iii. design of the atst

III. Design of the ATST Advanced Technology Solar Telescope Construction Phase Proposal

III. Design of the ATST Page 57 of 174

III. DESIGN OF THE ATST

1. SYSTEMS OVERVIEW

The Advanced Technology Solar Telescope (ATST) is an all-reflecting, four-meter, off-axis Gregorian

telescope housed in a rotating enclosure. It can deliver up to a 300-arcsec field of view to either a

Nasmyth or a coudé observing station. Energy outside of this field is rejected from the system by a heat

stop located at prime focus, allowing manageable thermal loading on the optical elements that follow. The

telescope also includes an integrated adaptive optics system designed to provide diffraction-limited

images to the focal-plane instruments at the coudé observing station. Our design meets all science

requirements stated in the Science Requirements Document (SRD, ATST Document #SPEC-0001).

A persistent systems approach is essential to the success of a telescope like ATST. Systems engineering

works with project management, the scientific staff, and the other engineers to accomplish various

activities. In this chapter the emphasis will be on design requirements flow-down, error budgets, and

performance predictions. It will conclude with a top-level description of the telescope design that serves

as an outline and general background material for the subsequent detailed design descriptions. Other

aspects of systems engineering are discussed in Part IV of this proposal, Management of the ATST

Construction, Integration, and Testing (see Chapter 6).

1.1 THE FLOW-DOWN PROCESS

Systems engineering has been responsible for flowing the science requirements – as specified by the

scientific community – down to design requirements on the telescope. For example, the science that

ATST will perform requires a sharp image. Systems engineering must first list all of the telescope and

instrument subsystems that have the potential to cause the image to blur. These will include the quality of

the optical components (mirror figures and polish quality), telescope mount vibrations, and thermal

distortion of the air above the telescope enclosure, to name just a few. While the science requirement is

expressed in terms of the size of a point-source image, this constraint must be converted to a mirror-polish

specification, mount stiffness, and maximum allowed temperature variation on the enclosure skin to be

useful when designing the telescope and specifying the manufacturing tolerances of its components.

The process of flowing science requirements down to design requirements began with the ATST SRD.

That document established the top-level science requirements based on the solar community’s vision and

proposed mission of the telescope. These requirements lead directly to a set of critical science use cases

listed in the SRD, selected because they place the most stringent technical requirements on the telescope

and instrumentation. These use cases lead, in turn, to specific performance requirements placed on the

telescope and instrumentation. All of this has been spelled out in the SRD.

It has been the task of the engineering team to produce a design that meets the top-level telescope and

instrument requirements, and hence the science requirements. The process followed to get to an initial

design was highly iterative, involving a baseline concept that was proposed, tested against the top-level

requirements and error budgets, and modified until cost and performance requirements were met. The

following paragraphs highlight a few aspects of this process.

1.1.1 Science Use Cases

The Science Requirements Document describes 18 science use cases. They establish minimum

performance requirements on a variety of important parameters and operational modes that must be

supported by ATST to meet the science requirements. The detailed justification of each requirement is

included in the SRD. Table 1.1 lists the use cases that demand the highest performance in each area:



1.1.2 Telescope Requirements

The use cases described above and other scientific considerations lead to several direct top-level

requirements on the telescope and site that are justified in detail in the Science Requirement Document:

Resolution:

ATST must have a minimum four-meter aperture.

Table 1.1. Science Use Cases.

1. Spatial Resolution Half of the science use cases require observations at or near to the diffraction limit at visible wavelengths. These include Interaction of strong and weak magnetic fields Flux emergence and disappearance Dynamics of kilogauss flux tubes Internal structure of flux tubes / irradiance variations Magnetoconvection in sunspots Generation of acoustic oscillations Temperature and velocity of the photosphere and chromosphere Prominence formation and eruption Solar Flares

2. Field of View Most use cases require a field of view of several arcmin. The coronal use cases established the most stringent requirement, desiring 3 to 5 arcmin: Prominence formation and eruption Coronal magnetic fields Coronal plasmoid search Coronal velocity and density in active region loops Coronal intensity fluctuation spectrum

3. Wavelength Coverage Three use cases require observations in the thermal IR (12 m): Turbulent/Weak fields Dynamo processes in deep layers of the convection zone Solar Flares Four require observations at or near to the atmospheric UV cutoff at 300 nm: Dynamics of kilogauss flux tubes Internal structure of flux tubes / Irradiance variations Turbulent/Weak fields Hanle effect diagnostics Many use cases require or strongly desire simultaneous observations over broad wavelength ranges.

4. Spectral Resolution Many use cases require spectral resolution of 1 pm (picometer) or less. The most stringent use case requires 0.42 pm at 500 nm: Dynamics of kilogauss flux tubes

5. Polarimetric Sensitivity and Accuracy

The most stringent use cases require polarimetric sensitivity of 10-5

: Turbulent / Weak fields Hanle Effect Diagnostics

6. Scattered Light The coronal observations are the most demanding in terms of sky and instrumental scattered light near the limb of the sun, requiring excellent coronal sky conditions, and low instrumental scatter: Coronal magnetic fields Coronal plasmoid search Coronal velocity and density in active region loops Coronal intensity fluctuation spectrum On-disk observations of large sunspots also place requirements on in-field scattering: Magnetoconvection in sunspots

7. Observing modes The most demanding use cases involve active-region evolution, which require simultaneous observations with multiple instruments in both the visible and the thermal infrared: Dynamo processes in deep layers of the convection zone Solar flares



ATST shall include high order adaptive optics capable of deriving information from solar

granulation and other solar structure.

Photon flux and sensitivity:

ATST shall provide a minimum collecting area of 12 m2

Polarization sensitivity and accuracy:

The polarization sensitivity must be 10-5

with polarization accuracy of 510-4

Scattered light:

The scattered light from the telescope and instrumentation from angles greater than 10 arcsec

shall be 1% or less

The total instrumental scatter due to dust and mirror microroughness must be less than 2510-6

at

1.1 solar radii (1.6 arcmin from the limb of the sun)

Field of view:

The ATST shall provide a minimum field of view of 2 arcmin square at coudé, and 5 arcmin at

Nasmyth.

Wavelength Coverage:

The ATST shall cover the wavelength range from 0.30 to 28 m

Flexibility:

The ATST must accommodate simultaneous multi-wavelength observations at visible and IR

wavelengths.

It must be possible to carry out simultaneous observations with different instruments.

Image rotation introduced by the telescope must be counteracted by de-rotation, preferably

without additional reflections.

Maximum scientific productivity requires easy and fast (less than 30 minute) switching between

facility instruments.

Lifetime:

The ATST is expected to be the major solar ground based facility for a minimum of two decades.

The useful lifetime of ATST is expected to exceed 40 years.

Adaptability:

The ATST shall be designed with a minimum of limitations for future use and in a way that

allows future upgrades and the addition of new instruments.

Availability:

Scheduled engineering and maintenance should not exceed 10-15%.

Of the remaining time, ATST should set a goal of achieving telescope reliability that allows

observing during 97-98% of the available clear time (similar to the best nighttime telescopes). Location:

ATST should be located at the best affordable site in terms of seeing, sky clarity and sunshine

hours. This will maximize the telescope performance and minimize the cost of adaptive optics.

The science use cases also place several important derived requirements on the telescope:

Pointing and Tracking:

Absolute (blind) pointing shall be accurate to better than 5 arcsec.

Offset pointing shall be accurate to better than 0.5 arcsec.

Open-loop tracking stability must be better than 0.5 arcsec for one hour.

Active Optics:

Active control of the primary mirror figure will be required to achieve the necessary resolution.

The active-optics system must be run in an open-loop mode during coronal observations when

real-time wavefront information is unavailable.



1.1.3 Design Requirements

Many design requirements were derived directly from the science requirements; others were determined

from error allocations within several systems error budgets (see Section 1.1.4). The job of flowing

science requirements down to design requirements is often made easier because existing telescopes have

met similar requirements, and the engineering specifications of those systems are known and available.

Table 1.2 lists important subsystems that are constrained by a specific telescope requirement. Note that

the demand for high-resolution observations yields the greatest number of design requirements. The error

budget discussion that follows helps to show how some of these constraints were derived in detail. The

table also shows the current “compliance” status, noting briefly what features of our design allow us to

meet the more challenging requirements.

Table 1.2. Flow-down to Subsystems, and Compliance Status

Telescope Requirement

Subsystem Flow Down Design Compliance and Strategy

Resolution Constrains telescope mount drives, control, and thermal systems; coudé rotator and drive systems; pier design; M1 aperture size, figure, support system and thermal control; active optics performance; heat stop thermal control; M2 figure, mount, and thermal control; feed optics figures and thermal control; adaptive optics performance; guiding systems; tip-tilt performance; optical alignment; polarimetry optics; Nasmyth and coudé optics; instrument lab thermal control; science instrumentation; telescope control software; enclosure thermal system (both skin temperature requirements and ventilation requirements); support facility location and construction methods.

In Compliance – Challenging

4-m aperture

Diffraction-limited optical design

Active thermal control of components

Heat rejection at prime focus

Rigorous error budgeting applied to many subsystems

Photon flux and sensitivity

Constrains M1 aperture area, M1 ancillary equipment (mirror washing), and all mirror coating specifications, telescope and instrument optical designs (number of reflections).

In Compliance – Straightforward

4-m aperture

Polarization sensitivity and accuracy

Constrains mirror coatings, polarimetry analysis and calibration performance, and science instrumentation.


Pre-Gregorian Modulation

Pre-Gregorian Calibration

Optical designs

Charge-caching cameras

Scattered light Optical design, mirror polishing specifications, M1 mirror cover design, M1 ancillary equipment (mirror cleaning and washing), occulting system, mirror coatings, baffles and stops, enclosure thermal system (specifically the interior ventilation system and its relation to dust control), and coating and cleaning facilities.


Off-axis configuration

Nasmyth station

In-situ cleaning/washing

Active ventilation filtration. Prime-focus occulting



There are some technical risks associated with the more challenging requirements. These and the

project’s plans for risk mitigation are outlined in the management discussion (see Part IV of this proposal,

Chapter 8, Risk Assessment and Management).

Table 1.2. Flow-down to Subsystems, and Compliance Status (continued)

Field of view Constrains heat stop dimensions, thermal systems, occulting system, feed mirror dimensions, mirror cell dimensions, acquisition system, and instrument designs.


Optical design

Mechanical design

Wavelength coverage

Constrains mirror coatings, imaging system optical materials, science instrumentation, and restricts the general use of transmissive windows for thermal control of the light path.


All-reflecting design

Laminar-air coudé station isolation

Flexibility Constrains coudé rotator and drives, polarimetry analysis and calibration, coudé stations, instrument control system, science instruments, telescope control system, data handling system, and the observatory control system.

In Compliance

Software design

Modular components

Facility instrument concepts

Coudé station layout

Lifetime The ATST lifetime requirement affects all designs, requiring a high level of robustness or replacement and reconfiguration strategies.

In Compliance

Coudé station design

Modular component design

Adaptability The adaptability requirement drives many of the features of the feed optics, the Nasmyth and coudé observing stations, the science instruments, the instrument control system, telescope control system, and observatory control system.

In Compliance

Software designs

Modular component designs


Coudé station design

Availability The Availability requirement affects many systems that could impede observing efficiency. In particular, it drives the of mirror cleaning and washing facilities, acquisition and guiding, polarimetry analysis and calibration, Nasmyth and coudé platform configurability, optical enclosure ventilation system, mirror cleaning and coating facility.

In Compliance

In-situ mirror cleaning and washing

Control software


Coudé layout

Location The location of ATST constrains the details of the pier design, in-situ M1 mirror cleaning and washing (depending on local dust levels), mirror thermal control systems, enclosure thermal control (based on wind and temperature extremes), coudé room thermal control, site infrastructure, buildings, facility equipment, and coating and cleaning facilities (and their broader availability).

In Compliance

Design meets requirements at our selected site and alternate site

Site specific details have been rolled into the design since selecting Haleakala

Pointing and tracking

Constrains telescope mount, drive system, telescope control software.


Mount Design

Pointing Kernel selection

Active Optics Constrains M1 Mirror, M1 support structure, M1 controller, wavefront sensor, telescope control system, and AO (adaptive optics) system.

In Compliance – Challenging at high zenith distances

Thin meniscus mirror

120 axial supports

24 lateral supports

Dedicated wavefront sensor


III. Design of the ATST 2 of 174

1.1.4 Error Budgets

Error budgeting is fundamentally a systems-level issue. A given error budget will typically be distributed

across many disparate subsystems. These have been designed by different engineers and will be

fabricated by different vendors (Systems Error Budget Plan, ATST Document #SPEC-0009). Error

budgeting is a useful tool at all levels of design since it represents a means to negotiate design trades in

the broadest possible context. This process is central to the mission of systems engineering.

The highest priority error budgets developed to support the flow-down process involved delivered image

quality. Three different image quality error budgets were derived from science use cases:

Diffraction limited observations at 500 and 630 nm

The ATST shall provide diffraction-limited observations (at the detector plane) with high Strehl

(S>0.6 required, S>0.7 goal) at 630 nm and above during excellent seeing conditions (r0 (630 nm) >

20 cm) and S > 0.3 at 500 nm and above during good seeing (r0 (500 nm) = 7 cm).

Seeing limited on-disk observations at 1.6 m

[For] excellent seeing conditions, (r0 at 1.6 micron 100 cm)… Minimum requirement: 50%

Encircled Energy Diameter < 0. 15 arcsec.

This requirement is derived from science use cases performing on-disk observations, but it presumes

excellent seeing conditions when good images can be obtained over a wider field of view than can

typically be obtained with single-conjugate AO (adaptive optics). Hence, the aO (active optics

controlling M1 figure) and tip-tilt loops are closed, but the high-order AO system is not in use.

Seeing limited coronal observations at 1.0 m

Off-pointing up to 1.5 solar radii, wavelength 1 micron, excellent seeing conditions: r0 (1 micron)

50 cm, FWHM seeing limited PSF 0.4 arcsec. The minimum resolution required for coronal

magnetometry is 2 arcsec. The Telescope shall deliver the following image quality:

50% Encircled Energy Diameter < 0 .7 arcsec

85% Encircled Energy Diameter < 2 arcsec

This requirement applies to most coronal observations. “Off-pointing” implies that there is no

granulation present within the 5-arcmin field of view, so no wavefront information is available. The

AO loop (including tip/tilt) is open. Similarly, the aO loop is open, and the primary mirror figure

and telescope alignment must be corrected based on constant or repeatable errors via a look-up table

or function fit.

These three cases are of particular interest because they span the range of possibilities for wavefront

correction, as shown in Table 1.3. Complete error budgets with Monte Carlo simulations show that

ATST meets the science requirements for each of these cases. An example is shown in Section 1.1.5.

Table 1.3. Wavefront correction cases.

