Optical Design

1
Optical Design
2
The optics include

• On-Instrument Wave-Front Sensor
• Near-infrared Integral Field Spectrograph
• Image Slicer Feed (Focal Ratio converter and
  Coldstop).
• IFU.
• Spectrograph.
• Two alternative designs are presented having
  different
• Image Slicer Feed and IFU arrangements

3
3D IFU Design
4
Content IFU Design
5
Baseline Design Trimetric view
6
Baseline Design Side view
7
Alternative Design
8
Concentric IFU
9
Concentric IFU
10
Concentric IFU
11
Concentric IFU
12
Concentric IFU
13
Linear IFU
14
(No Transcript)
15
Linear IFU
16
Pupil Mirror Array
17
Baseline or Alternative Design?

• Both feasible.

• Offner coldstop can be better baffled but is
  more complex.

• Concentric system is less compatible with
  cryostat.

• Linear pupil mirror array can be a surface of
  revolution but must be stepped.

• Linear system requires a more complex collimator
  with more surfaces.

• Concentric system has better optical and
  throughput performance.

• Linear system can be better baffled.

18
Baseline Design Data
Field Mask Size 2 mm square (3.0 arcsec square
at f/16) Focal Converter Ratio 16 (f/16 -
f/256) Cold-Stop Diameter 4 mm Image Slicer
Field 30 mm square, 29 slitlets each 1 mm (0.1
arcsec) wide Pupil Mirror Array Pitch 0.25
(1.98 mm on 448 mm radius arc) Pupil Mirror Array
Fill Factor 0.82 Field Mirror Array Pitch 0.25
(1.86 mm on 420 mm radius arc) Field Mirror Array
Focal Ratio f/16 Collimator Focal Length 421
mm Collimator Beam Diameter 26.3 mm Ebert
Angle 30 Grating Angle 20 Spectral
Resolution 5300 (for a grating angle of
20) Camera Focal Length 288 mm Camera Focal
Ratio f/9 Detector Field 36.864 mm square
(2048 x 2048 x 0.018 mm pixel)
19
Diffraction Effects
20
Diffraction Effects

• For K 1.6 and l 2.5 mm
• Throughput loss is 3.
• Image profile wing attenuation at 3 pixels
  off-centre is 660.

21
Image Quality (for Baseline Design)
22
Cold Stop Pupil
10 mm Box
(Diffraction Limit is 34 mm at l 1 mm)
23
Image Slicer Field
Boxes are 0.512 mm (corresponding to one pixel in
spectral direction)
24
Pupil Mirror Array
Boxes are 40 mm
25
Field Mirror Array
Boxes are 0.032 mm (corresponding to one pixel in
spectral direction)
26
Collimator
Boxes are 0.032 mm (corresponding to one pixel in
spectral direction)
27
Camera
Boxes are 0.018 mm (one pixel)
28
Detector Field without Re-Focus
J1 Grating
H Grating
K Grating
Boxes are 0.018 mm (one pixel)
29
Grating Configuration
30
Configuration Options for Resolution and Field
Size

• 26.3 mm Beam (R5300) and Blaze-to-Collimator
  (3.0 arcsec Square Field).
• 19.7 mm Beam (R4000) and Blaze-to-Collimator
  (3.0 arcsec Square Field).
• 32.0 mm Beam (R5300) and Blaze-to-Camera (3.6
  arcsec Square Field).
• 24.0 mm Beam (R4000) and Blaze-to-Camera (3.6
  arcsec Square Field).

Blue option is the baseline Grey option makes
spectrograph too long
31
26.3 mm Beam and Blaze-to-Collimator

• H band (1.49 - 1.80 mm) is fitted to detector.

• K band coverage is restricted to 2.00 - 2.41 mm
  by detector.

• Resolving power is 5300 (with 20 grating angle).

• Field size is 3.0 arcsec square with 29 slitlets.

• Two-pixel spatial resolution is 0.082 arcsec.

• Spectrograph size is as shown in drawings.

This is the baseline and preferred option
32
19.7 mm Beam and Blaze-to-Collimator

• K band (1.95 - 2.50 mm) is fitted to detector.

