MICROSCOPE: preliminary data processing methods

R. Chhun, E. Hardy, A. Levy, M. Rodrigues, P. Touboul (ONERA)

G. Métris (OCA)

From Quantum to Cosmos 5 (11/10/2012)

2

Contents

• The EP test in space• MICROSCOPE mission and instrument• Mission scenario• Data description• Data post-processing• Scientific activity and ground segment

3

Equivalence Principle

All bodies, independently of their mass or intrinsic composition, acquire the same acceleration in the

same uniform gravity field

Gravitational mass = Inertial mass

Newton:

Universality of free fall

General relativity can not merge gravitation with quantic mechanicsA new model is necessary:New interactions? New particles?String theory and supersymmetry?Space dimension >4?All these approaches ask for a violation of the equivalence principle at some level (10-14 to 10-21….)

Cstmm

i

g ga amgm ig

Einstein, 1907GR, 1916

Today

4

EP Testing in Space

On ground: • Seismic vibration (Earth, human activity) ~ µg• Local gravity to be compensated• Earth gravity gradient variations• Duration and measurement frequency difficult to master

“Free fall” test in space:• Vibrations reduced by several orders of magnitude ~ nano-g• Field sources = Earth + reduced local gravity• Measurement frequency = very well known orbital frequency• Very long fall duration (several orbits)• Dedicated instrument with sufficient limited range because in free fall• Highly stable thermal conditions

5

The MICROSCOPE Test of Universality of Free Fall

Inertial or spinningsatelliteMeasurement axisProof masses:material 1 (Pt)material 2 (Ti)Acceleration

• Servo-controlled concentric masses in the same gravitational field

• Measure of acceleration needed to maintain the two masses on the same orbit

• Differential acceleration ≠ 0 EP violation

• 2 differential electrostatic accelerometers (2 pairs of masses : Pt/Pt & Pt/Ti)

Measurement axis

6

Star Sensor

Drag

Solar Panels

Microthrusterpods

CNES MYRIADE CNES MYRIADE MicrosatelliteMicrosatelliteLaunch scheduled for 2016Circular Orbit : 720 km Inertial orbit or with satellite rotationMission duration : 11 monthsMicrosat : 200 kg, 1m3

Payload budgets : 35 kg, 40 Watts4 3 cold gas microthrustersContinuous drag compensation system

and orbit/attitude controlPassive thermal control

MICROSCOPE Mission Main Parameters

7

Instrument Description

360 x 348 x 180 mm3 x 20kgMechanical thermal model

• 2 differential accelerometers = 2 pairs of test-masses relatively centered at better than 20µm after integration• Integrated inside highly insulated thermal case• Integrated inside a µmetal magnetic shield• Rigidly linked to the satellite star sensor

Vacuum System

Blocking System

External Acc Electrode Cylinders

External PM(PtRh10 or TA6V)

Internal PM (PtRh10)

Internal AccElectrode Cylinders

• Proof masses with spherical inertia minimize effects of gravity gradient

• Electrode cylinders in gold-plated silica thermal geometrical stability

• Electrostatic control of 6 degrees of freedom of each mass steady configuration

• Retractable stops mass blocked during launch limited electrical disturbances

• Tight housing with getter material vacuum (<10-5 Pa)

• Invar housing matched CTE with silica + magnetic

shielding complementary to µmetal global

magnetic shield

1 differential accelerometer

External inertial sensor

Internal inertial sensor

8

The measure The measure Earth, satellite, instrument, physics contributionsEarth, satellite, instrument, physics contributions

Stochastic and Tone Signals to be considered with a limited observation period and missing data Detailed Specifications for S/C Sub-Systems, Instrument Environment & Instrument Performances Accurate in orbit calibration A posteriori estimation and corrections

czcxcxcycy

cxcxcyczcz

cycyczczcx

KK

K

dndmeasquadcappdsatcddmeas MOgCorInTMK ,,,,0, )(..2

Measured Differential Acceleration

Bias differencelimited thermal

fluctuations

Common Mode Sensitivity Matrix (Inst. Scale Factor &

Attitude, Coupling)Estimated by calibration

or limited by construction

Earth Gravity gradient tensorComputed with

model, S/C position & attitude,

and removed

Inertia Tensor (Angular Velocity and

Acceleration)Minimized by AOCS

from SST & Inst. data

Differential Mode Sensitivity Matrix (Scale

Factor Mismatching & Misalignment)

