title of abstract a - ESA Conference Bureau

ICSO 2014
International Conference on Space Optics
Tenerife, Canary Islands, Spain
7 - 10 October 2014
C. Engel1, M. Ferrari1, E. Hugot1, C. Escolle1,2, A. Bonnefois2, M. Bernot3, T. Bret-Dibat4, M. Carlavan3, F.
Falzon3, T. Fusco2, D. Laubier4, A. Liotard3, V. Michau2, L. Mugnier2
Aix Marseille Université, CNRS, LAM (Laboratoire d'Astrophysique de Marseille), UMR 7326, France
Office National d’Etudes et de Recherches Aérospatiales (ONERA/DOTA), Châtillon, France
Thales Alenia Space, Cannes la Bocca, France
Centre National d’Etudes Spatiales (CNES), Toulouse, France
The next generation of large lightweight space telescopes will require the use of active optics systems to
enhance the performance and increase the spatial resolution. Since almost 10 years now, LAM, CNES,
THALES and ONERA conjugate their experience and efforts for the development of space active optics through
the validation of key technological building blocks: correcting devices, metrology components and control
strategies. This article presents the work done so far on active correcting mirrors and wave front sensing, as well
as all the facilities implemented. The last part of this paper focuses on the merging of the MADRAS and
RASCASSE test-set up. This unique combination will provide to the active optics community an automated,
flexible and versatile facility able to feed and characterise space active optics components.
Partnership. Started ten years ago, the LAM space active optics activities are undertaken in partnership with
Thales Alenia Space (TAS) and ONERA, in close collaboration with the french space agency (CNES). In the
frame of R&D projects, these partners conjugate their experience and efforts for the development of space
active optics through the validation of well identified building blocks: correcting devices, metrology
components and control strategies and algorithms.
Space active mirror. In 2012, the MADRAS project successfully resulted in the design, integration and
characterization of the first active correcting mirror for space applications, able to achieve a correction
performance below 3nm RMS for each low order aberration, and below 9nm RMS for random phase maps of
amplitudes around 1µm PtV. The MADRAS experiment reached a TRL4 (See Sec II.B and Laslandes 2013 [1]).
Wave front sensing and control. The second step, achieved in the frame of the RASCASSE project, consisted
in the design, development and characterization of wave-front sensors dedicated to extended and dynamic
scenes. Two WFS were built and characterized, in pupil and focal plane. This experiment achieved exquisite
performance, around 10nm RMS of precision on the measurement of specific phase maps blurring extended
scenes (See Liotard et al [2], Bonnefois et al [3], this conference).
A versatile facility. The space active optics facility at LAM consists in the different building blocks of these
two complementary experiments. Not only that the natural next step will be the merging of RASCASSE and
MADRAS to test an entire active correction loop. The bench is also a versatile facility able to test any
deformable mirror technology with a 20-200mm pupil, and to test any WFS working in pupil or focal plane.
Both MADRAS and RASCASSE projects were dedicated to the case of 2-3m class orbiting telescopes, with a
1deg FoV and a 100mm relay pupil were the active correction occurs. The set ups are constituted of these main
building blocks:
- Point and extended sources
- Telescope simulators (static and dynamic generation of calibrated WFE)
- Relay optics for beam expansion or compression and mechanical interfaces
- Active mirror
- Commercial pupil wave front sensors (Shack Hartmann)
- Customized homemade pupil and focal plane wave front sensors
- Imaging cameras
A. Sources and telescope simulator
The extended source is a 1280x1024pxs OLED screen (mono or polychrome) able to send different extended or
moving sources, with a spectrum between 450nm and 700nm. Additional pinholes are also used for alignment
purpose and performance characterization in the case of star pointing. A set of objectives and collimation lenses
allows to transfer the beam and adapt its size to the aberrations generators.
ICSO 2014
International Conference on Space Optics
Tenerife, Canary Islands, Spain
7 - 10 October 2014
Two different solutions are proposed to send calibrated wave front errors to the active optics system. The
dynamic solution, adopted for MADRAS characterization, is based on a magnetic ALPAO DM88 [4], which
performance is better than the requirements. The static solution adopted for the RASCASSE experiment consists
in a rotating wheel carrying different phase screens (SILIOS technologies [5]) with calibrated WFEs from 40 to
200nm RMS. Deviation from specifications is lower than 4 nm RMS for each masks.
