Transcript Telescope Errors NGAO - Caltech Optical Observatories

Keck Interferometer Support Trade Study
Christopher Neyman
Junichi Meguro
W. M. Keck Observatory
Viswa Velur
Caltech
Keck NGAO Team Meeting #5
March 7, 2007
Video Conference: WMKO,CIT,UC
Keck Interferometer Support Trade Study
•
3.1.2.1.5 Keck Interferometer Trade Study
– Consider the relative performance, cost, risk, & schedule of feeding KI
with NGAO or a repackaged version of the current AO system
– Decoupling of NGAO from interferometer support may simplify &
improve performance of NGAO
– The feasibility of maintaining a version of the two current AO systems
for KI use should be evaluated
– Complete when NGAO baseline architecture selected
2
Overview Slide/Outline
• Introduction
• Concepts for Interferometer support
– NGAO feeding interferometer
– Dual IF AO Systems
– Swapping Keck (I/II) AO with NGAO
• Cost of construction and operations for above concepts
• Conclusions and recommendations
3
Introduction
Why Are We Doing This?
• Current Keck Interferometer
– Visibility
• First fringes on AGN
– Nuller
• MRI (LGS → faint objects)
–
–
–
–
Planet detection
Young stellar objects
AGN
Galactic center: tests of general relativity
In general astrometry requires long time span (1 century)
Relatively short for galactic center (years)
Current project started in ~1993 with speckle, continuing with KII LGS
PI retires in ~2030
• Important not to over constrain NGAO design
4
NGAO Feeding Interferometer
• Legacy AO (arm A) and NGAO (arm B)
• Need to match polarization, rotation, fringe symmetry (Nuller only)
• Viswa Velur in consultation with Kent Wallace (JPL) looked at
polarization matching:
– Martin-Puplett polarization modulator
• Preferred over wave plates
• Basis of current KI nuller combiner
– “In general the polarization issue with interferometry is an intractable
problem”
• Use close to normal incidence (K mirrors problematic)
• Coating information is incorrect, proprietary, or unavailable
• Geometry match imperfect, tight tolerances, realignments frequent
– Still need to determine polarization shift for NGAO relay
Indiana Wells
New designs as available from Brain Bauman
5
Differential Phase Shifts Can Reduce Fringe Contrast:
a Real World Example
Both K1 &
K2 recoated
After recoating only
K1 AO image rotator
significant drop in V2
V2
Fall 2003
returned to nominal
values when K2 was
also recoated
Suspect phase shift
caused by different
protective layers on
top of silver coating
Only K1 recoated
6
Interferometer Only AO Systems
• Build two identical AO systems
– ESO MACAO (One for each 8-m)
• Options available
– Adaptive secondary
• High throughput system
• Expensive
• Facility investment: could be used for other AO systems and instruments
– Basement AO systems
• Some space in coude crypt
• 589nm transmission through basement windows (Silica yes, ZnS no)
– Mount AO system under current AO bench
• Room available
• Could use MEMs and other technologies to reduce size/costs
7
Interferometer Only AO Systems
AO enclosure
IF beam tubes
Primary mirror
cell
8
Interferometer Only AO Systems
AO enclosure
NGAO
IF AO system
IF beam tubes
Primary mirror
cell
9
Interferometer Only AO Systems
IF beam tubes
Cables
Cooling lines
10
Interferometer Only AO Systems
Need to move AO focus below Nasmyth deck
– Use large f/# secondary
• Have f/25 for both telescopes (Keck II in storage)
• Possibly requires more secondary exchanges
– Use 1:1 relay (Offner or others)
• Extra reflections (3-5 total for Offner)
– Move secondary toward primary
• 1 cm at secondary, 74 cm at Cassegrain focus
• Large spherical aberration (much larger than typical DM stroke)
– MEMs small foot print AO
• Mount in current enclosure near focus
• Relay light out
• Need ~2 arc minute field for LGS tip/tilt, Interferometer field small
– Adapting Bauman’s off-axis ellipse relay should address these issues
– See NFIRAOS petite
11
AO for Interferometry at ESO: 4 Identical AO Systems that
Only Feed the Interferometer
• 4 identical AO systems located below
each telescope
• Coude mirror m8 is AO bimorph
• ESO built 6 AO units in total
– 4 for IF
– 1 for IFU spec.
– 1 for IR echelle
• Possible cost savings
12
Swapping K(I/II) AO with NGAO
• Need to repackage current AO
–
–
–
–
Optical/mechanical bench
Electronics
Cooling lines
If NGAO is cooled, need to maintain NGAO temperature/seal
• Method to swap AO
– Cart: roll AO on/off deck
– Crane: lift AO on/off deck
• Historically, when instruments are damaged, it has occurred during
instrument exchanges
• Currently move DEMIOS in and out of beam
• Originally proposed for current Keck I AO system and rejected
13
Cost of IF Support Options
• Cost estimate (WAG level)
– AO secondary
• $7.6 M: $15.2 M (2 systems)
• source GEMIN GLAO study
– Conventional AO
• $2.2 M: $4.5M (2 systems)
– MEMS AO
• $1.7M: $3.5M (2 systems)
– Swapping NGAO with legacy AO
• No estimate done yet
14
Preliminary Recommendations
• Swapping Keck I/II with NGAO
– Hardware costs relatively low
– Cost of moving NGAO is higher than fixed NGAO
– Appears risky in operation
• Matching NGAO to Keck
– Challenging to match polarization states
• Two AO systems
– AO secondary on each telescope
• Elegant solution, but costly
• Utility of a shared capability could offset higher costs
– MEMs AO for each IF arm
• Relatively inexpensive
• Small footprint allows more mounting/packaging options
15