Estimated by calibration

Common mode acceleration

(S/C drag-free Control from Sensor

common data)

Instrumentnoise

2 2

2 2

2 2

y z x y z x z y

x y z x z y z x

x z y y z x x y

In

dzdxdxdydy

dxdxdydzdz

dydydzdzdx

KK

K

Coriolis tensor

0

00

z y

z x

y x

Cor

Quadratic residue

1

1

2

2

I

g

I

g

mm

mm

EP violation Masses excentring

9

Requirements and constraints

What must be done:• Acquire precise measurements for the EP test in varied experimental conditions:

• Inertial pointing instrument: different phases• Rotating Instrument: different velocities• Re-centered proof-masses or not

• Validate performance:• Accurate calibration in flight of the main parameters (sensitivity matrices…)• Fine characterization of instrument behavior with respect to thermal experimental

conditions…

Constraints :• Limited gas quantity for propulsion

• Need to establish priorities• No improvisation during mission flow

• Operational Security:• Need to foresee what to do and how to do it• Different EP test, calibration, characterization sessions: new session types not impossible but

limited • Mission scenario is a flowchart of sessions: flexibility at medium and long term

1010

Mission scenario: minimum and expected

S/C & payload Operation Verification & Adjustment

Preliminary Tests andEP Inertial sensor calibration

EP Tests with andwithout mass centering

Calibrations of bothEP and REF Instrument

REF Tests with andwithout mass centering

Additional EP Tests& calibration

REF Instrument new calibrationfor stability verif.

SUREF = Pt - Pt : SUEP = Pt - Tis

11

Data levels

• N0 level data : operational

• N1 leval dataCalibrated acceleration measurements per inertial sensor+ Complementary mission support data:orbit, attitude, acceleration and gravity gradient…

• (N1a) : reference calibration, fixed for the whole missionduration “raw” data for independent analysis

• (N1b) : Latest validated calibration phase data calibrated with as little “scientific bias” as possible

• (N1c) : calibration considered the most pertinent (time evolution, thermal evolution…) Best data but impacted by calibration choice

+ data necessary for the interpretation of the experiment…

• N2 level dataDifferential accelerations with correction of the gravity gradient effect among others + EP violation (or not)

• (N2a) : derived from (N1a), corrected of gravity gradient and excentrings estimated on ground• (N2b) : derived from (N1b), idem using excentrings estimated from test session data• (N2c) : derived from (N1c), idem using most relevant estimated excentrings and corrected of additional effects

with respect to the sensitivity of the instrument to its environment

Type Vol.%

Data description

ALIM 0.23 Battery charge, equipment current/voltage measurements, temperatures

SCAA/MCA

57.64 Attitude SST in J2000, Command outputs SCAA, guidance, combination TM T-SAGE for SCAA

CGPS 12.84 Cold gas propulsion system HK

OBC 0.01 Status bits of OBC and its components (FPGA, CPU…)

TTC 0.04 Antenna status

CU 28.73 T-SAGE 4Hz et 1 Hz data (angular et linear acc., position sensors, temperatures, reference voltages,

statuses….)

Thermi. 0.04 Platform temperature

GestionBord

0.16 TM, TC handling, modes, work plan, ….

Equipmt 0.01 SST raw data, inertia wheel

GNSS 0.3 GPS data (if present)

12

Data exploitation time scales

Different time scales with different flexibilities:• 1-week horizon: operational loop

• Verification of data integrity by automated processing• Mission program fixed (except potential stop or extension of a long session)

• 1-month horizon:• Preliminary analysis of data• Scenario still modifiable, in the frame of the predefined sessions

• 1-year horizon:• Detailed scientific analysis• Detailed performance analysis• Optimization of calibration processing• Application of data correction models

