This invention relates generally to optical alignment techniques for telescopes, for example space-based telescopes.
The optics of a telescope have to be accurately aligned with respect to each other in order for the telescope to work properly. Alignment can include placing the individual optics, placing segments of a single optic (large telescopes often have segmented primary mirrors), and controlling the shape of a deformable mirror. For diffraction limited telescopes, some of the optical elements have to be positioned with respect to one another to within a fraction of a wavelength of light. As used in this specification references to light include non-visible light e.g. infrared (IR) or ultraviolet (UV) light.
This specification describes alignment techniques for telescopes. The techniques described are useful for all telescopes but can be particularly useful for space-based telescopes.
A telescope may need some initial alignment, e.g. if it is unfolded and deployed in space. After the telescope is accurately aligned the alignment will need to be maintained. Alignment drift can be caused by many things including variations of temperature, variations of gravity vector, wind buffeting, and acoustic vibrations. Accurate alignment can be especially important for imaging applications as these may require a large-area flat field.
Most telescope structures employ passive techniques to deal with these variations. Some telescopes use a distant external light source, e.g. a star, to measure wavefront error i.e. alignment error. However this is not suitable, e.g., for a space-based telescope imaging the earth.
A telescope can be aligned using a flat mirror placed ahead of the telescope so that light from an internal source will return back through the telescope. If the flat mirror is perfect, the measured wavefront errors will only be due to the telescope. However this can be inconvenient, and the telescope cannot be used to view distant objects while the flat test mirror is in place.
McCann et al., “A new mirror alignment system for the VERITAS telescopes”, Astroparticle Physics 32 (2010) 325-329, describes a system for aligning a telescope using a star. The telescope has a segmented mirror and the mirror segments are referred to as “facets”. Vazquez-Montiel, “Method For Facets' Alignment For The High-Flux Solar Furnace At Cie-Unam In Temixco, Mexico. First Stage” http://www.concentracionsolar.org.mx/images/pdf/memorias/05_sp12 sv.pdf, also describes a solar concentrator with a segmented mirror in which the mirror segments are referred to as facets. Both these systems lack mirror facets as defined herein. Kishner, “High-bandwidth alignment sensing in active optical systems” Proc. SPIE 1532, Analysis of Optical Structures, 1 Dec. 1991, describes an optical alignment system that attaches retroreflectors to a primary mirror; since these reflect back along the incoming light path they do not need angular alignment. Zhao et al., “Simultaneous multi-piston measurement method in segmented telescopes”, Optics Express 24540, Vol. 25 (20), 2 Oct. 2017, describes the use of modulation transfer function sidelobes to measure mirror segment piston error. WO2021/018634 describes a diffraction grating-based alignment system for a telescope with a segmented mirror. Further background prior art can be found in U.S. Pat. No. 5,274,479 and US2018/0074236.
In one aspect there is therefore described a telescope with an optical alignment system. The telescope comprises a segmented primary mirror, the segmented primary mirror having a plurality of mirror segments. Each mirror segment has a segment surface defining part of an imaging surface of the primary mirror. The imaging surface of the primary mirror may, but need not be, the first surface of the telescope.
Each mirror segment has a mirror facet. A direction of light reflected from the mirror facet defines an orientation of the segment surface of the mirror segment, i.e. the facet has a fixed orientation with respect to the segment surface. Thus an orientation of the segment surface of the mirror segment can be determined from the direction of light reflected from the mirror facet.
In implementations the, i.e. each, mirror facet is configured to reflect light in a different direction to light reflected from the segment surface of the mirror segment. In general the light reflected from the mirror facets converges towards a metrology system.
A light source, in implementations an internal light source, is configured to illuminate the mirror facet of each mirror segment. In this context an internal light source may be a light source which is not located beyond a front aperture of the telescope.
As indicated above, a metrology system located to receive light from the light source after reflection by the mirror facet of each mirror segment. The received light characterizes an optical alignment of the mirror segments. The metrology system may include an imaging system, e.g. a camera, to capture an image of the received light, such as locations of the reflections from the mirror facets and/or interference fringes generated by overlapping reflections.
Some implementations of the telescope provide excellent sensitivity to various types of mirror segment misalignments. The telescope may be configured to automate the alignment process.
