lisa pathfinder and elisa: measuring differential ...• 1 million km michelson ifo with...

LISA Pathfinder and eLISA:

measuring differential acceleration for

gravitational wave astrophysics

William Joseph Weber

Università di Trento / INFN

for the LISA Pathfinder team

Iberian Gravitational Wave Meeting

Barcelona, 13 May 2015

Weber – 5th IGWM – Barcelona - 13 May 2015

Massive black hole merger science with eLISA

Hubble Space Telescope image of UGC1810 – UGC1813 merger in ARP273

• 106 M◉ black holes at core of merging galaxies

• Max GW frequency at mHz detectable only from space

• need to measure relative acceleration at femto-g/Hz1/2 • Free-falling test masses • Measurement of their relative acceleration

Demonstrate with

LISA Pathfinder


• Low frequency performance acceleration noise limited

Not required for detection of MBH but key for science observation

most SNR (100 – 1000) built up in last week before merger (2-5 mHz)

ability to locate source requires weeks – months observation (< 2 mHz)

eLISA Massive Black Hole Science and low

frequency performance

ACCEL NOISE IFO NOISE

Z = 3


GW detection as “differential accelerometry” measurement • Tidal accelerations between 2 or more geodesic reference test masses • Need reference masses in free-fall(*) and measurement of their separation

2

ha L h

Effective “GW acceleration”

2

m e a s u r e na x

distance measurement noise

fo r c e

fa

m

Stray force noise (imperfect geodesic motion)

(*) Force free except for known, calibrated control / suspension forces


eLISA link as a differential accelerometer (2 TM in distant SC)

LPF as a differential accelerometer (2 TM in 1 SC)

LPF similar to 1 axis of a geodesy gravity gradiometer

g1

g2

IFO doppler signal

c

cLtgtgcLththerer

/

2

/

gc

2

106 km

40 cm

GOCE

Difference in stray force per unit mass


• 1 million km Michelson IFO with free-falling TM

• transponder replaces mirrors (100 pW usable light)

• time delay interferometry

time offset phase data, synthesize equal arm IFO

• interferometer TM – TM measurement in 3 parts:

TM-SC SC-SC SC-TM

12 pm/Hz1/2

• Drag-free TM SC follows TM along sensitive axes

3 fm/s2/Hz1/2

eLISA Measurement Concept


1 Million km SC-SC link • Telescope 20 cm diameter • Laser 1064 nm YAG 2 W output • 5-25 MHz doppler beat frequencies • Frequency noise with unequal arm IFO: L 10000 km

Phase measurements shifted in time (TDI) reduce L 10000 km < 5m

• 12 pm/Hz1/2

Local SC – TM measurement • Drag-free control of SC at 3 nm/Hz1/2

• Cross-talk / reference frame: Measure TM – moving in 6 DOF – 6 pm/Hz1/2

• TM force noise per unit mass: 3 fm/s2/Hz1/2

• LISA Pathfinder tests all of this! requires 2nd test mass as reference


da LISA a eLISA/NGO • Single link IFO noise

< 10 pm/√Hz @1 mHz

• Single TM stray acceleration

< 3 fm/(s2√Hz) @ 0.1 mHz

• 3 non-contacting (“drag-free”) satellites

• 3 arms 2 arms

• 5 Mo km 1 Mo km

• eLISA design parameter space: 2-3 arms, armlength

• Spectaculare science in reach even with smaller, cheaper NGO design

• 3rd arm gives 2nd and 3rd IFO combinations

instantaneous polarization, redundancy / debugging


The LISA Pathfinder differential accelerometer

1 spacecraft drag-free control

2 test masses 38 cm separation along x science axis 1TM drag-free reference 2nd TM electrostatically forced 38 cm

x

Each TM inside gravitational reference sensor (GRS) 6 DOF capacitive position sensor + actuator

GRS1

GRS2

LPF tests force noise and local IFO measurement for eLISA


The LISA Pathfinder differential accelerometer

1 spacecraft drag-free control

2 test masses 38 cm separation along x science axis 1TM drag-free reference 2nd TM electrostatically forced 38 cm

x

Each TM inside gravitational reference sensor (GRS) 6 DOF capacitive position sensor + actuator

GRS1

GRS2

Interferometer to measure relative TM acceleration

IFO optical bench

LPF tests force noise and local IFO measurement for eLISA


Demonstrating eLISA differential accelerometer performance with LPF

• Why LPF can improve upon GOCE performance by more than a factor 1000 design, analysis, experiment

