|
My name is Thomas Zielinski, and I am currently a second year PhD
student at the Institute of Optics. I am working under the direction
of Professor James
R. Fienup on wavefront sensing risk reduction methods for next
generation telescope systems. Current work is modeled after the
James Webb Space Telescope (JWST).
 |
| The James Webb Space Telescope (JWST) |
| The JWST is
a large infrared, segmented-aperture space telescope scheduled
for launch in 2015. The large 6.6 meter primary mirror is
too large to fit in any current launch vehicle, and therefore
must be designed to fold in on itself. In the current design,
the primary will consist of 18 controllable hexagonal segments.
After unfolding, each one of these segments must then be precisely
aligned. For proper operation, these segments must be positioned
and maintained to an accuracy on the scale of billionths of
a meter. In order to determine misalignments an image-based
wavefront sensing algorithm will be used. Unfortunately, there
are several possible scenarios under which a conventional
variant of such an algorithm may fail. The goal of this research
is to develop and characterize phase retrieval and phase diversity
algorithms that are more robust under several non-ideal circumstances
and can be used to phase the telescope when the current algorithms
are not adequate, thereby reducing the risk of failure. |
|
|
Geometrical Ray Trace through the James Webb Optical Telescope
Element |
| |
|
Additionally, the development of such algorithms may serve
as an enabling technology for future missions. Several planned
NASA space telescope missions require image-based wavefront
sensing and control under conditions for which the primary
fine-phasing approaches for the JWST will not be adequate.
More advanced and robust algorithms will be required in order
to satisfy the alignment requirements of such missions.
|
Artist's rendition of a future formation flying telescope. |
|
| |
In order to ascertain the misalignemnts on the primary mirror,
several defocused images are taken as illustrated below.
 |
 |
 |
Aberrated Pupil |
Defocused Image #1 |
Defocused Image #2 |
These defocused images are then used in a nonlinear
optimization algorithm that minimizes the differenece between a
simulated version of these defocused images and the measured data
with respect to parameters controlling mirror alignment. An animation
of this optimization process converging to the correct solution
is shown below.
Our goals are to extend and characterize the performance
of such a method in situations with :
- Lower SNR (allows use of dimmer stars)
- Larger aberrations
- Stars with nearby companions
- Extended Objects
- Significant Jitter
- Wider Optical Bandwidth
- Under-Sampled Data
- Bad Pixels
|