In another aspect there is provided a method of aligning a telescope, in particular as described above, the method using a light source to illuminate the mirror facet of each mirror segment, and moving one or more of the mirror segments to align a reflection from the mirror facet of each mirror segment.
The aligning may include correcting tilt misalignments of the mirror segments, by tilting one or more of the mirror segments such that reflections from the mirror facet of each mirror segment are coincident. Also or instead the aligning may include correcting translational misalignments of the mirror segments along an optical axis of the telescope, by moving one or more of the mirror segments in piston such that reflections from the mirror facet of each mirror create interference fringes.
The mirror segment surfaces may also be provided with a diffraction grating, illuminated by a monochromatic light source to generate a diffraction pattern. The method may then include determining one or more characteristics of the diffraction pattern, and correcting a mirror segment misalignment by moving one or more of the mirror segments in response to the one or more characteristics. Various characteristics of the diffraction pattern may be measured, depending on the correction to be made.
These and other aspects of the invention are further described, by way of example only, with reference to the accompanying Figures.
In the figures like elements are indicated by like reference numerals.
In
A facet may have a flat or a curved surface. For example the surface of a facet may be curved, e.g. it may have spherical curvature, to focus light from the light source onto the metrology system. Typically an area of the facet 104 is small compared with the total reflective surface area of the mirror segment, e.g. less than 10% of the total surface area. Thus generally as used herein a facet is a part of the segment surface that directs light in a different direction to the segment surface e.g. it may have a focus that is displaced away from a focal surface of the telescope. As used herein, references to the segment surface are to the part of the segment surface defining part of the imaging surface of the primary mirror i.e. the part of the segment surface other than the facet. Sometimes this may be referred to as the main part of the mirror segment surface.
Various techniques may be used to fabricate a mirror segment with a mirror facet. For example diamond turning or other precision freeform CNC (Computer Numerical Control) manufacturing methods may be used.
Rays 114 indicate light from a distant object viewed by the telescope. These are focused by segmented primary mirror 102 towards a focal point 106 of the telescope where, for example, an imaging system 108, or other optical system, is located.
The telescope includes a light source 109, in implementations an internal light source, and a metrology system 110. Here an internal light source is one which is not located beyond a front aperture of the telescope. Light from light source 109 will be referred to as metrology light. In implementations the light source is a point source of the metrology light.
In the example of
In
The optical path defined by reflection of rays 114 at the primary mirror 102 is different to the optical path defined by reflection of rays 112 at primary mirror 102, because the metrology light is reflected by the facets whereas the rays 114 from the distant object are reflected by the main part of each mirror segment. In general the mirror facet 104 is configured to reflect light in a different direction to light reflected from the main part of the mirror segment surface.
In general the imaging surface of the primary mirror 102 is configured to focus light to a focal surface or focal plane. Typically but not necessarily it is the first surface of the telescope and defines a light collecting aperture for the telescope. There can be many optical elements between the light collecting aperture and an imaging region of the telescope; in the Figures herein these are omitted for simplicity.
One advantage of directing metrology light using reflective facets rather than, say, a diffractive approach is that reflection is compatible with use of a broadband light source. As described later a broadband light source can be useful in aligning the mirror segments in piston.
The metrology light can be used to determine an alignment of the mirror segments 102a-c, and in implementations also to automatically align the mirror segments to one another. Thus in implementations the telescope comprises a set of actuators 120a-c and each of the mirror segments (or at least all except one of the mirror segments) is provided with a respective actuator to move the mirror segment. For example a mirror segment may be moveable by an actuator in tip/tilt and/or piston.
In implementations a mirror segment actuator comprises a three point (“Maxwell-type”) mount with three linear actuator devices spaced around an edge of the mirror segment. In such an arrangement moving any one linear actuator device rotates the mirror segment in tip and/or tilt about an axis running through the points of contact made by the other two linear actuator device, and piston adjustments may be made by moving all three linear actuator devices in unison.
