2.1.

Impacts of Atmospheric Conditions

The first stage in the life of an instrument once it has been manufactured involved testing and storage on the ground. The structure of the telescope will need to be kept in conditions that mimic those encountered in space for alignment critical operations. Therefore, stability to variable ground conditions is beneficial. In addition to the standards of cleanliness inherent to any space instrument, there is the need to control for two more effects: temperature and humidity.

Thermal distortion is one of the operational challenges, which will be discussed later in this paper with more detail, though of course there is the need to maintain the stability of the test setup during alignment tests. The effect of thermal creep, particularly in materials with polymeric matrices, such as the common place carbon fiber-reinforced plastics (CFRP), also needs to be taken into consideration, as heavy structures under permanent load can creep out of specifications if the storage temperature is high enough. Depending on the material, a significant fraction of this creep may be recovered,⁹ but it still poses a danger to the structure’s repeatability.

Hygral expansion is also an important issue since any humidity absorbed by the materials will tend to outgass in space, eliminating the swelling of the structure on the ground. This is not a major concern for metallic or ceramic materials, or for their composites, but it can have an effect on materials with a polymeric base.^10,11 Most assembly operations are carried out in humid air, so the components need to be coated with a moisture barrier to prevent excessive absorption. A system that is sensitive to moisture will need to receive a “bakeout” treatment, which removes this expansion prior to operations on the ground. Another mitigation technique is to store the critical components in a dry atmosphere and only getting them out for short periods for alignment operations. Note that performing a bakeout prior to alignment may be good practice, but if the structure is allowed to swell before launch, the absorbed moisture may deposit in other surfaces of the spacecraft, compromising other subsystems.

2.2.

Gravity Release and Testing Procedures

Regarding their structural integrity, space structures in general can be considered to experience nearly no loads, with the notable exception of thermal loads due to temperature gradients. In deployable structures, the deployed configuration is not necessarily designed to support its own weight on Earth.^12,13 This forces the testing phase to use gravity offloading procedures to simulate deployment procedures and deployed-state performance. This is typically achieved by hanging the structure from several points that can move without friction parallel to the direction of the motion but compensate gravity or supporting it with rollers on the floor.

This is true of most deployable structures, such as antennas and solar panel assemblies. Optical structures are different since they have stringent three-dimensional precision deployment requirements. Therefore, it is important that the gravity offload system does not overconstrain the motion of the structure. This is difficult to achieve, because gravity, being a distributed force, creates stresses within a structure supported from discrete points, which can be mitigated by adding more support points, thereby negating the principle of exact constraint. Ideally, such a system would have zero stiffness in the vertical direction, which is an emerging property of certain structural configurations. A general description of zero-stiffness mechanisms is given by Schenk and Guest.¹⁴ Another possibility is to use a pressure-controlled flow to provide stable force output.¹⁵

There is an additional issue when testing the optical systems because large aperture mirrors also need to withstand their own weight during ground testing. Mirror materials, usually technical ceramics such as silicon carbide or metals, are very stiff, but the allowable surface figure error requirements during testing may pose a significant engineering challenge due to the need for special large mounts and cranes to move them around or test their performance.¹⁶ This is an old problem and not exclusive to deployable telescopes. One of the mitigation strategies is using simulations with different mounting boundary conditions, and extracting the zero-gravity sag from the deflections resulting from pointing the mirror upward and downward.¹⁷

2.3.

Launch and Deployment Failures

Launch is the most structurally challenging event faced by the spacecraft, with the possible exceptions of in-orbit collision or re-entry. Loads experienced during transportation and handling may also be an issue with large space structures in general.¹⁸ Development of space structures capable of withstanding these loads is a wide topic and covering it is well beyond the scope of this paper. An introduction to the subject was written by Wijker.¹⁹ Launch vehicle manufacturers typically present four main profiles to describe the mechanical environment within the launcher: static acceleration profile, separation shock response spectrum defined in the payload adapter, sine-equivalent vibration, and acoustic vibration.^20–22 These are standard loads used in the verification procedures for all spacecraft. Loads experienced during transportation and handling may also be an issue with large space structures in general,¹⁸ and they must be included in the analyses.

Like other deployable elements, deployable optics are built so that the stowed state is much stiffer and stronger than the deployed configuration. Going by eigenfrequency as a criterion, launch loads typically require first eigenfrequencies in the order of 100 Hz for the structure to survive, but deployed structures typically have first bending eigenfrequencies below 1 Hz. This imposes the need for hold down and release mechanisms and alternative load paths, which spare the optical elements from excessive loads. Structures typically fail due to yielding or rupture of its components, but it is advisable to keep in mind that microplasticity^8,23 effects can appear well below the nominal yield stress of materials.

2.4.

Microdynamics

Microdynamics is a term referring to a number of loosely related phenomena, all of which take place below the microscale threshold. In this paper, the term refers to the effects of friction in joints from a purely mechanical source. Other effects that are referred to as microdynamics by other authors, such as thermal snapping, creep, and microyielding are covered in other sections of this paper.²⁴ This separation is adopted to clarify the causes for each phenomenon, but even in this case, several microscale effects are comprised in this definition. These different effects are difficult to uncouple and observe independently in experiments, and even the terminology in the literature is neither clear nor consistent about it. An effort in this regard was presented by White and Levine,²⁵ who proposed a framework for the analysis of microdynamic effects. The scope of their definition of microdynamics is the same as in this work.

Microdynamics is also related to structural nonlinearity. Joints are known to present three essential kinds of nonlinearities in their behavior, namely freeplay, nonlinear elasticity, and hysteresis. These effects are illustrated in Fig. 2. These deviations from the ideally linear response are known to lower the eigenfrequencies of the structure with respect to a linear approximation, making them significantly less stable against vibration inputs.²⁷

Fig. 2

Three types of nonlinear behavior in mechanical joints.²⁶

The microdynamics of deployable trusses has been the focus of an intense research effort in the context of developing highly stable trusses, such as the one present in NuSTAR.²⁸ This type of joint-dominated structures experience sudden vibrations during operation, which are consistent with the sudden release of energy previously lost as a result of hysteresis.²⁹ This is very similar to the phenomenon of thermal snapping but can happen without the influence of a thermal load.³⁰ Peterson and Hinkle²⁴ also provide a rationale to lay out hysteresis requirements on large structures. For a given level of acceptable displacement, a stiffer structure is able to accommodate more hysteretic loss with acceptable stability.

The joints where these hysteretic losses occur usually rely on contact surfaces for the transmission of the deployment torque. An example of such mechanism is the ball-bearing hinge developed as part of the Origins program and presented by Lake et al.³¹ This was a joint designed to minimize nonlinear responses.

