Your SlideShare is downloading. ×
 Optical testing for giant telescopes
Upcoming SlideShare
Loading in...5

Thanks for flagging this SlideShare!

Oops! An error has occurred.

Saving this for later? Get the SlideShare app to save on your phone or tablet. Read anywhere, anytime – even offline.
Text the download link to your phone
Standard text messaging rates apply

Optical testing for giant telescopes


Published on

Published in: Education, Technology, Business

1 Like
  • Be the first to comment

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

Report content
Flagged as inappropriate Flag as inappropriate
Flag as inappropriate

Select your reason for flagging this presentation as inappropriate.

No notes for slide


  • 1. Optical Testing of Mirrors for Giant Telescopes Jim Burge College of Optical Sciences and Steward Observatory University of Arizona
  • 2. Today’s talk• Introduce giant astronomical telescopes planning to be built in the next decade• Summarize challenges of testing the mirrors for these telescopes• Present technologies and capabilities developed at the University of Arizona that enable measurements of the mirrors J. H. Burge University of Arizona 2
  • 3. The Giant Magellan TelescopeSite in Chile2018 first lightConsortium with UA J. H. Burge University of Arizona 3
  • 4. GMT Design 36 meters high, 25.3 meters across 25-m Primary mirror (f/0.7) 3.2-m segmented secondary mirror corrects for PM position errors deformable mirror for adaptive optics Alt-Az structure ~1000 tons moving mass 21.3 m azimuth disk 21 m elevation C-ring Steel + CFRP secondary support Instruments mount below primary at the Gregorian focusJ. H. Burge University of Arizona 4
  • 5. Key optical testing challenges for GMT primary mirrorsGMT Primary mirror segments – Off-axis asphere with ~15 mm aspheric departure – Segments must match in radius to work together correctly – For 25-m diameter parent, 36 m ROC is needed 14.5 mm departure from best fit sphere J. H. Burge University of Arizona 5
  • 6. GMT segments• Aspheric polishing with stressed lap• Shaping relies on feedback from surface measurement J. H. Burge University of Arizona 6
  • 7. Optical testing of GMT segments Heritage (LBT) GMT~1.4 mm aspheric departure ~14.5 mm aspheric departure Test wavefront defined to match aspheric shape of mirror roof Test opticsAxisymmetric No AxisymmetryTest optics at ~20 meters Light path defined by GMT is muchLight from optical test is only 200 larger 20 m (~3.5 meters across at the top of our tower)mm diameter near the test optics –allows direct measurement of testsystem J. H. Burge University of Arizona 7
  • 8. Interferometric surface measurementsinterferometer andreflective null corrector 3.75-m M1 fold sphere tilted 14.2°with CGH Interferometer interferometercomputer-generatedhologram CGH 25 meters od llipsi 75-cm mirror 0.75 m sphere rent e Sam of pa Axis GMT segment GMT segment Interferometer at M1 center M1 of curvature center of curvature J. H. Burge University of Arizona 8
  • 9. Test tower at Steward Observatory Mirror Lab New tower Original towerNew tower 28 meters tall, 80 tons of steel floated on 400 ton concrete pad accommodates other UA projects (LBT, LSST) lowest resonance of 4.8 Hz with 9 ton 3.75-m fold sphere + cell J. H. Burge University of Arizona 9
  • 10. CGH test of small optics system• CGH inserted into light coming from Sam for alignment test• Reflection back through system is used to verify wavefront• CGH mounted on invar plate with other references for M1 alignment J. H. Burge University of Arizona 10
  • 11. Active system alignment relies on laser tracker• Reference hologram is aligned to Sam. Then it is used to represent Sam.• Laser tracker used to measure locations of Sam, fold sphere, GMT• Fold sphere and GMT actively positioned to ~100 um Sphere Laser tracker Mounted Retroreflector Measures r, q, f to determine position to ~25 µm J. H. Burge University of Arizona 11
  • 12. Measurement of center segment Tilt the fold sphere to nadirCone defined bylight from outeredge of mirror 50 mm CGH compensates 20µm asphericCone defined by departurelight from edge of Vibrationcentral hole insensitive interferometer J. H. Burge University of Arizona 12
