MIPAS InOrbit Operations and Failure Recoveries

1
MIPAS In-Orbit Operations and Failure Recoveries

• J. Frerick
• Rhea Systems S.A. c/o ESA-ESTEC

2
Content

• Introduction
• The MIPAS experiment
• Interferometer layout
• Operational experience
• Phase I Launch April 2004
• Phase II Anomaly investigation
• Interferometer Speed control
• Classification of Velocity Errors
• Investigation Objectives
• Special tests
• Investigation summary
• Phase III Since January 2005
• Summary
• The Team

3
Introduction (1)The MIPAS experiment

• Michelson Interferometer for passive
  atmospheric sounding
• On-board ENVISAT (Launch March 2002)
• Limb sounding IR instrument

4
Introduction (2)The MIPAS experiment

• Targeting atmospheric species relevant for
• Ozone chemistry
• Green house effect

5
Introduction (3)The interferometer Layout

• Double slide interferometer
• 100 mm rails
• Nominal speed 25 mm/s

6
Operational Experience (1)Phase I Launch until
April 2004

• Very reliable, autonomously working instrument
• Easy going mechanics (interferometer, Stirling
  Cooler)
• High data quality (very) satisfied users
• Only a few un-planned unavailabilities
• However Since April 2003 there were signs of a
  degrading interferometer subsystem.
• Warnings from on-board monitoring (so called
  differential speed monitoring errors) with no
  operational consequences.
• Later Increasing number of so called velocity
  errors, which forced the instrument into
  Heater/Refuse mode, requiring manual recovery
  from ground.
• Substantial instrument unavailabilities.
• Specifically alarming in some anomaly cases the
  recovery showed a risk for a permanent failure

7
Operational Experience (2)Phase II The Anomaly
investigation

• Interferometer speed control

Reference Laser Fringe count (1 msec)
Absolute position sensor count (10.65 msec)
Heater refuse If v1 and/or v2 gt30 mm/s
Heater refuse If gt 15
Check Vopt
Check V1 , V2
If lt15
Check V1-V2D
Increase/Decrease currents of both, Slide 1 and
Slide 2
Increase/Decrease currents of Slide 1 and Slide
2 by D/2
If gt4, differential speed monitoring error
8
Operational Experience (3)Phase II The Anomaly
investigation

• Classification of velocity errors
• Turnaround (principally known)
• In sweep (unexpected)
• With critical initialization
• During a critical initialization, slide 2 has to
  pass the former launch lock position. An
  increasing number of attempts was necessary to
  move the slide across this point. This was
  regarded a risk of permanent failure.

9
Operational Experience (4)Phase II The Anomaly
investigation

• Investigation objectives
• To eliminate the source of the problem
• Reference laser
• Absolute position sensor
• Rails
• Bearings
• Temperature effects
• Any combination of all
• To eliminate/reduce risk of permanent failure
• How to avoid getting trapped by the launch lock?
• To reduce operational effort
• How to minimize manual operations/instrument
  recoveries?
• To minimize outage times

10
Operational Experience (5) Phase II The Anomaly
investigation

• Special tests
• Static Mode tests
• Moving the slides slowly, but individually and
  monitor drive force
• No conclusive results
• Reduced Resolution tests
• Better performance could have been a hint for
  rail depended problem at turnaround points (i.e.
  the end of the rails).
• No conclusive results regarding interferometer
  performance
• Positive effect on avoiding critical
  initializations.
• Tests on instrument redundant side
• To elude error sources like the reference laser,
  the absolute position sensor
• Similar performance as on nominal side ? switch
  back to nominal
• Tests in back up mode
• Single slide operations with one slide fixed in
  the middle.
• Does one slide have a better performance than the
  other?
• Not conclusive.

11
Operational Experience (6)Phase II The Anomaly
investigation

• Investigation Summary (1)
• Explanation of error
• Overall degradation of non-redundant parts
  performance (most likely bearings) in combination
  with inflexible control loop layout (it was
  impossible to change control loop thresholds by
  software).
• Risk of critical initializations could be
  minimized by optimization of spectral resolution.
• 41 of maximum (full) resolution turned out to be
  the optimum.

12
Operational Experience (7)Phase II The Anomaly
investigation

• Investigation Summary (2)
• Minimization of outage times/operational effort
• By performing a planned interferometer
  initialization every orbit.
• By reducing the duty cycle (as the performance
  indicators showed relaxation after longer periods
  of outage).

13
Operational Experience (8)Phase III Since
January 2005

• The concept is working
• Operations is stabilized (no more manual
  intervention due to velocity errors)
• The total outage due to velocity error is reduced
  from former 3 orbits (best case) to lt 0.5 orbit,
  nominally.
• Critical initializations occurred only once in
  almost 18 month.
• Reduction of spectral resolution only had limited
  scientific impact due to the originally
  implemented over-design.
• The number of interferometer errors is still high
  and even showing seasonal maxima
• Evidently, there is more than one error
  mechanism.
• However, the main conclusion remains the control
  system is too inflexible to cope with a changing
  environment.

14
Summary

• An interferometer control problem interrupted the
  so far very successful MIPAS experiment on-board
  ENVISAT in April 2004.
• Intense failure investigation was performed until
  end 2004, taking advantage of various and
  numerous special on-board test possibilities, the
  most important ones being
• Single slide/ backup mode tests
• Tests on redundant electronics
• Variation and finally reduction of stroke length
  (i.e. spectral resolution)
• Analysis concluded that a degradation in non
  redundant parts in combination with the
  inflexible design of the interferometer control
  caused the observed problems.
• A new operations scheme could be implemented
  successfully
• Minimizing risk of permanent failure
• Minimizing the operational effort for recoveries
• Minimizing the outage and thus leading again to a
  stable mission
• The new concept took advantage of a very flexible
• Instrument design
• The mandatory reduction of spectral resolution
  could be compensated without significant
  reduction of scientific data quality
• Mission planning design
• With reasonable effort the new, campaign
  oriented way replaced the former continuous
  operations.

15
The Team
ESA-ESTEC/PLSO Investigation/coordination
supported by Rhea Systems S.A.
ESA-ESOC Operations supported by Vega
Group SciSys GmbH
Industry/EADS-Astrium Performance
Evaluation/Investigation Supported by Officine
Galileo AEA Technology
ESA-ESRIN Data quality analysis Mission
planning supported by Rhea Systems S.A. SERCO
Spa