Development of Adaptive Optics system at laboratory
Narsireddy Anugu Click to edit Master subtitle style Roll No: 97/AIM/101006 Under the Guidance of Dr.J.P.Lancelot & Prof.
A.K.Saxena
7/12/12
Contents
Introduction Simulation of Kolmogorov turbulence numerically and experimentally its characterization Simulation of SHWFS Studies of SHWFS in the presence of turbulence Using turbulent simulator closed loop correction at lab Analysis of the results
7/12/12
Kolmogrov Theory of Turbulence in a solar Outer scale L0 Nutshell
Inner scale l0
h Wind shear
convection
h
ground
Big whorls have little whorls, Which feed on their velocity; 7/12/12 Little whorls have smaller
Kolmogrov Theory of Turbulence : Eddy Cascade
Assume energy is added to system at largest scales outer scale L0 Then energy cascades from larger to smaller scales (turbulent eddies break down into smaller and smaller structures). Size scales where this takes place: Inertial range. Finally, eddy size becomes so small that it is subject to dissipation from viscosity. Inner scale l0 L0 ranges from 10s to 100s of meters; l0 is a few mm
7/12/12
Wave Propagation in Turbulent medium
For monochromatic plane waves arriving from a distant point source with wave-vector k, we have The Turbulent layer 1. Scatters light 2. Perturbs Phase of the wave 3. Causes fractional Amplitude change with effect:
7/12/12
Atmospheric terms
Atmospheric coherence radius : (ro )
7/12/12
Images with Seeing
Image = Object Point Spread Function I = O PSF u= O ( u ) P ( u v) dv I( )
I ( f ) = ( f ).P( f ) O
OTF OTF . O TF effective = Telescope Atmosphere
7/12/12
Overcoming atmospheric seeing
Speckle imaging : which allows bright objects to be observed with very high resolution. Working outside the Atmosphere : Hubble Space Telescope and thus not have any seeing problems Adaptive optics : Systems that partially solve the seeing problem. Observations are usually limited to a small region of the sky surrounding relatively bright stars. Lucky Imaging : The technique relies on the fact that every so often the effects of the atmosphere will be negligible, and hence by recording large numbers of images in real-time, a 'lucky' excellent image can be picked out. This technique can outperform adaptive optics in many cases and is even accessible to amateurs. It does, however, require very much longer observation times than adaptive optics for imaging faint targets, and is limited in its maximum resolution.
7/12/12
Generation of Kolmogorov Phase screen
Motivation : To study astronomical objects propagation through atmosphere and analyze AO algorithms and system. Atmosphere can be simulated by different methods numerically
exp( 2 /i2 ) D r N ( )= 0.023( 0 / ) 2 2 ( + 0 )11/ 6 Van Karman Power spectral density
5/3
0 =
Where,
2 L0
i = 5 .9 2 l0/
f ( r ) = N ( ) e dk ikr
phase screen related to spectrum
Discrete Fourier transform of the square root of the PSD * random numbers. 7/12/12
Continued..
Fourier Transform method
7/12/12
Analysis and discussion of simulation results :
D( r =r ( ( 1 = ) < +>1 ) r r ( ) 6 82 ) .8 (
(5 ) /3
r r0
Phase structure function:
7/12/12
Study of point source in turbulence
At D/r0 = 1 & 5
D/r0=10 & 15
7/12/12
Corresponding PSF
At D/r0 = 1 & 10
D/r0=5 & 15
7/12/12
This OTF is calculated for long exposure images.
Normalized OTF at Different turbulent strengths
For this 100 short exposures (1millisecon d) images 7/12/12 added up
Adaptive Optics system
7/12/12
wave-front sensor
Requirements of WFS: WFS must work on white light incoherent sources. WFS must use the photons very efficiently. The WFS must be linear over the full range of atmospheric distortions. WFS must be fast.
