ACS by ewghwehws


									Attitude Control System/Guider
          WBS 2.3.5

               M. Lampton
       Space Sciences Laboratory
     University of California Berkeley
                 July 2002
     ACS: Overview

•   Two Attitude Control Systems
•   Science Driven Requirements
•   Recommendation from the last DoE review
•   Three Studies Completed
•   ACS Schematic
•   R&D Issues and Goals
•   R&D Schedule & Plan
•   R&D Manpower
•   Summary

Two Attitude Control Systems
• Booster ACS
   —manages attitude during climbout & staging
   —climb to initial circular LEO parking orbit
   —responsible for ascent to 2.5 Re elliptical transfer orbit
   —responsible for final ascent to 25 Re HEO apogee
• Payload ACS
   —performs safe hold maneuver w/ sun sensors
      • keeps solar panels toward sun
   —maintains limits for onboard angular momentum
      • dumps using hydrazine jets
   —performs all science target field acquisition
      • slews, settlings, stability while on target
      • must therefore be managed as science system resource
    Science Driven Requirements
•    Dynamic: Large Maneuvers
      — 6 degrees/minute slew rate
      — 30 minutes settling time after 180 deg maneuver
      — acquire Earth RF link each orbit; choose NEP, SEP, SNe, cal, ...
      — after we’ve performed a complete on-orbit calibration, must place
         attitude on target without need for science focal-plane data
           • 1 arc second absolute in pitch and yaw, 3 sigma
           • 2 arc second absolute in roll, 3 sigma
•    Static: No drift during 300 sec exposures
      — 0.01 arcsec RMS stability in pitch and yaw [specification]
      — 1 arcsec RMS stability in roll (goal: 0.5 arcsec)
•    Step: change position within CCD for science dither
      — step size 1 arcsec, accurate to 0.01 arcsec = 0.1 pixel (TBC)
•    Sun avoidance safehold mode uses redundant sun sensors
      — handles startup/reboot “empty statevector” situations
      — handles conflicting gyro/statevector situation
Recommendation from the last DoE review

       Section 8, OTA; Section 9, Spacecraft:
        We recommend that the OTA system design permit a
        fast steering mirror to be included in the ACS
        trade space, and that this trade be fully explored
        in terms of risk and net mission cost."

   •    Three studies have now been completed
   •    All show that pointing jitter will be few milli-arcsec RMS
   •    We concur that this issue is extremely important
   •    We will enlarge the ACS study to include dynamic modelling
   •    We are now including ACS performance as a budget item in our systems
        engineering management of PSF
   •    A trade study “Fast Steering Mirror” is posted at

     Three Studies Completed
•   Secroun et al “High-accuracy small FOV star guider” Exp.Astron. v.12 2002
     — density of guide stars is sufficient for NEP and/or SEP
     — sensor noise <0.007 arcsec RMS single frame at 30frame/sec
     — sensor noise <0.002 arcsec RMS < 1 Hz
•   NASA/GSFC Integrated Mission Design Center study June 2001
     — existing sensors: gyro packages “as flown” + planned focal plane guider
     — “rigid” spacecraft: all flexural modes > 5Hz
     — disturbance terms and spectrum estimates from
          •   wheels without vibration isolators
          •   payload flexure
          •   antennas
          •   solar wind pressure torque
          •   propellant slosh decay
     — did not study improvements using Kalman filter or vibration isolators
     — first estimate (27 June 2001) gave 0.057 arcsec RMS, no Kalman filter
     — revised to ~0.005 arcsec RMS depending on Kalman filter & vibration isolator
•   S/C manufacturer measured data on orbit, ACS comparable to SNAP
     — observed 0.005 arcsec RMS using existing wheels, specialized sensor package
     — predicts “about the same” for SNAP configuration                             6
               ACS Schematic
Disturbances                        Dynamics

                          Coarse star trackers
                          Coarse sun sensors
  STATE VECTOR                                     Controller
    s/c attitude                                                   Jets
                           Coarse/fine gyros
                           GigaCam guider                        Torquers

    Environment:           Cassegrain guider
    orbit, Sun,
    Earth, stars....            noise            commands data

                  What is the disturbance torque spectrum?
                  What are the various sensor noise spectra?
                  What is the closed-loop response?
    Disturbance Torques
•   Momentum wheels
     — four wheels
     — run at various RPMs
     — strong sinewave unbalance disturbances
     — broadband torque noise from bearings
     — broadband torque noise from DAC drive
     — will be somewhat reduced by vibration attenuator mounts
•   Shutters
     — impulse unbalance torques; predictable, broadband
•   Solar radiation torque
     — essentially continuous DC torque
•   Thruster unloading of angular momentum
     — impulse, predictable, broadband
•   Maneuver settling effects
     — structural
     — propellant slosh
     — antenna motion aftereffects