Active Optics Loop Tip-Tilt Loop High-order AO loop

Diffraction limited observations

Closed Closed Closed

Seeing limited on-disk observations

Closed Closed Open

Seeing limited coronal observations

Open (Look-up table)

Open Open



The fundamental difference between these three observing modes is the range of spatial frequencies that

can be corrected in each case. It is useful to look at the problem in terms of power spectral density (PSD)

of spatial frequency errors in the wavefront delivered to the focal plane. For example, if optical surface

polishing errors are analyzed in this way, it is found that they obey a power law distribution over a very

broad range (five orders of magnitude) of spatial frequencies extending from dimensions near to the full

aperture all the way down into the realm of surface microroughness. Throughout this range the slope of

the PSD on a log-log plot is roughly –2. This is shown schematically in Figure 1.1.

When the active optics loop is closed allowing figure errors to be compensated, the power at the lowest

spatial frequencies is reduced considerably. Switching on the adaptive optics system causes further

improvement at higher frequencies. This analysis suggests that it is useful to allocate errors separately

within each of four frequency regimes. Table 1.4 defines the four frequency regimes.

With these definitions in place, it is possible to use the same error tree for all three observing cases, and

modify the error allocation according to the active controls available.

Table 1.4. Frequency regimes.

Definition Spatial Period (mm) Description

Low 4000 800 Active Optics Influence Range

(10 actuators across the 4-m primary)

Intermediate 1 800 200 Adaptive Optics Influence Range (100 mm sub aperture on primary)

Intermediate 2 200 4 Uncorrectable figure errors

High 4 down Microroughness

Figure 1.1. Periods and spatial frequencies are relative to the four-meter entrance pupil.



Telescope Delivered Image Quality w site stats Haleakala

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.0 0.8 1.6 2.4 3.2

50% Encircled Energy (arc sec)

Pro

ba

bilit

y

Site Seeing

Bottom Up

Telescope & Instrument

Figure 1.2. Telescope delivered image quality with site statistics.

A specific example of how these conventions are used to flow down to design requirements is shown in

Table 1.5.

Table 1.5. Flow down example.

M2 Errors apportioned by frequency band

Low Int. 1 Int. 2 High Total Correction applied

140 30 8 2 144 None (Manufacturing Spec.)

70 30 8 2 77 aO loop open (with look-up table)

2 30 8 2 31 aO loop closed

0 2 8 2 8 AO loop closed

Taken in combination, the three delivered image quality error budgets place limiting constraints on many

telescope subsystems, and hence form the basis of their design requirements.

1.1.5 Performance Predictions

The error budgets maintained for ATST are used in two different modes. The first mode represents

snapshots in time, assuming specific observing conditions. The image-quality science requirements, for

example, specify seeing conditions that are good or excellent. We make assumptions about other free

parameters, like ambient temperature, wind speed, and zenith distance, and these values are entered as

constants. Normally we must adopt “worst case” values when the science requirement does not include

these details. Details are contained in the System Error Budget Plan, ATST Document #SPEC-0009.

The second mode used in the error budgets brings additional information into the error calculation. This

includes distributions of expected parameter values that affect the image quality. For example, wind-

speed statistics are available for the Haleakalā site. With these distributions in place, Monte Carlo

simulations can be performed to randomly select wind speeds weighted by the probability functions

(histograms). By looking at thousands of system manifestations, it is possible to predict the fraction of

the time that the telescope system will deliver images of a given quality.

Wind speed is a particularly interesting case to study because of the diverse ways in which wind affects

the final image. As wind speed increases the flushing it provides to the enclosure and mirrors will

improve seeing by sweeping away the warm, turbulent boundary layer. The vents in the enclosure will

allow some fraction of this flow to flush the telescope mirrors and mount structure, again improving self-

induced seeing. In all of these cases more wind is better. Wind can also degrade performance, however.

The pressure of the wind on the thin M1 mirror will cause it to deform, resulting in poorer images. The

wind also excites vibrations in the telescope mount assembly. For these cases the higher the wind-speed,

the worse the performance.

The Monte Carlo simulations built into

the error budget spreadsheets allow all of

these effects to be analyzed with

underlying wind-speed probability

distributions, and any other error

parameters for which probability

distributions exist or can be estimated.

Figure 1.2 shows the results of such an

analysis for the seeing-limited error

budget using both seeing and wind-

velocity probability distributions for

Haleakalā. The green curve shows that

ATST will meet the 0.15 arcsec

requirement most of the time. The blue



curve includes seeing statistics for the site, and the red includes telescope effects. These give a more

accurate prediction of how often we will meet the requirements. Similar positive results are produced for

the adaptive optics and coronal cases (ATST Document #SPEC-0009).

1.2 DESIGN OVERVIEW

The design for ATST that has emerged from the requirements flow down is described in detail in the

chapters that follow. We describe each subsystem in some detail starting with the requirements place

upon it, followed by a general description of the design. The top-level organization of these subsystems

is as follows:

Telescope Assembly Enclosure

Wavefront Correction Systems Support Facilities and Buildings

Instrument Systems Remote Operations Building

High Level Controls and Software

While most aspects of these subsystems can be discussed in isolation, several important features of the

ATST design span several subsystems: the optical design and its overall performance, thermal control,

and special features of the design that facilitate coronal observations. Each of these will benefit from a

brief systems-level description. The individual components are discussed in more detail later.

1.2.1 Optical System Design

The ATST optical design has two features that distinguish it from nighttime telescopes with similar

aperture size: it is off axis and Gregorian. Both of these features are included in the design in direct

response to the science requirements, and have many practical advantages as well.

The basic idea of the off-axis configuration is

shown in Figure 1.3. M1, the primary mirror,

is a four-meter section of a 12-meter parent

parabola. The parent is shown as a wire-frame

structure across the bottom of the figure, but

only the solid shaded part on the left is used in

ATST. Similarly M2, the secondary mirror, is

a 0.62-meter section of a two-meter parent

ellipsoid. Only the solid shaded section on the

top right is used. The red beam filling the

primary represents a point source at zenith.

The underlying Gregorian design has the

feature that an image is formed in front of M2,

offering an opportunity to reject most of the

energy in the concentrated beam before

introducing it onto M2 and the optics that

follow. There is an 80-mm diameter image of

the sun formed at prime focus. Only a 13.5-

mm diameter circular section of this image (five arcmin unvignetted) is passed through the heat stop and

on to the Gregorian focus. The rest of the energy is reflected away or absorbed by a liquid-cooled heat

stop. As a result, the irradiance incident upon M2 is roughly the same as that on M1 (i.e., the same as any

object lying in direct sunlight). For an aluminum reflective coating that is 88% efficient only about 40

watts of power is absorbed by M2, which is manageable.

M1

M2

Gregorian

Focus M1

M2

M1

M2

Gregorian

Focus

Figure 1.3. Basic off-axis configuration.

Prime Focus



The stray-light performance of such a telescope is improved by the lack of diffraction around M2 or a

spider structure supporting it. This is critical to coronal observations close to the limb of the sun (see

Section 1.2.3 below). Also, the beam is unobstructed by M2, so the full four-meter aperture is available.

Both the heat stop and the secondary mirror are outside of the beam incident on M1, reducing the effect of

any residual heat plume on seeing performance. The off-axis design also simplifies the delivery of cooling

and other utilities to the heat stop, limb occulter, and M2 since these services can be provided without

crossing the beam. Studies carried out by the ATST project and commercial vendors during the design

and development phase have shown that the off-axis optical elements do not represent any significant

technical challenge or unmanageable risk.

The ATST needs a total of sixteen mirrors to deliver the beam to the Nasmyth and coudé instrument

stations. They are organized into five different groups based on their function and relationship to the

telescope’s mechanical components.

The OSS Gregorian Optics include M1 and M2 which form an image at the f /13 Gregorian focus. They

are attached to the Optical Support Structure (OSS), which is the element of the telescope mount that

moves to track the sun as its altitude changes (see section 2.1.2). The Gregorian focal plane is not

intended for focal-plane science instrumentation, but is instead used to inject calibration sources and

targets. These mirrors will be coated with aluminum to optimize UV performance.

The OSS Nasmyth Transfer Optics include M3N, M4N and M5N which relay the Gregorian focus to the

Nasmyth observing station (Figure 1.4). They are attached to the optical support structure. M3N is a flat.

M4N and M5N are off-axis paraboloids. The Nasmyth focal plane is the first science observing station.

It is a zero-magnification transfer of the f /13 Gregorian image to a location on the altitude axis of the

optical support structure. This station will be used primarily for infrared coronal spectro-polarimetry and

other observations requiring minimum telescope polarization and maximum throughput. While the

Nasmyth image quality is not diffraction limited over the full field of view, it meets the science

requirements for the observations to be performed there. Observing at the Nasmyth station precludes use

of the coudé observing station since some of the coudé mirrors need to be moved so as not to interfere

Figure 1.4. The Nasmyth Transfer Optics.

Altitude Axis

M3N

Gregorian Focus

M5N

M4N



with the Nasmyth beam. These mirrors are also coated with aluminum to optimize UV performance.

The remaining eleven mirrors are used to transfer the Gregorian focus across the altitude and azimuth

axes and down to the coudé observing station (Figure 1.5). At noted previously, it is not possible to

observe simultaneously at the Nasmyth and coudé observing stations. All of the remaining mirrors are

coated with protected silver to optimize the visible and near-IR throughput at coudé.

The OSS Coudé Transfer Optics include M3 and M4. They serve to place the beam over the altitude axis

and, like the Nasmyth transfer optics, they are attached to the optical support structure. M3 is a folding

flat, and M4 is an off-axis ellipsoid that transfers the Gregorian focus to an unused focal plane above the

coudé room. M4 also re-images the telescope’s entrance pupil onto M5, the fast steering mirror. M3,

because of its location close to a focus, is also part of the quasi-static alignment system (see section 3).

The Mount Base Assembly Coudé Transfer Optics include M5 and M6. Their primary function is to

transfer the beam over the azimuth axis. M5 also serves as the fast steering mirror for the wavefront

correction subsystem (see section 3). While folding the beam vertically down the azimuth axis could be

Figure 1.5. The ATST coudé optical train (left) with exploded views of the OSS and mount base

transfer optics (top right) and the coudé feed optics (lower right).



accomplished with one mirror at a 45 angle, this significantly increases the size and hence the mass of

the mirror which would degrade its ability to make fast tip-tilt corrections to the wavefront. M6 is close

enough to the pupil image that it is used as a compensator in the quasi-static alignment system. Both

mirrors are attached to the mount base assembly, which caries the optical support structure, and rotates to

track the sun in azimuth.

The Coudé Rotator Optics include M7 through M13. They have several important functions, including

beam folding, wavefront conditioning (to remove aberrations inherent in the off-axis Gregorian

telescope), and beam scaling to form a correctly sized pupil image on M9 (the wavefront correction

subsystem’s deformable mirror described in more detail in section 3). These optics rotate with the coudé

observing station.

The output of this set of optics is a collimated optical beam rather than an image. This is desirable due to

the varying focal-ratio and beam-size requirements of the coudé focal-plane instrumentation. Each

instrument will form its own, optimized image. The collimated beam is also ideally suited for beam

splitting and bandpass selection when multiple instruments are in use simultaneously since beam splitters

and filters inserted into a collimated beam do not introduce significant optical aberrations as they would

in a converging beam. The beam diameter emerging from M13 has a diameter of about 320 mm.

The characteristics and functions of the sixteen mirrors that make up the ATST optical system are

summarized below in Table 1.6.

Table 1.6. Summary of the functions of characteristics of the ATST mirrors

Mirror Function Diameter Optical Characteristics

OSS Gregorian Optics

M1 Primary mirror 4240 mm Off-axis concave parabola

M2 Transfer f /2 prime focus to f /13 Gregorian,

compensate for misalignment, and provide fast

steering during coronal observations.

635 mm Off-axis concave ellipsoid

OSS Nasmyth Transfer Optics

M3N Folding mirror 220 mm Flat

M4N Nasmyth transfer mirror 1, f /13 to f /13 390 mm Off-axis concave parabola

M5N Nasmyth transfer mirror 2, f /13 to f /13 370 mm Off-axis concave parabola

OSS Coudé Transfer Optics

M3 Folding mirror, quasi-static alignment

compensator

160 x 114 mm Flat

M4 Transfer f /13 Gregorian to f /50 intermediate

focus

450 mm Off-axis concave ellipsoid

Mount Base Assembly Coudé Transfer Optics

M5 Fast steering mirror for coudé observations 220 mm Flat

M6 Quasi-static alignment compensator 275 mm Flat

Coudé Rotator Transfer Optics

M7 Wavefront conditioning mirror 540 mm Off-axis concave



hyperboloid

M8 Folding mirror 450 mm Flat

M9 High-order deformable mirror 220 mm Nominally flat

M10 First corrector mirror 275 mm Off-axis concave ellipsoid

M11 Second corrector mirror 175 mm Off-axis convex ellipsoid

M12 Folding mirror and AO beam splitter 275 mm Partially silvered flat

M13 Folding mirror 450 mm Flat

1.2.2 Thermal Control

The ATST design addresses the need to control so-called “self induced seeing,” which is seeing that

results from the presence of the telescope and telescope enclosure. This component of seeing tends to

contain higher spatial frequencies than atmospheric wavefront distortions, so the adaptive optics system is

less effective at correcting it when present. We have paid careful attention to this problem, particularly

because observations will be carried out during daylight hours when the sun can heat exposed

components to well above ambient air temperatures.

One of the common methods of controlling seeing at existing solar telescopes involves evacuating the

beam column within the optical system. This scheme is not possible with ATST because of the

requirement to observe simultaneously over a broad range of wavelengths, thus precluding use of an

entrance window at the top of the column. Instead ATST will use the approach of actively controlling the

temperature of insolated components such as telescope mirrors and the enclosure’s outer skin. Our

various studies and experiments have shown that if the temperature of these components can be

maintained close to or slightly below the ambient air temperature, self-induced seeing can be controlled

within the error-budget allocations. The details of this process are discussed item by item in the following

chapters.

Another aspect of the general thermal-control strategy involves active or passive flushing of surfaces

within or near to the optical beam. This is provided either by natural winds that enter the enclosure

through passive ventilation gates, or forced ventilation provided by fans that can be operated during

periods of little or no wind. These systems will be discussed in more detail below in the enclosure

description (Section 6).

1.2.3 Coronal Capabilities

The case for including coronal capabilities in the ATST design has already been made in the scientific

justification for the facility. In summary, while space-based instruments can generally deliver very sharp

images of the sun, it is the four-meter aperture of ATST that will make it a unique and powerful tool for

coronal science. An aperture of this size will allow scientists to do high-resolution coronal spectroscopy

– and hence polarimetry – to probe the magnetic properties of off-limb coronal features.

As we noted in the error budget discussion in Section 1.1.4 above, the requirement on image quality for

coronal use cases is relaxed considerably relative to the diffraction-limited observations. The diffraction

limit for a four-meter telescope operating at 1 μm is 0.06 arcsec, while the SRD specifies 0.7 arcsec (both

in terms of 50% encircled energy diameter) for coronal use cases. The emphasis here shifts to observing

relatively faint features that are very close (as close as 5 arcsec) to the bright limb of the sun. In addition

to the requirement to build ATST at a site with low levels of atmospheric scatter, several features of the

design presented in the following chapters are included specifically to address coronal science needs:



A Nasmyth platform is included, providing a faster beam with a minimum number of reflections

(described in Section 4.1.2) and minimum telescope polarization.

