacement of the motor winding during shock, evaluated by analysis, is about 1 mm. This shows that the contact is likely. However, the shock will only create a contact between the motor winding and the housing during a very short period of time. This is considered acceptable (the time is short enough to avoid any damage). The shock test performed on the assembled RE13 motor showed that indeed no effects from such contact were observable. Brush material Although brushed motors have the advantage of not needing complex commutation electronics, they determine the lifetime of the motors and cause the life to be much lower than for an equivalent brushless motor. Unfortunately the most common brush materials for high current applications used in industrial motors (copper or silver graphite mixtures) require the presence of water vapor and oxygen to build up a hardened patina on the wear surface of the brush. If the water vapor is not present, then extremely rapid wear can occur. The situation with no water vapor applies on Mars and was noticed during qualification testing for the Viking lander camera scan mirror motors in the mid 1970’s (Mutch 1978, p. 17). Twenty years later for the Sojourner rover the problem was circumvented by the use of precious metal brushes without lubrication in otherwise nearly standard RE-16 motors. This is a simple solution for situations where required torques are low and only limited lifetimes are required. Testing for Sojourner showed that under load and start-stop conditions motor failure occurred after about 30-40 million revolutions (Braun 1998) which was more than sufficient for the planned mission.

Document

AMS_2012.pdf

6 For the MER mission a more extensive testing program was undertaken to test various different brush materials. Some of the early testing has been described in detail in (D. E. Noon 1999) and (Reid, Braun and Noon 1999), and it consisted of testing motors from various manufacturers with brushes made of mixtures of copper or silver graphite with impregnation with various different lubricants. For the flight models, the RE20 was flown with a copper graphite brush with 8% MoS 2 content (from Le Carbone) whereas the RE25 was flown with silver graphite brushes and 5% MoS 2 content (from Shunk). The success of these rovers (and by implication the motor and brush choice) has been well documented in the general press. Although the motors have not been completely trouble free (e.g., one failed drive motor on Spirit and one drive motor with periodic increased current draw on Opportunity), they have worked well enough to allow, in particular the Opportunity mission to significantly surpass the expected distance covered. The ~35 km driven by Jan 2012 represents ~65 million motor revolutions. This is well within the estimated lifetimes of several hundred million revolutions obtained from testing – although it is worth noting that the testing demonstrated a very large scatter with the (D. Noon 2001) report stating that the life could be as low as 10 million revolutions, which was indeed the case for Spirit’s failed motor. It is clear from the published reports that the selection of the correct brush material is critical to obtaining good lifetimes. Unfortunately explanations as to why significantly different brush mixtures were chosen for the RE20 than for the RE25 motors are not available in the published literature. What has been published shows a wide variation in the performance of similar brush materials with differing motor loads. Given that insufficient understanding of these issues was available either within maxon or ESA, it was decided to undertake a new test program for the RE13 motors. Table 4: Brush breakage test results showing values (in newtons) where the brushes failed. It is coincidence that types 1-3 have exactly the same mean and standard deviation. No. Type 1 Type 2 Type 3 Type 4 Type 5 Type 6 Type 7 Mean 9.5 9.5 9.5 10.8 9.1 7.2 5.9 SD 1.0 1.0 1.0 1.1 0.6 0.8 1.3 For the test, 24 RE13 motors based on type 118626 were used. Differences to the standard motor were the inclusion of Braycote 601EF lubricated bearings, a fiberglass re-enforced winding, an extended length shaft, and the various brush materials to be tested. The brush materials shown in Table 5 were tested, in each case three motors were built It is of course not only important that the brushes have a long lifetime in use, but that they also are structurally strong enough to withstand launch vibrations and shocks. A simple mechanical strength test was undertaken using a standard maxon procedure that is used to quality control standard brush lots. The brush is attached to a holder through its normal mounting hole, a rod is placed into the brush hook and force applied to pull on the brush arm. The force needed to break the arm off is shown in Table 5 where it can be seen that for the copper-based brushes there is no difference in strength. For silver-based brushes, increasing amounts of MoS 2 weaken the brush. However, the effect is probably not large enough to cause a problem; for comparison the maxon standard material (type 8 in Table 5) is specified to withstand a force of 8 N, other standard materials have specified values as low as 4 N. The brushes were constructed using the same mold as for the standard RE13, with a contact cross sectional area of 2.4 mm 2. Standard springs were also used that yield a contact force of 12.5 N-cm-2 with new brushes (as the brushes wear, the force reduces as the spring uncoils). The motors for the ExoMars application are required to operate for testing and qualification on Earth in a standard Earth atmosphere as well as for short periods in vacuum while travelling to Mars and of course extended operation on Mars in a 5-10 mbar CO 2 atmosphere. Hence the test program was specified with the segments as shown in Table 6. The motors were to be operated at the maximum working point defined by the specifications of 2.3 mNm load at a rotation speed of 8000 rpm.

Document

AMS_2012.pdf

7 Table 5: Brush material types tested. Motor group number is used elsewhere in this report to refer to a particular brush material type. Group 7 with a very high silver content was an attempt to produce a brush that was optimized for vacuum operation. Motor Group Brush Material 1 50% Cu 45% C 5% MoS 2 2 50% Cu 40% C 10% MoS 2 3 50% Cu 35% C 15% MoS 2 4 50% Ag 45% C 5% MoS 2 5 50% Ag 40% C 10% MoS 2 6 50% Ag 35% C 15% MoS 2 7 85% Ag 3% C 12% MoS 2 8 50% Cu 50% C Figure 6: Motor holder setup. The eddy current brake magnets are visible on the inside. Table 6: The test sequence followed. The motor revolutions column shows the approximate number of revolutions the motors made during each test phase (based on actual number of hours run and the approximate speed settings). Period Gas Pressure No. of motor revolutions 3 days 78% N 2, 21% O 2 + other traces ~1000mbar 30 million 2 weeks CO 2 5-10mbar 165 million 1 week none <10-6mbar 60 million Figure 7: Current draw for all motors in 8 mbar CO 2 test. Figure 8: Motor currents during the 2nd week in 8 mbar CO 2 atmosphere. The gaps in the data are caused by the logging PC crashing - the motors continued to run in these gaps.

Document

AMS_2012.pdf

8 In order to simulate a near worst-case thermal situation, the motors were mounted in a plastic Polyoxymethylene (POM) holder (see Figure 6) so there would be minimal thermal conductivity. Eddy-current brake disks were used to provide the required load. Each motor was individually monitored and its current draw logged. At the end of each test phase, 2 of the 3 motors with a particular brush type were opened up and pictures of the brushes and collectors taken under a microscope. The 3 rd motors of each brush type were not opened so as to retain the worn material inside the motors. The motors that were opened had all loose brush material vacuumed out before re-assembly. Other than vacuuming, no attempt to clean them was made; in particular the collector slots were not cleaned (other than vacuuming). The initial 2.5-day run under normal Earth atmospheric conditions produced the main result that the group 7 motors all failed within the first 14 hours. Inspection of these brushes showed that the cause of failure was the complete wear of the brushes. Thus these brushes are shown to be unsuitable. The following one week of operation in an 8 mbar CO 2 atmosphere was performed with the motors operating with loads of between 1.65 mN-m and 2.05 mN-m and speeds of between 8500 rpm and 9200 rpm. The differences were caused by a mixture of motor manufacturing tolerances (all motors shared the same power supply) and variations in how the eddy current brakes responded variably to running in a vacuum after being set under atmospheric conditions (brake adjustment after being placed in the vacuum chamber was not possible). The first week of operation in an 8 mbar CO 2 atmosphere also produced a number of failed motors. The first motor (equipped with maxon standard brushes) failed after 1.5 days followed a few days later by additional motors. By day 6, all group 8 motors (standard brushes), all group 1 motors, all group 2 motors, and 2 of 3 group 4 motors had failed. These were the groups with lower MoS 2 content. Additionally, all group 3 (Cu 15% MoS 2 brushes) showed increase current draw. We therefore decided to interrupt the test and open all failed motors to inspect the brushes. It immediately became clear upon inspecting the brushes that the motors were failing due to the complete wear of the brushes. As the brushes come to the end of their life, the section in contact with the collectors becomes wider causing several collectors to be in contact with the brush at the same time. This leads to winding segments being short circuited and hence a higher current draw. One problem that had not been anticipated was that as the motors start to fail they dramatically heat up. In the case of motor 8.2, which had a temperature sensor on it, the housing reached over 180°C (implying Figure 9: Motor current from the best performing brush types where no deterioration is seen after two weeks near continuous operation. Figure 10: Motor currents during operation in vacuum. Brush type 5 all failed within 25 minutes of starting the test!

Document

AMS_2012.pdf

9 a winding temperature of >200°C!). This caused the POM support structure to melt and hence their brake disks to contact the brake magnets. Thus no significance can be attributed to the presence or variation of spikes and other features in the current curves once the motors start to fail. The major results of this first week of testing are that copper-based brushes do not work as well as silver-based brushes for high load situations like this, and that in both material types more MoS 2 results in slower wear rates. Brush Type 2 (Motor 2.2) Brush Type 4 (Motor 4.3) Brush Type 6 (Motor 6.2) After 1 week CO 2 After 1 week CO 2 After 2 weeks CO 2 Figure 11: Pictures of the RE13 brushes after 1 week operation in Earth atmospheric conditions (top row) and 1 week operation (2 weeks for type 6) in an 8 mbar CO 2 atmosphere (bottom row). Since the results from the first week of tests were now understood, we decided to run the second half of the CO 2 test at a lower voltage to test a different working point. The eddy current brakes were set to a value corresponding to a load of 0.65-0.8 mN-m at 8000 rpm. Since the ExoMars specification of 2.3 mN-m includes all possible margins, actual operation would be expected at a lower power setting so it is also important to test at these lower current settings. The results from the second week of CO 2 testing clearly re-enforce the results of the first week. The silver-based brushes work considerably better than the copper-based brushes and the higher the quantity of MoS 2, the longer the brush lasts. No significant wear is present in group 6 brushes, even after 2 weeks of operation (see Figure 9). It seems likely that some significance can be attributed to the result that both motors 3.3 and 5.3 (i.e. those which were not opened and hence retained the worn brush dust) started to fail before those motors which were opened and cleaned, however the effect is not large. The results from the vacuum testing were consistent with what was been reported by (D. E. Noon 1999) in that failure is extremely rapid and is caused by brush wear residue accumulating between the collector bars and causing partial shorts. Type 5 motors were already starting to wear out at the end of the CO 2 testing as they were showing increased current draw. Despite this, we were surprised at how fast they failed; after less than half an hour of running in vacuum all three group 5 motors had failed.

Document

AMS_2012.pdf

10 Figure 12: Type 6 collector after the final vacuum test - wear levels are very low, however the collector bars are partly shorted together by material stuck between them. Figure 13: Example from type 5 brush where material can be seen shorting the collector bars. Key Problems for Brushless Motors For the BrushLess DC (BLDC) motors, the brush wear problem of course is not present. However, commutation for a BLDC motor needs a feedback device to tell the control electronics when to commutate (sensorless operation using back-EMF rather than a physical feedback device is not suitable for start stop operation or where high loads can be expected on startup). The most common way of performing this feedback is using hall sensors; however these are vulnerable to radiation damage. The ExoMars mission specifies a total dose threshold of 10 krad, defined as the end of life dose deposited under 2 mm of Al-eq solid sphere shielding with a safety factor of 2. In order to provide additional safety margins, the motor is specified with redundant hall sensors, each set being capable of operating the motor alone. Although a number of radiation hard hall sensors are available, these suffer from the major problem that they are US sourced and hence covered by ITAR. Additionally the sensors that we were aware of are all larger than ideal for a Ø22-mm motor. We therefore decided to test several non-US sourced sensors that maxon uses in various motor lines to see if a standard type could be used. Figure 14 shows the test setup where 3 examples of each type of hall sensor to be tested, mounted on a PCB, were placed in front of a rotating permanent magnet. During the test, the sensors were in continuous operation. The permanent magnets were rotated using standard RE13 and RE25 motors, thereby simultaneously confirming that the radiation caused no problems for the two main types of brushed motors to be used (no effects were expected and none were seen). The radiation dose was delivered over a period of 90 hours at a rate of 0.079 Gy/min, yielding a total dose of 38 krad (Si). The sensors were additionally retested after an annealing period of 100 hours (at 25°C) to confirm no further changes. Table 7 shows that only the Infineon TLE4945 sensor passed the test with no measurable degradation of the switching level. This result is not unexpected since the Infineon sensor is of bipolar design whereas the others are CMOS which is known to be more sensitive to radiation. Subsequent low temperature testing of the Infineon sensors was undertaken to ensure their operation at -55°C (the manufacturer only rates them to -40°C). With the successful passing of this test, these sensors were selected for design into the EC22 motors.

Document

AMS_2012.pdf

11 Figure 14: Test setup for hall sensor irradiation with test configuration of 4 PCBs, each with 3 hall sensors of one type, mounted in front of four motors with permanent magnets attached to their output shafts. Table 7: Results of Hall sensor irradiation Sensor type Behavior during irradiation Switching level after annealing Melexis US3881LSE Output voltage started dropping after ~20 krad 2.2 V Melexis US4881LUCOutput voltage started dropping after ~20 krad 0 V Infineon TLE4945 No change observed 5 V Allegro A3230 No change observed 0 V Qualification Testing In order to qualify the motors for operation on Mars, a test sequence was established to fully test all phases of a motor’s life, including ground testing, launch, travel to Mars and of course operation on the Martian surface. The major steps of the test sequence are shown in Figure 15. Table 8: Shock requirements for ExoMars mission The “motor recurrent measurements” shown in Figure 15 was a defined sequence of measurements, including winding resistance, dielectric strength and inductance, housing capacitance and motor torque constant that was repeated after every major test step. Sterilization tests The ExoMars program has established a document describing the planetary protection requirements that applies to “all ExoMars spacecraft elements” (Kminek 2007). For the purposes of this qualification, a simple test consisting of 60 hours at 125°C in a dry N 2 atmosphere was considered adequate. Figure 15: Motor qualification test sequence Shock Response (Q=10) Frequency Launch from Earth Landing on Mars 100 Hz 25 g 10 g 300 Hz 400 g 30 g 2,000 Hz 1500 g 30 g 10,000 Hz 1500 g 30 g

Document

AMS_2012.pdf

12 Shock Tests The requirements for launch and landing shocks are given in Table 8. Only the launch levels were tested as shown in Figure 16. Vibration Tests An initial resonance search was performed consisting of a sine sweep with frequency 5-2000 Hz, sweep rate of 2 oct/min ±5% with an amplitude of 0.5 g. Next, a sine sweep at the specified rate (2 oct/min ±5%) and at the amplitude shown in Table 9 was performed for 2 minutes for each of the three axes. The motors were not operated during this test (reflecting the launch conditions). Finally, a random vibration test was run, also for 2 minutes per axis for all three axes. Table 9: Vibration parameters for both RE13 and EC22 motor tests Sine vibration Random Vibration Input Freq.(Hz) Amplitude Input Freq. (Hz) Amplitude 0-20 15.5 mm peak 20-100 +6 [dB/oct.] 20-100 27.2 g 100-400 0.45 [g2/Hz] 400-2000 -6 [dB/oct.] Figure 16: Picture of RE13 & EC22 test motors attached to a stiff mounting cube (note the feedback accelerometers a ttached to both motors in the left picture) on the shaker (left) and on a plate for the shock test (right). The powder actuated nail gun used to impart the shock is visible at the right. Operating Cycles The central part of the qualification tests for both motor types were low temperature, 8 mbar CO 2 and vacuum lifetime tests with representative loads were performed as shown in Table 11. In each case, the test setup consisted of a brake motor (whose braking load could be adjusted via a variable resistor across the windings) and the motor being tested as shown in Figure 17. A bearingless torque sensor was used to couple the two motors together and provide a reading of the actual braking force being applied. The speeds and loads used during the lifetime tests were the reference “worst case” loads given in the ExoMars specifications.

