1.

INTRODUCTION

The MTG (Meteosat Third Generation) program is a collaboration with EUMETSAT and ESA to perform and innovate the next generation of meteorological satellite system.

The constellation is made up of six satellites MTG-I and MTG-S. Each has one of the two key instruments: on the one hand, the Flexible Combined Imager (FCI) and, on the other, the Infrared Sounder (IRS).

The previous meteorological satellite, MSG, is a satellite in constant rotation able to obtain:

As for MTG-FCI, which is a fixed satellite with a more complex thermal environment, it will be able to obtain:

One of the main challenges on the evolution between MSG and MTG is higher optical resolution of the instruments. The optical solution for these instruments is a TMA (Three mirrors Anastigmat) which include three off-axis mirrors in a small volume with a good optimization of main optical aberrations. Optical quality for each mirror in terms of precision and stability are the two criteria guaranteeing high performance of the instrument.

Figure 1-1:

TMA optical path

Thales SESO participates in the development and manufacture of opto-mechanical components including thermal hardware on both FCI and IRS instruments. Regarding mirrors, Thales SESO has designed, produced and tested all the FCI and IRS telescopes mirrors with eight different designs including six off-axis mirrors.

In terms of final telescope quality, the driver is to have mirrors perfectly aligned. The alignment is based on the telescope WFE optimization. As the three mirrors are off-axis, it is impossible to find the real optical axis of the aspheric surface during the alignment. Therefore a very accurate knowledge of virtual position of optical axis versus a real mechanical referential is needed. From the delivery of the equipped mirrors as manufactured at Thales SESO level, alignment of the telescope requires a good knowledge of the optical interface (namely the apex of the mirror for aspherical mirrors) toward its mechanical reference frame. Accuracy of this knowledge directly impacts:

In this process, both the optical reference frame and mechanical one have to be agreed together with the customer to have a process allowing the customer to reposition the mirror in its theoretical position before starting the alignment process.

In the frame of MTG FCI and IRS TO mirrors, the choice from the customer (OHB) was to build the mechanical reference frame of each sub-assembly from conical holes machined on the side of the mirrors (3 conical holes per mirror).

For other programs the choice is made of the mechanical reference frame in the interface plane: references on the mechanical parts (MFDs) interfacing the mirror assembly to the telescope for example.

The challenges are then:

Whatever the procedure it has to be applicable for small mirrors to very large one.

On MTG the procedure was applied for the different mirrors ranging from about 60 mm diameter to 400 mm diameter. Also Thales SESO succeeded to apply the same procedure on the TANGO demonstrator for Thales Alenia Space, with a diameter up to 1.5 m.

2.

MANUFACTURING

Thales SESO has a strong experience in the design and manufacture of complex opto-mechanical components ranging from large space mirrors (including Ø 1m50 very lightweighted TANGO M1 zerodur mirror at 25 kg/m² and down to M2 MTG TO Ø 60 mm mirrors).

2.1

Anticipate the manufacturing stages

Upstream of manufacture, the definition of the design must take into account many aspects related to the critical subjects of manufacture such as reference areas for measuring the component.

To guarantee the centering and orientation specifications of each component, we guarantee the consistency of the measurement reference system during the product life cycle.

Anticipating design changes during manufacturing is a key step in ensuring the centering specification (between 50 microns and a few hundreds of μm depending on the mirror sensitivity regarding the decentering) of the component’s optical surface. The second point is to design and develop specific tools for the measurement and integration of complex parts (To guarantee a fast and stable integration of the component before and during the measurement).

Figure 2-1:

View on the difficulties of access to measure the mechanical reference frame after the final integration (M2 mirror FCI)

3.

MEASUREMENTS

Before carrying out a measurement campaign, Thales SESO makes sure to offer either a standard measurement solution or, as here, to carry out a specific procedure. For this sequence we must carry out a characterization campaign of each final optical component surface

The sequence in place is the same for all mirrors, only the technical programming parameters obviously change depending on the size of the component and the measurement accuracy required.