Ingham and Crawley²⁷ investigated the modal behavior of another deployable truss structure. Very small strains, below $1\mu ϵ$, were shown not to affect the modal shapes of the structure, therefore respecting its linear behavior. Strains above this level, however, did induce a shift in the eigenfrequency and an increase in the observed damping ratio. The researchers conclude that the nonlinear structural damping mechanisms in the joints do not activate for strains below this boundary. An effort to incorporate similar effects was reported by Coppolino et al.,³² including nonlinearity, stiffness uncertainty, and snapping in joints. The authors created a toolbox to define properties at component level and simulate whole structures in the 100- to 500-Hz frequency band.

Another effect within the microdynamics classification is the so-called microlurching, described by Warren et al.,³³ who found that joint-dominated structures subjected to transient disturbances consistently “lurch” to a new static position once the vibration dissipates. Repeating this event a sufficient number of times makes the structure reach an “equilibrium zone,” which was found to be extremely repeatable. This behavior allows positioning a large deployable structure with very high repeatability by using intentional, transient disturbances. The underlying cause of this behavior is understood to be the progressive release of residual strain energy stored at the frictional interfaces prior to the intentional excitation.

2.5.

Microvibration

Microvibration, also referred to as jitter in the literature when it affects spacecraft attitude, is the presence of small oscillations propagating through the spacecraft structure. In most cases, these vibrations pose no threat to the survival of the said structure, but they can alter pointing of the system and cause misalignment of optical components, in which case it is referred to as wavefront error (WFE) jitter.³⁴ In fact, this type of vibration is widely considered one of the largest threats to the pointing stability of optical payloads.^35–37

Reaction wheel assemblies are generally the largest source of vibration in most systems.^36,38 The dynamics of this phenomenon are usually modeled with an unbalanced rotor model with the first harmonic appearing at the reaction wheel speed. Subsequent harmonics, which may be of similar importance, appear as a result of other imperfections in the wheel assembly. Examples of these defects are inhomogeneity of the wheel’s mass, worn or irregularly shaped bearings, stick-slip behavior, or freeplay. One may argue that friction effects are also similar or even the same as those described as microdynamics in this text, but for the purposes of this paper, the reaction wheel is a “black box” with an output vibration signature. Magnetic-bearings reaction wheels have been proposed to mitigate these problems.³⁹ Other sources of vibrations in spacecraft are the turbulent flow of coolant or fuel, propulsive burns, composite microcracking, and any other mechanism with moving parts. Another source of jitter is micrometeoroid or orbital debris impact, which causes a transfer of momentum from the impactor to the spacecraft.⁴⁰

In general, any release of energy through vibration would fall into this category, but again sources that are specifically due to microdynamics or thermal disturbances will be treated separately in this paper in Secs. 2.4 and 2.8.

In general, the jitter environment of an instrument will largely determine the stiffness and damping requirements of the structure. Note that both stiffness⁴¹ and damping⁴² of a material are temperature-dependent and therefore the dynamics of the structure is subject to a change depending on its operational temperature. Measurement of these properties may also require special instrumentation capable of operating at extreme temperatures.⁴³ In practice, a stable structure is typically as stiff as required and as lightweight as possible. These conflicting requirements define the trade-off of structural design, as more stiffness requires either more load-carrying material or a more efficient use of it. However, the specific stiffness of materials or structural depth is not easily scalable with the size of the observatory.^16,44 Structural depth also requires either a large volume or complex deployment mechanisms, which add hysteresis and uncertainty to the deployment.

2.6.

Thermal Cycling and Creep

Thermal effects can produce misalignment of components. This can happen in a reversible way, as is the case of thermal expansion, or irreversibly, in case of creep. Both effects are critical to the operation of space instruments. In addition to misalignment, the refractive index of lenses, beam splitters, and other refractive elements can change as a function of temperature. Refractive elements, however, are not usually part of the OTE of deployable space instruments, which is the focus of this paper.

The main drivers of the thermal environment of a spacecraft are the heat dissipation of its components, its injection and operational orbits, and its ability to reject or absorb radiation. A spacecraft in orbit cannot evacuate the heat it produces or receives by any other means than radiation heat exchange. Ideally, it would be possible to size and align the structure of a space optical system such that the instrument would reach radiative equilibrium with its environment at its nominal alignment. This is not possible because of the dynamic nature of heat inputs to the system and the uncertainty in its modeling.

Thermal expansion is described by the material’s coefficient of thermal expansion (CTE). Thermal expansion over wide ranges of temperature, however, does not behave in a linear way. CTEs reported in the literature are usually specified for a certain temperature, such as ambient temperatures. This is not necessarily the operating temperature of the structure. Some telescopes have a certain temperature range required to operate, such as thermal IR telescopes,⁴⁵ while others are indifferent to it, but their temperature is defined primarily by their environment. Therefore, it is of primary importance to utilize the correct temperature to linearize the thermal expansion behavior of the material. In general, thermal expansion is an undesirable effect, which makes very low CTEs desirable when choosing materials. In some cases, high CTE materials can be used for passive compensation techniques.²

Another effect of thermal expansion is thermal warping or bending, which is the result of the combination of thermal expansion and the uneven distribution of temperatures in a bulk component. Even if a certain component achieves perfect radiative or conductive equilibrium with its environment, there can be a temperature gradient within the material, which makes a region expand or contract more than others. This is typical of situations where a component receives heat from one side and emits it on a colder side. This induces a global bending of the component. This bending may be completely acceptable if it has been modeled and included in the design previously. However, changes to these gradients will negate this compensation. Both the magnitude and variability of these gradients are diminished by high conductivity materials for given boundary conditions. Homogeneous temperature changes also allow an easier definition of the structure’s thermal center, which may assist in the modeling stages. Materials with high thermal conductivity, which favor a homogeneous temperature distribution, are therefore desirable.

Taking these two effects into account, a coefficient of thermal warping can be defined as $\alpha /\kappa$, with $\alpha$ being the materials’ CTE and $\kappa$ its thermal conductivity. This gives a measure of how different materials would tend to warp^46,47 for a given heat transfer situation. Another figure of merit is described by Bely² and defined as $\kappa /\alpha C_p\rho$. This parameter includes a correction for thermal diffusivity, which is the conductivity divided over material’s density $\rho$ and its specific heat $C_p$. This is done in order to describe how quickly the steady state is reached after a change in the boundary conditions of the thermal property. Note the behavior of this parameter is inverse to the aforementioned coefficient of thermal warping.

