Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://xmm.vilspa.esa.es/calibration/documentation/epic_cal_meetings/200904/kirsch_09-03-TTDs.pdf Äàòà èçìåíåíèÿ: Tue May 13 17:47:00 2014 Äàòà èíäåêñèðîâàíèÿ: Wed Apr 13 13:35:00 2016 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: star trail

ESOC

XMM recoveries in 2008 and the new operations strategy re. antennae switching
XMM-Newton Deputy Spacecraft Operations Manager
with Inputs from the MOC Flight Control Team European Space Agency (ESA) European Space Operations Centre (ESOC)

Dr. Marcus G. F. Kirsch

March 2009
M.Kirsch, XMM-Newton European Space Operations Center Page 1


menu

ESOC

· XMM-MOC changes · ESAM · R/F switch anomaly
M.Kirsch, XMM-Newton European Space Operations Center Page 2


XMM-Newton

ESOC

M.Kirsch, XMM-Newton European Space Operations Center Page 3


XMM-MOC
Spacecraft Operations Manager M. Kirsch
as of 01/04/09 (Currently M. Schmidt)

ESOC

Project-Support 0.2

PA-Support 0.1

4
Spacecraft Operations Engineers N.N. (D-SOM) J. Martin (OBDH/RF) U. Weissmann (P/Th) M. Pantaleoni (AOCS)

6
Spacecraft Controllers

Infrastructure support G/S Simulator MCS S/W maintenance 1.5 2.8 Flight Dynamics Support

(shared with Integral)

+

1
Analyst D. Weber

M.Kirsch, XMM-Newton European Space Operations Center Page 4


menu

ESOC

· XMM-MOC changes · ESAM · R/F switch anomaly
M.Kirsch, XMM-Newton European Space Operations Center Page 5


Events leading to ESAM #6
Attitude & Orbit Control System

ESOC
· EVENT: LCL Trip-off CAE-A (which drives the RCS). · ACTION: CAE-A unit switched on but the Unit Configuration WORD B in the ACC was not re-initialized. · EFFECT: ACC could not talk to the CAE-A the RCS was not driven. In the following Reaction Wheel Bias the thrusters didn't fire and the S/C entered in ESAM.

CAE: Control and Actuation Electronics RCS: Reaction Control System
· · · · 2008-10-05 09:49:25 (279): OOL P1104 (LCL STA CAE-A) OOL 2008-10-05 09:49:33 (279): OOL A5010(MACS ERRORS) OOL 2008-10-05 10:12:25 (279): CRP_AOC_4019( Unexpected LCL switch on or off primary branch) executed step5 TC P3041(LCL CAE A ON) sent 2008-10-05 10:20:26 (279): SCC reset as per step 3 of CRP_AOC_4025( ACC and errors) MACS errors still present but not anymore evident in OOL

in the anomalies

M.Kirsch, XMM-Newton European Space Operations Center Page 6


ESAM #6: entry
· wheels try to compensate the lack of thruster firing · Roll angle hits -5 deg threshold in SAS window ESAM

ESOC

Anomalous behaviour ESAM

Roll Angle Pitch Angle

FDE criteria SAS Window Threshold ± 5º

ESAM controller brings back Sun in the FOV within 3 min

rpm

deg

SAS Window Exceeded !

Wheels try to compensate the lack of thrusters fire 5 10

ESAM TRIGGERED

5

10

15 M.Kirsch, XMM-Newton
Space Operations Center Page 7

Time [min]

Time [min] ropean Eu


ESAM #6 recovery
· ESAM entry 2008-10-06 at 23:10:07 · Recovery into IPS after RCS-A/CAE-A failure (CRP_AOC_1528) · RWB tested with CAE-B / RCS-B

ESOC

all nominal - B branch healthy S/C able to manoeuvre and 2 days 13 h 2 days 13 h

· Slew to perigee attitude with RCS-B put to safe attitude

· Test of CAE to health status of CAE-A Ad hoc procedure, tested on the Simulator, was used to swap from CAE-B to CAE-A. · RWB with RCS-A CAE-A was declared healthy

· Back to nominal operations with RCS-A. · S/C on target on DOY 283, 2008-10-09 at 12:18:00
M.Kirsch, XMM-Newton European Space Operations Center Page 8


ESAM #6 follow up
· Fuel:
­ Consumption for ESAM: 320g ­ Consumption per month: 500g ­ Remaining Fuel after ESAM: 84 kg Fuel margin until 2019 (reserving fuel for 2 ESAM/year)

