2.1

Telescope design requirements

The optical requirements the telescope design has to satisfy are reported in Table 1.

Table 1.

Summary of the telescope optical requirements.

The collecting area is related with the minimum intensity (magnitude) of the observable targets. A collecting area of at least 0.6 m² is required, which implies an entrance pupil of the order of 1 m in diameter.

The required final as-built quality of the telescope, i.e. diffraction limited at 3 μm over a FoV of 30” , is equivalent to ask for an RMS wavefront error (WFE) of 220 nm.

To guarantee the required throughput without increasing the size of the primary mirror, that is the entrance pupil of the telescope, the optical design has to be unobscured. The unobstructed solution also assures the energy in the PSF is primarily contained inside the first Airy disk and not spread towards the secondary rings.

The characteristics of the instruments following the telescope, i.e. the FGS and the AIRS [8], have determined the wavelength coverage (0.55-8 μm) and the global FoV of the telescope. A diffraction limited FoV is needed for the target observations, a slightly larger FoV with suitable optical performance is required to allow the FGS to acquire the star and thus guiding the instrument, and a slightly larger unvignetted FoV is needed for background subtraction.

2.2

Telescope design characteristics

The baseline telescope design is an afocal unobscured eccentric pupil Cassegrain telescope (M1 and M2) with a recollimating off-axis parabolic tertiary mirror (M3). All the mirrors share the same optical axis. An M4 plane mirror is redirecting the exiting beam parallel to the back of M1 where the optical bench is located and the instrument will be mounted (see Figure 1) [8].

Figure 1.

Scale drawing of the telescope – view in Y-Z plane.

The center of the FoV of the telescope is inclined of 0.1° (in the YZ plane) with respect to the optical axis of the telescope defined by the mirrors common optical axis.

The system aperture stop/entrance aperture is located at the M1 surface. The M1 aperture is an ellipse with major/minor axes dimensions of 1100 mm x 730 mm. The complete characteristics of the optical design are summarized in Table 2a, while in Table 2b the telescope mirror parameters (radius of curvature, conic constant, off-axis, etc.) are described.

Table 2.

(a) Summary of the telescope optical design characteristics. (b) Mirrors parameters description.

2.3

Telescope optical performance

The raytracing analysis and design optimization have been done by means of the raytracing software Zemax^®. Being the telescope afocal, the spot diagrams can be derived using an ideal focusing paraxial lens with a defined focal length, or using the afocal image space option appropriate for systems with collimated output. Note that the spot diagrams obtained with this second method have their size expressed in milliradians.

The nominal diffraction PSF at 3 μm wavelength has an Airy radius respectively of 0.2 mrad and 0.29 mrad in the X and Y directions. A picture of the expected theoretical PSF is depicted in Figure 2a. The telescope RMS wavefront error is always less than 26 nm over the 30” nominal telescope FoV (see Figure 2b); this value is well below the telescope diffraction limit at 3 μm, i.e. 220 nm.

Figure 2.

In (a) PSF calculated at the telescope FoV centre for a wavelength of 3 μm depicted over a 1 mrad square box. In (b) RMS wavefront error field map calculated for the 3 μm wavelength over the 30” nominal telescope FoV. Units are λ.

3.1

Parameters for material selection

The selection of the substrate for the mirrors, in particular for the primary mirror which is the largest one, has to take into account many parameters. The first is the availability of the material in the proper form and size, since the ARIEL primary mirror is of about 1 m in diameter. Then the characteristics related to the handling/manufacturing/polishing process and, in particular, to the required surface accuracy and surface roughness have to be accounted for.

Use of an overcoat can be considered for materials that are difficult to polish bare. In addition the mirror has to be mounted and integrated in the telescope structure, so the coupling of the mirror substrate coefficient of thermal expansion (CTE) with that of the mechanical structure has to be borne in mind.

For the ARIEL payload, the impact of the temperature change between the ambient temperature, i.e. the one at which the elements will be manufactured, aligned and tested, to the operative temperature, of about 50 K, has to be considered. The selected material has to scale uniformly with temperature without static or transient distortion and at the proper operative temperature it’s required to be as stable as possible and with minimum gradients, the dimensional stability is very important since few adjustments can be made for static or transient distortion once in operation. A high thermal conductivity is required coupled to a as low as possible CTE at the operative temperature, that is a high thermo-optical stability is required.

Other parameters are related to the requirements on the minimum throughput that impacts on the mirror reflectivity and thus on the choice of the coating. It is worth considering which coatings can be deposited on the material. The temperature range of use and storage should also be accounted for in matching, as well as, the CTE of the coating to the substrate. If an overcoat is present also that thermo-mechanical coupling has to be evaluated.

Because the diameter of the primary mirror is quite big (~ 1 m), the effects of the gravity changes and the impact of the lightweighting of the substrate has to be taken into consideration together with the allowable total mass of the telescope. Therefore, the specific stiffness and specific strength of the material selected should be high in order to optimize this aspect of the design.