A high thermal diffusivity and therefore fast response to thermal variation may not be desirable if the system is designed to be heavily damped through the use of a large thermal mass or latent heat storage. In that case, these compound figures of merit for a material may have their meaning inverted or may be ignored. However, the use of high thermal mass systems typically imply larger inertial mass and volume, partially negating the benefits of deployable systems in space applications. A compromise solution may be found in the use of phase-change materials, which allow additional heat to be used in a reversible phase transition without the need for a large mass.⁴⁸

2.7.

Thermal Flutter

Thermal flutter can be understood as a dynamic effect caused by cycles of thermal warping, exciting vibration modes of the system. The Hubble Space Telescope famously experienced a disturbance in its pointing whenever undergoing eclipse due to quick heating and deformation of the deployable solar panels, which excited structural modes.^49,50 Deployable optics, as discussed in Sec. 2.3, tend to have lower eigenfrequencies, possibly by 2 orders of magnitude, and tend to be more exposed to thermal fluxes than traditional monolithic telescopes, which are encased in rigid bodies. A way to mitigate this effect is keeping heat fluxes steady by means of a thermal shield or particular orbit selection, or diminishing the aforementioned effects of thermal expansion. For missions that orbit Lagrange points, the solar heat flux remains constant and therefore flutter is not a concern. Thermal flutter can affect the pointing stability of instruments or cause instability of the optical system itself, introducing wavefront jitter.

The overall phenomenon can be explained with a boom exposed to solar fluxes. Limited thermal conductivity will establish a steep temperature difference between the exposed and shaded sides, which causes the former to expand more than the latter. If the process can be regarded as quasistatic, no dynamic effect will occur. However, booms might be poorly insulated and have small thermal inertia, in addition to low eigenfrequencies. Boley⁵¹ proposed that the coupling between thermal fluxes and vibration modes may be assessed through the parameter:

2.8.

Thermal Creaking

The other major coupled effect is referred to in the literature as thermal creaking or snapping. This refers to differential heating of contact interfaces causing a vibration as stresses built up at the interface are violently released.

Kim⁵⁵ studied the interaction between this phenomenon and the dynamics of spacecraft. This phenomenon happens primarily in the joints of deployable structures due to the presence of contact interfaces. In essence, the energy release mechanisms are the same as described in Sec. 2.4, but the driver of the stress accumulation is the constrained thermal expansion. In terms of mitigation, the same recommendations as exposed in Sec. 2.4 apply.

Since differential thermal expansion is the driver of this phenomenon, limiting the temperature difference across a contact interface with the same materials on both elements can prevent slippage. This, however, is not easily achieved in mechanisms with nonconforming contact, as it is usually the contact itself that acts as a thermal interface. These contact interfaces in precision applications usually rely on point or line contacts,⁵⁶ which minimizes the effective thermal contact area. In contact interfaces that have different materials, the CTE mismatch between them will drive slippage proportional to a bulk temperature change of the whole joint, even if the temperatures across the interface are the same.

3.1.

Large Lagrange Point 2 Observatories

The L2 observatories are missions that employ telescopes that do not fit in existing launchers and orbit the second Sun–Earth Lagrange point. Their mission is to look into the universe for a number of scientific enquiries. They have the largest aperture sizes, which need to be deployed in order to fit the launcher, and the largest focal length, which is also deployed. Two such missions are present in the literature: The JWST and the Large Ultraviolet and Infrared Surveyor (LUVOIR). Their thermal environment in L2 is stable save for fluctuations in solar output, which eliminate concerns of thermal flutter and mitigate temperature variations. The large structures of these missions and their use of many mirror segments, however, exacerbate wavefront stability challenges. Their stringent WFE budget, resulting directly from the scientific requirements, imposes the need for more complicated alignment and vibration isolation mechanisms. Owing to their complexity and institutional support, these are the most well-documented and extensively researched projects reported so far in the literature.

In these missions, microdynamics is a major concern due to the large number of interfaces, joints, and latches involved, and much of the knowledge presented in Sec. 2.4 is derived directly from investigations performed to develop JWST.

3.1.1.

James Webb Space Telescope

The JWST is the largest and most complex deployable optics instrument built so far. Here, a summary of the key thermomechanical elements of its OTE is presented. JWST will operate in orbit around the second Earth–Sun Lagrange point (L2). This provides a stable thermal environment compared to the eclipse cycles, which occur in orbit around Earth. Figure 3 shows an exploded view of the JWST’s OTE plus the integrated science module and the thermal management system, and a schematic of the sunshield.

Fig. 3

(Top) Exploded view of the main elements of JWST, including the integrated science module and the thermal management unit.⁵⁷ (Bottom) Schematic representation of JWST’s sunshield.⁵⁸

Optical design

JWST is a three-mirror anastigmat (TMA) with a segmented and actuated primary mirror (M1), 6.5 m across, an actuated secondary (M2), and a fast steering mirror incorporated in the exit pupil for line-of-sight disturbance correction. The dimensions of the primary mirror and the sunshield exceed those of the largest launchers in the market. The telescope observes in visible and IR wavelengths, being diffraction limited at 2 μm. The need for operating wavelengths in the thermal IR range also imposes operational temperatures in the order of 40 to 60 K, to prevent emissions from the telescope from affecting the signal.

Describing JWST or any of its subsystems in full is well beyond the scope of this paper. A wealth of information can be found in the literature. Howard et al.^59–64 describe in more detail the optical design and the linear optical model developed to run structural–thermal–optical performance analysis on it.^34,65 A summary of Howard’s papers was published in 2011.⁶⁶ Here the most relevant aspects of its design will be outlined: mirrors, deployment mechanisms, fine actuation mechanism, and the sunshield.

