Tiptilt options Trade Study Report on Standalone TT

Tip/tilt options Trade Study Report on Stand-alone T/T vs. DM on T/T Stage (WBS 3. 1. 2. 2. 13) Brian Bauman December 12, 2006

Status • Study started, perhaps 20% complete • Scheduling/manpower issues resolved last week; now can spend more time

Options considered so far • • • Pair of flat mirrors Tilting OAP 1 + move field lens Under-DM platform Can we use secondary (future or existing)?

Pair of tip/tilt mirrors • Coordinated moves of mirrors to affect pointing, but not centering • Place in the telescope focus space • Mirrors would be ~250 -300 mm in diameter (3. 6 -5. 0 kg)—may be in range of PI off-the-shelf stages, e. g. , PI-518. TCD (more later) • If separated by 500 mm, then tilts necessary are ~1. 5 mrad; consistent with PI-518. TCD (need to calculate resonant frequency) • Packaging could be interesting • May merit follow-up if throughput/emissivity penalties are acceptable

Tilting OAP 1 • Accommodating ± 2 arcsec of tip/tilt slews pupil around by about 1. 5% (about 0. 7 -1. 0 subaperture) – – • • • Plate scale: 1. 375 arcsec/mm 1. 45 mm @ telescope focus OAP focal length ≈ 3 meters Beam is steered by 1. 45 mm / 3 m ≈ 0. 5 mrad, cf. 33 mrad, f/15 cone On-axis aberrations generated by tilting OAP: 80 nm rms of astigmatism per arcsec on sky…. not too bad. – Correcting on-axis aberrations brings off-axis performance approximately back to pre-tilt level Even accommodating only 1 arcsec of tip/tilt slews the pupil by ~0. 3 -0. 5 subaperture Results in time-dependent illumination pattern on DM/WFS/other pupils; DM-WFS registration not affected Perhaps closed-loop performance penalty? Could mitigate by stopping down aperture on both inside and outside of the “annulus” For reference, mirror would weigh about 16 -25 lbs (7. 3 -11. 4 kg), without lightweighting (cf. 5 kg limit for PI-518)

Tilting OAP 1 + field lens • Need 6 mm of motion on field lens to steer cone by 0. 5 mrad • Not really practical

Scaled-down CILAS TMT mirror • TMT DM specs/features – 360 mm pupil mirror – 73 x 73 actuators – 41 kg • Assume DM scales down for NGAO – 64 actuators across – 315 mm diameter (within range of DM sizes considered during Indian Wells) – About 31 kg • Very rough assumptions, but enough to get going

Kinematic vs. gimballed mount • Gimballed obviously most desirable, but CILAS design of integrated gimbal disheartening • Is kinematic mount sufficient? Example with 315 mm DM – Tilt required on 315 mm DM (worst case) is 70 arcsec = 350 μrad for 2 arcsec tilt on sky – If center of mirror is ~200 mm from the axis, then Abbe motion translation is 200 mm*350 μrad =70 μ, which is small (2%) compared to interactuator distance of 3. 5 mm – Seems practical but should quantify performance penalty • If lever arm is 200 mm, then stroke required is 70μ— consistent with the larger stroke PI actuators (120μ); could be reduced with smaller lever arm

PI stages under DM • Discussions with PI have indicated that the question is not whether it could be done, but how much it will cost…. modulo moment-of-inertia concerns below • Awaiting more information about best approaches/using previous designs • Largest PZT actuators – can pull 3500 N, push 30, 000 N – About $10 K each – Resonant frequency ≈(1/2π)√(k. T/m) = (1/2π) √(240 N/μ)/30 kg=450 Hz (30 kg per actuator probably pessimistic); seems OK – Better moment-of-inertia/angular acceleration calculation pending, but I’m pretty concerned about it so far – depends on temporal tip/tilt power spectrum assumptions

Mirror sizes supported by other vendors • • Ball: ~ 2” Newport: 2” Axsys: ~2” Optics. In. Motion: 2 x 3”, up to 4” custom