7/12/12
Continued
Shack Hartmann
Shearing interferometry
7/12/12
Pyramid wavefront sensor
Curvature wavefront sensor
Shack Hartmann sensor simulated results with WCOG at D/ro=5
7/12/12
Centroid algorithms
COG
WCOG
Weighting function chosen is Gaussian
S =,y y wx ) ( , ) w I xx ( , y S =,yx yW,x ) ( , ) x xx I ( y S =,yx yW,x ) ( , ) y yx I ( y
W x y ) (= e p (( x ) y 2 ) c (, 1/2 )x 2 + ( y ( x )/2 2)
2 c
2 S x =,y x yW x y ( , ) x xx I (, )
2 S y =,y x yW x y ( , ) y yxI (, )
X = S/ c S w x Y = S/ c S w y
=( S) x S Sxx S * S S S 1/ w S * + * S x w y w * y y y
7/12/12
Spot realization
Gaussian
) 2 J (u P ( ) = 1 0 u u
7/12/12
Sinc2 function
Ai ry di sk
Different noise simulation
7/12/12
Photon noise , turbulent noise (D/ro=8), read out noise & total
Centroid calculation with WCOG at D/r0 =3 and 5
7/12/12
Comparison of Centroid algorithms
At different At turbulent different S/N strengths ratio
7/12/12
Continued.
At different subaperture sampling of spot
7/12/12
The spot pattern at the focal plane of a S-H sensor
Gaussian spot array (10 x 10)
Airy disk array (10 x 10)
7/12/12
Random noise affected spot array
Gaussian spot array (10 x 10)
Airy disk array (10 x 10)
7/12/12
Array pattern at different turbulence strengths D/ro=1 D/ro=5
7/12/12
Phase reconstruction
WCOG algorithm has been using for the centroid calculation phase reconstruction : Model Approach ( Zernike Polynomial ) Zernike polynomials are orthogonal polynomials defined over unit radius of circle. Expression of Wavefront: and
The derivatives of the Zernike polynomials can be expressed as a linear combination of Zernike polynomial (Noll, 1976)
7/12/12
In Matrix notation
Shack Hartmann sensor simulated results with WCOG at D/ro=15
7/12/12
Low cost turbulence simulator in lab
Fabrication of phase screens by spraying multiple layers of ordinary hair spray onto a glass substrate Hair spray
7/12/12
Experimental setup used for phase screen characterization
7/12/12
Phase screens obtained at lab and numerical simulation
AtAt D/r0 = 9 =3
7/12/12
Characterization of phase screens
Characterization has been done by measuring ro value with two methods 1.The model representation ( Zernike approach) W *( ) = ) r aZ r (
i i i 2 <ai= >
Ni
5/6D *( ) r0
2.OTF method
O F ( = 4( T ) exp( 3.4 ) ) r0
5/3
7/12/12
Summary of the results
Modal approa ch
7/12/12
Experimental setup for SHWFS
Reference
Distorted wavefront At D/r0=9
7/12/12
Comparison of Phase variance with theoretic model
D < a 2i > = N i *( )5/6 r0
7/12/12
Schematic and experimental layout for closed loop AO
7/12/12
Lenslet Spots
Reference
At turbulent strength (D/r0=20)
7/12/12
Wavefront constructed
No Turbulence
7/12/12
Zernike Coefficients
No Turbulence turbulence
7/12/12
When closed loop Adaptive optics system is on
7/12/12
Closed loop AO Demonstration
AO on AO off (D/r0=20)
7/12/12
Summery
Project was carried out in three phases Study of Atmospheric turbulence experimentally and numerically Study on Shack-Hartman sensor Closed loop correction with AO kit in the presence of turbulence simulator Numerical simulation has been done with Matlab-2010a with AMD Athlon 2.7GHz computer . Airy disk spots are used for Shack Hartmann wavefront sensor study. Centroid calculation with IWCoG. For phase reconstruction 21 Zernike modes has been used.
7/12/12
THANK YOU
7/12/12