Mirror technology

The JWST mirrors are made of beryllium O-30-H.⁶⁷ While beryllium has a relatively poor thermal stability at ambient temperatures, its total thermal expansion is very small when cooled to JWST’s cryogenic operational temperatures. It is also very lightweight, strong, and has good thermal conductivity, compared to other optical materials. Material was removed from the mirror cores to produce a rib structure that is both lightweight and stiff. When compared with the competing material, ultralow expansion (ULE™) glass, beryllium was selected for its superior technical properties. This decision was made despite the fact that beryllium posed higher risks of production delays and cost overruns.⁶⁷

Deployment mechanism

The segmented mirrors are mounted in a central, static frame, and two “wings,” which have to deploy over a 103-deg angle and latch in place, powered by a stepper motor. Both the fixed and mobile frames are made of carbon fiber composites, specifically tuned for thermal stability through manipulation of the material CTE.⁶⁸ This is done in order to minimize temperature dependence of the final positions of the mirrors. The development team observed the principles previously laid out to avoid microdynamics response. Nonconforming interfaces were used so as to avoid any load transfer through friction and subsequent microslippage. However, stiffness requirements for some interfaces made it necessary to add redundant nonconforming contacts or oversized friction joints.⁶⁹

Actuation mechanism

Deployment repeatability requirements for this system, from the launch stowed position to the nominally deployed operational position, are in the order of a few millimeters.⁶⁹ This is a very large error compared to the allowable WFE in practice. The mirror positioning is corrected by the actuation mechanism down to a step size smaller than 10 nm. This is achieved through the use of a two-stage actuation mechanism, with a coarse range of 20 and a fine range of $2\mu \mathrm{m}$. The actuators are mounted on a beryllium delta frame and interface with the mirror in a hexapod configuration, which provides 6 DOFs position control. An additional actuator, coupled to beryllium struts, also provides radius of curvature correction.⁷⁰ The overall control architecture is described by Scott Knight et al.⁷¹ The actuators are powered by a so-called gear motor, composed of a stepper motor, a resolver, and a gear head. The bearings in the motor are a limited lifetime item, sized for the expected mission lifetime.⁵⁷

Sunshield

Passive cooling of the JWST’s OTE will be achieved mainly through the use of a large, deployable sunshield, as shown in Fig. 3, which stands between the Sun and the OTIS, which is the OTE plus the Integrated Science module. The sunshield is composed of five layers of Kapton E with a vapor-deposited aluminum (VDA) coating on their inner faces. In the outer faces, the two outer layers are coated with a silicon optical solar reflector coating, and the three inner faces with the same VDA. The layers are arranged with a dihedral angle, which allows radial rejection of heat both from the spacecraft and from sunlight.⁵⁸ The sunshield is intended to receive roughly 300 kW of thermal power on its hot side but let less than 0.05 W pass on to the cold side.⁷² The sunshield is deployed via four hinged booms and two additional telescopic booms. These six booms are attached to tensioning mechanisms, which are in charge of separating and stretching the different layers of the shield to its designated geometry.

3.1.2.

Large Ultraviolet Optical Infrared Surveyor

LUVOIR is not, strictly speaking, a definite project but rather a response to the call for proposals initiated by NASA as part of its decadal survey on astronomy missions. Similar concepts for very large telescopes operating in the ultraviolet, visible and near infrared have been proposed prior to the current LUVOIR design. Two notable concepts are the Advanced Technology Large-Aperture Space Telescope (ATLAST) and the High Definition Space Telescope (HDST), but these offer less engineering analysis.⁷³ At the same time, LUVOIR is one of the competing proposals for prioritization of the next large observatories, alongside the Origins Space Telescope (OST), Habitable Exoplanet Imaging Mission (HabEx), and the Lynx X-ray Observatory. Of these concepts, only LUVOIR has seen extensive development in the direction of a segmented aperture deployable architecture like that of the JWST. However, a concept for a smaller version of HabEx called HabEx Lite was proposed⁷⁴ featuring a segmented but nondeployable primary mirror. OST also is also meant to use a non-deployable, segmented primary mirror.⁷⁵

LUVOIR has been proposed in different formats, including one featuring a monolithic 8 m primary mirror. However, most of the proposals follow the trend toward a segmented aperture reminiscent of that of JWST but substantially larger. This would allow to employ the lessons learned during development of JWST but also tightens the alignment requirements by a factor of approximately 4 based on a diffraction limit at 500 nm.¹⁶ Proposals exist for primary mirrors of sizes 9.2, 11.7, and 16.8 m.⁷⁶ Any of these proposals exceeds the total diameter of any current launcher fairing. The latest report available⁷⁷ further develops a 15-m aperture concept as LUVOIR-A, which can be seen in Fig. 4. Another possible architecture described in that report, LUVOIR-B, is based on an 8-m primary aperture, off-axis telescope.

Fig. 4

LUVOIR-A conceptual architecture.⁷⁸

As its name implies, this telescope would operate primarily in the optical range, with capabilities in ultraviolet and shortwave IR, therefore, being a true successor to the Hubble Space Telescope. This means that it would be a “warm” telescope, without the need to create a cryogenic environment. Even so, operation of this telescope will require launching it into an orbit about the L2 Lagrange point, and a large sunshade to prevent sunlight from affecting the measurements. The requirements for this sunshade are likely to be simpler than those of JWST’s sunshield.⁷⁹

The architecture for the current LUVOIR-A concept consists of a 120-segment primary mirror aperture, with all segments controlled in 6 DOFs. The use of ULE™ segments in a closed back structure makes these segments stiffer than their JWST predecessors, and so the shape actuator can be dropped.⁷⁸ Like in JWST, a fine steering mirror would provide line of sight correction. LUVOIR-B is an off-axis TMA telescope with 55 mirror segments, also controlled in 6 DOFs. The deployment mechanisms of either concept are broadly based on the wings.

In addition to controlling the rigid body motions of the primary mirror segments and the secondary mirror, LUVOIR is also meant to control the temperature of each individual segment via the use of heater and diffuser plates mounted behind each segment.⁸⁰

3.2.

Earth-Orbiting Observatories Deploying along Optical Axis

These missions are small satellites intended to deploy an element along its main axis with the intention of achieving longer focal lengths than would otherwise be possible in a constrained space. However, not deploying the primary aperture, the achievable aperture and therefore achievable image quality is limited. This can be mitigated by having these instruments fly on very low altitudes, which makes their service life limited.

This is the simplest possibility for deployment, and it has attracted interest from companies and universities because it is potentially the cheapest and fastest technology for development purposes. From a mechanical perspective, these configurations suffer from low eigenfrequencies in the bending of elements along the optical axis. This makes it difficult to keep the alignment of the main optical elements and precisely control their attitude. However, not having a segmented aperture, these systems would not need to cope with wavefront stability issues due to alignment of mirror segments. This architecture also simplifies the design process of a baffle for straylight and thermal control, which can be simply attached to the deployable element. The reduced amount of deployment mechanisms also reduces the induced microdynamic instabilities.

3.2.1.