Document

AMS_2012.pdf

13 Table 10: CVCM per item Temperature RE13 EC22 -25°C 0.22 mg 0.57 mg -50°C 0.37 mg 0.86 mg -75°C 0.42 mg 0.92 mg Table 11: Low temperature, CO 2 & vacuum tests Motor Test step Temperature range Load torque & rpm No. of cycles Approx. no. of motor revs. EC22 RE13 Low temp. cycles -120°C to +25°C none 10 non-operating 0 EC22 Operating cycles -55°C to 40°C 25Nm@12,000rpm 10x as in Figure 19 7.2 million EC22 Lifetime test (8 mbar CO 2) 25Nm@12,000rpm 500x as in Figure 19 360 million EC22 Vacuum test 25Nm@12,000rpm 10x as in Figure 19 7.2 million RE13 Operating cycles 2.3mNm@8,000rpm 10x as in Figure 19 4.8 million RE13 Lifetime test (8 mbar CO 2) 2.3mNm@8,000rpm 150+200 as in Figure 18 22.4 million RE13 Vacuum test 2.3mNm@8,000rpm 3x as in Figure 19 1.4 million Figure 17: Test configuration for the EC22 (a similar configuration was used for the RE13) Outgassing performance To get a good performance from the motors in terms of particulates contamination and outgassing, which is essential for a mission like ExoMars that is searching for traces of life, the motors were baked out at 125°C for five days (the maximum allowable temperature of the windings). Following the qualification tests described above, the motors underwent outgassing tests at the ESA/ESTEC materials laboratory. The tests consisted of raising the temperature from 25°C to 125°C in steps of 25°C every 24 hours in a pressure of 10 -7 mbar. The Collected Volatile Condensable Material (CVCM) and the Total Mass Loss (TML) were measured by using four Quartz Crystal Microbalance (QCM) plates at various temperatures as shown in Table 10. Particulates emission from the brushes was a major concern for the RE13 motor since these are carbon based so several design features were added to make the exit path for wear particles from the motor as long as possible but to simultaneously ensure the minimum amount of material vents through the bearings during launch. However, the outgassing test results showed that there was negligible emission of particles from the brushes. The overall TML and CVCM levels were found to be very low.

Document

AMS_2012.pdf

14 Figure 18: Reference cycle for RE13 Figure 19: Reference cycle for EC22 (and certain RE13 tests) Figure 20: TML for RE13 Figure 21: TML for EC22 For the RE13 and the EC22, the TML, measured at 125°C, is 17 mg/item and 122 mg/item, respectively. Figures 20 and 21 show the TML extrapolated to different temperatures to predict more realistic operating scenarios. For the EC22, there was some concern related to an aromatic amine and a plasticizer. The TML and CVCM results showed that the aromatic amine and the plasticizer have been clearly removed by the bake-out. The low TML and CVCM levels measured for the RE13 and EC22 during the outgassing tests have resulted in a discussion within the ExoMars project as to whether the current requirement to encapsulate the motors can be dropped. This would save a significant amount of mass on the ExoMars mission, due to the considerable number of motors involved in the mission. Qualification Results Both motors passed the sterilization, vibration and shock tests with no noticeable problems (as determined by the recurrent measurements taken after each test step). The RE13 performed without problems through the life test in CO 2. However, during the vacuum test, after just 2 cycles (representing just 2 hours operation), increasing current draw was observed and after the 3 rd cycle the test was aborted to prevent damaging the motor and destroying any evidence of the cause of the problem. Upon strip down of the motor, it was clear that the cause of the problem was as

Document

AMS_2012.pdf

15 had been seen in the previous brush wear tests and was caused by worn brush/collector material sticking between the collector bars and partially shorting the windings. No other major problems were seen during the strip down. Although the EC22 completed all the life tests, including the vacuum part of the test sequence, two major problems were encountered. Most serious was the failure of two hall sensors after the “operating cycles” test listed in Table 11. In order to allow the sequence to proceed, the rest of the testing (including the main lifetime in CO 2 test) was performed using a sensorless controller (i.e., one that relies on back EMF to commutate). All the remaining hall sensors were still operational at the end of the test sequence. Upon strip down it was also noticed that the rear flange had worked loose (by about 0.5 mm) and had resulted in the partial failure of both bearings (due to extra forces on them since the rotor was no longer perfectly aligned to the stator. It is unlikely that the motor would have run for much longer in this condition. Planned Future Work Current work is concentrating on understanding the cause of the rear flange movement (for which a fairly simple mechanical design improvement is expected to provide a solution) and the more complex issue of why the hall sensors failed. So far, considerably more extreme temperature cycling on several additional PCBs has failed to reproduce the problem. The next stage of the development program, which is expected to be performed during 2012, is to develop motor gearbox combinations using the motors described in this paper as well as the missing types listed in Table 1 that have not yet been qualified. This will also be an opportunity to confirm the effectiveness of the design changes to the EC22 to remove the two identified weaknesses. Summary The work presented here has shown that, with appropriate modifications, the COTS motors RE13 and EC22 are capable of standing up to the launch vibration and shock and a landing on Mars and then functioning for the full planned ExoMars mission in the environmental conditions found on Mars. The necessity of a qualification campaign as described here is made clear by the failures that were encountered in the hall sensors and the weakness in the rear flange design of the EC22. With suitable modifications in these areas, we expect a trouble-free qualification of these motors when attached to the gearboxes for the next stage of development. Acknowledgements This work was performed in collaboration between maxon motor, RUAG Space and Selex-Galileo and was funded by an ESA development contract. References Braun, David and Noon, Don. "Long Life" DC brush motor for use on the Mars Surveyor Program. 32 nd Aerospace Mechanisms Symposium, 1998. Ezell, Edward C., and Linda N. Ezell. On Mars: Exploration of the Red Planet 1958-1978. Vols. SP-4212. NASA, 1984. Kminek, G. Planetary Protection Requirements. ESA Doc. EXM-MS-RS-ESA-00005 Iss.3 Rev.1, 2007. Kminek, G. Planetary Protection Requirements. ESA Doc. EXM-MS-RS-ESA-00005 Iss.3 Rev.1, 2007. Mutch, Thomas A. The Martian Landscape. Vols. SP-425. NASA, 1978. Noon, Don E. "Motor Brush Testing for Mars and Vacuum" 33rd Aerospace Mechanisms Symposium 1999 Noon, Don. Motor Brush Development Testing Results. Unpublished, 2001. Reid, Lisa K., David F. Braun, and Don E. Noon. "Robotic Arm and Rover Actuator Systems for Mars Exploration." New Perspectives in Mechatronics Systems. Winterthur, 1999.

Document

AMS_2012.pdf

17 Rotary Percussive Sample Acquisition Tool (SAT): Hardware Development and Testing Kerry Klein*, Mircea Badescu*, Nicolas Haddad*, Lori Shiraishi* and Phillip Walkemeyer* Abstract In support of a potential Mars Sample Return mission, an Integrated Mars Sample Acquisition and Handling (IMSAH) architecture has been proposed to provide a means for Rover-based end-to-end sample capture and caching. A key enabling feature of the architecture is the use of a low mass Sample Acquisition Tool (SAT) that is capable of drilling and capturing rock cores directly within a sample tube in order to maintain sample integrity and prevent contamination across the sample chain. As such, this paper will describe the development and testing of a low mass rotary percussive SAT that has been shown to provide a means for core generation, fracture, and capture. Introduction As part of a potential Mars Sample Return campaign NASA and the European Space Agency are mutually working on a Mars 2018 Joint Rover Mission to potentially send a rover to Mars in order to perform in-situ investigations as well as collection of Martian samples for a return to Earth upon a subsequent mission. As such, it is foreseen that a key NASA payload contribution is the development of a Sample Acquisition and Caching subsystem capable of acquiring Martian rock cores and soil samples that could be cached within a return canister. Once the samples have been successfully cached within the return canister, the canister would be placed on the Martian surface. A follow-on mission element would then utilize a fetch rover to pick up the return canister and place it within a Mars Ascent Vehicle which would be capable of inserting the canister into a passive orbit around Mars. A third and final element of the campaign would then rendezvous with the return canister and return it to Earth [1], [2], [3]. Figure 1. Integrated sample acquisition and caching prototype subsystem. In support of the development of the Sample Acquisition and Caching subsystem, the Integrated Mars Sample Acquisition and Handling (IMSAH) architecture was developed in order to advance the key * Jet Propulsion Laboratory, Pasadena, CA Proceedings of the 41st Aerospace Mechanisms Symposium, Jet Propulsion Laboratory, May 16-18, 2012 SHEC TDD SAT

Document

AMS_2012.pdf

18 elements necessary for end-to-end sample generation and containerization. The IMSAH architecture has been presented in depth in previous publications [4], [5], [6], [7], and is characterized by three major subelements as shown in Figure 1: 1. Tool Deployment Device (TDD) 2. Sample Acquisition Tool (SAT) 3. Sample Handling Encapsulation and Containerization (SHEC). The corresponding operational process of the specified hardware as it pertains to the IMSAH architecture is depicted in Figures 2 and 3 and defined by the following operational needs [4] [7]:  Sample transfer between the coring tool (SAT) and the caching mechanism (SHEC) is to occur by means of bit change-out  Acquire samples into individual sample tubes in order to preserve sample integrity and minimize the risk associated with handling cores of unknown geometry.  Utilize a rotary percussion mechanism for the Sample Acquisition Tool in order to reduce subsystem mass and maximize efficiency. The use of a rotary percussive coring tool allows for successful coring at a reduced weight on bit (i.e., lower arm preload), minimizes bit walk due to spindle rotation, and allows for robust hole start when compared with rotary only alternatives.  The coring tool deployment, alignment, preload, and feed would be performed using a five degree-of-freedom (DOF) robotic arm. By using the specified deployment arm the system has enough DOFs to provide tool alignment and accommodate modest rover slip. Figure 2. IMSAH coring tool deployment Figure 3. Bit change-out and sample transfer configuration Of the three IMSAH sub-elements, this paper will focus on the development of the rotary percussive Sample Acquisition Tool (SAT) with an emphasis on the tool’s mechanism designs, testing challenges, and lessons learned.

Document

AMS_2012.pdf

19 Sample Acquisition Tool Description & Requirements The proposed IMSAH architecutre allows for a less complex tool to be developed than what has previously been investigated or developed as it allows for the use of the TDD for both tool deployment, alignment, preload, and feed. These previous tools required stability tines due to the use of an integral 1-DOF linear feed mechanism which in turn results in the need for greater drilling preload and therefore greater demands on the tool deployment system. Furthermore, the use of the TDD as the linear feed allows for the employment of a sprung linear compliance stage between the turret and the coring tool to provide for both dynamic isolation of linear motion as well as extended linear range of motion during operation under light arm preloads. In order to satisfy the IMSAH operational needs, as described in the previous section, the SAT was designed to provide for autonomous core generation, core fracture/retention, and bit change-out. The resulting functions necessary to perform these operations have been identified and listed as requirements as follows: 1. Acquire rock cores with approximately 1 cm in diameter by 5 cm in length. 2. Acquire at least 20 rock cores for return. 3. Acquire samples from Kaolinite, Santa Barbara Limestone, Siltstone, Saddleback Basalt, and Volcanic Breccia. 4. Be able to eject a bit that is inadvertently stuck in a rock 5. Be robust to anomalous cores that may be broken in the bit and/or at the bit opening. 6. Account for catastrophic slip conditions where it is presumed the rover experiences a significant shift in position while the SAT tool is in the ground. 7. Cores need to be of appropriate quality and suitable for caching It is important to note that during the development effort the determination of the tool’s performance as it pertains to core quality was a qualitatve assessment that binned the generated cores into three categories as described below and represented in Figure 4:  Good – full length cores or in a few segments  Acceptable – mostly segments, discs, and/or pucks  Bad – Powder and/or small chunks, stratigraphy not maintained Figure 4. Assessment of core quality based on qualitative parameters.

Document

AMS_2012.pdf

20 The subsequent sections will provide detailed information associated with the development and testing of the tool as well as resultant lessons learned. The sections are arranged to give a general overview of the SAT followed by greater detail of the four SAT subassemblies/mechanisms. Sample Acquisition Tool (SAT) Design & Testing SAT Assembly Overview Figure 5 provides a general overview of the completed sampling tool. The developed SAT is a rotary percussive coring tool comprised of four mechanical subassemblies – the Spindle Percussion Assembly (SPA), the Core Breakoff Assembly (CBO), the Magnetic Chuck Assembly (MCA), and the Core Bit Assembly (CBA). The mechanisms are driven by a total of three dc brushless motors mated to gear heads which provide the necessary actuation to complete the critical functions for core generation, fracture, and capture. Figure 5. Rotary Percussive Sample Acquisition Tool (SAT) Figure 6. Core Quality Results In order to validate the tool’s unit level functionality a series of verification and validation tests have been performed using a rock test suite that encompasses a variety of rock types that are analogous to Martian rocks (as specified in the requirements) and have been used in the past to qualify Martian surface sampling hardware. The results of the testing have shown the tool can successfully generate, fracture, and capture rock cores within a sample tube for all of the rocks within the test suite while maintaining an appropriate level of core quality, see Figure 6. SAT General Hardware Development Lessons Learned Early on in the tool’s development it became apparent, as with most R&D efforts, that schedule and resources were going to be severely limited. As such, it was assessed that the tool design should be somewhat modular, and implement mechanisms and corresponding subassembly interfaces such that each mechanism could be tested and operated independently of the others. Doing so provided for several advantages:

Document

AMS_2012.pdf

21  Mechanisms could be developed relatively independently of the others as long as interfaces were maintained and negotiated.  Mechanisms could be tested at a subassembly level allowing for early performance/capability investigations prior to tool integration.  Resultant modularity allows for relatively easy assembly/disassembly of tool during anomalous behavior investigations and allows for the isolation of possible suspect mechanism behavior. Spindle Percussion Assembly (SPA) Design & Development SPA Overview Figures 7 and 8 provide an overview of the Spindle Percussion Assembly (SPA) within the SAT. The SPA is a linked spindle/percussion mechanism that provides the rotational DOF necessary to drive the Core Bit Assembly through the spindle drivetrain. In turn, the rotational DOF is translated to axial motion through the use of a cam and lever which drives a striker mass and provides the necessary impact energy to facilitate rock fracture. Since the tool development was intended to be a single point design, a linked spindle percussion mechanism was chosen early in the development life cycle because it allowed for a reduction in the number of required actuators and a lower tool mass. Figure 7. General overview of SPA within the SAT In order to verify the SPA’s ability to generate the necessary impact energy, a standalone test was devised using high-speed video to monitor the striker velocity upon contact with the anvil for a given spindle speed (see Figure 9). Due to the linked spindle/percussion mechanism, the impact energy is presented as a function of spindle speed. The initial SPA design implemented a striker mass of approximately 51 g based on preliminary percussion development tests. However, due to a design error in the lever stroke length the resultant impact energy was approximately 20% less than the intended design point resulting in the need to increase the striker mass to approximately 61 g. However, during additional testing it also became apparent that the losses downstream of where impact energy was being measured were greater than

Document

AMS_2012.pdf

22 initially expected. As resources were not available to fully investigate the system losses, a final striker mass of 121 g (approximately 2x the 61 g striker mass and the largest striker than could be implemented within the assembly constraints) was selected in order to ensure that an appropriate level of margin could be maintained during drill operations in the harder rocks of the proposed test suite. Figure 8. General schematic of SPA Figure 9. Impact energy at a range of spindle speeds for a variety of striker masses. SPA Lessons Learned Several design choices were selected early in the development life cycle because they allowed for either a reduction in the number of actuators or resulted in a lower tool mass. In addition, a fair amount of effort was spent on trying to provide an optimized system in terms of generated impact energy vs. performance. However, in retrospect several noticeable drawbacks cropped up by pursuing these routes:  Linked mechanisms provide for reduced flexibility in terms of which “knobs” can be turned when investigating tool performance and capability. If the required mechanism capability is clearly defined (i.e., flight like requirements are already known) this may be less of an issue. However, during early development efforts this is not necessarily the case and can actually limit one’s ability to investigate anomalous behavior due to either ill-defined requirements or test parameters.  Due to the overall complexity of the mechanism, trying to provide an optimized system is very challenging especially due to the number of variables that can affect performance. During early development efforts greater emphasis should be placed on ensuring a high degree of capability rather than optimization as the tool will be utilized extensively to derive “flight” capability requirements. The optimization could then more successfully be implemented during the flight development effort. This is especially true for sampling mechanisms. Striker

