Transcript SR_POEM - www.cfa.harvard.edu/

A 15 Minute WEP Test
(SR-POEM)
R.D. Reasenberg, E. Hirose, B. Patla,
J.D. Phillips, E.M. Popescu
Smithsonian Astrophysical Observatory
Harvard-Smithsonian Center for Astrophysics
Cambridge, Massachusetts, USA
and
E.C. Lorenzini
Faculty of Engineering
University of Padova, Italy
Q2C4
9/20/09
September 2009
Reasenberg et al., SAO
Bremen
Slide 1 of 33
Mission Concept
• For a single pair of substances, σ(η) ≤ 10-16.
– 1000 fold advance over present best result.
• WEP test in sounding rocket payload.
– Experiment duration 400 to 800 s.
– Payload ≈ 200 kg.
• Non-recoverable payload (like orbiting payload).
• Low cost (not like orbiting payload).
9/20/09
Reasenberg et al., SAO
Slide 2 of 33
Instrument Concept
• Derived from POEM (which was
derived from JILA test).
• 2 test mass assemblies (TMA)
observed by 4 tracking frequency
laser gauges (TFG).
• Double difference observable:
– Observations made from co-moving
reference plate.
– Difference of observations yields
quantity of prime interest.
9/20/09
Reasenberg et al., SAO
Slide 3 of 33
Experiment Concept
• TMA (about 1.0 kg) in free fall for 40 s per drop.
• Experiment includes 8 drops.
• Payload inversion between drops.
Inversion cancels most remaining systematic errors:
In payload frame, gravity from local mass is fixed
and WEP signal reversed by inversion.
Earth’s gravity gradient is symmetric and thus the
same after inversion (except for higher order term
which is too small to matter.)
9/20/09
Reasenberg et al., SAO
Slide 4 of 33
Experiment Housing
Precision instrument
inside vacuum chamber
inside 14 inch payload
tube. Not shown here
are the two vacuum ports
at the upper end of the
chamber, the capacitance gauge electrode
sets, and the TFG optics.
9/20/09
Reasenberg et al., SAO
Slide 5 of 33
Tracking Frequency Laser Gauge (TFG)
• Developed around 1990 for POINTS.
– Recent further development.
– Now being developed under NASA-APRA.
– See Phillips & Reasenberg, RSI, 76, 064501, 2005.
• Based on Pound-Drever-Hall locking
– Converts distance change to frequency change
• Easily and reliably measured.
– Hops to new fringe when it runs out of range.
• Has advantages over traditional heterodyne laser gauge.
(5 applicable here)
9/20/09
Reasenberg et al., SAO
Slide 6 of 33
TFG: Six Advantages
1. Intrinsically free of the nm-scale cyclic bias characteristic of
the heterodyne laser gauge (e.g., Hewlett-Packard, Zygo).
2. Uses one beam, not two: simplifies the beam launcher.
3. Distance changes converted to Δ(radio frequency): more
stable and more easily measured than RF phase.
4. Able to operate in a resonant cavity: improves precision,
suppresses misalignment error and supports servo-based
alignment.
5. Suppresses polarization error from non-normal incidence in a
cornercube (nm scale).
6. Measures absolute distance with a minimum of added cost or
complexity.
9/20/09
Reasenberg et al., SAO
Slide 7 of 33
Laser Gauge Alignment
9/20/09
Reasenberg et al., SAO
Slide 8 of 33
Wavefront Sensing
Sampas-Anderson
(Morrison, Meers, Robertson, Ward)
θ
pick-off
E0
E0
SB1
mirror1
mirror1
CR
SB2-
CR
SB1
θ
SB2+
mirror2
QPD1
fold
mirror
mirror2
QPD
QPD2
1) Only Carrier resonates with TEM00.
2) Reflected light is used.
1) Carrier resonates with TEM00,
and SB2+ resonates with TEM10.
2) Transmitted light is used.
3) Used by LIGO.