Dobson Space Telescope

Another project, which saw development during the 2000s, was the Dobson Space Telescope (DoST), by a team of researcher at TU Berlin. This project was reported on by Segert et al.^81,82 They claimed VHR capabilities from a Microsat platform flying at 550 km, and dual EO and near-Earth object observation purposes. The primary mirror aperture was of 0.5 m for its baseline mission. The operational wavelength is not detailed in the papers, but it is understood to cover the visible range. The project was reported on over a series of papers, including several experimental setups and a tentative launch date in 2012. The members of the team moved on to create the startup Berlin Space Technologies GmbH, which aimed to take the project to the market, but the company does not list deployable telescopes as part of their product line. The project was eventually discontinued because it was not deemed commercially viable.

In the case of the DoST architecture, only the secondary mirror is stowed for launch and deployed in orbit. This allows for a longer focal length with a significant reduction in size, at least in one dimension. The deployment is achieved with a rigid deployable truss, which pushes the secondary mirror 1.1 m away from its stowed position, though this mechanism is not explained in detail. The fact that there is no segmented aperture makes misalignment much less critical. The DoST implements an active optics strategy with a 6 DOFs-actuated secondary mirror. A baffle incorporated in the secondary mirror structure also provides cover for the primary mirror, and the active optics mechanism is expected to bring the errors due to deployment repeatability and thermal expansion from 1 mm to $1\mu \mathrm{m}$. The DoST concept is shown in Fig. 5.

Fig. 5

Artist impression of the deployed DoST.⁸³

3.2.2.

Collapsible Space Telescope

Yet another project for deployable optics can be found in the Collapsible Space Telescope (CST) project, a proposal for a deployable secondary mirror mounted on a coiled mast.⁸⁴ The operational wavelength of this telescope is not detailed, other than covering the visible range. A baseline primary aperture of 152.2 mm was proposed to fly at an altitude of 250 km. The overall architecture of this system is similar to that of the DoST and is shown in Fig. 6. Once released, the strain energy stored in the coils pushes the secondary mirror away from the primary. This technique does not affect the aperture of the entrance pupil and therefore does not increase the achievable diffraction-limited resolution but makes integration of a baffle easy and dramatically increases the achievable focal length. This project is scarcely reported and does not have any continuation beyond the original paper, which is a valid conceptual design case but lacks sufficient analysis to back up its feasibility.

Fig. 6

CST concept, showing the deployment sequence.⁸⁴

3.2.3.

Surrey Satellite Technologies Limited deployable telescope

Surrey Satellite Technologies Limited (SSTL) proposed a telescopic optical barrel, shown in Fig. 7 coupled to the secondary mirror of a Cassegrain telescope. The system performs at a GSD of 1 m and is built upon heritage of their Carbonite-2 satellite but with a deployable secondary mirror. This is estimated to lower the volume requirements of each system inside the launcher fairing, thereby allowing for more deployments with less launches. Gooding⁶ proposes a case study with a 500 km baseline orbit, though no details of operational wavelength, focal length, or primary aperture are provided.

Fig. 7

Concept of the SSTL telescopic deployment barrel.⁶

The secondary mirror is spherical, which restricts the alignment procedure to a 3 DOFs kinematic problem. The same optical barrel used to deploy the secondary mirror provides protection for the primary mirror and is a straylight management tool. The barrel deployment is powered by a motor that drives a lead screw per barrel section into a V groove. The primary mirror is monolithic, and so the maximum aperture is still limited by the available volume within the fairing. The focal length, however, can be drastically increased similarly to that of the CST and the DoST.

3.2.4.

Picosatellite for Remote Sensing and Innovative Space Missions

The Japanese Picosatellite for Remote Sensing and Innovative Space Missions (PRISM)⁸⁵ is, to the best of the author’s knowledge, the only deployable optics spacecraft whose launch has been confirmed. It was successfully inserted in a 660-km circular orbit in 2008, as a technology demonstrator for a nanosatellite class imager. Unlike all other examples cited herein, PRISM is a refractive system with an aperture of 90 mm.⁸⁵ The overall system is shown in Fig. 8.

Fig. 8

Schematic representation of PRISM, showing the coil booms that push the focusing lens away. The baffle is omitted for clarity.⁸⁵

The optical system is pushed away from the main instrument housing by means of a collapsible boom, similar to the mechanism described in the case of the CST.⁸⁶ This allows the system to achieve a much longer focal length than would otherwise be possible. The coils are also used to pull a baffle, which provides straylight control and a degree of thermal protection to the structure. The detector array is mounted in a focusing mechanism, which can adjust its position to correct for focus errors caused by the flexible deployable structure.⁸⁵ The low stiffness structure made special correction of its line of sight jitter necessary,⁸⁷ but successful imaging with 30-m GSD was achieved.

3.3.

Earth-Orbiting Observatories with Segmented Apertures

These systems are proposed to deploy the primary mirror and may or may not deploy the secondary mirror. In this way, they can increase both the focal length and the aperture of the optical system beyond what is achievable with a conventional telescope of the same size. They are also smaller than the L2 observatories, although eclipses are present in their thermal environment. The alignment of primary mirror segments is therefore critical, and disturbances are frequent. The solution is generally the use of active optics mechanisms, which constantly correct the alignment.

The challenges of telescopes that only deploy along their optical axis also apply here, but they are aggravated by the difficulty to integrate a baffle to mitigate the large amounts of straylight from albedo originated outside of the field of view and the thermal influence of the solar flux.

3.3.1.

Large Aperture Telescope Technology

The Large Aperture Telescope Technology (LATT) project is reported by Marchi et al. (2008) in several papers. Its purpose was described by Hallibert and Marchi⁸⁸ as pushing the critical technology readiness level (TRL) of large active mirrors, taking advantage of existing experience with active secondary mirrors for ground-based observatories. The project was finalized in 2015, considering the technology for large-aperture active mirrors to be TRL 5 under the European Space Agency standard.

LATT proposed a design of an afocal telescope with 4 m of aperture diameter to take differential absorption Lidar measurements around the 935-nm wavelength. Unlike the rest of the telescopes mentioned herein, this telescope’s purpose is not imaging, and it doesn’t operate within the visible spectrum.⁸⁹ The design featured a segmented primary mirror with active segments made with a CFRP core and a thin sheet of Zerodur. A novel feature of the design is the use of voice coil actuators to make the large aperture mirror correct its shape. In addition, electrostatic locking is proposed as a means of holding the thin sheet to the substrate, providing a strong load-path to resist launch loads.

