fine_tune_data/train_data.csv

text
"### Human: What is the title of the symposium hosted by NASA Langley Research Center and Lockheed Martin Space Systems Company, as mentioned in the context?### Assistant:NASNCP-2006-2 14290  3 sth Aerospace Mechanisms Symposium  Compiled by  Edward A. Boesiger  Lockheed Martin Space Systems Company, Sunnyvale, California  Proceedings of a symposium hosted by  the NASA Langley Research Center and  Lockheed Martin Space Systems Company and  organized by the Mechanisms Education Association  held at the Williamsburg Maniott Hotel  Williamsburg, Virginia  May 17- 19,2006  May 2006 {'source': 'AMS_2006.pdf', 'page': 1}"
"### Human: What is the purpose of the NASA Scientific and Technical Information (STI) Program Office, as stated in the provided context?### Assistant:The NASA STI Program Office . . . 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 Office plays a key part in helping NASA  maintain this important role.  The NASA STI Program Office is operated by  Langley Research Center, the lead center for NASA’s  scientific and technical information. The NASA STI  Program Office provides access to the NASA STI  Database, the largest collection of aeronautical and  space science STI in the world. The Program Office is  also NASA’s institutional mechanism for  disseminating the results of its research and  development activities. These results are published by  NASA in tlie NASA STI Report Series, which  includes the following report types:  0 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 compilations of  Significant scientific and technical data and  information deemed to be of continuing  reference value. NASA counterpart of peer-  reviewed formal professional papers, but having  less stringent limitations on manuscript length  and extent of graphic presentations.  0 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.  0 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, projects, 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 that complement the STI  Program Office’s diverse offerings iiiclude creating  custom thesauri, building customized databases,  organizing and publishing research results .. . even  providing videos.  For more information about the NASA STI Program  Office, see the following:  Access the NASA STI Program Home Page at  http://www.sti. nasa.gov  E-mail your question via tlie Internet to  help@sti.nasa.gov  Fax your question to the NASA STI Help Desk  at (301) 621-0134  Phone the NASA STI Help Desk at  (301) 621-0390  Write to: .1  NASA STI Help Desk  NASA Center for Aerospace Information  7 12 1 Standard Drive  Hanover, MD 2 1076- 1320 {'source': 'AMS_2006.pdf', 'page': 2}"
"### Human: What is the name of the symposium hosted by NASA Langley Research Center and Lockheed Martin Space Systems Company, as mentioned in the document?### Assistant:NASA/CP-2006-2 14290  3 8th Aerospace Mechanisms Symposium  Compiled by  Edward A. Boesiger  Lockheed Martin Space Systems Company, Sunnyvale, California  Proceedings of a symposium hosted by  the NASA Langley Research Center and  Lockheed Martin Space Systems Company and  organized by the Mechanisms Education Association  held at the Williamsburg Marriott Hotel  Williamsburg, Virginia  May 17- 19,2006  National Aeronautics and  Space Administration  Langley Research Center  Hampton, Virginia 2368 1-2 199 {'source': 'AMS_2006.pdf', 'page': 3}"
"### Human: What is the contact information for the NASA Center for Aerospace Information (CASI)?### Assistant:The use of trademarks or names of manufacturers in the report is for accurate reporting and does not  constitute an official endorsement, either expressed or implied, of such products or manufacturers by the  National Aeronautics and Space Administration.  Available from:  NASA Center for Aerospace Information (CASI)  7 12 1 Standard Drive  Hanover, MD 2 1076- 1320  (301) 621-0390 National Technical Information Service (NTIS)  5285 Port Royal Road  Springfield, VA 22 16 1-2 17 1  (703) 605-6000 {'source': 'AMS_2006.pdf', 'page': 4}"
"### Human: What is the name of the organization that hosts the Aerospace Mechanisms Symposium (AMS)?### Assistant: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,  the National Aeronautics and Space Administration and Lockheed Martin Space  Systems Company (LMSSC) share the responsibility for hosting the AMs. Now in its  38fh symposium, the AMS continues to be well attended, attracting participants from  both the US. and abroad.  The 38fh AMs, hosted by the Langley Research Center (LaRC) in Williamsburg,  Virginia, was held May 17, 18 and 19, 2006. During these three days, 34 papers were  presented. Topics included gimbals, tribology, actuators, aircraft mechanisms,  deployment mechanisms, release mechanisms, and test equipment. Hardware displays  during the supplier exhibit gave attendees an opportunity to meet with developers of  current and future mechanism components.  The high quality of this symposium is a result of the work of many people, and their  efforts are gratefully acknowledged. This extends to the voluntary members of the  symposium organizing committee representing the eight NASA field centers, LMSSC,  and the European Space Agency. Appreciation is also extended to the session chairs,  the authors, and particularly the personnel at LaRC responsible for the symposium  arrangements and the publication of these proceedings. A sincere thank you also goes  to the symposium executive committee who is responsible for the year-to-year  management of the AMs, including paper processing and preparation of the program.  The use of trade names of manufacturers in this publication does not constitute an  official endorsement of such products or manufacturers, either expressed or implied, by  the National Aeronautics and Space Administration.  ...  111 {'source': 'AMS_2006.pdf', 'page': 5}"
"### Human: What are some of the challenges faced during the development and qualification of mechanical systems for spaceflight, according to the paper by James Wincentsen?### Assistant:CONTENTS  ... Symposium Schedule ............................................................................................................................... vi11  Symposium Organizing and Advisory Committees ................................................................................... xii  Precision Linear Actuators for the Spherical Primary Optical Telescope Demonstration Mirror ................ 1  Jason Budinoff & David Pfenning  The CRISM MotodEncoder Assembly and Diaphragm Bearing Assembly Design ................................. 11  Jeffrey Lees & Ed Schaefer  Gear Teeth Particles and Bearing Failures .............................................................................................. 25  William Greenwood & Jeffrey Dabling  Failure of Harmonic Gears During Verification of a Two-Axis Gimbal for the Mars  Reconnaissance Orbiter Spacecraft ....................................................................................................... .37  Michael Johnson, Russ Gehling & Ray Head  Stacer Driven Deployment: The Stereo Impact Boom ............................................................................. 51  Robert Ullrich, Jeremy McCauley, Paul Turin, Ken McKee & Bill Donokowski  Heritage Adoption Lessons Learned: Cover Deployment and Latch Mechanism .................................... 65  James Wincentsen  Problems and Product Improvements in a Qualified, Flight Heritage Product ......................................... 75  Chuck Lazansky & Scott Christiansen  SoftRide Vibration and Shock Isolation Systems that Protect Spacecraft from  Launch Dynamic Environments ................................................................................................................ 89  Conor Johnson, Paul Wilke & Scott Pendleton  Summary of the New AlAA Moving Mechanical Assemblies Standard .................................................. 103  Brian Gore  Lessons Learned From the Development, Operation, and Review of Mechanical  Systems on the Space Shuttle, International Space Station, and Payloads .......................................... 11 3  Alison Dinsel, Wayne Jermstad & Brandan Robertson  Reliability and Fault Tolerance in ISS Thermofoil Spaceflight Heaters .................................................. 127  Victor Bolton  Development, Pre-qualification and Application of an Active Bearing Preload System ........................ .133  Simon Lewis & Martin Humphries  Development of a Dual Mode D-Strut@ Vibration Isolator for a Laser Com'munication Terminal ........... 141  Dale Ruebsamen, James Boyd, Joe Vecera & Roger Nagel  Design and Testing of a Low Shock Discrete Point Spacecraft Separation System .............................. 149  Pete Woll & Daryn Oxe  V {'source': 'AMS_2006.pdf', 'page': 7}"
"### Human: What are the factors affecting the nut factor in a constant torque spring, as discussed in Robert Postma's paper ""Mechanical Design of a Multi-Axis Martian Seismometer""?### Assistant:Faying Surface Lubrication Effects on Nut Factors ................................................................................ 157  Deneen Taylor & Raymond Morrison  Torque Loss and Stress Relaxation in Constant Torque Springs .......................................................... 163  Robert Postma  Mechanical Design of a Multi-Axis Martian Seismometer ...................................................................... 169  Franck Pecal, Nicolas Paulin, Doug Mimoun & Gabriel Pont  Commercial off-the-shelf Components in Reaction Wheels .................................................................. 187  Andrew Haslehurst  Design of the ATMS Scan Drive Mechanism ......................................................................................... 197  Curtis Allmon & Dave Putnam  Lessons Learned From the Windsat BAPTA Design and On-Orbit Anomalies ..................................... 209  Steve Koss & Scott Woolaway  JWST NlRSpec Cryogenic Light Shield Mechanism .............................................................................. 223  Kathleen Hale & Rajeev Sharma  Development Tests of a Cryogenic Filter Wheel Assembly for the NlRCam Instrument ....................... 229  Sean McCully, Charles Clark, Michael Schermerhorn, Filip Trojanek, Mark O’Hara,  Jeff Williams & John Thatcher  Cryogenic Nano-Actuator for JWST ....................................................................................................... 239  Robert Warden  Space Shuttle Body Flap Actuator Bearing Testing for NASA Return to Flight ...................................... 253  Tim Jett, Roamer Predmore, Michael Dube & William Jones, Jr  Bearing Development for a Rocket Engine Gimbal ................................................................................ 269  Christian Neugebauer, Manfred Falkner, Ludwig Supper & Gerhard Traxler  Effect of Test Environment on Lifetime of Two Vacuum Lubricants Determined by  Spiral Orbit Tribometry ........................................................................................................................... 283  Stephen Pepper  Influence of Oil Lubrication on Spacecraft Bearing Thermal Conductance ........................................... 291  Yoshimi Takeuchi, Matthew Eby, Benjamin Blake, Steven Demsky & James Dickey  Mars Exploration Rover Potentiometer Problems, Failures and Corrective Actions .............................. 303  Mark Balzer  Mechanism Development, Testing, and Lessons Learned for the  Advanced Resistive Exercise Device ..................................................................................................... 31 7 .  Christopher Lamoreaux & Mark Landeck  Radarsat Range Adjustment Mechanism Design .................................................................................. 331  Xilin Zhang & Sylvain Riendeau  vi {'source': 'AMS_2006.pdf', 'page': 8}"
"### Human: What is the purpose of the Come-Along Tool Development for Telerobotic In-Space Servicing of the Hubble Space Telescope, as described in section 345?### Assistant:Come-Along Tool Development for Telerobotic In-Space Servicing of the  Hubble Space Telescope ....................................................................................................................... 345  Jonathan Penn  Planetary Airplane Extraction System Development and Subscale Testing .......................................... 359  John Teter Jr  ""Digital"" SMA-Based Trailing Edge Tab Actuators for Aerospace Applications ..................................... 373  Robert McKillip Jr  Development of a Forced Oscillation System for Measuring Dynamic Derivatives  of Fluidic Vehicles .................................................................................................................................. 387  Bo Trieu, T. Tyler, 6. Stewart, J. Charnock, D. Fisher, E. H. Heim, J. Brandon & S. Grafton  vii {'source': 'AMS_2006.pdf', 'page': 9}"
"### Human: What are the precision linear actuators used for in Session I of the AMS 2006 symposium?### Assistant:SYMPOSIUM SCHEDULE  WEDNESDAY, 17 MAY 2006  8:OO  8:OO  9:oo  9:30  1 1 :45  1 :oo Wednesday Presenters' Breakfast - Terrace Room  CHECK-IN AND REFRESHMENTS - Auditorium  INTRODUCTORY REMARKS - Auditorium  James Wells, Host Chairman  NASA Langley Research Center, Hampton, VA  Stuart Loewenthal, General Chairman  Lockheed Martin Space Systems, Sunnyvale, CA  CENTER WELCOME  Stephen G. Jurczyk, Deputy Director  NASA Langley Research Center, Hampton, VA  SESSION I - ACTUATORS  Stephen Sandford, Session Chair  NASA Langley Research Center, Hampton, VA  * Precision Linear Actuators for the Spherical Primary Optical Telescope Demonstration Mirror  Jason Budinoff & David Pfenning, NASA Goddard Space Flight Center, Greenbelt, MD  The CRISM Motor/Encoder Assembly and Diaphragm Bearing Assembly Design  Jeffrey Lees & Ed Schaefer, Johns Hopkins University Applied Physics Laboratory, Laurel, MD - Gear Teeth Particles and Bearing Failures  William Greenwood and Jeffrey Dabling, Sandia National Laboratories, Albuquerque, NM - Failure of Harmonic Gears During Verification of a Two-Axis Gimbal for the Mars  Reconnaissance Orbiter Spacecraft  Michael Johnson, Jet Propulsion Laboratory, Pasadena, CA; Russ Gehling & Ray Head,  Lockheed Martin Space Systems, Denver, CO  LUNCH BREAK - Box lunch is provided  SESSION II - MECHANISMS  Ted Hartka, Session Chair  Johns Hopkins University Applied Physics Laboratory, Laurel, MD  Robert Ullrich, Jeremy McCauley, Paul Turin, Ken McKee & Bill Donokowski, Space Sciences  Lab, University of California, Berkeley, CA  e Heritage Adoption Lessons Learned: Cover Deployment and Latch Mechanism  James Wincentsen, Jet Propulsion Laboratory, Pasadena, CA - Problems and Product Improvements in a Qualified, Flight Heritage Product  Chuck Lazansky & Scott Christiansen, Starsys Research Corp., Boulder, CO Stacer Driven Deployment: The Stereo Impact Boom  2:30 BREAK  ...  Vlll {'source': 'AMS_2006.pdf', 'page': 10}"
"### Human: What are the key considerations for designing a low shock discrete point spacecraft separation system, as discussed in the presentation by Dale Ruebsamen et al.?### Assistant:2~45 SESSION 111 - “BIG PICTURE”  John McManamen, Session Chair  NASA Johnson Space Center, Houston, TX  SoftRide Vibration and Shock Isolation Systems that Protect Spacecraft from Launch Dynamic  Environments  Conor Johnson, Paul Wilke & Scott Pendleton, CSA Engineering, Inc., Mountain View, CA  Summary of the New AIAA Moving Mechanical Assemblies Standard  Brian Gore, The Aerospace Corporation, El Segundo, CA  * Lessons Learned From the Development, Operation, and Review of Mechanical Systems on the  Space Shuttle, International Space Station, and Payloads  Alison Dinsel, Wayne Jermstad & Brandan Robertson, NASA Johnson Space Center, Houston,  TX  411 5 SESSION IV - POSTER PREVIEW  Michael Johnson, Session Chair  Jet Propulsion Laboratory, Pasadena, CA  * Reliability and Fault Tolerance in ISS Thermofoil Spaceflight Heaters  Victor Bolton, The Boeing Company, Houston, TX  * Development, Pre-qualification and Application of an Active Bearing Preload System  . Simon Lewis, European Space Tribology Laboratory, Warrington, Cheshire, U.K.; Martin  Humphries, Sula Systems Ltd., Wotton-under-Edge, Gloucestershire, U.K.  * Development of a Dual Mode D-Strut@ Vibration Isolator for a Laser Communication’Terminal  Dale Ruebsamen, James Boyd, Joe Vecera & Roger Nagel, Honeywell Defense and Space,  Glendale, AZ  Design and Testing of a Low Shock Discrete Point Spacecraft Separation System  Pete Woll, NEA Electronics, Chatsworth, CA; Daryn Oxe, Lockheed Martin Space Systems  Company, Sunnyvale, CA  * Faying Surface Lubrication Effects on Nut Factors  Deneen Taylor, NASA Johnson Space Center, Houston, TX; Raymond Morrison, The Boeing  Company, Huntington Beach, CA  * Torque Loss and Stress Relaxation in Constant Torque Springs  Robert Postma, The Aerospace Corporation, El Segundo, CA  6:30-9130 RECEPTION & DISPLAYS - ADAMS BALLROOM/PROMENADE OF THE MARRIOTT  Invited component and software suppliers display current products and provide tutorials.  THURSDAY, 18 MAY 2006  7:15 Thursday Presenters‘ Breakfast - Terrace Room  7:45 Light Refreshments - Auditorium  8:15 SESSION V - GIMBALS - Auditorium  William Jones Jr., Session Chair  Sest, Inc., Middleburg Heights, OH  * Mechanical Design of a Multi-Axis Martian Seismometer  Franck Pecal & Nicolas Paulin, EADS SODERN, Limeil Brevannes, France; Doug Mimoun, IPGP,  Saint Maur, France; Gabriel Pont, CNES, Toulouse, France  * Commercial off-the-shelf Components in Reaction Wheels  Andrew Haslehurst, Surrey Satellite Technology Ltd, Guiford, Surrey, U.K.  Design of the ATMS Scan Drive Mechanism  Curtis AIlmon & Dave Putnam, Lockheed Martin Space Systems Company, Sunnyvale, CA  ix {'source': 'AMS_2006.pdf', 'page': 11}"
"### Human: What are the potential causes of the Mars Exploration Rover's potentiometer problems, failures, and corrective actions, as discussed in the presentation by Mark Baker?### Assistant:10:15  10:30  12:oo  1 :oo  3:30  3:45 0 Lessons Learned From the Windsat BAPTA Design and On-Orbit Anomalies  Steve Koss, Naval Research Laboratory, Washington, D.C.; Scott Woolaway, Ball Aerospace &  Technologies Corp., Boulder, CO  BREAK  SESSION VI - JWST  Casey DeKramer, Session Chair  Swales Aerospace, Beltsville, MD  JWST NlRSpec Cryogenic Light Shield Mechanism  Kathleen Hale & Rajeev Sharma, NASA Goddard Space Flight Center, Greenbelt, MD  0 Development Tests of a Cryogenic Filter Wheel Assembly for the NlRCam Instrument  Sean McCully, Charles Clark, Michael Schermerhorn, Filip Trojanek, Mark O'Hara,  Jeff Williams & John Thatcher, Lockheed Martin Space Systems Company, Palo Alto, CA  0 Cryogenic Nano-Actuator for JWST  Robert Warden, Ball Aerospace & Technologies Corp., Boulder, CO  LUNCH BREAK - Box lunch is provided  SESSION VI1 - BEARINGS & POTS  Din0 Christopoulos, Session Chair  Raytheon Space & Airborne Systems, El Segundo, CA  Space Shuttle Body Flap Actuator Bearing Testing for NASA Return to Flight  Tim Jett, NASA Marshall Space Flight Center, Huntsville, AL; Roamer Predmore, Swales  Aerospace, Beltsville, MD; Michael Dube, NASA' Godciard Space Flight Center, Greenbelt, MD;  William Jones, Jr, Sest, Inc., Middleburg Heights, OH  Bearing Development for a Rocket Engine Gimbal  Christian Neugebauer, Manfred Falkner, Ludwig Supper & Gerhard Traxler, Austrian Aerospace  GmbH, Vienna, Austria  Effect of Test Environment on Lifetime of Two Vacuum Lubricants Determined by Spiral Orbit  Tribometry  Stephen Pepper, NASA Glenn Research Center, Cleveland, OH  Influence of Oil Lubrication on Spacecraft Bearing Thermal Conductance  Yoshimi Takeuchi, Matthew Eby, Benjamin Blake, Steven Demsky & James Dickey, The  Aerospace Corporation, El Segundo, CA  * Mars Exploration Rover Potentiometer Problems, Failures and Corrective Actions  Mark Baker, Jet Propulsion Laboratory, Pasadena, CA  BREAK  SPECIAL PRESENTATION - NASA Engineering & Safety Center, Mechanical Systems Super  Problem Resolution Team  John McManamen  NASA Johnson Space Center, Houston, TX  The NESC is an organization established in the wake of Space Shuttle Columbia accident to  provide an independent look at high risk, complex technically issues. Presentation will include  problems the SPRT has been engaged in and focuses in more detail on an ongoing investigation  regarding the Space Shuttle Solid Rocket Booster Holddown Post stud hang up problem and the  completed assessment of the Orbiter Rudder Speedbrake gear micro-pitting problem.  530-1 0130 SYMPOSIUM BANQUET AT JAMESTOWN SETTLEMENT  5:30  1O:OO Bus leaves Marriott to Jamestown  Bus leaves Jamestown for hotel  X {'source': 'AMS_2006.pdf', 'page': 12}"
"### Human: What is the name of the session chair for Session Vlll, which is scheduled to take place at 8:15 AM on Friday, May 19th, 2006, in the Auditorium?### Assistant:FRIDAY, 19 MAY 2006  7:15  7:45 Light Refreshments - Auditorium Friday Presenters’ Breakfast - Terrace Room  8:15 SESSION Vlll - TOOLS & AIRCRAFT - Auditorium  Gerard Migliorero, Session Chair  ESNESTeC, Noordwijk, The Netherlands  Mechanism Development, Testing, and Lessons Learned for the Advanced Resistive Exercise  Device  Christopher Lamoreaux & Mark Landeck, NASA Johnson Space Center, Houston, TX  Radarsat Range Adjustment Mechanism Design  Xilin Zhang &. Sylvain Riendeau, MDA Space, Inc., Ste-Anne-De-Bellevue, Canada  * Come-Along Tool Development for Telerobotic In-Space Servicing of the Hubble Space  Telescope  Jonathan Penn, Swales Aerospace, Beltsville, MD  Planetary Airplane Extraction System Development and Subscale Testing  John Teter Jr., NASA Langley Research Center, Hampton, VA  “Digital” SMA-Based Trailing Edge Tab Actuators for Aerospace Applications  Robert McKillip Jr., Continuum Dynamics, Inc., Ewing, NJ  * Development of a Forced Oscillation System for Measuring Dynamic Derivatives of Fluidic  Vehicles  Bo Trieu, T. Tyler , B. Stewart, J. Charnock, D. Fisher, E. H. Heim & J. Brandon, NASA Langley  Research Center, Hampton, VA; S. Grafton, Vigyan, Inc., Hampton, VA  11 :15 PRESENTATION: An Overview of LaRC  11 :45 TECHNICAL SESSIONS CONCLUSION  PRESENTATION OF THE HERZL AWARD  CLOSING REMARKS  Edward Boesiger, Operations Chairman  Lockheed Martin Space Systems Company, Sunnyvale, CA  BUSES DEPART HOTEL FOR LaRC TOUR 1130  2:00-4:00 NASA LANGLEY RESEARCH CENTER FACILITY TOUR  xi {'source': 'AMS_2006.pdf', 'page': 13}"
"### Human: What is the name of the symposium organizing committee chairperson who is from NASA LaRC?### Assistant:SYMPOSIUM ORGANIZING COMMITTEE  James E. Wells, Host Chair, NASA LaRC  Robin Tutterow, Host Co-Chair, NASA LaRC  Stuart H. Loewenthal, General Chairman, Lockheed Martin  Edward A. Boesiger, Operations Chairman, Lockheed Martin  Carlton L. Foster, NASA MSFC  Claef F. Hakun, NASA GSFC  Christopher P. Hansen, NASA JSC  Wayne Jermstad, NASA JSC  Patrice Kerhousse, ESNESTeC  Alan C. Littlefield, NASA KSC  Edward C. Litty, JPL  Fred G. Martwick, NASA ARC  Donald H. McQueen, Jr., NASA MSFC  Wilfred0 Morales, NASA GRC  Robert P. Mueller, NASA KSC  Fred B. Oswald, NASA GRC  Minh Phan, NASA GSFC  Donald R. Sevilla, JPL  Mark F. Turner, NASA ARC  SYMPOSIUM ADVISORY COMMITTEE  Obie H. Bradley, Jr., NASA LaRC (ret)  Robert L. Fusaro, NASA GRC (ret)  Ronald E. Mancini, NASA ARC (ret)  Stewart C. Meyers, NASA GSFC (ret)  William C. Schneider, NASA JSC (ret)  xii {'source': 'AMS_2006.pdf', 'page': 14}"
"### Human: What is the purpose of the SPOT research testbed, according to the document?### Assistant:Precision Linear Actuators for the Spherical Primary Optical Telescope  Demonstration Mirror  Jason Budinoff* and David Pfenning*  Abstract  The Spherical Primary Optical Telescope (SPOT) is an ongoing research effort at Goddard Space Flight  Center developing wavefront sensing and control architectures for future space telescopes. The 03.5-m  SPOT telescope primary mirror is comprise9 of six 0.86-m hexagonal mirror segments arranged in a  single ring, with the central segment missing . The mirror segments are designed for laboratory use and  are not lightweighted to reduce cost. Each primary mirror segment is actuated and has tip, tilt, and piston  rigid-body motions. Additionally, the radius of curvature of each mirror segment may be varied  mechanically. To provide these degrees of freedom, the SPOT mirror segment assembly requires linear  actuators capable of <lO-nm resolution over a total stroke of 5 mm. These actuators must withstand high  static loads as they must support the mirror segment, which has a mass of -100 kg. A stepper motor  driving a differential satellite roller screw was designed to meet these demanding requirements. Initial  testing showed that the actuator is capable of sub-micron repeatability over the entire 6-mm range, and  was limited by 100-200 nm measurement noise levels present in the facility. Further testing must be  accomplished in an isolated facility with a measurement noise floor of <5 nm. Such a facility should be  ready for use at GSFC in the early summer of 2006, and will be used to better characterize this actuator.  Introduction  Future large (>6 m) space telescopes such as the James Webb Space Telescope, SAFIR, and beyond  require segmented primary mirrors to package into launch vehicle payload fairings of diameters less than  their apertures. Requisite architectures to align or “phase” the individual segments into a single optical  surface after launch and deployment are required. Current techniqyes used on large ground-based  telescop:s such as Keck include precision segment edge sensors and various types of wavefront  sensors . However, phasing a large number of segments requires a significant amount of computing  resources which can reduce observing efficiency. Maintaining a “phased” array of mirror segments in a  challenging environment such as low-earth or L2 orbits remains to be seen. The SPOT research testbed  will explore a new method of phasing segmented mirrors. The SPOT telescope architecture has possible  application to a robotically assembled telescope for ISS, as shown in Figure 1.  2t-  F  Figure 1. A possible application of the SPOT telescope architecture: a telescope mounted on the  “top” (zenith) end of the 21 truss  Wavefront Sensina and Control  A relatively recent technique utilizing image-based wavefront sensing has been pursued by the GSFC  optics branch, Code 551. By placing a point source and camera at the center of curvature of a spherical  * NASA Goddard Space Flight Center, Greenbelt, MD  Proceedings of the 3gh Aerospace Mechanisms Symposium, langley Research Center, May 7 7-79,2006  1 {'source': 'AMS_2006.pdf', 'page': 15}"