Document

AMS_2012.pdf

23 Magnetic Chuck Assembly (MCA) Design & Testing MCA Overview Figures 10 and 11 provide an overview of the Magnetic Chuck Assembly (MCA). The main design driver for the MCA was associated with the desire to allow for a passive breakaway of the Core Bit Assembly under a rover slip condition during drilling. The implementation of a passive release also required that a low-profile separation interface be maintained to prevent an off-nominal loading condition in the event of a hardware hangup or snag occuring during interface separation As such, the MCA design utilizes two permanent magnets (one fixed and one with a rotational degree of freedom) to create a magnetic interface that can be turned on (engaged) or turned off (disengaged) while maintaining a very low-profile separation interface. Figure 10. General overview of MCA within SAT Upon testing, the MCA was shown to function as intended and the bit could be passively released under a given separation load of 100 N at the bit tip. In order to utilize the magnetic chuck approach, several compromises were required in terms of material selection. In order to prevent unwanted interactions between the MCA and adjacent mechanisms, several torque coupling interfaces and drive shafts were machined from non-magnetic A286 which could not be hardened as much as desired. In anticipation of the accelerated wear that would occur on these compromised interfaces, additional material was provided to allow for increased torque coupling wear during operation while preserving the tool’s functional lifetime. As expected, accelerated wear was noticed during testing at the specified torque coupling interfaces (see Figure 12). MCA Lessons Learned In order to employ a new and novel approach for bit retention it became apparent that several common mechanism design issues would need to be considered:  More often than not some type of compromise (e.g. accelerated wear of low hardness materials at torque interfaces due to material selection contraints) often results that may not present itself until the hardware is well into its verification and validation testing. As such, recognizing this

Document

AMS_2012.pdf

24 compromise early in the design processes may allow for a better anticipation of the actual hardware performance and possible mitigation paths.  This also clearly highlights the need to address hardware durability early in the design lifecycle as well as providing a means for graceful degradation. Figure 11. General schematic of MCA Figure 12. Accelerated wear at torque couplings Core Bit Assembly (CBA) Design & Testing CBA Overview Figures 13 and 14 provide an overview of the Core Bit Assembly (CBA). The CBA is used to provide the mechanical constraints for the Core Break-Off Mechanism and to accept the impact energy and the rotational input from the Spindle Percussion Assembly. The main component of the CBA is the custom drill bit which is supported by a duplex bearing pair and is free to translate axially within a sleeved bushing to allow maximum transmission of the applied impact energy to the rock. The drill bit is based on a COTS coring bit tooth configuration with two external helical flutes for cuttings removal. Due to the choice of implementing a magnetic chuck, a large overall chuck interface was required for the CBA. This in turn resulted in a large bit housing which inadvertently reduced the tool’s capability to accommodate large surface irregularities in the rock as well drilling full-depth holes at off-normal angles (see Figure 15). Furthermore, the large diameter has downstream impacts to the overall IMSAH architecture volume and mass as the SHEC houses the spare bits.

Document

AMS_2012.pdf

25 Figure 13. General overview of CBA within SAT Figure 14. General schematic of the CBA Figure 15. Tool cannot accommodate large surface irregularities or drilling at off-normal angles due to minimal clearance between bit housing and rock. Minimal clearance at full-depth hole

Document

AMS_2012.pdf

26 During the tool development it was presumed that the amount of rock that needs to be removed as well as bit tooth wear play a significant factor in the tool’s rate of penetration performance. Since the ID of the bit is based on a desired core size of 9.9 mm (0.39 inch), the amount of material that is removed during drilling (and the corresponding bit face surface area) is driven by the coring bits outer diameter. Furthermore, because the coring bit houses the Core Break-Off mechanism understanding the tool’s sensitivity to bit diameter would be useful for follow on development efforts. Although resources were not available to fully investigate the sensitivity of the tool to bit diameter, two bounding cases were developed: 1. Min Bit OD – 22.6 mm (0.89 inch) 2. Max Bit OD – 24.4 mm (0.96 inch) It is clear from the test data shown in Table 1 that there appears to in fact be a significant improvement in rate of penetration even with a minor reduction in the bit face surface area of only 16%. For the purpose of this investigation relatively unworn bits were used during the rate of penetration investigation in order to limit the amount of test variables. Table 1. Rate of penetration of coring tool utilizing 121g striker at 500 rpm Avg Drilling Rate [mm/min] 22.6 mm (0.96") Bit 24.4 mm (0.89") Bit Limestone 26.2 59.6 Kaolinite 49.3 162 Siltstone 6.3 27.9 Saddleback Basalt 2.4 7.9 Volcanic Breccia 1.3 2.7 CBA Lessons Learned During the CBA design effort, the modularity of the tool’s sub-elements allowed for a relatively independent development from its mating hardware; however, it also resulted in a failure to check the implications of an increase of the mating chuck interface at the tool’s assembly level resulting in a significant reduction in capability to drill holes that are off normal from the rock surface.  As such, remember the big picture – resultant interface growth that is satisfied at the mechanism level, i.e. black box interfaces to black box as intended, might still allow proper mechanism function, but upset higher level functionality In addition, it was recognized that since this was an early development effort with a low maturity level, a significant portion of the verification and validation effort should focus on understanding the tool’s performance and capability sensitivity as it relates to bit diameter.  Therefore, during early development efforts with hardware that has a low maturity level, significant emphasis should be placed on the desire to understand the hardware’s performance sensitivity to varying requirements, especially if the results may be used as inputs on subsequent higher fidelity tools.

Document

AMS_2012.pdf

27 Core Break-Off (CBO) Mechanism Design & Testing Figure 16. General schematic of CBO CBO Overview Figure 17 provides an overview of the Core Break-Off (CBO) Mechanism. The core break-off device is a pinching/cleaving mechanism that uses two opposing 45o wedges that are symmetrically split about the cleaving plane. The mechanism is actuated through the CBO drivetrain located within the SPA and is designed to fracture cores that range in diameter from 9 mm to 10 mm. As alluded to above, the need to fracture the generated core close to the parent rock required the CBO to be nested within the ID of the drill bit. As such, during nominal drilling operations the CBO is disengaged from the CBO drivetrain in order to allow for unobstructed operation of the Core Bit Assembly. Once the desired core is generated, the CBO drivetrain interfaces to the CBO mechanism thru a novel axially compliant lead screw that translates axially to interface with the CBO torque interface as well as provide the rotational input to fracture the core. During testing, the CBO was able to fracture all generated cores as intended. However, a significant problem did present itself during testing. In addition to providing a means for core fracture and capture, the CBO also provides a method for retaining the sample tube within the pinch tube by way of flexure retention fingers. During testing, these retention features began to fail at high stress concentration regions located at the base of the flexures. This failure method was initially missed during the design phase as the flexures were only analyzed under a low static load which showed significant margin. However, this load case did not appropriately account for the repeated cyclical loading generated by the SPA and the resultant failure of the flexure by fatigue. CBO Lessons Learned As previously mentioned, due to the vibratory environment generated by the percussive mechanism several retention features ended up failing to retain the sample tube. Therefore:  In a vibratory environment, stress concentration regions should be analyzed for fatigue failure even if the resultant cyclical load is small and static results show high margin.

Document

AMS_2012.pdf

28 Conclusion In support of the development of the Integrated Mars Sample Acquisition and Handling (IMSAH) architecture, a low-mass Sample Acquisition Tool (SAT) has been developed. As with most R&D efforts, the project’s schedule and resources were limited so a modular approach to the tool’s development was implemented such that individual subassemblies/mechanisms could be independently developed and tested from the others prior to integration. In addition, these same schedule constraints resulted in early design choices that limited the tool’s flexibility to investigate some anomalous behavior due to several linked mechanisms as well as a premature attempt at performance optimization. As such, during early development efforts with low maturity levels, it became readily apparent that greater emphasis should be placed on the desire to understand the hardware’s performance sensitivity to varying requirements. In doing so, more useful feedback can be provided to allow for better informed decisions on subsequent higher fidelity developments efforts. With that said, the developed SAT has demonstrated the necessary capability to autonomously generate, fracture, and capture rocks cores as necessitated by the proposed sampling architecture. Acknowledgements The research described in this publication was carried out at the Jet Propulsion Laboratory of California Institute of Technology under contract from the National Aeronautics and Space Administration (NASA). References [1] R. Mattingly, S. Matousek, and F. Jordan, “Continuing Evolution of Mars Sample Return,” IEEE Aerospace Conference, paper #1238, March 2004 [2] R. Mattingly and L. May, “Mars Sample Return as a Campaign”, IEEE Aerospace Conference, paper #1805, March 2011 [3] P. Backes, J. Aldrich, D. Zarzhitsky, K. Klein, and P. Younse “Demonstration of Autonomous Coring and Caching for a Mars Sample Return Campaign Concept”, IEEE Aerospace Conference, March 2012 [4] P. Backes, R. Lindemann, C. Collins, and P. Younse, “An Integrated Coring and Caching Concept,” IEEE Aerospace Conference, paper #1675, March 2010 [5] P. Younse, C. Collins, and P. Backes, “A Sample Handling, Encapsulation, and Containerization Subsystem Concept for Mars Sample Caching Mission,” International Planetary Probe Workshop (IPPW-7), June 2010 [6] N. Hudson, P. Backes, M. DiCicco, and M. Bajracharya, “Estimation and Control for Autonomous Coring from a Rover Manipulator,” IEEE Aerospace Conference, March 2010 [7] P. Backes, P. Younse, M. DiCicco, N. Hudson, C. Collins, A. Allwood, R. Paolini, C. Male, J. Ma, A. Steele, P. Conrad, “Experimental Results of Rover-Based Coring and Caching,” IEEE Aerospace Conference, March 2011 [8] H. Price, K. Cramer, S. Doudrick, W. Lee, J. Matijevic, S. Weinstein, T. Lam-Trong, O. Marsal, and R. Mitcheltree, “Mars Sample Return Spacecraft Systems Architecture,” IEEE Aerospace Conference, March 2000 [9] A. Okon, “Mars Science Laboratory Drill, “40th Aerospace Mechanisms Symposium, Cocoa Beach Florida, May 2010

Document

AMS_2012.pdf

29 Lock & Release Mechanism for the CHOMIK Penetrating Device and its Tribological Properties Marcin Dobrowolski* and Jerzy Grygorczuk* Abstract A unique geological low velocity penetrator CHOMIK for the Russian Phobos Sample Return mission has been developed at Space Research Centre PAS (SRC PAS). One of the most important goals of the mission is to collect a soil sample from Phobos and deliver it to the Earth. The sample will be collected from the surface layers of the Mars’ satellite by the penetrator and deposited in a container that is going to land in 2014 in Kazakhstan, encased in the Russian re-entry capsule. Apart from sampling, CHOMIK will perform thermal and mechanical measurements of Phobos’ regolith. The penetrator itself is an improved version of the MUPUS instrument for the ESA’s Rosetta mission. A new lock and release (L&R) mechanism for the instrument has been developed to meet the Phobos lander requirements. The penetrator is held in place by a multi-arms mechanism and released using Dyneema string melting system. This actuation method provides reliable operation with negligible shock and no special handling requirements. Introduction The Space Research Centre PAS has been involved in the development of low-speed, hammer-driven penetrators since 1996. It was strictly involved with the MUPUS project for the Rosetta mission to the 67P/Churyumov–Gerasimenko comet. Rosetta spacecraft was launched in March 2004 and will reach the comet in 2014. Recently, in 2010 SRC was invited to take part in the Phobos Sample Return mission and develop a penetration system capable of collecting a sample of the Mars’ satellite soil. Such a device has been delivered to Russia in April 2011 and the Phobos-Grunt spacecraft has been launched in November 2011. One of the most important parts of the hammer-driven penetrator is its insertion device, which constitutes about 90% of the mass allocated for the whole penetrator. The insertion device consists of free suspended elements in one degree of freedom; e.g., hammer or counter-mass that is accelerated in opposite directions during operation in microgravity environment (law of conservation of momentum). The launch of the rocket and landing operations are dangerous for a delicate counter-mass suspension and, therefore, it was locked in the stowed configuration by a specially dedicated L&R mechanism. Such a unit has to hold the device steadily and to move the penetrator head in a predicted direction after release. The operating conditions of the system are about -160°C for MUPUS and about -100°C for CHOMIK in high vacuum environment. For a long time, from launch to landing which might take up to ten years, all components have to keep stable parameters of work. L&R mechanism is an extra element of the penetrator so it needs to minimize its mass and power consumption. Based on the abovementioned requirements, a method of holding and releasing of the penetrators was developed based on the Dyneema string melting system. The Dyneema fiber (SK65) is a multi-filament fiber produced from UHMW-PE with the following main characteristics: high strength, low density, low elongation at break and negative coefficient of expansion during heating up. Table 1 shows main Dyneema fiber properties compared to steel. * Space Research Centre of the Polish Academy of Sciences, Warsaw, Poland Proceedings of the 41st Aerospace Mechanisms Symposium, Jet Propulsion Laboratory, May 16-18, 2012

Document

AMS_2012.pdf

30 For long duration exposure, Dyneema fiber can be used from cryogenic conditions up to a temperature of 70°C. A not very high melting temperature (~150°C) of the fiber makes it possible to cut the Dyneema string using a heating element operated by a small amount of electric power. Preliminary investigations have shown that good results can be achieved using a small resistor as a heating element and a 60° string angle of contact. Tests have shown that Dyneema string shrinks during the warming up process providing better contact to the resistor. The actuation time depends mainly on the local temperature. High reliability and repeatability ratio was observed in the presented actuation method. Table 1. Comparison of the Dyneema fiber and steel selected properties Material Density kg/m 3 Tensile strength GPa Young’s modulus GPa Elongation at break % Melting temperature °C DYNEEMA®SK65 970 3 95 3.6 149 Steel 316L 8000 0.6 200 45 1370 The general rule in our designs is using Dyneema string only as an actuator element not as a main holder element. For a string of 0.5-mm diameter, the tensile load capability is about 350 N but we reduce the cord’s intensity of stress using leverage to eliminate the necessity of using very tough heating elements. A simple way of putting into use the listed principles is shown on the Figure 1 in MUPUS lock and release mechanism pictorial drawing. Figure 1. MUPUS penetrator head lock & release mechanism principle of operation The MUPUS penetrator head, being a part of the counter-mass, has three cylindrical appendages that are placed in three bearings of the structure. The bearings are crossed to provide a zero degrees of freedom grip. When stowed, the penetrator head is held by the clamping of the latch component into its bracket. The proper tension between the cylindrical appendages and bearings is provided by an elastic flat spring and adjusting the gap between beam and structure. Melting of the Dyneema string releases the opening of the latch component and the unlocked bracket allows the motion of the penetrator head in the deployment direction, as it is indicated in Figure 1 by the red arrow.