ESOC

· ·

LCL trip procedure corrected (CRP_AOC_4019) ESAM and main AOCS contingency procedures review and test with simulator ongoing Start mini simulation campaigns in 2009: consolidate/maintain the engineering know-how for platform and instruments
M.Kirsch, XMM-Newton European Space Operations Center Page 9

·


menu

ESOC

· ESAM · R/F switch anomaly
­ Recovery ­ failure analysis and operations strategy
M.Kirsch, XMM-Newton European Space Operations Center Page 10


Failure Occurrence 18/10/2008
Sequence of Events · 14.19: LOS Santiago · 15.27: Time Tag Command RFDU SWA POS X · 15.37: Expected AOS Villafranca no signal

ESOC

TM/TC gap
M.Kirsch, XMM-Newton European Space Operations Center Page 11


Spacecraft situation
· Configuration
­ ­ ­ ­ Nominal platform configuration Attitude safe for a long time Wheel speeds safe for a few days Payload configuration in a safe state

ESOC

· Operational Constraints
­ Wheel Speed limit violation predicted 19/10 ­ Autonomous Momentum Dumps ­ Orbit effected commanding into the blind more difficult ­ Sun Aspect Angle violation predicted 24/10 ­ ESAM entry ­ Orbit effected commanding into the blind more difficult
M.Kirsch, XMM-Newton European Space Operations Center Page 12


Failure Assessment on 18 Oct
· Ground station failure excluded ­ No failure identified at Villafranca ­ Santiago also did not find the signal despite of search pattern

ESOC
no no no

· ·

Error in orbit predicts excluded ­ New orbit determined and data consistent based on last known status before LOS Time tag command failure excluded ­ Excluded after sending relevant commands several times into the blind ­ CDMU failure not considered as this would not explain why there was no signal

·

Transmitter switch-off by SEU excluded ­ First considered as a likely failure but then discarded as commanding into the blind did not have the desired effect Failure of Transmitter excluded ­ When switching the 2nd transmitter into the blind and still no signal though there should be visibility

·

no

no

·

Failure of R/F Switch ­ Considered possible ­ Linked to the last planned satellite operation ­ Considered as possible failure for INTEGRAL

possible
M.Kirsch, XMM-Newton European Space Operations Center Page 13


RF Switch

ESOC

M.Kirsch, XMM-Newton European Space Operations Center Page 14


Additional Measures on 19 & 20 Oct

ESOC

· Repetition of standard contingency procedures, commanding into the blind
­ Commanding switch position ­ Transmitter 1 and 2 commanded on

· Different sweep procedures applied to get either Receiver 1 or 2 or both Receivers in lock · Different Virtual Channels (VC0 and VC63, i.e. different command chains)
M.Kirsch, XMM-Newton European Space Operations Center Page 15


Optical detection

ESOC

· October 20th: amateur astronomers at Starkenburg observatory imaged track of XMMNewton. Observations were also performed by two other telescopes, German radar and US Space Command No catastrophic event such as an explosion, thruster malfunction, or collision with space debris or a meteorite.

Stakenburg Observatory
·Newtonian

Telescope ·Aperture 450 mm ·Focal Length 2000 mm ·CCD Camera

· Satellite on predicted path · No visible debris · No indication of light fluctuation caused by S/C spin

M.Kirsch, XMM-Newton European Space Operations Center 16 Page 16


Radar and optical follow up
· ESA Space Debris Telescope in Tenerife 1 m aperture, 0.7º field of view, Ritchey-ChrÈtien focus, 2,401 m altitude, 28.2º North, 16.3º West CCD: 4096 x 4096 Pixel; ~ 2s integration time, ~ 19s readout time Limiting magnitude: 19 ­ 21 mag ( object of ~ 15 cm size in GEO); 120º of the GEO-ring are visible

ESOC

·

· ·

·
· TIRA (Tracking and Imaging Radar) at Wachtberg, Germany (lat: 50.62° North, 7.13° East) Monostatic L­band tracking radar and high resolution Ku­band imaging radar supported by one 34 m parabolic antenna This was the very first occasion for FGAN to track an object in more than 8000km range. XMM was at 22,000km (perigee) when FGAN started tracking

Zimmerwald telescopes (CH) Owned and operated by the Astronomical Institute, University of Berne, 950 m altitude at 46.9° North, 7.5° East