• Other bands are extended.

• Resolving power is 4000 (with 20 grating angle).

• Field size is 3.0 arcsec square with 29 slitlets.

• Two-pixel spatial resolution is 0.082 arcsec.

• Spectrograph size is reduced to 75 of baseline
  version (easy to fit).

33
24.0 mm Beam and Blaze-to-Camera

• K band (1.95 - 2.50 mm) is fitted to detector.

• Other bands are extended.

• Resolving power is 4000 (with 20 grating angle).

• Field size is 3.6 arcsec square with 35 slitlets.

• Two-pixel spatial resolution is 0.122 arcsec.

• IFU and collimator size is reduced to 91 of
  baseline version.

• Camera size is reduced to 62 of baseline
  version.

34
R5300 Gratings for the 26.3 mm Beam Configuration
35
Optional Gratings for the 26.3 mm Beam
Configuration
36
R4000 Gratings for the 19.7 mm Beam Configuration
37
Optional Gratings for the 19.7 mm Beam
Configuration
38
IFU Manufacture
39
Issues

• All three components diamond-machined.

• Image Slicer machining is simple but fanning
  accuracy is demanding.

• Arrays are proposed as monoliths to eliminate
  relative alignment problems.

• Lathe turning allows genuine spheres and toroids
  (surfaces of revolution).

• Lathe turning requires centres-of-curvature to
  lie in a straight line, or complex jigging.

• Straight line condition not met for either the
  concentric or linear system.

• Jigging introduces alignment and boundary
  placement problems.

• Fly-cutting avoids these problems but surface
  approximation errors must be acceptable.

• Stepped surfaces (Linear IFU) introduce
  additional problems.

40
Image Slicer 29 slices each 1 mm x 30 mm Fanning
increment 0.127 Fanning tolerance 0.002
41
Fly-Cutter Envelope
42
(No Transcript)
43
Results

• Aberration is strongly dependent on displacement
  from origin of coordinate system.

• Aberration is minimised by locating coordinate
  system origin in centre of pupil.

• For the drawn configuration the aberration
  envelope for the end elements is 0.70 x 0.48
  pixels.

• Aligning the fly-cutter axis perpendicular to
  the array blank reduces this to 0.34 x 0.01
  pixels.

• Additionally tilting the blank can reduce the
  aberration envelope to 0.02 x 0.01 pixels.

44

• Pupil Mirror Generation.
• Needs three-axis translation.
• No steps between mirrors.
• 1 mm diamond tip radius.

45
Conclusions

• Fly-cutting can produce good surface figure.

• But it requires a machine with three
  translational axes.

• These machines are available, but are not common.

• The University of Bremen are interested and
  competitive.

• P-OE have experience with IFUs, but only operate
  lathes.

• Method not applicable to the stepped (linear
  IFU) pupil array.

• But lathe turning is an option because
  straight-line condition is met.

• Tip radius should be large (1 mm) for good
  smoothness.

• This is a difficulty for the stepped array.

46
Alignment
47
Some Basic Principles

• Relative adjustment of IFU channels is avoided
  by precise construction methods.

• Location of components largely determined by
  machine registration.

• Adjustment largely restricted to orientation of
  selected components.

• Errors measured by means sighting methods and
  recorded images.

• Camera lenses mounted in a single barrel with
  precision fits.

Comments

• Procedure is yet to be designed.

• The unusual and difficult requirement is the
  need to cope with small channel elements.

48
Throughput
Specification is 15
49
AR Coatings

• Bare gold is the safe default for reflective
  surfaces.

• Bare silver is an alternative.

• Both are suited to silica and aluminium alloy
  substrates.

• Protected silver and gold are options (Denton
  FSS-99 and FSG-98).

• LaserGold is an option for the diamond machined
  surfaces (may reduce scatter).

• MgF2 is assumed for the low-index refractive
  materials.

• Proprietary Janos coatings are assumed for the
  ZnSe and Sapphire.