  • 13. 3.75 m fold sphere• UA produced mirror, mounted at the top of the tower• Shape is actively controlled based on surface measurements from the center of curvature Polished, measured at the Cast in the Mirror Lab Mirror Lab spinning oven Coated at Kitt Peak J. H. Burge University of Arizona 13
  • 14. Operational test system J. H. Burge University of Arizona 14
  • 15. Scanning pentaprism test Uses the “magic” of the pentaprism Pentaprism rail lies in plane perpendicular to parent axis. Image at CCD CCD camera at focus of paraboloid A is Axis of pa fp re rent p pa Fixed reference Scanning rabo a lo pentaprism pentaprism lo with beamsplitter id d laser parent paraboloid mateColli Off-axis mirror Scanning pentaprism measures slope errors Scanning pentaprism test as implemented by producing collimated beams parallel to for GMT off-axis segments. Pentaprism parent axis. Displacement of focused spot is rail is suspended from tower. measured with camera in focal plane. J. H. Burge University of Arizona 15
  • 16. Pentaprism test of 1.7 m off-axis NST mirror• 1/5 scale GMT pentaprism test (f/0.7 off axis paraboloid)• The pentaprism test only samples lowest order aberrations• The PP results corroborate the results from interferometry! interferometric test pentaprism measurement nm surface (Peng Su) 113 nm rms 102 nm rms J. H. Burge University of Arizona 16
  • 17. Laser Tracker Plus Guide initial figuring laser tracker & distance-measuring interferometers (DMI) sphere-mounted retro- laser tracker reflector for laser tracker DMI laser and remote receiversPSD 10% BS DMIs Retroreflector for interferometer and position sensing detector (PSD) assemblies in 4 places at edge of DMI retroreflector mirror T. Zobrist J. H. Burge University of Arizona 17
  • 18. Ball positioning system J. H. Burge University of Arizona 18
  • 19. Comparison of Laser Tracker with Interferometer Laser Tracker nothing subtractedInterferometry nothing subtracted Difference nothing subtracted 2.0 μm rms (200 point sample) 2.6 μm rms 1.3 μm rmsfocus & astig subtracted focus & astig subtracted nothing subtracted 0.44 um rms 0.44 μm rms 0.53 umrms 0.53 μm rms 0.48 umrms 0.48 μm rms J. H. Burge University of Arizona 19
  • 20. SCOTS : Software Configurable Optical Test SystemConcept from Roger Angel for measuring solar reflectorsApplied to GMT measurements (Peng Su, Bob Parks, Tom Zobrist)
  • 21. Measurement Principle• SCOTS measures slope variations by looking at the reflection of a screen (computer monitor) from the surface under measurement.• The measurement brightness variations are used to calculate surface slopes• Integrate slope maps to get full surface maps Line scanning Fringe projection Image at CCD Monitor image
  • 22. Implementation for GMTUses reflection from large fold sphere Looking down at Sam andVideo monitor and CCD placed upward SCOTS looking near intermediate focus CCD
  • 23. SCOTS Comparison with Interferometry Wow! Principal test SCOTS difference21 polynomials subtracted 21 polynomials subtracted 0.16 μm rms 0.32 μm rms 0.35 μm rms All maps show surface error in microns. Circle indicates 8.4 m diameter.
  • 24. The Giant Magellan Telescope• Uses proven mirror technology• Challenges of optical testing has been solved• Primary segment #1 nearing completion• The casting of the second primary mirror segment is currently being planned J. H. Burge University of Arizona 24
  • 25. Large Synoptic Survey Telescope8.4-m apertureUses three-mirror design to attain3.5° field of view15-sec exposures to survey the sky30TB of data per night!Plan for 2016 operation in Chile J. H. Burge University of Arizona 25
  • 26. LSST PM-TM• Spin cast blank at UA• PM curvature from spinning, TM shape cut in with diamond generated J. H. Burge University of Arizona 26
  • 27. Null test of M1, M3 Separate interferometric optical tests being developed for TM and PM Each test is aligned to mirror using two methods: - Laser tracker - Projected references from computer generated hologramsJ. H. Burge University of Arizona 27
