LSST Alignment

1
LSST Alignment

• C. F. Claver (NOAO)
• J. H. Burge (UofA, Steward Obs.)

2
Outline

• LSST Imaging requirements
• Specific Concerns
• Initial Alignment Concept
• Final Alignment Simulation
• Maintaining Alignment
• Summary

3
Specific Concerns

• Fast optics demanding image quality
• Primary is f/1
• Final system is f/1.25 51?m per arcsecond
• Image error budget is tight
• 0.5? site 0.33? total system ? 0.6? delivered
• LSST optics 0.12?
• Remaining error budget 0.30?
• Complex system, many degrees of freedom

4
Initial Alignment

• Precision Metrology
• Fabrication
• Surface-Substrate
• Substrate-Fiducials
• Assembly
• Optical fiducials to structure
• Commercial Solutions
• Differential
• 0.06 resolution
• 9µm 2µm/m accuracy
• 3µm 1µm/m repeatability
• Absolute
• 20µm 1.1µm/m accuracy
• 7µm 1µm/m repeatability
• Range
• 35m

5
Initial Alignment Concept

• Optics within view of laser ranging
• Two locations
• Redundant measurement
• M1
• Instrument
• Precision limited by fiducial placement during
  fabrication

6
Final Alignment

• What are the issues
• Can the wavefront be sensed well enough to
  converge on system specs?
• How field many positions?
• What signal-to-noise?
• Is there a unique solution?
• Do we care?

7
Simulation Process
Initializemodel
500 models
Perturb degrees of freedom
Record initial merit function
Measure wavefronts
Add noise
5 times
Solve for best imagery
Record solved merit function
8
Degrees of Freedom

• Rigid Body
• 5 DoF for each
• Primary M1
• Secondary M2
• Tertiary M3
• Surface Deformation
• M1 12 modes
• M2 6 modes
• M3 6 modes

39 degrees of freedom
9
Tolerances

• Rigid Body
• M1
• Translation ?0.5mm in each of 3 axes
• Tilt ?60? about 2 axes ? to optical axis
• M2 M3
• Translation ?0.25mm in each of 3 axes
• Tilit ?60? about 2 axes ? to optical axis
• Surface Deformation
• M1
• Z4-Z9 ?3.25?m (?5 waves _at_650nm)
• Z10-Z15 ?1.30?m (?2 waves)
• M2 M3
• Z4-Z9 ?1.30?m (?2 waves)

10
Wavefront Measurement

• 60 measured quantities
• 5 Field Positions
• Zernike Terms Z4-Z15
• In Situ Curvature Sensing
• Ad-hoc noise model
• 0.05?m minimum uncertainty
• 20 of Zernike term in quadrature

11
Simulated Curvature Sensing

• Perturbed Optical model
• Before and after extra-focal images
• EF wavefront solution
• Compare against Zemax wavefront error

12
EF vs Zemax Wavefront

• Good agreement between calculated Zemax wavefront
  and recovered wavefront.
• Recovered wavefront falls mostly within error
  envelope.

13
Solve for the degees of freedom

• Using Zemax
• Each DoF becomes a variable
• Optimize model using noisy wavefronts as targets
• Evaluate differences between solution and
  unperturbed model
• Correct perturbed model using calculated
  differences
• Record weighted spot size merit function

14
Initial Perturbed State

• Merit Function
• 9 field positions
• weighted by effective area
• 500 Models
• Mean 118 ?m
• STDEV 39 ?m
• Max 245 ?m
• Min 22 ?m

15
Monte Carlo Results

• Put stats here

16
Merit Function Final Distribution

• 91.2 lt 10?m after 5 iterations
• All models once below 10 ?m, stay!

17
Remaining Issues

• Maintaining Alignment
• At what frequency are measurements required
• How often to update solutions