The deployment of these mirrors relies on elastic memory composite hinges. The authors proposed that the error inherent to this technology could be compensated by means of the actuators. In addition, the design featured an inflatable baffle, which covered the entire system in order to prevent straylight from falling in the detector.⁸⁹ The complete concept can be seen in Fig. 9. An alternative design was proposed by Thompson et al.⁹¹

Fig. 9

System architecture of the LATT baseline telescope, showing the cylindrical baffle.⁹⁰

Its purpose it to push the critical TRLs for a near-infrared (NIR), very large aperture telescope. Their progress reports a very lightweight active mirror, and an optical design featuring an $7\text{-}{\mathrm{m}}^2$ deployable collection area. Though some of the technical requirements of this project have been published, there is no publicly available information about its phasing budget. Details on its operational orbit are not explicitly reported, though calculations for straylight reported by Mazzinghi et al.⁹² point to a 450-km altitude.

3.3.2.

Deployable Petal Telescope

Utah State University’s Space Dynamics Lab built and tested the so-called Deployable Petal Telescope (DPT),⁹³ a Cassegrain telescope, which can be mounted on a 3-U CubeSat platform, deploying both its primary and secondary mirrors. This endeavor, however, is scarcely reported on the literature, and so actual feasibility and performance of the system is difficult to ascertain. A video is still available online, in which the primary mirror segments are seen to unfold in a flower-like fashion, whereas the secondary mirror, mounted on a rail, deploys away from the instrument housing.

The DPT is a Cassegrain-type telescope with a 200-mm aperture, deployable segmented primary mirror, which unfolds in a similar fashion to flower petals. The prototype mirror is a smaller version and has flat at the tip to aid in alignment, which makes its effective aperture smaller. The secondary mirror is mounted on a rail, which linearly extends away from the primary. Both mechanisms are described to be fully passive via a spring load. The conceptual imaging system is diffraction limited at 632.8 nm, achieving a 1.3-GSD from a 500-km orbital altitude. In the architecture reported by Champagne et al.,⁹³ there are no external baffles, but a collapsible baffle is installed in the space between the two mirrors. This would mean the telescope is largely exposed to heat fluxes from the Sun. Unlike most of the other deployable telescopes, the DPT does not implement an active optics mechanism to correct for misalignment or nonrepeatability. The authors trust the assembly to be stable enough in 5 DOFs, excluding the tip motion, which is adjusted through a mechanism attached to the back of the mirror. This mechanism was tested for deployment repeatability and the results showed that most of the surface error came from the individual segments, which the authors expected could be improved. The test setup for the DPT is shown in Fig. 10.

Fig. 10

Picture of the DPT test setup, not including the internal baffles.⁹⁴

3.3.3.

Deployable Space Telescope (UK Astronomy Technology Centre)

A more recent attempt to use CubeSats for VHR EO is reported by Schwartz et al.⁹⁵ The purpose of this project is to reach a 39-cm GSD from a CubeSat platform flying at 350 km altitude. This system is diffraction limited at 550 nm and has a 300-mm aperture. The authors provide quantified measures of the sensitivity to misalignment of the system and propose an active optics actuation system to get the mirrors aligned within the tolerances. The necessary metrology proposed is a sharpness optimization algorithm, which drives the active optics. The system can fit in a 1.5U CubeSat standard unit. The authors do not describe yet the sensitivity of their concept to thermoelastic deformation as a result of the orbital transients.

The primary mirror active optics acts in tip-tilt and piston directions. Three motors are coupled to the mirror by means of a flexure system connected to a steel shaft. This shaft acts as the connection to the mirror substrate and also integrates a torsion spring that powers the deployment. The active optics strategy is able to obtain a surface error of 25 nm under laboratory conditions.⁹⁶ The authors do not report on a particular deployment strategy for the secondary mirror. In the latest design, as shown in Fig. 11, a baffle is included in between the primary and secondary mirrors for straylight attenuation, which seems to leave the telescope exposed to thermal fluxes like in the DPT case.

Fig. 11

Schematic representation of the adjustment mechanism of the deployable telescope proposed by Schwartz et al.⁹⁵ showing an additional internal baffle.

3.3.4.

Deployable Space Telescope (TU Delft)

The TUD DST project was proposed by Kuiper in 2012 and has been running ever since. Dolkens proposed the optical design and a ray-tracing tool to assess its performance. From that, the top–down misalignment budgets were derived and presented in Ref. 97. In addition to the optical analysis, several Master of Science theses have been published by the TU Delft,^98–102 detailing the evolution of the mechanical design from the first iteration by Dolkens to the latest developments reported by Dolkens et al.¹⁰³

The DST, as shown in Fig. 12, is a TMA telescope with a four-segment primary mirror of 1.5 m aperture, diffraction limited at 550 nm, and flying at 500 km altitude. These segments are actuated in piston and tip-tilt directions, which are regularly identified as the most critical DOFs in both this and other proposals. Aberrations caused by warping of the mirror shape are corrected by a deformable mirror installed in the exit pupil of the OTE. This deformable mirror, however, cannot correct for the WFE caused by a dephasing of the primary mirror. This misalignment is controlled by means of a so-called PistonCam, installed in the intermediate image plane, tracking the sharpness of the image at the mirror edges. This information is fed to a control algorithm, which drives the active optics mechanism.

Fig. 12

Artistic rendition of the TUD DST.

The mechanical design includes a baffle capable of limiting the variability in the thermal environment and a low hysteresis compliant rolling element hinges, so as to comply with the guidelines presented in Sec. 2.^99,102 This active optics mechanism was proposed by Pepper.¹⁰¹ In addition to providing exact constraint and actuation of the mirror substrate to its support plate, the active optics actuator acts as a primary load path holding the mirror through launch with acceptable strength margins. The mechanism consists of four actuators in push–pull configuration, which move an intermediate plate. This intermediate plate is in turn constrained by means of a hexapod mount to the mirror substrate.

Successful operation of the system relies on three layers of increasingly strict tolerances.⁹⁷ The deployment mechanism should be accurate enough to reach a coarse alignment in the micron range. The system is then actuated to a nominal position in the order of $\lambda /20$, with short WFE jitter disturbances kept below the order of $\lambda /100$, with $\lambda$ being the diffraction-limited wavelength of the telescope. A deployable structure capable of meeting these requirements taking into consideration all the effects mentioned in this paper and that of Edeson et al.⁸ is currently under development.

3.3.5.

Deployable Optics Model Experiment

The Deployable Optics Model Experiment (DOME), reported by Peterson and Hinkle, was a structural mechanics experiment related to a concept differential absorption Lidar deployable telescope based on the requirements for the Ozone Research through Advanced Cooperative Lidar Experiments mission.^104,105 The baseline design was a segmented aperture telescope with a 2.55-m diameter consisting of an hexagonal monolithic core and six additional petals. An artist impression of the concept is shown in Fig. 13. Since this was a technology development project, no specific details of orbit or operational wavelength are available to the knowledge of the authors.