The materials that can be used for the ARIEL M1 mirror are rather limited. Those identified for the trade-off through brainstorming within the ARIEL Consortium and a review of the literature on other space telescopes [9] of a similar scale are:

Glass materials have very good polishing properties, but are difficult to mount. Beryllium is very difficult to handle. SiC structures suffer from moisture release. An aluminum mirror mounted on an optical bench with the same material allows for automatic temperature compensation.

3.2

Criteria for material assessment

Six criteria have been used to assess each possible material against:

The material properties assumed for each of the five possible materials are shown in Table 3.

Table 3.

Material properties.

(*)

This value is for ambient T CTE

As a conclusion of the material trade-off study for the telescope mirrors, and specifically for M1, considering the consortium provision of the telescope, the optimum solution is a telescope with mirrors and structure made from aluminum 6061 T651 alloy [7]. The viability of using aluminum as the baseline material for the telescope mirrors has been assessed during the phase A by producing the pathfinder M1 mirror.

5.1

Pathfinder mirror program (PTM) characteristics

The baseline scope for the PTM was to verify and assess, on a full-size demonstrator, the critical manufacturing technology to be adopted for the primary mirror realization. Given the time constraints and available funding, the project agreed to adopt a nominal spherical shape and focus on the demonstration of the diamond machining process used for the preparation of the Aluminum substrate for the subsequent figuring and polishing steps. As such, the surface error and roughness requirements for the PTM were relaxed with respect to the flight requirements but still very representative for the diamond machining step. In summary, the primary objectives of the PTM were the following:

5.2

PTM design and manufacturing

In order to minimize mass and complexity, the backside plane of the PTM is tangent to the nominal paraboloid at its decentering point. In this configuration, the mirror is geometrically symmetric. The nominal optical shape of the mirror is spherical with a radius of curvature of 2,400.725 mm, with a tolerance of ±2 ‰. A margin of 10 mm on the external dimensions has been added to account for possible edge effect during machining and polishing.

Triangular pockets ensure 36% light-weighting, setting the overall mass at approximately 156 kg. The mechanical interface configuration is very preliminary and consists of three pads equally spaced 120° apart and protruding from the external rim of the mirror. Each pad has a 12.5 mm through-hole each. A preliminary planarity requirement of 0.005 mm was set for these interface pads.

The first manufacturing steps consisted of standard CNC milling runs to create the light-weighting structure, the lateral perimeter with the interface pads and the spherical front surface (Figure 3). In order to leave enough material for the subsequent diamond machining step, the spherical surface was machined with a radius of curvature 0.7 mm shorter than nominal. The substrate was measured at Media Lario on the CMM, yielding a radius of curvature of 2,394.154 mm and residual error of 30 μm RMS or 190 μm PV. The substrate was then ready for diamond machining of the optical surface. Diamond machining was performed by LT-Ultra Precision Technology GmbH, Germany. The size of the mirror required some tuning of the machining strategy and speed in order to ensure adequate quality of the diamond tool throughout the entire machining process. In the end, the mirror was delivered to Media Lario according to the agreed schedule and quality (Figure 4).

Figure 3.

Front (left) and backside (right) of the PTM mirror prototype after milling.

Figure 4.

Two views of the PTM mirror surface after diamond turning.

5.3

PTM measurement results

In order to minimize the gravity sag during the final CMM measurement, the mirror was accommodated on a 9-point isostatic support system consisting of a simple whiffletree structure (Figure 5). This set-up had previously been analyzed to check that the gravity sag were at least one order of magnitude smaller than the surface figure error specification of the mirror. According to our simulations, the gravity sag with a 9-point isostatic support is 0.09 μm RMS, with 21 μm impact on the radius of curvature. We calculated that a more complex 18-point support would have only marginally reduced the gravity sag to 0.07 μm RMS

Figure 5.

The PTM mounted on the designed whiffletree structure to counteract gravity effects.

The CMM measurement showed a residual error of 1.98 μm RMS or 11.6 μm PV on the diamond machined optical surface (Figure 6). This value is approximately double the goal set for the PTM. However, we remark that half of the measured value is a low-frequency astigmatism error that can be reduced with further tuning of the diamond machining process and easily corrected by the subsequent figuring and polishing steps. The radius of curvature was measured at 2,403.404 mm, that is 1.1 ‰ longer than nominal, confirming that the substrate meets the requirement of the next figuring and polishing steps. The measurements are summarized in Table 4.

Figure 6.

Surface figure error of the PTM on the 9-point support: total residual error (left) and with astigmatism removed (right).

Table 4.

Measurements results of the PTM after diamond machining.

In order to experimentally assess the gravity sag, the PTM surface was measured both on the 9-point support and on the three standard interface pads. The latter is shown in Figure 7, while Table 5 compares the measurement results in the two configurations. The residual error of the mirror on its interface pads is smaller because the center of the mirror drops a little under gravity thus smoothing the low frequency error of the mirror.

Figure 7.

PTM measured surface error map when mounted on its 3 pads.

Table 5.

Results of the shape accuracy measurements of the ARIEL PTM optical substrate supported on its nominal interface pads.