In implementations the actuators 120a-c are controlled by control signals from an actuator control system 130 coupled to the metrology system 110, to align the mirror segments. More particularly the metrology system is configured to capture an image of the light from the light source after reflection by each of the mirror facets, and the actuator control system is configured to control the set of actuators dependent upon the captured image to move the mirror segments to align the reflections from the mirror facets of the mirror segments.
The mirror segments may be aligned to stack sub-images from the mirror segments and/or to co-phase the mirror segments, i.e. to align the mirror segments to reduce optical path differences (OPDs) between the segment surfaces of the mirror segments and a focal point of the telescope to less than λop, where λop is a wavelength of operation of the telescope. This is described in more detail below.
As a preliminary to an alignment procedure for the telescope, depending on an initial degree of alignment, each of the mirror segments may be scanned in tip/tilt so that the image sensor of the metrology system sees a reflection from each facet. The mirror segments may also be co-focused i.e. moved in piston to align the mirror segments to within a depth of focus of the mirror segments e.g. ˜ 0.1 mm. This may, but need not, involve using the reflected light from the mirror facets. For example, where the facets have a curved surface to focus the metrology light onto the metrology system the mirror segments may be moved in piston to maximize the contrast of an image of the metrology light source from each facet, e.g. by minimizing a gradient of the Modulation Transfer Function (MTF) of the image of the light source from the facet.
The automatic alignment procedure starts by using the actuator control system to control the set of actuators, dependent upon the image of the metrology light captured by the metrology system, to tilt the one or more mirror segments to align the reflections from the mirror facets of the mirror segments such that the reflections are laterally coincident with one another in the captured image, i.e. to stack the sub-images from the mirror segments. (Although “tip” and “tilt” are generally used to refer to rotation about two orthogonal axes in this specification sometimes “tilt” is used as shorthand to refer to rotation about either axis, for conciseness.)
In some implementations a position error can be calculated and the corresponding mirror adjusted to correct this. In some implementations a closed-loop control system can be used to control the actuators to align the sub-images.
Ideally tip/tilt misalignment of the mirror segments should be reduced to less than λop/D where D is a (maximum) lateral dimension of the primary mirror 102, for diffraction-limited angular resolution. A fine-alignment process can be used to improve correction of the misalignment. This involves separating each of the sub-images in order that the lateral position of each can be more accurately measured, as accurate measurement of their positions becomes difficult when they start to overlap.
One way to do this is to use a shutter to select individual mirror segments to determine the sub-image position. Another way is to spatially separate the sub-images using an additional detector e.g. a second image sensor, in a defocussed plane. This latter approach has an advantage that all the mirror segments can be measured and corrected simultaneously; it also facilitates fine-alignment process using a closed-loop control system. In such arrangements the metrology system can include a first image sensor to capture the image of the received light from the light source after reflection by each of the mirror facets, and a second an image sensor configured to capture a defocussed image of the received light. When the reflections are coincident with one another the reflections captured by the defocussed image are spatially separate from one another.
Once the sub-images have been stacked the mirror segments can be co-phased to reduce the OPD between them due to misalignment in piston i.e. misalignment along an optical axis, e.g. z-axis, of the telescope. More particularly the actuator control system controls the set of actuators dependent upon a captured (focused) image of the light from the light source, to move the mirror segments in piston to obtain interference fringes in the captured image.
The interference fringes are created by interference between the overlapping sub-images of the metrology light from each of the mirror segments. In implementations, in order to use interference fringes to reduce the OPD between mirror segments the light source employs a broadband light source, e.g. a white light source such as a white LED, so that fringes only appear when the OPDs are matched.
More specifically in implementations the light source has a coherence length that is less than 1000 μm, 100 μm or 10 μm, for example of order 1 μm. Fringes only appear when the OPDs are matched to within the coherence length of the light source. In implementations where the mirror segments are already aligned to within a depth of focus of the telescope the light source may have a coherence length that is less than the depth of focus of the telescope. In one example, where the facets define a spherical curved surface, the fringes can appear at a centre of curvature of the facets.
In implementations the actuator control system is configured to control the set of actuators to maximise a measure of visibility of the interference fringes. The visibility of the interference fringes may be defined in a standard manner, e.g.
where Imax and Imin are respectively the maximum and minimum fringe light intensities. As described below, in some implementations the measure of visibility of the interference fringes comprises a measurement of a height of one or more sidelobes of a modulation transfer function of the interference fringes.