13

Influence of the observation window

TF TF

• Aliasing: a perturbation at any frequency has a component at the EP violation frequency (fEP)

EP violation signal:

Projection rate of a perturbation on the EP violation signal at fEP:

EPEP

dEP

SSSS

,,

)sin( EPEPEPEP tAS

)sin( dddd tAS Disturbance signal:

10-5 10-4 10-3 10-2 10-1

10-8

10-7

10-6

10-5

10-4

10-3

10-2

10-1

100

fdfréquence du signal perturbateur

Pro

ject

ion

max

imis

ée p

ar ra

ppor

t aux

pha

ses

Gabarit à fEPInertiel - 120 orbites

ProjectionSpécification

fEP

14

Main disturbances: due to the orbital or spin motion, n1forb+n2fspin→ choice of fspin and Tobservation to have minimal projection rates:

- Tobservation = k1.Torb- Tobservation = k2.Tspin

Choice of the measurement duration and spin frequency, special specifications

To distinguish forb and fEPk2 = 73

to reduce the noisek1 = 20k1 = 120Spinning modeInertial mode

10-5 10-4 10-3 10-2 10-110

-8

10-6

10-4

10-2

100

fd

fréquence du signal perturbateur

Pro

ject

ion

max

imis

ée p

ar ra

ppor

t aux

pha

ses

Gabarit à fEP

Inertiel - 120 orbites

ProjectionSpécification

fEP

Influence of knowledge and realization error:- Knowledge on the orbital frequency : 2.10-8rad/s (↔ 100m for the orbit altitude)- Command error on the spin frequency: 3.10-8rad/s- Realization error on the inertial pointing: 1.10-8rad/s

3.3182 3.3182 3.3183 3.3183 3.3183 3.3183 3.3183 3.3184

x 10-4

10-7

10-6

10-5

10-4

2forbestimée

2forbréèlle

Not 0!

15

Non-rectangular windows

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1Fenêtres temporelles

t

fenêtre rectangulairefenêtre de Hannfenêtre de Hammingfenêtre de Blackman

-0.05 -0.04 -0.03 -0.02 -0.01 0 0.01 0.02 0.03 0.04 0.050

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1Transforméee de Fourier des fenêtres

f

fenêtre rectangulairefenêtre de Hannfenêtre de Hammingfenêtre de Blackman

10-5 10-4 10-3 10-2 10-110-18

10-16

10-14

10-12

10-10

10-8

10-6

10-4

10-2

100

fd : fréquence du signal perturbateurP

roje

ctio

n m

axim

isée

par

rapp

ort a

ux p

hase

s

Gabarit à fEP

Inertiel - 120 orbites

fenêtre rectangualairefenêtre de Hammingfenêtre de Hannfenêtre de Blackmanspécification

10-3.79 10-3.78 10-3.77

100

fdfréquence du signal perturbateur

Pro

ject

ion

max

imis

ée p

ar ra

ppor

t aux

pha

ses

Gabarit à fEPInertiel - 120 orbites

fenêtre rectangulairefenêtre de Hammingfenêtre de Hannfenêtre de Blackmanspécification

Comparison between different windows: important improvement, except around fEP

TF

16

Error budget

• Rejection rates induced errors

• Attitude control error budget

• Pointing, angular velocity and angular acceleration control performance / specifications among ~60 contributions

• Mission error budget

Error general form (in mesd)

0 Md x T x OdfpO in DF res.acc. 0.00E+00 0.0% 0.00E+00 0.0%1 1/2 x (I+Mc) x (T x res + Tres x ) 7.27E-16 0.0% 7.78E-16 13.9%2 1/2 x ( TTres+resTres