50
Concentric IFU Design
51
Linear IFU Design
52
Ghost Images
Specification
53
Scattered Light

• Diamond machined surfaces are likely to be a
  major source of scatter.

• Expected surface roughness is 10 nm RMS.

• TIS for this is negligible (0.006 at l 1.6 mm).

• Major concern is near-angle scatter from
  surfaces not at field foci.

• The only potential offender in the IFU is the
  Pupil Mirror Array.

• Diamond machined surfaces employed elsewhere may
  also be a problem.

• Ameliorating measures large tip radii, grade
  L111 alloy and LaserGold coating.

• Near-angle scatter effects have not been
  quantified.

• Grating is a major source of scatter.

54
Tolerances
55
Performance Criteria

• Gemini instrument wavefront error allocation is
  124 nm RMS (S 0.80 _at_ 1.65 mm)

• Wavefront error for 0.04 arcsec 50 EE is about
  200 nm RMS (S 0.56 _at_ 1.65 mm)

56
Preliminary Optical Error Budget
Assumptions

• All mirrors diamond machined with 40 nm figure
  error.

• Refractive surfaces have a figure error of 20 nm
  RMS.

• Wavefront error proportional to ratio of beam
  diameter to surface diameter.

• No diffractive pupil spread.

• 200 nm RMS wavefront error is acceptable.

57
Optical Tolerance Budget for Concentric IFU System
58
Camera Lens Mounting Tilt
r1
e
d
q
r2
59
Camera Lens Tilt Error
Lens r1 (mm) r2 (mm) q (deg) 1 157 220 0.013 2 -1
77 757 0.005 3 95 -82 0.002 4 109 207 0.016 5 -84
-206 0.019 Additional element wedge 0.003
60
Camera Error List
Surface Tilt (deg) Rad (mm) Sep (mm) Fig (fringes
_at_ 0.63 mm) 1a 0.013 0.15 0.10 0.5 1b 0.013 0.20 0
.05 0.5 2a 0.005 0.17 0.10 0.5 2b 0.005 0.50 0.05
0.5 3a 0.002 0.02 0.025 0.5 3b 0.002 0.02 0.10 0.5
4a 0.016 0.10 0.05 0.5 4b 0.016 0.20 0.10 0.5 5a
0.019 0.10 0.05 0.5 5b 0.019 0.20 - - Red-Bold
Extra constraint applied to standard values of
0.1 (ZnSe lens)
61
Camera Performance
Condition Wavefront Error Strehl Ratio (nm
RMS) (l 1.65 mm) With Design Errors
Only 36 0.98 With Construction Errors
Only 28 0.99 With Design and Construction
Errors 46 0.97 No pupil diffraction spreading
included Back-Focus compensation employed
62
Conclusions

• Wavefront error for system may be high.

• Worst offenders are diamond machined surfaces.

• Avoid diamond machining for sensitive surfaces
  (except pupil array)?

• Assumed errors for small diamond machined
  surfaces may be pessimistic.

• Camera errors are acceptable but ZnSe lens is
  demanding.

• Other camera lenses can be re-optimised after
  after ZnSe lens is made.

• Normal construction methods for high quality
  optics are acceptable.

63
Risks

• Scatter from diamond machined surfaces.

• Baffling effectiveness.

• Angular registration errors in Image Slicer.

• Viability of machining method for monolithic
  mirror arrays.

• IFU alignment.

64
Summary of Optical Issues

• Offner relay or incidental pupil for Cold Stop?

• Concentric or Linear IFU?

• R5300 or R4000 resolving power?

• 3.0 or 3.6 arcsec square field size?

• Focus control?

• Diamond machining for non-IFU mirrors?

65
NIFS plus ALTAIR

• Artificial star from ALTAIR.
• Differential flexure comparison.
• Phase Maps, common path optics in NIFS.

66

• NIFS common path optics.
• Telescope focus -30mm to fill 2mm entrance
  aperture.
• Common path optics near focus see large changes
  in footprint.

67

• Phase maps from NIFS will be small.
• We can make phase maps by warming the cryostat
  and moving the detector about -9mm. Our manual
  focusser is designed for about -3mm.

68
End