In principle monochromatic metrology light of just two different wavelengths will generate interference fringes with a varying visibility, but then the appearance of fringes does not define a unique position for each mirror segment. The use of multiple monochromatic wavelengths can narrow down the range of uncertainty. The degeneracy is broken by using a broadband source as then fringes only appear if the OPDs match to within a coherence length of the light, which is very small. Thus the system can be arranged so that there are no fringes if the OPDs do not match to better than lop. The broadband light source may e.g. have a FWHM (Full Width at Half Maximum) bandwidth of greater than 10 nm, 50 nm or 100 nm. A shorter coherence length, broader bandwidth source can provide a more accurate piston error measurement.
To identify a position when an OPD is matched all the mirror segments except one may be fixed and the position of one of the mirror segments scanned in piston until fringes appear. These fringes are formed when the OPD between the scanned mirror segment and one of the other mirror segments is sufficiently small, and comprise parallel fringes oriented perpendicular to a line between the mirror segments. When the mirror segment is scanned another set of parallel fringes is obtained, rotated at an angle to and crossing the first set of fringes, corresponding to a rotational angle between pairs of mirror segments generating the fringes and generating a criss-cross pattern. Successive mirror segments may be scanned in this way generating a pattern of dots when the OPDs are all matched.
Thus the actuator control system may control the set of actuators to obtain overlapping sets of fringes in which each set of fringes has an orientation direction defined by an orientation of a respective pair of the mirror segments. The actuator control system may control the set of actuators to obtain overlapping fringes of multiple different orientations.
As explained above, as a mirror segment is scanned an interference pattern emerges. Where the light source comprises a broadband light source the zeroth fringe is uniquely identified, as then the zeroth fringe has a unique location of maximum visibility.
However in one variant, rather than using broadband light the light source may comprise a source of three or more different monochromatic wavelengths, or a source of at least two separate wavelengths one of which is tunable. Then interference fringes may be sought, e.g. by the above scanning process, using two of the wavelengths and when interference fringes are found one of the wavelengths may altered, by substituting another of the monochromatic wavelengths or by changing a tuning of the wavelength. When the zeroth order position has been located the visibility of the fringes should be substantially unchanged, but if a different fringe order has been found the fringe visibility will reduce or the fringes disappear.
Thus where the light source comprises light of two different monochromatic wavelengths the degeneracy can also be broken by changing one of these wavelengths. For example the actuator control system may be configured to control the set of actuators dependent upon the captured image to move the mirror segments in piston such that the interference fringes remain substantially stationary when the wavelength of one of the different monochromatic wavelengths, e.g. the wavelength of a tuneable light source, is changed.
As described above, one way to measure the visibility of the fringes is to use the modulation transfer function (MTF) of the metrology system. As the MTF is defined by the (magnitude of the) Fourier transform of the PSF (Point Spread Function), where the light source is a point source the MTF can be determined from a Fourier transform of the light captured by an image sensor (in a focused image plane) of the metrology system. More specifically the height of the MTF sidelobes is a measure of the visibility of the fringes, and the visibility can be maximised by maximising the sidelobe height.
For a phase difference β between a pair of mirror segments, due to piston misalignment, the OPD between the mirror segments is 2β (due to the reflection). As β increases the fringe phase (position) varies cyclically with period λ/4, and with broadband light the fringes are gradually smeared out.
The sidelobe height, H, can be calculated as:
where A(λ) is the spectrum of the metrology light intensity and the integral is over a wavelength range for the metrology light, e.g. from a minimum wavelength λmin to a maximum wavelength λmax that together define wavelength limits of the metrology light source.
In one example alignment process mirror segment pairs are systematically scanned in piston to find interference fringes. For example with N mirror segments this can involve keeping a reference mirror segment stationary and scanning the remaining N-1 segments. Thus the set of actuators may be controlled to move multiple each of the N or N-1 mirror segments sequentially in piston to obtain interference fringes from each pair of mirror segments amongst the set of N.