T) 9.02E-19 0.0% 1.37E-15 24.4%3 Instrument self gravity (wrt thermal stability) from finite elements model 6.00E-14 1.3% 2.00E-16 3.6%4 Satellite Gravity in DF residual acceleration 0.00E+00 0.0% 0.00E+00 0.0%5 Md x Tsc x OdfpO in DF res.acc. 0.00E+00 0.0% 0.00E+00 0.0%6 1/2 x (I+Mc) x Tsc x 1.21E-13 2.6% 3.04E-16 5.4%7 Md x 2 x OdfpO in DF res.acc. 0.00E+00 0.0% 0.00E+00 0.0%8 1/2 x (I+Mc) x 2 x 1.62E-17 0.0% 8.05E-17 1.4%9 Md x d/dt x OdfpO in DF res.acc. 0.00E+00 0.0% 0.00E+00 0.0%

10 1/2 x (I+Mc) x d/dt x 2.02E-13 4.3% 4.05E-16 7.2%11 in DF residual acceleration 0.00E+00 0.0% 0.00E+00 0.0%12 1/2 x (I+Mc) x x d/dt 4.15E-19 0.0% 2.91E-20 0.0%13 in DF residual acceleration 0.00E+00 0.0% 0.00E+00 0.0%14 1/2 x (I+Mc) x d2/dt2 3.13E-16 0.0% 1.05E-18 0.0%15 in DF residual acceleration 0.00E+00 0.0% 0.00E+00 0.0%16 1/2 x (I+Mc) x (T+Tsc+2+d/dt) x 4.39E-16 0.0% 3.41E-18 0.1%17 d x c 5.72E-14 1.2% 0.00E+00 0.0%18 c x d 3.31E-14 0.7% 2.76E-16 4.9%19 d x (c + Binst) 1.80E-13 3.8% 0.00E+00 0.0%20 c x (d + Binst) 5.00E-13 10.6% 0.00E+00 0.0%21 d x (c + Binst) 2.05E-13 4.3% 0.00E+00 0.0%22 c x (d + Binst) 1.41E-13 3.0% 0.00E+00 0.0%23 Accelerometer measurement noise Instrument error budget 3.29E-12 69.5% 0.00E+00 0.0%24 Bias sensitivity to SU temperature variation dBinst/dTmec x Tmec 6.06E-13 12.8% 2.02E-15 36.1%25 Bias sensitivity to FEEU temperature variation dBinst/dTelec x Telec 1.01E-13 2.1% 1.01E-15 18.0%26 dd/dTmec x Tmec x (c) 8.58E-14 1.8% 4.02E-16 7.2%27 dc/dTmec x Tmec x (d) 6.32E-18 0.0% 4.54E-20 0.0%28 dd/dTelec x Telec x (c) 1.43E-14 0.3% 2.01E-16 3.6%29 dc/dTelec x Telec x (d) 1.05E-18 0.0% 2.27E-20 0.0%30 dKd/dTmec x Tmec x (c+Binst_c) 1.44E-13 3.1% 4.80E-17 0.9%31 dKc/dTmec x Tmec x (d+Binst_d) 1.20E-13 2.5% 3.69E-16 6.6%32 dKd/dTelec x Telec x (c+Binst_c) 1.20E-13 2.5% 1.20E-16 2.1%33 dKc/dTelec x Telec x (d+Binst_d) 1.00E-13 2.1% 1.00E-15 17.9%34 dd/dTmec x Tmec x (c+Binst_c) 1.23E-13 2.6% 5.79E-16 10.3%35 dc/dTmec x Tmec x (d+Binst_d) 8.49E-14 1.8% 4.00E-16 7.1%36 dd/dTelec x Telec x (c+Binst_c) 2.05E-14 0.4% 2.89E-16 5.2%37 dc/dTelec x Telec x (d+Binst_d) 1.41E-14 0.3% 2.00E-16 3.6%38 Bias sensitivity to SU thermal gradient variation dBinst/dgradT x gradT 1.20E-12 25.4% 2.00E-15 35.7%39 Satellite positioning 1/2 x dT x 5.27E-17 0.0% 3.00E-16 5.4%40 Timing error 1/2 x T x 3.54E-17 0.0% 1.59E-16 2.8%41 Synchronisation error ep x t x c + ep x t x c 5.28E-19 0.0% 1.10E-18 0.0%42 DF residual acceleration Md x Resdrag 9.95E-14 2.1% 5.00E-16 8.9%43 Magnetic field (Earth+local) effect Analysis 8.00E-14 1.7% 4.00E-16 7.1%44 Radial measurement noise introduced by data processing (d+d)est x ninst_c 3.20E-13 6.8% 0.00E+00 0.0%