• How LPF will demonstrate this


From pico-g/Hz1/2 to sub-femto-g/Hz1/2: LPF design innovations

LEO mm/s2(Terrestrial)

residual atmosphere

L1 nm/s2 (spacecraft g) 0 for eLISA !!

deep space

GOCE LPF

• 2 TM, 40-50 cm baseline gradiometer in a drag-free spacecraft • pm/Hz1/2 displacment sensitivity

300 gm TM 100 mm gaps

discharge wire

2 kg TM 3-4 mm gaps no contacts

Actuation force noise «accelerometer range»

GRS surface force noise

tough caging UV discharge

need IFO


eLISA / LISA Pathfinder gravitational reference sensor (GRS)

VACT1

VACT2

VM

VAC

100 kHz L

L • 46 mm cubic Au / Pt test mass (2 kg)

• 6 DOF capacitive “gap” sensor w/ AC readout

• Audio freq electrostatic force actuation avoid DC voltages

• Large gaps (3 – 4 mm) limit electrostatic disturbances

• High thermal conductivity metal / sapphire

limit thermal gradients

• High vacuum (p = 2 mPa) limit Brownian, radiometric effects

• UV photoelectric discharge

• Caging mechanism for launch lock

• Defines TM environment

• nm/Hz1/2 measurement on all axes

• 10 pm/Hz1/2 interferometer used on x axis

• nN actuation forces (mN non-science)

• Flight heritage for eLISA


LPF IFO displacement metrology < 9 pm/Hz1/2

• First high-precision interferometer coupled to free-falling TM in space • Demonstration of eLISA local position measurement • Zerodur optical bench with 4 heterodyne interferometers

o12 (x2 – x1) TM - TM +1, h1

o1 (x1 – xSC) TM - SC

+, h

+ frequency and phase stabilization IFO


LISA Pathfinder as a differential accelerometer

Control

TM1 TM2

o12

o1 g1 g2

Drag-free: thrust SC to follow TM1 (null o1, 1 Hz BW)

Electrostatic suspension: force TM2 to follow TM1 (null o12, 1 mHz BW)

m

F

m

Fg

12

We want to measure differential force per unit mass:



TM1 TM2

o12

o1

ESSCp

SCp

FxxmFxm

xxmFxm

2

2

222

1

2

111

SCxxo

xxo

11

1212

Newton’s Eqns: IFO Readouts :

Commanded control forces

Elastic coupling

12

2

21

2

1

2

212ˆ oo

m

Fog

ppp

ES

IFO Acceleration

g1 g2

• Produce differential acceleration time series

• Spacecraft coupling term (stiffness) subtracted (also for LISA)

m

F

m

Fg

12

«gravitational observable» differential force per unit mass

g estimator:



TM1 TM2

o12

o1

ESSCp

SCp

FxxmFxm

xxmFxm

2

2

222

1

2

111

SCxxo

xxo

11

1212

Newton’s Eqns: IFO Readouts :

IFOggg ˆ

g1 g2

• differential acceleration time series

• Spacecraft coupling term (stiffness) subtracted (also for LISA)

12

2

21

2

1

2

212oo

m

Fog

ppp

ES

g

1

12

n

n

12

2

21

2

1

2

212nnn

ppp

IFO Noise

IFOg

m

F

m

Fg

12

«gravitational observable» differential force per unit mass


LISA Pathfinder control, displacement and acceleration

TM1 TM2

o12

o1

«working in acceleration» • use dynamic parameters but not transfer function • remove long transients

Free particle «measure acceleration»

Accelerometer «measure control force»

12

2

21

2

1

2

212ˆ oo

m

Fog

ppp

ES


xxxm

Fxg

pSC

ES

2

21

2

IFOp

Cggoo

m

Fo

12

2

21

2

12

Calibrating the LPF differential accelerometer

• Modulate satellite drag-free and electrostatic setpoints

• Fit acceleration to commanded force (FC), measured positions o12, o1

+ other measured disturbances • Background force (g) and IFO noise

o1 o12

Actuator calibration

factor dynamical stiffness

Background force + IFO noise


Surface

Bulk

3D sensing + actuation

Torsion pendulum small force

measurements (lightweight TM)

LISA Pathfinder flight test

LPF Instrument limit

IFO Actuation noise

LPF performance and TM acceleration noise sources for eLISA


Torsion pendulum upper limit

eLISA Spec

LPF Spec

Tungsten fiber (Q = 3000)