The coronal Near IR Spectro-polarimeter is housed in a cryostat to reduce thermal background

levels (described in Section 4.2.3).

An active occulting system is provided to block light from the photosphere while observing the

corona (described in Section 2.3.2).

Many aspects of the optical and mechanical design of the telescope (including the off-axis

configuration discussed above) are driven by the desire to keep stray light to a minimum.

This last bullet point addressing stray-light control is critical to the success of ATST. The science

requirements for coronal observations at the Nasmyth focus place strict and challenging constraints on the

stray-light performance. Detailed stray-light analysis on the ATST design confirms what experience has

shown with existing coronal instruments operated at excellent sites: when observing close to the limb of

the sun, instrumental scattered light is dominated by scatter off the telescope’s mirror. When mirror

surfaces are clean, the underlying microroughness imparted by the polishing process limits the

performance. By specifying an RMS microroughness of 2 nm, the science requirements can be

comfortably met. This level of polish is routinely achieved using modern polishing techniques, so this

requirement presents no manufacturing challenge.

It is found, however, that close attention must be paid to keeping the surfaces clean during coronal

observations. If only 0.01% of the mirror surface is covered by dust, the dust contribution to instrumental

scatter overtakes mirror microroughness to become the dominant factor. The length of time necessary to

reach or exceed the 2510-6

at 1.1 solar radii requirement will vary considerably depending on weather

conditions, but even under average conditions this contamination level can be reached in just a few days

of uncontrolled exposure. The ATST has adopted several strategies to mitigate and control dust

contamination and its effects:

Occulting of the sun’s disk at prime focus. The ATST design includes a reflecting occulter that rejects the

sun’s photosphere at the prime-focus image, passing only the corona. (See “Active Occulter Insert,” in

Section 2.3.2 below.) For the coronal instruments operating at the Nasmyth focus, this eliminates dust on

M2 and subsequent mirrors as significant contributors since the bright on-disk radiation never reaches

those mirrors. It also reduces stray light due to diffraction around the Lyot stop located just down stream

of M2, again because no direct power from the sun’s photosphere ever reaches that point in the optical

system.

In-situ cleaning and washing system. This system – part of the M1 mirror mount assembly – incorporates

a CO2 snow cleaning capability that is efficient and convenient for use as often as required during coronal

observations, and a wet washing station that can be used without removing the mirror from its cell. (See

“Cleaning and Washing System” in Section 2.2.2 below.)

Closed vent gates. As noted above, coronal observations have much-relaxed image-quality requirements

compared with on-disk diffraction-limited observations. Hence, the open vent gates and active flushing

that must take place during high-resolution observations is unnecessary, allowing better short-term

control of dust infiltration.

Block Scheduling. Whenever possible, coronal observations will be block scheduled to take maximum

advantage of freshly coated or recently cleaned optical surfaces and good observing conditions (dark sky

conditions and low dust levels).



Figure 2.1. Telescope Mount Assembly

2. TELESCOPE ASSEMBLY

The Telescope Assembly is comprised of (1) the telescope mount assembly; (2) the M1 assembly; (3) the

heat stop assembly; (4) the M2 assembly; (5) the feed optics; (6) system alignment; and (7) the

acquisition system. These are described in detail in this chapter.

Wherever possible, the designs of the Telescope Assembly and its subcomponents were based on

previous successful large telescope systems, including the structural layout, the servo and control systems,

the pier concept, and many of the mechanical subsystems. Where it was impossible to emulate existing

telescope designs, we have verified the ATST Telescope Assembly design by a variety of proven

methods. For example, full static and dynamic finite element (FE) studies have been performed on the

overall structure and pier. Transient thermal analyses, computational fluid dynamics (CFD) studies, and

various other calculations have been performed as well to support the design.

In addition to these analyses, a variety of potential telescope fabricators and vendors were involved in

design evaluation studies from early in the project. The purpose of these evaluations was to review the

subassembly designs, suggest technical improvements, and provide fabrication cost estimates. The results

helped refine the overall system design, and concentrated the project’s efforts on reducing technical risk

and improving overall telescope performance. Industry involvement of this type is especially useful in

addressing manufacturing concerns and logistics early in the design process.

2.1 TELESCOPE MOUNT ASSEMBLY

The Telescope Mount Assembly (TMA) appears in Figure 2.1. It provides structural support for the

major optics and instruments of the ATST observatory. It includes a variety of mechanical subassemblies,

bearings, controllers, drives, and equipment that are

used to point, track, and slew these optics and

instruments during science observations. The TMA is

comprised of six major components: (1) the Mount

and Drive System; (2) the Nasmyth rotator and drive

system; (3) the Coudé Rotator and Drive System; (4)

the Mount Control System; (5) the Pier, and (7)

Ancillary Mechanical Systems. These six items are

described in detail, below.

2.1.1 Telescope Mount Assembly Design Requirements

The Telescope Mount Assembly serves a number of

important roles and functions during science

observations. Of these, five are considered to be top-

level, or most important to the performance of ATST.

These top-level functional requirements are as

follows:

Optics Mounting: The TMA provides precise and

stiff mounting interfaces for the M1 through M13

mirror assemblies, heat stop, occulter assembly,

polarimetry optics, the acquisition system, and the

telescope alignment system. The nominal positions

and allowable deflections for all of these mounted

assemblies are derived from the Static and Dynamic

Optical Alignment specifications, which are derived



from the delivered image quality error budgets. The worst-case (i.e., most stringent) error budget terms

were then used to design the structural members and mounting points of the Telescope Mount Assembly.

A series of detailed FE analyses was performed to optimize and validate the design and mounting

interfaces.

Nasmyth Instrument Interface: The TMA provides a precise and stiff mounting interface for the

Nasmyth instrument rotator. The specifications of this interface were derived from the requirements of the

Nasmyth instrumentation, including such factors as geometric size, overall mass, required mounting

stiffnesses, thermal considerations, stray light requirements, and a variety of handling and operational

concerns.

Coudé Lab Instrument Interface: The TMA provides a precise and stiff mounting platform for all the

coudé-lab instruments. Structural designs and stiffnesses were flowed down from the respective error

budget provisions into the design. The overall layout and detailed design of the Coudé Rotator was

validated in terms of these requirements by way of FE analyses.

Pointing, Tracking, and Slewing: The TMA provides for accurate and repeatable pointing, tracking and

slewing of the ATST optics and instruments over their required full ranges of travel. The specifications

for pointing, tracking, and slewing are based on a combination of direct flow-down from the SRD and

from derivations of the delivered image quality error budgets (e.g., drive jitter).

Throughput, Thermal, Stray Light: The TMA provides for an unobstructed optical path from the sun to

the Nasmyth instrument station and to the coudé instrument station. It does this without imparting excess

thermal input into the beam (i.e., degrading seeing) or adding deleterious stray light into the science light

paths. A number of thermal and stray-light analyses were performed to verify that the TMA layout met

these requirements.

In addition to these top-level requirements, there are a number of second-level functions that the TMA

provides. For example, the TMA is designed to allow for periodic removal of the major optics for

servicing operations (e.g., M1 stripping and recoating). The TMA also provides a variety of features and

safety systems designed to protect personnel and the telescope from damage (e.g., failsafe M1 cover; GIS

interface, etc.). The complete specifications and design for the Telescope Mount Assembly, including all

the top-level and second-level requirements are outlined in the TMA Specifications Document (ATST

Document #SPEC-0011).

2.1.2 Telescope Mount Assembly Design Description

Mount and Drive System: The mount is

comprised of two major structural elements:

the Optics Support Structure (OSS), which

rotates about the altitude axis, and the Mount

Base, which rotates about the azimuth axis.

These are shown in Figure 2.2. The OSS

provides mounting interfaces for the M1

Assembly, the heat stop, the M2 Assembly,

feed optics M3 and M4, and the Nasmyth

rotator and instruments. The mount base

provides interfaces for the M5 and M6 coudé

transfer optics.

The structural design of the Mount employs

large steel weldments that have been stress

Figure 2.2. Telescope Mount Assembly, showing the OSS,

access platform, Mount Base, Azimuth Track and Bearings.



relieved prior to final machining. This technique has been very successful on many other recent telescope

projects, such as Gemini, SOAR, and WIYN. Detailed FE analyses were used to verify the basic design.

The Mount structure is configured with minimal bolted joints. This is critical to minimizing non-

repeatable errors that can affect telescope performance. The overall Mount layout provides high stiffness

(i.e., minimize static and dynamic flexure), is resistant to vibrations (e.g., wind-induced resonances), and

allows for direct load paths from the supported optic assemblies down into the structure, the bearings, and

ultimately the concrete pier and the ground. The major subcomponents of the Mount are described as

follows.

Optics Support Structure: The OSS is a performance-based design, configured to accommodate the

large bending loads of the off-axis optical layout. The M1 Assembly alone weighs more than 12,000 kg,

with its center of gravity cantilevered four meters horizontally from the altitude axis. This structural

challenge is met by an optimized two-piece layout of the OSS.

The bottom portion of the OSS carries the large off-axis loads via an arrangement of large square and

rectangular steel tubes that are optimized for bending loads. This system maximizes stiffness by making

the best use of the relatively stiff section properties (i.e., large area-moment of inertias) of the rectilinear

shapes. The upper portion of the OSS, in contrast, utilizes round steel tubing members to reduce weight. It

also results in minimal thermal mass, and reduces airflow obstructions on the upper portion, which

improves thermal flushing and minimizes the cross-sectional areas and coefficients of drag of the upper

OSS.

The upper portion of the OSS is joined to the lower via non-slip bolted joints. The two-piece

configuration eases manufacturing and helps facilitate transportation from the fabricator to the ATST

observatory site. When bolted together, the complete assembly is partially self-compensating for relative

M1-to-M2-to-altitude axis displacements as the OSS rotates from zenith to horizon. It also is extremely

stiff, allowing only minimal deflections and rotations of the heat stop assembly, M1-M2 optic assemblies,

and the Nasmyth instruments. The M1 Assembly is installed into the OSS from underneath via a

specialized handling/lifting cart.

Mount Base: The mount base assembly is a large machined weldment constructed of plate steel of

moderate thicknesses. All welds are full penetration-type welds. To reduce the possibility of structural

creep and/or hysteresis, all components are thermally stress-relieved after welding and prior to final

machining.

The mount base coudé transfer assemblies (M5 and M6) are supported on a reinforced vertical column

that rises from the top of the plinth up to the altitude axis. The static and dynamic structural performance

of the Mount Base has been verified with FE analyses.

Bearings and Azimuth Track: The mount bearing system is comprised of three major parts: i) the

azimuth bearings; ii) the altitude bearings; and iii) the hydraulic supply system that is located in the

support and operations building (S&O building). The azimuth bearings are comprised of two major

subassemblies: a) the axial bearing shoes; and b) the radial guide pads. The axial bearing shoes are

mounted to the underside of the mount base and bear vertically downward against the top surface of the

azimuth track. There are four of these shoes, each located in a stiff corner of the mount base. The altitude

bearing system is comprised of large bore rolling element bearings that are sized to stiffly carry the OSS

radial and axial loads, while minimizing start-up and running friction. The bearings are designed to be

replaced in-situ and fully lubricated by periodic preventative maintenance operations.

The azimuth track is a large diameter welded and machined steel structure that serves a number of

purposes. First and foremost, it provides a stiff and smooth surface upon which the azimuth axial

hydrostatic bearings ride. It also provides mounting surfaces for the azimuth drive ring gear assembly (on

the O.D. of the track) and the azimuth brake rotor (on the I.D. of the track). The track itself is constructed



of steel plate and is built in segmented sections that allow into relatively easy transportation and handling.

These sections are designed in such a manner as to facilitate reassembly at the observatory site and future

replacement of track segments in the event of damage. The azimuth track is designed to carry the axial

loads of the mount structure above it, via the azimuth axial bearings, transferring this load directly into

the pier. The azimuth track also serves as a large collection trough for the oil spent by the azimuth axial

hydrostatic bearings.

Mount Drive System: The mount drive system is comprised of the drive assemblies, encoders, fiducials,

brakes, and over-travel stops that allow slewing, pointing, and tracking of the mount structure. The

altitude and azimuth axes of the mount are driven by way of a gear-type drive system. To help minimize

the number of spare parts required, the same drive motors, amplifiers, and tachometers are used for both

axes of the mount. Drive motors are brushless DC-type. The design of the mount drive system is dictated

primarily by the pointing, tracking, and slewing requirements of ATST. The top-level specifications are

as follows:

Blind pointing < 5 arcsec

Offset-pointing < 0.5 arcsec

Tracking stability < 0.5 arcsec/hr

Tracking rate = solar rate

Slew speed = +3º/sec.

On the azimuth axis, a large precision helical-cut ring gear is mounted to the outside diameter of the

azimuth track assembly. This gear is built in replaceable sections that help facilitate shipping and

assembly on site. The gear segments are keyed together to ensure that there is no cogging or other drive

error when the pinion gears pass over the joint and so that assembly/reassembly can be readily facilitated.

The altitude drive system is comprised of two pairs of drive motors and gear heads. One pair is located at

the bottom of the inner surface of the +X-side mount column. The other pair is in a similar location on the

-X-side mount column. Large ring gear segments are bolted to the outer radius of the altitude disks.

Mount Thermal Control: The telescope mount has a large amount of surface area, much of it above the

level of the primary mirror. It is therefore

important that the mount temperature track

the ambient air temperature closely to avoid

self-induced seeing. The mount temperature

is controlled primarily by shading it from

direct solar radiation with the enclosure. In

the absence of solar heat loads, the main

source of temperature differences is the

thermal inertia of the mount. As the ambient

temperature rises in the morning and falls in

the late afternoon, the mount temperature

will lag the ambient air temperature by an

amount that depends upon the wind speed.

Wind passing over the mount helps to

reduce the lag time. In addition, provision is

made for drawing ambient air through the

interior of structural members to reduce the

thermal equilibration time.

Figure 2.3. The Nasmyth Rotator Structure.



Nasmyth Rotator and Drive System: The Nasmyth Rotator Structure is comprised of two major

components: the inboard support assembly and the outboard support assembly (Figure 2.3). These two

components provide support and rotation of the Nasmyth instrument. The range of travel of the Rotator is

± 270 degrees. The Nasmyth rotator drive system is comprised of the drive unit, encoder, fiducials, and

brakes that allow slewing, pointing, and tracking of the Nasmyth instrument. The drive system for the

Nasmyth structure is comprised of a direct-drive brushless DC-type motor (e.g., Kollmorgen) built into

the rotator structure inboard support assembly. SOAR has used a similar system with good results on its

Nasmyth rotator, as also have the SOLIS

telescope project on its mount drives.

Amplifiers are fully-programmable

devices from the same manufacturer as

the motor.

Coudé Rotator and Rotator Drive

System: The coudé rotator structure is

comprised of the coudé frame, the

support tower, the coudé bearings, the

coudé floor system, and the instrument

support beam assemblies (Figure 2.4).

