Adaptive Optics Simulations

Introduction

Adaptive optics requires simulations in order to predict the performance of the system before it is actually built. Different metrics can be used for this prediction. Most commonly, we use the Strehl ratio, the Ensquared Energy in a pixel of a given size or the full width at half maximum of the point spread function. One also uses simulations characterize the components to be used (for example, how many deformable mirrors, with how many actuators...).

Tools

Hardware tools

This year, cluster power was increased by increasing number of machines. Now, we have a total of 76 3.2 GHz slaves, and a dual 3.0 GHz Xeon master. Each slave has 2 GB of RAM, so a total of about 150 GBytes of RAM. The PCs are connected with a Gigabit switch.

Below is a picture of the cluster:

Software tools

We use mainly a parallel C code written in house. The code is written in such a way that it minimizes the memory use on each PC. Therefore, all matrices are distributed among the different machines.

We also use, for quick analytical calculations Cibola, written by Brent Ellerbroek. It is freely available here.

Examples of recent simulation tool upgrades

Recently, several features have been added to our main simulation tool. The pictures below show some of these improvements:

The figure on the left shows the elongation of the spot produced by a laser guide star, when the laser launching telescop is located at the edge of the telescope pupil.

The above image (left) shows the semi-analytical simulation of a 2 stage AO system, designed to achieve a high contrast, to find extra-solar planets. The plot on the right shows the optimal modal gains for a 32x32 sub-aperture AO system (diamonds) and the (arbitrarily scaled) singular values of the interaction matrix.

Simulated systems

AOF-GALACSI: Simulations for the Adaptive telescope review
AOF-GRAAL: Simulations for the Adaptive telescope review
FALCON: Code comparison between Cibola, ONERA, OdP
VLT-Planet Finder: Analysis on stroke, number of permissible dead actuators, PYR
MAD: Various performance estimates, set up of comparison with lab
HOT: Phase screen parameters defined
Owl Blue Book: Single Cconjugate AO, Ground Layer AO, Multi-Conjugate AO, Laser Tomography AO, Multi-Object AO (MOMFIS), High Contrast AO (EPICS)

About 50% of our time is spent on ELTs and 50% on VLT activities

Available PSFs

50m ELT PSFs: (Science and AO working groups and ELT design study, WP9500, password protected, please contact M. Le Louarn to have access). Here you can find PSFs for GLAO and LTAO systems, and for comparison, the corresponding seeing limited PSFs. The simulations were made on a 50m telescope, at different wavelengths and for a seeing of 0.8" (at zenith, 0.5 µm).

42m ELT PSF: coming soon.

People

The simulators
Miska Le Louarn	General AO simulations
Christophe Verinaud	High contrast imaging, Pyramid WFS
Visa Korkiakoski	Optimal control