45 Radial bias sensitivity to SU temperature variation introduced by data processing (d+d)est x dBinst_c/dTmec x Tmec 1.36E-13 2.9% 6.40E-16 11.4%

46 Radial bias sensitivity to FEEU temperature variation introduced by data processing (d+d)est x dBinst_c/dTmec x Telec 2.26E-14 0.5% 3.20E-16 5.7%

47 Radial bias sensitivity to SU thermal gradient variation introduced by data processing (d+d)est x dBinst_c/dgradT x gradT 1.36E-13 2.9% 3.20E-16 5.7%

48 Accelerometer angular noise x acc_ang 5.52E-15 0.1% 0.00E+00 0.0%49 Angular attitude control variations x d/dt 1.06E-14 0.2% 3.00E-17 0.5%50 Angular bias sensitivity to SU temperature variation x dacc_ang/dTmec x Tmec 3.18E-16 0.0% 1.50E-18 0.0%51 Angular bias sensitivity to FEEU temperature variation x dacc_ang/dTmec x Telec 7.42E-16 0.0% 1.05E-17 0.2%52 Angular/linear coupling noise x angular accelerometer bias x acc_ang 1.48E-13 3.1% 0.00E+00 0.0%53 Angular/linear coupling noise x attitude control bias x d/dt 8.49E-15 0.2% 0.00E+00 0.0%54 Ang/Lin coupling sensitivity to SU temperature variation d/dTmec x acc_ang x Tmec 8.91E-14 1.9% 4.20E-16 7.5%55 Ang/Lin coupling sensitivity to SU temperature variation d/dTmec x d/dt x Tmec 5.09E-15 0.1% 2.40E-17 0.4%56 Ang/Lin coupling sensitivity to FEEU temperature variation d/dTelec x acc_ang x Telec 0.00E+00 0.0% 0.00E+00 0.0%57 Ang/Lin coupling sensitivity to FEEU temperature variation d/dTelec x d/dt x Telec 0.00E+00 0.0% 0.00E+00 0.0%58 K2c x appc x appd 4.20E-13 8.9% 1.29E-15 23.0%59 K2d x (appc

2 + appd2) 5.05E-13 10.7% 1.68E-16 3.0%

TOTAL (direct sum)/3 5.55E-15 99.0%TOTAL (quadratic sum) 3.71E-12 78.3%

SPECEP 8.93E-16

tone @fep (m/s²) in 2mesd

Coupling sensitivity to FEEU temperature variation

Alignment sensitivity to SU temperature variation

Alignment sensitivity to FEEU temperature variation

Scale Factor sensitivity to SU temperature variation

Scale Factor sensitivity to FEEU temperature variation

Alignment (not thermal stability)

Coriolis (wrt thermal stability)

PM average and relative positions (wrt thermal stability)

noise (m/s²/Hz^1/2) in 2mesd

4.73E-12

noise (m/s²/Hz^1/2) in 2mesd

5.60E-15

Quadratic factor

tone @fep (m/s²) in 2mesd

Coupling (not thermal stability)

Coupling sensitivity to SU temperature variation

Earth Gravity Gradient

PM average and relative accelerations (wrt thermal stability)

Satellite Gravity Gradient

Centrifugal acceleration

Angular acceleration

Scale Factor (not thermal stability)

TOTAL (direct sum)/3 5.55E-15 99.0%TOTAL (quadratic sum) 3.71E-12 78.3%

SPECEP 8.93E-16

4.73E-12

noise (m/s²/Hz^1/2) in 2mesd

5.60E-15

tone @fep (m/s²) in 2mesd

17

Measurement loss

• Teletransmission errorsInformation from Picard mission:• frequency: about 100 events over 10 months• duration: from seconds to hours