The coudé frame is octagonal is shape

when viewed in plan. It is constructed of

large wide-flange and box beam

structural elements. These items are

joined together in large weldment

assemblies that are then bolted together

with high-strength bolts and precision dowels. The final configuration of this combination of weldments

and mechanical fasteners is intended to minimize the number of bolted joints yet still allow for relatively

easy and straightforward reassembly at the Site within the confines of the complete pier structure.

To accommodate the requirement for maximum configurability of the coudé lab (i.e., the positions and

orientations of the instruments located on the rotator), instrument support beams (ISB) and stand-offs are

used. The ISBs are supported on the flanges of the horizontal beams of the coudé frame. These ISBs can

be moved (slid) into the correct position and then locked down via a flange clamp at each end. The ISBs

are sized to fit flush within the coudé frame, while still maximizing their stiffness.

Coudé Rotator Drive System: Because the sizes and inertias of mount structure and the coudé rotator

structure are very similar, identical drive system components are used wherever possible. This simplifies

the final design effort, and also minimizes the number of spares that are required during operations.

Requirements were derived primarily from the worst-case pointing, tracking, and slewing requirements:

Blind pointing < 5 arcsec;

Offset-pointing < 0.5 arcsec;

Tracking stability < 0.5 arcsec/hr;

Tracking rate = nominally solar rate (dependent upon site);

Slew speed = +3º/sec.

Pier: The telescope pier is a large, steel-reinforced monolithic concrete structure that supports the mount

structure and transfers its loads into the soil. The telescope pier also supports coudé rotator in a similar

manner. The pier includes the foundations and interface to the soil. The pier also includes the stationary

floors, stairs, ladders, man-lift, crane, and other equipment attached directly to it.

Figure 2.4. Coudé Rotator.



Mount Control System: The Mount Control System (MCS) provides control software for the telescope

mount assembly. It is an integral part of the ATST telescope mount. The MCS operates all associated sub-

assemblies, including azimuth and altitude drives, coudé rotator, Nasmyth rotator, and thermal

management. It is controlled by the Telescope Control System (TCS) for all operations except low-level

engineering activities and safety interlock situations. The MCS is directly connected to the GIS to

perform safety operations.

Ancillary Mechanical Systems:

M1 Cover Assembly: The M1 Cover provides three key functions on the TMA:

1. Impact damage: The M1 Cover protects the primary mirror from damage whenever the telescope is

not observing the sun. Dropped tools and other types of impact damage are a significant danger over

the 40-year lifespan of an observatory the size and complexity of ATST. The M1 Cover is designed to

survive an impact load of a falling 2.5 kg weight released from a height of 15 m, and to carry the

weight of up to three workers simultaneously walking on top of it.

2. Contamination control: The M1 Cover keeps dust and other airborne particulates from collecting on

the mirror at night and during non-operational periods. A slight overpressure of dry air is maintained

underneath the cover, to reduce the infiltration of contaminants.

3. Thermal safety system: The M1 Cover serves as an integral part of the ATST thermal safety system.

The Cover is designed to rapidly close in an emergency event. This closure results in an interruption

of the light path from the sun to M1, and thereby causes a safe removal of focused light from reaching

the Heat Stop.

The M1 Cover design, shown in Figure 2.5,

is based on a traditional folding panel

system common to many modern large

telescopes. Aluminum honeycomb panels,

joined with full-length hinges, ride on

guide rails. The M1 Cover is opened via an

electric drive, and then held in that position

via an electromagnetic clutch. In the event

of a power failure, the clutch disengages,

and large fail-safe springs passively close

the cover in approximately 15 seconds.

Rotary dampers are used to smooth and

control the spring closure of the cover.

The M1 Cover is designed to operate in

any orientations of the OSS. When open,

the cover folds back into a low-profile

package inboard of the mirror, to minimize

air flow obstructions.

Cable Wraps: Powered cable wraps are employed on all three axes of the TMA (Mount Altitude, Mount

Azimuth, and Coudé Rotator Azimuth; see Figure 2.6) to manage the system utilities throughout their

ranges of travel. These utilities include AC and DC power lines, copper and fiber signal and data lines,

coolant supply and return hoses, compressed air and nitrogen lines, and various communication links.

Powered wraps minimize non-repeatable torque inputs that can affect the pointing and tracking

performance telescope by isolating utility cable stiction and slip from being input into the telescope

structure. The wraps are mechanically isolated from the telescope, and they are continuously slaved to

follow the respective TMA axis rotation.

Figure 2.5. M1 Cover



Sensor Arrays: The Telescope Mount Assembly is outfitted with an array of functional, safety, and

diagnostic sensors that are distributed on and around the system. These sensors are used to continuously

monitor the performance and health of the TMA, and to provide feedback to the Telescope Control

System (TCS), Observatory Control System (OCS), and Global Interlock System (GIS) systems. The

sensor arrays include (1) Thermal Sensors; (2) Vibration Sensors; (3) Wind Velocity Sensors; and (4)

Bearing Health Sensors.

2.2 M1 ASSEMBLY

The M1 Assembly (Figure 2.7) is the heart of the

ATST telescope; it contains the primary four-meter

diameter off-axis mirror (M1) that is the first element

in a chain of optics that collects and focuses the solar

energy into high-resolution images. The assembly

consists of the four-meter diameter M1, the axial and

lateral support system for M1, the M1 cell, the M1

thermal control system, M1 cleaning and washing

system and M1 control system. The M1 assembly

defines the position of M1 and maintains its optical

figure under the operating conditions of changing

gravity load, thermal conditions and wind loading.

The M1 support system also has active optics

capability, to slowly adjust the figure of M1 to

compensate for a wide range of effects, including

changes in zenith angle and thermal conditions.

2.2.1 M1 Design Requirements

The SRD defines the aperture of the telescope as four meters. The M1 is consequently sized at 4.24

meters diameter to yield a 4.0-meter clear aperture after the necessary baffling of the outer edge and

accommodation of the inclination angle of M1 with respect to the incoming solar beam. The optical

quality of M1 is critical to maintaining the required solar image quality; a surface figure of 32 nm rms

must be maintained over the operational limits of 0 to 80 zenith angle (changing gravity vector),

thermal conditions and wind loading. A comprehensive M1 Assembly Specification has been developed

to address these requirements (ATST Document #SPEC-0007).

Figure 2.7. M1 Assembly.

Figure 2.6. Azimuth (left) and Altitude (right) cable wraps.



120 AXIAL

SUPPORTS

24 LATERAL

SUPPORTS

Figure 2.9. Axial and Lateral Support Point Locations

2.2.2 M1 Design Description

M1 Mirror: The M1 Mirror is a 4.24-meter diameter constant thickness meniscus, approximately 100

mm in thickness. The physical configuration is an off-axis paraboloid with a focal length of 8000 mm,

effectively part of a 12.1-meter diameter f/0.67 parent paraboloid as shown in Figure 2.8.

The material for M1 is ultra-low expansion fused silica or glass-ceramic; this choice was driven by the

large temperature gradients from the front to the back of M1 due to solar loading and thermal control.

Two materials are available at the four-meter scale for an M1 blank: ULE from Corning, Inc., New York

and Zerodur from Schott Glasswerke of Germany. Both materials have a long and well-established

history of use in astronomical and solar telescopes.

M1 Support System: The function of the M1 Support System is to support the weight of the M1 and

maintain its nominal surface figure to within 80 nm RMS over the operational zenith angles and thermal

environments of the telescope; it also defines the position and orientation of M1. In addition, the support

system makes small changes to the surface figure of M1 by applying active optics correction forces

through the axial support actuators.

The support system is composed of an array of axial

supports and an array of lateral supports. The axial

support system consists of 120 discrete support

actuators arranged in five concentric rings on the

back of the mirror. The body of each axial support

actuator is attached to the M1 cell and a support rod

from each actuator is attached to a load spreader

bonded to the back of M1. The lateral support system

consists of 24 discrete support actuators arranged

around the periphery of M1. The body of each lateral

actuator is attached to a bracket at the outside edge of

the M1 cell and a support link from each lateral

actuator is attached to an Invar pad bonded on the

outer edge of M1. Figure 2.9 shows the arrangement

of axial and lateral support points and arrows that

represent the force vector applied by each support.

Ø12100 PARENT PARABOLOID

4000

(4100)

100

ALL DIMENSIONS IN MM.

Ø4237

SURFACE A

SURFACE B

ATST PRIMARY MIRROR

PARENT

PARABOLOID

GEOMETRICAL AXIS OF

ATST PRIMARY MIRROR BLANK

AXIS A,

GEOMETRICAL AXIS OF

PARENT PARABOLOID

A A

SECTION A-A

Figure 2.8. M1 Mirror.



Finite element analysis was used to optimize the axial and lateral support systems, a process that has been

used for the last two decades for the design of modern astronomical telescopes. However, whereas

nighttime telescopes optimize for best performance at zenith and allow slow performance degradation

with increasing zenith angle, ATST requires a high level of performance at all zenith angles, with

optimum performance at zenith angles of 65° to 80° to match the best atmospheric seeing profiles

measured at the chosen Haleakala site. First, the radial location and force value of the five rings of axial

support actuators was determined. The goal in this process was to reduce the deflections of the optical

surface of M1 caused by gravity at a zenith pointing position to an acceptable level that meets the error

budget requirements for image quality. During the optimization, it was assumed that the force applied by

all the axial actuators in any given ring is equal (this clearly follows from the rotational symmetry of M1).

Also, the nominal force values for each of the five rings were limited to a range to allow one actuator to

service all rings.

Since the back of the meniscus mirror is

smooth and continuous, there were no

constraints on the radial locations of the

five rings. The optimization was successful

in yielding a surface figure accuracy of 18

nm rms with a force value of 180 N for

ring 1 and 320 N for rings 2 through 5.

These force values are for the passive

support of M1; small changes up or down

from the nominal value will be made for

active optics correction. Figure 2.10 shows

a plot of the optical surface deformation of

M1 at zenith pointing on the 120 axial

supports.

A similar optimization routine was carried

out for the lateral support system. In this case M1 is at the position for a telescope zenith angle of 80º,

oriented in a near vertical, slightly over-hanging configuration. The 24 lateral support points on the

perimeter of M1 are equally spaced 15º apart and symmetrical about the vertical axis as shown in Figure

2.9. This arrangement was chosen to allow adequate clearance between actuators. During the optimization

process, the force level and vector orientation of each support was varied. Figure 2.11 shows a plot of the

optical surface deformation of M1 at a zenith angle of 80° with active optics correction applied.

50 nm

-80 nm

50 nm

-80 nm

Figure 2.10. M1 Optical Surface Deformation for a Zenith

angle of 0°

Figure 2-11. M1 Optical Surface Deformation for a Zenith angle of 80°

Figure 2.12. M1 Cell Configuration



M1 Cell: The M1 cell is a steel structure, with a structural plate on the rear and a honeycomb rib

structure attached to the rear plate (see Figure 2.12). The cell will support the axial and lateral support

mechanisms and the air distribution system for the thermal control of M1, and interface to the OSS. The

axial support mechanisms extend through holes in the rear plate of the cell and are serviceable from the

rear of the cell. The overall design of the M1 cell is proven, having been used on many existing nighttime

telescopes such as ESO NTT, WIYN and Gemini.

Thermal Control System: The M1 receives the largest amount of solar heat load of any optic and also

contributes the most to localized “seeing” that can degrade the quality of the solar image. The purpose of

the M1 thermal control system is to remove the solar energy that is absorbed by M1 and to maintain the

M1 optical surface temperature as close to ambient as possible. Both analysis and empirical observation

has shown that maintaining the optical surface of M1 at or slightly below ambient temperature, combined

with wind flushing of the surface will minimize local seeing to acceptable levels.

The M1 thermal control system is an array of air

jets, or tubes, located behind the rear surface of

M1 that direct conditioned air against the rear

surface. Several hundred air jets are fed by a

network of larger distribution tubes; the array of

air jets and distribution tubes are divided into six

zones with each zone fed by its own fan and

liquid/air heat exchanger (see Figure 2.13).

Considerable analysis has been performed to

determine the effectiveness of the thermal control

system under different conditions. There is a

significant time lag between when a change in

cooling air temperature is applied to the backside

of the mirror and when this change is seen on the

front optical surface due to the thickness and

relatively poor thermal conductivity of the M1

substrate. However, the diurnal change in

temperature over the observation period is

predictable and daily temperature models will be developed and used as base curves to drive the mirror

temperature. This approach, combined with wind flushing across the optical surface of M1, allows the

temperature of the optical surface to be maintained within 1 to 3 ºC below the ambient air temperature.

Cleaning and Washing System: The M1 Assembly will have the capability of cleaning the M1 on a

daily basis and in-situ washing of M1 on a periodic basis. A CO2 dispersal device will be attached to the

M1 cover for cleaning of M1 at the beginning of each day (Figure 2.14). This will be done with the

telescope in a near horizon pointing position; particulates and other matter removed from the surface of

M1 during the cleaning operation will be collected by a vacuum trough and removed from the area.

During in-situ washing, the telescope is moved to a near horizon pointing position and a collection trough

is positioned at the lower edge of the mirror to collect cleaning effluent (Figure 2.15). The optical surface

of M1 is washed, rinsed and subsequently dried with an air knife that is mounted on the telescope mount

next to the M1 cell.

Figure 2.13. M1 Air Jet Distribution System



M1 Control System: The primary tasks of the M1 Control System are to control the application of active

forces to M1 and to control the M1 thermal management system. The system accepts input mirror figure

information at up to 10 Hz and blends and averages this figure information at up to 0.1 Hz. It also controls

the temperature of the front side of M1 and the aperture stop to within a pre-determined range around

ambient temperature. The system will also store and apply a 24-hour thermal profile estimation to be used

in the thermal control of M1. The M1 Control System also provides status information at up to 10 Hz and

interfaces to the TCS, GIS and OCS.

2.3 HEAT STOP ASSEMBLY

The main purpose for arranging the ATST optics

in a Gregorian optical configuration is to reject

energy at prime focus before the concentrated

beam is directed onto M2. The full 32-arcmin

solar image is formed at prime focus, but only a

five arcmin part of that image is allowed to

proceed onto subsequent elements of the optical

system (Figure 2.16). The rest is reflected,

trapped, and the heat is pumped away.

In many observing scenarios (on-disk and off-

disk coronal), the heat stop assembly is simply

required to block the occulted field (OF) and pass

the field of view (FOV). Observations very near

the solar limb, however, require the prime focus

occulter to quickly and actively track the solar

limb.

2.3.1 Heat Stop Design Requirements

The heat stop assembly provides five top-level functions. They are as follows:

Block OF: The heat stop assembly blocks (reflects) solar disk light at prime focus over an area

sufficient to allow off-pointing as much as 2.5 solar radii (SRD requirement, approximately 82

arcmin).

Pass FOV: The heat stop assembly allows a 5-arcmin FOV to pass to M2.

Figure 2.14. CO2 dispersal as M1 cover opens.

Figure 2.15. Edge seal and collection trough for M1

washing.

Figure 2.16. Cross sectional view of the heat stop

reflecting cone and trap when the telescope is pointed at the center of the sun.