9/20/09
Reasenberg et al., SAO
Slide 9 of 33
Comparison of Methods
Aspect
Wavefront
Sensing
SampasAnderson
Detection port
Reflected
Transmitted
Sidebands
1 (share with locking)
2 (need extra SB)
Photodetectors
2 (tilt & lateral shift)
1 (uses quadrature
detection)
Mirrors / cavity
2
3
Gouy phase setting
Yes (long path or extra
optics)
No
Electronics
Low frequency
RF, more complex
9/20/09
Reasenberg et al., SAO
Slide 10 of 33
TMA Suspension System
• Can observe and control 6 degrees of freedom.
– Capacitance gauge sensing.
– Electrostatic forcing.
– All active during setup and inversion; off during WEP
measurement.
• Coriolis acceleration: measure E-W velocity.
– Transverse position measurements made before and after WEP
measurements.
9/20/09
Reasenberg et al., SAO
Slide 11 of 33
Capacitance Gauge Plates
The capacitance gauge is the
sensing portion of the TMA-SS
Dimensions are in cm.
Plan of capacitance gauge electrodes.
9/20/09
Interelectrode gaps are 0.25 cm.
The long bars are drive electrodes
and the large rectangle is the
sense electrode. Not shown,
grounded shield around the sense
electrodes.
These electrodes are deposited on glass
plates (ULE or Pyrex) that are later joined
to form a box. Any significant open areas
are filled with grounded shields, and the
small remaining gaps are covered with
resistive material, which leaks off charge.
A ground plane on the back of each plate
covers the drive electrodes but not the
sense electrode (because it would add
capacitance and thus decrease
sensitivity.)
Reasenberg et al., SAO
Slide 12 of 33
CG Electrode Alternatives
• Metal plates, insulated from and attached to, a stable
conductive housing.
– Insulators are well-hidden behind metal electrodes.
– Facilitates attachment of leads.
TMA
9/20/09
Reasenberg et al., SAO
Slide 13 of 33
POEM Capacitance Gauges
Collaboration with Winfield Hill, Rowland Institute at Harvard
Vacuum
TMA
+
-
Cal.
9/20/09
ADC
Correlator
24 bit
100 kHz
s/w in PC
f1, f2, …, f6
+
-
~ f
1
Estimates of 6 positions
per TMA, at 1 kHz
Moving
Reasenberg et al., SAO
Static
Slide 14 of 33
Possible Capacitance Gauge for SR-POEM
Two of six position measurement circuits shown
f2
Vacuum Chamber
+
ADC
TMA
f1
+
9/20/09
Correlator
f2
Reasenberg et al., SAO
ADC
Correlator
f1
Slide 15 of 33
Charge on TMA
• TMA potential at separation  few × 100 mV.
• Measure by applying DC field to capacitance gauge
electrodes and neutralize, e.g., with UV LED.
Sun, et al. LISA-LIGO Charging Workshop, 2007.
– Before reaching altitude at which drag is low (800 km).
– Make TMA voltage  rms variation over surface.
• Effect of small constant charge on TMA cancels in payload
inversions.
9/20/09
Reasenberg et al., SAO
Slide 16 of 33
Surface Potential Variation
• Classical solution: gold, graphite (Aerodag).
• Recent LISA work by Robertson et al.
– Class. Quantum Grav. 23 (2006) 2665-2680
– New materials studied: Au over Nb, diamond-like carbon, TiC,
indium tin oxide (ITO), Au over ITO.
– Many achieve 1 to 2 mV rms wrt mean.
• Measurements done with 3 mm Kelvin probe.
– Needs to be smaller and more sensitive.
– They, GSFC and PNNL are investigating improvements.
9/20/09
Reasenberg et al., SAO
Slide 17 of 33
Surface Potential Variation for SR-POEM
• Assume: 1 mV rms, 4 mm spacing (top and bottom)
=>2×1.7×10-16 g.
• Measured time variation of surface potential at 1/140 Hz,
averaging over 4 cycles:
δa=1.2×10-17 g.