  • 28. The Thirty Meter TelescopeLed by Universities of California (formerly known as the California Extremely Large Telescope)Plan for Mauna Kea site2018 first light J. H. Burge University of Arizona 28
  • 29. TMT optical challenges3.1-m convexsecondary mirror3.5-m flattertiary mirror30 meter f/1 primarymirror, made of 492hexagonal segments J. H. Burge University of Arizona 29
  • 30. The European Extremely Large Telescope42-m apertureEuropean Southern Observatories2018 First light J. H. Burge University of Arizona 30
  • 31. E-ELT telescope 6 meter convex aspheric secondary mirror 42-meter f/1 segmented primary mirror(plus other mirrors) J. H. Burge University of Arizona 31
  • 32. Measurement of large flat mirrors TMT tertiary: Flat mirror: 3.5 x 2.5 meter Conventional test of large flats uses Ritchey-Common test, not practical Reference mirror (spherical) Flat surface under testMeasurement difficulties are solved by technologiesdeveloped and proven at UA J. H. Burge University of Arizona 32
  • 33. 1-m vibration insensitive interferometer• Commercial instantaneous Fizeau interferometer (uses 2 circularly polarized beams)• Modified to use external 1-m reference• Demonstrated on 1.6-m flat mirror• World’s best large flat! H1000 Fizeau interferometer Flat Surface by Stitching Method - power/astigmatism removed 0.1100 0.09 Fold flat200 0.08300 0.07 0.06400 0.05500 0.04600 1 m reference flat 0.03700 0.02 Large flat mirror800 0.01 100 Spowl] Rotary table [R. 200 300 400 500 600 700 800 33 7 nm rms J. H. Burge University of Arizona 33 (J. Yellowhair, P. Su)
  • 34. UA Fizeau test for TMT tertiary Monte Carlo analysis to evaluateLayout of subapertures 3 nm rms typical coupling of alignment, noise to measurement noise surface reconstruction 5.5 nm rms measurement accuracy with proven UA hardware! Mode 1 Mode 2 Mode 3 Mode 4 Mode 5 Mode 6 2.9 nm rms 2.4 nm rms 1.4 nm rms 1.1 nm rms 1.4 nm rms 0.8 nm rms Mode 7 Mode 8 Mode 9 Mode 10 0.8 nm rms 0.8 nm rms 0.5 nm rms 0.8 nm rms 4.6 nmall modes: all modes RSS for rms for Residual modes 4.6 nm rms fitting all from 3 nm rms 3 nm rms residual (C. Zhao) J. H. Burge University of Arizona 34
  • 35. Scanning pentaprism test for flat mirrors Feedback mirror UDT (Alignment AC) Fixed prism Coupling Scanning (reference ) wedge prism ELCOMAT (Measuring AC) Autocollimator system Shutters Mechanical supports• Demonstrated performance is 0.2 µrad rms• Power measurement for 1.6-m flat was 11 nm rms J. H. Burge University of Arizona (J. Yellowhair) 35
  • 36. Measurement of large convex aspheres Convex secondary mirrorsTMT secondary mirror in telescope TMT secondary: 3.1-m LSST secondary: 3.4-m E-ELT secondary: 6-m Conventional test of such mirrors uses the Hindle test, not practical at these sizes LSST SM blank : ULE J. H. Burge University of Arizona 36
  • 37. Swingarm Optical CMM• Scans surface with optical displacement probe• Continuous arc scans create profiles• Profiles stitched together to give surface maps• Works for convex or concave surface probe and arm alignment rotary stages stage convex asphere probe trajectory optical axis axis of rotation center of curvature J. H. Burge University of Arizona 37
  • 38. Surface maps from SOC data for 1.4-m off axis asphere Pattern of 64 scans Repeatability ~ 6 nm rms/scan rms=6nm 0.02 0.015 0.01 Surface measurement 0.005 in µm 0 -0.005 -0.01 -0.015 -0.02 -50 -40 -30 -20 -10 0 10 20 Encoder angle in degrees Interpolated data 896 term reconstruction Low order terms removed 75 nm rms Grid map power removed rms=0.07245um 78 4 terms removed, rms=0.07857um nm rms 6 nm 43 terms removed, rms=0.0065192um rms 0.4 0.06 20 0.4 2050 0.3 40 40 0.04100 0.3 60 60150 0.2 0.02 80 0.2 80200 100 100 0.1 0.1 0250 120 120 140 0 140 0 -0.02300 160 160 -0.1350 -0.1 -0.04 180 180400 200 -0.2 200 100 200 300 400 50 100 150 200 50 100 150 200 Repeatable errors of 34 nm rms are calibrated out (P. Su) J. H. Burge University of Arizona 38
  • 39. Measurement with Swingarm Optical CMM• Original system built at Optical Sciences was made with 3.8-m capacity.• This was designed to be integrated with 4-m polishing machine• Performance is expected to be < 20 nm rms at this size J. H. Burge University of Arizona 39