Track Solar Limb: During near-limb coronal observations the heat stop assembly occults limb

light while actively compensating for telescope shake and atmospheric seeing. It must permit

coronal observations as close as 5 arcsec of the solar limb.

Remove Irradiance Load: The heat stop assembly removes the prime focus irradiance load of up

to 2.5 MW/m2 from the optical path.

Minimize Self-Induced Seeing: The heat stop assembly introduces no more seeing than the error

budgets allow. Experiments and scaling laws for small hot objects near M2 indicate insensitivity

for seeing-limited observations (e.g., Beckers and Melnick, 1994, and Zago, 1997). A reasonable

bottom line requirement is that surface temperature must be kept within some 10 ˚C of the

ambient air temperature.

In addition to the five top-level requirements, there are a number of second-level requirements such as

easy periodic removal of major subsystems for servicing and replacement, and safety systems to protect

personnel and the telescope from damage.

The complete specifications and design for the heat stop assembly are outlined in the Heat Stop Design

Requirements Document (ATST Document #SPEC-0003).

2.3.2 Heat Stop Design Description

The heat stop assembly consists of a reflector assembly, an active occulter insert, a coolant loop, a plume

control system, a beam dump, and control and interlock systems.

Reflector Assembly: The Reflector Assembly is the heart of the heat stop assembly (Figure 2.17). The

assembly is designed to remove high heat flux with minimum temperature rise in a compact package. The

reflector cone is the first component encountered after M1 and sees the solar image at prime focus, nearly

2.5 MW/m2 irradiance at noon. The reflector is made of a highly conductive, high strength alloy of

copper, coated with AlMgF2 to provide high reflectivity. The reflector is cooled from behind by an array

of liquid jets.

Outside the reflector assembly lies the safety shield, a ceramic ring mounted on the M1 side of the heat

stop assembly to provide passive irradiance protection in the event of a tracking or power failure.

Figure 2.17.Three-D view of the reflector

assembly and heat trap.

Figure 2.18. The Active Occulter Insert



Figure 2.19. M2 Assembly.

Active Occulter Insert: The Active Occulter Insert (AOI) is a small conical device that fits into the

conical interior of the Reflector Assembly (Figure 2.18). The AOI provides a solar limb shaped occulting

edge in the center of the FOV for observations very near the limb (as close as 5 arcsec). The AOI rotates

to follow the solar limb and senses and tracks the solar limb at rates of several tens of Hz. The poor

quality of the prime focus image away from the center of the field of view will require that the radius of

the occulting edge be somewhat greater than that of the sun’s image.

Plume Control System: The Plume Control System (PCS) creates a flow of air across the Reflector

Cone that sweeps away buoyant flows. The PCS consists of a blower, a getter, supply fans, and ductwork.

Coolant Loop: The Coolant Loop supplies the Reflector Assembly with coolant of the proper

temperature, pressure, and flow rate. The Coolant Loop consists of a heat exchanger that transfers energy

to the primary coolant supplied by System Services, a secondary coolant that is both highly effective at

transferring heat and compatible with optics and mirror coatings, pumps that circulate the coolant, an

accumulator that stores a sufficient volume of coolant to supply the HSA during emergency conditions,

pressure relief valves, and an array of functional, safety, and diagnostic sensors.

Beam Dump: The Beam Dump absorbs the irradiance reflected

from the Reflector Plate. The irradiance at the Beam Dump is

spread out over a large area cooled by liquid from System

Services.

Control and Interlock Systems: The Heat Stop Control System

(HSCS) is responsible for the control and coordination of the

HSA, including the pump speed, system pressure, coolant

temperature, and the sundry sensors. The HSCS provides all

software interfaces for these components to the TCS, OCS, and

GIS. The HSA Local Interlock Controller provides an

independent safety override.

2.4 M2 ASSEMBLY

The M2 Assembly (Figure 2.19) contains the 635-mm diameter off-axis mirror (M2) that is the second

element in a chain of optics that collects and focuses the solar energy into high-resolution images. The

assembly consists of M2, the M2 positioning system composed of a hexapod and fast tip-tilt mechanism,

the M2 thermal control system and the M2 control system. The M2 positioning system defines the

position of M2 and maintains its position under the operating conditions of changing gravity load, thermal

conditions and wind loading. The M2 positioning system also provides fast tip-tilt motion to compensate

for some aspects of atmospheric seeing.

2.4.1 M2 Design Requirements

The M2 is sized at 635-mm diameter to yield a 5-arcmin unvignetted field after the necessary increase in

diameter to eliminate outer edge effects. The optical quality of M2 is critical to maintaining the required

solar image quality; a surface figure of 32 nm rms must be maintained over the operational limits of 0 to

80 zenith angle (changing gravity vector), thermal conditions and wind loading. A comprehensive

Design Requirements Document (ATST Document #SPEC-0008) has been developed to address these

requirements.

Interface requirements for the M2 Assembly include interfaces to the OSS, the M2 lifter used to install

and remove M2 from the assembly, the TCS and the required utility services.



Figure 2.20. M2 Configuration

2.4.2 M2 Design Description

The M2 is a 635-mm diameter structured mirror, approximately

75-mm in thickness at the center. It consists of a continuous

facesheet with a triangular rib pattern on the backside (Figure

2.20). Bosses are provided at three areas to allow attachment of

the mounting flexures. The optical configuration is a concave

off-axis ellipsoid with radius of curvature of 2081 mm.

The baseline material for M2 is silicon carbide. This choice was

driven by the requirement for extremely low mass and high

stiffness to achieve the desired fast tip-tilt motion; in addition,

the excellent thermal conductivity of silicon carbide minimizes

optical surface deformations under the solar load. Finite element

thermal and structural analysis was employed to evaluate

several potential materials and it was shown that silicon carbide

has the best optical and thermal performance under the ATST

operating conditions. Figure 2.21 shows a global figure change

of only 40 nm P-V for a silicon carbide M2 during the peak

solar load.

There are many different processes for fabricating silicon carbide mirror substrates, but special attention

has been given to CVD (chemical vapor deposition), reaction bonding and sintering since each of these

processes has a demonstrated capability of producing a blank of the required size. ATST personnel have

been in contact with major silicon carbide manufacturers during the design and development phase to

determine the most feasible, lowest risk and cost effective methods of M2 blank fabrication.

Positioning System: The function of the M2 positioning system is to support the weight of M2 and

define its position and orientation over the operational zenith angles and thermal environments of the

telescope. In addition, the positioning system provides fast tip-tilt motion of M2.

Figure 2.21. Global figure change for SiC substrate during peak solar load - 40 nm P-V



The M2 positioning system is composed of a commercial off-the-shelf hexapod, a fast tip-tilt mechanism

and a three-point flexure support that connects M2 to the fast tip-tilt mechanism. The hexapod provides

six degree of freedom movement to allow x-y positioning, focus, tip-tilt and rotational orientation of M2

with respect to M1. The fast tip-tilt mechanism provides rigid body motion of M2 at a rate of up to 10 Hz

and an amplitude of 5 arcsec to counteract certain types of atmospheric seeing.

Thermal Control System: Similar to M1, M2 receives a significant amount of solar heat load and also

contributes to the localized “seeing” that can degrade the quality of the solar image. The purpose of the

M2 thermal control system is to remove this solar energy that is absorbed by the optical surface of M2

and to maintain the M2 optical surface temperature as close to ambient as possible.

The M2 thermal control system is an array of air

jets, or tubes located behind the rear of M2 that

direct conditioned air into each triangular pocket

of the structured mirror (Figure 2.22).

Approximately 140 air jets are utilized, fed by a

manifold system that provides conditioned air

from a fan and liquid/air heat exchanger.

Ancillary Equipment: The ancillary equipment

consists of the necessary utilities services for the

M2 assembly. This includes the conditioned

coolant for the M2 thermal control system.

Control System: The M2 Control System

monitors and controls the M2 positioning

system, the M2 thermal system and the M2 fast

tip-tilt system. Control of the positioning system involves taking wavefront correction input and making

the necessary changes in the position of M2 by moving the appropriate actuators on the hexapod. Look-up

tables will also be utilized to compensate for slow and predictable changes in the position of M2 with

respect to M1 due to deflections in the optical support structure over changing zenith angles.

2.5 FEED OPTICS

A train of smaller reflective optical components, both flat and powered, are used to transfer the solar

beam from the Gregorian focus of the moving telescope assembly to the stationary observing floor, then

down to the coudé observing rooms. Their specific functions were discussed above in Section 1.2.1, and

illustrated in Figure 1.5. Their optical properties were summarized in Table 1.6.

As with M1 and M2, all of these mirrors receive significant solar heat loads, and active cooling and

thermal control will be necessary at various levels to maintain their optical surface temperatures at or near

ambient air temperatures.

Most of the feed optics in the ATST are lightweighted blanks made from high conductivity material such

as silicon carbide that are impingement cooled from the rear. The material properties combined with thin

stiffening ribs and facesheets allow the required heat transfer with only small gradients through the optic

and a much faster response time which allows working fluid temperatures closer to ambient.

Finally, there are a few optics where the absorbed flux is so high that only water cooling will provide

sufficient heat transfer coefficients to maintain optical surface temperatures within their limits. These

optics are expected to be fabricated of silicon carbide or high conductivity copper using either pin-post or

channel flow heat exchanges behind the optical surface. Many examples of this type of design can be

found in synchrotron optics, where significantly greater heat transfer efficiency is required.

Figure 2.22. Thermal control air jet nozzles directed

toward back of M2.



2.6 SYSTEM ALIGNMENT

The liberal use of off-axis optics within the ATST design presents special challenges for both initial

installation alignment and maintaining alignment of the optics within supporting structures that deflect

due to gravitational and temperature changes. These challenges are dealt with by a combination of initial

alignment procedures that ensure the optics are positioned accurately and by an active system that

maintains bore-sight and wavefront quality during operation.

2.6.1 Initial Alignment

System alignment components are provided to facilitate initial rough alignment of the telescope,

realignment after optical elements are removed and reinstalled as part of operational maintenance

procedures, and realignment at other times as required. These would include theodolites, targets, laser

positioning systems, and portable wavefront sensors.

Initial alignment of the optics must be done relative to the mechanical constraints of the azimuth and

altitude axes of the telescope. Therefore, the procedure for populating the observatory with optics starts

with the mirrors that define that portion of the beam path. These would include the OSS and mount base

assembly transfer optics for both the coudé and Nasmyth optical systems (see table 1.6 in section1.2.1 for

an overview of the optical components).

Next M1 and M2 are installed and adjusted in position relative to the transfer optics. This is first done

using mechanical datums referenced to optical metrology testing of the mirror figure, and positioning

them with accuracy better than 100 microns. After this, progressive nighttime optical testing at the prime

focus, Gregorian focus and then intermediate (coudé) or Nasmyth focus will be possible to confirm or

adjust alignment of the optics. Gross M1 deformations and alignment errors producing coma and other

low-order aberrations apparent in the point-spread function will be removed based on guidance provided

by the optical alignment sensitivity analysis at various field points. Wavefront analysis will then be used

while observing bright stars – again depending on the sensitivity analysis for guidance – to complete

alignment of the M1 through M6/MN5, and initial calibration of the active optics system.

Installation of the remaining coudé mirrors will than be done first using mechanical datums relative to the

intermediate focus and telescope azimuth rotation and optics testing at night will continue until the entire

optical train is populated, prior to working with extended sources and the thermal issues that come with

working during the day.

2.6.2 Active Alignment

Once initial alignment is achieved it must be maintained. Finite element analysis of the telescope’s

optical and mechanical structures has given us guidance about what sort of flexure to expect as the

telescope tracks the sun. It is clear that while our mount will be stiffer than many modern nighttime

telescopes, it is still insufficient to maintain correct positioning and alignment of the optics as the

temperature and gravity vector changes. We have demonstrated that wavefront errors can be

compensated by repositioning M2, and bore-sight can be maintained with two flat mirrors in the coudé

transfer optics. The information necessary to do this will be provided by one on-axis wavefront sensor,

and a minimum of three off-axis wavefront sensors.

What complicates this quasi-static alignment algorithm is the simultaneous presence of M1 figure errors

with similar spatial and temporal frequencies to those caused by telescope misalignment. These two

sources of error must be separated from each other in the data from the suite of low-order wavefront

sensors observing multiple field points. We have modeled the telescope optics to allow realistic

perturbations to both mirror figure and system alignment, allowing us to evaluate correction algorithms.



The work performed to date has shown no difficulty maintaining the required alignment error levels using

realistic deformations and compensators.

2.7 ACQUISITION

The Acquisition System provides a full-disk image of the sun that can be used by the observer to select

and acquire solar features. It is a small auxiliary telescope that will yield one-arcsec resolution images

through a relatively narrow-band H-alpha filter. Other alternative filters (such as one centered on the

calcium K-line) may also be provided.

In addition to the acquisition function, this system may furnish the most basic level of guiding feedback

during observations (limb guiding) to the tracking system to allow limb guiding during off-disk coronal

observations.

3. WAVEFRONT CORRECTION

The Wavefront Correction system includes the Adaptive Optics system and the wavefront-sensing

elements of the active optics system. The individual subsystems include:

A high-order adaptive optics (HOAO) system. This subsystem corrects atmospheric seeing at > 2

kHz rates. The baseline design has a 1369-actuator Deformable Mirror (DM) and a fast tip/tilt

mirror. The wave-front sensor is a correlating Shack Hartmann sensor with 1280 subapertures.

The approach builds on the very successful AO systems deployed at the Dunn Solar Telescope.

Correlation Trackers. Both the Nasmyth and coudé stations will be equipped with tip/tilt sensors

that can be used to provide image motion compensation at a fast rate. Either driving the

secondary mirror for the Nasmyth focal station or the tip/tilt mirror of the coudé transfer optics

accomplishes this task.

Active optics (aO) systems. The main task is to correct slowly changing aberrations that may

arise from gravitational and thermal deformations of the telescope structure. One of the main

objectives of the system is to keep the figure of the primary mirror within the allowed tolerances.

The secondary mirror will also be used as an active element, for example, to correct focus terms.

Alignment. The ATST’s off-axis optical system alignment requires wavefront measurements at

several points within the extended field of view. These multiple field wavefront sensors will be

available at both Nasmyth and coudé stations.

Blending. Information from different wavefront sensors (e.g., AO and aO) will be conditioned

and combined by the Wavefront Correction Control System (WCCS), which then drives the

appropriate corrector elements.

The ATST has several correctors and sensors for wavefront correction including: Quasi-static alignment

(QSA) for keeping the entire optical path– most importantly M1 and M2– aligned in closed-loop. The

active optics (aO) system’s main function is to keep the figure of M1 within spec compensating for

deformation due to gravitational and thermal distortions. Tip/tilt devices are provided for image

stabilization. High order adaptive optics (HOAO) correct atmospheric and internal seeing and residual

optical aberrations.

The sensors are:

Nasmyth active optics wavefront sensor for both QSA and M1 figure.

Nasmyth correlation tracker sensor.

Coudé aO wavefront sensor for QSA and M1 figure.