"### Human: What is the goal of the SPOT internal research & development program initiated in 2003?### Assistant:mirror, a direct measurement of its surface wavefront error is possible. Taking various defocused images  at the center of curvature and using an iterative transform solver, the phase error of the reflected  wavefront can be determined3. From the phase error tip, tilt, defocus, and other Zernike terms (currently  truncated to the first 15 terms) can be recovered. This information is used to position the mirror segments.  To further develop this approach to phasing mirror segments, the SPOT internal research & development  project was started in 2004.  u,  B - P  or, - 0""  E$ c)  -  0  (1) Fuewue Caineia hlteiface 8 /  Souice at ROC  (3) WFS Alpritluils (DSP) _---_---___  DSP Processor  Figure 2. SPOT Testbed Schematic  Nanometer-level positioning of -1 00-kg mirror segments was required, as well as a high-load nanometer  displacement actuator to mechanically bend the mirror segments to adjust their radius of curvature. As the  program had limited funding, low-cost actuators were designed to meet these requirements.  SPOT Background  The Spherical Primary Optical Telescope (SPOT) is a GSFC internal research & development program  initiated in 2003. The goal of the SPOT effort is to develop a robust architecture which will reduce the cost  of large-aperture, segmented primary mirror space telescopes. The SPOT telescope architecture is based  upon two key technology developments: 1 ) a high-rate, center of curvature, iterative transform phase-  diversity phasing algorithm, and 2) a low-cost mirror segment. The SPOT demonstration telescope is a  03.5-m segmented spherical primary. The primary consists of 6 identical hexagonal segments measuring  876 mm point-to-point, in a 1-ring configuration, without a central segment. However, only 2 segments are  being fabricated for this effort. Two segments are the minimum amount required to successfully  demonstrate the phasing architecture. Each segment has rigid-body position control in tip, tilt, and piston.  Each segment also has mechanical radius-of-curvature control. Some of the relevant requirements for the  SPOT mirror segments are given in Table 1.  2 {'source': 'AMS_2006.pdf', 'page': 16}"
"### Human: What is the maximum weight of each mirror segment in the SPOT mission, assuming a mass of -50 kg each?### Assistant:Table 1. Pyrexm Mirror Segment Requirements  Tipmilt Range  Tipmilt Resolution  Tipmilt update rate  Position Hold  Actuator Thermal Stability  Static Load Mirror Requirement I Value I Units I Note  Size I 0876 I mm (inch) 1 Point-to-point hex  f 2.0 degree  0.05 arcsecond As allowed by focus resolution  1 HZ  0 Amp Power-off hold  35 kg Mirror mass -1 00 kg, assume 3 Micronddeg As small as practicable I Focus update rate I I  I I I I actuators I  To provide rigid-body positioning in tip, tilt, and defocus of the segments, which will weigh -50 kg each, a  tripod mechanism with custom actuators was designed at GSFC.  Design of the Segment Assembly Tripod  The mirror segment must have 3 rigid-body degrees of freedom: tip and tilt rotations and piston, a vertical  translation. A segment assembly is shown in Figure 3.  Figure 3. The SPOT Mirror Segment Assembly Tripod {'source': 'AMS_2006.pdf', 'page': 17}"
"### Human: What is the number of degrees of freedom in the SPOT tripod system, according to Grubler's mobility criterion?### Assistant:Kinematics of the Seqment Assemblv & Grubler’s Mobilitv Criterion  The mirror requires only tip tilt and piston adjustment, 3 degrees of freedom. A hexapod would provide 6  degrees of freedom; but we don’t need 6. Therefore, a tripod was selected to provide the rigid body  motions required. The end joints of each strut must constrain a number of degrees of freedom. For  example, a ball-in-socket joint constrains 3 translations but is free to rotate, allowing 3 rotations. From  kinematics, Grubler’s mobility criterion states that F, the number of degrees of freedom in a system, can  be defined by:  i= 1  Where  h = 6, the degrees of freedom in the space the mechanism will be operating in  n = the number of links in the system  j =the number of joints in the system  f, =the degrees of freedom allowed (unconstrained) at the th joint  For the SPOT tripod, each leg consists of 2 links and 3 joints. Ground is considered a rigid link, and the  mirror is the “end effector” or output link. For this system, the following values are used:  h = 6, we shall consider the system exists in 6 degrees of freedom  n = 8, ground and the mirror are each one link, and each leg has 2 links  j = 9, each leg has 3 joints (base, linear, and upper) x 3 legs  f, = 2, base joint of leg 1, XY flexure allowing 2 rotations  f2 = 2, base joint of leg 2, XY flexure allowing 2 rotations  f3 = 2, base joint of leg 3, XY flexure allowing 2 rotations  f4 = 1, linear joint of leg 1, allowing 1 translation  f5 = 1, linear joint of leg 2, allowing 1 translation  f6 = 1, linear joint of leg 3, allowing 1 translation  f7 = 2, upper joint of leg 1, XY flexure allowing 2 rotations  fa = 2, upper joint of leg 2, XY flexure allowing 2 rotations  f9 = 2, upper joint of leg 3, XY flexure allowing 2 rotations  Using the above values, the system degrees of freedom are calculated as:  i=l  The number of system degrees of freedom is 3, corresponding to tip, tilt and piston. The use of XY  flexures, allowing f, (for I = 1 ..3, 7..9) = 2, is justified. Ball joints could be used, but the additional passive  degree of freedom at each leg (roll) would have to be subtracted out of the Grubler criterion equation to  keep F = 3. The XY flexures will allow hysteresis-free angular motion at the cost of increased force  proportional to displacement.  Actuator Design  The actuator is shown in Figure 4.  4 {'source': 'AMS_2006.pdf', 'page': 18}"
"### Human: What is the step size of the Phytron 500 stepper motor in the SPOT linear actuator, as indicated in Figure 5?### Assistant:Figure 4. The SPOT linear actuator, shown with a six-inch ruler for scale.  Figure 5. An exploded view of the SPOT linear actuator  Stetmer Motor/Harmonic Drive Gearhead  The HD14 1OO:l harmonic Drive gearhead  Phytron 500 step per revolution or 0.72 degree step size  3-phase stepper motor, 2.5 Amp winding  Agilent HEDL 5540 500 line (A quad B = 2000 counts) incremental encoder on motor output ZSS 52.500.2.5.Kl -HEDL-HDl4/100  5 {'source': 'AMS_2006.pdf', 'page': 19}"
"### Human: What is the effective lead of a differential satellite roller screw, as calculated using the formula provided in the context information?### Assistant:Figure 6. The Phytron ZSS 52 / HD14 actuator  Differential Satellite Roller Screw  Several options exist for rotary-to-linear motion: lead screw, ball screw, or roller screw. Generally the most  precise of these is the satellite roller screw. A differential roller screw was selected. After several months  of vendor interaction, a differential roller screw was selected and sized. The smallest, readily available  precision roller screw has a pitch of 0.5 mm. A lead of this size can produce 10 nm steps using the 0.72""  stepper motor and 1OO:l harmonic drive.  Using a differential roller screw, the effective lead can be reduced by 2 orders of magnitude, but at the  cost of a stroke limitation to -6 mm. The theoretical attainable step size drops to 0.4 nm (see below). A  RollvisTM differential satellite roller screw utilizes equal thread pitch on the nut and the shaft, but varies the  nut/shaft thread pitch diameters and the number of starts on the nut and shaft. The effective lead of such  a differential satellite roller screw can be calculated by:  Where  D, = Nut thread pitch diameter, mm  D, = Shaft thread pitch diameter, mm  P =thread pitch in threaddmm  N, = number of thread starts on nut  N, = number of thread starts on shaft, negative for opposite handedness to nut starts  For the SPOT actuator, the values were varied within reasonable limits until a minimum value for effective  lead was found. Using the following values:  D, = 29 mm  Ds=19mm  P = 0.5 threadslmrn  Nn = 4  Ns = -6 {'source': 'AMS_2006.pdf', 'page': 20}"
"### Human: What is the minimum effective lead of the roller screw, as determined by the experiment?### Assistant:The minimum effective lead, Lff, was found to be 0.02 mm per revolution, or -21 microns per revolution. A  bind condition determines the total stroke, which for this differential roller screw is -6 mm. Rollvis Swiss  S.A., a Swiss manufacturer of precision roller screws, fabricated the roller screw as Model  RV160/19,02.R1.604350, custom designed for maximum resolution.  The 5 rollers roll around the shaft and are held in a rotating retainer ring at each end of the nut. A sun gear  at each end of the nut and mating roller gears at the ends of each roller keep them in proper clocking as  they rotate around the shaft. The shaft, nut and rollers are 410 stainless steel. The lubricant for the roller  screw is lsoflex Topas NCA 52, manufactured by Kluber Lubrication. It is a synthetic oil with a calcium  thickener. The differential satellite roller screw is shown in Figure 7 and 8.  Figure 7. Custom Differential Satellite Roller Screw  Figure 8. Nut end details showing the timing gear teeth  Helical Couding  A standard flexible shaft coupling from Helical Products Company, Inc. was used to couple the motor  output shaft to the satellite roller screw shaft. Such couplers allow torque to transmitted despite small axial  misalignments between the shafts. A model HRM-125-12mm-12mm coupling was used. This coupler  uses 2 pairs of cup-point set screws to secure the motor and screw shafts. The coupling is 17-4PH H900  stainless steel. The coupling is shown in Figure 9.  7 {'source': 'AMS_2006.pdf', 'page': 21}"
"### Human: What is the static load capacity of the Helical Coupling Bearing 906HCDUL, as per the given context?### Assistant:Figure 9. Helical Coupling  Bearinas  Barden 1 O6HCDUL back to back duDlex oair. ABEC-9. SAE52100 steel  30-mm bore diameter, 15-degree contaci angle, 14 7.1 4-mm (9/32"") diameter 440C balls, 36-kg (80-lb)  heavy preload  Static load capacity 1005 kg (221 6 Ib)  Machined p ca  Winsorlube oil  Figure 10. Barden 106HC Duplex bearing pair  Lea End Flexures  XY flexures  Crossed flexure  Torsional rate  Fat iq u e/c ycl e I if e  8 {'source': 'AMS_2006.pdf', 'page': 22}"
"### Human: What is the value of the modulus of elasticity (E) for the SPOT tripod, as per the given context?### Assistant:Where  K = torsional spring rate  r = radius of notch cut  b =thickness of flexure section  t = width of notch E = Modulus  For the SPOT tripod, each leg consists of 2 links and 3 joints. Ground is considered a rigid link, and the  mirror is the “end effector” or output link.  Conclusion  Further testing to fully characterize nanometric step size, repeatibilty, and linearity must be accomplished  in a quiet facility. The actuator performance will also be measured with an actuator built into a mirror  segment. Actuator positioning performance will be indirectly measured by mirror radius of curvature  change per commanded step.  References  1.  2.  3.  4.  5.  6.  7.  8. 9. Budinoff, Jason G. “SPOT Mirror Segment Assembly Requirements  - Revision C, March 2004  Howard, Joseph ”Optical Design Study for NASA’s Spherical Primary Optical Telescope” SPIE  Budinoff, Jason G, Michels, Gregory J. “Design & Optimization of the Spherical Primary Optical  Telescope (SPOT) Primary Mirror Segment” SPIE 5877-42  Dean, B, Smith, S, Budinoff, J. “Image-Based Wavefront Sensing for the Control of Space Optics”  USAF AMOS Maoi Technical Conference, September, 2000  Chanan, G., Troy, M., Ohara, C. “Phasing the Primary Mirror Segments of the Keck Telescopes:  A Comparison of Different Techniques” Proc. SPIE, 4003, 188-201, 2000  Martinez, L.M., Yaitskova, N., Dierickx, P., Dohlen, K. “Mach Zender Wavefront Sensor for  Phasing Segmented Telescopes”  Tsai, Lung-Wen, ‘Robot Analysis: The Mechanics of Serial and Parallel Manipulators” 01 999 John  Wiley & Sons, Inc.  Paros, J.M., Weisbord, I. “How to Design Flexure Hinges” Machine Design vol 37, ppl51-156,  1965  Lobontiu, Nicolae “Compliant Mechanisms: Design of Flexure Hinges” 02002 CRC Press LLC 5524-1  9 {'source': 'AMS_2006.pdf', 'page': 23}"
"### Human: What is the purpose of the gimbal bearings in the CRISM motor/encoder bearing assembly?### Assistant:The CRlSM Motor/Encoder Assembly and Diaphragm Bearing Assembly Design  Jeffrey Lees"" and Ed Schaefe;  Abstract  This paper will describe the thin section angular contact bearings and WS2 dry film lubrication used on the  compact Reconnaissance Imaging Spectrometer for Mars (CRISM) motor/encoder and diaphragm  bearing assemblies.  Introduction  CRISM will use targeted observations to search for evidence of aqueous activity and to characterize the  geology and composition of surface features' on Mars (Figure 1). Global measurements acquired  repeatedly throughout the Martian year will provide information on atmospheric water vapor, CO, and  aerosols complementary to that from other MRO instruments.  The Optical Sensor Unit (OSU) consists of an optical system, a cryogenic system, and focal plane  electronics gimbaled about a single axis to allow scanning over k60"" from nadir. Its mechanical design  builds on proven technology from previously successful APL instrument designs. The base housings are  fabricated from titanium that provides high stiff ness and thermal isolation within the same component.  The gimbal bearings are a precision assembly designed to operate in a -60°C environment. The gimbal is  driven directly by a brushless DC motor paired with a BE1 20-bit incremental position encoder. The  encoder disk is co-mounted directly to the bearing shaft beside the motor rotor and the read heads are  mounted to the bearing housing alongside the motor stator. The electrical signals and purge are passed  through a twist capsule in the center of the motor/encoder bearing assembly. A second bearing pair is  mounted in a parallel diaphragm bearing housing that provides high stiffness in the lateral directions to  the gimbal axis and flexibility along the gimbal axis to compensate for differential expansion of the  instrument and spacecraft. The spectrometer housing is passively cooled to -90°C using a flexible link to  the anti-sunward radiator. The anti-sunward radiator passes through the center of the diaphragm bearings  in a thermally isolated mount, and thus rotates with the OSU; its FOV is independent of gimbal position.  4  I  Figure 1. CRISM instrument  Johns Hopkins University Applied Physics Laboratory, Laurel, MD  Proceedings of the 38 Aerospace Mechanisms Symposium, Langley Research Center, May 17- 19,2006  11 {'source': 'AMS_2006.pdf', 'page': 25}"
"### Human: What is the minimum operating temperature for the CRISM IR focal plane?### Assistant:CRISM Thermal Design  The CRlSM thermal design provides both active cooling of the IR focal plane to cryogenic temperatures  to reduce dark current and passive cooling of the spectrometer housing to -80°C for low background.  Simultaneously it keeps the electronics section near -40°C (Fig. 2). Cryogenic cooling is provided by three  Ricor K508 integral Stirling cryocoolers. The multi-cooler configuration requires “thermal switching”  between coolers. A cryogenic diode heat pipe assembly consisting of heat pipes and a thermally isolating  mounting assembly connects the active cooler with the focal plane while isolating if from the two dormant  coolers2. Each of the three diode heat pipes is connected to the focal plane on one end and to a cooler on  the other. The focal plane electronics are mounted in the bottom of the gimbal housing and are  maintained at -40°C. The housing along with tantalum plates provide EM1 shielding and minimize cable  length to the focal planes. Table 1 lists the CRISM expected flight operating temperatures.  Table 1. CRISM Flight Operating Temperatures  -40”c 7, I Component ~ I I Min. I Max. I  mousina I “C I -1  IR FPA 120  Crvo-Coolers I “C I -25 I 20  c T  / 113K -20°C 4- -7 7 -1 20°C  I  Figure 2. CRISM Thermal Zones  12 {'source': 'AMS_2006.pdf', 'page': 26}"
"### Human: What type of bearings are used in the CRISM instrument to rotate the OSU from nadir?### Assistant:CRISM Bearing System  The CRlSM instrument rotates the OSU do"" from nadir. The OSU weighs 20.9 kg (45.9 Ibf) and is  supported by the motor/encoder (Figure 3) and the diaphragm (Figure 4) bearing assemblies. The motor-  encoder side used a thin-section angular-contact duplex pair mounted back to back and was designed to  take the non-axis moments, the entire thrust load, and its share of the radial loads associated with launch  event (Figure 5). These bearing were designed for a 266.9-N (60-lbf) axial preload. The diaphragm side  used a thin-section angular-contact duplex set mounted face to face and was designed to take only the  radial loads associated with the launch event. The diaphragm bearings were designed for a 66.7-N (15-  Ibf) axial preload.  7 Spacer/Seal  Bearing  Spacers  Switch Titanium  Spindle  2x Limit  Titanium  Housing Motor Stator  Motor Rotor - Encoder Disk  Purge  Fitting  - Twist  Capsule i  Figure 3. CRISM MotorEncoder Assembly  Preloading  Angular contact bearings should be preloaded as lightly as necessary to achieve the desired results. A  duplex pair is a pair of bearings that have a pre-determined amount of preload built into them. This was  accomplished by grinding the inner or outer ring a sufficient amount to eliminate all internal clearance  within the bearing commonly referred to as the preload offset. There are, however, several disadvantages  to preloading bearings:  Increased running torque  0  0 Sensitivity to misalignment Sensitivity to differential thermal expansion  13 {'source': 'AMS_2006.pdf', 'page': 27}"
"### Human: What is the purpose of preloading in CRlSM bearings, as described in the given context?### Assistant:Parallel Corrugated  Diaphragms  m  I  s 7 1  I k- Bearings  Flex link  L Interface  r /- Titanium Housing  Figure 4. CRISM Diaphragm Bearing Assembly Anti-Sunward  Radiator  The CRlSM duplex bearing pairs were separated by titanium spacers so that the preload offset would  remain constant over temperature. However, the difference between the 440C inner and outer rings and  the titanium shaft and housing resulted in a reduction of clearance as temperature decreased (Figure 6).  A reduction of clearance results in a decrease of the contact angle. However, the bearings are only going  to experience substantial axial loads during the launch. The bearings were tested to -196°C and  continued to rotate freely. The few disadvantages of preloading are more than offset by the following  advantages:  Reduces axial and radial runout of the rotating shaft. Required for the encoder disk to read head  alignment  Reduces the shaft deflection under load and improves its assembled stiffness  Removes free play in the bearing set, keeping the bearing set loaded in-order to avoid skidding of  the balls  Minimizes the peak stresses that occur during the maximum loading events by ensuring the load  on the bearings is shared by more balls in each bearing 0  Decreases bearing noise  In addition to the axial preload, the CRlSM bearings employed a light interference fit, 12.7 pm (0.0005 in)  in the bearingkhaft fit and 15.2 pm (0.0006 in) in the bearing/housing fit.  14 {'source': 'AMS_2006.pdf', 'page': 28}"
"### Human: What is the clearance at the LMC (Leading Motor Coupling) of the CRlSM (Cryogenic Refrigeration Linear Motor) motodencoder spindle, as shown in Figure 5 of the provided document?### Assistant:Encoder Disk-  Interface  Bearing -  Bearing 7  Spacer  Bearin -  0.80  0.70  h C  ~ 0.60  0 9 .-  v  $ 0.50  C  (d 5 0.40 !!!  -  (d  2 0.30 E  C -  0.20  0.10 Bearing -  Retainer  Figure 5. CRlSM motodencoder spindle  - -  L - Clearance at LMC  I I l~l~l~l~l~l~l~l~l  -200 -180 -160 -140 -120 -100 -80 -60 -40 -20 0 20  Temperature (C) 12.5  12.0  h w  Q) U v  e,  w  11.5 2 -  c 0  (d  C c  6  11.0  10.5  Figure 6. CRlSM motor/encoder bearing clearance  15 {'source': 'AMS_2006.pdf', 'page': 29}"
"### Human: What are the three choices of dry film lubrication considered for the CRlSM motor/encoder and diaphragm bearing assemblies?### Assistant:CRlSM Bearing Lubrication  The thermal analysis of the CRlSM motor/encoder and diaphragm bearing assemblies indicated that they  would operate at -40°C and -90°C respectively. The analysis indicated that the diaphragm bearing  assembly would be too cold for oil or grease lubrication. Our tests showed that -40°C was beyond the  acceptable temperature range for Pennzane synthesized hydrocarbon based oils and greases and -90°C  was beyond the accepted range for perfluorinated polyether based Brayco oils and greases. It was  desired to use identical lubrication in both bearing assemblies for the following reasons:  0  a To simplify bearinghubrication testing  0 To have nearly identical bearings and identical lubrication in both bearing assemblies  Once the bearings and lubrication were selected, an single qualification test could be applied to  both assemblies  The three choices of dry film lubrication that were considered were:  1. Ion plated lead  2. Sputtered MoS2 (molybdenum disulfide)  3. WS2 (tungsten disulfide)  Ion-plated lead bearings were successfully used in the Compact Remote Imaging Spectrometer (CRISP)  tracking mirror assembly on the Comet Nucleus TOUR (CONTOUR) spacecraft. The CRISP tracking  mirror assembly bearings were a Barden precision angular contact duplex pair with a bronze cage. They  were purchased through BE1 as part of their motor/encoder assembly. This was done since they were the  same bearings used in the BE1 motor/encoder and they had used an identical set previously in the  SABER3 instrument on the TIMED spacecraft. However, we were unable to procure ion-plated lead thin-  section bearings at the time. If it were possible to procure them in the future, we would definitely  recommend trying them.  There is a great deal of literature touting the benefits of sputtered MoS,. Dry film lubrication tests for MoS2  and WS2 were conducted for the CRISP cover and release mechansim3. Sputtered MoS2 is supposed to  work great in vacuum, however, moisture absorption can cause severe performance degradation. Most of  the CRISP cover and release mechanism qualification testing was conducted in ambient conditions  including a test on NASA’s Low Gravity Experiment aircraft (the “Vomit Comet”). Tests comparing  sputtered MoS2 and WS,, showed WS2 to be clearly superior to sputtered MoS, for this application.  Sputtered MoS2 seemed to exhibit tremendous stiction in tests conducted in ambient conditions where as  stiction was virtually undetectable with WS2. In fact, we believe that the cover and release mechanism  would not have worked at all using sputtered MoS, in ambient conditions. Sputtered MoS2 was also used  in the bearings for the MDlS instrument on the MESSENGER spacecraft launched in 200X. In addition to  the well known issues of sputtered MoS2, this instrument suffered two additional issues:  1. Smoothness  2. The sputtered MoS2 coating is a hard and brittle coating  The MDlS instrument was able to meet its performance requirements with the sputtered MoS2 coated  bearings. However, the initial flight assembly had to be replaced and a new assembly built up because  the sputtered coating had become cracked and resulted in large unacceptable torque spikes. Based on  our experiences, we would NOT recommend sputtered MoS2 coatings.  WS,, Dicronite, was chosen for the CRlSM motor/encoder and diaphragm thin section bearings.  However, we were unable to locate any specific examples or find any heritage on WS2-lubricated  bearings used in space applications in the available literature. However, based on our previous success  with WS2 on the CRISP instrument, we were optimistic that WS2 could also work in bearings. We also  knew that the bearing companies offered it as an option and that the WS2 coaters coated bearings. We  decided to conduct our own WS2 coated thin-section Teflon-toroid bearing test.  16 {'source': 'AMS_2006.pdf', 'page': 30}"
"### Human: What is the main reason why the MoS2 bearings did not feel smooth according to the text?### Assistant:CRISM Bearing Tests  Identical sets of sputtered MoS2-coated and WS,-coated thin-section Teflon-toroid angular contact  duplex-pair bearings were purchased simultaneously (Figure 7). Prior to running the tests, an initial set of  pictures using a scanning electron microscope were taken of both types of bearings (Figure 8 through  Figure 10). Figures 8 through 10 show that MoS2 appears to have a much rougher surface than WS2. We  believe this is why the MDlS bearings never “felt” smooth.  Figure 7. WS, and Mo S2 life test bearings  Figure Sb. MoS2 Coating  17 {'source': 'AMS_2006.pdf', 'page': 31}"
"### Human: What was the primary concern regarding the bearings after the test, based on the post-test inspection?### Assistant:rigure iua. w& boaring Figure IUD. ~os~~oaring  Following the test of the WS2-coated bearings in which 57952 cycles were completed, we were  concerned that the bearings seemed much rougher than they were prior to the test. After a visual  inspection of the bearings, a significant amount of debris in the bearings was found. Post-test pictures  were taken with a standard microscope, Figure 11 through Figure 13. There were several types of  contamination found:  0 White particles  0  0 Brown particles on the toroids  Brown film on the raceways  We were concerned that the contamination found in the bearings following the test would present a real  problem, both in terms of lifespan and smoothness. However, upon review of the data, the rougher feel  was not adversely affecting the motor control system. The bearings were disassembled and further  inspected. We could find no evidence of wear in the raceways or the WS2 coating. The life test was also  significantly longer than the expected lifetime of the instrument at Mars.  Bearinq Test Conclusion  The WS2 coating is more than adequate for a slow moving oscillating gimbal requiring dry film lubrication.  We were satisfied enough with the WS2 performance that we did not test the sputtered MoS2 bearings,  we had found the solution we were looking for. The Teflon toroids would likely continue to break down  and contaminate the bearings with additional particles that the instrument may no longer be able to rotate  as precisely as is required. We believe the Teflon toroids will ultimately be the life limiting factor of these  18 {'source': 'AMS_2006.pdf', 'page': 32}"
"### Human: What is the potential cause of a ""large"" gap occurring between a ball and toroid in thin-section Teflon-toroid-spaced bearings, according to the document?### Assistant:bearings. Although this may be an advantage for continuously rotating higher speed mechanisms acting  as an additional lubricant  Figure 11. Post life test bearing  Figure 12. Post life test bearing raceway Figure 13. Post life test b6amlllY mabcway  CRISM Bearing Anomalies  Bearina Ball Spacing  The APL Space Department has used different sizes of thin-section Teflon-toroid-spaced bearings on  several programs and tests. One feature that we have always noticed with these types of bearings is the  potential for a “large” gap to occur between a ball and toroid (Figure 14).  We attribute both the high running torque and the poor feel to the non-uniform ball-toroid spacing. We  believe that non-uniform ball-toroid spacing can result in some various amounts of pressure between all  the toroids and balls. This can result in unpredictable friction between the ball and toroid causing an  increase in the running torque. It may also result in a stick-slip situation between the ball and toroid  resulting in torque spikes or non-smooth rotation. Additionally, it could lead to rapid wear of the toroid.  The toroid wear particles could wind up in the raceways as contamination, also causing anomalous  19 {'source': 'AMS_2006.pdf', 'page': 33}"
