II. D. The Differential Delay Lines
| HOME | INDEX | SEARCH | HELP | NEWS |
|
|||||
|---|---|---|---|---|---|
II. D. The Differential Delay Lines |
|||||
|
|
To ensure the corrections of the errors measured by PRIMET and the FSU A, PRIMA must be able to bring the differential corrections after the normal delay lines. To realize it PRIMA will use the Differential Delay Lines (DDL) that will be located in the laboratory. Concerning this subsystem, we have to see what are its objectives, the technical requirements it has to perform, its general principle of working (and the comparison with the normal delay lines) and what are the eventual next milestones of this subsystem.. The DDL are located in the laboratory of the VLTI, just after the main DL and before FSU (cf fig.1):
Warning! The Netscape and Mozilla navigators don't recognize the greek letters, that could disturb the good reading of this page. |
Fig. 1: Link between the DDL and PRIMA
The differential delay lines shall compensate the small (0 to 65 mm) optical delay between both star OPD by introducing a supplementary delay (like the delay lines), and follow the evolution of this differential delay with earth rotation in real time.
Fig. 2: General Scheme of the DDL
It will be realized by using a retro-reflecting optical system placed on a translation device (cf fig. 2). The system could be similar to the one of the normal delay lines. However, it is aimed at compensating another kind of supplementary OPD. In a first phase of PRIMA, they will not be available yet and the facility will use the delay lines to replace them.
These DLL will bring some supplementary accuracy to the whole system, that are resumed in the requirements of PRIMA.
The DDL technical critical points are the required high stability in its movement of translation:
a) Maximal delay :
If we choose two stars separated by a = 60 arcsec with a baseline of B = 200 m, we can then calculate the wanted delay needed to compensate the created OPD (cf fig.3):
Fig. 3: DOPD between two stars seen from the telescopes T1 and T2 (with a baseline B)
thus we have DOPD = B.s (where s is the unit vector of the direction of the pointed star and B the baseline vector)
And so DOPD = B.sin a = 5.82*10^-2 m ~ 6 cm
PRIMA requires from the DDL a delay variation of at least 6cm. And if the design is compatible: a 12cm goal.
b) Identity:
In order to to limit the introduced internal differential OPD and to ensure the symmetry of the full PRIMA system, the DDL must be identical withi 5% on all characteristics.
c) Resolution:
The supplementary delay must be introduced with a great accuracy: the goal for the OPD resolution is to reach < 1 nm.
d) Absolute OPD repeatibility:
The repeatibility (i.e. the capability to reach the same position at two different times) has to be < 200 nm for the OPD (corresponding to 100 nm in the mechanical position).
e) The translation movement:
The translation system must be very smooth and straight, the DDL optical axis shall be guided within accuracy such that the optical axis remains in a cylinder of: < 15 µm diameter.
f) Vacuum environment:
A specific requirement for the DDL is that they must be able to work in vacuum environment (in particular in order to avoid LAD effects that introduce measurement biases).
g) Pupil relay:
The pupil must be relayed to the same position as without using DDL so that the use of DDLs is transparent on the optical point of view.
i) Bandwith:
The function of the DDL shall be preserved over the spectral range specified for VLT: 0.420 µm to 28.0 µm. For the operation of the delay lines, four spectral bands have to be distinguished within the specified spectral range, they are:
| VIS | 0.400 µm to 0.800 µm | mean 0.600 µm | Visible |
| NIR | 1.0 µm to 5.0 µm | mean 2.2 µm | Near Infrared |
| N-band | 8 µm to 13 µm | mean 10 µm | Thermal Infrared |
| Q-band | 16 µm to 28 µm | mean 20 µm | (possible later addition) |
The DDL shall fulfil the performance specified on the Near and the Thermal Infrared and be compliant, as a design goal, in the Visible. Moreover, all DDL mirrors shall be coated with protected silver or hard gold. At delivery, the DDL (excluding the possible vacuum vessel windows) shall meet the following efficiency requirements in the following table, over the specified spectral range:
|
Wavelength range
|
Absolute Troughput
|
|
|
a
|
0.4 < l
< 1 µm
|
> 0.80 (design goal)
|
|
b
|
1 < l
< 2.5 µm
|
> 0.80
|
|
c
|
2.5 < l
< 28 µm
|
> 0.90
|
|
d
|
l metrology
|
> 0.80
|
Until here the detailled principle and the design of the DLL has not been fixed yet. However it will probably have similarities in some point with the principle of the main delay lines with some improvements due to their small size.
a) The general principle:
The DDL have to receive both of the beams of the Primary and Secondary stars (PS and SES) and to send them to the FSU. The normal delay lines (DL) of the VLTI were composed of 4 holes: two for the entrance of the PS and the SES (these at the top of the DL) and two for the exit (the two others): cf photos on the delay lines page. To receive and send a beam, the DL use the "Cat's eye" system fixed on a mechanical system in charge to translate them on an axis (cf fig.4).
Fig. 4: The Cat's eye
The incident beam arrives on a parabolic mirror and sent to its focus, where is the Variable Curvature Mirror (VCM), a spherical mirror with a variable curvature that sends the beam back to the parabolic mirror at the exact point where it can exit of the "Cat's eye" (cf fig.4).
For the moment, without the DDL a solution could be found by using the normal delay lines. Indeed, instead of sending the PS and SES to the DDL (each with one entrance and one exit), we could send the PS to one DL and the SES to an another DL, and the same thing for the beams from the other telescope. Finally, by using 4 different DL we could correct the additional delay calculated by PRIMET. But even if such system is operationnal, il would need first the use of 4 different DL in one side, and in another side we would have clearly less good perfomances. Indeed the two beams don't "look" any more the same mirrors and the total optic path of the beams would be larger in this case including such significant problems: for example not any more the same vibrations in the same time, not the same conditions of turbulence and index variations ( dOPD would be larger because the air index of refraction is not good well known) etc...
b) The possible design:
The DDL have not been created yet but they will probably feature a cat's eye design, mounted on straight and stiff air bearings with a small piezo-actuator on the cat's eye secondary mirror for fine adjustements. The DDL design could also be combined with a beam compressor capability in order to gain reflections. Indeed the beams coming from the telescopes are 80 mm in diameter while most of the instruments accept only 18 mm beams. A compression is so needed and could be perfomed independantly from or combined with the DDL. THe other solution would be to have large delay lines what would be more expensive and present more problems (size, weight, vibrations of the mirrors...).
|
A consortium formed by the Observatory of
Geneva, Switzerland - Nevec, Leiden, Netherlands - Max Planck Institut fur
Astronomie, Heidelberg, Germany are working on a proposal to provide to
ESO the DLLs as a contribution in kind. The proposal should be submitted
to ESO council by the end of 2003 and in case of approval, the development
should take 2 to 2.5 years.
More informations about the DDL"PRIMA Technical Description and Implementation", F.Derie, F.Delplancke,A.Glindeman,S.Leveque, S.Menardi,F.Paresce,R.Wilhelm,K.Wirenstrand,Workshop "Hunting for Planets ", Lorentz center, Leiden University, 3-6 June 2002. Slides of the presentation.
For the other publications, see the part: References and links.
|
|
|
|||
|
|||