·

·

M.Kirsch, XMM-Newton FGAN TI ter European Space Operations CenRA ra Page 17


Orbit Determination Results
· Although observation data only covered a period of 4 hours after perigee passage, it was sufficient to determine an accurate orbit Orbit determination was independent from any operational data Orbit determination accuracy was sufficient to clarify that the orbit was nominal (matched the predicted one within 1km RMS)
USED OBSERVATIONS TIRA TIRA, ZIMSMART, ZIMLAT TIRA, ZIMSMART, ZIMLAT, ESASDT
A [KM

ESOC

Comparison: predicted and measured orbit

·

·

]

E

[-]

I [DEG]

[DEG] 102.8368 ±0.0196 102.8258 ±0.0002 102.8261 ±0.0001

[DEG]

V [DEG]

66849.6 ±198.4 66926.7 ±23.0 66929.3 ±2.8

0.5825812 ±0.0011727 0.5830662 ±0.0001321 0.5830820 ±0.000015

57.7900 ±0.0157 57.7797 ±0.0002 57.7794 ±0.0001

116.3677 ±0.0283 116.4064 ±0.0032 116.4058 ±0.0009

8.0323 ±0.0193 8.0096 ±0.0031 8.0100 ±0.0009

M.Kirsch, XMM-Newton European Space Operations Center Page 18


Observation Data
· · Despite of the enormous range, FGAN was able to generate 1.5 hours of tracking data Fusion of data from optical telescopes and radar was performed for the very first time in ESA
Sensor TIRA at FGAN Zimmerwald telescopes ESASDT Data Type Range, azimuth, elevation Right ascension, declination Right ascension, declination

ESOC
Arc [2008/10/20] 17:28:43 18:07:06 17:23:33 18:30:15 21:17:13 21:35:12

· ·

While weather conditions on Tenerife allowed only 20 minutes of observation, the Zimmerwald telescopes were able to observe for about 1 hour High frequent magnitude variations were not observed revealing that the Light Curve 99066A attitude was nominal
9 10 11 12 13 14 15 0 500 1000 1500 2000 2500 3000 3500 4000

Magnitude

4500

Seconds past 20081020 17:23:33

M.Kirsch, XMM-Newton European Space Operations Center Page 19


Optical data
· · · Astrometry is precise down to a level of a few arcsec Post-processing and determination of object states was done within 1 hour

ESOC

ESA Space Debris images re-confirmed that XMM was intact and no debris (> 10cm) was visible in its vicinity (150km)

·

The telescope was tracking with XMM's angular velocity Three overlayed frames showing XMM as dots and stars as trails (2-second exposures)

·

M.Kirsch, XMM-Newton European Space Operations Center Page 20


Radio Contact
· · · ·

ESOC

On 21 Oct a very weak signal was picked-up by ESA's 35-m antenna in New Norcia (West Australia) Doppler information could be collected to confirm the orbit derived from the radar and optical tracking data Measurement of the signal attenuation (-55 db) indicating a failure of the switch information concerning the required uplink power Commanding not possible via New Norcia, uplink power not sufficient Commanding not possible via Canberra, frequency not supported
35m ESA New Norcia

M.Kirsch, XMM-Newton European Space Operations Center Page 21


Final Recovery
· 22nd Oct: Configuration of Goldstone to be able to support XMM · Validate procedures on simulator · Uplink command to put RF switch into = position using high uplink power of Goldstone · Check Satellite status · Update Platform and Payload configuration to put satellite into a safe state

ESOC

M.Kirsch, XMM-Newton European Space Operations Center Page 22


menu

ESOC

· ESAM · R/F switch anomaly
­ Recovery ­ failure analysis and operations strategy
M.Kirsch, XMM-Newton European Space Operations Center Page 23


RF
· ·

switch failure

ESOC

For downlink hemispherical coverage the nominal operating transmitter will be interconnected to the earth pointing antenna by the Transfer Switches (RF switch). The Transfer Switches, that interconnect the transponders to the antennae, are of pulse latching type and operate in break-before-make mode. Each Transfer Switch will be capable of performing minimum 50.000 switching cycles and was tested to 10.000 hot switchings by INTEGRAL in October 1999 (same switch as XMM1999 Newton) Status Switch on 06-10-2008 ­ Total number of switchings: 3486 ­ Number of switchings left: 6514 On 18 October, XMM-Newton failed to establish contact with the Villafranca 2 (Vil-2) ground station upon emerging from a nominal period of non-visibility between the Santiago ground station in Chile and Vil-2 in Spain. Reason: RF switch SW-A stuck in open position after Time-Tag command execution
M.Kirsch, XMM-Newton European Space Operations Center Page 24

·

·

·

·


Possible causes
1. application of simultaneous switch commands not possible with s/c or ground design 2. extremely high vibration and shock levels no more possible 3. Imperfect switch commands never achieved during pre-launch ground testing (waiver issued) (however, imperfect pulse was demonstrated during ESOC Lab testing, switch stuck in neutral position at about 50% of the commands) 4. Mechanical problem · (switch cannot be moved at all) no; switch was moved back to +z position · switch cannot be moved to one position (-z) because of broken contact possible; needs testing (?)