"### Human: How does the alignment of toroids in angular-contact ball bearings affect their performance, according to the text?### Assistant:torque spikes. Grease and oil lubricated bearings may exhibit the same problems as the CRISM dry film  lubricated bearings to a far less noticeable degree due to the grease or oil between the ball and toroid.  The evidence for this conclusion with the CRlSM bearings is two fold:  1. The significant amount of Teflon contamination found in the bearings following the run-in  procedure described in Appendix 1  2. The drastic change in torque level and smoothness following the high-pressure air blow through  cleaning procedure described in Appendix 1  A  Figure 14. Thin section bearings, same size, different manufacturer  ReDroducible Torque SDikes  Another type of torque spike was also noticed with the CRlSM bearings that could be easily reproduced  based on the operation of the spindle. During the run-in procedure in Appendix 1, unidirectional rotation  resulted in extremely smooth running torque. However, CRISM was intended to oscillate do”. A  reproducible torque spike occurred following a change in direction of rotation. Rotating the spindle  backwards and forwards, sometimes referred to as “safe cracking”, torque spikes, equal to or greater than  the nominal running torque, would result within several degrees of rotation following the reversal. A ball  can never remain rolling between surfaces that form an angle to each other5. All angular-contact ball  bearings create an angle between the two raceways. Therefore, as the bearings rotate, the balls produce  a gyroscopic motion in addition to rolling. Pressure or drag friction between the toroid and the adjacent  balls also seems to deflect the toroids based on the direction of rotation. Thus, the gyroscopic motion of  the balls and pressure between toroids and adjacent balls appears to cause to cause the toroids to align  themselves based on the direction of rotation. This behavior appears to be attributable to the vast majority  of toroids aligning themselves (Figure 15). When the spindle reverses, as do the balls, the toroids flip and  align themselves in the opposite direction (Figure 16). As the toroids flip and align themselves in the  opposite, direction, a torque spike resulted. This effect was significantly reduced following the cleaning  procedure. Once the balls and toroids became more evenly spaced, the drag friction between the toroid  and adjacent balls was reduced, thus reducing the ability of the toroids to align themselves. This was also  noticed following the Christmas holiday. As the bearings sat over the holiday, pressure between the balls  and toroids either slightly re-spaced the balls or caused the Teflon to cold flow resulting in bearings the  felt much better after having sat for an extended period. However, following another run-in, the bearings  quickly resorted to their pre-holiday behavior.  20 {'source': 'AMS_2006.pdf', 'page': 34}"
"### Human: What was the main issue with the ""feel"" of the flight bearing assembly, according to the text?### Assistant:L Figure 15. Toroids in “down” position  Unfortunately, we never photodocumented this effect, thus, it was extremely difficult to find pictures of the  flight hardware demonstrating this effect clearly. Figure 15 clearly shows all the toroids uniformly aligned  in the “down” position. Figure 16 shows most of the toroids in aligned in the “up” position, including most  of the ones at the top of the bearing. This effect can not be reproduced with a single unmounted bearing.  CRISM Bearing Assembly Procedure  The CRlSM bearings were removed from the manufacturer’s packaging and visually inspected to verify  that they conformed to the documentation and were marked properly (Figure 17). Once we were sure that  the bearings were correctly marked and free of contamination, they were assembled as follows:  1. The bearings and spacers were stacked and aligned per drawing and documentation (Figure 18)  2. The bearings and spacers were placed in an assembly fixture (Figure 19) specifically designed to  keep them aligned during assembly  3. The bearing retainer was placed on the assembly (Figure 20)  4. The bearing shaft (Figure 21) was cooled in liquid nitrogen  5. The cooled bearing spindle was quickly assembled in the bearing assembly fixture. Weights were  placed on the assembly to ensure that it remained seated against the top bearing.  6. The bearing assembly was quickly placed in a N2 purged vessel and allowed to equilibrate for 24  hours (Figure 22)  7. The spindle bearing assembly was then visually inspected and checked for “feel”. It was noted  that there was a significant amount of pressure on the spacers between the two bearings on both  the Engineering Test Unit (ETU) and the flight unit. The ETU and the flight assembly were the first  bearing assemblies with spacers that we had assembled using the liquid nitrogen technique.  Previously, the bearings were duplex pairs without spacers. We were concerned that with the  shaft being significantly cooler than the bearings and spacers, that as it warmed and expanded,  there would be very little pressure between the bearings and spacers or worse, a gap. However,  that never materialized  8. The entire spindle bearing assembly was then cooled in liquid nitrogen  9. The cooled spindle bearing assembly was quickly assembled in the titanium  motor/encoder/bearing housing  10. The assembly was then purged and allowed to equilibrate for 24 hours  11. The assembly was visually inspected and checked for “feel”  12. We were not happy with the “feel” of the flight assembly and took the steps outlined in Appendix 1  There were two main issues with the flight bearing assembly, 1) high running torque; 2) torque spikes that  made the assembly not feel smooth as it rotated. We were confident that contamination was not the  problem from pre- and post-assembly inspections causing either of these problems.  21 {'source': 'AMS_2006.pdf', 'page': 35}"
"### Human: What is the name of the instrument used for CRISM (Clouds and the Radiative State) riignr uearings as mentioned in Figure 17, according to the source ""AMS_2006.pdf"" on page 36?### Assistant:A  Figure 17. CRISM riignr uearings  I  I.: .. I .. ,C ,  . ' , -,. c I. ;  riyure IO. Deuririya stuFncu uriu uiiy~icu  s- a  22 {'source': 'AMS_2006.pdf', 'page': 36}"
"### Human: What was the average torque measurement for the bearing assembly during the run-in phase, as shown in Table 2?### Assistant:Appendix 1  Date Operation  12/04/2003 Initial torque measurements Table 2 lists the operations and measurements made to the flight motor/encoder bearing assembly. Most  measurements were an average running torque. Some measurements were a peak torque when there  was a significant torque spike to the running torque. On several tests where many measurements were  made, an average value is reported and denoted by and Avg. in the Torque column.  Torque  mN-m (oz-in)  111.6 (15.8) Table 2. Bearing Assembly Operations  12/16/2003 11 1.6 (1 5.8) nominal  190.7 (27.0) peak  Run-in Q60 RPM, 19 minutes (1 140 Revolutions) 143.0 (20.25) Avg.  Run-in Q60 RPM, 25 minutes (1500 Revolutions) 381.3 (54.0)  Cool off 254.2 (36.0)  12/24/2003-  1 /4/2004 Loosen retaining nut  Run-in Q60 RPM, 20 minutes (1200 Revolutions)  Tighten retaining nut, 17.0 N-m (1 2.5 ft-lb)  Stored in N2 purged container in cleanroom 158.9 (22.5)  190.7 (27.0)  1 /5/2004  Changed torque measuring method for all future  measurements.  Peak torque measurements  ~ Loosen retaining nut and re-measure torques  Set-1 238.7 (33.8) Avg.  Set-2 317.8 (45.0) Avg  1 /6/2004 Run-in Q60 RPM, 20 minutes (1200 Revolutions)  Peak torque measurements  The best these bearings have ever “felt”  Removed retaining nut, rotor, and spacer 148.3 (21 .O) Avg. {'source': 'AMS_2006.pdf', 'page': 37}"
"### Human: What type of instrument is CRISM, according to the reference?### Assistant:Acknowledgements  This work was performed at Johns Hopkins University Applied Physics Laboratory under a contract with  the National Aeronautics and Space Administration. The authors would like to thank all those who helped  design, fabricate, integrate and test the CRISM instrument. 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 Johns Hopkins University Applied Physics  Laboratory, Laurel, MD.  References  1. S. Murchie et. al., “CRISM (Compact Reconnaissance Imaging Spectrometer for Mars) on MRO (Mars  Reconnaissance Orbiter)”, Instruments, Science, and Methods for Geospace and Planetary Remote  Sensing, SPlE Vol. 5660, pp. 66-77.  2. Bugby, D., J. Garzon, 8. Marland, 6. Stouffer, D. Mehoke. M. Fasold, “Cryogenic Diode Heat Pipe  System for Cryocooler Redundancy,” SPlE Optics and Photonics Conference, Cryogenic Optical Systems  and Instruments XI, San Diego, CA, 31 July -4 August, 2005.  3. J. Lees, E. Schaefer, “Design and Testing of the CRISP Tracking Mirror Cover and Release  Mechanism”, Proceeding of the 36‘h Aerospace Mechanisms Symposium, Glenn Research Center, April,  2002, NASNCP-2002-211506, pp.63-76.  4. Esplin, Roy. A satellite-based multichannel infrared radiometer to sound the atmosphere (SABER). In  Optical Remote Sensing of the Atmosphere held in Salt Lake City, Utah, 5-9 February 1995. pp. 130-1 32.  5. Harris, Rollina Bearina Analvsis, 3rd Edition, John Wiley & Sons, Inc., pp.451-452.  24 {'source': 'AMS_2006.pdf', 'page': 38}"
"### Human: What is the purpose of the ratchet wheel in the safety mechanism of missile warheads, according to the abstract?### Assistant:Gear Teeth Particles and Bearing Failures  William H. Greenwood* and Jeffrey G. Dabling*  Abstract  Torque is transmitted from rotary solenoids to rotate drive arms that advance a ratchet wheel as part of a  safety mechanism in missile warheads. The small volume constraint led to single gear teeth to transmit  the torque from the rotary solenoid. High contact forces and compliant gear teeth caused many fine  particles to be generated at the rubbing surfaces of the gear teeth. The particles were pulled into the ball  bearings of the adjacent solenoids causing early failure while having no ill effect on the ball bearings of the  drive arms. A temporary solution of custom plastic shields allowed the prototype units to proceed to  environmental and flight tests. A subsequent build replaced the gear teeth with ball bearing followers.  Introduction  Missile warheads and bombs usually have safety mechanisms [l ] to prevent unintended explosions in the  event of accidents. These safety mechanisms are small and made of stainless steels for high temperature  integrity in the event of accidents involving jet fuel or propellant fires. The safety mechanisms often have  ratchet-type wheels, and are usually driven by rotary solenoids (Figure 1).  Figure -  Mounting plate, ratchet wheel, and rotary solenoid  * Sandia National Laboratories, Albuquerque, New Mexico  Proceedings of the 3dh Aerospace Mechanisms Symposium, Langley Research Center, May 17- 19,2006  25 {'source': 'AMS_2006.pdf', 'page': 39}"
"### Human: What is the purpose of using molybdenum disulfide (MoS2) on the gear teeth in the new generation safety mechanism?### Assistant:A new generation of safety mechanisms is nearing production. The safety mechanism of Figure 1 is  among the new generation and transmits torque from the rotary solenoid to the drive arms by way of  gears. Due to the limited stroke of the solenoid and the volume constraint, the gears are actually single  gear teeth cut as part of the solenoid rotor and drive arm (Figure 2).  Figure 2. Rotary solenoid and single gear teeth  Earlier generations of these safety mechanisms often used full gears to transmit torque from the ratchet  type wheels to other shafts for thousands of ratchet wheel cycles. One life test of the first generation of  safety mechanisms involved 10,000 ratchet wheel cycles; the main gear had teeth that were no longer  involutes but appeared triangular. Much wear debris was adjacent to the gear but did not migrate to ball  bearings or cause operational failure.  The first units of the new intent safety mechanism were operated in January 2003 and failed to operate  after only a dozen to a few hundred ratchet wheel cycles, as shown in Table 1. A failure to operate at such  low number of cycles was unexpected. The original units had molybdenum disulfide (MoS2) applied to the  gear teeth, but had no lubrication on the ball bearings of the solenoid or the drive arms. Molybdenum  disulfide is used since the mechanisms must sit for decades unused and then be called upon to function.  The safety mechanisms are usually hermetic units and have a nitrogen and helium atmosphere, although  the ratchet cycle tests in this report were all performed in the laboratory atmosphere as part of early  qualification. The ball bearings are generally lubricated but a new process for the bearings was not yet  ready, so the early ratchet wheel cycle tests were performed with dry (non-lubricated) ball bearings.  26 {'source': 'AMS_2006.pdf', 'page': 40}"
"### Human: What is the probable cause of the observed higher than expected minimum operate voltage in the ratchet wheel units, according to the context information provided?### Assistant:Table 1. Ratchet wheel rotations to failure, MoS2 on gears, no MoS2 on bearings  Powder at gear teeth  Initial investigation  The energized stroke of the solenoid is on the order of five milliseconds and the spring (de-energized)  return of the drive arms is also on the order of five milliseconds. The initial investigation by high-speed  video of the mechanism in operation showed erratic operation times for the inboard rotor during energized  strokes when the unit was close to failure. Failed units were disassembled and visually examined. Drive  arm bearings and drive springs showed no problems, and no rubs were detected between the arms and  adjacent surfaces. The wear zone on the gear teeth showed a powdered layer rather than a burnished  appearance as shown in Figure 3. The powdered layer was 0.1 to 0.2-mm thick at the edges and 0.05 mm  or thinner in the center. The powdered layer was composed of many micron size particles and was readily  scraped or wiped away from the gears. In addition, many small dark particles were observed on the  mounting plate under the gear teeth and at stop pins. The quantity of particles was much greater than  observed in previous safety mechanisms.  An additional troubling test was a higher than expected minimum operate voltage. The minimum operate  voltage occurs when the solenoid torque is nearly equal to the drive arms spring torque. The solenoids  were designed to operate the mechanism as low as 16 volts and up to the maximum supply voltage of 22  volts. The ratchet wheel cycle tests are operated at 22 volts to give more solenoid torque and greater  forces and loads on the bearings and pins. From the measured solenoid torque and the measured drive  spring forces, the ratchet wheel units were expected to operate at 14 volts but when initially assembled  were measured to operate at about 15 volts. This higher minimum operate voltage indicated a small loss,  in the solenoid bearings, gear teeth or drive arm bearings, and the gear teeth contact between the  solenoid and the drive arms was suspected as the probable cause. The powdered layer that was  observed at the gear teeth contact in the first wheel cycle tests gave more evidence than the gears had  substantial sliding contact and not the rolling contact desired in well-formed gear trains. The solenoid  torque at 22 volts is approximately double the solenoid torque at 15 volts. The failure to operate after  dozens or a hundred ratchet wheel cycles at 22 volts means an apparent frictional loss increased from  near zero at the start to approximately 50% of the solenoid torque. If the large frictional loss was solely at  the gear teeth, then slower and erratic spring return times of the drive arms and solenoid rotors were also  expected but not seen in the high-speed videos.  The powdered layer at the gear contacts was obviously troubling, but it did not appear that the gear teeth  interaction caused such high loss of torque. From the failed units, a solenoid rotor was manually stroked  and no high friction was measured. Several failed units were cleaned of the powdered layer at the gear  teeth contact, and the minimum operate voltage improved somewhat to 18 to 20 volts, but was still  significantly higher than the initial 15-volt measurement. The powdered layer at the gear teeth did not  appear to be the high fiction loss that caused the units to stop at 22 volts.  27 {'source': 'AMS_2006.pdf', 'page': 41}"
"### Human: What is the cause of the rapid bearing degradation observed in the failed ratchet wheel units, according to the text?### Assistant:al  Figure 3. Gear tooth wear zone has a powdered layer.  The solenoids from failed units were rotated manually and some roughness in rotation was noted.  However, a quasi-static torque test of the solenoid at 16 and 22 volts showed nearly the same torque as  measured before the installation into the ratchet wheel assembly, but the measured friction at zero volts  was somewhat higher. The solenoid bearings in failed units were visually examined at 1 Ox magnification,  and the balls in the bearings in the inboard rotor had a roughened or a worn surface (Figure 4). The  solenoid ball bearings were replaced, and the solenoids were re-assembled into their failed ratchet wheel  units. The rebuilt units were then measured for minimum operate voltage and found to be 15 volts, the  same as new units. The parts that degraded as the ratchet wheel units were operated were the solenoid  bearings. The rebuilt units (with new solenoid bearings) were then cycle tested again and failed at similar  cycles as those of Table 1.  The solenoid bearings could be degrading due to high dynamic loads or from particles from the gear teeth  entering the bearings during operations. The axial load from the magnetic forces when the coil is  energized was calculated to be 71 N (1 6 Ibf) compared to the dynamic load rating of 11 6 N (26 Ibf). The  small oscillatory solenoid stroke of 0.2 rad (12 degrees) is a more severe service than continuous  rotations since only a small portion of the balls see the load repeatedly. Therefore, the high axial magnetic  force could be degrading the unlubricated bearings in such a few cycles. However, the innermost bearings  of the rotors support all of the axial forces; whereas the failed units showed bearing wear to be most  severe in the inboard rotor and approximately equal from the innermost and outermost bearings of the  inboard rotor. The rapid bearing degradation did not appear to be caused by the relatively high dynamic  axial load.  28 {'source': 'AMS_2006.pdf', 'page': 42}"
"### Human: What is the primary cause of bearing degradation in the solenoid ball bearings, according to the text?### Assistant:,A ri F- A  PA  Figure 4. Solenoid ball bearings before (left) and after (right) life cycle test.  When ball bearings fail from high loads, the initial signs are many small pits as bearing material spalls,  then the ejecta from the pits are rolled into the bearing surfaces that cause more localized high loads and  accelerated degradation and eventually a non-rotating (seized) bearing. As the bearings were visually  scrutinized, many patches or scabs were observed and some indentations but not pits. The scabs  appeared to be a particle or perhaps several particles that are rolled into the balls as the bearing is  operated. The scabs were foreign particles and not bearing ejecta. The parts that degraded and caused  the early wheel cycle failure were the solenoid bearings; the cause of the bearing degradation was the  particles from the gear teeth. The problem resolution was to sharply reduce the particles generated at the  gear teeth.  Attempts to Reduce the Gear Teeth Particles  The gear teeth were lubricated with a proprietary molybdenum disulfide process. A tribological  investigation showed some alumina particles imbedded in the surface where the solid lubricant was  applied. The lubricant applicator did indeed prepare the surfaces with alumina beads. A possible cause  was identified; some of the alumina beads from the surface preparation were imbedded into the gear  surface and caused grinding at the mating gear surfaces and generated the great amount of metal  particles.  Gear teeth surfaces were cleaned of the MoS2 and alumina beads and were then life cycle tested, but the  same failure to operate after a few hundred ratchet wheel cycles was observed (Table 2). The post-  mortem inspection again showed many metal particles at the gear teeth wear zones and scattered on the  nearby mounting plate and stop pins. The inboard solenoid rotor had much rougher manual rotation than  the outboard rotor. The wear zone on the inboard rotor gear tooth was broader at the base as shown in  Figure 3, rather than a more parallel wear zone observed in the outboard rotor gear tooth.  29 {'source': 'AMS_2006.pdf', 'page': 43}"
"### Human: What is the material composition of the gear teeth in Table 2, according to the text?### Assistant:Table 2. Ratchet wheel rotations to failure, no MoS2 on gears, no MoS2 on bearings  CP008  CP009 12 Feb 2003 180 Rough manual rotation  14 Feb 2003 260 Rough manual rotation  Several units with the wear resistant outboard gear were tested with the same disappointing result: the  units failed to operate after a few ratchet wheel cycles, and there was a thick powdered layer at the wear  zone on the gear teeth (Table 3 and Figure 5). The wear zone on the new outboard gear appeared to be  even heavier than the wear zone on the trials from Tables 1 and 2. The rotors spun roughly after the life  test. The bearings were observed to have a roughened surface on the inboard rotor, but only slightly  roughened on the outboard or stator bearings. The long stringers observed in Figure 5 on the edge of the  gear are composed of many fine particles that are magnetized. These magnetized stringers appear at the  edges of other parts as well as the gear teeth. A schematic of the solenoid is shown in Figure 6. The  solenoid bearings from left to right will be numbered 1 through 4 in many of the descriptions to follow. The  outboard rotor bearings are 1 and 2 while the inboard rotor bearings are 3 and 4.  Figure 5. Wear particles generated due to a Nitronic 60 gear tooth interface  30 {'source': 'AMS_2006.pdf', 'page': 44}"
"### Human: What is the main conclusion drawn from the third replication in Table 4 regarding the effectiveness of MoS2 on gear teeth wear?### Assistant:Table 3. Ratchet wheel rotations to failure, Nitronic 60 gear, no MoS2 on gear or bearings  Date Wheel cycles p  CP011 20 Feb 2003 72 Bearings 3 and 4 bad, bearings 1 and 2 fair  CP037 20 Feb 2003 280 Bearings 3 and 4 bad, bearings 1 and 2 fair  CP034 21 Feb2003 220 Bearing 3 very bad  CP045  CP054  CP021 Outboard rotor Outboard coil  -- 27 Feb 2003 110 I  27 Feb 2003 90 Inboard rotor rotates very roughly  03 Mar 2003 51 8 Inboard rotor bearing seized Stator  24.9 mm  L Inboard coil  Inboard rotor  I_- 19.2mm -4 Inboarc \  Ball bearings i gear  Figure 6. Double rotary solenoid; outboard and inboard rotors move independently.  A new trial was performed with MoS2 to the applied gear teeth, but with the surfaces prepared by acid  etching rather than alumina bead blasting as in the original condition of Table 1. The desire was a drastic  reduction of wear particles at the gear teeth. Again, low ratchet wheel cycles to failure were measured and  a heavy powdered layer at the gear teeth was again observed (Table 4). The third replication that failed at  wheel cycle 518 was encouraging, but the two replications that failed at low cycles indicate that the  condition of Table 4 is similar to the first three conditions. The unit that failed at 51 8 cycles was examined  carefully in the post-mortem, but resulted in no clues regarding its much longer operating life.  Table 4. Ratchet wheel rotations to failure, MoS2 on gears without grit blasting  31 {'source': 'AMS_2006.pdf', 'page': 45}"
"### Human: What was the temporary solution proposed to prevent particle entry into the solenoid bearings due to the lack of success in decreasing particle generation at the gear teeth?### Assistant:A Temporary Solution - Bearing Shields  As our early efforts to decrease particle generation at the gear teeth were unsuccessful (Tables 2-4), we  looked for a quick solution; so that environmental and flight tests could continue. A shielded bearing was  suggested as a way to prevent the gear teeth particles from entering the solenoid bearings. However,  there is not room for standard shielded bearings. The inboard rotor is 3.8-mm thick; the original bearings  are 1.6 mm thick, which leaves a 0.6-mm gap between the ball bearings. A shielded bearing of the same  inside and outside diameters has a thickness of 2.4 mm that for a pair of bearings is wider than the rotor.  The temporary solution was to make thin washers, 50 microns (.002) thick, out of acetate plastic that had  the same inside and outside diameters as the bearings and were placed next to the bearings of the  inboard rotor. Plastic was used since they could be made quickly. One shield was placed on the inboard  rotor bearing next to the outboard gear (bearing 4) as shown in Figure 7, and the other shield was placed  between the inboard rotor bearing (bearing 3) and the stator bearing (bearing 2).  These shields were placed in test units with no MoS2 on the gears or bearings and had encouraging  results as shown in Table 5. Finally, most units run at least 500 wheel cycles. The initial solenoid torque  tests showed no increase in friction due to addition of the plastic shields. After the life tests, the shields  showed a darkened annulus where the bearings were sliding, but the plastic was not grooved. A workable  solution appeared possible. The shields appeared to be protecting bearings 3 and 4 since tested units  now had some bad bearings at position 1 with some good bearings at position 4.  -- I  Figure 7. Plastic shields were used to keep particles out of the solenoid bearings  32 {'source': 'AMS_2006.pdf', 'page': 46}"