  • 40. Fizeau interferometry for large secondary mirrorsReference surface on test plate has concave matching asphere for off axis portion ofthe secondary mirrorWe have also developed concept for using a spherical reference, corrected with CGH J. H. Burge University of Arizona 40
  • 41. TMT primary mirror30–m f/1 near paraboloid492 segments1.44-m segments, 45 mm thickEach is supported by 27-point whiffle tree, warped with 21 actuatorsSegment tip/tilt/piston adjusted with 3 position actuators, based on edgesensors J. H. Burge University of Arizona 41
  • 42. E-ELT primary mirror• 42-m f/1 near paraboloid• 984 segments1.43-m segments, 50 mm thickEach is supported by 18 or 27-point whiffle treeSegment tip/tilt/piston adjusted with 3 position actuators, based onedge sensors J. H. Burge University of Arizona 42
  • 43. The challenges of measuring the TMT, E-ELT PM mirror segments• Off-axis aspheres, with different prescription (curvature changes from center to edge)• To work together, the radius of curvature must match. Power is treated as a figure error• Must be efficient, limiting setup, alignment, and test time• Measurement accuracy of 5 nm rms is requiredOther important issues:• Efficient fabrication of hundreds of mirror segments• Complex support for each segment• Active shape and position control for each segment J. H. Burge University of Arizona 43
  • 44. CGH Fizeau test for primary mirror segments Mirror Test Plate Test wavefront segment First order from CGH Measurement Reflects from segment Common CGH Reference wavefront CGH Zero order from CGH Reflects from reference sphere Collimator Projection diffuser Lens m=1 Return : common path Both wavefronts coincide The difference between these gives the shape error in the segment m=0objective CCD camera Focusing lens Aperture Blocked by aperture: m = 0 from segment m = 1 from sphere Convex reference Aspherical Reference and test wavefronts come to sphere surface focus and pass through aperture All other orders and reflections are blocked J. H. Burge University of Arizona 44
  • 45. CGH Fizeau test• Common path – low noise• Radius matching is easy, all segments compared with the same reference• Detailed engineering analysis for TMT, E-ELT predicts 14 nm rms overall accuracy, 5 nm rms after some low order correction J. H. Burge University of Arizona 45
  • 46. Full scale demonstration of CGH Fizeau test Test convex, off axis aspheres by Reference measuring though the substrate, OAP sphere Measurement Common CGH CGH Collimator diffuser m=1 Lens m=0objective CCD camera Blocked by aperture: m = 0 from OAP m = 1 from sphere Aspheric surface Spherical surface J. H. Burge University of Arizona 46
  • 47. UA achieved very low noise measurements with CGH Fizeau systemExcellent fringe visibilityExcellent spatial resolutionOverall accuracy of < 4 nm rms< 1 nm rms noise per measurement (average of 50 maps)Verified effects of straie in glass to cause < 1 nm rmsLargest sources of error: ghost reflection in CGH, coating irregularity on fold flat J. H. Burge University of Arizona 47
  • 48. Comparison of Fizeau, SOC for off-axis aspheres • The Fizeau test was budgeted as < 3.3 nm rms uncertainty, after correction for low order terms. • SOC measurements of the OAPs are consistent with this. Fizeau SOCRaw data Astigmatism and coma from alignment were not needed to be controlled accurately 109 nm rms 117 nm rms Difference After removing low order terms 14 nm rms 16 nm rms 7 nm rmsLargest errors in Fizeau came from coating defect on large fold flat 1 nm rms ghost fringes 1 nm rms J. H. Burge University of Arizona 48
  • 49. What next?Combine several elements:• Fizeau interferometry with spherical reference, corrected by CGH• Vibration insensitive interferometry using polarizationDesign for such a test for LSST was presented last summer.Prototype work is underway. (M. Dubin) J. H. Burge University of Arizona 49
  • 50. Conclusion• University of Arizona technology is enabling the Giant Magellan Telescope and the Large Synoptic Survey Telescope• We are prepared to support TMT and E-ELT if those projects move forward. J. H. Burge University of Arizona 50