Fig. 13

Artist impression of the Lidar telescope, which was used as baseline by the DOME project.¹⁰⁶

The purpose of the project was to characterize the behavior of a single deployable petal with a simple deployment mechanism. The mechanism consisted on a strutted hinge, which latched upon reaching the deployed state. Repeatability of the latch was identified as the largest source of deployment nonrepeatability in the mechanism. This meant that the largest effort went to develop a latch with very stringent repeatability requirements.⁴⁴ The project was expected to end with the single petal test carried out on this structure coupled to an ultrastable metrology frame. Some results were reported in Ref. 106, but no definite conclusion could be found for the project.

The baseline mirror material for this system is a mixture of CFRP and ULE™.^11,105 The former provides a thermally stable and stiff base, but it does not provide a surface of enough quality. A thin layer of ULE™ is used for that purpose and bonded to the composite mirror core. The mirror core is protected from the effects of moisture through the use of a moisture barrier.¹¹

3.3.6.

Deployable In-Space Coherent Imaging Telescope

The Deployable In-Space Coherent Imaging Telescope (DISCIT) is a project supported by MIT Lincoln Laboratories and the U.S. Air Force. In contrast to the efforts in deployable optics, which were discussed previously, reports related to this project are much sparser in the information about their optical design, focusing more on the development of high-precision deployment mechanisms. However, a ray-trace schematic of the system can be seen in Fig. 14. The researchers intend to reduce the complexity of deployment mechanisms, which can meet optical precision requirements, compared to the complicated deployment mechanism of the JWST.¹⁰⁷

Fig. 14

Optical ray tracing schematic of DISCIT’s baseline optical architecture.¹⁰⁷

The baseline optical design is not thoroughly described in the literature available on this project. The objective is a 0.7-m effective sparse aperture Cassegrain telescope.¹⁰⁷ The expected performance of such a system or its operational environment are not clarified either. Some renders related to this project can be found online showing a multimirror arrangement with nondeployable secondary mirror and a deployable, segmented primary. The results of the experiments on the tape spring hinge reported in Ref. 108 show that tape spring hinges can achieve micron-level repeatability and also provide interesting results about their dimensional stability under changes in temperature and humidity.

DISCIT itself does not incorporate any type of baffle as of the current date. A previous project within MIT Lincoln Laboratories, with the involvement of silver, does investigate the deployment of optical barrel assemblies using similar tape spring hinge technology.^109,110 Though this is a separate development to DISCIT, it does mean the authors are aware of the need to baffle sunlight falling onto the telescope.

3.3.7.

Autonomous Assembly Reconfigurable Space Telescope

Another notable research project is the Autonomous Assembly of a Reconfigurable Space Telescope (AAReST). It has been a long-term student project with students from Caltech supported by the Jet Propulsion Laboratory, and it is expected to be launched in 2020. More than a functional telescope, AAReST is a technology demonstrator for large aperture deformable mirrors and, more importantly, in-orbit reconfiguration of optical segments. The proposed mission has a 0.4-m aperture and operates at the wavelength range of 465 to 615 nm. The main feature of this mission is the detachment of 3U CubeSats called “MirrorSats,” which will fly away from the main spacecraft, “CoreSat,” achieve reattachment to it at a different location and operate at a 650-km altitude.¹¹¹ This sequence is shown in Fig. 15.

Fig. 15

View of the reconfiguration mechanism of AAReST, also showing the integrated tape spring hinge.¹¹²

The CoreSat mounts two fixed mirrors, whereas the MirrorSats hold a thin-shell CFRP deformable mirror as described by Steeves et al.¹¹³ The large deformable mirrors are mounted on a 3 DOFs platform (piston, tip-tilt) for coarser rigid body motions, providing very high authority control. Reattachment of the MirrorSats is achieved through an electromagnetic docking mechanism, which makes the MirrorSats fall into a Kelvin clamp, which provides a repeatable reattachment.¹¹² Another valuable element to this mission is the foldable boom, which holds the imaging camera. Mallikarachchi and Pellegrino¹¹⁴ described the manufacture of the hinges present on these booms. Cutouts on monolithic CFRP booms are made and then bent until the fold. Once the holding force is removed, the strain energy contained in the fold is released, making the boom go back to its original shape with acceptable repeatability.

3.3.8.

UltraLITE

Ultralightweight Telescope (UltraLITE),¹¹⁵ also called deployable optical telescope (DOT),¹¹⁶ is a deployable TMA telescope developed mainly by the Air Force Research Laboratories. This proposed design spawned a series of structure experiments to validate several elements of its architecture. The design featured a deployable tower holding the secondary mirror and three deployable circular mirrors. These mirrors were notable for their very lightweight design, owed to the use of a CFRP core and a thin ULE™ shell, which also acts as an active mirror.⁴⁶ This development program focused extensively on active vibration controllers¹¹⁷ and the design of a very stiff deployment structure through the use of hybrid CFRP, including high and intermediate modulus fibers.¹¹⁵ Several effective apertures of the telescope were reported throughout the technology development program, with a testbed of 1.7 m being built.¹¹⁷ More system design options are discussed by Powers et al.¹¹⁸ for a baseline size of 5 to 6 m apertures flying on high altitudes, in the order of 15,000 km. The precise details of the final design could not be found as to its operating altitude or wavelength, though the latter is understood to be in the visible range.

Catanzaro et al.⁴⁶ mentions active heating as thermal control to keep the desirable stability of the telescope. There is no indication that a sunshield or baffle was proposed in the latest embodiment, although early concept schematics showed a deployable one.¹¹⁸ This is not elaborated in the texts found during this survey. An artist impression of the concept is shown in Fig. 16.

Fig. 16

Artist impression of the UltraLITE telescope.⁴⁶

4.1.

Deployment Mechanisms and the Need for Active Optics

There are 2 orders of magnitude differences between achievable deployment repeatability and the allowable WFE for high-quality imaging in visual and NIR ranges. Deployment repeatability depends on the specific technology, which locks the structure in place, and on the size of the structure, with typical values in the order of a few micrometers.^33,119 For comparison, WFE requirements for diffraction-limited optics are in the order of 10s of nanometers.¹²⁰ In addition, in-orbit disturbances can affect the stability of the system in several timescales. Therefore, there is a need for active correction of at least the most sensitive DOF, which can be determined via optical sensitivity analysis. In the most demanding applications, full 6 DOFs per element control and shape control are required. The total stroke of such a system must be matched to the magnitude of the foreseeable disturbances, and its resolution must be smaller than the allowable WFE.