"### Human: What is the main cause of failure in the solenoid bearings as shown in Tables 5 and 6?### Assistant:Table 5. Wheel rotations to failure, plastic shields over bearings, no MoS2 on gears or bearings -- --  unit # Date Wheel cycles --  CP046 03 Mar 2003 524 Failed, bearing 1,3,4 - bad, bearing 2 - good  CPO14 03 Mar 2003 520+ No failure, bearing 1 - bad, 2 - fair, 3, 4 - good  CP020 04 Mar 2003 820+ No failure, bearing 1,3 - bad, , bearing 2,4 - good  CPO16 13 Mar 2003 404 Failed, bearing 1, 4 - bad, bearing 2 ,3 - fair  Ball bearings with MoS2 lubricant were received and put into solenoids with the plastic shields next to the  inboard ball bearings and were cycle tested with very favorable results (Table 6). A fix or a solution to our  early wheel cycle failure was found. The solenoid used in the first test of CP028 showed an unrelated  assembly issue, resulting in a frictional rub between the inboard rotor and the stator. After being rebuilt  with an additional 13-micron shim added under the inboard rotor but with the original bearings, the second  cycle test of CP028 performed very well.  Table 6. Wheel rotations to failure, plastic shields over bearings with MoS2, no Mo~ on gears  Unit#m U ate wh eel cycles  CP046 21 Mar2003 101 4+ No failure, bearing 1, 2, 3, 4 - good  CP048 25 Mar 2003 1 ooo+ No failure, bearing 1,2 -fair, bearing 3,4 - good  CP028 25 Mar 2003 52 Failed, bearings - good, inboard rotor rub  CP028 26 Mar 2003 1 ooo+ No failure, bearing 1, 3, 4 - good, bearing 2 - fair  As previously discussed, the dynamic axial load was 71 N versus the manufacturer’s rated load of 116 N.  Since our load condition is a small reversing stroke rather than continuous rotation, the duty is more  severe than the rated condition. The dry film lubrication used in the bearings of the units of Table 6  appears beneficial when compared to the dry bearings of Table 5 and confirms the bearing manufacturer’s  recommendation that lubrication becomes more important as the bearings are more severely loaded.  The drive arm bearings are very close to the gear teeth but did not show any wear when compared to the  solenoid bearings. Many of the gear particles stayed on the gear teeth as shown in Figures 3 and 5, but  some of the particles will migrate to other parts as shown in Figures 7 and 8. The electric coils in the  solenoids produce a magnetic flux field when the coils are energized. The flux field is in the shape of a  toroid that makes a complete loop about the coil from the inside coil diameter to the outside coil diameter.  Most of the flux passes through the rotor and stator due to the high permeability of electrical iron, but  some flux passes through the air between the races of the bearings. Loose magnetic particles will align on  these flux lines and then move along the flux lines to high magnetomotive force. Some of the particles are  pulled into the bearings where they become scabs as the balls roll aginst the races.  The large axial loads on the solenoid rotors are present only when the coils are energized. The wear  particles are pulled into the bearings and rolled flat when the coils are energized; the rotors are returned to  their home positions by helical extension springs. Hence, the slower and erratic motion of the rotor was  observed during the energized stroke only and not during the return stroke. Likewise, a manual rotation  and a zero-volt solenoid torque test would not show the large friction that is present when the coil is  energized. The inboard rotor motion was erratic and caused the operational failure since the inboard rotor  bearings are very close to the gear teeth and the resulting wear particles. When the plastic shields are in  place as shown in Tables 5 and 6, the inboard bearings (3 and 4) are not consistently worse than the  outboard bearings (1 and 2).  33 {'source': 'AMS_2006.pdf', 'page': 47}"
"### Human: What is the Hertzian contact stress at the gear teeth, according to the text?### Assistant:Why so many gear teeth particles?  The tribological investigation continued to evaluate material pairs or lubrication that would decrease the  gear particles. Some standard pin on disk tests at a high contact stress showed long life of the MoS2  lubricant and little particle generation. This caused us to calculate the Hertzian contact stress [2] at the  gear teeth. The contact stress was 753 MPa (109 ksi). This exceeds the yield strength of the Hiperco 50A,  365 MPa (53 ksi), used for the inboard rotor, the wear resistant stainless steel, Nitronic 60, 414 MPa (60  ksi), used for the outboard gear in one of the trials, and the drive arms PH 13-8 Mol condition H1150 , 620  MPa (90 ksi). The contact stress was undoubtedly higher on the single cantilevered gear tooth of the  inboard rotor since it deflects during loading and the wear zone on the tooth was more triangular than  rectangular. Figure 3 shows just such a wear zone on the drive arm gear tooth that mates with the inboard  rotor gear tooth. The single gear tooth was analyzed for bending stresses at the base during the initial  design but the Hertzian contact stresses were not calculated for the contact load at the gear teeth. The  high contact stresses mean plastic deformation at the load zones, non-rolling contact and resultant  particle generation.  Long term solution - eliminate the gear teeth  The plastic shield was a fix that permitted further testing of the first build. However, the many particles  were undesirable and created a potential for other problems. A long-term solution was desired where few  particles were generated. Increasing the gear pitch or the length of the gear teeth to decrease the contact  stress by a factor of two or three was not possible due to the volume constraints. In addition, the single  gear tooth on the inner rotor and the drive arms made these expensive parts and difficult for part  34 {'source': 'AMS_2006.pdf', 'page': 48}"
"### Human: What was the reason for rejecting the design change from a single solenoid to two solenoids side by side?### Assistant:acceptance. Changing the solenoid design from a single shaft to two solenoids side by side was  considered, but rejected as too drastic a change.  The change selected was to use ball bearings as followers between the solenoid and drive arms. The  space was limited but the small 1.0-mm by 3.2-mm ball bearing could be made to fit. See Figure 9 and  compare to Figure 2. The contact load was high but within the manufacturer's limits of 4,000 MPa (580  ksi). This change was implemented on the subsequent group build, and has been successful with long  ratchet wheel cycles before mechanism failure. There are a few dark particles after a ratchet wheel cycle  test, but they are located at stop pins and not at the bearing followers that actuate the drive arms.  '*  I --3  L 4  1  Figure 9. Ball bearings now transfer the torque from the solenoid to the drive arms  Conclusion  Due to volume constraints and the desire for a simple assembly, single gear teeth were made as integral  part of arms and rotors in a small safety mechanism. The gear teeth were discovered in the prototype to  generate many very small particles that were pulled into the bearings of the adjacent solenoid and caused  very early failures in life tests. The cause of the particle generation was contact stresses exceeding the  yield strength of the gear teeth. The lesson learned was to prioritize our efforts in analysis and failure  mode prediction on new or unfamiliar design concepts that are incorporated into a mechanism. An  ancillary lesson is to consider and evaluate additional concepts when moderate or high risks are identified  in the base design.  35 {'source': 'AMS_2006.pdf', 'page': 49}"
"### Human: What is the definition of a Unique Signal Stronglink (USS) according to Plummer and Greenwood (1993)?### Assistant:References  1. D.W. Plummer and W.H. Greenwood (1993). A Primer on Unique Signal Stronglinks, SAND93-0951,  Sandia National Laboratories, Albuquerque, NM.  2. J.E. Shigley and C.R. Mischke (1 989). Mechanical Engineering Design, 5'h edition, McGraw-Hill, New  York City, NY.  36 {'source': 'AMS_2006.pdf', 'page': 50}"
"### Human: What are the limitations of the heritage material choices made for the harmonic gear components that failed during life testing of an actuator assembly on the Mars Reconnaissance Orbiter spacecraft?### Assistant:Failure of Harmonic Gears During Verification of a Two-Axis Gimbal for the  Mars Reconnaissance Orbiter Spacecraft  Michael R. Johnson', Russ Gehling**and Ray Head**  Abstract  The Mars Reconnaissance Orbiter (MRO) spacecraft has three two-axis gimbal assemblies that support  and move the High Gain Antenna and two solar array wings. The gimbal assemblies are required to move  almost continuously throughout the mission's seven-year lifetime, requiring a large number of output  revolutions for each actuator in the gimbal assemblies. The actuator for each of the six axes consists of a  two-phase brushless dc motor with a direct drive to the wave generator of a size-32 cup-type harmonic  gear. During life testing of an actuator assembly, the harmonic gear teeth failed completely, leaving the  size-32 harmonic gear with a maximum output torque capability less than 10% of its design capability.  The investigation that followed the failure revealed limitations of the heritage material choices that were  made for the harmonic gear components that had passed similar life requirements on several previous  programs. Additionally, the methods used to increase the stiffness of a standard harmonic gear  component set, while accepted practice for harmonic gears, is limited in its range. The stiffness of  harmonic gear assemblies can be increased up to a maximum stiffness point that, if exceeded,  compromises the reliability of the gear components for long life applications.  Introduction  The Mars Reconnaissance Orbiter Mission  During its two-year primary science mission, the Mars Reconnaissance Orbiter will conduct eight different  science investigations at Mars. The investigations are functionally divided into three purposes: global  mapping, regional surveying, and high-resolution targeting of specific spots on the surface. This detailed  mapping of the surface of Mars will provide future landed missions with the high resolution data required  to land safely in a desired area. The instruments on board the MRO spacecraft consist of five types:  cameras, a spectrometer, a radiometer, a radar, and engineering. Refer to Figure 1 for an overall view of  the MRO spacecraft's science deck.  Cameras  HiRlSE (High Resolution Imaging Science Experiment)  This visible-camera can reveal small-scale objects in the debris blankets of mysterious gullies and details  of geologic structure of canyons, craters, and layered deposits.  CTX (Context Camera)  This camera will provide wide area views to help provide a context for high-resolution analysis of key  spots on Mars provided by HiRlSE and CRISM.  MARC1 (Mars Color Imager)  This weather camera will monitor clouds and dust storms.  Spectrometer  CRISM (Compact Reconnaissance Imaging Spectrometer for Mars)  This instrument splits visible and near-infrared light of its images into hundreds of ""colors"" that identify  minerals, especially those likely formed in the presence of water, in surface areas on Mars not much  bigger than a football field.  ' Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA  ** Lockheed Martin Space Systems, Denver, CO  Proceedings of the 38 Aerospace Mechanisms Symposium, Langley Research Center, May 17- 19,2006  37 {'source': 'AMS_2006.pdf', 'page': 51}"
"### Human: What is the purpose of the Radiometer MCS instrument on board the Mars Reconnaissance Orbiter?### Assistant:Radiometer  MCS (Mars Climate Sounder)  This atmospheric profiler will detect vertical variations of temperature, dust, and water vapor  concentrations in the Martian atmosphere.  Radar  SHARAD (Shallow Radar)  This sounding radar will probe beneath the Martian surface to see if water ice is present at depths greater  than one meter.  Enaineerinq  The engineering instruments facilitate spacecraft navigation and communications.  High Gain Antenna  *,  II J VADTR Tnstniment Deck  Once the science phase is completed (two years after the mapping orbit is established), the MRO mission  enters a second phase, communications relay. In this phase, the communication equipment on-board  MRO will be used as a communications relay between the Earth and landed crafts on Mars that may not  have sufficient radio power to communicate directly with Earth on their own. This capability allows landed  crafts to use smaller antennas with reduced mass, improving the lander’s science complement potential.  Due to the mapping nature of the mission, the instrument deck of the spacecraft must always be facing  the surface of Mars. Additional pointing requirements include maintaining sun pointing of the solar panels  and keeping the High Gain Antenna Earth pointed for communication purposes. The solution to this  extreme panel and antenna pointing choreography was to put a two-axis gimbal at each of the  appendages: two solar array wings and one High Gain Antenna.  The path from Earth to Mars orbit and mapping of the surface consists of launch, cruise, orbit insertion,  aerobraking, and mapping phases. The spacecraft configuration of the solar array and the High Gain  Antenna are different for each of these phases. During launch, the solar array wings are folded in half and  the High Gain Antenna is positioned directly over the spacecraft bus to fit into the launch vehicle fairing  (Figure 2). Once MRO is launched and in the cruise phase of the mission, the solar array and High Gain  Antenna are pointing in roughly the same direction to capture sunlight and communicate with Earth  (Figure 3). For Mars orbit Insertion and aerobraking, the appendages are moved slightly from the cruise  configuration to produce an aerodynamically stable configuration. The High Gain Antenna and the solar  array wings are the predominant source of atmospheric drag on the spacecraft and must be positioned to  keep the spacecraft stable throughout the aerobraking maneuvers, which last for about 6 months (see  Figure 4). For the mapping phase of the mission, the solar array wings and High Gain Antenna are almost  38 {'source': 'AMS_2006.pdf', 'page': 52}"
"### Human: What is the purpose of the gimbal actuators on the Mars Reconnaissance Orbiter (MRO)?### Assistant:continuously articulated so the wings remain sun pointed and the High Gain Antenna maintains a lock on  Earth. This continuous motion must be performed while the spacecraft maintains precise pointing for high  resolution imaging and high-speed data transmission to Earth (see Figure 1 for a mapping configuration).  Because of the stringent pointing stability requirements, the gimbals were required to be exceedingly  smooth and quiet.  The different phases for the MRO appendages lead to a large range of requirements for the two-axis  gimbal assemblies. The gimbals must be capable of carrying launch loads through their output bearings  for the launch phase. The cruise phase is very benign with minimal load on the gimbals. The orbit  insertion and aerobraking phases put a significant load on the output bearings and gears, since the loads  have a significant component in the backdriving direction for the gimbal actuators. Once the spacecraft is  in the mapping phase, the high resolution capability of the instruments on board require that the gimbal  assemblies do not produce any significant disturbance to the spacecraft platform while they are  continuously scanning to maintain the required pointing of the attached appendage. The gimbals must  withstand all of these load combinations and still maintain extreme pointing accuracy and smooth  operation once at Mars. In addition to the smooth motion, the lowest structural frequency in the mapping  configuration is determined by the natural frequency of the deployed appendages. A major contributor to  the frequency of the deployed appendage is the gimbal actuator output stiffness for each axis.  Figure 2. MRO Launch  Configuration Figure 3. MRO Cruise  Configuration Figure 4. MRO Orbit  Insertion & Aerobraking  Configuration  Gimbal Actuator Confiauration  Each two-axis gimbal consists of two identical gimbal actuators, structurally connected with application  specific components. The core of the actuator is a 130-mm diameter two-phase brushless dc motor with a  large number of poles in order to maintain smooth rotor velocity. The brushless motor is commutated  using a resolver with the same number of poles as the motor to simplify the commutation logic. The motor  directly drives the wave generator of the output harmonic gear component set through a bellows coupling.  The bellows coupling was used to minimize speed ripple that would cause disturbances while operating.  The harmonic gear is a size-32 HDC, standard-cup-type unit. The flexspline is mounted to the actuator  housing and provides the torque reaction mount. The circular spline is mounted in a pair of preloaded  angular-contact ball bearings. A multi-speed output resolver is installed between the angular-contact  bearings for a compact assembly that measures the output position to the accuracy required for the MRO  mission pointing. This arrangement of drive components provides a zero-backlash actuator with minimal  mechanically generated disturbance sources and applies all of the externally generated loads directly to  the harmonic gear teeth. A photo of a completed flight gimbal actuator assembly is shown in Figure 5.  39 {'source': 'AMS_2006.pdf', 'page': 53}"
"### Human: What is the primary reason for selecting a size-32 cup-type harmonic gear component set for the MRO spacecraft?### Assistant:The need for the harmonic gear to take the external loads led to the selection of the size of the harmonic  gear in concert with the required output stiffness. The available volume and mass allocation for the  actuators was minimal on the MRO spacecraft, requiring use of the smallest gears possible to achieve the  required load and stiffness capability. The loads from the Mars orbit insertion and aerobraking phases  needed to be taken into account along with all other applied loads during the mission. The mass of the  solar array wings and the antenna assembly loading the output gear teeth in the acceleration environment  of orbit insertion was one source of applied load. The force from aerodynamic loading on the large area  array and antenna during aerobraking was another source of loading. These conditions together defined  the magnitude of the loads that would be applied to the harmonic gear output teeth in flight. It was  determined that the applied loads could be carried with appropriate margins by a size-32, cup-type  harmonic gear component set.  The output torsional stiffness of the actuator axes affects the spacecraft dynamics. The spacecraft  sensitivity to jitter disturbances and the attitude control system authority dictated a minimum natural  frequency for the deployed solar array wings and High Gain Antenna. Since the gimbal actuator was a  significant contributor to the overall appendage stiffness and the harmonic gear teeth were the load  reacting devices, an output stiffness was required of the chosen harmonic gear that exceeded its  standard specification significantly. The magnitude of this stiff ness increase for the selected size of  harmonic gear was within the range of experience for this type of application on other programs. Since  there was no significant difference in the stiffness requirement when compared to other heritage  programs, this was not considered a significant risk to the program.  The selection of the harmonic gear materials from the available set involved comparing the MRO  requirements with those of previous flight programs in order to maintain as much heritage as possible. A  fundamental tenet for this and many programs was to use only corrosion resistant materials in all space  mechanisms as is commonly done in the medical, semiconductor, and food processing industries. The  materials of the various pieces of the MRO harmonic gear component set are listed in Table 1. This  combination of materials had been used successfully in several programs with stiff ness and total lifetime  revolutions requirements that were similar to MRO. Table 2 lists the other heritage applications of the  same material combination with similar functional requirements. To maximize the life cycle capability of all  40 {'source': 'AMS_2006.pdf', 'page': 54}"
"### Human: What was the operating temperature range required for the Braycote@ lubricant family, and why was it not used in the Penzane lubricant family?### Assistant:of the actuator components, Penzane 2001-3PbNp Oil was utilized in the motor bearings and Rheolube  2004 grease was chosen for the lubricant throughout the harmonic gear. Other flight applications with the  same material combinations for the harmonic gear components had also used Braycote@ because of a  much lower operating temperature requirement than MRO. Since the required operating temperature  range did not necessitate the use of bray oil or grease, the Penzane lubricant family was chosen since it  tends to demonstrate more life capability over the Bray series when the operating temperatures are  moderate.  Table 1. Commercially Available Materials for  Note: The Melonite process is a Nitrocarburizing case hardening per SAE-AMS-2753B  Hardware Verification  The motor and gearbox assembly was the same for all of the six axis applications on MRO, allowing one  life verification 6 be performed on the worst case loaded design that would encompass all of theother  five axis applications on MRO. The life verification program for these actuators consisted of operating a  harmonic drive assembly alone in ambient environment with standard lubrication, followed by operation of  an assembled flight-like actuator assembly in a vacuum with thermal cycling and a representative wiring  harness for loading of the output gear.  The lone harmonic drive gear assembly was operated in a standard gear test fixture used at the harmonic  drive vendor. This was performed at ambient pressure and temperature in a bath of low viscosity  commercial oil at an input speed of 1750 RPM and with an applied load on the output that matched the  cable loading from harnesses. The wave generator was driven for 32 million revolutions with no sign of  unusual wear or failure of the bearing or the gear teeth. This was done without incident to a rotational life  of five times the flight requirement of 6.2 million input revolutions, indicating harmonic gear rotational  lifetime was a low risk.  Next an Engineering Development Unit (EDU) actuator was operated in a flight-like configuration. The  EDU actuator used the flight housings with output hardware that supported the High Gain Antenna cable  management system. The cable management system was incorporated to cycle the cabling as well as  provide flight like output loading to the harmonic gear. The High Gain Antenna application was chosen  because it has the largest number of cables across any of the gimbal axes. The EDU actuator was  operated over a total output angle of 340 degrees. The EDU actuator was run in a thermal/vacuum  environment, with the temperature slowly cycled from -25°C to +40°C at the rate of one thermal cycle  every 18 hours. The actuator was driven at a motor rate of 125 RPM for approximately 3.8 million input  revolutions, and then run at 65 RPM for the remainder of the time. The EDU motor was driven with an  industrial stepper motor driver, severely limiting visibility into the performance of the actuator. The stepper  motor driver was used for this operation for several reasons: flight actuator drivers were not available for  the start of the running, the motor is a two-phase brushless dc (not three phase), and the rotary life was  believed to be low risk so the limited visibility was not considered to be significant. The EDU life setup is  shown in Figure 6.  41 {'source': 'AMS_2006.pdf', 'page': 55}"
"### Human: What is the material used for the Heritage Hardware from Previous Programs with Similar Stiffness and Life requirements as the MRO Applications in Table 2?### Assistant:Figure 6. Engineering Development Unit Configuration in Vacuum Chamber  Flexsp I i n e Table 2. Materials and Surface Finishes of Heritage Hardware from Previous  Programs with Similar Stiffness and Life requirements as the MRO Applications  Lubricant Program  (Harmonic Drive Size)  MRO  (size 32) 15-5 PH H1150  Melonite  15-5 PH H1150  Melonite  15-5 PH H1075  Melonite  15-5PH H1075 (size 25) Penzane  2001 -3Pb Rheolube  2004  Braycote 602  Bray 81 52 oil  Penzane  2001 -3Pb Rheolube  2004  Penzane  2001 -3Pb Rheolube Circular Spline  Program #5  (size 32) 15-5PH H1075  2004  15-5 PH H1075 15-5 PH H1150 Rheolube 2000 with  Melonite 3% lead Napthenate 15-5 PH H1075  15-5PH H1150  15-5 PH H1150  Melonite  (sbe 40) I Melonite I I 2001 -3Pb Rheolube  42 {'source': 'AMS_2006.pdf', 'page': 56}"
"### Human: What was the maximum input revolutions reached by the EDU actuator before it started showing signs of improper operation?### Assistant:The EDU actuator operated up to 6.1 million input revolutions, when the output telemetry indicated the  actuator was not following the input signals properly. Later analysis revealed there were indications of  improper operation as early as 4 million input revolutions that were not diagnosed due to limitations in the  test setup with the stepper motor driver. The EDU actuator was removed from the chamber,  disassembled, and inspected. Figures 7, 8, and 9 show the condition of the harmonic gear teeth at this  inspection. Note that the tooth profile was completely obliterated and damaged across the entire width of  the flexspline teeth and most of the width of the circular spline.  I  i Intacl  Figure 7. EDU Circular Spline Life Damage Figure 8. EDU Flexspline Life Damage  Figure 9. Magnification of EDU Flexspline Life Test Damage  43 {'source': 'AMS_2006.pdf', 'page': 57}"
"### Human: What was the result of testing the harmonic assembly from each of the two remaining Melonite process groups with thermal cycling, output load, and input speed of 130 RPM?### Assistant:Once the dramatic failure of the harmonic gear assembly was seen, a detailed review of the  manufacturer’s documents uncovered that there were some problems with the Melonite coating on the  group of flexsplines that included the life unit, labeled Lot B. Two additional Melonite processing groups of  flexsplines had been received at the time of the failure, labeled Lot C and Lot D. Operation of these units  was started to determine if the Melonite processing on the EDU life test unit was the source of the failure.  A harmonic assembly from each of the two remaining Melonite process groups was placed in a harmonic  gear fixture (not in the gimbal actuator) and operated in vacuum with thermal cycling over a 12-hour  period, an output load, and an input speed of 130 RPM. Additionally, one of the two units was tested  using the same Penzane family of lubricants as the EDU and the other was tested using Braycote@ 602.  Both assemblies failed at approximately two million input revolutions. Figure 10 shows the failed splines  that were tested with the Penzane family of lubricants. Figure 11 shows the failed splines that were tested  with the Braycote@ 602 lubricant. A significant result of these two tests was that the lubricant type made  no difference at all, with both assemblies failing at nearly the same number of revolutions with the same  type and level of damage.  I  Figure 10. Circular Spline and Flexspline from Harmonic Only Test #1  Tested with Penzane  Figure 11. Circular Spline and Flexspline from Harmonic Only Test #I2  Tested with Braycote@ 602  44 {'source': 'AMS_2006.pdf', 'page': 58}"