Coudé LOAO wavefront sensor for a correlation tracker image stabilizer.



Figure 3.1. Wavefront correction block diagram

Coudé HOAO wavefront sensor for high order image corrections.

The correctors are:

M1 has 120 axial actuators that support the mirror and are able to correct M1 figure errors up to

Zernike spherical mode.

M2 is mounted on a hexapod and thus has 6 degrees of freedom for correcting telescope

alignment and focus. It is also used for image stabilization when the telescope is in Nasmyth

mode.

M3N has slow tip/tilt motion for telescope alignment. M3N is used when the light is fed to the

Nasmyth station.

M3 has tip/tilt motion for telescope alignment. M3 and after are used when the light is feed to the

coudé lab.

M5 is a fast tip/tilt mirror for image stabilization correcting both motion from the atmosphere and

telescope shake at coudé.

M6 has slow tip/tilt motion for telescope alignment.

M9 is a deformable mirror used to correct high frequency high order seeing affects.

The mount will accept offset commands to keep the image stabilization mirrors centered in their

travel.

Each of the wavefront sensors has a computer that processes the wavefront information in real time and

outputs information to one or more correctors. This is done either directly where high bandwidth is

required (e.g., DM) or through the Wavefront Correction Control System (WCCS).

The WCCS is a supervisory computer that coordinates all the wavefront correction systems. It accepts

commands from the Telescope Control System and passes them on to the appropriate system. It

determines which of the sensor systems controls which of the correctors. In some cases it blends

information from different sensors. An example is QSA, where input from the HOAO wavefront sensors

and the aO wavefront sensors is combined and passed on to a reconstructor that drives the appropriate

corrector elements. A solution to the inverse problem of building the QSA reconstructor is described

elsewhere in this conference. Figure 3.1 shows a functional block diagram of the ATST Wavefront

correction system.



3.1 ADAPTIVE OPTICS

Adaptive optics (AO) is a critical technology that is essential in achieving the science goals of the ATST.

AO will enable diffraction-limited imaging to resolve the fundamental scales in spectroscopic and

polarimetric observations of solar fine structure, which generally require long exposures. Compared to

nighttime AO, solar AO faces a number of different challenges, and solar AO systems are in some aspects

technically more challenging than nighttime AO. The main challenges are the inferior daytime seeing, the

fact that solar astronomers mostly observe at visible wavelengths – although infrared observations are

becoming increasingly important – and the solar wavefront sensor that has to work on low-contrast,

extended, time-varying objects such as solar granulation.

3.1.1 Adaptive Optics Design Requirements

The requirements for diffraction-limited observations are discussed in detail in the SRD, and listed in

above in Section 1.1.4. In summary the requirements are as follows:

The ATST shall provide diffraction-limited observations (at the detector plane) with high Strehl

(S>0.6 required, S>0.7 goal) at 630 nm and above during excellent seeing conditions (r0 (630 nm) >

20 cm) and S > 0.3 at 500 nm and above during good seeing (r0 (500 nm) = 7 cm).

1. The wavefront sensor must be able to lock on granulation and other solar structure, such as pores

and umbral and penumbral structure.

2. Time sequences of consistent image quality are required for achieving many of the science goals.

Spectral or spatial scans often suffer from varying image quality during the scan. The AO system

shall provide consistent image quality during varying seeing conditions (time scales of seconds)

often encountered during the day-time.

3. The AO system shall correct residual (not corrected by active optics) optical aberrations and self

induced and atmospheric seeing to the performance levels specified in the SRD. Mirror seeing or

internal seeing in general must be avoided and any “residual” local seeing components must be

correctable by adaptive optics.

4. The AO system shall be robust enough to perform during transparency fluctuations typically

encountered in thin cirrus clouds.

The detailed AO systems parameters and specifications that flow from these requirements depend heavily

on the site characteristics, such as median r0, range of temporal fluctuations of r0, Greenwood frequency,

and isoplanatic patch size. Now that Haleakala has been selected as the site for ATST, good estimates of

these parameters are now in hand.

For the purpose of defining the baseline AO system we originally used the average of the median r0

values at two sites, measured at the telescope height. The average r0 is about 10 cm. The bandwidth

requirements for the AO system were derived from subaperture tilt spectra measured at the Big Bear Solar

Observatory (BBSO) site and the Sac Peak site using a wavefront sensor (WFS) with 10-cm subaperture.

The power spectra were used to estimate the Greenwood frequencies over a range of seeing conditions. A

detailed performance error budget analysis was performed to define the baseline AO system using the

average median r0 of 10 cm and the average Greenwood frequency (~32 Hz). These design parameters are

being adjusted now that Haleakalā has been selected and the telescope height has been set.



3.1.2 High-Order Adaptive Optics Design Description

The system design is modeled very closely after the successful high-order AO system operated at the

Dunn Solar Telescope (DST). The system is based on a correlating Shack Hartmann wavefront sensor and

uses a parallel processing approach using commercially available digital signal processors (DSPs).

The ATST HOAO system will be located in the coudé observing station. It will run at approximately a

2000 Hz frame rate with a resulting –3dB bandwidth of at least 200 Hz. It will consist of a Shack-

Hartman wavefront sensor with a 40×40 square grid lenslet array, an 800×800 pixel camera, a 64 DSP

processing system and a deformable mirror with a 41×41 square grid of actuators. It measures the

wavefront by cross correlating an array of subaperture images with a reference image.

Wavefront Sensor: The WFS is a correlating Shack-Hartmann WFS similar to the one successfully used

for the DST and BBSO AO systems. The principle of the WFS is shown in Figure 3.2. The telescope

aperture is sampled by an array of lenslets, which forms an array of images of the object (e.g.,

granulation, sunspots). Cross-correlations between subaperture-images and a selected subaperture-image,

which serves as reference, are computed. These cross correlations are shown in Figure 3.2, upper right.

By locating the maximum of the cross-correlation we determine the displacement of the images with

respect to the reference, thereby measuring the local wavefront tilts.

The pixel resolution in the SH-WFS images and the 2-D cross-correlation, respectively, is typically 0.5

arcsec. The FOV is about 1010 arcsec. Image displacements are computed to subpixel precision by

fitting a parabola to the correlation peak using and interpolating between pixels. A tilt map is shown in

the lower right corner of Figure 3.2. We use a modal reconstructor to derive the actuator drive signals.

Figure 3.2: Principle of correlating Shack-Hartmann wavefront sensor



The optical portion of the wavefront sensor will consist of a field stop, lenses, filters and a camera.

Following a beam splitter that brings a small percentage of the telescope light to the HOAO system is a

lens that brings to a focus the surface of the sun on to a square field stop that is approximately 10 arcsec

square. A second lens puts a pupil image on the lenslet array. The area of the pupil image on each of the

lenslet array lenses corresponds to a 10 cm by 10 cm area on M1. Following the lenslet array is another

lens that collimates the light onto a zoom lens and the camera. The images from each of these

subapertures will fill 20×20 pixels on the camera.

Wavefront sensor camera: A custom development is needed to produce the approximately 800×800

pixel, > 2 kHz frame rate wavefront sensor camera. The output from the camera will be 32 digital

channels each running at 40 Mpixel per second.

Wavefront Sensor Prossessing Unit: The wavefront sensor/reconstructor processing unit is also well

within the capabilities of existing technology. Utilizing improved DSP technology the ATST processing

unit can be built with a very moderate increase in size and complexity (Figure 3.3). However, we believe

that a parallel processing approach will be essential for the ATST AO not only to achieve the processing

rates required but also the high data throughput needed to achieve the > 2 kHz update rate that will be

needed for this system.

The camera channels will feed into a custom interface that will sort the pixels into subaperture order and

output them on 16 channels directly to 16 blocks of 4 DSPs (baseline: Analog Devices TigerSharc). Each

of the 64 DSPs will cross correlate a reference image with 20 of the 1280 subaperture images. The

multiplication of the resulting vector of image shifts with the pre-computed reconstruction matrix will be

performed by the DSPs as well. A servo algorithm will be applied to drive the 1369 DM actuators in

closed loop. Tip and tilt are separated out and corrected by the fast tip/tilt mirror M5. The HOAO system

will also provide time-averaged low order Zernike terms, which will be off-loaded to aO and possibly

QSA through the WCCS. The WCCS provides these averaged Zernike terms to the TCS for M1 figure

Host Computer/HOAO Control System

DSP

To

Mirror

Inter-

face

SMART

INTERFACE

Camera

To

DSPs

Sorts

Pixels

Into

Subaperture

Images

CAMERA

800x800

32 ports

40 MHz

2000 fps

Fast

Tip/Tilt

Mirror

Deformable

Mirror

FTT Out

Wavefront

Correction

Control

System

64 DSPs

Switch

From Correlation

Tracker

Figure 3.3. The AO real-time processors



correction, M2 for telescope focus correction and the WCCS for blending with information from the aO

system for QSA.

Deformable Mirror: The deformable mirror (M9) will be a continuous phase sheet DM with 1369

actuators. The mirror is 221 mm in diameter and the actuators are on a 4040 square grid. Each actuator

will have a stroke of at least 5 m. The actuators are located in the corners of the subapertures (Fried

geometry) yielding a total of 1280 useful subapertures. One of the leading DM manufacturers (Xinetics

Inc.) considers the ATST DM a straightforward extension of their 941-actuator off-the-shelf DM system.

Thermal control — The solar irradiance imposed on the front surface of the DM is approximately 10

kW/m2. To maintain the DM front surface at acceptably cool temperatures (<5˚C above ambient

temperature), the stock DM structure will be modified to allow convective cooling of the faceplate.

Coolant (air or helium) enters the rear of the DM and impinges normally on the faceplate backside before

exiting the DM radially. In addition, the faceplate will be sealed to prevent coolant leaks, and a broadband

high reflectivity coating is used on the faceplate. External to the DM, the coolant is recirculated through a

heat exchanger that transfers thermal energy to System Services facility coolant (ethylene glycol or

similar). We are currently working with Xinetics Inc. on developing detailed thermal models and

designing the modifications necessary to the DM to include the thermal control system.

Fast-Steering Mirror: The fast-steering mirror (M5) will have a range ± 1 mrad in both tip and tilt. The

mirror will be made from light-weighted material (SiC). The mirror is 220 mm in diameter. The controller

will have position feed back from the mirror and a closed-loop bandwidth of at least 100 Hz.

Thermal control: The peak solar irradiance imposed on the front surface of the tip-tilt mirror is

approximately 9 kW/m2. To maintain the front surface of M5 at acceptably cool temperatures

(approximately 3 ˚C above ambient temperature), an air jet array cools the rear of the SiC substrate. A

flexible boot encloses the air jet array and prevents the cooling air from passing into the optical beam. In

addition, M5 is coated with a high reflectivity coating to reduce the absorbed solar heat flux. The M5

cooling air is recirculated through a heat exchanger that transfers thermal energy to System Interconnect

facility coolant (ethylene glycol or similar).

3.2 ACTIVE OPTICS

The Active Optics (aO) system includes a dedicated wavefront sensor, wavefront-sensor camera, and the

aO controller. It provides feedback to the M1 axial support system that allows the optical figure of M1 to

be maintained to the desired level. This subsystem is also responsible for maintaining the quasi-static

alignment (QSA) of the telescope using the six degrees of freedom provided by the M2 hexapod and bore

site via slow tip tilt (- correction) of M3 and M6.

3.2.1 Active Optics Design Requirements

As noted in the M1 design requirements (Section 2.2.1 above) a surface figure of 80 nm rms must be

maintained over the operational limits of 0 to 80 zenith angle (changing gravity vector), thermal

conditions and wind loading. Error signals must also be generated by this system that will maintain

quasi-static alignment of the telescope to within 70 nm rms wavefront distortion.

3.2.2 Active Optics Design Description

The ATST will have two identical, multiple field active optics wavefront sensors - one at coudé lab and

one at Nasmyth. The aO wavefront sensors will measure wavefront errors averaged over many

atmospheric realizations. This will provide information about slowly varying (quasi static) aberrations

due to optical misalignment and/or M1 figure errors. FEA analysis and optical modeling show that a low

order wavefront sensor that reliably measures about 11 Zernikes is sufficient for this task. For QSA,



measuring the wavefront at multiple positions in the field of view is necessary to have enough

information to determine which optical elements need to be adjusted to keep the telescope in alignment.

The main objective is to distinguish between M1 figure errors and error due optical misalignment (e.g.

M2 de-center). The basic idea is to use wavefront sensor measurements from different field points

distributed over a large field of view. M1 figure errors cause wavefront errors that are constant across the

field while optical misalignment in general produces field dependent aberrations. Simulations indicate

that three positions are adequate, the center of the field and two adjacent corners. Although, it would be

straightforward to add additional field points if it turns out to be of advantage, e.g., provides better S/N.

The update rate of the aO will be about 30 seconds.

Wavefront Sensor: The optical design of the aO wavefront sensor is very similar to the AO system with

the exception that a large field of view will be imaged by each lenslet to enable wavefront sensing at

multiple field points. There are some similarities between the aO wavefront sensor approach and the solar

MCAO wavefront sensor approach. The field stop will be about 2-arcmin square pending the results of an

ongoing analysis of sensitivity as a function of separation of the field points. This means that the 2-

arcmin by 2-arcmin subaperture images will be sampled with 192×192 pixels and a camera that has at

least 1546×1344 pixels is required. Figure 3.4 shows the subaperture images as they will be imaged onto

the camera. The small grey squares are the 32×32 pixel subfield areas within each subaperture image

where the three wavefronts will be sensed.

Wavefront Sensor Camera: Wavefront measurements are required once every 30 seconds. About 300

individual wavefront measurements will be co-added in order to average out the seeing. This means that

aO wavefront sensor camera and processing unit will have to be run at about 10 frames per second. 2k×2k

Cameras that run at tens of Hz frame rate are commercially available. The processing power needed for

the aO task is also available off-the-shelf. A Zernike decomposition of the seeing averaged wavefront

measured at different field points will be output to the Wavefront Correction Control System. The WCCS

will pass the values on to the Telescope Control System for M1 figure, the M2 for focus and M1, M2, M3

and M6 for telescope alignment and bore-sight. If the light is at Nasmyth, the alignment will use M3N

instead of M3 and M6.

2’ =

192

pixels

20” =

32 pixels

Lenslet array with pupil outline Subaperture images with field points

Figure 3.4. Active Optics wavefront sensor.



3.3 WAVEFRONT CORRECTION CONTROL SYSTEM

To coordinate the various components of the wavefront correction system, a Wavefront Correction

Control System (WCCS) will be used. It will enable the Nasmyth or coudé systems depending on where

the light is to be used. It will control which sensing system will control which corrector.

The major task of the WCCS is blending information from the operating wavefront correction systems to

determine corrections for M1 figure and adjustments to the telescope optics for alignment. For instance,

when both the aO and HOAO systems are running at coudé, the aO system is measuring the wavefront at

the center and two corners of the field. However the HOAO is correcting the center of the field so the

HOAO provides the correction it is making to the center of the field to the WCCS which it then combines

with the information from the aO system to determine corrections for M1 figure and quasi-static

alignment.