Fused Silica fiber (Q = 800000)

GRS surface force noise: overall upper limits with torsion pendulum

• Attribute ALL unexplained torque noise to interaction between GRS – TM

• Convert to equivalent acceleration surface force noise upper limit

• LISA-like TM suspended inside prototype GRS with FEE

• Within factor 2 of LPF spec at 1 mHz • Many known forces measured to eLISA goal


LPF instrument limit: noisy compensation of spacecraft self-gravity

2/1

/

2/12 2

VVaACTSgSVF

«accelerometer dynamic range» problem Noise in “DC” force applied to compensate local g Not present in eLISA!!

g1 g2

Design specs

Fx

N

• Expect < 8 fm/s2/Hz1/2 at 1 mHz • full analysis complicated by uncorrelated voltage fluctuations, act

Current measurements 2-3 worse

Current spacecraft mass balance calculations look better!

1/26-2/1

/

2

/Hz102

nm/s 65.0

VVS

g


Opt bench + GRS ~ -50 nm/s2

(TM – TM) ~ -2 nm/s2

Spec 0.65 nm/s2

Model ~ 10 nm/s2 level

compensate by GRS internal balancing masses

(IBM) 1.8 kg W

Gravitational balancing to g < 1 nm/s2

• No full ground test: analysis (1/r2) + mass + position measurements

• In flight: measure 3 trans + 6 rot gravitational balances + measure gravitational stiffnesses


LISA Pathfinder performance: with and without x-actuation

LPF best est

LPF spec

• Expect to beat LPF spec at all frequencies • If gravitational balancing at spec level, likely limited by actuation that

is not present in eLISA


LISA Pathfinder: avoiding actuation fluctuations with free-fall mode

compensate average DC force imbalance by applying a large impulse followed by free-fall (parabolic flight!)

x

• Example: Apply 100 x average needed force for 2.5 s, followed by 247.5 s free-fall • Analyze only free-fall data, avoid impulses

window to avoid gaps (windowed PSD) Iteratively extract noise to fill gaps (standard PSD) Low pass / decimate / detrend, set gap data to 0 (standard PSD)

• Can we recover intrinsic force noise at 10-3 -- 10-4 Hz in a series of 250 s experiments?

Grynagier, CQG 26 (2009) 094007

10 mm


Testing free-fall mode on the ground • Can we recover intrinsic force noise at 10-3 -- 10-4 Hz in a series of 250 s experiments?

• Data with gaps (spectral leakage) • Huge increase in dynamic range non-linearity, timing issues

• What are the conditions in which free-fall / impulse mode can be used?

• Force levels, displacement range, time of flight

On-ground torsion pendulum test: • Rotated torsion pendulum forced to center • Flexibility to choose:

• DC force (pendulum rotation) • Elastic coupling (applied voltages) • Flight and impulse times


LPF best est eLISA TM accel noise spec

LISA Pathfinder performance prediction

LISA Pathfinder should guarantee most of LISA science

(performance needed to observe calibration binaries with SNR=1 in 1 year)


LPF GRS

Infinite gap model

Factor 13 in noise power

Noise source: Residual gas impacts Excess damping + Brownian noise in constrained volume

F

x

| |F

PRL, 103 140601 (2009) , Phys Lett A, 374 3365 (2010)

Single TM gas damping noise 4/1

0

2/1

1/222/1

30Pa 2Hz/fm/s 8.1

mpS

a

m • vent to space • could be largest stray force noise • no direct pressure measurement (but indirect through T effects)


LPF: a physical model of limits to measuring differential acceleration In – orbit and on-ground

Example: electrostatics inside GRS

gap

11

x

C

CqEqF

x

TOT

xxx

• stray electrostatic fields (patch effects) • cosmic ray charging

Largest interaction: TM charge + average residual E field [PRL 108:181101 (2012)]

)(),(

1

SjTMi

ji

ij

x

xVV

x

C

x

C

q

V

x : equiv. single electrode potential

UV discharge of TM Charge noise requires compensation of average stray field Operation with non-zero q requires low fluctuations in stray field


• Torque noise with charged TM • Single electrode (500 mm2) < 12 mV/Hz1/2

• OK for eLISA (marginally)

Measure /compensate dF/dq (dN/dq)

• Measure / compensate 100 mV < 1 mV

• Need to measure dF/dq by varying charge (not with modulated VTM)

Uncompensated

Compensated

dN/dVTM (modulate VTM)

dN/dq (charge bursts)

Stray electrostatics (on ground):

mHz 1at V/Hz 801/22/1

m

x

S

Measure noise in dF/dq (dN/dq)

Tests of UV discharge, charge measurement


Measure compensate dF/dq with UV TM charge bursts and modulated voltages

• Performance and debugging

• Several times and for two TM

Measure long-term TM charge fluctuations

• How well / how often do we need to balance field?