This is indeed the overall consensus in the deployable optics literature, with projects that either focus heavily on the development of novel active optics concepts or acknowledge it as an essential enabler to achieve its goals. This holds true, regardless of the overall size of the proposed telescope and its environment. Table 1 shows how the majority of systems include some type of active correction, although some do not specify whether or not they do. Lake et al.¹²¹ highlighted the importance of a trade-off between structural stiffness requirements and authority of the control system. Most active optics mechanisms allow for more DOFs than the six per element pure rigid body kinematics by exerting some control over the surface shape. This can be achieved at the primary optical element or through the use of a deformable mirror as is the case of the TU Delft DST. JWST implements one extra DOF per mirror segment for radius of curvature control of the primary mirror. AAReST and LATT propose a fully deformable primary mirror with an undefined number of DOFs, which allows much more control authority.

Survivability of these large active mirrors can be achieved through electrostatic locking of the face sheet during launch.⁹⁰ With this technology, extremely low areal densities, below $20\mathrm{kg}/{\mathrm{m}}^2$ can be achieved, however, at the cost of complexity of the active optics mechanisms. However, embodiments such as those in the LATT or AAReST projects are not yet capable of achieving diffraction-limited performance in visible wavelengths.

4.2.

Thermal and Vibration Control

Thermomechanical stability is achieved through the systematic removal of external influences on the telescope at multiple stages of the mission design. Selecting the environment of the telescope is a choice with many potential variables and trades. Deployable optics in the literature is proposed for either LEO or L2 orbits, parallel to the split between EO purposes and astronomy. Missions in LEO would be primarily for EO, with a focus on lightweight systems that can be produced cheaply, fast, and potentially in larger numbers to provide higher temporal and spatial resolution. L2 is too far away to be a good candidate for EO, but a prime location for astronomy due to a stable thermal environment and constant Sun illumination. Astronomy requires much larger apertures with more stringent optical quality requirements, which results in more complex and sensitive systems in a comparatively milder environment. This divide is clear in Table 1, with the largest apertures and number of controlled DOFs predictably belonging to the large L2 observatories.

In L2, a single sunshield is able to greatly reduce the straylight and thermal effects caused by sunlight, although such sunshield has been a critical part of JWST’s development and represents a system of considerable complexity and scale on its own. LEO is a very dynamic environment by comparison due to the need to cope with cycles of dayside and eclipse. A notable exception to this would be the particular case of dawn–dusk Sun-synchronous orbits whose illumination is constant. It is also more populated by orbital debris from previous missions. A large aperture in LEO comes with the need to deploy a baffle to protect it from sunlight if alignment is to be maintained.

In the literature regarding deployable telescopes, thermal stability is addressed mainly through the material choice rather than temperature control of the support structures. LUVOIR is a notable exception to this rule, where active heating is used. Active heating is the simplest way to control the temperature of the optical elements, but it has important drawbacks. Namely it adds to the power requirements of the telescope, is incompatible with cryogenic telescopes, and it increases modeling complexity due to the potential of heat radiating onto other optical assemblies. Eisenhower et al.⁸⁰ estimated that for the ATLAST concept, 526.8 W of power would be needed to maintain the correct temperature.

Though this may be attributed to the immaturity of most of the proposals in the literature, there has not been extensive description of a thermal management system and its impacts, with the exception of JWST and LUVOIR.⁷⁷ The Delft DST team also considers this a major concern for development.^122,123 This has caused Table 1 to be very incomplete on that aspect.

As for vibration control, there is a need to eliminate or insulate the sources. One area in which this may be achieved is the design of mechanisms that do not transmit any loads through friction²⁶ and the substitution of classical bearing hinges and joints with compliant mechanisms, which eliminate undesired effects cited in Sec. 2.4, such as creaking, backlash, freeplay, and wear.¹²⁴ Owing to the need of pointing the telescope toward the target, reaction wheels will likely be always a necessity, and therefore the microvibration requirements outlined in Sec. 2 must be observed. Deployable optics would benefit from magnetic-bearing or otherwise low-noise reaction wheels, although these are more complex and expensive.

4.3.

Lightweight Mirrors

Deployable telescopes, compared to monolithic telescopes, have more flexible structures and therefore have comparatively lower eigenfrequencies, which can be partially offset by the use of extremely lightweight, stiff mirrors. Mirror areal density is influenced by total aperture of the segment, available core depth, material choice, and core geometry. Lightweight mirrors can be achieved by removing material from a mirror blank, which achieves bending stiffness by leaving a complex pattern of stiffeners behind the face sheet. This process is limited by manufacturability and local buckling failure modes, if the stiffeners are made excessively thin. Baiocchi and Stahl¹⁶ cite other concerns that may dominate over areal density when deciding on a mirror design, such as stiffness, complexity, and cost.

As for the thermal environment, telescopes operating at cryogenic temperatures benefit from the small cumulative shrinkage of beryllium at cryogenic temperatures, its low density, and high strength.⁴⁷ In the case of warm telescopes, beryllium has a relatively high CTE and is typically outperformed by other high performance materials, such as Zerodur, ULE™, and Silicon Carbide (SiC). Zerodur and ULE™ have a long usage history in both ground and space telescopes due to their extremely low CTE at ambient temperature, but they present poor thermal conductivity and are brittle compared to metals or SiC. Another candidate family of materials is aluminum alloys. Aluminum has excellent programmatic and mechanical properties, but it also boasts a much higher CTE than its competitors.¹²⁵ However, its good thermal conductivity and the possibility of making its supporting structure out of the same material makes it possible for a full aluminum telescope to have a degree of inherent athermalization, provided the temperature is homogeneous across the entire system. There is little evidence in the literature suggesting that this is enough to offset its high CTE in large aperture telescopes.

Table 2 summarizes the thermal and mechanical properties of these materials. The $\rho$ is the material density, $E$ is the Young’s modulus, and ${\sigma }_{\mathrm{Max}}$ is the maximum allowable stress.

Table 2

Properties of typical mirror materials.2,47

CFRP is used as a base material that can be mixed with an ultralow CTE material, which can be polished in order to profit from the high specific stiffness and thermal stability while preserving the optical surface quality.

The alternative to these stiff mirrors are large active mirrors with very small structural depths, which are controllable through electromechanical actuation to correct for its large initial surface errors, as shown by the LATT and AAReST projects. These are prime examples of the trade-off between mass and control complexity shifting to the latter. So far this approach of very high authority mirror control has not been demonstrated in flight, but experiments have been performed.