When the light is at Nasmyth and the telescope is on disk, the WCCS will enable the Nasmyth correlation

tracker to sense image motion. The correction values will be sent to M2. The WCCS will also enable the

Nasmyth aO system to measure the average wavefront in three field positions. This information will be

sent to the WCCS which will provide quasi-static alignment correction values for M2 and M3N and M1

figure correction values.

When the light is at coudé and the telescope is on disk, the WCCS will enable the coudé aO system. If

the telescope is in diffraction limited observing mode the AO system will also be enabled for HOAO and

either the correlation tracker or the HOAO system for image stabilization. The WCCS will blend

information from both the aO and AO systems for quasi-static alignment and M1 Figure.

4. INSTRUMENTATION

The description of ATST instrumentation is divided into two parts: the instrument lab facility and the

focal-plane instrumentation.

4.1 INSTRUMENT LAB FACILITY

The Instrument Lab Facility is a set of common components that directly support science instruments and

observers. The dominant requirements that affect the design of the instrument lab facility are derived

from the telescope requirements associated with resolution, polarization sensitivity and accuracy,

flexibility, adaptability, and availability. These were discussed in Section 1.1.2.

4.1.1 Polarimetry Analysis and Calibration

The Polarimetry Analysis and Calibration system is used both to modulate the beam for determining the

polarization state of solar features, and to calibrate out polarization introduced by the telescope. Because

polarimetry is nearly ubiquitous in observational solar physics, ATST provides polarimetry analysis and

calibration at the facility level, rather than making it part of the requirement for each instrument. The

science requirement for sensitivity (10-5

relative to intensity) and accuracy (510-4

relative to intensity)

place strong constraints on this system, and ultimately dictate the methods and strategies that will be used

to do polarimetry.

Polarimetry at the Nasmyth Station: Ideally the polarization introduced by the telescope and focal-

plane instruments should be kept under 1%. This goal is only approached at the Nasmyth focal plane.

This will be the observing station for the coronal module of the Near-IR Spectro-polarimeter (see

description below in Section 4.2.3), and other instruments for which the spatial and spectral resolution

requirements are relatively low.



Figure 4.1. Block diagram of a DID.

Time-multiplexed polarization modulation and analysis will be used at all observing stations because it is

versatile, and the issues are well understood. The initial ATST facility-level modulators will include

piezo-elastic modulators (PEMs), ferroelectric liquid crystal (FeLC) modulators, and rotating retarders.

These modulators, as well as calibration polarizers, will be mounted in the Gregorian Optical Station

(GOS), which is located near (and just above) the Gregorian focal plane. It consists of two large rotating

wheels with up to five positions each. The One of the wheels is in the Gregorian focal plane; the other

precedes it.

Telescope polarization at the Nasmyth station will be calculated to the required accuracy using a detailed

Mueller-matrix model of the telescope that includes the properties of aluminum coatings and oxide

overcoatings. This model will be tested and improved by actual measurements. Polarization introduced

by the focal-plane instrument will be measured using calibration polarizers in the turret assembly.

Seeing and tracking errors introduce small changes to the sun’s image that, when sampled too slowly, can

be misinterpreted as polarization of the solar source. Several strategies will be implemented to reduce

this effect. The first is dual beam, or spatial modulation. This will be implemented by using a polarizing

beam splitter as a polarization analyzer, spatially separating the s and p polarization states. Both beams

can be analyzed independently but simultaneously to derive the four polarization states in a way that is

less sensitive to seeing effects. It has the added advantage of utilizing all of the light introduced onto the

polarization analyzer.

Spatial modulation alone still has shortcomings. The two beams will generally not pass through the

instrument along exactly the same optical path. Differential aberrations may then become important.

Furthermore, spatial modulation requires that different detectors or detector areas sample the two beams,

which makes the measurements susceptible to differential-gain effects. Whenever possible the ATST

polarization strategy will also include rapid polarization modulation (>1 kHz) and charge-caching camera

systems. These specialized cameras, dubbed DIDs for Demodulating Imaging Detectors, will be

optimized for highly sensitive and precise differential

imaging. Chopping between the four linearly

independent polarization states can be performed at

speeds in the kHz domain to provide virtually

simultaneous images without the need to read out the

array at kHz frame rates (Figure 4.1). All independent

image planes are observed with the same physical

pixel on the detector, which renders normalized

differences between image planes insensitive to the

gain of individual pixels. DIDs will have a 100%

geometrical fill factor and quantum efficiencies

approaching unity. The technology can be applied to

silicon to cover the 200 to 1100 nm wavelength range,

and to infrared-sensitive materials such as HgCdTe for

the 1000 to over 10,000 nm wavelength range.

Rockwell has expressed an interest in providing these hybrid detectors to ATST. Another possible

alternative is AIM, an independent entity of AEG Infrarot-Module, GmbH. They were part of a proposal

to the EU to develop a DID for landmine detection. These detectors are considered to be a modest and

achievable extension of the detector technology built into the most recent upgrade to the successful

Zurich imaging polarimeters (ZIMPOL II).

Polarimetry at the Coudé Station: To meet the science requirements for polarimetry of solar features

on the sun’s disk ATST must provide large instruments, slow beams, diffraction-limited imaging, and

adaptive optics to correct the wavefront. Instruments that support these observations will reside at the

coudé stations. The uncompensated oblique angles of the transfer optics below the Gregorian focus will



introduce considerably more telescope polarization than is present at the Nasmyth station. This dictates

that modulation and analysis must be performed near the Gregorian image prior to these strongly

polarizing telescope mirrors. This allows the same GOS polarimetry components to be used for both

Nasmyth and coudé observations.

It will not be practical to pass two orthogonally polarized beams through the adaptive optics system.

Thus, the coudé instruments can perform spatial modulation only after transitioning through the

telescope’s feed optics. The DID strategy described above will also be available at coudé. Based on

experience with existing strongly polarizing telescopes, we expect that ATST will meet polarization

science requirements at coudé as well.

The individual components provided as part of the ATST polarimetry analysis and calibration system are

summarized in Table 4.1.

Table 4.1. Polarimetry Analysis and Calibration System Components.

Component Location

Achromatic Linear Polarizer Calibration Wheel, Turret Assembly

Achromatic Retarder Calibration Wheel, Turret Assembly

UV Linear Polarizer Calibration Wheel, Turret Assembly

UV Retarder Calibration Wheel, Turret Assembly

UV/Vis PEM Modulator Wheel, Turret Assembly

Rotating Achromatic Retarder Modulator Wheel, Turret Assembly

UV Rotating Retarder Modulator Wheel, Turret Assembly

Visible FeLC Modulator Wheel, Turret Assembly

IR FeLC Modulator Wheel, Turret Assembly

Vis/IR LCVR Modulator Wheel, Turret Assembly

Achromatic analyzer Gregorian Focus Wheel, Turret Assembly

8 Polarizing Beam Splitters Nasmyth or coudé stations

8 Linear Polarizers Nasmyth or coudé stations

5 FeLCs Nasmyth or coudé stations

4.1.2 Nasmyth Station

The Nasmyth Station includes transfer and re-imaging optics, mounting fixtures, and connections to

utilities provided via System Interconnects.

This observing station is mounted to the

Nasmyth Rotator Structure, which is part of

the Telescope Mount Assembly (Figure

4.2), and provides a plate scale of 3.95

arcsec/mm at f/13. Hence, the required 3-

arcmin field of view has a diameter of 46

mm, and the goal five-arcmin field has a

diameter of 76 mm. This station is suitable

for instrumentation that is relatively

compact, has relaxed spatial and spectral

resolution requirements, and can tolerate a

changing gravity vector. The Nasmyth

station has the advantage of providing an

image with the minimum number of

reflections. Thus, the Nasmyth station

provides the best throughput and the lowest

levels of scattered light and telescope

polarization. It has the disadvantage

Figure 4.2. The Nasmyth Observing Station.



(compared with the coudé station) of no high-order wavefront correction, images that are not diffraction

limited over the full field of view, a changing gravity vector while tracking the sun, and more restrictive

instrument size and weight limits.

4.1.3 Coudé Station

The Coudé Station includes the optical tables, imaging optics, standardized mounting fixtures, camera

systems, and a connection to utilities provided via Interconnects and Services. The majority of facility

instrumentation and visitor instruments will be operated at this station. It has the advantage of full

wavefront correction provided by the upstream high-order AO system, diffraction-limited images, and a

large horizontal platform that can accommodate multiple, large instruments in a constant-gravity location.

Six facility instruments will eventually be permanently installed at the coudé station. These are the

Visible Broadband Imager, Visible Spectro-polarimeter, the coudé module of the Near IR Spectro-

polarimeter, the Visible Tunable Filter, the Near IR Tunable Filter and the Thermal-IR Polarimeter and

Spectrometer. These are discussed in Section 4.2.

To accommodate the science requirement for a flexible, adaptable facility that can be quickly configured

for a variety of diverse experiments, and to accomplish this in a cost-effective manner, ATST has adopted

a strategy that calls for a high level of standardization in our approach to instrumentation wherever

possible. This will be particularly true for observations performed at the coudé station. For example, all

facility scientific instruments being fed by a particular beam will conform to a prescribed optical height

above the horizontal optical tables. This will allow common mounts with focus and decenter motions for

cameras, filters, and other incidental optics that will be provided as part of the laboratory instrument

facility. Visitor instruments will be strongly encouraged to adopt the same standards, thus giving them

access to facility components whenever possible.

Solar observations generally utilize multiple cameras in the course of a single observation. This need is

derived from the wavelength diversity requirements placed on the instrumentation, and the need to make

efficient use of the available light. The necessary cameras will be available “off the shelf” as part of the

ATST instrument lab facility. This will allow scientist to take maximum advantage of observational

targets of opportunity. This strategy of using uniform camera and controller systems will also minimize

the cost of developing multiple camera systems and the software that runs them.

The camera systems that will be provided as part of the initial instrument lab facility are listed in Table

4.2. These same systems are available for use at the Nasmyth focus as well.

Table 4.2. Initial Camera Systems.

Camera Type Format Readout Count For use with

Fast CCD 1k1k 100 Hz 5 Vis Spectro-polarimeter, Vis Tunable Filter

Large CCD 4k4k 5 Hz 7 Vis Spectro-polarimeter, Vis Tunable Filter, Broadband Imager, Slit-jaw viewer

Large Hybrid IR 2k2k 5 Hz 5 Vis Spectro-polarimeter, NIR Spectro-polarimeter, IR Tunable Filter

Visible Light DID 2k2k 25 Hz 5 Vis Spectro-polarimeter, Vis Tunable Filter

Near IR DID 2k2k 25 Hz 4 Vis Spectro-polarimeter, NIR Spectro-polarimeter, IR Tunable Filter

4.1.4 Instrument Lab Thermal Control

The Instrument Lab Facility Thermal Control system maintains air and component temperatures within

the bounds necessary to meet self-induced seeing error budget. The coudé lab will be held at a constant,

uniform temperature while the telescope assembly above it will track the ambient outside air temperature.

Over much of the year there will be a large volume of warm air beneath the much cooler air within the



telescope enclosure, producing a thermal instability. Many of our science use cases require simultaneous

observations spanning wavelength ranges from the visible to beyond the glass cutoff wavelength in the

near infrared. This precludes using a window to separate the two environments. Instead, the coudé

stations will be isolated via a laminar “air curtain.”

4.1.5 Instrument Control System

The strategy outlined for modular instrumentation components places a large burden on the instrument

control software. This is solved using the concept of a virtual instrument which is “assembled” both in a

hardware and software sense as part of the observing setup procedure. The details of the virtual

instrument and other important aspects of the instrument control system are discussed in the high-level

controls chapter of this document (see Section 5.2.5).

4.2 SCIENCE INSTRUMENTS

The science instruments envisioned for ATST and their relative priorities (from the SRD) are listed in

Table 4.3. The first four instruments will be built as part of the construction phase. The priority five and

six instruments – and perhaps others – will be built early in the operations phase.

Table 4.3. Science Instruments.

Priority Instrument Fore-Optics Dispersing System Detector System

1 Visible Broadband Imager Phase Diversity Interference Filters Visible

2 Visible Spectro-polarimeter Visible Polarization Analyzer

Medium Dispersion Spectrograph

Visible or special

3 Near-IR Spectro-polarimeter (Disk and Corona)

Near-IR Polarization Analyzer

Medium Dispersion Spectrograph

Near-IR

4 Visible Tunable Filter Polarization Analyzer Visible Tunable Filter Visible

5 Near-IR Tunable Filter Polarization Analyzer Near-IR Tunable Filter

Near-IR

6 Thermal-IR Polarimeter & Spectrometer

Polarization Analyzer Medium resolution, cold grating

Thermal-IR

7 Visible/Near-IR High-Dispersion Spectrograph

Visible/near-IR high-dispersion spectrograph

Visible and near-IR

Swiss Contribution

UV-Polarimeter Polarization modulation system

Visible spectrograph or narrow-band filter

ZIMPOL Detector System

4.2.1 Visible Broadband Imager

Design Requirements: The primary science requirement of the Visible Broadband Filter instrument

(VBI) is to obtain the highest possible spatial and temporal resolution image sequences from the ATST.

The study of small-scale magnetoconvective processes both inside and outside of sunspots requires spatial

resolutions on the order of 0.01 arcsec and temporal cadence values of 5 seconds or less. The spatial

resolution requirement is near the projected ATST diffraction limit, thus the VBI cannot significantly

degrade the image quality delivered by the telescope. The VBI must also be capable of making broadband

filter images in a range of scientifically important visible spectral bands on a rapid cadence. This drives

the design to use simple, high optical-fidelity, thin-film interference filters for spectral selection. Table

4.4 lists the baseline science requirements of the VBI derived primarily from the SRD and secondarily

from optomechanical design considerations.

Instrument Description: The baseline VBI consists of an optical relay unit (collimating and camera lens

systems), one or more rotating filter wheel mechanisms containing the broadband interference filters, and

a focal plane camera mounting system. Each of these systems will be designed to be removable,

replaceable, and/or reconfigurable within a horizontally mounted enclosure.

Table 4.4. VBI Instrument Requirements.