ESOC

5. Electrical problem · full or partial coil failure to move to ­z position possible; needs testing (?) · impact of SEU (Single Event Upset) caused imperfect pulse (?) very low possibility · other component failures that could lead to degraded pulses M.Kirsch, XMM-Newton none European Space Operations Center identified Page 25


RF switch testing (ESOC Lab)
Objective:
1. Demonstrate whether it is possible to reproduce the suspected neutral position under a variety of failure conditions. 2. The EM RFDN switches with serial number 001 & 002 were removed from the XMM/Integral RF suitcase and being regarded as representative of the flight model switches and used for the test

ESOC

Conclusions:
· Switch with serial number 001 can be put into neutral position with reduced pulse duration with a suspected bias in one direction. · Reducing the amplitude of the pulse appears to cause the switch not to operate at all. · Losses through the switch are consistent with flight anomaly observed values · Switch with serial number 002 cannot be made to fail, but there appears a possible mechanical bias when switching with a reduced pulse by the clicking of the switch on one direction only until switchover

M.Kirsch, XMM-Newton European Space Operations Center Page 26


Recommendations
R 1:
Establish Operational Concept without the use of any RF switch · use of only 1 transmitter · use both transmitters permanently on or TX1 always on and TX2 on/off · alternate operations of transmitters TX1 on/TX2 off and TX1off/TX2 on

ESOC

R 2:
Use back-up RF switch · switch can only be commanded with direct commands · switch cannot be commanded by use of time-tag commands · therefore if switch gets stuck, safety time-tag commands to move switch back to previous position cannot be used

R 3:

Test of RF-A switch

· most likely cause of anomaly is an imperfect pulse ·multiple time-tag commands to move switch would cover both cases Management permission for a test of R2 and R3 is required.
M.Kirsch, XMM-Newton European Space Operations Center Page 27


Operational concept: Alternate switching of transmitters
· Provides Omni-directional coverage:
­ Lock uplink on receiver to be used ­ Switch on TX to be used ­ Switch off TX no further in use
T1 R1 T2

ESOC

R2 R2

T1

R1

T2 T2

R2

almost no impact to the mission and easy to coordinate with mission planning and network activities operational since beginning of December 2008

T1

R1

T2

R2

T1

R1

T2

R2

M.Kirsch, XMM-Newton European Space Operations Center Page 28


Implementation
· · · · ·

ESOC

FD generates PSF SOC modifies PSF by introducing antenna hand over windows SOC generates POS MOC generates EPOS and from EPOS automatically hand over command stacks Spacon loads command stacks for handover
45 40 35
Time after Perigee (h)

30 25 20 15 10 5 0 1630

Sw ap to +Z Sw ap to -Z Sw ap to +Z Sw ap to -Z Sw ap to +Z Apogee Christmas 40.000 km 40.000 km real

1650

1670

1690
Rev

1710

1730

1750

M.Kirsch, XMM-Newton European Space Operations Center Page 29


Conclusion
· Full mission can be conducted by using the option alternating the 2 transmitters. This option is in operation since beginning of December 2008

ESOC

­ Resumption of science operations after the R/F anomaly took place in Rev1631 (starting on 03/11/08) and full orbit operations (which allowed recovering the standard science efficiency) were resumed as of Rev-1649 (09/12/08). ­ After the definition of the new strategy for operations by the MOC, the SOC modified internal planning tools to include the new constraints. This has allowed fully utilising the available science window and operating with the efficiency achieved before the anomaly.

· ·

Full mission can also be resumed by using the back-up switch. Also the option having 1 transmitter on all the time and switching on the other transmitter when needed is possible but less favourable since more complex. Only to be used as back-up. Should one of the two transmitters fail, the RF switch (prime or redundant) needs to be used again
M.Kirsch, XMM-Newton European Space Operations Center Page 30

·