"### Human: What is the estimated contact stress in MPa for the harmonic gear teeth based on rough mathematical analyses performed in the document?### Assistant:Since the lubricant type made absolutely no difference in the life capability of the harmonic drive gear  teeth, it was clear the problem was related directly to the material combination of the components and  their internal stress level. The appearance of the failure surfaces gave the impression of a possible galling  condition at work, but it was not clear if galling was the initiator of the failure or a consequence of the  damage once the failure had been initiated. Rough mathematical analyses were performed to estimate  the contact stress at the harmonic gear teeth from the preload and under the applied load in the  operation. The constant external load in the harmonic fixture runs represented the worst-case load from  the cable management system and was responsible for a roughly 40% increase in tooth contact stresses  over the preload. The estimated contact stresses in the EDU life test were in the realm of 750 MPa. The  galling threshold listed for a 15-5 PH stainless steel contact pair is around 14 MPa. This indicated that the  contact stresses compared to the galling threshold for the selected materials was a strong candidate for  the cause of the failures. To minimize the schedule time to a solution, the next group of tests used  Nitronic 60 (another available harmonic gear material) with a listed galling threshold value greater than  345 MPa. Nitronic 60 was identified as a candidate in addition to other standard commercial materials for  the harmonic gear assemblies, like nodular iron. Also, the output stiffness of the assembly was reduced  so the internal harmonic gear preload (and with it the internal tooth stresses) could be reduced as well. At  this point in the project schedule, the flight solar array panels had been fabricated and their stiffness was  measured, allowing reduction of the stiffness margin for these panels in the MRO spacecraft stability  analysis. This made it possible to reduce the gimbal actuator stiffness requirement significantly.  To investigate these issues within the remaining program schedule, three readily available harmonic drive  assemblies were procured with different material combinations and tested on the harmonic gear fixture.  The material combinations consisted of a unit with a nodular iron circular spline and an E4340 flexspline.  The second unit had a Nitronic 60 circular spline and an E4340 flexspline. The third assembly was  composed of a Nitronic 60 circular spline with an E4340 flexspline processed with a Melonite surface.  Additionally, the internal preload of the harmonic gear (to obtain the new required output stiffness) was  reduced in order to lower the gear tooth internal stresses. Table 3 lists all of the life test units and the  material and lubrication configuration for each.  F Table 3. Selected Harmonic Gear Material and Lubricant Configurations  te: LM stands for Lockheed Martin  45 {'source': 'AMS_2006.pdf', 'page': 59}"
"### Human: What is the difference in revolution life to failure between the EDU Assembly #1 unit and the LM Harmonic Assembly #1, based on the results shown in Table 4?### Assistant:The test units for LM Harmonic Assemblies #3, #4, and #5 were tested in the harmonic drive test fixture,  in the environment, with the applied load and input speed listed in Table 4. The stiffness of the units was  measured at the start of the testing as shown in column 2 of Table 4. The life tests already discussed are  listed in Table 4 for completeness.  Test Unit Table 4. Life Unit Stiff ness, Environments, Conditions, and Revolutions at Failure  Initial Stiffness  with 45/140 N.m Applied Approx.  Applied Torque Load Input Revolutions  (N-mlrad) Environment on Output Speed to failure  vendor  #I (in actuator)  LM Harmonic  Assembly #1  (same unit as  Harmonic drive  vendor Unit #1)  LM Harmonic  Assembly #2  LM Harmonic  Assembly #3  LM Harmonic  Assembly #4 Approx. 68,000  > 56,000  Approx. 68,000  69,000/75,000  56,000  55,600/60300  56,500 (28 N.m)  43,800 (28 N.m)  54,600/61,000  43,800 (28 N.m)  37,300 to 46,300  (28 N.m) Ambient  Vacuum, -25""C/+4OoC  18 hour cycles  Vacuum, -1 5""C/+40°C  12 hour cycles  Vacuum, -1 5""C/+40°C  12 hour cycles  Vacuum, -1 5""C/+4OoC  12 hour cycles  First 3M revs  Vacuum, -1 5""C/+40°C  12 & 24 hour cycles  Next 3M revs  Vacuum, +23""C  constant  Next 8M revs  Vacuum, -1 0°C  constant  Vacuum, -lO""C/O""C  constant  Vacuum, -1 O""C/+l 0°C  constant 11.3 N.m  HGA Cable  Harness  11.3 N.m  11.3 Nom  11.3 N.m  11.3 N.m  11 -3 N.m  HGA Cable  Harness 1750 RPM  125 RPM  & 65 RPM  130 RPM  140 RPM  140 RPM  140 RPM  140 RPM  60 RPM No Failure  4M  2M  2M  4M  No Failure Q  14M  No Failure Q  9.1 M  No Failure Q  13M  Key results of the above are as follows:  0 The high input speed of the harmonic gear only operation performed at the harmonic drive vendor  allows the lubricant to support high contact stresses that would otherwise result in complete  failure of the gear teeth.  The difference from the EDU Assembly #1 unit and the LM Harmonic Assembly #l was the  Melonite lot and coating details. The results of the LM Harmonic Assembly Unit #1 showed that  the Melonite coating was not involved in the failure.  The applied output load, while small when compared to the maximum torque capability of the  harmonic gear, made a difference in the revolution life to failure. Note that the EDU life unit  exhibited failure around 4 million revolutions, while assemblies #1 and #2, with a constant applied  load, failed at 2 million revolutions. This difference could be due to the EDU assembly #1 using a  cable wrap harness, with a variable load depending on output position, compared to a constant  load. 0  0  46 {'source': 'AMS_2006.pdf', 'page': 60}"
"### Human: What was the material selection for the circular spline and flexspline in Harmonic Assembly #4, and why was this material selection critical in achieving the required output stiffness without compromising the reliability of the harmonic gear assembly?### Assistant:LM Harmonic Assemblies #1 and #2, with failure at a similar number of revolutions using very  different lubricants, indicated the lubricant was not a significant player in the failure mechanism.  This eliminated the lubricant as a variable in further failure investigation.  The LM Harmonic Assembly #3, with nodular iron and a reduced preload, had a longer life to  failure than the stainless steel, supporting the galling hypothesis. Note the regions of damage in  Figure 12.  The LM Harmonic Assemblies #4 and #5 used Nitronic 60 for the circular spline, the highest  galling threshold material that could be obtained in a harmonic gear assembly. The late date of  the testing permitted a reduction of the output stiffness to two-thirds of the initial values used for  the gimbal actuators. The stiffness of the unit in Harmonic Assembly #4 was reduced as the  operation progressed and a method of setting the harmonic gear preload was established. The  result of the run was no failure at all with some minor wear of the harmonic gear teeth, as shown  in Figure 13. The schedule dictated that the first successful combination be used, so the final  material selection was a Nitronic 60 circular spline and a flexspline of E4340 with no additional  surface processing. Note that schedule dictated changing more than one variable at a time.  The setting of the internal preload of the harmonic gear assembly during assembly at the  harmonic drive vendor was critical in achieving the required output stiff ness without compromising  the reliability of the harmonic gear assembly. 0  0  Figure 12. Circular Spline and Flexspline fromHarmonic Only Tex#3  .A n.ri+h +ha ,-ir,.nalmr rnlins A  =:, ('.  Figure 13. Circular Spline and Flexspline trom harmonic vnty I esr lf4  Note the slight wear region on the circular spline showing where the flexspline was engaged with it.  47 {'source': 'AMS_2006.pdf', 'page': 61}"
"### Human: What is the purpose of increasing the output stiffness of a harmonic drive assembly, according to the text?### Assistant:Harmonic Gear Assemblv Internal Preload Settinq  The cup type harmonic gear assembly has zero backlash due to the angle of approach, along the rotation  axis, of the flexspline teeth relative to the circular spline teeth (Figure 14). This arrangement leads to a  two-sloped stiffness curve of the output of a harmonic gear. As torque is applied to the gear, the cup  flexes to allow more of the width of the teeth to engage with the circular spline. As more tooth area is  engaged, more load sharing occurs and the stiffness increases. Once the angle between the teeth of the  splines is reduced to near zero, the stiffness has reached its maximum value. As the torque is increased  from this point, the stiffness is related to deflection of the individual teeth and the change in roundness of  the housing and circular spline. Figure 15 shows a stiffness curve for a typical harmonic gear assembly  prior to increasing the internal preload for stiff ness improvement.  Flexspline  tooth angle - Cup flexes to allow more  tooth face engagement / /Flexspline  Circular Spline  tooth angle  Figure 14. Harmonic Drive Cross Section Showing Significant Source of Variable Stiffness  Note the largest stresses on flexspline teeth occur at the point of initial contact, the open end of the cup.  Drawing courtesy of Harmonic Drive, LLC  Increasing the output stiffness of a harmonic drive assembly involves increasing the diameter of the wave  generator in the area where it forces contact between the teeth of the two splines. This is accomplished  by using a different wave generator plug with a larger major diameter of the oval. As the diameter is  increased, the flexspline cup is deflected in the direction of engaging more of the face of the teeth. This  has the same effect as increasing the torque on a nominal unit in the low stiffness region. As the wave  generator plug size is increased, the low stiffness region gets smaller. This trend continues until the  stiffness curve is essentially straight. At this point, a further increase in the diameter of the wave  generator plug will increase the overall gear assembly stiffness and significantly increase the internal  tooth stresses. Figures 16 through 19 show the how the shape of the stiffness curve changes with  different wave generator plugs in a harmonic drive assembly. The low stiffness region is very evident in  Figure 16. A larger plug, after insertion into the wave generator bearing may nearly eliminate the low  stiffness region, as seen in Figure 17. Figures 18 and 19 show the next two larger size wave generator  plugs, without a significant change in the shape of the stiffness curve. The desired operating point for the  flight plug is smallest wave generator plug that exhibits a fairly straight curve. If there is any uncertainty  between units, the smaller one would always be installed to guarantee that the gear teeth were not being  jammed together with high, and unknown, internal stresses. The wave generator plug used in this  example would be Figure 17’s.  48 {'source': 'AMS_2006.pdf', 'page': 62}"
"### Human: What is the typical harmonic gear assembly stiffness shown in Figure 15, and how does it vary with rotation angle?### Assistant:Measured Initial Harmonic Drive Stiffness  Rotation Angle  Figure 15. Typical Harmonic Gear Assembly Stiffness  Note two regions of stiffness: low near zero torque and larger at high torques.  The following figures are from measured data on the fliaht harmonic aear assemblies:  (Note: the following figures' axes are rotated relative to Figure 15)  .,~ __-......._.I *_. . :, r. . ,. . .. . . . . . . . . . . . .. . , ..  I  Applied Torque Applied Torque  Figure 16. Stiffness Curve Showing  Low Stiffness Region Figure 17. Stiffness Curve with Minimal  Low Stiffness Region  49 {'source': 'AMS_2006.pdf', 'page': 63}"
"### Human: What is the maximum stiffness enhancement achievable without compromising the reliability of the gear assembly for long life applications, according to the text?### Assistant:ji  Applied Torque Applied Torque  Figure 18. Next Larger Plug Size Over Figure 17 Figure 19. Next Larger Plug Size Showing  Showing a Small Change in Shape Showing Little Change  Conclusions and Lessons Learned  The most significant result of the failure investigation was determining that the internal stresses due to the  preload and the cable harness loading caused the premature failure of the EDU harmonic gear assembly.  In order to maintain reliability for long life applications, increasing the wave generator plug size (a service  provided by the harmonic drive vendor) is an acceptable technique up to the point where the low stiffness  region is eliminated. This is the maximum stiff ness enhancement achievable without compromising the  reliability of the gear assembly for long life. Any further increase in the wave generator plug size will  increase the stiffness at the cost of reducing the life of the unit. For minimal life applications, increasing  the stiffness beyond this point may still be acceptable.  Stiffness enhanced harmonic gears are very sensitive to the externally applied load and test environment.  Since the failure mode is galling, the presence of any gas (nitrogen, for example) severely compromises  the test results. Life capabilities from previous heritage programs had been successful and so the initial  gear material for the MRO gimbal actuators was considered acceptable and robust. However, some of  the heritage operation had been performed in nitrogen, instead of vacuum. When enhanced stiffness is  required in a harmonic gear application and it is not being used in a preload configuration represented by  Figure 16, the situation is sensitive to possible galling of the harmonic gear teeth. Performing harmonic  gear component operation at loads above the planned level to increase the tooth contact stresses should  be considered. This will demonstrate if internal stress margin exists in the hardware. Also note that  running a harmonic gear at a high input speed to reduce the operating time is not adequate. The high-  speed condition may function with no incidents while low speed operation may catastrophically quit  functioning. Finally, operating a unit at nominal contact stress levels to a larger number of revolutions  than planned is a necessary, but not complete, margin demonstration program. A catastrophic failure may  be lurking just a few megapascals away from the nominal value.  References  Harmonic Drive, LLC web site, www.harmonic-drive.com/suppor-t/principals.htm, description of harmonic  gear assembly operation  Acknowledgements  This work was performed at Lockheed Martin Space Systems, Denver, Colorado under a spacecraft  system contract to the Jet Propulsion Laboratory, California Institute of Technology, under a contract with  the National Aeronautics and Space Administration. 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, the Jet Propulsion Laboratory, Pasadena, California, or  Lockheed Martin Space Systems, Denver, Colorado.  50 {'source': 'AMS_2006.pdf', 'page': 64}"
"### Human: What is the length of the boom when it is stowed for launch, according to the text?### Assistant:Stacer Driven Deployment: The Stereo Impact Boom  Robert Ullrich’, Jeremy McCauley*, Paul Turin*, Ken McKee* and Bill Donokowski*  Abstract  The Impact’ Booms carry 3 scientific instruments each on the twin NASA Stereo2 satellites. When stowed  for launch the boom is 1.54 m in length, when deployed the boom extends to 5.80 m. The boom consists  of 5 concentric graphite epoxy (GlE) tubes extended by the spring force of a Stacer. The Stacer is  fabricated from a flat strip of Elgiloy spring material, rolled with a constant diameter and fixed helix angle.  It supplies the motive force for deployment, and requires no external power once released. The deployed  boom exhibits excellent rigidity, the natural frequency first mode occurring at 1.96 Hz. Discussed is the  implementation of a Stacer to deploy the 5 segment telescoping boom and some of the activities  performed during its design, qualification and testing.  Mission Introduction  The NASA Stereo mission consists of twin, three-axis stabilized satellites orbiting and viewing the Sun in  the plane of the ecliptic at -1 AU. Spacecraft A (Ahead) will be sent into an Earth preceding path with an  Earth-Sun-Spacecraft angle increasing at e rate of 22’ per year, while Spacecraft B (Behind) is sent into  an Earth lagging orbit, also at a rate of 22 per year. The imagers on board will yield true ‘stereoscopic’  views of coronal mass ejections, while other instruments perform concurrent in-situ measurements of a  large portion of the electro-magnetic spectrum. The telescoping boom was conceived to interface 3  instruments from the Impact suite: the Magnetometer (Mag), the Solar Wind Electron Analyzer (SWEA),  and the Supra-Thermal Electron - Downward looking instrument (STE-D) to the Stereo spacecraft. The  program requirements demanded a new concept, as existing hardware was deemed too expensive or  unsuitable. The boom was initially developed via three ‘proof of concept’ models for the tube locking  mechanism, and a final mock up using the Stacer to deploy four concentric, telescoping graphitejepoxy  tubes from the center of the fixed 5th tube. An engineering model (EM) was then built to verify end to end  design via qualification testing. The challenge for this mechanism was demonstrating that the design met  the requisite GEVS SE3 force (torque) margin. Two flight models (FMs) were then produced, with the EM  being refurbished as a flight spare. The FMs are currently mounted to the spacecraft and mission I & T is  progressing. Launch is scheduled for May 2006 from Cape Kennedy on a Delta II.  Figure 1 Magnetometer, SWEA and STE-  D mounted on the end of the Istowed) Figure 2. Magnetometer  ’ Space Sciences Lab, University of California, Berkeley, CA  ’ In-situ Measurements of Particles And Coronal mass ejection Transients  * Solar - TErrestrial RElations Observatory ’ See References Section  Proceedings of the 3dh Aerospace Mechanisms Symposium, Langley Research Center, May 1 7- 19,2006  51 {'source': 'AMS_2006.pdf', 'page': 65}"
"### Human: What is the minimum boom length required for the Impact mission to avoid sensing the spacecraft, as per the EMVEMC design?### Assistant:The Science Flowdown Requirements  The requirements for the Impact boom were based on the scientific needs of the three instruments  mounted to it: the Mag (built at Goddard Space Flight Center); the SWEA, supplied by CESR, Toulouse,  FR; and the STE-D, supplied by UCB - SSL. The magnetometer for this mission is very sensitive: the  magnetic (B) field at 1 Au heliocentric orbit is - 3 to 4 orders of magnitude smaller than near Earth. This  slight field strength was a driver for the EMVEMC design for the spacecraft and the devices near it. To  avoid 'sensing' the spacecraft, the Mag needed to be 3 meters away from it. This requirement set the  minimum boom length. To ensure a low magnetic signature from the boom assembly, no ferritic alloys  were allowed for its construction.  Titanium screws were used to mount the magnetometer to its tray on the 4'h tube element and the tray  itself is made of carbon impregnated PEEK, a high-strength, conductive engineering plastic. To lower  Mag exposure to any eddy currents present in the harness or structure of the boom, the mounting tray  offset the Mag 200 mm from the nearest metal on the tube. Additionally there were not to be any other  instruments closer than 1 m to the Mag. To allow accurate inter-experiment correlation of data, the  angular alignment accuracy and repeatability requirement has an allowable deviation of <0.88' (52.5  arcmin) between the Impact boom mounting feet and the magnetometer  housing from the stowed condition to the deployed state, for the two axes  that form the mounting plane of the Mag.  The second experiment mounted on the Impact boom is the Solar Wind  Electron Analyzer (SWEA, supplied by CESR, Toulouse, FR). The SWEA  has two variably charged hemispheric surfaces that attract electrons into  an anode assembly which counts them as they impinge on to it. The SWEA  would have a limited field of view when mounted directly to the spacecraft  deck, hampering its ability to characterize the electron regime in the  volume around it. Proximity to the spacecraft also causes deflections of the  electrons due to the almost unavoidable static fields that develop near the  spacecraft surface. Since this effect is difficult to model, a better solution  was found. The implementation of the SWEA on the Stereo mission is  extremely good: it is mounted on the extreme end of the boom, allowing a  full 2n radians x 135O field of view (FOV). The demands of the SWEA for  mounting to the boom are not complex: power lines, command lines supplied to it, data return lines from  it. Figure 3- SWEA 1 STE  The final instrument is the STE, a Supra-Thermal Electron detector, mounted on the side of the SWEA  pedestal. It needs a clear 80' x 80° field of view looking along the plane of the ecliptic at a 45' angle  (aligned with the Parker spiral), and to stay at - -4O'C.  The fixed base of the Impact boom is pointed towards the Sun for all science activities, and the only off-  points scheduled are for momentum dumping. The boom deploys away from the Sun, so that there is  minimal solar input to the boom suite, giving very low operational temperatures for the instruments.  Thermal control was a large concern for the instrumenters.  52 {'source': 'AMS_2006.pdf', 'page': 66}"
"### Human: What is the purpose of the aluminum rings bonded to each end of the Impact Boom's Gr/E tubes?### Assistant:e  Figure 4. Solid Model Section View of Stowed Impact Boom  Impact Boom Mechanism  The Impact Boom consists of five concentric Gr/E Tubes, ranging from 50 mm to 210 mm in diameter,  with a pair of aluminum rings bonded to each end. Each ring pair contains three lock pins, pointing  outward at the Sun-ward end, and inward pointing at the release mechanism end, and three sockets, in  their corresponding orientations. The pins are spring loaded, and have rollers mounted in their tips. When  released, the Stacer spring extends the tubes until the end of travel where the pins drop into sockets,  locking the assembly rigid. The mounting feet are integral to the outermost rings on the 0210-mm tube.  There is a spool for the electrical harness while stowed, a flyweight brake to govern the deployment  speed, a shape memory alloy release mechanism with pretensioning adjustment, deployment assist rods  and kick springs to initiate the deployment. Combs at each end hold the tubes in alignment prior to  deployment, and during vibration/launch. A provision for individually adjusting the combs to remove any  play in the stowed tubes due to fabrication tolerances is also provided. The design was performed in  SolidworksTM, utilizing its 3-dimensional solid modeling and multiple configuration capabilities.  Tube Details  q-  .I I The use of telescoping concentric tubes is not a new idea. Each telescope  application brings its own set of challenges however. For the Impact  Boom, the tubes needed loose tolerances on their cylindricity callout to  allow for simple tube manufacture. The deployment / locking scheme  required compliance regarding the inter-tube fit since there is a relatively  low force available from the Stacer. The boom needed to be very rigid  when deployed, so locking pins were utilized at end of travel, rather than  relying on spring force to hold them in place. The tubes have three  longitudinal concave grooves equispaced  about their circumference, running their  length, with a precise profile that doesn’t jam  the rollers. These grooves kept the pins on  track to be aligned with the sockets at the far  end of the tube. The tubes are a five-layer  Gr/E composite ~ designed to be quasi-  isotropic: three 0 -90 layers interleaved with  two 45’ layers of 0.1 2-mm woven epoxy pre-impregnated material (Fiberite  Hy-E 1034C prepreg). The tubes were fabricated by Vision Composites of  Signal Hill, California on internal mandrels, with a slight taper to enhance  ease of extraction after cure. The cure regimen was specified to be ‘dry’: the  ratio of epoxy to carbon filament was held to a minimum to ensure low  surface resistance. This was achieved by using a higher autoclave pressure  with a slower ‘warm up to cure’ temperature ramp and a thick layer of  absorbent over the bleeder sheet. The process determination was somewhat  lengthy, however, the final result met requirements. ’  Figure 5. Tube  End Detail  Figure6 Partially  Deployed Tubes  53 {'source': 'AMS_2006.pdf', 'page': 67}"
"### Human: What was the maximum stress value predicted for the glue line in the FEA analysis, and how did it compare to the published maximum stress value of the glue?### Assistant:At both ends of each tube there are inner and outer interlocking aluminum rings that ‘sandwich’, and are  bonded to, the Gr/E. After an extensive search, Loctite Hysol 9309NA was used to form the bond. The  thermal environment for the boom is rather severe: it will be in the shadow of the spacecraft for all of the  science activity; thermal analysis estimates put the operating low temperature at 30K. There is very little  data for epoxies at this temperature, so we performed an FEA for the bond between the aluminum and  the Gr/E to establish what parameters minimized the stresses in the glue, the weakest part. Several  cases for the glue design were examined: thickness of the bond, edge conditions of the bond and effects  of the aluminum ring thickness (internal and external) on the joint stresses. Optimized, the glue line was  1 vo. mue. 1 vo. YUU  Figure 7. FEA of Aluminum Ring - Glue Line - Gr/E - Glue Line - Aluminum Ring  chosen to be -0.4-mm thick, with a fillet onto a tapered aluminum ring edge. This best case predicted a  stress value of 11OvMPA (16 k~i)~, which exceeded the glue published maximum stress value of 38 MPa  (5.5 ksi)’.  There was concern that the joint would not be sound after exposure to the thermal gradient, so an actual  test was needed. A test GR/E tube / aluminum ring  assembly was fabricated, with a large cantilever  mass attached at the extremity providing -2X  expected loads for this test. This test ‘tube’  assembly was installed in a cryogenic liquid  helium chamber, which was then installed into a  cryogenic liquid nitrogen chamber in turn was  placed inside of a refrigerated chest. We  performed a multiple cycle thermal test (in a dry air  environment), utilizing four candidate gl!es. The  cold temperature was set to 25K (-248 C), and  nominally, with no crazing or cracking and was warm was 150K (-1 23’C). The 9309 performed  accepted for use. As a side note, only one of the tested epoxies exhibited any signs of thermal distress.  As a hedge against exceeding our thermal predicts, small solar absorbers were attached to the two joints  in the mid boom. These little flags raise the expected temperature by -5 degrees, buying some margin for  the assembly (Figure 6). I  Figure 8’ Epoxy Thermal Test Chamber  Besuner Consulting, Madera, CA 93638  Loctite Hysol Applications Note, April 101, Loctite Aerospace, Bay Point, CA 94565 4  54 {'source': 'AMS_2006.pdf', 'page': 68}"