– Very conservative estimate.
• Good enough, and further …
– Temperature in SR-POEM 100- to 1000-fold more stable (below).
=> SR-POEM has stable vacuum environment.
• Need additional testing under SR-POEM conditions.
– [Does voltage at inversion cause change of potential?]
9/20/09
Reasenberg et al., SAO
Slide 18 of 33
Caging (Uncaging!)
• Synergy with LISA.
• Clean metals tend to cold weld.
– Welds can reach a large fraction of strength of the metal.
• Candidate design concepts:
– Non-stick materials with possible separate ground point.
• Graphite gas bearing to push off.
• R-S-H compounds (long chain thiol or mercaptan).
• S → Se ?
– Contact at bottom of hole to hide the surface potential of contact
area.
9/20/09
Reasenberg et al., SAO
Slide 19 of 33
Thermal Stability
• Two concerns:
– Direct effect on apparent double-difference acceleration
measured with TFG (thus on η).
– Indirect by moving payload mass.
• Direct effect made small by:
– Use of ULE glass for precision structure.
– Layered passive thermal control.
– Symmetry of thermal leaks.
Thus far, unable to find a problem.
9/20/09
Reasenberg et al., SAO
Slide 20 of 33
Thermal Time Constants
• The precision instrument hardly sees the external
temperature changes.
• Vacuum chamber gold coated inside and out.
– Emissivity, ε = 0.02.
• Payload tube (ε = 0.1) to chamber,
τ = 1.5 x 105 s.
• Chamber to metering structure (ε = 1),
τ = 1.4 x 105 s.
• Chamber to TFG plate (ε = 1),
τ = 5.5 x 105 s.
9/20/09
Reasenberg et al., SAO
Slide 21 of 33
Thermal leaks
Vacuum flange to TMA plate,
radiative: 6 mW/K
Support ring to TMA plate,
conductive: 56 mW/K
Vacuum flange to support
ring, conductive: similar
Vacuum flange to TMA plate,
total: 34 mW/K (τ=105 s)
Cables?
9/20/09
Reasenberg et al., SAO
Slide 22 of 33
Earth’s Gravity Gradient, I
at 1000 km altitude
• For a TMA 1 mm (z) from payload CM plane, it moves
1.6 μm in 40 s. (2 x 10-8 ms-2 = 2.7 x 10-9gh)
(We are investigating how closely the CM can be placed to the
expected plane of the TMA.)
• For a TMA 0.1 mm radially (ρ) from payload cm, it moves
(radially) 80 nm in 40 s. (1 x 10-10 ms-2)
• These motions are very predictable and change slowly
with payload trajectory.
• Symmetry: rising vs falling part of trajectory.
– Nearly perfect!
9/20/09
Reasenberg et al., SAO
Slide 23 of 33
Earth’s Gravity Gradient, II
at 1000 km altitude
• Objective: matched conditions in pairs of drops.
– Each pair has instrument in both orientations.
• Use data with same altitude range from each drop.
– Symmetrically placed with respect to apogee.
• A small term remains.
– 3 x 10-11 m (per micron of TMA relative centering error)
– Gravity gradient = 2.7 x 10-7 g/m (Earth only, 1000 km alt.)
– Acceleration error < 8 x 10-18 g (per micron …error)
9/20/09
Reasenberg et al., SAO
Slide 24 of 33
Systematic Error Estimation
(major contributions)
9/20/09
Class of Error
Sources
Position Measurement
Optics/alignment
Gravitational Effects
Local masses, Earth
Magnetic Effects
Permanent & induced moments
Electrostatic Effects
Charge on TMA, Patch effect
Gas
Out-gassing, Radiometer effect
Kinematics
Coriolis
Reasenberg et al., SAO
Slide 25 of 33
♦ Optics
Notes
Error (g)
1. Misalignment of measuring beam
1.0×10-19
2. Transverse velocity w. curvature of V=25 nm/s ; r=150m
flat mirror
1.0×10-22
4. Reference laser wavelength drift
Δν/ν=10-9 (quad, 40 sec) ;
ΔL=10-4m
1.5×10-17
1% gravity Model error
4.2×10-17
♦ Local Gravity
TMA/shelf 4 mm,
Δ(spacing) = 300 nm
9/20/09
Reasenberg et al., SAO
Slide 26 of 33
♦ Permanent Magnetic Moment
Notes
Error (g)
Shielding factor =103 ,
Magnetic moment of TMA =10-7 Am2
►Need for higher shielding factor, low
magnetic moment (LISA: 10-8 Am2).
