Transcript The secondary and tertiary mirros for the TMT

Thirty Meter Telescope
Secondary and Tertiary Mirror Systems
International Symposium on Photoelectronic
Detection and Imaging 2011
International Colloquium on Thirty Meter
Telescope
Virginia Ford
25 May 2011
TMT.OPT.PRE.11.046.DRF01
1
M2 and M3 Systems
Secondary Mirror
(M2) System
Tertiary Mirror
(M3) System
Primary Mirror (M1) System
TMT.OPT.PRE.11.046.DRF01
2
M2 and M3 Systems
M2 Positioner and
Control Electronics
M3 Cell
Assembly
M2 Cell
Assembly
M2 System
M3 Positioner and
Control Electronics
TMT.OPT.PRE.11.046.DRF01
M3 System
3
M2 and M3 Cell Assemblies: Passive
Passive support systems – non-actuated optical surfaces
M2 and M3 optical surface errors will be corrected by M1 segments
– Low-spatial frequency errors are correctible by M1
Residual fitting errors are within error budget
– Image blur from beam footprint motion on M3 is acceptable
– Off-axis image shear errors are acceptable
EXAMPLE: M2 GRAVITY PRINT-THROUGH PATTERN CORRECTED BY M1 SEGMENTS
UNCORRECTED
M2
SURFACE ERROR
CORRECTION
UNCORRECTED M2
SURFACE
CORRECTED WITH M1
SEGMENTS: PISTON,
TIP AND TILT (PTT)
nm
RMS
120
11
CORRECTED WITH M1
0.8
SEGMENTS: WARPING
HARNESS AND PTT
TMT.OPT.PRE.11.046.DRF01
Surface Peak to Valley = 530 nm
M1 SEGMENT
PISTON, TIP, TILT ONLY
4
Surface Peak to Valley = 54 nm
M2 and M3 Mirror Substrates
Substrate requirements:
– Low-speed movements → ultra light weight substrate not required
– High thermal stability required
Thermally-induced surface distortions should cause spatially smooth
errors
– Resistant to chemicals used to clean and strip the coating
Substrate selected:
– Very-low expansion glass or glass ceramic
– Smoothly varying, low CTE
– Resistant to cleaning and stripping chemicals
– Meniscus-style substrate form – gravity print-through creates low
spatial frequency, smooth, correctable surface errors
– Existing technology – no development, low-cost, low-risk
TMT.OPT.PRE.11.046.DRF01
TMT.OPT.TEC.11.067.DRF01
5
M2 and M3 Mirror Polishing Surface Error
Requirements
Surface error requirements are controlled as a function
of spatial frequency
– Low-spatial-frequency allowances are generous to ease
polishing and metrology
Well correctible by M1 segments
– Mid-spatial-frequency allowances have tighter limits
Partially correctible by M1 segments
Ability to correct decreases as spatial frequency increases
– High-spatial-frequency allowances are tightly controlled
Not correctible by M1 segments or the Adaptive Optics (AO) system
TMT.OPT.PRE.11.046.DRF01
6
M2 and M3 Mirror Polishing State-of-the-Art
TMT M2 and M3 mirrors are larger than existing equivalent
mirrors that have been produced
– Recent advances in polishing techniques and metrology enable their
fabrication
M2 State of the art:
– MMT convex 1.7 m diameter secondary mirror:
f/1.3; asphericity 330 μm; radius of curvature 5151mm; conic constant 2.7
– TMT convex 3.0 m diameter secondary mirror:
f/1.0; asphericity 885 μm; radius of curvature 6228mm; conic constant 1.3
M3 State of the art:
– 2 m diameter flat (University of Arizona – reported SPIE 2007)
Metrology: subaperture Fizeau interferometry and scanning pentaprism
resulting in 3nm rms measurement uncertainty
– TMT: 3.5m x 2.5m ellipse
TMT.OPT.PRE.11.046.DRF01
7
Secondary Mirror Polishing Challenges
Since M2 is convex, it cannot be tested from its centerof-curvature
– Must fabricate very large test optics or use smaller test optics
with subaperture stitching
M2 is highly aspheric
– Polishing tool fitting challenges
Especially at the edge of the mirror, the polishing tool size must be
small
The tight requirements on high spatial frequency optical surface
errors provides challenges to small tool polishing
TMT.OPT.PRE.11.046.DRF01
8
M2 Mirror Polishing Challenges
Asphericity causes tool fit issues:
– M2 optical surface is extremely non-spherical
(for tool misfit < 1 μm)
large spherical tool
small spherical tool
aspherical surface
TMT.OPT.PRE.11.046.DRF01
9
M3 Polishing Challenges
General issues with large flats:
– Edge roll-off
Created by area of polishing tool that is unsupported at edge of part
Elliptical shape issues
– Polish mirror as roundel then cut ellipse
Must remove stress relief springing errors caused by cutting
or
– Cut ellipse then polish
Asymmetrical forces during polishing cause surface errors
3.594 m
TMT.OPT.PRE.11.046.DRF01