"### Human: What is the purpose of the locking pins in the Impact Boom, and how do they provide rigidity to the boom during deployment?### Assistant:Lock Din details  Once the rings are bonded to the tubes, the locking pins / rollers are installed into precision radial bores in  these inner and outer rings. When the segment locks, the pins are pushed into the sockets by custom  wound torsion springs, two for each pin. The ‘arms’ of the springs also  provide alignment for the pins, keeping the roller in the groove. The pin /  roller combination allows any tube dimensional change to be insignificant  during the deployment, as the spring compliance takes up any bumps or  dips. When the tube section reaches end of travel, the locking pins are  pushed into tapered holes, causing the tubes to become rigid with  respect to each other. Guide ramps are provided at the end of travel to  ensure that the pins are aligned with the sockets. Each locking pin has a  slight taper (i.e., <20° included angle) that fits into the corresponding  tapered socket. This gives a ‘self-locking’ feature to the pin, offering  increased rigidity and prevents the pins from backing out under slight  vibrations. With six lock pins engaged per joint, three inward acting, three  outward, the boom exhibits great rigidity, and offers redundancy in the  event that one pin (or more, up to three maximum, as long as they are not in same ring) does not lock. Figure 9. 50-mm Tube  End Lock Pins (EM)  Rollers are fitted to the tips of the locking pins to minimize deployment drag when rolling in the tube’s  grooves. Repeated deployments have shown no signs of wear to the tube or the rollers. To provide  conductivity, the tapered portion of the pins and sockets were Alodined, while the sliding cylindrical parts  were Type Ill black anodized to give good wear and low friction sliding properties. The Gr/E exhibits a low  surface resistance too, enabling the boom to easily meet the surface resistance requirement of <lo8  ohms per square, throughout its stroke. The drag was measured to be 3.1 N on average for the  assemblies. The main function of the rollers is to keep the tubes aligned during deployment, so that the  pin engagement is virtually guaranteed at the end of stroke.  Shape Memorv Allov Release (SMAR) details  The SMAR uses the interesting phase change properties of a 50% titanium - 50% nickel alloy (trade  named Nitinol initially) to provide the actuation of the Impact Boom. This device, pioneered by TiNi  Aerospace‘ in cooperation with UCB-SSL, takes advantage of the -4% dimensional change in the drawn  alloy wire when heated above its transformation temperature to let a ball detent assembly release a large  spring loaded retracting pin. Since there was a large design load (50 Gs), >2.5 kN retraction force was  needed. A force amplifier was added to the TiNi standard P50 (-200-N [50-  Ib] pin puller). The force amplifier contains a stack of Belleville washers,  preloaded and held by the P50 pin in another ball detent assembly,  providing a final pull force exceeding 3 kN. When an electric current is  passed though the Ti-Ni wire, it changes phase, elongates, releasing the  primary pin, which then retracts and releases the main pin, which retracts  with great force, allowing the Stacer to deploy. The main benefit of using an  SMAR, aside from increased safety as no explosives are used, is that the  flight unit can be tested over and over again (hundreds of cycles), and is  simply resetable with a hand tool, with no temperature or time dependant  constraints. I  1 a  Flvweiaht Brake Figure 10. Shape  After the SMAR has been triggered, kick springs push the tubes out of the  combs and the deployment assist device pushes the Stacer out of the  canister with a force of -90 N, giving the assembly a good initial velocity.  The Stacer continues to provide force throughout the travel, so the deployment velocity would continue to  increase until a balance between drag and push is achieved. This balance is never reached by the boom, {'source': 'AMS_2006.pdf', 'page': 69}"
"### Human: What is the purpose of incorporating a means to limit deployment velocity in the Impact Boom, according to the given text?### Assistant:increase until a balance between drag and push is achieved. This balance is never reached by the boom,  so the deployment velocity reaches a ‘run-away’ condition rapidly, with the possible issues of lock pin  shearing, ring-tube separation or other damage as consequences. As with every Stacer, a means to limit  deployment velocity is incorporated. For the Impact Boom, a flyweight brake mechanism is attached to Memory Alloy Release  Assembly  TiNi Aerospace, San Leandro, CA 94577 6  55 {'source': 'AMS_2006.pdf', 'page': 69}"
"### Human: What is the purpose of the flyweight brake on the Impact Boom, and how does it work?### Assistant:P % the Stacer via a woven Dacron (parachute shock) cord. Similar to the device found on (old) dial  telephones that prevented the dial from being rotated faster than an old telephone exchange could count,  the flyweight brake supplies braking force proportional to the rotational speed of its weight assembly. If  the force (speed) increases, the brake shoes are centripetally accelerated against the brake drum,  increasing the braking force and slowing the rotational velocity. Over a wide range of forces, the brake  typically can control the speed to *lo%. For the Boom, a deployment  velocity of -0.5 m/s was chosen. This allows a certain momentum to  build, but is slow enough to avoid shearing damage to the lock pins at  the end of travel.  Harness and SDOOI  The power and electrical signals between the data processing unit and  the instruments are carried by a cable routed down the center of the  tubes, and is stowed on a spool for launch. This harness is a custom-  fabricated conductor assembly consisting of seven coaxial cables and  five twisted shielded pairs. Built of silver-plated copper with Gortex@’  dielectric, the harness is wound onto a bobbin when stowed and is  pulled off when the boom deploys. This ‘straight through’ design  provides greater signal strength, higher reliability and allows longer  harness length as there are no slip rings or other connections between  the data processing unit and the instruments. Care is taken to prewind  the harness to avoid kinking or ‘birdnesting’ when stowing. I  Figure ,,. Flyweight  Brake 81 Harness Spool  The Stacer  The Stacer is a rolled, constant helical pitch, fixed diameter flat spring. The strip width, thickness, roll  diameter, and helical pitch are selectable for each application, allowing each Stacer to be tailored for  optimum properties. Stacers range in size from <1 m to >10 m in length, from 4 mm to 55 mm in diameter  at the tip, and can provide extensive force from almost nothing to 200 N. Trade studies can balance mass  versus length, force, etc. In the last 30 years, more than 650 units have been utilized in aerospace  applications, from sounding rocket sensors to gravity gradient booms with large masses on the end. What  makes the Impact Boom unique is the use of the Stacer as a spring ‘motor’ without using it as the  structure or sensor surface. Most applications have the Stacer with the sensor(s) mounted directly on it,  or the Stacer as the sensor, for example as an antenna (a total of six 6-m-long beryllium copper Stacers  are used on the Stereo satellites for the Swaves experiment in this manner).  I  .- -I  Figure 12. Picture of a Stacer  To accommodate the Mag requirement of low magnetic signature for the boom, Elgiloy@ was selected as  the spring material over the more traditional beryllium copper (Be-Cu). This alloy was chosen to minimize  any eddy currents that could be developed between the SWEA / STE and the spacecraft. Originally  invented in the late 1940’s in Elgin, IL for use in watches, it has been used for exacting Stacer  applications several times. Its internal resistance is higher than copper, and cuts down the eddy currents  accordingly. It has a higher modulus (E), and can provide greater force in the same physical volum7e as  the Be-Cu. Elgiloy is a cobalt ‘super-alloy’, having an E -1 90 GPa and a yield strength of -1 600 MPa .  7 Matweb, http://www.matweb.com/search/SpecificMaterial.asp? = Elgiloy  56 {'source': 'AMS_2006.pdf', 'page': 70}"
"### Human: What is the typical helix angle of the Stacer coils, and how does it contribute to the inter-coil friction and stability of the structure?### Assistant:At deployment, the formation of the Stacer starts with the initial coil winding out of the storage canister  onto a cylindrical tip piece, which is slightly larger than the free coil diameter of the Stacer. Thus the  Stacer grabs the tip piece tightly, and the subsequent coils ‘stack’ up on the prior, producing the  characteristic spiral appearance. The typical helix angle provides for significant overlap, such that a  section taken at any point along the Stacer would yield at least two thicknesses of strip material. Since  the ‘outer’ layer of strip is rolled to the same diameter as the inner layer, the outer grips the inner with a  force normal to the surface. So between layers, significant inter-coil friction exists and prevents inter-coil  slipping for small disturbances. This gives the Stacer one of its more useful properties: it behaves as a  thin walled tube for small displacements, with similar bending strength and stiffness. If a larger  displacement occurs, the coils slip, dissipating the strain energy, serving as a friction damper. The  damping ratio value is typically 5 - 15% for the non-slipping regime, and can reach 30 - 40% with the  slipping. Of course, the displacement limit is buckling, as any tube would experience when taken beyond  its yielding strength.  As described, the motive force for the deployment is a Stacer. When compressed (stowed) it is a very  compact package: it fits in a cylinder 050 X 130-mm long. When the Stacer is stowed, the strip is flexed  into the canister, laying each coil inside its predecessor, and wound tightly to the outside of the can.  When released, this stored strain energy is reclaimed, giving the motive effort needed to move the tubes  along their path. The Stacer generates a higher force at the beginning of stroke, -46 N for this  application, and the force curve dropped to 1 N at the end of stoke (this was an isolated minimum value  obtained from one force test). The force that the Stacer provides is shown in the polynomial fit curve in  Figure 13.  Impact Boom Stacer Push Force  50  25  20 3 0 1 2 3 4 5  Deployed Length (m)  Figure 13. Stacer Force  57 {'source': 'AMS_2006.pdf', 'page': 71}"
"### Human: What is the purpose of the deployment initiator in the Stacer system, as described in the context information?### Assistant:For the purposes of torque (force) margin analysis, the initial push is 45 N, final thrust 18 N (the lowest  value obtained). This lower value was used to bound the design force available for deployment. The  Torque (force) Ratio (tR) requirement from GEVS SE (Sec. 2.4.5.3) is:  t~ = tavaill trequird 2 3-0  and the Torque (force) Margin (tM) requirement from GEVS is: (Equation 1)  tM = tavail trequired - 1 2.0 (Equation 2)  for systems dominated by resistive torques due to friction. This assumes worst case for the boom, taking  the lowest force for the Stacer and applying it to the entire stroke. There is additional margin as there is a  significant mass at the end of the boom SWEA / STE-D which contributes momentum towards full  deployment stroke. Using the given values it can be shown that the force available, the minimum Stacer  force of 18 N, divided by the force required, the tube drag of 3.1 N yields a torque ratio of 5.8, and a  torque margin of 4.8. The Stacer satisfies the force requirements by analysis. Still, the device must show  functionality to prove that manufacturing has been in accordance with design.  The graph shows the need for a deployment initiator. The stowed Stacer is in a ‘meta-stable’ condition. If  left by itself, it would partially deploy in either direction, therefore a back plate on the canister is required.  To ensure that it deploys a deployment assist device (DAD) is incorporated. The final upturn in the force  curve is an artifact of how the force was measured. The Stacer in this case is 5-m long, and when it is  fully deployed, the coils have tightened onto themselves. The force value was taken at the moment the  Stacer began to slip back into the canister. For this case, the coil needs to be expanded significantly, and  requires greater effort.  There is an additional use for the Stacer after deployment as the secondary  EMVEMC shield. While each of the conductors in the harness is shielded, the  mission’s low noise requirement demanded a second, ‘over-shield’ for all  conductors. Since the harness runs down the center of the Stacer, the Stacer  was tied to ground, and serves this purpose.  Deployment Sequence  Deployment is initiated when a TiNi Aerospace shape memory alloy release  device (SMAR, Model P50-810-1 RS) is triggered causing the restraint pin to  pull out of the tail of the Stacer tip piece. To give the stacer and tube  deployment an initial ‘kick’, a deployment assist device (DAD) is incorporated  between the SMAR mount plate and the 50-mm tube base. The DAD consists  of three long coil springs compressed when stowed, and when released  provide -90 N of push at the very beginning of the stroke. After the first  100 mm of travel, the initial coils of the Stacer are fully formed around the tip  piece, and the flyweight brake has been spun up to speed. At this time the  DAD has completed its stroke. The Stacer is attached to the base of the 50-  mm tube via a swivel, allowing the Stacer to wind down while extending,  recapturing the strain energy stored when the Stacer was wound ‘out’ against  the canister. At the end of the 50-mm tube travel, the six lock pins pop into  their sockets, and transfer the Stacer push force, as well as momentum, to the  90-mm tube, pulling it along until it latches; the process continues with the  130-mm tube and the 170-mm tube, and finally the entire four tube rigid  assembly locks onto the 210-mm tube, which is fixed to the spacecraft. While  the actual sequence follows this description fairly closely, occasionally the  tube drag would cause one or another tube to partially deploy. There is no  provision or requirement for any tube to deploy in any set sequence. To  control the velocity of the tubes during deployment, the flyweight brake is  attached to the 50-mm tube via a lanyard, limiting the speed of deployment to  -0.5 m/s, giving a total deployment time of -10 sec. There are position {'source': 'AMS_2006.pdf', 'page': 72}"
"### Human: What is the maximum speed of deployment for the impact boom when it is attached to the 50-mm tube via a lanyard?### Assistant:attached to the 50-mm tube via a lanyard, limiting the speed of deployment to  -0.5 m/s, giving a total deployment time of -10 sec. There are position  alignment blocks for stowed (launch) condition holding the tubes aligned 1. i -.  L-he-  Figure 14. Impact  Boom after a thermal  vacuum deployment  58 {'source': 'AMS_2006.pdf', 'page': 72}"
"### Human: How did the Impact Team determine the sufficient force margin for extension of telescoping sections throughout the Boom's stroke during testing?### Assistant:relative to each other, and carrying the vibration loads. These also incorporate ‘kick’ springs to aid in their  deployment, and to alleviate any possible “stiction” from the alignment blocks.  The boom is not retractable once deployed. Re-stowing is achieved by removal from the spacecraft, and  hand retraction of each set of pins followed by each tube segment being (de) telescoped; after which the  Stacer is compressed into its canister; and the harness and flyweight brake are rewound. Finally, the  SMAR is reset reinstalled, and preload is set.  Verification  The Impact Boom’s qualification activities were based on GEVS SE, as modified by JHU-APL for mission  specific needs. The test regime selected for the Stereo mission was Protoflight, meaning new (unflown or  non-heritage) hardware is tested with a combination of prototype (EM) levels (i.e., temperature or  vibration) with flight (FM) durations. This method is typically used to shorten development times by  eliminating the engineering / qualification model fabrication and test  period. However, the Impact Team did build up an engineering model,  and tested all 3 assemblies to the protoflight levels. The main changes  and additions pertinent to this paper: Level 300 cleanliness, UV + Visible  light inspection, no silicones used for fabrication, and testing for silicone  residuals. Vibration levels were taken from the Delta II user’s manual*  modified by APL analysis for the ‘stacked’ configuration. Stringent  EMVEMC levels were levied, due to the extremely sensitive radio  receiver and magnetometer on board. r  Testing procedures were standard NASA mission fare. The main tasks to  be performed for this application were: demonstration of sufficient force  (torque) margin for extension of telescoping sections throughout the  Boom’s stroke; thermal design validation at 25K (discussed previously);  and thermal vacuum cycling and deployment verification at hot and cold  operational temperatures.  As the team worked on the testing it became clear that Stacer thrust  force is not easy to measure accurately or repeatably. The deployment of  a Stacer is a ‘stick - slip’ affair: and once stopped, it sticks, then when  released slips, giving a wide range of force values due to the hysteresis  built into the inter coil friction. For consistency, the force value used at  any point was the force needed to start the Stacer being pushed back  into the canister, after overcoming the ‘stiction’. This does not accurately  convey the sliding force, but is as close as can be statically measured.  Attempts to measure Stacer force dynamically were fruitless. Another  difficulty lay in measuring the drag from rollers and harness. Each tube  has a 1.1 -m stroke, and pulling steadily for that distance vertically while  monitoring force is a challenge. The weight of the tube assembly was  subtracted from the pull out force, giving the drag value.  Finally it was seen that proving force margin analytically was not  conclusive as the uncertainties in each measurement, when combined,  exceeded the margin requirements. A different path was chosen: show  that the boom deploys while using 113 of the available Stacer force. By definition, there is sufficient  margin. This is how the boom was verified. Figure 15. Thermal  Vacuum Chamber  Delta II Payload Planner’s Guide, The Boeing Company, Huntington Beach, CA 92647 8  59 {'source': 'AMS_2006.pdf', 'page': 73}"
"### Human: How many times did the EM deploy vertically during testing, and what was the purpose of these deployments?### Assistant:When testing deployables, the desire is to prove beyond question that the mechanism will deploy in  space, however, it must be tested here on the ground first. How many times? GEVS provides a minimum,  and each program defines how many additional operations. This brings up  wear margin: the design must show that it is sufficiently robust to survive  testing and flight without degradation. The EM served this purpose, getting  many deployments more than the FMs did. After identifying these values, a  test plan was developed, reviewed and implemented.  Testing large deployables in a simulated space environment is difficult, and  ensuring that the test actually verifies functionality is critical. Deploying the  boom horizontally was initially considered since it is easier to develop a 6-m-  long test rig that rests on the floor. After a few small efforts in this orientation,  it was realized that the only way to ensure that roller drag on the tubes was  representative of actual orbital deployment would be to deploy the boom  vertically. While several interim off load pulley systems were used, all the  verification deployments were performed on the thermal vacuum gantry.  To this end, a tall vacuum chamber was designed and built to allow the tests  to be performed (Figures 14 & 15). Inside the chamber ‘chimney’ a gantry  that allowed a counterbalance pulley system to provide G negation was  installed (Figure 16). The distance from the top of the boom to the pulley was  maximized to provide the least possible restorative (centering) force to the I  sections of the boom during deployment.  Figure 16. Gantry  Detail  Counterweiaht DescriDtion  To demonstrate the force (torque) margin, the masses to be used for the counterbalance force had to be  chosen to show that the Stacer would be energetic enough to deploy the boom. The mass of each of the  tubes (in flight configuration) was added to give neutral balance, plus the Stacer neutralization mass  (determined by bare Stacer vertical deployments to be 164 9). This mass (5214 g) was decreased by 213  of the Stacer minimum force (31 N * 0.67 = 20.7 N, converted to kg: 20.7 N / 9.98 kgm/s2 = 2.07 kg) and  subtracted from the counterweights. All 10 verification deployments were ‘force margin’ deployments and  were successful. After deployment, each boom was inspected for wear, with no signs of degradation of  rollers or Gr/E. The EM has been deployed -20 times, and is still in good condition.  Initially, the counterweights far exceed the G negation requirements for deployment as only one tube is  being deployed, while it is being pulled by the counter weight for 4 tubes. This is not invalid for our testing  needs, as the flyweight brake dissipates the extra force, keeping the velocity in correct range. The area of  interest is the very end of travel, where the Stacer force is lowest, and the full mass is being acted on. It  was this point that the gantry design was built on.  Maanetometer Alicmment Verification  After deployment, the alignment of the Mag needed to be measured to determine compliance with the  specification. A very accurate digital level (resolution: 0.01 degree a.02) was used for this activity. The  measurements were taken relative to the origin and coordinate system established in the Interface  Control Document (ICD)’ for the boom. The relative angles for the mounting feet for the X - 2 and X - Y  planes were recorded at the mounting foot. The angles for the same planes (translated out to the Mag  tray) were then measured. The difference was taken, yielding the deviation of the Mag tray from the  mounting plane. Figure 17 shows the measured differences of the deviations for the FMs for each plane.  See Appendix A for ICD Details 9  60 {'source': 'AMS_2006.pdf', 'page': 74}"
"### Human: What is the average deviation in the X-Y plane of the magnetometer alignment for Flight Model 1, according to Figure 17?### Assistant:I Impact Magnetometer Alignment  0.090  0.085  0.040 0.045 0.050 0.055 0.060 0.065 0.070 0.075  -1 Average Deviation in the X-Y Plane (degrees)  Circle Diameter = Std. Deviation  Figure 17. Magnetometer Alignment Error  For the final determination of pointing error, the root of the sum of the squares was calculated to give the  magnitude of the deviation (the directiqn is not of interest as long as the specification is met), Flight Model  1 measured deviation ""used up"" 0.10 of its allowable 0.88O margin, while FM2 used 0.11 . The Impact  Boom deploys accurately, repeatably, and exceeds requirements by a large margin.  Lessons Learned  Desian for test.  The final validation of any design is complete when the device performs as expected in its orbital  environment. GEVS SE gives guidelines, developed over many programs, as to what tests must be run,  and how much extra (or over-) testing is needed. For the boom to be tested, a vacuum chamber was  required to be built in a vertical orientation. The initial plan had been to use a horizontal track, using an  existing facility. This plan did not give sufficient demonstration of the booms ability to deploy in a straight  line as the track would have given alignment to the sections through out their travel. Additionally, the drag  induced by the lock pin rollers when the tubes deploy horizontally far exceeded the Stacer deploy force.  At the time the decision was made to go vertical, the Stacer should have been sized to allow a non-  counterweighted deployment. There would have been a mass hit, but being able to leave off the gantry  would have been a great savings, as the chamber would not have had to be as tall.  Safetv  One person was iniured durina initial installation of the stacer into its canister. As the final Dortion was  stowed, the safety ietaining pinwould not fit into the hole provided for it. When additional force was tried,  the operator lost control of the stacer and it deployed in an uncontrolled manner to -1 m, when it was  grabbed, cutting their finger deeply through two sets of latex gloves.  Some points arose from the review of the accident:  a) Check fit all safety related parts, sub-assemblies and fixtures. The size + tolerance of the hole  for the safety pin were too small after plating to allow the pin to be inserted. This would have been an  easy test, prior to assembly.  b) Have back up hardware and personnel: the operator was working virtually alone, and had no  back up person there to hold the stacer while the pin was being inserted. After the finger was cut, a  colleague from another part of the lab had to run to help control the stacer and finger damage. Another  61 {'source': 'AMS_2006.pdf', 'page': 75}"