The different SILIOS phase screens allow sending WFE with only form content, only mid or high spatial
frequency contents, polishing errors and different combinations of these WFEs. This strategy is fundamental to
characterize the WFS performance on low order modes estimation with and without high order modes. The
amplitude of WFE are defined by the TAS system studies and directly etched on a glass polished transmitting
plate. Additional laser cut pupil masks can be added close to the pupil plane to simulate the telescope spiders
and secondary mirrors obscuration.
Point source (pinhole)
& extended source (OLED screen)
Phase screens
on rotating wheel
Collimating lens
+ motorized
translation stage
Fig. 1 Left: Illumination unit, Right: telescope simulator and zoom on the spider pupil
B. The MADRAS active mirror: closed loop performance.
The design and performance of the space active mirror are extensively described by Laslandes et al in [1]. The
active mirror is based on boundary actuation, ie the actuators influence is transmitted through the edges of the
mirror in order to avoid any actuator print through effect (see Fig. 2).
The Zerodur mirror is hold by an Invar warping harness and actuated with classical PZT. With a useful diameter
of 90mm, a total volume of 80x200x200mm3 and a mass of 4.0kg, the system is able to generate 24 modes with
a correction performance of 5nm RMS per mode and less than 10nm RMS on random phase maps.
Fig. 3 right shows the closed loop performance of the active mirror on each mode, while the specifications and
FEA results are presented on the left. Opto-mechanical interfaces can be adapted to receive and characterize any
active mirror from 20mm to 200mm in diameter.
Fig. 2. The MADRAS CAD and equipped prototype
ICSO 2014
International Conference on Space Optics
Tenerife, Canary Islands, Spain
7 - 10 October 2014
Fig. 3. Left: RMS of modes to be generated and performance obtained after FEA optimization.
Right: Closed loop performance.
C. The RASCASSE wave front sensors
RASCASSE aimed to design, realise and integrate two types of wavefront sensors, and compare their
performance in the case of extended moving scenes for different scenarios. The scenarios are defined by the
luminosity parameters, the contrast in the scene, and the content and amplitude of the WFE to be measured. The
bench was designed to provide a 1x1deg² FoV with an optical quality of 60nm RMS and a WFE variation lower
than  /100 over the entire field. Each optical component have been characterized using Fizeau interferometer
(accuracy<1nm RMS, repeatability<1 nm RMS). The static wavefront error induced by all lenses is less than 10
nm RMS.
The pupil WFS is a customized Shack Hartmann WFS, designed for the specific scenarios. It is made of a field
stop, a pupil relay and a 10x10 µlenses matrix. (Fig. 4 Right)
The focal plane WFS is based on the phase diversity method, and uses a lateral shift splitter to feed the camera
with two twin images, with a focus difference of 1.2 The Phase diversity channel is telecentric to provide a
constant magnification on the two images (Fig. 4 Left).
Both WFS use an ORCA2 camera with a water cooling system. Fig. 5 shows the images obtained with an
entrance grid of 7x7 point sources for the phase diversity, and a point source for the SH channel. Initial
performances were characterized using an extended scene: grid of 7x7 points covering all the field of view.
Initial aberrations are less than 60 nm RMS as specified. Variations on the field of view are around λ/25 with a
major contribution coming from filed curvature.
Lateral shift splitter
10² µlenses matrix
Pupil relay
Field stop
Focusing lens
Fig. 4 Left: The phase diversity channel. Right: The Shack Hartmann channel
ICSO 2014
International Conference on Space Optics
Tenerife, Canary Islands, Spain
7 - 10 October 2014
Fig. 5 Left: Phase diversity image obtained with a grid of 7x7 point sources; Middle: Shack-Hartmann
image obtained with a point source; Right: Total WFE over the field.
D. Automation of optical configurations, acquisitions and data storage
Specific GUIs were developed for both benches. The MADRAS GUI allows calibration of the system and
closed loop control as well as a display of actuators commands, WFS signal and imaging cameras.