• Coating cracking• due to temperature changes (Earth / Space vacuum)• frequency: for each of the four satellite sides, about 6 times

when the side faces the Earth• duration: 0.5-0.75s 2 - 3 measurement points

• Tank cracking• worst case, depending on gas pressure• frequency: for each of the 6 tanks, about 43 times/orbit• duration: 0.5s 2 measurement points

18

Measurement loss

• Without measurement loss:

TF(S) = TF( )*TF( )• With measurement loss: replacement by zeros

TF(S) = TF( )*TF( )• With measurement loss: replacement by the mean value of the measurement before and after the interruption

TF(S) = TF( )*TF( )

19

Short duration loss

- one measurement loss

- simulations with different duration

- respect the specifications up to 1 minute

No measurement loss

First method: replacement by zeros

Second method: replacement by the mean value before and after the interruption

20

Division into subsections

16% Not acceptableInertial session: 120 orbits2,9% AcceptableSpinning session: 20 orbitsProbability of failureActivity

With the measure replacement procedure, session failure criterion: more than 1 measurement loss of duration > 1 minute per session

For data losses > 1 minute in inertial session (and < 1 orbit): → inertial session of 120 orbits divided into several subsections

- Tsubsection,i = niTorb (rejection of the main perturbations)

Need to consider an alternative correction method

+ additional orbits

21

Division into subsections

0,07%40.4%32,5%216%1

Probability over 120 orbits

Number of data loss > 1 minute and < 1 orbit

Success probability > 99,9%

10-5 10-4 10-3 10-2 10-110-8

10-7

10-6

10-5

10-4

10-3

10-2

10-1

100

fdfréquence du signal perturbateur

Pro

ject

ion

max

imis

ée p

ar ra

ppor

t aux

pha

ses

Gabarit à fEP en inertiel

Cumul de 4 sessions séparées de 1 orbite

20 orbites + 20 orbites + 40 orbites + 40orbites20 orbites + 20 orbites + 20 orbites + 60 orbitesSpécification Method is robust up to 3

losses considering a worst case distributionOver 4 losses, success depends on losses distribution

3 losses, 2 distributions

22

Data post-processing summary

• Influence of the observation window• Perturbations at any frequency can have a component at the EP

frequency• Main perturbations frequencies: adjusted to have minimal projections,

but effects are amplified by frequency knowledge and realization• Numerical estimation of the projection rate: compatible with the

specification• Influence of the measurement losses on the projection and

rejection rate• Numerous very small losses or one larger loss up to 1 minute:

replacement by the mean value of the measurement before and after the interruption

• Losses > 1 minute: division in separated subsections for the inertial sessions

• Acceptable probability of success for the mission• Possibility of accuracy improvement using Hann or Blackman

window.

23

Operational and scientific organization

GEX: Group of ExpertsSPG: Scientific Performance GroupSWG: Science Working group

24

Ground segment architecture

SPG

25

Development of the scientific activity around MICROSCOPE

Scientific objectives:To obtain and validate results, to give the data as much credit as possible

• Phase 1: completing the existing team• Contribution to the present scientific team to reinforce/double the expertise for

the exploitation and validation of the experimental data• Contribution through a phenomenological approach allowing to test new

theories using MICROSCOPE experiment data in relation with the present scientific team

• Exploitation of data provided by the mission for scientific objectives other than the EP test

• Phase 2: critical and independent exploitationAfter validation of experimental data, exploitation and interpretation of MICROSCOPE data, independent of the present team

26

Conclusion

Cooperation and data sharing are necessary to reinforce validation and open up the scientific exploitation

Scientific teams already envisaged will be contacted starting this year MICROSCOPE Colloquium II scheduled 28th and 29th of January, 2013 (TBC):

detailed presentation of the mission and data invitation to scientists to present their concrete interests and competences

The organization of the different operational/scientific groups is formalized and is to be validated by CNES

The schedule for the setting of this community and the activities is proposed till launch mid-2016

The MICROSCOPE SWG is actor in:The exploitation of experimental data, The phenomenological developments and the exploitation of the EP test,The exploitation of data for other scientific exploitations,The encouragement toward the scientific community for the exploitation of the MICROSCOPE

mission