Document

AMS_2012.pdf

31 CHOMIK lock & release mechanism description CHOMIK L&R mechanism is modified and improved compared to MUPUS’. First of all, it has been designed for different dynamic loads during launch, able to withstand 10g of vibration acceleration and 40g shocks. Such loads result in a relatively high penetrator holding force, which in this case was about 400N. General requirements also influenced the instrument structure that has to contain electronics boards, penetrator guide, and the lock and release unit as a compatible part. The movement direction of the penetrator head after releasing also underwent changes. In MUPUS, the penetrator head has to retreat from the locked elements. In CHOMIK, the penetrator head has to operate inside its holding elements. The instruments with the penetrator head’s movement directions pointed out and their lock & release mechanisms are shown in Figure 2 and Figure 3. The selection of the melting elements determined the parameters of Dyneema string: 0.5-mm diameter with 10-15 N tension force safe loading. Such a system required 1:40 leverage to withstand the penetrator’s head holding force. Figure 4 illustrates a kinematic diagram of the developed symmetric multi-arms system. The penetrator holding force is transmitted through the crank arms to a slider by the binary links. An angular shaped slider has a sliding joint with a possible displacement of 18 mm. It is blocked by the holders in the stowed position. During release, initially connected by Dyneema string, holders rotate, setting the slider free. The slider’s linear movement cause the crank arms’ rotation. Three pairs of adjustable holding pins separate from the penetrator’s sockets creating 10-mm gaps between the lock and the penetrator. Figure 2. MUPUS penetrator with deployment system (left) and its lock and release mechanism in the stowed position (right)

Document

AMS_2012.pdf

32 Figure 3. CHOMIK penetrator with electronic box (left) and its lock and release mechanism in the stowed position (right) – penetrator is removed Figure 4. A planar kinematic diagram of the lock and release mechanism There are three types of springs responsible for the mechanism release: (I) torsion springs generate torque T to the crank arms rotating them by 18°, (II) flat C-shaped springs guide and produce a pushing force F1 to a slider, (III) flat springs, which are integrated with holders and execute two functions: kick-off and Dyneema string tensioning with F2 force.

Document

AMS_2012.pdf

33 1 - PENETRATOR 2 – SOCKETS 3 - HOLDING PINS 4 - CRANK ARMS 5 - BINARY LINKS 6 - DYNEEMA STRING 7 - SLIDER 8 - HEATERS (RESISTORS) 9 - HOLDERS 10 - STRUCTURE Figure 5. Practical realization of the mechanism The tight string protects the holders after their displacement caused by the Dyneema cord elongation of about 3.5%. The design of the mechanism is shown in Figure 5. The complete lock and release mechanism weighs 150 grams and occupies a volume of 81x82x36 mm. Two redundant melting elements have 110 Ω of resistance and the actuation power voltage is 28V. Two connectors provide confirmation of the holding arms release. The main challenge during the mechanism development was the elimination of the self-locking effect. To avoid such a situation, the design process was preceded by detailed kinematic studies, and then a structural-thermal model was assembled and tested. The investigation showed some problems with slider elements that potentially could produce a non-symmetric release. This issue has been solved by blocking one of the kick-off spring’s motions and implemented in the qualification model. Tests also showed that it is very important to place all moving arms in their suitable positions during the arming process. Figure 6 presents the mechanism’s arming equipment which is helpful to provide symmetric arms’ settings and to generate the preliminary tension making it possible to install the Dyneema string. The mechanism successfully passed the vibration resistance test, shock test and linear overload test. A functional test has been carried out in the vacuum-thermal chamber. The temperature on the instrument structure was about -150°C and vacuum of about 10 -6 Pa. The Dyneema string application needs 6 seconds for release in those simulating Phobos environmental conditions, compared to less than 1 second at room temperature (laboratory air). Shock generated during release is very low compared to those created by a pyrotechnic actuator.

Document

AMS_2012.pdf

34 Figure 6. Mechanism arming equipment (left) and L&R mechanism with Dyneema string installed (right) High-speed camera record has shown that the maximum speed of the holding pins is about 4 m/s during releasing (at room temperature). The time period, from Dyneema string melt to the crank arm full rotation, is about 5 milliseconds; then the crank arms rebound several times dissipating energy, and stop after about 52 milliseconds. As it is shown in Figure 7, the first crank arms movement is very equal, which is the result of accurate positioning of the mechanism element during Dyneema string installation using the arming equipment. Figure 8 shows selected moments of the mechanism action. Frame A shows the locked mechanism. In Frame B, the Dyneema string is melted but holders still occupy the slider lateral surface. In Frame C, the slider is shifting resulting in the binary links pull and that makes the crank arms rotation possible. Frame D shows a mechanism in the fully released position. The moments captured on the presented frames has been marked in Figure 7. Figure 7. Holding pins displacement during mechanism release 3436384042444648505254 0 0,005 0,01 0,015 0,02 0,025 0,03 0,035 0,04 0,045 0,05 0,055 0,06 0,065 0,07Time [s]Displacement [mm] QM left crank arm position QM right crank arm positionABD C

Document

AMS_2012.pdf

35 Figure 8. Lock & release mechanism action - penetrator is removed Tribological tests All rotating and sliding joints in the mechanism (Figure 9) have to work in very low temperatures and vacuum, with a small movement range and with only one, short time actuation. This influenced the slide bearings selection. Titanium alloy with titanium nitride layer and polyimide-based plastic Vespel SP1 were chosen as mating parts materials. Tribo-components made of Ti6Al4V alloy (shafts, pins, latches) with Vespel SP1 (bushings, sliders, guides) have been used in many SRC PAS mechanisms, including the CHOMIK device. This selection was based on our experience, previous material properties tests, and the ability to manufacture good quality layers on parts with a complicated shape by our partners.

Document

AMS_2012.pdf

36 Figure 9. L&R mechanism with its exemplary plastic-metal slide componentsFigure 10. Pin-on-Disk tribometer mounted in the SRC thermal-vacuum chamber To develop a unit without a self-locking effect, a coefficient of friction for selected mating plastic-metal parts had to be known. Another unknown was a repeatability of the mechanism behavior during room and low temperature operation. Adequate tests were conducted in a newly built vacuum pin-on-disc tribometer (Figure 10). Ti6Al4V alloy with titanium nitride layer as a pin while Vespel SP1 as a disk were used. Figure 11. Coefficient of friction versus sliding distance for Ti6Al4V pin with TiN layer and Vespel SP1 disk for vacuum environment: room temper ature (left), very low temperature (right) The tests were conducted in the following conditions: (1) pressure: 0.01-0.04 Pa, (2) load: 50N, (3) sliding speed: 0.1 m/s, (4) contact area: 10 mm2. Figure 11 (left) shows coefficient of friction versus sliding distance at room temperature, whereas Figure 11 (right) is a diagram of coefficient of friction for -80°C. Tribological tests in the Phobos-simulated environment showed that the coefficient of friction for Vespel SP1 and Titanium alloy with titanium nitride layer is low, about 0.22. Moreover, friction properties are stable and on the same level, both at room or low temperature. 00,10,20,30,40,5 0 5 10 15 20 25 30 35 Distance [m]Coefficient of Friction 00,10,20,30,40,5 0 5 10 15 20 25 30 35 Distance [m]Coefficient of Friction

Document

AMS_2012.pdf

37 Conclusions Development of the CHOMIK L&R mechanism provided successive arguments that thin Dyneema string melting method can be applied for releasing medium-high loaded space devices. The alike systems are developed at SRC PAS taking into account the superior reliability of the melting process. Medium-higher loads need a high leverage mechanism, which is the main subject of the presented paper. The conducted tests proved it. One of the lessons learned is that in case of the high leverage mechanism, additional springs counteracting the self-locking of the mechanism must be applied. Usage of a system that counteracts the cord elongation is also recommended. The investigations that have been made confirm that the Dyneema string melting actuation method is very reliable and Vespel SP1 and titanium alloy with titanium nitride layer are suitable mating materials for low temperature applications. References 1. Grygorczuk J., Dobrowolski M., Wisniewski L., Banaszkiewicz M., Ciesielska M., Kedziora B., Seweryn K., Wawrzaszek R., Wierzchon T., Trzaska M., Ossowski M. “Advance mechanisms and tribological tests of the hammering sampling device CHOMIK. ” The Proceedings of ESMATS” (September 2011). 2. Grygorczuk, J., Banaszkiewicz, M., Seweryn, K., Spohn, T. (2007). MUPUS Insertion device for the Rosetta mission. Journal of Telecommunications and Information Technology (1/2007), pp50-53. 3. Grygorczuk, J., Banaszkiewicz, M., Kargl, G., Kömle, N., Ball, A.J., Seweryn, K. (2009). Use of hammering to determine cometary nucleus mechanical properties. Penetrometry in the Solar System II (Eds: Günter Kargl, Norbert I. Kömle, Andrew J. Ball, Ralph Lorenz), Proceedings of the 2nd International Workshop on Penetrometry in the Solar System, Graz, 25th-28th September, 2006, Austrian Academy of Sciences Press, pp93-107. The paper is supported by the Polish national grant project 791/N-ROSJA/2010/0 and ESA PECS project No. 98 105

Document

AMS_2012.pdf

39 Harmonic Drive™ Gear Material Selection and Life Testing Jeffrey Mobley* and Jonathan Parker* Abstract Sierra Nevada Corporation (SNC), along with NASA Goddard Space Flight Center (GSFC), tested several Harmonic Drive gears for their longevity and performance to determine the best material combinations for a mission, requiring an actuator life nearly ten times previously qualified. Accelerated life testing in both ambient and vacuum environments revealed two suitable candidates for the flight build. The combination of a Nitronic 60 circular spline and 15-5PH H1075 flexible spline was ultimately chosen. It was surprising though, that 15-5PH H1075 on both circular and flexible splines outperformed the heritage material combination (Melonited Flex Spline against a 15-5PH Circular Spline) and also could have been used in the flight build. Introduction The NASA Global Precipitation Measurement (GPM) mission required actuators for its High Gain Antenna System (HGAS) and Solar Array Drive Assembly (SADA) with a life significantly higher than the required life of heritage actuators previously qualified by SNC for the Lunar Reconnaissance Orbiter (LRO) mission and the Solar Dynamics Observatory (SDO) mission as shown in Table 1. One of the life-limiting components in the actuator design is the Harmonic Drive gear. Wear was observed in the circular spline teeth during the post-life test teardown inspection on the SDO and LRO actuators. This wear was determined to be acceptable for those missions, but when extrapolated out to the life requirements of the GPM mission, the effect of tooth wear on performance became uncertain. The heritage LRO and SDO actuators used T-Cup Harmonic Drive gears manufactured by Harmonic Drive LLC in Peabody, MA in order to provide a large through hole in the actuator. The drives had a 200:1 gear ratio and utilized a combination of a Melonite TM treated Flex Spline and 15-5PH Circular Spline. This particular material combination had not been life tested by either SNC or Harmonic Drive LLC for the number of cycles required for the GPM mission. Table 1: Required Life Comparison Application Cycle # of Cycles (for 1X Life) Total Degrees SDO 360° (continuous) 2,500 900,000° LRO 180° CW, 180° CCW 5,700 2,052,000° GPM SADA 250°CW, 250° CCW 17,500 8,750,000° GPM HGAS 190° CW, 190° CCW 48,600 18,468,000° Harmonic Drive™ is a registered trade mark of Harmonic Drive LLC * Sierra Nevada Corporation, Durham, NC Proceedings of the 41st Aerospace Mechanisms Symposium, Jet Propulsion Laboratory, May 16-18, 2012

Document

AMS_2012.pdf

40 Hardware Configurations Based on discussions between Harmonic Drive LLC and SNC, the following material combinations were selected for comparative life testing, with the goal of extending the life of the Harmonic Drive gear in order to meet the GPM HGAS requirement:  15-5PH H1075 Circular Spline against Melonited 15-5PH H1100 Flex Spline o Most similar to heritage o Removed weight reduction cuts for more repeatable performance o Reduced the size of the pre-melonite grit blast media for better surface finish o Increased the circular spline hardness from H1150 to H1075 for better wear resistance  Nitronic 60 Circular Spline against Melonited Flex Spline o Same as above except for Circular Spline material o Nitronic 60 recommended by Harmonic Drive LLC based upon testing reported in Reference 1  Nitronic 60 Circular Spline against 15-5PH H1075 Flex Spline o Same as above except for flex spline material o Melonite removed to result in less abrasive flex spline o Flex Spline hardness increased to H1075 for better wear resistance  15-5PH H1075 Circular Spline against 15-5PH H1075 Flex Spline o Chosen for test because the components were available from the other combinations o Low expectations due to potential for galling/cold welding from identical materials in contact All Harmonic Drive gear configurations were customized HDT-25, 200:1, T-cup component sets manufactured in Peabody MA by Harmonic Drive LLC and all were of the same dimensional design. All combinations used the same wave-generator bearing material and design, and each was identically lubricated with Pennzane 2001-3PbNp oil and Rheolube 2004 grease per standard procedures. Table 2 summarizes the tested configurations with assigned part numbers. Table 2: Material Combinations Part # Circular Spline Material Flexible Spline Material 36638-1 15-5PH H1075 15-5PH H1100 w/ Melonite 36638-2 15-5PH H1075 15-5PH H1075 36638-3 Nitronic 60 Annealed 15-5PH H1100 w/ Melonite 36638-4 Nitronic 60 Annealed 15-5PH H1075

Document

AMS_2012.pdf

41 Ambient Life Test Plan The Life Test plan is shown in Figure 1. A standard run-in was performed by Harmonic Drive LLC on each Harmonic Drive component prior to the start of any testing. Following run-in, each Harmonic Drive gear was characterized by measuring starting torque and torsional stiffness. The starting torque of the Wave Generator input was measured in six equidistant positions with the Circular Spline fixed and no load applied to the Flex Spline output. Torsional stiffness was measured in six equidistant locations at the Flex Spline output with the Circular Spline and Wave Generator input held stationary. Once each Harmonic Drive gear was assembled into its life test fixture, the torsional stiffness test was repeated at 25 equidistant locations to characterize the baseline assembled condition. The ambient life test was conducted to simulate 2X the required mission life and was conducted in air at room temperature. The Harmonic Drive gear input was driven at a nominal input speed of 52.36 rad/s (500 rpm). The Harmonic Drive gear output was loaded to 2.93 N-m (26 in-lb) with a friction brake (see Figure 2) to simulate the application frictional load plus the load required to accelerate the specified inertia. The life test was run alternating between clockwise and counter-clockwise directions for approximately 24 hours at a time. At the conclusion of the life test, the Harmonic Drive gear had completed over 51,300 output revolutions in each direction (39,936,000° of total travel). Following the ambient life test, the units were again tested for torsional stiffness while assembled in the test fixture. The units were then disassembled, inspected, cleaned, and inspected further with findings documented below. Figure 1: Test Flow Figure 2: Ambient Life Test Setup

Document

AMS_2012.pdf

42 Ambient Life Test Results Torsional Stiffness As shown in Figure 3, the Harmonic Drive gears with a Melonite Flex Spline (36638-1 & 36638-3) showed a decrease in torsional stiffness (5 to 8%) compared to the 15-5PH Flex Splines which showed a slight increase (1% to 5%). As shown in Figure 4 and Figure 5, the shape of the post-life stiffness curves for the Harmonic Drive gears with the 15-5PH Flex Splines were very comparable to the pre-life stiffness curves. However, the shape of the post-life stiffness curves for the Harmonic Drive gears with the Melonite Flex Splines showed a significant reduction in low torque stiffness when compared to the pre-life stiffness curves. The reduction in stiffness on the Harmonic Drive gears with the Melonited Flex Splines was an early indication of increased wear in those drives. Teardown Inspection Each Harmonic Drive gear was disassembled to examine its components for wear and debris generation. Circular Splines Figure 6 illustrates the differences between the Circular Splines at the end of the life test. Configurations with the Melonite Flex Splines (36638-1 & 36638-3) contained metallic particles packed into the roots of the Circular Spline teeth. Numerous small fragments of metal were found in the roots of 36638-1 but were not as prevalent in 36638-3 even though the amount of wear was similar. The metal fragments were found to be fine metallic wear debris that had packed into the roots of the teeth and been formed into slivers as seen in the photo of the 36638-1 Circular Spline. Additionally there was a noticeable wear step on the 36638-1 & 36638-3 Circular Splines. The non-Melonite Harmonic Drive gears (36638-2 & 36638-4) showed only a slight wear step at the gear mesh and had no accumulation of metallic debris in the roots of the gear teeth. The lubricant color was darker than seen on the 36638-2 & 36638-4 flexible splines, further indicating an increased presence of wear debris. Flex Splines Figure 7 and Figure 8 illustrate the differences between the teeth and ID of the Flex Splines at the end of the life test. Configurations with the Melonite Flex Splines (36638-1 & 36638-3) showed very little difference between pre and post- life for both the teeth and the ID. The non-Melonite Flex Splines (36638-2 & 36638-4) teeth were visually worn in the area of highest contact, but the wear was comparable to that seen on their respective Circular Splines and not nearly as significant as seen on the 36638-1 & 36638-3 Circular Splines. The ID of the non-Melonite Flex Splines (36638-2 & 36638-4) were visually worn in the area of contact with the wave generator OD, but there were no signs of galling and the amount of wear was acceptable for a 2X life condition. Wave Generator Figure 9 and Figure 10 illustrate the differences between the OD and balls of the Wave Generators at the end of the life test. All Wave Generators showed some polishing on the OD from contact with the Flex Splines. The Melonited Flex Splines in 36638-1 and 36638-3 showed a higher level of polishing but not enough to cause any concern. The wave generator lubricant was lighter in the non-Melonite drives (36638-2 & 36638-4) indicating less wear particles present. Harmonic Drive LLC Inspection The Harmonic Drive gears were thoroughly cleaned and sent back to the manufacturer to measure the tooth profiles. A total of four teeth were measured on each flexible and circular spline and compared to the theoretically perfect tooth. The Harmonic Drive gears using the Melonite Flex Spline (36638-1 & 36638-3) showed much more deviation from the theoretical tooth profile than the non-melonite flexible splines. The teeth on the non-Melonite drives (36638-2 & 36638-4) were still generally within the acceptable manufacturing standard tolerance range, with 36638-2 (15-5/15-5) showing the least amount of deviation.