"### Human: What safety measures were implemented to prevent damage to the Impact Boom during its deployment?### Assistant:point for spring loaded deployables: don’t rely on a single safety path. A lock down plate in front of the  stacer assembly would have caught the stacer, preventing damage. These recommendations have been  implemented in the procedures, and stowing fixture used now: the stowing procedure has a minimum  requirement of two persons to proceed.  Materials SDecifications  Often materials can be useful beyond supplier’s data, one only needs to verify what limits to the previous  testing exist. The search for a cryogenic temperature suitable glue was fruitless. In the end a test was  performed to establish suitability, after much effort. The question to add after “At what temperatures does  the product perform satisfactorily?” is “Has it been tested beyond that?” A fair number of days could have  been saved by realizing the manufacturers don’t have all the information.  Marain  While most programs have a margin requirement, it is good to carry some margin for additional mass  demands while designing. This is almost rhetorical. The proposal mass for the instruments at the end of  the boom was 1.2 kg. After deliberation, it was determined that the data would be significantly better if  pre-processing were done closer to the detector, so additional circuit boards were added out at the end,  raising the deployed end mass to 2.2 kg. This drove the size of the release mechanism from being ‘off the  shelf‘ to a new, custom version, requiring additional testing, with the usual learning curve associated with  new mechanisms. The entire structure needed ‘beefing up’ to accommodate the added loads.  Summary  The Impact Boom has completed qualification and acceptance verification for use in flight for the NASA  Stereo mission. This application has shown the use of a Stacer spring can be implemented for major  deployables as a motive force, not only as a sensor or sensor support. This represents a major cost  savings from traditional motor driven deployables, with their associated high cost electronics. Currently,  the launch is planned for 26 May 2006, with deployment of the boom occurring within a 3 - 30 day window  after launch.  Acknowledgements  Thanks go to Dr. Janet Luhmann, Principle Investigator for the Impact Suite, and the entire Impact Boom  Team at the University of California Berkeley, the Space Sciences Lab. Also to NASA Goddard Space  Flight Center for working with us to get this new hardware off the ground, and to all the reviewers whose  inputs and criticisms helped make the boom robust enough to survive qualification. Thanks also go to the  Stereo crew at APL for their continued support for the program.  62 {'source': 'AMS_2006.pdf', 'page': 76}"
"### Human: What is the location of the Impact Boom on the Stereo ""B"" Spacecraft, as specified in Figure 18 and referenced in Appendix A of the document?### Assistant:I .-  Figure 18. Stereo “B” Spacecraft: Impact Boom location  References  Space System Reliability and Safety Office, Code 302; “General Environmental Verification Specification  for STS and ELV, Rev A; (June 1996), National Aeronautics and Space Administration, Goddard Space  Flight Center, Greenbelt, MD 20771; June, 1996; 32.4.5.3 - 2.4.6.2  Appendix A  1543.1 [60.757 Overall inc Blanket at each end  Figure 19. Impact Boom ICD Detail: “B” Stowed  63 {'source': 'AMS_2006.pdf', 'page': 77}"
"### Human: What is the deployed position of the SWEA Deployed Position marker in meters, as indicated by the magentometer reading of -1267.4?### Assistant:4260 [I 67.72-1  (Deployment Travel)  r SWEA Deployed Position m ‘i  2 r ” r SWEA Stowed Position  +X ___)  - 1267.4 [49.907  (from Boom 4299.5 Origin [I to 69.272 Magnetometer) ------I - 5803.1 1228.477 (Overoll incl Blankets at each end) -1  DETAIL B  Boom Origin  (Bottom of Insulator Pad)  Figure 20. Impact Boom ICD Detail: “B” Deployed  64 {'source': 'AMS_2006.pdf', 'page': 78}"
"### Human: What is the purpose of the Latch mechanism in the CDLM, as described in the document?### Assistant:Heritage Adoption Lessons Learned: Cover Deployment and Latch Mechanism  James W incentsen.  Abstract  Within JPL, there is a technology thrust need to develop a larger Cover Deployment and Latch  Mechanism (CDLM) for future missions. The approach taken was to adopt and scale the CDLM design as  used on the Galaxy Evolution Explorer (GALEX) project. The three separate mechanisms that comprise  the CDLM will be discussed in this paper in addition to a focus on heritage adoption lessons learned and  specific examples. These lessons learned will be valuable to any project considering the use of heritage  designs.  Introduction  Because of future JPL mission needs for meter class space telescopes, there was an internal technology  demonstration to develop a complete mechanism set for single-time deployable cover to protect the  optics. Because the task was tracked for a fast technology development, the decision was made to fully  adopt the GALEX cover deployment design early in the project cycle to save cost and schedule. For  reference, the baseline design aperture opening diameter was 0.83 m (32.7 in) and the outer diameter of  the instrument was 1.1 m (43.3 in) in diameter, roughly twice the size of GALEX.  CDLM Overview  The three mechanisms that comprise the CDLM are the Latch, Hinge, and two Energy Absorbers. When  the Latch releases the cover, two push off springs and the Hinge mechanism rotate the cover  approximately 4.66 rad (267 deg) and impact the crushable honeycomb filled Energy Absorbers.  Deployment time is approximately 3.4 seconds. The Hinge mechanism is un-dampened. After  deployment, the cover remains against the canister. An overview of the instrument and placement of the  mechanisms are presented in Figure 1.  Latch Mechanism  The Latch mechanism, attached to the cover, employs a Starsys paraffin thermal actuator as the prime  mover. An interfacing slotted Latch Arm is affixed to the Cover Ring. The Latch Arm is spring loaded to  rotate away from the aperture opening upon release. A detail view of the latch area is shown in Figure 2,  the Latch Arm is shown in Figure 3, and a cross section of the Latch mechanism is shown in Figure 4.  The Locking Piston passes through and retains the Latch Arm. The Push Piston has a small-diameter tip  and passes through the Latch Arm slot. To deploy the cover, the actuator heater is energized, which  translates both the Push Piston and Locking Piston, forcing the Locking Piston clear of the Latch Arm.  Once the Locking Piston is clear of the Latch Arm, the Push Piston slips through the Latch Arm as it  rotates. The mechanism locks open by means of torsion spring loaded arms that snap into a grove on the  Locking Piston. Microswitches sense the motion of the arms and provide telemetry of the mechanism  state.  Power to the actuator is discontinued once one of three criteria are meet: Hall effect sensors mounted on  the Energy Absorbers register a deployed cover (discussed later), PRTs mounted on the actuator reach a  maximum temperature, or a time limit circuit is exceeded. Both the temperature and time are based on a  look up table derived from thermal vacuum Latch test data. After power is terminated and the actuator  * Jet Propulsion Laboratory, Pasadena, CA  Proceedings of the 38th Aerospace Mechanisms Symposium, Langley Research Center, May 17-19,2006.  65 {'source': 'AMS_2006.pdf', 'page': 79}"
"### Human: What mechanism is responsible for resetting the actuator in a Push Piston return spring, as mentioned in the context information?### Assistant:cools, a Push Piston return spring resets the actuator for further ground testing. Resetting of the cover  and Latch Arm is manually performed.  Latch Mechanism  Cover 7  Hinge  Mechanism 7  Energy  Absorber  Mechanisms  Energy  Absorber  Structure Figure 1. Cover Deployment  L and Latch Cover Ring  Magnets  Instrument  Canister  Microswitches  -  LatchArm - 1  Figure 2. Latch mechanism and Cover/Cover Ring  66 {'source': 'AMS_2006.pdf', 'page': 80}"
"### Human: What is the purpose of the torsion arms in the Latch Arm assembly, as shown in Figure 3?### Assistant:Figure 3. Latch Arm assembly, rotated  Push Piston and  actuator return spring ot in Locking Piston  Torsion Arms  Figure 4. Latch mechanism cross section  6- Hinge  Mechanism  Figure 5. Hinge Mechanism  67 {'source': 'AMS_2006.pdf', 'page': 81}"
"### Human: What is the purpose of the Hinge mechanism in the GALEX Energy Absorber?### Assistant:Hinge Mechanism  The Hinge mechanism works by means of two sets of nested compression springs acting against a lead  screw / carrier nut combination. A clever and complex design, the Hinge is fully single fault tolerant. For  some failure modes, such as the loss of a spring, the Hinge is two-fault tolerant. A graphic of the Hinge  mechanism is presented in Figure 5 and a cross section is shown in Figure 6.  In the cover-closed position, the compressed springs react against the Hinge mid section and the spring  support end, which are restrained from rotation by shoulder screws and rollers running in slots in the  housing. Attached to the spring support end is the lead screw, which is also restrained from rotation. The  lead screw passes though a carrier nut, which is constrained from translation, but allowed to rotate. As  the lead screw translates, the carrier nut rotates. The cover is attached to the clevis, which in turn is fitted  on roller bearings on the carrier nut. Pins fixed in the carrier nut and clevis engage during rotation, driving  the cover open. Each side of the hinge mechanism works independently of the other. If one side jams, the  Clevis rotates freely on the bearings. If a Clevis bearing freezes, the carrier nut and Clevis can rotate on  the inner bearing.  rn Should screw /  Rollers  Spring Support End  Housing L  Clevis Translating  Rotating  Fixed I Lead Screw  Figure 6. Hinge mechanism cross section  Energy Absorber Mechanism  The GALEX Energy Absorber used a compression spring in combination with a ratcheting plunger. Once  the plunger was pushed in, the ratchet held the plunger and compressed spring fixed. Hall effect sensors  imbedded in the striker and magnets mounted on the cover provided telemetry of a deployed cover. The  cover magnets also served to latch the cover open. One disadvantage of the GALEX design was that  after each cover deployment test, the Energy Absorber had to be disassembled to be reset. It was  requested by the project to simplify the GALEX Energy Absorber design with replaceable, crushable  honeycomb.  We desired to keep the new energy absorber function as similar as possible to the GALEX mechanism  due to the support structure design. The housing and Hall effect sensor striker were left relatively  unchanged, but the compression spring and ratcheting device were replaced with crushable honeycomb  core. The honeycomb core is bonded to a simple disposable aluminum plate which is attached to the  plunger. The push rod and Hall effect sensor striker are then attached to the plunger, creating the plunger  subassembly. As the magnets mounted on the Cover impact the striker during a deployment, the core is  crushed against the Energy Absorber End Cap. See Figure 7 for an exploded view of the crushable  honeycomb Energy Absorber mechanism and Figure 8 for a graphic of the Plunger subassembly.  The Energy Absorber honeycomb core is replaced after each cover deployment. The Push Rod is  segmented in two pieces with a left-hand thread so that the Hall effect sensor striker and front segment  68 {'source': 'AMS_2006.pdf', 'page': 82}"
"### Human: What is the crush load required to absorb the energy equivalent to the potential energy of the cover at its closed position?### Assistant:Push Rod can be removed from the plunger subassembly without stressing or twisting the sensor leads.  The mechanism is reset by first removing the striker and Front Cap and then the Plunger subassembly is  removed from the housing and the core replaced.  During prototype cover deployment tests, several types of honeycomb were experimented with, for the  goal of reducing cover bounce-back and obtaining the cleanest cover capture. Aluminum core, 3/8-5052-  1 .O, foil thickness 0.01 8 mm (0.0007 in) and crush strength of 172 KPa (25 psi), trimmed to three cells  was found to work well, however the 50.8-mm (2-in) long samples available came close to bottoming out.  New core, 88.9-mm (3.5-in) long, was selected to allow for enough travel with margin (see following  paragraph for more detail). Deployment with this core proved to be the best and the cover bounce back  was limited to approximately 0.04 rad (2.5 deg) (the cover does not rebound off the striker; instead the  Plunger is pulled back to its limit stop). Because of the core’s long length compared to its cross-sectional  area, the core appears to initially buckle uniformly, then folds between the mid-section to base. Crush  tests were conducted with the core and the force required to continue crushing gradually dropped from  the initial buckling. This result corresponded well for this application as the impact force falls off  considerably after some energy is absorbed. See Figure 9 and Figure 10 for before and after cover  deployment images of 50.8-mm (2-in) and 88.9-mm (3.5-in) core respectively.  The kinetic energy of the deployed cover and energy to be absorbed is equal to that of the Hinge  mechanism compression spring’s potential energy at the cover closed position. The average crush load  multiplied by the crushed length gives the energy absorbed, and thus, with some extra length for margin,  defined the length of required core.’ Figure 11 is Hexcel’s honeycomb crush strength curve, which  illustrates the peak load, average crush load, and the energy absorbed. Core used in the Energy  Absorbers were pre-crushed slightly to remove the peak force spike. Crush test data from the three-cell,  88.9-mm (3.5-in) long core is presented in Figure 12.  Housing  Endcap 7  Figure 7. Energy Absorber exploded view  69 {'source': 'AMS_2006.pdf', 'page': 83}"
"### Human: What is the length of the three-cell core in Figure 9, according to the context information provided?### Assistant:Figure 9. Three-cell core, 50.8-mm (2411) long  I  Figure 10. Three-cell core, 88.9-mm (3  70 {'source': 'AMS_2006.pdf', 'page': 84}"
"### Human: What is the crush load of Hexcel Honeycomb at a displacement of 1 inch (as shown in Figure 11)?### Assistant:Peak Lead  crush load  bsd eMnat.d by precnuhing  Bottomed * I  I  I  I  I  I  I  I 1 out 8 mn.  Area under curve I  energy absorbed  Deflection  Figure 11. Hexcel Honeycomb crush strength curve’  318-5052-,0007, 3 Cell, 3.5in Samples  Displacement vs Force  Sample  -1  -5  Displacement [in]  Figure 12. Force / Displacement graph for 3-cell, 88.9-mm (3.5-in) core  71 {'source': 'AMS_2006.pdf', 'page': 85}"
"### Human: What is the main reason why a heritage design adoption may impose unnecessary limitations, constraints, or failure modes on interacting mechanisms or systems?### Assistant:Heritage Adoption Lessons Learned  Sufficient review of a heritage design is necessary before adoption as heritage designs may impose  unnecessary limitations, constraints, or failure modes on interacting mechanisms or systems. Additionally,  a flight readiness review of a heritage design is necessary, as it can not be assumed that all necessary  information regarding the design, such as as-built drawings, assembly instructions, test procedures and  test data, are available. Of particular interest are the GALEX heritage adoption lessons learned during the  CDLM technology development.  GALEX cover deploy mechanism was obviously built to a redlined set of drawings; however, these  drawings were not available during this task. As GALEX was an aggressively “faster, better, cheaper”  mission, it is thought that resources were not available to complete the drawing package. Much of any  schedule savings in using the heritage design was exhausted going through more than 70 drawings to  look for, and correct, interference, material, and lubrication issues. Some issues were found only after  fabrication and assembly, necessitating the rework or re-fabrication of built parts. Accurate build and  assembly histories are required to adopt heritage designs.  Very limited test data, such as cover deployment time and cover impact force, was available from GALEX.  The lack of test data necessitated the building of a schedule intensive deployment test fixture and  mockup cover and duplicating cover deployment tests. Fortunately, a spare Latch mechanism and an  engineering model Hinge mechanism were available for testing. An image of the deployment fixture is  shown in Figure 13.  Mockup  cover  Early adoption of the Hinge mechanism restricted the design of the crushable honeycomb Energy  Absorbers. Not until the deployment test fixture and prototype energy absorbers were built and tests run  did it become apparent that the cover impact force was much lower than expected. If the honeycomb  crush strength was too high, the cover would bounce off the Energy Absorbers as the impact force  dropped off during impact. Conversely, the honeycomb had to be strong enough to resist the Plunger  subassembly inertial loads during vibration. It was found during testing that the impact force necessary to  sustain honeycomb crushing, and to effectively keep the cover from rebounding, required the Hinge  torque output to be increased (by the use of larger springs). By increasing the un-dampened Hinge torque  output, the amount of kinetic energy to absorb was also increased. This was further complicated by  adopting the same GALEX composite cover thickness, even though the cover grew substantially in size.  An alternative design, with a stiffer cover, could have placed the crushable honeycomb on brackets close  to the hinge axis where the impact forces would be higher, allowing for a more stable and compact  72 {'source': 'AMS_2006.pdf', 'page': 86}"
"### Human: What potential single point failure was introduced by the rotation of the Latch arm in the GALEX mission, and how could it have been avoided in an alternate latch mechanism design?### Assistant:honeycomb design and reducing the torque output and energy absorption. The overall design would have  been much simpler, reducing drawing, fabrication, and installation costs and schedule with the removal of  the Energy Absorber structure and simplification of the Hinge mechanism. Additionally, the crushable  honeycomb would have been replaceable without disassembly of the Energy Absorber mechanism.  GALEX placed the Latch mechanism on top of the cover due to mounting surface constraints and  contamination control. However, the proposed configuration could have allowed the Latch mechanism to  be mounted to the Cover Ring instead of on the cover. This would have removed the necessity of routing  Latch cabling over the Hinge, eliminating cable parasitic torque drag. Additionally, the simplified cabling  could have received cable stiffening micrometeorite shielding. With the Latch mechanism mass removed  from the cover, the crushable honeycomb could have been placed close to the hinge axis instead of along  the canister.  Cover deployment depended on Latch Arm rotation. The GALEX latch arm was positioned close to the  instrument aperture opening and was rotated away from the aperture to reduce stray light issues. The  rotation of the latch arm rotate introduced some potential single point failures. After the GALEX latch arm  was adopted, it was discovered than an earlier mission that originally designed the Latch mechanism  instead utilized a fixed latch arm with a slot in the path of cover deployment. Because the proposed  design did not face the same stray light issues as GALEX, a potential failure mode was unnecessarily  adopted and additional work was necessary to reduce risk.  The Latch mechanism is zero-fault tolerant in some cases as it depends on a single actuator (with  redundant heaters) and a single set of pistons to translate prior to cover deployment. An alternate latch  mechanism design using two Starsys pin pullers in a toggle type configuration could have been more  desirable and would have been fully single fault tolerant. Additionally, the mechanism would have been  less complex and would possibly have fewer parts and less expensive to fabricate.  Both the Latch and Hinge mechanisms were complex, with many tightly toleranced parts. These  mechanisms were expensive and schedule intensive to fabricate. Simpler alternate designs described  above would have potentially saved more schedule than building the heritage designs.  The Hinge mechanism is only capable of cover deployment with the Hinge axis aligned with gravity  (vertically). It would have been preferable to incorporate a hinge mechanism that was capable of  deploying the cover in any orientation as the instrument will not be positioned vertically during I&T and  ATLO, making an end-to-end test impossible. Instead, cover deployment tests will be conducted before  CDLM delivery to I&T. Once in I&T and ATLO, first motion tests will verify Latch mechanism functionality  and the cover will be sweep to ensure there are no obstructions.  The Energy Absorber honeycomb was finalized before flight-like cable was installed over the Hinge axis  during prototype cover deployment tests. It was expected that the cable could be wrapped in a way to  provide a positive torque to aid the cover rotation, but was initially found not to be possible. Further  prototype deployment tests with the cable showed that the honeycomb needed to be changed.  The crushable honeycomb Energy Absorber prototype effort cost significant schedule. In keeping the  Energy Absorber housing similar to GALEX (and thus limiting the redesign of the support structure), the  Energy Absorber still must be disassembled to be reset after a cover deployment test. This disassembly  is only moderately less time consuming than a GALEX Energy Absorber reset.  While the residual end-of-travel Hinge torque output and magnetic latch provide enough force to keep the {'source': 'AMS_2006.pdf', 'page': 87}"
"### Human: What is the main advantage of using a positive latch mechanism for the spacecraft cover, according to the given text?### Assistant:is only moderately less time consuming than a GALEX Energy Absorber reset.  While the residual end-of-travel Hinge torque output and magnetic latch provide enough force to keep the  cover captured during spacecraft maneuvers, a positive latch mechanism would have been more  preferable.  73 {'source': 'AMS_2006.pdf', 'page': 87}"
"### Human: What is the purpose of performing trade studies of designs similar in function to that of the proposed heritage design?### Assistant:Conclusion  Heritage designs offer projects an attractive means of reducing cost and schedule. However, without a  through review and investigation of the state of readiness, a heritage design may impose unnecessary  limitations and restrictions, in addition to cost and schedule impacts. It is recommended that trade studies  be completed of designs similar in function to that of the proposed heritage design prior to adoption. Only  then, potential shortcomings of the heritage design may become apparent.  Before adopting a heritage design:  Perform trade studies of designs similar in function to that of the heritage hardware. Look for best  solution.  Thoroughly review heritage drawing package for completeness  Verify the heritage design will meet project requirements  Review heritage test data and test plans and verify they meet current projects requirements. If  they do not, study impact and feasibility of revised testing.  Review heritage design for failure modes. It can not be assumed that all modes were found, or  that new modes will not be introduced.  Perform all prototype testing with as flight-like hardware configurations as possible  Acknowledgements  The research described in this paper was performed by the Jet Propulsion Laboratory, California Institute  of Technology, under contract with the National Aeronautics and Space Administration.  The author gratefully acknowledges the contributions of JPL team members: Mark Baker, Kevin Burke,  Keith English, Ted Iskenderian, Mike Johnson, Ellyn McCoy, Doug Packard, Don Sevilla, and Brad  Swenson.  References  1. “HexWeb Honeycomb Energy Absorption Systems, Design Data”, Hexcel Corporation, March 2005.  74 {'source': 'AMS_2006.pdf', 'page': 88}"
"### Human: What is the primary function of the QWKNUT mechanism, according to the abstract?### Assistant:Problems and Product Improvements in a Qualified, Flight Heritage Product  Chuck Lazansky’ and Scott Christiansen*  Abstract  This paper will discuss improvements to an existing, qualified, flight heritage launch restraint and release  mechanism. The changes made to the design are the result of customer feedback, test failures, and  optimization of performance margins. Specific test failures and their resolutions will be discussed. Impacts  to field units, process controls, product heritage, and qualification status will be summarized. Conclusions  and lessons learned will include aspects of what “qualified product” means and insights around what is  required to improve a product based on lessons learned through production and customer use.  Introduction  A Launch restrainthelease device must meet a demanding set of requirements to be reliable and robust.  Most importantly, the device must never release prior to command (during ground handling,  transportation, launch, etc.) and must always release when properly commanded. A combination of  redundancy and robust design features are typically used in pursuit of meeting these demanding  requirements. The QWKNUT has been designed with these goals in mind, and has been shown through  qualification and flight use to meet these requirements.  Figure 1. Gen 1 QWKNUT Mechanism  Product Description  The QWKNUT is a device which utilizes a segmented nut to maintain and release an axial preload. The  device accepts a standard, hardened, %-28 bolt to carry the nominal 13345-N (3000-lbf) load. Preload is  released when the four nut segments are opened by activating the mechanism. Release of the device is  initiated by a redundant pair of shape-memory alloy (SMA) wires within the QWKNUT, which are linked  directly to the latch. The QWKNUT requires an electrical pulse (3-5 Amps, -10-75 msec) similar to that  used for pyrotechnic release devices. The pulse causes resistive heating of the SMA wire above its  transition temperature, resulting in a strain of the wire and release of the latch.  Functionally, the mechanism can be separated into two parts: the preload-bearing part, and the latch-  release part. The preload-bearing part consists of four nut segments (Figure 3). Axial load from the bolt  exerts a radial load on these nut segments, which are retained radially by a set of rollers and a bearing  outer race. This outer race has slots which correspond to each roller, such that rotation of the race allows  the rollers to drop into these slots, allowing the nut segments to open and release the bolt. The outer race  is preloaded with redundant coil springs acting to rotate the race and release the device.  Starsys Research Corporation, Boulder, CO  Proceedings of the 3d’ Aerospace Mechanisms Symposium, Langley Research Center, May 17- 19,2006  75 {'source': 'AMS_2006.pdf', 'page': 89}"