3.1

Definition of the measurement means

Thales SESO is equipped with the high-precision Leitz three-dimensional machine equipped with a non-contact sensor that can measure flat or free-form optical surfaces.

Figure 3-1:

view of the PMM-C Leitz Machine

This machine is able to take measurements of optical parts up to 1600 mm and guarantee measurement accuracy of 0.9μm+L/650mm.

Thales SESO is able to accurately link the location of the optical surface with the location of the mechanical referential. Firstly, by measuring each mechanical part with conventional ball probes (flatness and location interfaces) and by measuring each optical surface without contact @Precitec probe. This non-contact sensor makes it possible to scan optical surfaces at high speed without losing measurement precision. The @Precitec sensor reduces measurement uncertainties due to probing. It enables a large number of acquisitions to be acquired while guaranteeing good measurement repeatability

Figure 3-2:

View of the @Precitec Sensor

3.3

Accuracies of characterization measures

Through a specific metrology scheme, including accurate scanning of the optical surface Thales SESO delivers to the customer a reliable and accurate location of the optical reference frame of each sub-assembly toward its mechanical reference frame.

Thales SESO developed a dedicated software to build the tolerance budget of the inspection setup, depending on the mirror design. Accuracy includes using dedicated software to match the “as built” mirror (mechanical references vs optical axis) to the theoretical one.

From these relative location, the customer is able, in its assembly process, to “plug” the sub-assembly directly in its nominal position to start the alignment process.

With this approach and corresponding accuracy, the mirror in final use can be positioned on the telescope with an optical surface centered to its theoretical location with a WFE residue deviating less than a few hundred of nm from expectation.

This was confirmed both on MTG mirrors (FCI and IRS), for which starting from a perfectly characterized mirror location, the alignment could start directly with interferometric measurement. Therefore it could go directly to the very accurate adjustment without need for preliminary alignment going through sophisticated procedure with intermediate steps and verifications.

For MTG, a combined effort (OHB and TSESO) was made to achieve a precise opto-mechanical characterization between as-built apex position and mechanical interfaces.

This procedure allowed OHB to reach for example a WFE of 330 nm rough from mounting on the overall telescope assembly for the first FCI telescope as mentioned in some mail reproduced below (mail form OHB 30/05/2018):

“I wanted to share some good news from Oberpfaffenhofen with you! The OHB AIT team has mounted the mirrors and the WFE is of the order of 330 nm (which is quite a good result for the first placement ☺). I take it as a proof that our efforts on the characterization were well invested, so my special thanks goes out to those who were working on this slightly tedious subject! “

Figure 3-3:

View of the preliminary alignment at FCI telescope level (mail form OHB 30/05/2018)

4.

RESULTS

The results obtained by Thales SESO make it possible to correlate with great precision the location of the optical surface with the location of the mechanical reference frame. Given the good knowledge of Thales SESO measurements, for MTG FCI, starting with the preliminary alignment at 330 nm shown above, OHB was able to converge in one week (between May 30^th 2018 and June 7^th 2018) to the final performance of 68 nm RMS whereas the target was in the range 100 nm at this step, including alignment residues (mail from OHB 08/06/2018).

Figure 4-1:

View of the final performance at FCI telescope level (mail from OHB 08/06/2018)

The same approach for TANGO demonstrator (1.5 m large telescope) leads to the following comment from Thales Alenia Space : The active telescope concept allows to simplify the assembly and integration phase. Several complex alignment ground support equipment are no more required. For instance, the telescope integration and alignment of the secondary mirror becomes one order of magnitude relaxed and is managed by theodolite means during mechanical integration thanks to the M2 5 Degree of Freedom mechanism. (Thales Alenia Space mail January 2021).

5.

CONCLUSION

Alignment and integration off-axis mirrors are two complex works that require many iterative loops to ensure that each opto-mechanical component is properly measured and integrated to reply customer specifications. Our main results concern a good match between the measurements carried out at Thales SESO and the alignment prediction then the instrument integration loop at the customer.