             Attitude Sensors
•   Gyroscopes
     — most common choice for relative attitude control
     — many kinds: mech, fiber optic, hemispherical resonator, ...
     — essential in executing maneuvers
     — work in any attitude, including Earth-blocked, solar, lunar....
     — require updating to track drifts and offsets
     — excellent reliability where low accuracy is OK
     — poor reliability where high accuracy (mas) is required
         • HST experience
         • problems are lubricants, wear, ... ?
•   Coarse sun sensors
     — essential for startup and safety shutdown modes
•   Coarse star sensors
     — essential for attitude location after maneuvers
•   Fine guider within instrument focal plane
     — within Gigacam: best possible tracking accuracy (shutter open)
     — at cassegrain focus: good accuracy, independent of shutter

Guider CCDs
located within

 Guider CCDs
 located within
 rear metering
 structure, on
 optical axis

    Fine Guider Worksheet

Video guider CCD frame rate                             30   frames/sec
Guider pixel size                                      100   milli arcsec
Guider read noise                                       30   e RMS
Integral QE * dLambda                                  150   nm
Telescope aperture                                       2   meters
Telescope efficiency                                   0.7
Guider pixels per chip                 1024x1024             pixels
Number of guider chips                                  4
Sky area for guide stars                200x200              arcseconds

RESULTS FOR TWO CASES...               Typical field                    Poor field
brightest star, V mag                               13                              16
Percentile among all fields analyzed              50%                             95%
photon flux/m2 sec nm                            631.0                            39.8
photoelectrons/frame                            6934.2                           437.5
RMS jitter, in pixels, one frame                 0.007                           0.073
White noise bandwidth, Hz                       15.000                          15.000
RMS jitter, in pixels, per root Hz               0.002                           0.019
1-D RMS jitter, 1Hz BW, milli arcsec             0.191                           1.875

Sensor Noise Spectra

       R&D Issues and Goals
•   Issue: static ACS performance affects science
     — jitter hurts signal-to-noise ratio, hence exposure times
     — jitter causes varying PSF and therefore hurts weak lensing science
     — varying PSF complicates data pileline processing: subtraction etc
     — Budget specifies 10 milliarcseconds RMS 1-D; goal is 5 mas
•   Issue: dynamic maneuver settling times could become an issue
•   R&D Phase: what ACS trades need to be explored?
     — hardware: wheel vibration mitigation? DAC resolution?
     — Do we need precision gyros?
     — software: Kalman filter? loop bandwidth and sample rates?
•   R&D Phase: continue gathering ACS performance data from vendors
     — SKIRU-V gyros for Gravity Probe B; IMDC; others
•   R&D Phase: create and validate a model ACS at Berkeley
     — ongoing structural FEM models for SNAP
     — perform guider validation work
     — invite outside experts to review our findings
       R&D Schedule & Plan
•   Schedule driver: Science Team SNR & PSF trades
     — Now! ... and continuing throughout project lifetime
•   Schedule driver: S/C component selection, cost, reliability
     — in time for Phase B, preliminary design
•   Payload flexural requirements are probably NOT a driver
     — Fminimum > 5Hz satisfactory from IMDC; presently > 10Hz
•   FY 2003: concentrate on static target jitter performance
     — create dynamic model including ACS feedback, weightings, ...
     — implement Kalman filter, explore its optimization
     — gather and integrate wheel dynamic noise disturbance spectra
     — include known FEM structural resonances
     — evaluate simplified video CCD for centroid error distribution
     — explore need for precision gyros benefit to static jitter
•   FY 2004: concentrate on dynamic performance
     — upgrade model to include telescope flexural behavior
     — install simplified slosh model
     — explore large angle settling behavior
     — estimate maneuver times required for model mission profile
     — evaluate sun-avoidance safe hold mode                           14
  R&D Phase Manpower

• Existing staff
   —Pankow + Besuner for FEM
   —Pankow + Lampton for sensor noise spectra
   —Pankow + student for S/C disturbance spectra
• New hires (fractional time)
   —S. Harris for CCD guider validation
   —R.Abiad for control loop modelling
• Subcontract labor
   —ACS consultant (2mmo)

• ACS is manageable
   —experienced engineers have examined our requirements
   —estimates of static & dynamic performance appear
   —milli arc second attitude control has been accomplished
   —“rigid spacecraft” approach makes this possible
• ACS behavior will receive systems engineering treatment
   —cost benefit trades of the budgeted items
   —requires a rather detailed dynamic model
   —will be done in-house
• Will seek subcontractor support to validate our efforts
   —performance of existing controllers, Kalman filter, ....
   —precision gyros, lifetime, benefits, costs.


To top