Document

AMS_2012.pdf

43 Figure 3: Pre and Post Ambient Life Stiffness Figure 4: Pre- Life Minimum Stiffness Curves (Radians vs. In-Lb)

Document

AMS_2012.pdf

44 Figure 5: Post- Life Minimum Stiffness Curves (Radians vs. In-Lb) Figure 6: Circular Spline Teeth, Post Ambient Life Metallic Fragments

Document

AMS_2012.pdf

45 Figure 7: Flex Spline Teeth, Post Ambient Life Figure 8: Flex Spline ID, Post Ambient Life

Document

AMS_2012.pdf

46 Figure 9: Wave Generator OD, Post Ambient Life Figure 10: Wave Generator Balls, Post Ambient Life

Document

AMS_2012.pdf

47 Ambient Life Test Results Summary The two Melonite Harmonic Drive gears (36638-1 and 36638-3) were eliminated from contention due to the reduction in torsional stiffness at the end of life and the presence of relatively large particles of metallic debris in the gear mesh. The two non-Melonite Harmonic Drive gears (36638-2 & 36638-4) were relatively comparable. 36638-2 visually looked slightly better, (primarily with regard to the circular spline teeth and gear lubricant) but the 36638-4 was nonetheless acceptable. The concern over how 15-5PH against 15-5PH would perform during vibration and in a vacuum environment, resulted in the 36638-4 configuration (15-5/Nitronic) being chosen as the least risky path forward for the flight build. A qualitative assessment of the key parameters of the post- life inspection is shown in Table 3. In parallel with building flight hardware with the chosen configuration, the program decided to further this study by performing a vacuum life test comparison between the non-Melonite drive configurations (36638-2 & 36638-4). Table 3: Qualitative Summary of Ambient Life Test Results 36638-1 15-5/Melonite 36638-2 15-5/15-5 36638-3 Nitronic/Melonite 36638-4 Nitronic/15-5 Stiffness Change Poor Good Poor Good Circular Spline Tooth Wear Poor Average Poor Average Flex Spline Tooth Wear Good Average Good Average Flex Spline Cup Good Average Good Average Wave Generator OD Average Average Average Average Gear Teeth Lube Poor Average Poor Average Wave Generator Lube Good Average Good Average Vacuum Life Test Plan Based on the ambient life test results and concerns regarding the validity of a life test in ambient air, a second round of testing was conducted in a vacuum. The testing was repeated on new Harmonic Drive gears with the best performing configurations from the ambient test, 36638-2 and 36638-4, in a vacuum environment combined with thermal cycling. Testing was performed per the same flow shown in Figure 1 except as noted below. Baseline torsional stiffness testing was performed on the vacuum test Harmonic Drive gears prior to life testing as previously done with the ambient test to characterize the pre-life performance. The vacuum life test consisted of 100,000 cycles of reversing 180° output revolutions in a <5.0x10 -5 Torr vacuum environment, cycling between 0 and 40°C. The input to the Harmonic Drive gear was set to 52.36 rad/s (500 rpm). A magnetic particle brake was used to apply a constant 2.93 N-m (26 in-lb) friction load to the output for the duration of the test. Following the ambient life test, the units were again tested for torsional stiffness while assembled in the test fixture. The units were then disassembled, inspected, cleaned, and inspected further with findings documented below.

Document

AMS_2012.pdf

48 Vacuum Life Test Results Torsional Stiffness As shown in Figure 11, there was very little difference between the post- life torsional stiffness of the two configurations. The 36638-4 saw a 5% reduction in torsional stiffness while the 36638-4 saw a 9% reduction in torsional stiffness, but based on the stiffness plots seen in Figure 12 , neither drive saw a significant reduction in low torque torsional stiffness that was seen on the Melonited drives during the ambient life test. Teardown Inspection Each Harmonic Drive gear was disassembled to examine its components for wear and debris generation. Circular Splines Figure 13 illustrates the differences between the Circular Splines at the end of the life test. The Nitronic 60 Circular Spline (36638-4) only showed a change in surface finish at the edge of flex spline contact, whereas the 15-5PH Circular Spline (36638-2) showed a slight wear step. Neither drive displayed significant generation of metallic debris or any buildup of debris in the roots. The gear mesh lubricant was lighter than seen after the ambient life test most likely due to reduced oxidation in vacuum. Flex Splines Figure 14 and Figure 15 show the teeth and ID of the Flex Splines at the end of the life test. The Flex Spline teeth of both drives were minimally worn in the area of highest tooth contact. There was no sign of galling or cold welding in the 36638-2 configuration (15-5PH against 15-5PH). The ID of the Flex Splines were visually worn in the area of contact with the wave generator OD, but there were no signs of galling and the amount of wear was acceptable for a 2X life condition and comparable to the ambient life test. Wave Generator Figure 16 and Figure 17 show the OD and balls of the Wave Generators at the end of the life test. Both Wave Generators showed some polishing on the OD from contact with the Flex Splines, but results were comparable to the ambient life test and to be expected for 2X life. Again, the wave generator lubricant was slightly lighter than seen in the ambient life test (less oxidation) and the 36638-4 configuration was slightly darker than the 36638-2 configuration. GSFC Material Analysis Very slight wear was observed on the flex spline teeth of both assemblies which appeared to be slight burnishing of the tooth surface. No pitting or scoring was visible. Grease from both assemblies exhibited evidence of metallic wear. Infrared spectroscopic analysis of the grease was inconclusive in determining a ‘winner’. In all the samples, a weak intensity carbonyl band was detected, indicating very little grease/oil degradation. The greater amount of metallic wear debris in the flex spline-circular spline contact of the 15-5PH & 15-5PH combination suggests the 15-5PH & Nitronic 60 is a slightly better configuration

Document

AMS_2012.pdf

49 Figure 11: Pre and Post Vacuum Life Stiffness Figure 12: Pre and Post Life Torsional Stiffness Plots

Document

AMS_2012.pdf

50 Figure 13 – Circular Spline Teeth, Post Vacuum Life Figure 14 – Flex Spline Teeth, Post Vacuum Life

Document

AMS_2012.pdf

51 Figure 15: Flex Spline ID, Post Vacuum life Figure 16: Wave Generator OD, Post Vacuum Life Figure 17: Wave Generator Balls, Post Vacuum Life

Document

AMS_2012.pdf

52 Vacuum Life Test Results Summary The vacuum environment had little to no effect on the wear rate of the two Harmonic Drive gears tested. The slight differences in wear between the two configurations are not significant, and changes in wear between vacuum and ambient tests can be attributed to normal variation from unit to unit. The results of the vacuum life test supported the decision by the program to use the 15-5PH against Nitronic 60 Harmonic Drive gear configuration (36638-4) for the GPM program. Surprisingly, the 15-5PH against 15-5PH configuration (36638-2) would have also been an acceptable choice. A qualitative assessment of the key parameters of the post life inspection is shown in Table 4. Table 4: Qualitative Summary of Vacuum Life Test Results 36638-2 15-5PH / 15-5PH 36638-4 Nitronic / 15-5PH Stiffness Change Good+ Good- Circular Spline Tooth Wear Good- Good+ Flex Spline Tooth Wear Good Good Flex Spline Cup Average Average Wave Generator OD Average Average Gear Teeth Lube Average Average Wave Generator Lube Average Average Conclusion Accelerated Harmonic Drive gear life testing validated the concern that the heritage material combination of a Melonite Flex Spline running against a 15-5PH Circular Spline was not optimal for an extended life requirement of nearly ten times previously qualified. An alternate combination of a Melonite Flex Spline against a Nitronic 60 Circular Spline was also ruled out based upon test results. After completion of both Ambient and Vacuum accelerated life tests, two material combinations displayed test results indicating a high level of confidence in passing the increased life requirement. After joint deliberation between SNC and NASA GSFC, the combination of a 15-5PH H1075 Flex Spline and a Nitronic 60 Circular Spline was chosen for incorporation into the GPM HGAS and SADA actuators. Despite industry concerns over identical stainless steels operating in contact, the material combination of 15-5PH H1075 against 15-5PH 1075 performed nearly identically to the combination selected and would have been a high confidence selection as well. References 1. Johnson, Michael R., Russ Gehling, and Ray Head “Life Test Failure of Harmonic Gears in a TwoAxis Gimbal for the Mars Reconnaissance Orbiter Spacecraft.” Proceedings of the 38th Aerospace Mechanisms Symposium. May, 2006.

Document

AMS_2012.pdf

53 New Supplier - Hardware Duplication – Some Pitfalls Edwin Joscelyn* Abstract Companies, for many reasons, often need to task a new supplier to duplicate the design and manufacture of a product that has alr eady been qualified and flown on one of their earlier systems. This paper deals with such a situation. The conflicts that arise when required to include characteristics other than those defining the specific performance and dimensional requirements of the product is explored. While this paper uses the “reverse engineering” of a motor as the means to describe some of these problems, the basic conflicts and conditions can apply to other areas and products. Introduction The task was to duplicate the design of 3 different motor configurations. Although the performance specifications were clear and well written, there were no other mechanical design details provided other than the outline dimensions, winding location, and weight. The real challenge came from additional non-performance requirements that were specified. These requirements stemmed from the fact that these motors pre-existed from an earlier build cycle and the program was extremely sensitive to anything which differed from this earlier build in the new design. Winding profile, material, torque constants, resistance, inductance, harmonic distortion, detent torque, and drag torque parameters could not differ from the earlier designs. In addition some of the dimensions had tight tolerances on the OD and ID of the designs (e.g. 13 µm / 0.0005 inch). While these conditions alone might not seem to be in themselves presenting that difficult a design situation certain underlying factors proved otherwise. In many cases, the task of “reverse engineering” is greatly simplified when there is a unit available which allows for physical inspection and disassembly of the product and the ability to make performance parameter measurements. This was not the case for this exercise. Designing a duplicate and identical motor under these conditions is daunting at best. Attempting to get all of the performance parameters within specification is problematical. For example, one generally does not design a motor for a specific inductance. Inductance generally falls out as a by-product of the size of the motor, number of winding turns, laminations, etc. The motor’s torque constant Kt ( N-m/amp) is generally determined by the motor’s dimensions, materials, air gap and winding turns. The Kt is directly proportional to the number of turns while the inductance is determined by the number of turns squared. So, one can see the dilemma of attempting to converge these two parameters. This was only one of the issues…getting a pure sinusoidal BEMF waveform (< 2% harmonic distortion for the 3, 5 th and 7th) was another of the major difficulties to solve. We will attempt to walk through the iterative and somewhat frustrating process of arriving at a successful series of motors. This paper is not intended as a design review but rather as lessons learned when unrelated design restrictions are added to the design process. The work began in May 2008 and was completed in January 2010. * Aeroflex, Hauppauge, NY Proceedings of the 41st Aerospace Mechanisms Symposium, Jet Propulsion Laboratory, May 16-18, 2012

Document

AMS_2012.pdf

54 Description of motors Let’s call the 3 motor configurations T, G and H to depict the three different motor types. Within each configuration there were minor design changes for various applications. Every motor was to be redundant. T – This motor configuration was a traditional design with the rotor on the inside and the stator on the outside. The design required a 0.75 slot per pole configuration, i.e. being an 18-slot, 24-pole, 3-phase, redundant machine. This configuration was necessary to meet a requirement that the phases were to be separated by 120 degrees mechanical and the primary and redundant phases to be separated by 60 degrees mechanical. The stator O.D. was given as 9.1427 to 9.1440 cm (3.5995 to 3.6000 inches) and the rotor I.D was 4.8 cm (1.9 inches). Two different stack lengths were required for two designs within this configuration. The motors were to be sinusoidally commutated using resolver and drive electronics which were not part of the task at hand. G - This configuration was an “inside-out” design, the rotor was on the outside and the stator was on the inside. The configuration for all 3 variations of the G motor design was as above, 24 pole, 18 slot, 3 phase, and redundant. The rotor specified O.D. was 12.6 cm (4.98 inches) and the stator I.D., which mounted to the program’s equipment, was given as 6.6040 to 6.6053 cm (2.6000 to 2.6005 inches). The different variations for this configuration were, again, in the stack length and motor performance parameters. H – These motors were also of an “inside-out” configuration with the same slot design as in the G motor. The primary difference was size and performance parameters. The required rotor O.D. was 15.51 cm (6.105 in) and the specified stator I.D. was between 11.430 to 11.431 cm (4.5000 to 4.5005 inches). Initial Major Design Restrictions and Design Concerns Restrictions: The restrictions listed below are not necessarily required by a designer to comply with the performance parameters, and were partly responsible for some of the difficulties encountered during the design process. It was assed that these restrictions were a mandatory requirement of the customer for achieving an “identical" motor. 1 - Magnet material must be Samarium Cobalt 2 - Lamination material must be Carpenter 49 (High Nickel Steel) 3 - Primary and Redundant windings must be spaced at 120 degree intervals for each of the three phases alternating every 60 degree increment 4 - Stators were not to have a skew 5 - 0.75 slot per pole configuration with 18 slots and 24 poles was required. Had these restrictions not been applied, several design techniques, such as 2.25 slot per pole design (24 poles 54 slots) without the angular separations, which were a more standard configuration for us, might have been used. Another familiar configuration, like a 72 slot (3 slots/pole) with a skew, might have also been employed. Given some degree of dimensional freedom, the Aeroflex zero-cog approach might also have been used thereby guaranteeing a low distortion sinusoidal BEMF waveform with very low drag and detent results. In other words, allowing a manufacturer to use more familiar designs, techniques, and materials while meeting the specific performance and environmental conditions can often lessen problem areas encountered than with the requirement to match exactly the mechanical and material design details as manufactured by another company.