All operations on the RASCASSE test bench have been automated for remote control:
 change of optical configurations: rotation of the phase wheel, translation of the collimation lens to move the
focal plane,
 control of the extended source: flux adjustment and change of picture,
 modification of the camera’s integration time,
 data storage, with automatic naming and recording of acquisition parameters (date, configuration, phase
mask, etc…)
E. Long term stability
The uncertainties on long term were determined using representative sequences, including rotation of the phase
wheel and translation of the collimation lens. Several sequences of more than 24 hours with more than one
hundred changes of optical configurations were done.
For this long term characterization, an extended scene was used: grid of 7x7 points covering all the field of
view. Using these acquisitions, the uncertainty at 3σ was estimated for astigmatism, coma and focus.
In addition, stability of the image position was checked by controlling the distance from the field of view center.
Concerning aberrations, the uncertainty meet the requirement of λ/100 (Tab.1) Variations of distance from the
center of field of view is less than a pixel, so that the image deformation can be neglected.
Tab. 1 Synthesis of uncertainties
Uncertainty at 3σ
±2.5 nm RMS
±1 nm RMS
±3,5 nm RMS
Distance from center of FOV
±5.5 µm
ICSO 2014
International Conference on Space Optics
Tenerife, Canary Islands, Spain
7 - 10 October 2014
RAPACE stands for Real-time Automated Platform for Active Correction of Earth orbiting imagers. This
facility aims at merging the MADRAS and RASCASSE test beds in order to propose an integrated facility to
the community able to test and characterize any type of active mirror or WFS (see a view of the facilities on
Fig.6). Work is ongoing to upgrade the existing hardware and software to tend to a versatile and flexible
operating system. The RAPACE preliminary requirements are listed in Tab. 2. The following building blocks
are addressed in priority:
Telescope simulator. The block corresponding to the telescope simulator must now merge both proven
technologies, namely ALPAO DM88 and SILIOS phase screens, to provide either static or dynamic calibrated
WFEs. An effort on the optical design is ongoing and will use the existing characterized pupil relays.
Active mirror and WFS. The LAM space active optics facilities is now an automated platform able to hold and
characterize any active mirror and wavefront sensor, for point or extended sources, with minor modifications on
the beam expanders and compressor if necessary.
Closed loop. The natural next step is the merging of the two benches in order to demonstrate a closed loop
performance using the RASCASSE customized WFS instead of commercial ones used on MADRAS.
Automation. A general effort is put on the harmonization of the control panels and software, in order to provide
a versatile bench able to test any component.
Tab. 2 RAPACE preliminary requirements
System requirements
Number of pupil relays
Active optics pupil relay size
[20mm – 200mm]
Temperature control
Vibrations control and local turbulence
Image stability <1pix
Bench specifications
Static WFE
<60 nm RMS
WFE variation vs FOV
WFE stability
Stray light
30 to 150
WFE generation specifications
Low order modes
>1µm PV for each of the first 36 Zernikes
WFE generation accuracy
<5 nm RMS per mode
Wave front sensing specifications
WFS type
Pupil and image WFS
WFS accuracy
<10nm RMS on random phase maps
Correction specifications
WFE correction accuracy
<5 nm RMS per mode on the first 17 Zernikes
< 10nm RMS on random phase maps
ICSO 2014
International Conference on Space Optics
Tenerife, Canary Islands, Spain
7 - 10 October 2014
Fig. 6. Up: The MADRAS bench with point source and dynamic telescope simulator, beam expanders and
pupil relays, interferometer, commercial SH WFS and imaging cameras.
Down: The RASCASSE bench with static telescope simulator, beam expanders and pupil relays, extended
illumination unit, and two parallel optical paths for WFS simultaneous comparison.
This work has been financially supported by the FUI program MADRAS (Fond unique interministériel), and
the CNES for the RASCASSE program. C. Escolle PhD is also supported by ONERA and the CNES.
[1] M. Laslandes, E. Hugot, M. Ferrari, C. Hourtoule, C. Singer, C. Devilliers, C. Lopez et F. Chazallet,
«Mirror actively deformed and regulated for applications in space: design and performance,» Opt. Eng.
52(9), 091803 (2013),
[2] Liotard et al, “Wave-front sensing for space active Optics: RASCASSE project”, this conference
[3] Bonnefois et al, “Results on the RASCASSE project”, this conference
[4] ALPAO – adaptive optics, www.alpao.com
[5] SILIOS, www.silios.com