Apodized phase plates & Shaped pupils

Transcription

1 Apodized phase plates & Shaped pupils Surprising similarities & key differences Carlotti Alexis & Mamadou N Diaye Combining Coronagraphs and Wavefront Control - Oct. 6-1, Lorentz Center, Leiden 1

2 'Direct' apodizers Shaped pupil (SP) Apodized phase plate (APP) Amplitude mask Phase mask Achromatic Two sided - Partially achromatic Two-sided with vector-app Spergel, Kasdin, 21 Kasdin, Vanderbei, Littman, 23 Carlotti, Vanderbei, Kasdin, 21-2 Codono, 24 Kenworthy, Quanz, Meyer & 5 co-authors, 21 Snik, Otten, Kenworthy & 4 co-authors, 212

3 2D SP APP Pupil Image Optimization problem: Maximize: N X N X A(xi, yj ) i=1 i=1 With two constraints: -1 < A(xi, yj ) < 1 E(uk, vl ) 2 < 1 with {xi, yj } in Pupil c with {uk, vl } in Dark zone Phase mask w/ real amplitude: APP with two symmetric dark zones 3

4 Two-sided: similar performance In some cases, APP have a slight advantage over SP especially in 'extreme' cases (small IWA, very high contrast ) APP SP Angular distance (lambda/d) Shaped pupil PSF (log scale) TAiry=28% PSF ; CObs=.3 ; CST=8 ; IWA=6 ; ALPHA=18 Throughput: 33% TAiry=32% SP [,1] APP [-1,1] lambda/d 4

5 Hundreds of SP designed for specific telescopes but no global characterization yet Clear aperture Gemini/VLT-like Segmented WFIRST-AFTA => New paper submitted in Sept. 214! - trade-offs: IWA vs Contrast vs Obscuration, etc. - sensitivity to low-order aberrations - Benefits of APLC-like mode vs standalone mode (Carlotti, N Diaye, Riggs, Kasdin, Soummer, Vanderbei, Pueyo, Zimmerman) 5

6 Standalone vs. APLC Standalone 1/3 of masks have 5-13 contrast gain w/ APLC NACO APLC SPHERE APLC configuration provides higher contrast! Higher gain for smaller IWA (proportionally more energy next to FPM) 6

7 Closed-form approximations Tring = -.79 CObs IWA -.74 OWA C Tbowtie = -.54 CObs IWA +.11 OWA C Great for early design process Bow-tie T CO IWA OWA C T CO IWA OWA C Ring T CO IWA OWA C T CO IWA OWA C e.g.: W/ a bow-tie region, one order smaller C <=>.5 λ/d larger IWA 7

8 Moving energy around - Energy moved as close to the star as possible - Some regions are better suited than others: W/ real amplitude masks W/ APP Original PSF Ring Bow-tie 8

9 Optimizing the phase? You may optimize Re & Im, instead of phase "Non physical" cost functions work, e.g.: NX NX (A R (x i,y j )+A I (x i,y j )) i=1 j=1 Phasers are limited to inscribe square: I 1/ p 2 R 1/ p 2 9

10 One-sided dark zone More throughput, smaller IWA.5 PHASE ; CObs=.3 ; CST=8 ; IWA eff =2.5 ; ALPHA=6 3.5 PHASE ; CObs=.3 ; CST=8 ; IWA eff =2.5 ; ALPHA= π/2 -π/ APP is made of! phase steps lambda/d PSF ; CObs=.3 ; CST=8 ; IWA =2.5 ; ALPHA=6 eff lambda/d PSF ; CObs=.3 ; CST=8 ; IWA =2.5 ; ALPHA=18 eff Left: 6 dark hole: 43% throughput lambda/d lambda/d Right: 18 dark hole: 3% throughput

11 Application to WFIRST-AFTA 2.9 λ/d IWA, 1-8 contrast (5x1-9 in 2 nd image plane) 26% Airy throughput for 18 region (~ same for 12 ).5 APP phase ; Throughput=26% ; IWA=2.5 λ/d ; C= APP PSF (log scale) = Angular extent! w/ SP for AFTA 1 2 units of pupil diameter 1 1 Angular distance (lambda/d)

12 IWA vs. Throughput Throughput for 6 dark zone, 15% obscuration IWAmin depends on contrast, not on obscuration: IWAmin = 2 λ/d for 1-6 IWAmin = 2.5 λ/d for dark hole : Clear advantage for APP if planet location is known Airy throughput (%) Throughput vs. IWA ; CObs=.15 ; ALPHA=6 APP 1-6 APP 1-8 Disc. - SP 1-6 Charact. - SP 1-6 Disc. - SP 1-8 Charact. - SP 1-8 CST=6 CST= Effective IWA (λ/d) 12

13 IWA vs. Throughput Throughput for 18 dark zone, 3% obscuration (15% higher for 15% obscuration) IWAmin depends on contrast, not on obscuration: IWAmin = 2 λ/d for 1-6 IWAmin = 2.5 λ/d for dark hole : Clear advantage for APP even in discovery mode Airy throughput (%) Throughput vs. IWA ; CObs=.3 ; ALPHA=18 APP 1-6 APP 1-8 CST=6 CST=8 Charact. - SP 1-6 Discovery - SP Effective IWA (λ/d) 13

14 WFE effects on APP 1% bandwidth Standalone configuration APLC configuration Obs: 15% - IWA: 1.5l êd - ALF: 6 Normalized intensity in log scale Two groups Contrast in Dmag tip tilt defocus astigmatism astigmatism2 coma coma2 trefoil trefoil2 spherical rms WFE error in l unit 14

15 COMPASS simulations E-ELT pupil ; r=16cm ; windspeed=2m/s ; Roof-sensor ; 64x64 DM + tip-tilt DM ; Obs. at 3.5 μm (METIS).5 sec of exposure time (25 x 5ms simulated frames) W/ APP W/ SP (for HARMONI) -1 IWA = 2 λ/d 39 mas OWA = 24 λ/d 47 mas residual spikes 15 Plume? -5-6

16 Missing segments presented at the Exoplanets with E-ELT workshop in Garching (Feb. 214) With 3 (randomly) missing segments:! 1% chance to have < 1-6 mean contrast! 5% chance to have < 1-5 mean contrast! 9% chance to have < 3x1-5 mean contrast Possible solutions: - design coronagraphs for missing segments (each night ) - use micro-mirror arrays with 1k x 1k mirrors (see Zamkotsian et al.) - adaptive liquid crystal plate? - use adaptive optics to accommodate (but how much stroke?) - 1 DM? 2 DMs? Stroke? Number of actuators?