Document

AMS_2012.pdf

55 Initial Concerns 1 - In many new design cases, existing laminations and their characteristics can be referred to laminations that have already been established and proven in previous designs. In this case the lamination designs needed to be designed from scratch. The task of deriving the exact geometry of the lamination, so important to duplication, was doubtful. The 0.75 slot/pole format would require investigating the magnetic flux details of the laminations using magnetic FEA analysis tools, which unfortunately is not always that precise. How close would we come? 2 – How would the resistance vs. inductance winding parameters balance against each other, both a function of the motor geometry as well as the copper windings? 3 - What is the BEMF waveform for the specified winding profile? Will it meet the 2% distortion requirement? Major Parameters for T Motor The development of the T motor was moderately painless and is illustrated to show how the T motors development was hoped to have also proceeded with the G and H configurations. It was considered painless because the performance parameters were more easily obtained without major modifications to any of the restrictions outlined above. Also, the T motor being the smallest of the 3 configurations made tooling for test easier to handle than the larger motors. Anyone having the experience of mounting rotors and stator with high energy magnets has sooner or later gotten his fingers bitten during the insertion of the rotor into the stator process if the proper tooling was not available. Special tooling “jacks” needed to be designed to allow the assembly into the test fixtures. The basic performance for one of the T motors is listed below: * Kt > 35 N-cm/amp (49 in-oz/amp) * R = 2.6 ohms within 10%, L = 8 mH within 25% * Friction torque < 1.8 N-cm (2.5 in-oz) at 10 RPM * Detent torque < 3.5 N-cm (5 in-oz) A view of this motor is shown in Figure 1. Figure 1. T Motor A significant obstacle to the development and delivery effort was that Carpenter 49 lamination material was not available in the sheet stock we needed which required a special order with a long delivery schedule. It was decided to make the early prototypes using M-15 lamination material, which was

Document

AMS_2012.pdf

56 available and whose characteristics were well known to us. The plan was to learn as much as possible before the Carpenter material was received. The usual manufacturing procedure to make a lamination stack is to stack and bond a number of 0.18 or 0.36-mm (0.007 or 0.014-inch) thick laminations, previously punched to shape, to make the desired stack height required. Since these laminations were of a new design, the technique used for these unique shaped laminations was to bond square sheets of 0.18-mm (0.007-inch) lamination stock to produce the height needed and cut the proper shape with an electrical discharge wire machine (EDM). The sheets are welded on the edges to make the electrical contact needed for the process. The non-recast EDM process produces a lamination stack with very well defined and smooth edges to the dimensions required. Some follow up machining, is required to obtain the 13 µm (0.0005 inch) tolerances required. As a conservative strategy the thought was to start with a projected high Kt and a high detent prediction as a conservative approach to become familiar with the 0.75 slot per pole configuration. Using the measured data from a prototype M-15 lamination motor, we would then modify the air gap and turns to bring both parameters into specification. The FEA Kt and detent predictions as well as their ultimate measured parameters are shown in Figures 2 and 3. Figure 2. T Motor FEA Kt Prediction Initial T motor Kt analysis and measurement data (Note: the program does not allow for separate primary and secondary windings so the 116 N-cm/amp (164 in-oz/amp) value is for both primary and secondary energized). Figure 3. Initial T Motor Detent Analysis and Measurement As can be seen the detent prediction was 28 N-cm (39 in-oz) and the measured detent on the prototype was 25 N-cm (35 in-oz). The Kt prediction and the measure Kt were spot on with 58 N-cm/amp (82 in-oz/amp). Both of these were very high as expected compared with the required values of 3.5 N-cm (5 in-oz) and 35 N-cm/amp (49 in-oz/amp). With these actual values increasing the air gap from 0.635 to 1.52 mm (0.025 to

Document

AMS_2012.pdf

57 0.060 inch) was predicted to achieve the detent and Kt parameters needed. The Carpenter material was now on hand and with the following air gap modification the results in Figures 4 were obtained with no surprises and were very close to the required values. Figure 4. T motor Kt and detent with 1.52-mm (0.060-inch) mangetic air gap To get some perspective on the windings, in order to achieve the required 2.6-ohm resistance it was necessary to use a winding of 5 wires of #29AWG and 3 of 28AWG for a total of 8 strands per coil. The inductance was fortunately within the range required (8 mH). It did, however, take several iterations of the winding turns to arrive at the final values of the design. So far so good! The T motor BEMF curves are shown in Figure 5. The harmonic content was measured using the FFT functions of the scope (not shown). Figure 5. BEMF waveforms with 1.52-mm (0.060-inch) magnetic air gap

Document

AMS_2012.pdf

58 The spectrum analysis showed the harmonics to be less that the 2% limit. This motor performed as expected and promised similar expectations that the G and H motors would be equally accommodating…so it was hoped. Major Motor Parameters for G Motor The G and H motors were of an inside out design (rotor on the outside). This was not considered as an obstacle for obtaining the same results as was obtained for the T motors. The form factor for the G motors is illustrated in Figure 6. Figure 6. G motor Major Specification Parameters * Kt > 61 N-cm/amp (87 in-oz/amp) * R = 1.17 ohms within 10%, L = 3.58mH within 30% * Friction torque < 11 N-cm/amp (15 in-oz/amp) at 105 RPM * Detent torque < 19 N-cm (27 in-oz) * % Harmonic Distortion < 2% The initial lamination stack was already made with M-15 material and was not used since the Carpenter material had now arrived. (This becomes significant later on.) The G motor, as with the T motor, began with a known smaller than expected air gap of 0.635 mm (0.025 inch). The FEA Kt was predicted to be 79.44 N-cm/amp (112.5 in-oz/amp) and the measured was 80.36 N-cm/amp (113.8 in-oz/amp) on this first prototype. The detent was predicted as 65 N-cm (92 in-oz). and the detent was measured at 64 N-cm (90 in-oz). The wave form, however, was not looking too good as seen in Figure 7:

Document

AMS_2012.pdf

59 Figure 7. G motor BEMF waveform The shape of the BEMF wave looked less like a perfect sine wave and harmonic distortion was viewed as a potential problem. The gap was then opened to 1.27 mm (0.050 inch) to lower the detent as well as the Kt with the hope this would also improve the distortion issue. The results were as follows: Kt dropped 80.36 N-cm/amp (113.8 in-oz/amp) to 65.2 N-cm/amp (92.4 in-oz/amp) Detent dropped from 69 N-cm (98 in-oz) to approximately 18 – 20 N-cm (25 – 28 in-oz) Figure 8. Motor with 1.27-mm (0.050-inch) air gap The waveform became a little less peaked but nevertheless was still apparently distorted and a concern. This was caused by the 5th harmonic and was above the 2% spec and was unacceptable.

Document

AMS_2012.pdf

60 Figure 9. G motor Harmonic distortion It was postulated that the lamination shape was perhaps deficient and that also shaping of the rotor magnets would perhaps contribute to improving the harmonic distortion. Inspection of the FEA analysis (Figure 10) did not show magnetic saturation in any part of the lamination material. Figure 10. Flux Density Comparison of Lamination Materials There were several other issues other than the distortion to address such as inductance and detent. It was hoped that correcting these deficiencies would help correct the distortion issues. A brief description of the inductance and detent improvement efforts are summarized in bullet form:  The inductance was too high so the turns were reduced from 45 to 38 (inductance is a function of the turns squared). To compensate for the loss of Kt due to the turns reduction, decrease the air gap. This worked as intended for the inductance and Kt, but the detent was now 3 times higher than required due to the air gap reduction.  A small skew was introduced into the lamination structure which reduces detent. The detent was now OK but the Kt dropped below the desired level due to the skew.  Used higher energy product Neodymium magnets instead of samarium cobalt. Kt improved but still too low  Bring coils to 40 turns to improve Kt with inductance still OK; attempt to go back to Samarium Cobalt magnets…Kt just 0.7 N-cm/amp (1 in-oz/amp) below the minimum; inductance marginal….parameters too marginal to atte mpt production. Went back to180 C Neodymium, ultimately used for production.  Back to distortion issue, solving may resolve the Kt, detent and inductance issues.

Document

AMS_2012.pdf

61  Distortion remained above the 2% level for all of the above modifications Some of the lamination styles that were tried to reduce the harmonic distortion are shown in Figure 11. Recognize that this is a time consuming process to build the stacks, EDM the laminations, machine, and coat with an insulator as well as wind and insert the coils. Figure 11. Pole shoe shaping, tooth gap, and magnetic span variations were tried. FEA analysis alone was not considered a viable approach alone in that the difference of distortion was only about 1%. Shapes tried were: wide boot height to narrow boot height, wider tooth widths, wide tooth to tooth gap to narrow tooth to tooth gap, wide magnet to magnet spacing to narrow magnet to magnet spacing - all provided no change in harmonic distortion to bring values below the 2% limit. There were also models made with the edges of the magnets rounded. At some point defeat must be acknowledged and a call for help initiated, be it self imposed or externally suggested. The suggestion was to call in help from an recommended and established consultant. This was accomplished and was followed by an impressive 14-page report from the consultant. The report offered a solution. The proposed solution was so deceptively simple that it was welcomed with great enthusiasm. At the same time, the report gratifyingly duplicated the non-sinusoidal BEMF results we were observing, as well as duplicating other motor parameters, so high expectations were anticipated. The solution based on a detailed FEA analysis (using a different analysis tool than the ones we were using) was to make the magnets as flat rectangles rather than the having the normal surface curvature mirroring the stator curvature as in traditional designs. A test motor was then wound, magnets purchased, and tested with the wonderful results that all of the parameters, including harmonic distortion were now within specification. A bonus feature was the detent was reduced from 18 to 9 N-cm (25 to 13 in-oz) giving a good margin. What then followed was the production manufacture and testing of the modified rotor design. Total disbelief ….the motor performed as poorly as it had before the rotor magnet change. What had happened? Our thoughts went back to the CP49 and the M-15 materials. We always felt (just experience), despite the FEA analysis showing the reasonable flux paths within the two materials (Figure 10) that the M-15 was the proper way to go given past experience with the CP49 Hi nickel materials annealing variations. What had happened is that the flat magnets were mistakenly tested with the original M-15 stator that, as you recall, was originally made but never used. The M-15 lamination stack had gotten mixed up with the CP49 lamination stack. To prove that the model that was successful with the flat magnets was using M-15, a known M-15 stator stack was then wound and tested. Back to the good results tested!

Document

AMS_2012.pdf

62 Tests with older curved rotors using the M-15 stator also showed the same excellent results indicating that the issue was with the CP49 material and not anything else. The lesson learned here is that FEA analysis is a path towards good designs but is not the end all. Practical machines with practical materials must ultimately be the final say. We stayed with the flat magnets because we had purchased the entire programs supply. In retrospect, had we had been able to have gone with the M-15 material (usually our preferred material for this type of motor) most of these delays and perturbations relative to the harmonics would not have occurred or been necessary. A photo of the G motor is shown in Figure 13. The data in Figure 12 shows the distortion improvement with the M-15 material from 2.79% to 1.57%. Fundamental 18.4 volts; 5 th Harmonic at 0.26 volts Harmonic distortion = 0.260/18.4 = 0.0157 = 1.57% Figure 12. M-15 Harmonic Distortion Data for the M-15 Material Figure 13. Photo of the G Motor - Rotor is encapsulated in titanium ring Major Motor Parameters for H Motor The H motor, also an inside out design, presented with a whole different set of issues. The initial design using CP49 and Samarium Cobalt magnets resulted so enormous a difference between performance and requirements that an initial drastic step needed to be taken before any iteration would be possible.

Document

AMS_2012.pdf

63 Figure 14. Geometry of the H motor The Kt, detent and distortion below represent the performance requirements: * Kt > 56 N-cm/amp (80 in-oz/amp) * R = 0.84 ohms within 7%, L = 2.28 mH within 30% * Friction torque < 19 N-cm (27 in-oz) at 272 RPM * Detent torque < 15 N-cm (21 in-oz) * % Harmonic Distortion < 8% The initial lamination stack was made with CP49 material. The waveform and detent were much worse than seen of the G motor and the decision to use M-15 at the outset was immediately made. The result for the first CP49 H motor is shown in Figure15. The Kt was measured at 183 N-cm/amp (259 in-oz/amp) and the detent was over 141 N-cm (200 in-oz). Figure 15. BEMF Curve with CP 49 and 0.635-mm (0.025-inch) air gap The first attempt using M-15 material resulted in a Kt of about 79.8 N-cm/amp (113 in-oz/amp) with the detent being approximately 64 N-cm (91 in-oz). The decision to skew the stator was made to reduce the detent and smooth the waveform with the results shown in Figure 16.

Document

AMS_2012.pdf

64 Figure 16. BEMF for H motor with M-15 Material Again, exploring the flux density differences between the FEA analyses of the materials (Figure 17), there again did not seem to be much difference to cause this great disparity of performance for the H motor due to material as seen between Figures 15 and 16. Figure 17. Comparison of lamination material for the H motor This motor was most critical in terms of balancing the flux carrying capability of the laminations with the amount of copper area available affecting Kt, inductance and resistance. The motor needed a full complement of 40 turns (6 of #28 and 2 of #29 AWG wire) to conform to the specification requirements (Kt, resistance and inductance). A bonus of the skew was that the detent torque was reduced to about 1/3 of specified allowable which was well received. The wire would not fit within the area needed for the lamination design. (Several tooth widths were tried to arrive at the final configuration). Not a quick task when the programming, EDM process, lamination coating and winding time factors are considered. As a result, the motor could not be wound and inserted in the normal production manner. Normally, a string of coils per phase is wound on a mandrel, loosely tied to hold each coil of 40 turns together and then inserted into the stator as whole coils placed wire by wire into each stator slot. In this case the slot fill factor is so high that this would not allow the coil bundles to fit into the slots, just missing by a few percent. The stator needed to be hand wound, turn by turn, into the slots allowing each turn to be tightly fit into the slot maximizing the space available with just the right length of wire to make the resistance specified. Under these conditions it is necessary to have a very patient and amiable technician in your employ!

Document

AMS_2012.pdf

65 Figure 18. Photograph of the H motor with the lamination skew Summary With the above trades and compromises, all of the motors styles ultimately made their way into successful production cycles. In summary, what was needed to be changed from the 5 initial restrictions (which were not specific to the performance requirements) in order to meet the performance criteria is as follows: 1 - Magnet material must be Samarium Cobalt - In one case (G motor) higher energy product Neodymium magnets needed to be substituted. 2 - Lamination material must be Carpenter 49 (High Nickel Steel) – In two cases silicon steel was needed to replace the high nickel steel. 3 - Primary and Redundant windings to be mechanically spaced at 120 degree intervals for each of the three phases alternating every 60 degree increment - Preserved 4 - Stators were not to have a skew – In one case (H motor) a partial skew was required for detent and waveform conformance. 5 - Parameters must meet the tight specification values – In one case some relief was required for the inductance values 6 - 0.75 slot per pole configuration with 18 slots and 24 poles was required. - Preserved The major lesson learned from this project is to have give and take discussions between parties to determine which of the specification requirements are truly pertinent to the performance as opposed to those which would be nice to have based on existing hardware.