2.536 m
10
M2 and M3 Metrology Challenges
Full aperture acceptance testing requires impractical
large test optics or subaperture stitching
– M2 possibilities for acceptance test:
Fizeau Interferometry
Hindle Sphere or Hindle Shell or Aspheric test plate system
– M3 possibilities for acceptance test
Ritchey-Common test
Fizeau Interferometry
TMT.OPT.PRE.11.046.DRF01
11
M2 and M3 Mirror Cell Assemblies
Axial Support Systems
Have explored many axial support designs including
from 9 to 60-point active and passive
– 18-point passive whiffletree supports are sufficient provided
some axial print-through can be polished out
– 60 point active support: better than required and costly
TMT.OPT.PRE.11.046.DRF01
12
M2 and M3 Mirror Cell Assemblies
Lateral Support Systems
Passive lateral whiffletree concept study – one concept
– 3 meter diameter, 100mm thick mirror
Gravity
direction
LOAD DIRECTIONS
(3 PLACES)
NELSON & CABAK
Gravity
direction
Lateral Support
Performance
3 point
12 point
RMS surface error
11.7 nm
4.6 nm
Peak to Valley surface error
583 nm
171 nm
TMT.OPT.PRE.11.046.DRF01
13
Support Print-through Polish-out
Polishing-out a portion of support gravity print-through
improves performance
Optimize the portion that is polished out – the bias
– wavefront variance through the zenith angle operating range
– atmospheric degradation effects
With the polish-out portion (bias) optimized for a zenith
angle ~30°, surface errors are reduced significantly
Bias Implementation:
– During the surface polishing of M2 and M3 with use of small tools
– After polishing is completed by sending M2 and M3 to a
specialized vendor
Large facility is needed:
– Large enough Ion Beam Finishing (IBF) facility
– Large enough Magneto Rheologic Fluid (MRF) facility
TMT.OPT.PRE.11.046.DRF01
14
M2 18-point axial support performance
– polish-out biased or non-biased
M2 rms print through vs zenith angle
with and without polished-out bias
140.0
axial support with bias
120.0
lateral support with bias
rms surface error (nm)
combined support with bias
100.0
combined support no bias
80.0
60.0
40.0
20.0
0.0
0
10
20
30
40
50
Zenith angle (°)
TMT.OPT.PRE.11.046.DRF01
60
70
JERRY NELSON
15
M2 Positioner
Requirements:
– Travel Range of Motion:
POSITIONER
CONTROL SYSTEM
FIXED PLATFORM
Translation: ±15 mm
Rotation: ±413 arcsec
– Positioning accuracy:
Translation: 3 μm
Rotation: 0.1 arcsec
– Jitter:
Translation: 1 μm RMS
Rotation: 0.05 arcsec RMS
Similar performance to:
HEXAPOD ARM
MOVING PLATFORM
– Subaru Prime Focus Unit Hexapod
– Blanco DECam Hexapod
TMT.OPT.PRE.11.046.DRF01
16
M3 Positioner
Rotation Axis (θ)
Tilt Axis (Φ)
Tilt Cradle
Rotation
Cradle
Tilt Actuator
Counterbalance
weights
Rotation
Bearing
Cable
Wrap
Not shown: Positioner Control Electronics
TMT.OPT.PRE.11.046.DRF01
17
M3 Positioner
Requirements:
– Positioning accuracy:
– Jitter:
Rotation: 0.1 arcsec rms
Tilt: 0.1 arcsec rms
Rotation: 0.2 arcsec
Tilt: 0.2 arcsec
– Motion of M3 during tracking is a function of instrument location
and zenith angle:
Graphs show instrument locations on +X Nasmyth platform only
Rotation of M3, θ
rotation angle, θ (deg)
15
10
5
0
-5
-10
-15
-20
-25
-30
Tilt of M3, Ф
Instrument
locations
-14º az
0º az
6º az
0
50
100
telescope zenith angle (deg)
tilt angle, ψ (deg)
-28º az
Instrument
locations
50
48
46
44
42
40
38
36
34
32
-28º az
-14º az
0º az
6º az
0
20
40
60
80
telescope zenith angle (deg)
18
Summary
TMT M2 and M3 Systems are challenging but can be
fabricated
Conceptual designs have been developed that meet
requirements
Polishing and metrology are challenging but several
facilities in the world are capable of meeting these
challenges
TMT.OPT.PRE.11.046.DRF01
19
Acknowledgements
The TMT Project gratefully acknowledges the support of
the TMT partner institutions. They are the Association of
Canadian Universities for Research in Astronomy
(ACURA), the California Institute of Technology and the
University of California. This work was supported as well
by the Gordon and Betty Moore Foundation, the Canada
Foundation for Innovation, the Ontario Ministry of
Research and Innovation, the National Research Council
of Canada, the Natural Sciences and Engineering
Research Council of Canada, the British Columbia
Knowledge Development Fund, the Association of
Universities for Research in Astronomy (AURA) and the
U.S. National Science Foundation.
TMT.OPT.PRE.11.046.DRF01
20