"### Human: What is the purpose of the redundant compressive springs in the QWKNUT mechanism, according to the text?### Assistant:The outer race is prevented from turning by two latch pawls (called “toggles”) that engage the OD of the  race. The latch-release paH of the mechanism consists of a lever-arm which reacts the radial load from  the toggles, and retains them in the outer race (Figure 2). Nominally, the lever arm is held against a stop  feature, in the latched position, by a retention spring. The lever arm is acted on directly by the SMA wire,  which upon activation rotates the lever arm and frees the toggles, causing release of the mechanism.  REDUNDANT COMPRESSIMJ  SPRINGS =ATE TCRQUE  TO ROTATE OUTER RAE  ROTATES OUTER RACE  LATCHED RELEASED  Figure 2. Latch-Release Section Detail  OUTER RACE  ROLLERS DROP  INTO NOTCHES  NUT SEGMENTS ARE  SEGMENTED NUT FREE TI7 OPEN  LATCHED RELEASED  Figure 3. Preload-Bearing Section Detail  A key feature of the QWKNUT is that it is fully resettable. The reset procedure entails using a small tool to  rotate the outer race back towards its latched position. The reset process is the exact reverse of the  release sequence. Once the outer race is rotated, the rollers move radially inward closing the nut  segments. The toggles then re-engage the outer race, and the lever arm, which is spring loaded, moves  over both toggles to latch the mechanism. The straightforward reset process and multiple use capability is  an attractive feature in a release device, allowing a “test as you fly” approach to ground verification. For  additional ease of ground use, the original version of the QWKNUT contains cutoff switches, which  discontinue power to the SMA circuits once release occurs. The switches are not rated for Flight use, and  additional lead wires are included which bypass the cutoff switches for Flight.  76 {'source': 'AMS_2006.pdf', 'page': 90}"
"### Human: What were the four significant test failures that occurred during the product life of the QWKNUT, and how did these failures contribute to the evolution of the design?### Assistant:History of Product  The product was developed and fully qualified in 1999 with two test units. The initial production run of 4  units was delivered, and successfully flown, with the FalconSatl program. Following the initial run,  roughly six production runs were completed for various programs, totaling roughly 50 units. During this  time, we became aware of improvement opportunities as we worked to support customers with specific  flight applications. These were primarily usability issues related to damage during handling and operation,  and included:  0 Cutoff switch arm breakage  0 Reset difficulty  0  0  0 Strain relief of lead-wires SMA wires burned by excess pulse energy  Mechanism damage due to bolt over-engagement  Some of these issues were addressed with engineering changes during these first 6 production runs. In  2004, engineering proposed more extensive design changes and an internal research and development  effort was initiated. Production of the original design (Gen 1) was to continue until the new design(s) (Gen  2 and 3) and the delta-qualification program was completed. There were two versions of the new design  to be qualified to suit different programmatic needs. The Gen 2 design did not include an in-line current  shut-off feature. The Gen 3 design included a redundant, in-line current shut-off rated for Flight use as  well as ground operation. While the above designs were being developed, the final planned production lot  of the Gen 1 design was in process (circa May 2004). During this time, we encountered a failure in  vibration testing of several Gen 1 Engineering units at levels well below qualification.  Test Failures  Four significant test failures have occurred during the product life to date. These failures occurred  sequentially and were linked together as the process of investigation and resolution were carried out for  each. As each of the failures was investigated, our understanding of the mechanism increased. In some  cases, the failures required an exhaustive process of properly addressing any and all affected hardware  in the field. As painful as it was, the investigation process created invaluable opportunities to roll  improvements into the Gen 2 and 3 designs, requiting in an extremely robust final product.  Starsys’ FRB (Failure Review Board) process involves a systematic approach to failure resolution,  utilizing tools such as failure trees, fishbone diagrams, and detailed tracking of closure actions. Regular  reporting of progress and senior technical oversight continue until root cause and corrective action plans  are determined, and the investigation is closed. From June of 2004 until November of 2005, the  QWKNUT was under nearly continuous FRB activity for each of the failures described. These failures are  an example of how a qualified design can fail, and evolve, over the product life-cycle.  Engineering Unit Vibration Failures  In June of 2004, an Engineering Unit QWKNUT (Gen 1) undergoing development vibration testing  released the nominal 13345-N (3000-lbf) preload under the relatively low ievel of 12 Grms compared to  the 35 Grms qualification level. In reviewing the test setup, there were a number of potential causes  including cross-axis noise on the vibration table, and a non-standard, massive bolt interface. However,  two more EDU units at Starsys Research were vibration tested and also showed an intermittent release of  preload at levels below the 35 Grms qualification. We initiated an FRB action to investigate the anomaly.  With roughly 38 Gen 1 flight units delivered to customers and staged for flight use, there was concern  regarding the scope of the anomaly and status of units in the field.  The engineering team generated a complete failure tree. As potential causes were eliminated, the failure  tree pointed to the latch release portion of the mechanism. High-speed video of the failure event showed {'source': 'AMS_2006.pdf', 'page': 91}"
"### Human: Can you explain the mechanism of failure observed in the high-speed video footage, taking into account the movement of the lever arm during vibration and its sudden full movement towards release?### Assistant:tree pointed to the latch release portion of the mechanism. High-speed video of the failure event showed  the lever arm moving slightly during vibration with respect to the toggles, and then moving suddenly, and  fully, in the direction of release. A cause for this behavior could not be found initially. Physical and  dimensional inspection of the parts did not reveal any clear discrepancies. The engineering analysis  77 {'source': 'AMS_2006.pdf', 'page': 91}"
"### Human: What was the cause of the inconsistent FTR measurements observed during production, according to the investigation?### Assistant:showed relatively low inertial forces on the lever, which could not overcome the lever retention spring and  drive the lever in the direction of release.  Tests were developed to assess latch performance. The lever retention spring force was measured, and  also the net force required to release the lever arm from the toggles (force-to-release, or FTR). The  results of the testing were that some units showed low lever spring force (up to 50% below nominal), and  this was strongly correlated to vibration failure at 35 Grms. All units which failed in vibration showed  spring force below nominal values. The low spring force was traced to yielding of the spring caused by  installation damage, as well as overstress during manual releases. Though low spring retention force was  strongly correlated to vibration failure, it did not explain the physical cause of the lever motion under  vibration.  The true root cause was elusive. Driven in part by the need to resume production, the vibration  performance was initially addressed by increasing latch retention spring force. This was accomplished  with the addition of a second retention spring. The dual-spring design provided strong margin for load-  holding under vibration, and was not vulnerable to handling damage. Validation testing of the proposed  change supported the assertion that release margins and vibration margins were better balanced with the  second spring, and that cycle life was not impacted. A formal Delta-Qualification test program of the new  configuration was planned in parallel with production. The latch performance tests were also added to the  production process as additional screens. However, when production resumed, the in-process testing  showed inconsistent results for FTR screening. There was variability within a single mechanism that could  not be explained. Production was stopped, and the investigation resumed with the focus on the dry-film  lubrication layer at the latch interface. The lubrication was to be removed from the lever of a production  unit to determine the source of the inconsistent FTR measurements.  Drive springs (2X) provide bias  torque for Flywheel  ! ! !  ! !  ! !  ! ! !  i  Section through latch area of QWKNUT “%._. ,,...- During launch Lever  :’ Spring preloads Lever On orbit SMA wire  pulls Lever Arm off  Lever Arm creates  &/;&n force n Toggle of is Lever  perpe dicular to the Lever  Toggle Pivot  reacts torque  to the Body t  Figure 4. Latch-Toggle Section Detail  Failure Mechanism  The physical cause of the failure was discovered when lubrication was removed from the critical latch  surface on the lever. The shape of the surface did not meet the intent of the design. The curved surfaces  contained “flat” areas approximately 0.76 mm (0.030 in) in width, at each end (Figure 5). These flats were  not easily seen with the dry-film lubrication present. The presence of the flats immediately explained the  78 {'source': 'AMS_2006.pdf', 'page': 92}"
"### Human: What was the root cause of the vibration anomaly in the device, according to the investigation?### Assistant:intermittent vibration behavior and the variability in release force measurements. If the toggle is on the  “flat” area of the lever, a tangential driving force results and leads to release of the device. If a toggle is  located on the curved surface of the lever, no driving force is present and no release occurs (Figure 4).  The width of the flats corresponds closely to the nominal location of toggle contact when the device is  latched. In this case, small variations in toggle position on the critical surface caused the variability in  latch performance. The physical cause was confirmed when the discrepant lever arm in the failed unit  was replaced with a conforming part from another lot, and the unit passed 3 repetitions of vibration  exposure (35 Grms, 3 minutes).  Root Cause  The root cause of the failure was found to be inadequate form control of the critical latch surface. Though  the latch surface was depicted on the drawing with a continuous curvature, the dimensional controls were  not adequate to ensure this. A total of 5 production lots of lever arms were received. The first 4 production  lots of parts met the design intent. The 51h lot was procured from a new supplier who used a different  machining process, which did not produce the intended surface at the critical latch point. The inspection  process did not detect the different geometry since the drawing did not specifically control the feature  shape. The inadequate drawing controls, combined with variation in manufacturing process (change of  vendor, machine, or even machinist) created the opportunity for the discrepancy. Fortunately, the  discrepant lot of levers was confined to the current production run, and no field units were affected.  Continuous curvature between 45” surfaces \  Burnish mal  indicating Tc  contact poin  Lever from previous GOOD lot - MoS,  coating intact; continuous curvature of  radius is evident “Flats  Burnisl  indicat  contac  Lever from discrepant lot - MoS,  stripped down to anodize  Figure 5. Lever Surface Discrepancy  Corrective Actions  The corrective actions for the vibration anomaly included adding proper engineering controls to the critical  lever surface to ensure design intent is met. The curved surface is now tightly controlled on the drawing  using runout and true position callouts. 100% inspection of this surface is required per an inspection  process defined on the drawing, followed by engineering review before parts are certified and released.  Process controls were also added at the assembly level to verify proper latch performance. These include  measurement of retention spring force and force required to release the latch and defined pass/fail criteria  (Figure 6). The latch release point is also screened against pass/fail criteria. The dual-spring design  change was implemented (and delta-qualification tested) to improve margins for vibration performance,  and prevent handling damage. Finally, acceptance level vibration testing under nominal preload is  performed on all units.  Once the lever surface was restored to the intended geometry, the device has proven to be extremely  robust with respect to vibration exposure. Delta qualification tests were performed at greater than 42  Grms with the lever arms intentionally shimmed 0.508 mm (0.020 in.) in the direction of release (to  approximately 70% of the release point) and all units passed the test.  79 {'source': 'AMS_2006.pdf', 'page': 93}"
"### Human: What is the likely cause of the vibration anomaly in the QWKNUT Latch In-Process Test Fixture, according to the text?### Assistant:Figure 6. QWKNUT Latch In-Process Test Fixture  Discussion and Lessons Learned  The vibration anomaly is an example of how insufficient controls on critical features will, in time, cause a  failure. The first four production lots were not discrepant, and there was no indication of a shortcoming in  the design -- the problem was simply dormant. Design intent must be clearly communicated in  engineering drawings. This requires a thorough understanding of which features are critical to  performance. If adequate engineering controls are not in place, the discrepancy will eventually emerge  with enough repetition (or opportunity). Parts from different lots may both meet the same drawing  requirements but perform differently in a mechanism. Proper identification and inspection of critical  features is essential, and in-process screening at the assembly level provides redundancy in detection of  variation.  Failures are sometimes created as a result of two or more prevailing conditions. Care must be taken to  ensure root cause and all contributing causes are identified before the investigation process is concluded.  Low spring force was initially thought to be the cause because with a certain minimum spring force, even  the discrepant lever geometry could not cause a vibration failure. Increasing spring force did resolve  vibration release, but did not create consistent latch performance. Proper performance under vibration  loads is now ensured by a combination of geometric drawing controls, a rigorous inspection plan,  assembly level screening tests, and finally ATP. These added controls, and a deeper understanding of  the design, were some benefits of the investigation process.  Failure to Release during ATP  Following the resolution of the vibration anomaly, production of the QWKNUT resumed with the vibration  corrective actions fully implemented. The ""Gen 1 A QWKNUT configuration now contained dual-springs  for lever retention, and flight units were assembled in parallel with Delta-qualification activity. During  acceptance testing, a flight QWKNUT failed to release preload during an ambient functional test. The unit  operated nominally using the primary SMA circuit, but did not release when the secondary SMA wire  circuit was activated. The testing was stopped and the FRB team was again convened.  Physical observation of the failed unit helped guide us quickly to the physical cause. The lever on the  activated side of the mechanism had disengaged properly from its toggle, but on the opposite side, the  lever was still engaged with its toggle (Figure 7). This prevented the outer race from rotating and  releasing the mechanism. The lever arm appeared to be tilted at an unusual angle, and one end was in  contact with the underside of the cover, which is directly above the lever. The mechanism was in a  partially released condition. In this condition, the mechanism could be released with further travel of the  lever arm. Minimal added force was required, on either side of the lever, to pull the lever free from the  engaged toggle. When the partially released mechanism was activated with the opposite (redundant)  SMA wire circuit, it released full preload. This distinguishes the condition from a ""jam"" in which increased  force is encountered with mechanism movement. What we observed was a stable lever position at a point  in the travel which should have resulted in complete mechanism release.  80 {'source': 'AMS_2006.pdf', 'page': 94}"
"### Human: What is the cause of the failure mechanism described in Figure 7, according to the text?### Assistant:;MA Wire Fc  EL== 3  Figure 7. QWKNUT in Partially Released State  Failure Mechanism  The tilted lever arm in the failed unit had been stroked about 1.78 mm (0.070 in) from its stop. Previous  testing had shown that only 1.02 mm (0.040 in) of lever travel will release the mechanism. It was  determined that the addition of the second spring changed the force balance on the lever arm during  actuation. The result was the observed tilting motion of the lever during release, which extended the  nominal latch release point from 1.02 mm (0.040 in) to over 1.78 mm (0.070 in). Figure 8 shows how  forces acting on the lever, including the added spring force, contribute to the anomalous tilting motion.  The extended release point is only part of the failure mechanism.  The SMA wire stroke expected from analysis and early development tests was 2.29 mm (0.090 in). It was  unclear why the unit had not released even with the extended release point. This led us to measure  actual SMA wire strokes for the entire lot using high-speed video. Measurement of the 7 units in the failed  lot showed a mean stroke of 1.85 mm (0.073 in) and a variation of k0.305 mm (0.01 2 in), well below the  expected 2.29 mm (0.090 in). The unit which failed had the lowest recorded stroke in the lot (1.55 mm, or  0.061 in). If the failed unit had the expected SMA stroke it would have released, but the stroke was not  available. The extended release point in combination with reduced SMA wire stroke produced the failure.  Both are required for a partial release to occur. This failure is essentially one of low performance margins.  The stroke margin on the failed unit was less than zero.  In addition, the partial release behavior was only displayed when the mechanism was under preload.  However, the process screenings were all performed with no applied preload, and the anomaly was not  detected until test. All units in the failed lot displayed the partial release behavior during screening once  preload was applied. In addition, the forces required to release the latch increased significantly with the  presence of preload. A majority of units in the lot showed unacceptably low actuation force margin.  The described anomaly was present only on units with the dual return springs. Single spring units were  not affected. Fortunately, though we did not have the margins we expected in the dual-spring design,  acceptable margins were confirmed in the single spring units. This limited the scope of affected units in  the field.  81 {'source': 'AMS_2006.pdf', 'page': 95}"
"### Human: What was identified as a root cause of the release anomaly in the dual spring design of the QWKNUT?### Assistant:-1 SMA Wire FO - I- -  Figure 8. Forces Acting on Lever Arm  Root Cause  The change to the dual spring design created an anomalous lever motion during release. The tilting  motion that was introduced to the lever required significantly more stroke from the SMA wire than the  single spring design. The motion of the lever arm was not properly constrained in the design, but until the  second spring was added, there was no impact to performance. The tolerances in the latch assembly  allowed for the anomalous tilting motion, and this was a root cause of the failure.  The other root cause was variability in the manufacturing process. SMA wire stroke was found to be  lower, and more variable, than expected by previous test results and analysis. Wire stroke measurement  was not part of the manufacturing process for the QWKNUT. There was inherent process variability that  had not been characterized. The wire setting process had also drifted with ongoing production lots and  different operators since its development. Without in-process measurement of stroke, the drift was not  detected.  Finally, a contributing cause was that the in-process latch screening tests were not performed under load.  An assumption was made that load bearing part of the mechanism performed independently of the latch  portion, which was not correct. Preloading the mechanism for the screening tests takes more time, and  adds cycles, but it is necessary. In designing screening tests, care must be taken to ensure the tests  accurately represent true operating conditions.  Corrective Actions  Corrective actions for the Release Anomaly were:  1.  2.  3. A stabilizing washer was added to control lever arm release motion and eliminate ""tilt"" and the  partially released state.  The SMA wire setting process was improved to increase the mean stroke and reduce variation  (was reduced by 30%). Actual SMA wire stroke was to be measured on each unit.  New process controls were added to measure latch release point and SMA wire stroke, and  determine actual performance margins present in each flight unit. Existing process controls were  now to be performed with the mechanism under nominal preload, which better reflects true  operating conditions.  82 {'source': 'AMS_2006.pdf', 'page': 96}"
"### Human: What was the primary challenge of a release device according to the text?### Assistant:4.  5.  6. A process for adjusting stroke margin was developed, consisting of a shimming operation at the  stop feature for the lever. This allowed us to tune the latch release point to match SMA wire  stroke capability, so that acceptable stroke margin is achieved.  Lever retention spring force was tuned to a range, and controlled with a maximum. This ensured  adequate release force margins. Force margin is verified on each of unit as part of the production  process.  All field units with the dual lever return spring were recalled under a Technical Bulletin for  measurement of stroke and force margins. With some schedule impacts, all units which had been  built with dual springs were rebuilt and re-tested with the new stabilizing washer, process’  controls, and delivered with full margins for Flight use.  Discussion and Lessons Learned  The primary challenge of a release device is to maintain high performance margins for both pre-load  retention and operation. It is interesting that anomalies have been encountered in both these areas,  sequentially. The vibration anomaly focused attention on the preload retention aspect, and the analysis  showed that margins would be balanced with the dual spring. With the focus on improving vibration  performance, we added spring force and disturbed the release performance. Sometimes in an effort to  resolve one failure mode, we accidentally introduce another one. This is especially true when failure  investigations are performed under tremendous schedule pressure. In our case, the dual spring solution  was tested with greater than 400 life cycles prior to implementing the change, but the test unit had  positive stroke margin. Though the test unit passed, a variation in manufacturing process created a failure  of the revised design in a subsequent lot. The release anomaly is an example of unintended  consequences of a simple change, and insufficient controls on critical processes. The change to the  spring was necessary, and carefully implemented. But even with careful steps taken to validate the  change, there was an impact to mechanism performance.  When evaluating a qualified product, the variability of the production environment must be considered. It  is impractical and expensive to attempt qualification testing with all design tolerances at worst case.  However, worst case production results should be enveloped by the qualification program. In general, this  means that a production process control plan should be developed prior to qualification activities. A  qualified product must be manufactured with stable, controlled processes or future production lots will not  be similar to the original product.  Another lesson from this anomaly is that analysis of performance margins should always be verified by  actual measurement. Force and stroke margins on the QWKNUT were measured during development,  but not as part of the ongoing production process. The release anomaly led us to verify actual  performance margins, and ultimately these margin measurements were made a permanent part of the  production process. Stability of the performance margins over the product’s cycle life has also been  verified. The overall result is more robust hardware.  The investigation process was difficult and expensive, and impacted deliveries to several programs. It is  painful to have to recall flight hardware which has been delivered. However, the process has been  instrumental to the advancement of the design. We would not have discovered low SMA stroke and  variable performance margins had we not made the change to the dual spring. All 3 design generations  have been enhanced by the investigation of this anomaly.  Once the corrective actions were implemented, we proceeded to qualification of the “Gen 1 B design with  optimized design and manufacturing process.  Life-Test Pin Failure  Delta-Qualification of the Gen 1B design proceeded, using two test units which were rebuilt from the {'source': 'AMS_2006.pdf', 'page': 97}"
"### Human: What is the cause of the failure encountered during the extended life test of the Gen 1B design, according to the text?### Assistant:optimized design and manufacturing process.  Life-Test Pin Failure  Delta-Qualification of the Gen 1B design proceeded, using two test units which were rebuilt from the  failed lot in the release anomaly. The units passed all tests and proceeded into an extended life test, at  the end of which another failure was encountered. In each unit, one of the two spring posts (“flywheel  83 {'source': 'AMS_2006.pdf', 'page': 97}"
"### Human: What is the aerospace mechanisms symposia?### Assistant: An annual meeting of space mechanism experts. {'source': 'DM', 'page': 0}"