Document

AMS_2012.pdf

67 Development of the Vibration Isolation System for the Advanced Resistive Exercise Device Jason H. Niebuhr* and Richard A. Hagen** Abstract This paper describes the development of the Vibration Isolation System for the Advanced Resistive Exercise Device from conceptual design to lessons learned. Maintaining a micro-g environment on the International Space Station requir es that experiment racks and major vibration sources be isolated. The challenge in characterizing exercise loads and testing in the presence of gravity led to a decision to qualify the system by analysis. Available data suggests that the system is successful in attenuating loads, yet there has been a major component failure and several procedural issues during its 3 years of operational use. Introduction Resistive exercise (weightlifting) is prescribed for crew of the International Space Station (ISS) to maintain muscular strength and bone density during long duration space flight. However, the repetitive nature of exercise induces vibratory loads that can degrade the sensitive ISS micro-g environment and reduce its structural fatigue life. The Interim Resistive Exercise Device (iRED) was flown aboard the ISS from 2002 through 2008 as the primary means for resistive exercise. Ongoing maintenance issues, increased performance requirements, and the need for a Vibration Isolation System (VIS) eventually drove the need for a new clean-sheet exercise machine, called the Advanced Resistive Exercise Device (ARED) [1]. ARED creates resistance with a pair of vacuum cylinders that are connected through a variable length lever to an adjustable height bar as shown in Figure 1. The exerciser can stand, sit, or lie on the platform or bench (not shown) to perform a variety of exercises. * Formerly with NASA JSC, now with Apogee Engineering at the US Air Force Academy, Colo. Spgs., CO ** NASA Johnson Space Center, Houston, TX Proceedings of the 41st Aerospace Mechanisms Symposium, Jet Propulsion Laboratory, May 16-18, 2012 Vacuum Cylinders Figure 1. The Advanced Resistive Exercise Device Lever Arm Length Determines Load Pivot PointAdjustable Height Ba r Platform

Document

AMS_2012.pdf

68 The primary micro-g requirement is a frequency-dependent acceleration limit at the US Lab and Columbus module rack interfaces. Structural dynamics transfer functions are available to quantify the acceleration at those racks as a function of input load at the ARED location within the ISS. With the assumption of equal loading in all axes, the acceleration limit and transfer functions can be taken together to consider the requirement in terms of allowable load. A similar frequency-dependent load limit is required to preserve ISS fatigue life. The composite allowable load enveloping both micro-g and structural fatigue requirements is shown in Figure 2. This is a simplified and conservative view of the requirements, yet highly insightful. The graph highlights the most significant challenge to the design; a major ISS structural mode exists near 0.24 Hz which can be seen as a steady state allowable load of only 1.7 N (0.4 lbf) between 0.18 Hz and 0.28 Hz. As shown by the shaded area of the graph, this is well within the 0.09 Hz – 1.18 Hz range of exercise frequencies recorded during the ARED man-in-the-loop test (MILT) [2]. Analysis of the ARED in a non-isolated configuration suggests that exercise loads can be as a high as 67 N (15 lbf) at 0.24 Hz, requiring attenuation by more than an order of magnitude. Analytical Approach A rigid body dynamic model was created in the motion analysis software visualNastran 4D from MSC.Software Corporation. This software was chosen over general numerical computing codes for its native 3D visualization capabilities and ease of use. This made it possible to examine many different design concepts and better understand how they worked. Though somewhat controversial, it was felt that the cost and complexity of an actively-controlled isolation system wasn’t justified. Being able to visually simulate system dynamics was essential in convincing both technical and non-technical stakeholders alike that a passive isolation system would be sufficient. Model development focused on three distinct elements: a geometrical and mass model of the ARED, a geometrical and mass model of the exercising crew member, and a mathematical model of the VIS. Development of the ARED model was straightforward as solid geometry was imported from computer-aided design (CAD) software after being simplified into 12 major subcomponents by suppressing details like fasteners, fillets, and lightening pockets. Detailed mass properties of the subcomponents were retained from the original CAD model. Appropriate constraints were applied to the model to simulate fixed, rotational, and sliding joints. Capturing this level of detail kept the computing requirements manageable. The exercise model consists of a CAD model human constrained to simulate exercise motion when driven by displacement. Figure 3 illustrates how the human model and subcomponents of a non-isolated ARED Figure 2. Allowable Loads Requirement and Exercise Frequency Range Exercise Frequency Range 0.09 Hz – 1.18 Hz

Document

AMS_2012.pdf

69 both move throughout the stroke of a squat exercise. The y-axis is parallel with the long axis of ARED, the z-axis positive toward the exerciser, and the x-axis completes the right-handed Cartesian system. Exercise variables include type, stroke, crew size, and frequency. The exercise types and stroke lengths range from a squat moving nearly 90% of the body’s mass through 0.70 m (27.5 in) of stroke to a neck flexion that moves less than 6% of the body’s mass through less than 0.23 m (9 in) of stroke [3]. However, the squat, dead lift, straight leg dead lift, and heel raise are worst-case and envelope all other exercises in the analysis because they move the most amount of body mass over the greatest distance. The 95 th percentile American male represents the heaviest crew analyzed while the 5th percentile Japanese female is the lightest [4]. Data from the ARED MILT defines exercise stroke profiles and statistically quantifies exercise frequency. Figure 4 shows a box plot of the exercise frequencies for 3 of the worst-case exercises. The middle dash represents the average frequency, the box envelopes ±1 standard deviation, and the whiskers represent the extreme value recorded. Either idealized sinusoidal or real stroke profile data derived from testing can be used to drive the human exercise model. Use of idealized sinusoidal data tends to be more severe because spectral energy is focused into a singular frequency, whereas real stroke profiles capture the inherent variability in human exercise. The idealized sinusoidal approximation becomes worse at lower exercise frequencies that are typically distinguished by longer pauses between exercise cycles. The VIS model consists of mathematical definitions of spring and damping rates. Many different concepts were examined and several lessons learned: 1) The enveloping exercises only excited the system in 3 degrees of freedom (DOF) planar motion: Translation along the y- and z- axes, and rotation about the x-axis. Trading mounting rigidity for Figure 3. Non-Isolated ARED and Human Model Motion During Squat Exercise Y Z Figure 4. Box Plot of Frequency Statistics for 3 Exercises

Document

AMS_2012.pdf

70 isolation by reducing stiffness in only these 3 DOF is both sufficient and provides a certain level of safety when suspending nearly 567 kg (1250 lbm) in a micro-g environment. 2) Each enveloping exercise excites the 3 DOF differently. Squats and heel raises primarily excite the y-axis and rotation about the x-axis. Dead lifts, particularly the straight leg variant, tend to excite all 3 axes. 3) Decoupling the VIS DOF by lining them up with the ISS axes made the design of the system much easier. Changes to spring or damping rates in one axis only affect that axis. 4) Analysis lessons learned 2 and 3 combined with the knowledge that structural dynamics transfer function response varied in each axis allowed the stiffness and damping of the system to be optimized against the micro-g and structural fatigue requirements. A key principle in vibration isolation is choosing isolator properties such that the natural frequency of the system is below the forcing frequency. In this case the forcing frequency is the aforementioned exercise frequency range from 0.09 Hz to 1.18 Hz. The natural frequency, fn, is given by: ௡=1 ߨ݇ ݉)ݖܪ( Where k is the spring rate and m is the mass. Additional damping was added to more quickly dissipate energy in the system. Transmissibility, T, is a measure of the amplification of the isolation system or ratio of output to input. It is given by: ݂ ௡൰ଶ ݂ ௡൰ଶ ݂ ௡൰ଶ ቇଶ Where f is the forcing frequency and ζ is the ratio of damping to critical damping. Values of T > 1 signify an amplification of input load, whereas values of T< 1 signify an attenuation of input load. The chosen spring rates, percent damping, resultant natural frequencies, and amplification at various frequencies are shown in Table 1 for each of the 3 DOF. Table 1. VIS Properties and Isolation Qualities Axis Spring Rate Damping Natural Frequency Amplification, T (0.09 Hz ) Amplification, T (0.24 Hz) Amplification, T (1.18 Hz) Y 70 N/m (0.4 lbf/in) 11% 0.06 Hz 0.7 0.08 0.01 Z 175 N/m (1 lbf/in) 10% 0.09 Hz 1.9* 0.19 0.02 X (rotation) 0.6 N•m/deg (5.2 in•lbf/deg) 17% 0.05 Hz 0.4 0.08 0.01 * Amplification based on 0.11 Hz minimum deadlift exercise frequency since the squat exercise creates minimal excitation in z-axis

Document

AMS_2012.pdf

71 The spring rates, particularly in the y-axis, had to be extremely low to keep the natural frequency of the system from overlapping the exercise frequency range and to meet the micro-g and fatigue requirements. The z-axis motion is generally only a fraction of what is seen in either of the other two axes which is why the spring rate and the resultant natural frequency was increased to nearly encroach upon resonance in a worst-case scenario. This was necessary to minimize the motion envelope; one of several other issues that developed as a result of the low spring rates. Inertial accelerations applied to ARED and friction forces also became issues. While ARED is allowed to move in response to exercise, the inside of the ISS is a relatively limited space and there was a possibility that a crew member could hit their head while exercising. Travel limits were setup to prevent this from occurring. Another consideration was the effect of inertial accelerations applied to the ARED system as a result of events like reboost and docking. These events are enveloped by a 0.4 g load factor applied in any direction and cause the ARED to move across the motion envelope and collide with the end of travel limits. This potentially overloads the interface and exceeds a transient acceleration micro-g requirement. Much larger dampers, called snubbers, were sized and located at the travel limits to attenuate those loads. They were selected by calculating the velocity of ARED due to the inertial acceleration and sizing them to efficiently dissipate the energy. Friction had to be considered because, if too high, it had the effect of causing the entire system to “inch-worm” until it reached one end of the motion envelope, potentially reacting repeatedly against the end of travel limits and not re-centering. A special effort was made during the mechanical design phase to select components with low friction as well as to include its effects within the analytical simulation. Mechanical Design To stay within cost and schedule constraints many commercial off-the-shelf (COTS) components are used in the design of the VIS. It is a nearly symmetrical assembly composed of 2 plate assemblies connected by a beam as shown in Figure 5. Figure 5. The Vibration Isolation System VISARED Y Plate Z Plate Rotation HousingInterface to ARED Rotation Arm

Document

AMS_2012.pdf

72 Each side has half of the linear and rotational bearings that constrain motion to 3 DOF and the isolation components that attenuate exercise loads. Figure 6 shows that the Y Plate contains linear rails, springs, and dashpots that connect to the Z Plate, allowing motion in the y-axis. The Z Plate has the same components that allow the Rotation Housing to move in the z-axis. Lastly the Rotation Housing has a bearing to allow rotational motion and an arm to convert linear spring and damper forces into torque about the x-axis. The allowable deflection between the left and right plates was specified as ±0.051 mm (±0.002 in) to reduce the possibility of the VIS binding under load. This placed a burden on manufacturing since interface joints and bearing rail surfaces were held to extremely tight tolerances. To minimize misalignment due to accumulated tolerances, there are only 11 VIS structural components from one side of ARED to the other. The VIS was assembled in a fixture to verify alignment and set the spacing of the two halves. The linear bearings that constrain motion were chosen from amongst several candidates. The DualVee guide wheel design exhibited low friction, robustness, and ease of maintenance. Rails are made from 420C stainless steel while the bearings are made from 440C stainless steel, simplifying the material certification process. The low bearing friction is achieved by using a bearing with shields instead of seals, high quality Rheolube 2000 grease, and just enough bearing preload to maintain linear rigidity. The preload in the bearings is adjusted until the maximum load needed to move a given plate is 1.3 N (0.3 lbf). The double row angular contact design of the DualVee provides a robust bearing that can handle large static loads. Finally, by separating the bearing elements from the track as shown in Figure 7, maintenance is significantly reduced since dirt and debris attracted to the rails does not come in contact with the ball bearings. They also have a better tolerance for bearing misalignment than other linear bearing systems tested. Linear rails allow the Z plates to move in the y-axisLinear rails allow the Rotation Housin gto move in the z-axis A rotational bearing allows motion about the x-axisRails Dashpots Springs Figure 6. The Left Plate of the Vibration Isolation System Motion MotionMotion Figure 7. DualVee Bearing and Rail

Document

AMS_2012.pdf

73 Rotational motion of the VIS is constrained with a SKF double row angular contact bearing that also uses shields instead of seals and is lubricated with Rheolube 2000. Low rate springs attenuate exercise loads in all axes. 17-7 PH stainless steel extension springs are preloaded against one another to maintain a constant spring rate across the motion envelope and keep ARED centered. Compression springs were not favored because the combination of length and low spring rate would have required a support to prevent buckling. It was felt that the resultant friction would have reduced spring life, increased load variability, and perhaps resulted in a noise problem. The dashpots act to dissipate energy in the system by reacting to motion with a force proportional to velocity. Airpot dashpots were chosen for their simple design, use of an inert working fluid (air), adjustability, and advertised long life. A similar component is successfully used in the Active Rack Isolation System that pr otects ISS experiment racks from vibration. The dashp ots consist of a carbongraphite piston in a borosilicate glass cylinder with Rulon-lined ball joints at each end of the piston rod to allow for misalignment as shown in Figure 8. An adjustable screw orifice with a check valve adjusts the damping rate in the pull direction. The dashpots are wrapped with Kapton tape to contain debris in the event of breakage. The snubbers that attenuate impact loads from reboost and docking events are located to engage at the end of the motion envelope. Oil-filled models from ACE Controls were chosen because of their compact size and adjustability. The rotation axis snubbers are shown in Figure 9. Figure 8. Airpot Dashpot Adjustable Orifice Graphite Piston Glass Cylinde r Figure 9. Snubbers Mounted in the Rotation Housing

Document

AMS_2012.pdf

74 A lock-out pin inserted by the crew passes through the rotation arm and both plates, anchoring the VIS in the centered position to protect it from damage when not in use. In addition to creating simple crew interfaces, significant effort was expended in designing the VIS to be easy to inspect, repair, or even upgrade if needed. All of the isolation components are accessible by removing covers and serviceable using standard ISS tools. Individual components are replaceable instead of requiring large subassemblies to be kept as spares. The VIS attachment to the seat track in ISS is designed to accommodate misalignment due to initial tolerance variations or changes over time due to temperature or pressure fluctuations. This is accomplished by attaching at 3 points with sliding trunnions through spherical bearings as shown in Figure 10. This statically determinant mounting method does not transfer torque to the relatively torque-intolerant seat track. Installation was greatly simplified and crew can remove and replace the entire ARED and VIS system easily for relocation, emergencies, or storage access. Test and Qualification Approach Testing VIS performance is difficult in the presence of gravity. An earlier development program attempting to isolate the iRED tested the system lying sideways on an air bearing floor. The human subject was also supported sideways to perform exercises, which was very difficult and cast doubts on the validity of the results. A parabolic flight pattern on a reduced gravity aircraft provides a more realistic environment to perform exercises, but the window of weightlessness is too short and given the extremely low spring rates, there is significant risk of hardware damage during the pull-out phase of flight. Even testing on-orbit may not capture worst-case exercise scenarios due to natural human variability. While a test of system level performance was impractical, a test program was developed to: 1) Verify that components meet performance requirements and adjust those with variable settings 2) Life cycle test the system to identify and resolve potential issues 3) Gather component performance data for the system level qualification analysis model The test stand consists of a Motion Science MS700 electromechanical actuator (EMA), Sensotec Model 31 load cell, and a string potentiometer all mounted on a t-slot table as illustrated in Figure 11. Misaligned X YEach spherical bearing is attached to seat track and independently constrains 2 linear DOF. As a system they constrain all 6 DOF. If one spherical bearing (attachment point) moves, a combination of sliding (trunnion) and rotation (spherical bearing) at all 3 locations self-aligns the system. Figure 10. VIS Attachment to ISS Seat Track No Misalignment

Document

AMS_2012.pdf

75 Component level tests primarily involved the rails, dashpots, and snubbers. Several styles of rails were tested to find which provided the lowest friction. The rail bearings in the VIS are mounted on eccentric bushings to provide preload adjustment and tuned to provide rigidity while minimizing friction. The dashpot testing focused on finding the right setting to meet the target damping rate. Each model was cycled through 7 discrete constant velocities at each setting. The average force was recorded for each constant velocity, plotted, and fit to calculate the damping rate as shown in Figure 12. The dashpot behavior fit well to the expected characterization of force being linearly proportional to velocity. Figure 11. Vibration Isolation System Test Stand T-slot TableElectromechanical Actuato r String Potentiomete r Test Hardware Load Cell Figure 12. Component Test Results: Force vs. St roke at 27.9 cm/s (11 in/s) and Average Force vs. Velocity with Linear Fit for Y Dashpot Push Stroke (Relieved by Check Valve) Average Force at 7 Constant Velocities Pull Stroke Average Load = 5.1 N (1.11 lbf) Force vs. Stroke for 10 cycles

