A Milestone for the VLT Interferometer
First Underground Delay Line Now on Rails at Paranal
11 October 2000
Less than one month after "First Light" for the fourth 8.2-m YEPUN telescope (ESO PR 18/00), another special moment occurred at ESO's Paranal Observatory. This time, it was the first truly "underground" event, in the 168-metre long Interferometric Tunnel that has been dug beneath the platform at the top of the mountain. As one staff member remarked on this occasion, it was something like "the first scheduled trip of the Paranal metro"!
With the successful integration of the first Delay Line on Monday, September 25th, 2000, ESO has accomplished another important step towards the VLT Interferometer (VLTI). It will be followed by the integration of the second Delay Line by the end of November and the third is scheduled for February 2001; both are now in their final development phase in Europe.
"VLTI First Light" is then expected to take place soon thereafter, by means of two small special telescopes ("siderostats"). The combination of the light beams from two of the 8.2-metre Unit Telescopes will happen in mid-2001.
The VLTI Delay Lines
The VLTI Delay Lines form essential parts of this very complicated optical system. They serve to ensure that the light beams from several telescopes arrive in phase at the common interferometric focus. Details about how they function may be found in ESO Press Release eso9811.
In order to achieve the necessary performance, ESO has worked with two Dutch contractors, Fokker Space and TNO-TPD - Netherlands Organization for Applied Scientific Research - Institute of Applied Physics, to arrive at a totally new Delay Line concept. Another Dutch participant in the VLTI project is the Nova-ESO VLTI Expertise Centre (NEVEC), c.f. ESO Press Release eso0020.
The installation at Paranal
The last twelve months have been very busy for the integration team, with much preparatory work at the VLTI buildings for the final installation of the Delay Line systems.
The assembly of the translation mechanisms for the first two Delay Lines in the tunnel started in mid-2000. This included the alignment of their rails and supports to the extreme accuracy of about 0.25 mm over a total distance of 66.7 metres (ESO Press Photos eso0032a-b). To achieve such an unusually high precision, ESO - in collaboration with the French company FOGALE - developed a measurement system that is based on the water-level principle.
The delicate assembly and alignment of the critical sub-systems of the Delay Line were undertaken with the support of Fokker Space and TPD/TNO (ESO Press Photo eso0032e). Also for this, state-of-the-art methods were required in order to ensure a stringent performance of the system. This includes optical alignment of the optics with an accuracy at the arcsec level and positioning of the linear motors at the 0.01 mm (10 µm) level.
The Delay Line is one of the key systems in the VLT Interferometer. It is responsible for the compensation of the length of the optical path that is different from the individual telescopes.
Extreme accuracy needed
In the case of the VLT, this accuracy of the path length compensation must be within a tolerance of only 0.05 µm (0.00005 mm) over a distance of 120 metres. The present concept by ESO and the Dutch contractors is based on a retro-reflector (a "Cat's Eye") that is fixed on a carriage that runs on two stainless steel rails (ESO Press Photos eso0032c-d).
The motion on these rails is performed by a 60 metres linear motor and a piezo-transducer element. They are controlled by a laser metrology system that measures the instantaneous distances betwen the mirrors with the required accuracy.
This carriage is 2.5 metres long and weighs 250 kg. The total friction force is less than 50 grammes, thanks to the extreme accuracy of the rail alignment and special ball bearings. Because of this, the total power required for the Delay Line operation is only about 15 W.
The mirrors of the retro-reflector are made of aluminium by REOSC (France). They have been coated with a single layer of gold for the best possible reflection at infrared wavelengths.