Once fringes are found the OPD of a mirror segment pair is minimised by maximising the MTF sidelobe height e.g. by finely scanning a mirror segment position. This can be done by scanning just one, reference mirror segment in piston to identify where the sidelobe height is a maximum, Hmax, for each of the other N-1 (fixed) mirror segments, to measure the relative piston error between the scanned mirror segment and each of the other mirror segments. For each fixed mirror segment the piston location of Hmax (for the MTF sidelobe corresponding to interference between that mirror segment and the reference mirror segment) measures the piston error to the reference mirror segment (β=0). Thus in implementations the actuator control system moves one the mirror segments whilst keeping the other mirror segments fixed to obtain position correction information for each of the other mirror segments. The actuator control system can control the set of actuators using the position correction information to correct the relative positions of the mirror segments in piston, i.e. to correct the measured errors. This can reduce OPDs to β<<1.
In some implementations the light source includes a fine alignment light source to provide fine alignment light with a coherence length shorter than a coherence length of the metrology light used initially to find the fringes. The actuator control system may then be configured to use interference fringes from the fine alignment light after an initial coarse alignment, i.e. after obtaining interference fringes from the metrology light, to control the set of actuators to refine alignment of the mirror segments in piston.
Optionally, during the above-described co-phasing process tip/tilt alignment of the mirror segments may be rechecked at intervals, and corrected if needed.
An advantage of reflective facets is that they are compatible with the use of broadband metrology light. However a diffractive approach can monitor the main, imaging part of each mirror segment surface and, in implementations, substantially the entire imaging surface of the primary mirror. In some implementations both reflective and diffractive metrology may be used to obtain complementary information. In this case the mirror facets and diffraction grating may be arranged so that the sub-images/spots they respectively generate are spatially separated from one another at an image sensor of the metrology system, to avoid mixing signals from the reflective and diffractive approaches. Also or instead the separation may be achieved spectrally (using different wavelengths) or temporally (e.g. by time-multiplexing).
Thus one or more of the mirror segments in a telescope as described above may have a diffraction grating on the segment surface, e.g. part of a pattern of concentric rings that, as a whole, extends over the set of mirror segments. The diffraction grating may be faint, in that it may diffract only a small proportion of the metrology light e.g. less than 10%. Such a diffraction grating may also be formed by diamond turning, or by lithography e.g. e-beam or UV/optical lithography, or by using holography to manufacture a coating. The light source may include a monochromatic light source, e.g. a point source, e.g. a laser light source, for diffractive characterization of the primary mirror, and a diffracted light metrology system; these may be located at conjugate optical positions defined by the diffraction grating. The width of the lines and pitch of the grating may be chosen to direct light from the light source to the diffracted light metrology system.
The metrology system may include a diffracted light metrology system to characterize the diffracted light from the monochromatic light source for determining a degree of optical alignment of the telescope, and optionally for aligning the telescope.
For example, in implementations the diffracted light from each mirror segment generates a diffraction pattern comprising a spot in the metrology image plane and these spots can be stacked by moving the mirror segments in tip/tilt. This is similar to the previously described approach using reflective facets, but the diffraction grating can cover a larger area than a facet and generate a smaller spot in the captured image, facilitating more accurate alignment.
As another example a shape of the diffraction pattern, i.e. of an image of the diffracted light captured by the image sensor of the metrology system, may be used to determine a measure of decentre of one or more of the mirror segments from an optical axis of the telescope. Here decentre refers to translational misalignment of a mirror segment perpendicular to an optical axis of the telescope. Such decentring elongates the diffraction pattern (diffracted image/spot).
In another example a diffracted light metrology system may comprise a wavefront metrology system such as a Shack-Hartman wavefront sensor, to characterize a wavefront of the diffracted light. A Shack-Hartman wavefront sensor can measure the local tip-tilt of the wavefront from the displacement of a focused spot of metrology light, and hence can also be used to correct tip/tilt of the mirror segments.
No doubt many other effective alternatives will occur to the skilled person. It will be understood that the invention is not limited to the described embodiments and encompasses modifications apparent to those skilled in the art lying within the scope of the claims appended hereto.
Number | Date | Country | Kind |
---|---|---|---|
2202543.1 | Feb 2022 | GB | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/EP2023/054633 | 2/23/2023 | WO |