Document

AMS_2012.pdf

76 Snubber testing revealed a shortcoming of the EMA test stand in that it wasn’t able to supply a sufficient force at high velocities to characterize the snubbers. A simple pendulum-based test stand was developed as shown in Figure 13. It was easy to design, manufacture, and adjust to deliver the correct mass and impact velocity. Testing revealed that the snubbers are capable of dissipating the resultant energy from ISS reboost and docking events without overloading the ISS seat track interface. “Lead the fleet” life cycle testing is still underway with the EMA test setup. Each axis of the VIS is cycled through 300,000 cycles per year. Several observations have been noted to date: 1) The linear rails (of 420 stainless steel) have developed surface corrosion during life cycle testing after grease wore away. This may be an issue of test conditions being worse than service conditions as the test cycles start and stop in the same place, pushing the grease away at the cycle peaks. Regardless, the planned rail maintenance intervals were changed from once per year to 6 times per year as a measure of caution. 2) Graphite deposits were found on the cylinder walls after 100,000 cycles on the rotation axis dashpots. The manufacturer had seen this in other applications and there was no change in performance, so testing continued. 3) During testing of the rotation axis, debris was noticed on the rotation arm at the ball joint attachment. No changes in performance were apparent, so testing continued. The component test data was used to update the damping properties of the dashpots and friction properties of the rails within the analytical model. As the VIS design matured, its own mass properties were added to the model as well. A case matrix was developed to include runs with both sinusoidal and real stroke profiles, exercise frequencies at the minimums, maximums, averages, and ISS structural mode frequencies for all 4 of the major exercises for a total of 51 cases. The load and torque time histories from this for each case were sent to Boeing for a coupled loads analysis within the ISS structural dynamics finite element model. Snubbe rV Pendulum Weights Test Article Close-up of Test Article Load Cell Angular Encoder Figure 13. Pendulum Test Stand for VIS Snubbers

Document

AMS_2012.pdf

77 The results predicted successful performance of the VIS in both Node 1 and Node 3 of the ISS as shown in Figure 14 and the VIS achieved flight qualification. Operational Performance The ARED and VIS were flown to the ISS aboard STS-126 (ULF2) in November 2008. Assembly occurred during ISS Expedition 18 over 4 days in late 2009, with checkouts following. Figure 15 shows the ARED and VIS installed in Node 1 of the ISS with NASA Astronaut T.J. Creamer performing a dead lift exercise. On-orbit accelerometers have shown that the VIS is meeting its micro-g and loads requirements, though no localized vibration surveys have been done to validate analytical predictions. After 2 years of operational experience with no corrosion, the planned rail maintenance interval was changed from 6 times per year to 4 times per year. After additional evaluation the time interval between rail cleaning may be increased even more. While the VIS has performed very well overall, it has not been free of problems. A major mechanical problem and two procedural issues have been identified: Problem 1: After just 6 months of operation, a rotation axis dashpot failed at the swaged connection between the stainless steel connecting rod and aluminum rod end ball bearing. When the failure wasn’t immediately ARED at Node 3 vs Node 1 0.0010.010.1110100 0.01 0.1 1 10 100 Freq (Hz.)MicroGARED Req. Envelope ARED N1 Envelope ARED N3 Figure 14. Results of Coupled Loads Analysis for 51 Exercise Load Cases [5]

Document

AMS_2012.pdf

78 noticed, subsequent ARED use resulted in breakage of the dashpot’s glass cylinder as shown in Figure 16. The Kapton tape overwrap contained most of the debris. ARED was inoperable for a month until spares arrived. Since then, the swaged connection of the rotation axis dashpot has failed on 3 separate instances. However with spares available, downtime has been no more than a day. The failure appears to be most likely caused by several deficiencies in the dashpot design combined with the misalignment between the connecting rod ends as the rotation arm sweeps through its range of motion. A root cause analysis was undertaken to identify weaknesses in the design [6]. While the design does not have the robustness desired it was concluded that the dashpot has not been used in a manner that exceeded its specifications. Several design improvements for the rotation axis dashpot have been developed and approved for implementation: 1) Piston height has been increased to reduce binding that can cause excessive wear, generate debris, and increase friction. A groove around the piston has also been added to capture any debris that is generated. 2) The connecting rod ends have been redesigned to increase strength and wear resistance. Inhouse testing revealed several swages that tested at lower static loads than expected (but not below the design limit load.) Swaged connections are eliminated in favor of a stainless steel ball end threaded into the shaft. The ball ends will fit into a two-piece Vespel socket for durability. This failure was not observed during life cycle testing although a clue in debris generation at the ball joint was noticed as previously discussed. The life cycle test profile velocities and ranges are based on Figure 15. NASA Astronaut T.J. Creamer Exercises on ARED (Image Courtesy of NASA)

Document

AMS_2012.pdf

79 averages and do not capture peak velocities and hence the peak dashpot loads seen during operation on the ISS. Problem 2: Procedural lapses have resulted in the VIS being locked during exercise which is clearly detected by ISS accelerometers and violates micro-g requirements. This seems to happen mostly with new crew unfamiliar with ARED operation. New labels and training procedures have so far addressed these issues. Problem 3: Incorrect installation of the lock-out pin dislodged one of the rotation axis springs, but this was easily fixed by the crew. New labels and training procedures have so far addressed these issues. Lessons Learned Lesson 1: The use of motion simulation software was not only technically sufficient, but the visualization capabilities proved to be instrumental in bolstering confidence in a novel design. Lesson 2: Qualification by analysis can be a viable alternative if physical testing cannot provide a clear assessment of performance. However, careful attention must be given to risk mitigation. Figure 16. Broken Rotation Axis Dashpot (Image Courtesy of NASA) Broken Glass Broken Rod End

Document

AMS_2012.pdf

80 Lesson 3: Designing for ease of repair has been critical in addressing unforeseen issues. The rotation axis dashpots have proven to be easy to replace on orbit allowing for quick recovery from failures. This not only results in less time spent by crews on orbit, but by ground support crews as well since fewer procedures need to be developed. Lesson 4: Despite training and consideration of designing for humans, hardware may still be misused. Assume hardware will require on-orbit repair and design accordingly. Lesson 5: “Lead the fleet” life cycle testing should only be considered when the design is flexible enough that problems can be easily diagnosed when they arise and can also be easily fixed on-orbit. Lesson 6: Carefully consider life cycle test design for hardware used by humans. The profile used in testing the VIS revealed a corrosion issue that hasn’t developed on-orbit, but didn’t predict a hardware failure that happened after only 6 months of use. The profile is conservative in the number of cycles but doesn’t capture peak dashpot loads. Conclusions Despite several on-orbit problems, there have been numerous successes. A comparatively simple passive isolation system relying on COTS components was qualified by analysis and has demonstrated effectiveness saving untold development and sustaining engineering costs. Thorough characterization of load cases and attention to design for repair significantly mitigated risk and this has paid its dividends in reducing down-time while recovering from on-orbit failures. References 1. Lamoreaux, Christopher D., and Mark E. Landeck. "Mechanism Development, Testing, and Lessons Learned for the Advanced Resistive Exercise Device." Proceedings of the 38 th Aerospace Mechanisms Symposium , (May 17-29, 2006), pp. 317-330 2. Bentley, Jason R., et al. “Advanced Resistive Exercise Device (ARED) Man-In-The-Loop Test (MILT).” NASA TP-2006-213717, May 2006. 3. Guilliams, Mark, Mike Rapley, and Nahom Beyene. “Resistive Exercise Description Document.” JSC 29558, January 2002. 4. National Aeronautics and Space Administration, International Space Station Flight Crew Integration Standard, SSP 50005C, December 1999, pp. 3-45 5. Laible, Michael. “ARED Microgravity Assessment in Node 3 at Node 1 Port.” Boeing EID684-13762, November 2009. 6. Zamaitis, J.A. and G. Szymczak. “ARED X-Dashpot Performance Report.” EM-ARED-066, November 2011.

Document

AMS_2012.pdf

81 Passive Thrust Oscillation Mitigation for the CEV Crew Pallet System Matthew Sammons*, Cory Powell**, Joe Pellicciotti+, Ralph Buehrle++ and Keith Johnson++ Abstract The Crew Exploration Vehicle (CEV) was intended to be the next-generation human spacecraft for the Constellation Program. The CEV Isolator Strut mechanism was designed to mitigate loads imparted to the CEV crew caused by the Thrust Oscillation (TO) phenomenon of the proposed Ares I Launch Vehicle (LV). The Isolator Strut was also designed to be compatible with Launch Abort (LA) contingencies and landing scenarios. Prototype struts were designed, built, and tested in component, sub-system, and system-level testing. The design of the strut, the results of the tests, and the conclusions and lessons learned from the program will be explored in this paper. Introduction The Constellation Program aimed to send human explorers back to the Moon and beyond as part of the Vision for Space Exploration. The CEV, also called Orion, was the proposed human spacecraft capsule for the program. The launch environment of the proposed mission included the TO phenomenon described below. It had been determined that the effects of the TO event must be mitigated for crew safety and operational reasons. CEV Background Originally conceptualized as a six-man vehicle, the CEV was ultimately designed to support a four-man crew on trips to the Moon, Mars, and other destinations in the solar system as part of the Constellation Program. The interior of the Crew Module (CM) portion of the CEV contains a floating crew pallet structure supported by eight struts upon which the crew seats are installed. See Figure 1 for an image of CM interior design and crew seat pallet with struts. The goal of the NASA Engineering and Safety Center (NESC) team was to develop a Crew Impact Attenuation System (CIAS) to mitigate the TO effects on the crew. TO Description Like all launch vehicles, the launch vehicle dynamic environment is a significant input to the overall payload launch loads. The Ares I First Stage has a small oscillation in thrust at a frequency band centered at approximately 12 Hz ± 2.5 Hz, for about 10 seconds late in booster burn. Since the overall launch vehicle resonated with the input frequency of the boosters, the crew pallet struts had an additional requirement to mitigate the thrust oscillation resulting from the vehicle resonance. * ATK Aerospace Systems, Beltsville, MD ** NASA Goddard Space Flight Center, Greenbelt, MD + NASA NESC (Goddard Space Flight Center), Greenbelt, MD ++ NASA Langley Research Center, Hampton, VA Proceedings of the 41st Aerospace Mechanisms Symposium, Jet Propulsion Laboratory, May 16-18, 2012

Document

AMS_2012.pdf

82 Figure 1. CEV CM with Crew (left); Crew Pallet with Struts (right) Requirements The TO isolation solution was driven by the requirements and constraints imposed on it derived from evaluations of a spring-based isolation system’s impact on the crew in a TO event, LA scenario, and landing scenario. In addition, volumetric constraints and project constraints (resources, time) led to the design solution. This section will detail the results of the TO event, landing scenario, and LA evaluations. CM Interior Environment There are eight struts total: four in the X direction, launch axis, and two each in Z and Y directions. The CM baseline strut was modeled and tested using wire bender struts that absorb landing loads by dissipating energy through plastic deformation of steel wires – they have a one-time stroke during landing. As a pallet-based isolation system was selected (as opposed to seat-based), it was determined that the struts would need to be in series with the wire bender struts. Analyses showed that TO Isolators only needed to be present in the four X struts. Crew Impact from TO Event A requirement resulting from an investigation by the TO Focus Team and imposed by the Crew Office was to reduce crew response acceleration levels during the TO event to 0.25 g maximum at 12 Hz. This became the primary design driver for the isolation system. The NESC team used the Brinkley Dynamic Response model to assess the likelihood of injury to the crew in both the landing and LA scenarios. For more information regarding the Brinkley injury risk criteria, see Reference 1. Isolation Frequency Knowing the TO frequency band of 12 ± 2.5 Hz, a NASTRAN® coupled loads model was used to evaluate the effectiveness in g-reduction at different isolation frequencies. Following the theory of load transmissibility, if the pallet system frequency is greater than that of the input, dynamic amplification will occur. By dropping the pallet system frequency below that of the input, dynamic amplification can be eliminated and transmissibility can be less than 1.0. X Struts (+Z side) K = 600 lb/in spring X Struts (-Z side) K = 1200 lb/in spring +X +Y +Z 12 34 56 78

Document

AMS_2012.pdf

83 It was concluded that an isolation frequency of 5 Hz or less is required to reduce the crew acceleration responses to 0.25 g. A 4.5 Hz isolation frequency was selected as optimal, as it represents a good balance between deflection and transmitted dynamic load. This isolation frequency of the pallet is achieved by placing linear springs in series with the 4 X-axis struts. Strut Deflection During the TO event and while isolated at 4.5 Hz, the pallet will translate between 0.5 cm – 1 cm (0.2 in – 0.4 in). Therefore, the TO Isolator stroke was designed to allow for slightly more than the predicted max of 1 cm before hitting a hard stop. In order to eliminate Isolator deflection during ground operations, the mechanism was designed to be preloaded through launch until the TO event occurs between 3.25 g’s to 4.5 g’s quasi-static load. Volumetric Constraints Volumetric constraints were imposed by the available strut length and the crew size. It was determined that, being in-line with the CM struts, a maximum length of 39.4 cm (15.5 in) was not used for the landing event and allowable for the TO Isolator design. The cylindrical diametric size had to be kept to a minimum to ensure that it did not contact a crew member’s shoulder. Computer Aided Design models proved this to be less than 11.4 cm (4.5 in). Isolator Impact on Landing Loads LS-DYNA®, kinematic analysis software, was used to investigate the effect from the TO Isolator on landing loads experienced by the crew and compared to the baseline. The initial model consisted of CM struts in series with an isolation spring; subsequent models included a damper to address potential issues with LA. With a 30g input, it was found that introduction of the TO isolation springs does not significantly or detrimentally affect accelerations transmitted to the pallet. However, the isolation springs have a detrimental effect on the stroking of the baseline struts, which ideally should be held to a minimum. Preliminary analyses showed that the Isolators need to be locked out during a landing event to minimize CM strut stroke while keeping the Brinkley model’s injury risk probability to an acceptable level. Isolator Impact on LA Loads Although other launch analyses were done using NASTRAN®, the effect of the Isolators on crew loads in the contingency case of a LA were evaluated with LS-DYNA® software. This was done so that crew response of the abort event could be determined while keeping the Isolator kinematic motion in the model. Without damping, the seat accelerations were found to be unstable and grew without bounds. With just 44.5 N (10 lbf) of Coulomb friction in parallel with the Isolator, the accelerations were sufficiently attenuated and decayed once the abort loadings ended. Since actual damping levels throughout the crew module are unknown and difficult to determine it was not useful to attempt a detailed damping study for the isolation springs. Instead, it was decided to utilize a small amount of Coulomb damping in the model and then insure that the design of the isolation spring had a mechanism for providing a deterministic level of damping. TO Isolator Strut Design Design Overview The TO Isolator strut was designed as a passive spring and damper system that would be active during ascent and locked out during landing. It is mounted in-line with the X-axis CM struts. For hardware testing, the Isolator was mounted to the wire bender via a 1.905 cm – 16 (.750”-16) threaded interface, see Figure 2.

Document

AMS_2012.pdf

84 Figure 2. TO Isolator Strut In-line with Wire Bender Strut The principal details of the TO isolation design are:  Two 105 kN/m (600 lbf/in) springs (+Z) and two 210.2 kN/m (1200 lbf/in) springs (-Z) were required to achieve a balanced 4.5 Hz pallet system frequency.  Springs are extended 4.123 cm (1.625 in), preloading the Center Rod in compression, so that the Isolator does not unseat prior to the TO event. During launch, the rod will not unseat from the housing thereby keeping the springs out of the load path until just before the TO event’s quasi-static load is reached. The pallet then floats at 4.5 Hz for the range of 3.25 to 4.5 g’s.  During the TO event, cyclical motion of the Isolator will occur in the range of 0.5 cm – 1 cm ( 0.2 in – 0.4 in). The design allows for motion up to 1.59 cm (0.625 in).  In the preloaded condition, the Isolator design does not affect the baseline strut stiffness. Design Detail The overall Isolator architecture is a large machined spring housed between two hubs. Threaded into and protruding from the exterior of the Front Hub is a threaded rod end with a spherical ball joint at its end. The interior of the Front Hub has a large counter-bore upon which a Delrin® pad sits, providing a contact surface for the Center Rod, which is preloaded against the pad. The Center Rod passes through the center of the spring and screws into the Back Hub which also provides a threaded interface to the wire bender strut. The concentric alignment of the entire Isolator assembly is controlled by a Flanged Sleeve which is pinned to the Front Hub at assembly and bolted between the front hub and the machined spring. Inside the flanged sleeve are some tight tolerance Delrin® tube bushings which provide a slip fit guide for the Center Rod. Refer to Figure 3 for a cross-section view of the Isolator assembly. TO Isolator Threaded Interface Wire Bender Strut Load Cell (for testing)

Document

AMS_2012.pdf