► Does not cancel after inversion.
Mumetal: 5 kg
(inside vacuum chamber)
7.3×10-15
1. 1 mV on TMA w/ position offset 0.1mm
ΔV < 0.1mV, DC/surface
< 5.1×10-18
2. TMA off-center with C-gauge electrodes
~10-5 radian, 0.1mm offset
< 10-25
3. Surface Potential variation Vrms=1mV
ΔVrms=0.03 mV
3.0×10-17
♦ Electrostatic Effects
9/20/09
Reasenberg et al., SAO
Slide 27 of 33
♦ Gas
Notes
1. Radiometer effect
δT(Top-bottom) = mK
< 2.0×10-19
2. Out-gassing by TMA
10-8 Torr, pump rate = 10-2 m3/s,
top-bottom Δ rate diff = 0.1%
< 1.4×10-19
Error (g)
♦ Kinematics
Coriolis (inertial pointing)
δΩrms <3×10-8/s, v = 10-9 m/s
<10-17
g = is the acceleration due to gravity at 1000 km
Δ →change after inversion
9/20/09
Reasenberg et al., SAO
Slide 28 of 33
Local Gravity
Stabilization, I
Earth thermal radiation:
≈10-10 g => 0.6 μm in 40 s
Pseudo-drag-free instrument.
Useful because local masses
produce a much more complex
gravity field than the rest of the
spacecraft.
Alternative: Map the local
gravity field by making
measurements on the way up to
WEP-measurement altitude. Or
a real drag-free system.
Needs considerable study.
9/20/09
Reasenberg et al., SAO
Slide 29 of 33
Local Gravity Stabilization, II
• Earth gravity daz/dz = 2.7 x 10-7 g/m (well known)
• Local gravity daz/dz ≈ 10-8 g/m (poorly known)
– Require δz < 3 nm.
– Non-grav. force & ACS noise => ~1 μm.
• Servo to keep payload fixed w.r.t.
combined TMA.
– Control (zCM1+zCM2), θa, θb.
• TFG is sensor.
• Hexapod is actuator.
– Need initial ΔvCM<10 nm/s.
• Use local gravity model to correct for actual positions.
• In early stage of investigation.
9/20/09
Reasenberg et al., SAO
Slide 30 of 33
Why Does SR-POEM Work?
• TFG supports quick measurements.
– 0.1 pm/√Hz – requires development.
• Double difference observable.
– Symmetry maintained.
– Local gravity stabilization (if used).
• Payload inversions.
– Cancel systematic errors.
• Thermally benign environment.
9/20/09
Reasenberg et al., SAO
Slide 31 of 33
Concluding Comments
• Goal: σ(η) ≤ 10-16 for single pair of substances.
• Sounding rocket experiment is low in cost.
– Additional flights could test other substances.
If the sounding rocket had launched at the start
of this talk, the experiment would be over now!
9/20/09
Reasenberg et al., SAO
Slide 32 of 33
Two new post docs started this month:
Rajesh Thapa & Emanuele Rocco
http://www.cfa.harvard.edu/PAG
Papers and sounding-rocket proposal available.
[email protected]
617-495-7108
[email protected]
617-495-7360
This work has been supported by NASA-UG through grant NNC04GB30G.
It is now supported by NASA-ATFP through grant NNX08AO04G
9/20/09
Reasenberg et al., SAO
Slide 33 of 33