• Confirm charging models, low freq noise

• Correlation with radiation monitor

Full UV discharge testing

• Fast discharge and continuous

If necessary ...

• force noise with charged TM

• effect of single (bad) electrode

x = 27.4 +/- 0.5 mV (c2 =5.4 , 17 DOF)

LPF simulator

Stray electrostatics (In-flight):


Measurement science investigations with LPF

• Measure spacecraft g (3 translations and 6 rotations) • Force noise from actuation

• invert test mass roles, apply common mode force, • free-fall mode

• Measure temperature fluctuations down to 10-4 Hz – and below – and dF/dT • Measure magnetic environment and forces from applied B fields • Measurement and compensation of average stray E field (nulling dF/dq) • Measure low frequency charge fluctuations with long term charge test • Stiffness / SC coupling / crosstalk measurements

Create physical model for limits to achieving perfect free-fall

Talks: Ferran Gibert, Miquel Nofrarias


LISA Technology Package: Payload Integration • Integration of the LCA payload

completed (end April 2015)

• Currently at IABG (Munich) for integration with satellite

• On schedule for launch in early October 2015


LTP payload: IFO integration and testing • On-station thermal testing • integrated OBI with support structure • Fixed mirrors in place of TM • Success! 3 pm/Hz1/2 at 10 mHz


LTP payload: GRS integration and testing

Test-mass

Electrode Housing UV discharge fibers

(3 per TM)

TM caging (top + bottom)

Vent valve Gravitational balance mass (not shown)

Release mechanism (top + bottom)


LTP payload: TM / EH integration and testing

TM fixed (measured)

EH assembled around TM on hexapod

(aligned and measured)

• final cabling, electronics, part of vac chamber • EH moved around TM in 6 DOF • TM – EH position measured by CMM and by GRS electronics • Test of calibration, noise, alignment

alignment data used to align IFO


LTP payload: TM – EH calibration and noise testing

Sensing noise measure 2-3 nm/Hz1/2

sensing noise floor

sensor calibration All critical alignments of GRS

within 10 micron / 1 mrad


LTP payload: GRS vacuum and discharge test • UV discharge sensitive to surface

properties, cleanliness, bakeout

• Performed test of UV discharge after final bakeout with femto-amp electrometer measurement

• Next GRS venting will be in space!

• Observe UV emission from both TM and EH on both GRS (bipolar)

• Outgassing within GRS OK!


LISA Technology Package: GRS - IFO Integration

GRS vent valve closed evacuated TM caged No joint GRS / IFO operations possible before flight caging not designed to align TM to IFO

However, during integration IFO laser / metrology turned on allowed IFO contrast (IFO works!)

better than design specs


LPF Launch and Operations

• Launch with VEGA (5-6 th launch) into LEO from Kourou

• Propulsion module for orbit raising • One month cruise to L1

• Commissioning 2 months after launch

Decaging Gravitational balancing verification!

• 6 months science phase + possible extension


LPF Science Phase

Measure noise in g. And then explain it.

Dense schedule: noise runs interleaved with calibrations and tests on specific effects

Example timeline

• All procedures ready (and simulated) command sent by MOC (Mission Operations Center, ESOC, Darmstadt)

• Real time «scientific» data analysis (LTPDA) • On day N can change timeline for day N+3

STOC (Science Team Operations Center) ESOC/ESAC + remote


Timeline executed as part of 4 day SOVT «hardware in loop» simulation (February 2015) Test of ESOC ground segment mission operations Also test of STOC – analysis at ESOC and remote (50 people) [Real mission timeline won’t be so dense!]


STOC and MOC at work during SOVT...

... and dreaming of launch on October 2!!!

Weber – 5th IGWM – Barcelona - 13 May 2015 44

Thank you … and the LPF Team:

lisa pathfinder and elisa: measuring differential ...• 1 million km michelson ifo with...

Documents