Extreme stability: some lessons from Gaia - SpaceTec-CM · Alcione Mora et al. | Extreme stability: some lessons learned from Gaia | SpaceTec Madrid| 12/02/2016 | Slide 2 ESA UNCLASSIFIED

Issue/Revision: 1.0

Reference: Presentation Reference

Status: Issued

ESA UNCLASSIFIED - Releasable to the Public

Extreme stability:some lessons from GaiaAlcione MoraESA-ESAC Gaia SOC

RIA-SpaceTec workshop. CSIC Madrid (Spain) 12/02/2016

With contributions from U. Bastian, M. Biermann, F. Chassat4, C. Fabricius,J. Hernandez, R. Kohley, L. Lindegren, E. Serpell, I. Serraller and W. van Reewen

Alcione Mora et al. | Extreme stability: some lessons learned from Gaia | SpaceTec Madrid| 12/02/2016 | Slide 2


Outline

Preliminary results. Gaia still to be fully understood!

1.Gaia stability: architecture

2.Miscellaneous effects: focus, contamination,

micrometeoroids and microclanks

3.The Gaia basic angle. Importance, measurement, in-

orbit results, correlations

4.Some lessons learned



1. Gaia stability: architecture



1. Gaia stability: architecture

Passive thermal stability: sun shield, thermal tent Constant 45º sun aspect angle constant irradiation No moving parts (gyros, reaction wheels, antenna steering, …) Payload on top of service module (SVM), isolated via bipods Thermal disturbances in SVM (computers, transponders, …)

Airbus DS



2. Miscellaneous effects: focus, contamination and micrometeoroids



2. Focus evolution

1. Astrometric focus: Cramér-Rao PSF sharpness (derivatives)

2. The focus evolves during the mission. The rate gets progressively smaller

1. Some hypotheses: glue shrinkage, hysteresis, water contamination

3. Mitigation: adaptive calibration, occasional refocus

FoV1. FoV1

FoV2. FoV2

A. Mora



2. Throughput loss

Monitoring of response by comparison to Tycho-2 photometry

First decontamination (FOV2)

Second decontamination

C.Fabricius, commissioning



2. Micrometeorids and microclanks

Micrometeorids: L2 is not perfect vacuum!● Attitude discontinuities expected and calibrated

Micro clanks: 1-2 mas micro clanks per minute 7.5 nm 20 Si atoms● Calibration in place. What if target precision were 10-100x higher?

8F. Van Leeuwen, commissioning



3. The Gaia basic angle



3. The Gaia basic angle

Γ

Γ basic angle



3. Basic angle variations

Gaia aims at global astrometry (reference frame, stellar motions and parallaxes) at µas accuracy

The basic angle needs to be stable (or known) to corresponding precision

Gaia is largely self-calibrating (calibration parameters estimated from observations)

Low frequency variations (f < 1 / 2Prot): eliminated by self-calibration

High frequency random variations

● Averaged during all transits, not so harmful

Systematic variations synchronized with spacecraft spin

● Only partially possible to eliminate by self-calibration

● Residual variations could create systematic errors in astrometric results

● Thus high-frequency variations need to be monitored by metrology



3. Basic angle measurement: BAM

One artificial fixed star per telescope needed● Collimated laser beams directed to the primary mirrors

● Gaia telescopes generate the image

Relative AL centroid displacement basic angle variation

Single CCD AL centroid location precision● AF: ∆yAL ~ 40 μas single transit bright star limit

● BAM: ∆yAL <0.5 μas in 10 min (differential measurement)– ~20s/frame ∆yAL,1frame < 2.7 μas. ~15x better than bright stars!!

Many photons and sharp LSF needed● Artificial stars are interferometric patterns

Better than LIGO or eLISA at low frequencies (0.02 – 1 mHz)



3. BAM working principle

BAM measurement: differential interference pattern centroid

Airbus DS

0.5 μas = 2.4 prad = 3.6 pm = 0.66 μfringeOn ground state of the art: 1 milli-fringe



3. BAM phase and period variations

Fringe phase periodic shift: Sun synchronous, ~1mas (nm stability!) Fringe phase discontinuities: several per day Fringe phase mid-long term evolution Fringe period variability

A. Mora, commissioning



3. BAM phase periodic component

Periodic signal preliminar Fourier analysis● 6-12 harmonics of rotation period: mas µas● Slow temporal evolution + plenty of data can be characterised

Model input for the AGIS solution

L. Lindegren, LL-105




Sometimes related to on-board events

3. BAM phase discontinuities



One Day Astrometric Solution (ODAS): daily average (no periodicity) Some discontinuities are real. Long term evolution is not (but irrelevant)

3. BAM phase long term evolution

M. Biermann, FLS-033

ODAS



3. BAM period variability

E. Serpell, GAIA-ESC-TN-0066

Wavelength depends on laser temperatureand current

Power consumption depends on the sky!

• RVS LR-HR mode• VPU, PDHU power

Commissioning: ±0.005 K stability ~1/250,000

Now: ~1 mK~1/1,000,000 (no RVS LR mode)



3. BAM period variability

Laser temperature stability ~1 mK fringe period changes ~10-6 House keeping data affected by quantisation and time sampling

A. Mora



3. BAM fringe period vs position

1. Wavelet analysis: constant fringe period does not exist! NEAT, Theia2. Plane parallel fringe analysis works, but imperfect3. Changes are not homogeneous




A. Mora, AMF-015

3. BAM (non-) Gaussianity

BAM pattern: more complex than pure Gaussian beam inteference● Wavefront errors unavoidable can be modeled?



3. BAM fit residuals

24 h slow component: related to downlink (transponder + PDHU) Peaks: follow sky density (galactic plane) SVM computers thermoelastic efffect

● Most SVM perturbations modify basic angle. Rule of thumb: 100 μas/K

Reverse scalefor temperature

Airbus DS



E. Serpell priv. comm.

3. BAM vs spin restart

Spin restart after safe mode: variations appear very soon: few min BAM signal: periodic + transient Expected if thermoelastic



3. Ultimate stability: Earth shadow

1. The Sun strongest perturbations mas (= nm)

2. Avoided in L2 Earth’s shadow + nuclear power

1. No RTGs in Europe. Environmental concerns

2. Expensive, low power/weight ratio. Research?


NASA



4. Some lessons learned



4. Some lessons learned

Design for stability, but plan for instability:● State of the art in stability is mas, nm and mK● Future is µas, pm and µK. Models might not be ready● Do not play too much with the spacecraft. Routine is important

Telescope focus is not (that) stable refocus campaigns foreseen Water contamination obiquitous decontamination big perturbations Discontinuities in SVM and PLM micrometeoroids, thermal relaxation Minute changes in SVM important ensure constant power load HK is key high precision, resolution and, temporal frequency High precision metrology is essential and (never too) expenseve

● Calibration is difficult Simple models don't work● BAM data are an essential ingredient for Gaia data release 1



Thank you for your attention!

Stay tuned for data release 1!

http://www.esa.int/spaceinimages/Images/2015/08/Gaia_s_first_Hertzsprung-Russell_diagram

Extreme stability: some lessons from Gaia - SpaceTec-CM · Alcione Mora et al. | Extreme stability: some lessons learned from Gaia | SpaceTec Madrid| 12/02/2016 | Slide 2 ESA UNCLASSIFIED

Documents