A. Optical Requirements



Value Goal Priority Source Notes

Spectral Range 380-800 nm 330-1100 nm 1 SRD CN bandhead 388 nm

Field-of-View 3 arcmin 5 arcmin 1 SRD Unvignetted circular diameter

Spatial Resolution 0.02 arcsec 0.01 arcsec 1 SRD

Spectral Resolution 0.01 nm 0.01 nm 1 SRD Short exposure imaging

Beam Speed F/20-F/45 Variable 2 SRD Multiple plate scales

Scattered Light 10-2

I0 10-3

I0 2 SRD Sunspot umbral imaging

Instrumental Polarization

10-2

I0 10-2

I0 3 SRD No polarimetric capabilities required

Mounting Horizontal Multi-config 2 Coudé or Nasmyth focus

B. Interference Filter Requirements


Optical Quality /4 Φ25 mm /4 Φ100mm 1 A.3

Bandpass 0.01-0.1 nm 1 A.1 Varies with spectral region

Transmission 40% 60-70% 1 SRD Temporal resolution

Out-band Blocking 10-4

UV/IR 10-5

UV/IR 1 No active thermal control

Operating Temp. 17±3 ˚C 17±10 ˚C 1 Telescope ambient

Thermal Stability <0.01nm 1 No active thermal control

Mounting Parallelism

3—5 arcmin < 2 arcmin 2 A.3 Fringe avoidance

C. Mechanism Requirements


Filter wheel Speed 2 sec 1 sec 1 SRD Including settle time

Focus Lens Speed 2 sec 1 sec 1 SRD Temporal cadence < 5 sec

Camera Exposure <100 msec 10 msec 1 A.3 Freeze atmospheric seeing

The optical relay unit consists of a field lens and stop at the nominal telescope focal plane followed by a

collimating lens unit (CLU) which collimates the light prior to the interference filter stages. Following the

filters, the camera lens unit focuses the image plane onto the camera system. For a given camera pixel

size, no single focal ratio will provide a Nyquist sampled detector plane for all wavelengths in the VBI

spectral range. Therefore the VBI optical relay system will provide a range of focal ratios from F/20 to

F/45 in order to optimize a given detector’s sampling. The current concept uses a varifocal zoom lens to

provide a continuously variable focal ratio that will accommodate any camera system and pixel size.

Table 4.5. Nominal VBI Interference Filter Specifications

Filter

Central

Wavelength

nm

Bandpass

nm

CN molecular band 388.3 0.1

Ca II H & K lines 393.3 & 396.8 0.01

CH molecular G-band 430.5 0.1

Blue continuum 450.4 0.05

Green continuum 555.0 0.05

H-alpha 656.3 0.01

Red continuum 668.4 0.05

TiO sunspot bands 705.7 0.1

Ca II chromospheric magnetic 854.2 0.01

The VBI filter wheel unit will consist of up to three rotating mechanisms containing four broadband

interference filters each. The filter wheel mechanisms will be brushless DC motor driven with optical

encoders for positional readout. Each filter will be a three- cavity thin-film interference filter with a



nominal diameter of 10 cm. The large diameter is required to ensure a 3-arcmin field of view at the

nominal ATST focal length. A list of VBI spectral ranges, specified in the SRD, is shown in Table 4.5.

The availability of large-format (8k8k), high quantum efficiency (QE ~90%), fast-readout (~1 sec)

detectors is currently extremely limited. Thus it is not possible to design the VBI for a particular camera

that will meet the ATST spatial resolution, temporal cadence, and field-of-view science requirements. We

circumvent this by incorporating a flexible detector mounting using a three-axis linear-motion stage with

micron accuracy. Any given detector can thus be placed anywhere in the 8 to 10 cm diameter focal plane

to cover the full field. In addition, the VBI camera system incorporates two separate detector stages in a

split beam configuration. This allows two cameras to be independently focused and defocused for “Phase

Diversity” (PD) imaging, thus enabling instrumental aberration measurement as well as image restoration

to the diffraction limit.

For the FLI subsystem, we envision the VBI camera stages using one or two standard, readily available

4k4k CCD cameras, and a single filter wheel (or in the simplest cases, a single filter mounted directly to

the camera system input beam). As the commissioning phase of the telescope progresses, the FLI system

can be built into the larger VBI system.

4.2.2 Visible Spectro-polarimeter

Design Requirements: The Visible Spectro-Polarimeter (ViSP) is the instrument responsible for the

spectral analysis of the visible solar light and its polarization state, recording the wavelength dependence

of the full Stokes vector (I, Q, U, V) at each spatial point in the field of view.

In order to meet the science requirements, the ViSP must be able to:

Observe the small-scale magnetic elements (flux-tubes) in the solar photosphere with an angular

resolution of at least 0.05 arcsec, or about 40 km.

Cover a large field of view of at least 3 arcmin.

Routinely attain a polarimetric precision of 10-4

times the continuum intensity. In addition to this

requirement, it would be highly desirable to reach the 10-5

level at least in particular

configurations.

Minimize seeing-induced cross talk. It should be small compared to the polarimetric precision

quoted above.

Fully resolve spectral features, including those arising from hyperfine structure or magneto-

optical effects. The spectral resolution should be at least 3.5 pm at a wavelength of 600 nm.

Observe at least three different spectral ranges in the visible simultaneously (wavelength

diversity), in the range from 380 nm to 900 nm.

The ViSP should be able to operate simultaneously with the infrared spectro-polarimeter

(NIRSP). A compensator for atmospheric differential refraction is needed in order to ensure that

both instruments are observing the same field.

Instrument Description: The basic optical layout of the ViSP appears in Figure 4.3. The instrument

concept is based on modern spectro-polarimeters, with a slit that scans the field of view and a

spectrograph that images the slit spectrum on a 2D detector at each scanning step. However, it

incorporates significant technological advances coupled with an innovative design that are necessary to

fulfill its stringent requirements, far beyond those of any other present instrument of its kind.



The light beam first passes through a retarder with a time-dependent retardance (modulator) and later

through a polarization analyzer, which can be a linear polarizer or a polarizing beam-splitter depending on

the operation mode (see below). Reflections of the light beam along its path in the telescope introduce

significant instrumental polarization. In order to attain the polarimetric precision set forth by the science

requirements, the modulator and the calibration optics must be placed near the Gregorian focus. At this

location, the light beam has only undergone two reflections and the instrumental polarization can be

controlled at the level required.

Unfortunately, it is not possible to fulfill the wavelength diversity requirement and the 10-5

precision goal

simultaneously. In order to avoid the seeing-induced cross-talk, it is necessary to either run the modulator

at nearly kHz frequencies or to split the beam just before the detector into two beams with opposite

polarizations. In the first case (single-beam scheme), both the modulation and the analysis are done very

rapidly at the Gregorian focus. Subsequent reflections in the optical path to the detector will not affect the

measurements, since the beam has been analyzed already. This scheme, which permits highly accurate

polarization measurements, involves the use of FeLC. These devices are not achromatic and need to be

tuned to a specific wavelength. Therefore, it is not possible to meet the wavelength diversity requirement

with this setup. A dual-beam scheme, on the other hand, does not need such a fast modulation. The

relatively slow (tens of Hz) time modulation of two simultaneous images with opposite polarization is

used to correct the undesirable effects of seeing, at least to first order. The modulator and analyzer (a

polarizing beam-splitter) can be made achromatic over a broad range of wavelengths, which permits the

simultaneous observation of several spectral domains. The dual-beam setup has the analyzer at the end of

the optical path, right before the detector in the coudé focus. Multiple inclined reflections exist between

the modulator and the analyzer, introducing spurious polarization. Imperfections in the calibration to

correct such instrumental polarization and small residual cross talk from the seeing may compromise the

measurements at a level of 10-4

.

The ViSP will have three different operation modes: The single and dual beam modes described above

(for polarimetry at 10-5

and wavelength diversity, respectively), and a “hybrid” mode that combines

advantages from both schemes.

1. High Precision Polarimeter (HPP): The modulator is a FeLC and the analyzer is a linear polarizer,

both at the Gregorian focus. This mode allows for 10-5

polarimetry. The ViSP needs to be tuned to a

specific wavelength and cannot operate in combination with the NIRSP. HPP requires a charge-

caching device as a detector.

2. Fast Achromatic Polarimeter (FAP): This is the hybrid mode. It uses a fast (~1 kHz) rotating

wave plate as an achromatic modulator at the Gregorian focus. The analyzer is made of a FeLC

Figure 4.3. ViSP basic optical layout.



combined with a linear polarizer, and it is located at the coudé focus immediately before the detector

(one analyzer is required for each detector). Seeing-induced cross-talk is prevented by the fast

modulation, but multiple reflections between the modulator and analyzer complicate the calibration.

The FAP meets the wavelength diversity and 10-4

polarimetric precision requirements. FAP requires

charge-caching detectors, and it works with the NIRSP.

3. Slow Achromatic Polarimeter (SAP): The modulator is an achromatic rotating wave plate at a

frequency of ~10 Hz, located at the Gregorian focus. The analyzer is a polarizing beam-splitter before

the detector. Residual seeing-induced cross talk and calibration errors limit the polarimetric precision

to a few times 10-4

. This mode is capable of wavelength diversity. Conventional CCDs can be used

for detectors, and it works with the NIRSP.

The ViSP design is contained in one plane, allowing easy access to the components for upgrades and

adjustments. The slit width is adjustable to allow for various trade-offs between resolution and photon

flux. A turntable contains several gratings that can be selected to meet the needs of the observing

program.

4.2.3 Near-IR Spectro-polarimeter

Design Requirements: The Near-IR Spectro-Polarimeter (NIRSP) is the instrument responsible for the

spectral analysis of the near infrared solar light and its polarization state. As with the ViSP, this

instrument will record the wavelength dependence of the full Stokes vector (I, Q, U, V) at each spatial

point in the field of view. Infrared spectroscopy requirements are diverse because of the broad flux and

spectral resolution conditions inherent to photospheric and coronal physical conditions. In order to meet

these requirements we describe a modular NIRSP system that satisfies both the low flux, and lower

angular and spectral resolution requirements of the corona, and the higher resolution (spectral and spatial)

needs for observing the photosphere.

Our philosophy in designing the NIRSP has been to achieve ATST spectroscopic infrared science

observing requirements with multi-use optical components (and designs) wherever possible. For example,

coronal spectroscopy will be obtained almost exclusively from the Nasmyth focus, ahead of the many

reflections that bring light into the coudé instrument room. In many cases the spectral resolution for

coronal observing is dictated by the few-million degree temperature coronal line profiles. In practice the

necessary resolution is somewhat higher than what coronal line-widths dictate because we often need

sufficient spectral resolution to separate K and F coronal and scattered-light photospheric spectral features

(for example for calibration). This is achieved with our resolution 4104 coronal spectrometer. Spatial

resolution of even an arc second will provide revolutionary new information about the faint corona's

magnetic field. Disk observations must be obtained after the image is corrected by ATST adaptive optics

at the coudé focus. Thus we require resolution of at least 3105 with diffraction limited spatial resolution

here. As we illustrate below the full requirements can be achieved with common optical components, but

with distinct F/6.6 Nasmyth and F/40 coudé systems.

Instrument Description: Infrared detector technology is a significant driver for the NIRSP. While the

final instrument design will be decided in a year or more, we feel that the most prudent NIRSP concept

now should be based on the stable Rockwell detector HgCdTe technology. This proposal assumes

Hawaii-2-style 2k2k format detectors. Coronal science requirements also dictate the need to observe into

the thermal IR in order, for example, to reach the important 3.9 µm SiIX emission line for magnetic

diagnostics. With common camera elements we describe below a 1-5 µm range coronal and photospheric

NIRSP system. This also allows photospheric thermal IR spectroscopy without significant budget or

technical performance impact.

The NIRSP-C and NIRSP-G (coudé and Nasmyth) optical configurations are scaled versions of a

common reflecting Littrow design. These optics are described below in Table 4.6 and Figure 4.4. NIRSP-



G is illustrated here with the ATST-supplied focal relay that yields an F/6.6 input beam. The largest optic

is the collimator and camera 35 cm parabolic mirror. The grating is a standard R2 87 line/mm echelle, the

same grating used for NIRSP-C. Table 4.6 lists the optical specifications for both configurations. Pixel

binning is used to properly sample the larger NIRSP-G slit.

With slit choices described in Table 4.6 the diffractive NIRSP grating illumination is comparable to the

geometric illumination. NIRSP-G pixels will be binned 6×6 to yield 0.85-arcsec spatial and 242 mÅ

spectral resolution for coronal observing. This Littrow configuration minimizes off-axis angles from the

parabolic mirror so that the geometrical performance is close to diffraction limited.

Table 4.6. NIRSP Coudé and Nasmyth Optical Specifications

NIRSP-G NIRSP-C

Grating R2: 87 line/mm R2: 87 line/mm

Grating size 400x200mm 400x100mm

Collimator focal length 1.2m 3m

Focal Ratio 6.6 40

Plate scale 0.127 mm/arcsec 0.776 mm/arcsec

Slit Width 108µm 36µm

Pixel scale (at 1 μm) 242 mÅ 32 mÅ

Pixel scale (arcsec) 0.142" 0.046"

Design/Diffraction limited FOV 290" 290"

Hawaii-2 FOV 290" 47"

Cold blocking filter BW 1% 1%

System QE 5% 5%

Grating emissivity 50% 50%

While the optical configurations are similar, the mechanical structures are quite different. This results

from the vastly different background conditions for coronal and disk observing, i.e. NIRSP-C is not

intended for coronal observing conditions. The dashed lines in Figure 4.5 show the expected solar signal

and background versus wavelength assuming an optical table-mounted warm spectrograph, but with a 1%

bandwidth cold order-sorting and blocking filter in the IR camera Dewar module. The figure shows the

Figure 4.4. Optical layout for the NIRSP-G configuration. NIRSP-C is identical but scaled in length by a factor of approximately 2.5.



mean photospheric flux versus the purely

thermal background contribution. The

background is much lower than the solar

signal blueward of about 5.5 µm at which

point the background dominates the

photospheric flux. This shows that a

warm NIRSP-C will satisfy most

requirements for high resolution

photospheric observations.

Coronal observations with a warm

NIRSP-G configuration would be

dominated by thermal background

photons at all wavelengths longward of

about 1.9 µm. By cooling the grating and

spectrograph optics to approximately 85K

the NIRSP-G will allow coronal

spectroscopy up to a wavelength of at

least 4.5 µm. This is illustrated in Figure

4.5, which shows the coronal signal and

background for a warm and cooled

NIRSP-G configuration as described in

Table 4.6.

The lower spatial and spectral resolution

required for coronal observations also

permits a compact cooled spectrograph design. Our working concept is illustrated in Figure 4.6. The IfA

group recently designed, built, and commissioned an even larger cooled IR echelle spectrograph for the

AEOS telescope which is the basis for some of the schedule and cost estimates for the ATST version.

This NIRSP-G design uses (not shown) six stepper-based mechanisms utilizing vacuum cryogenic feed-

throughs. Additional manual external mechanical adjustments for M1 collimation and grating are also

provided. The design includes provisions for mounting a polarizing beam splitter behind the cold slit.

The slit viewer and integral slit-viewer filter wheel and the final science detector can be removed from the

NIRSP-G without warming the Dewar or breaking vacuum. The entire cylindrical volume of the NIRSP-

G is supported at both ends by ATST-supplied circular bearings, which provide for instrument rotation as

needed for the alt-az telescope configuration.

The NIRSP-C configuration will be constructed on a conventional optical table in the ATST coudé space.

The optical layout is identical to Figure 4.6 (except for scale). Infrared slit-viewer and final science

cameras for NIRSP-C are identical to the NIRSP-G cameras. Of course all mechanism controls and their

hardware and software interfaces will be common for all NIRSP components. NIRSP-C relies on the

ATST coudé room rotation and does not use separate instrument rotation control.

Figure 4.5. Signal and background flux calculations for NIRSP-C

(solid lines) and NIRSP-G (dashed lines). Assuming the optical configuration of Table 1 the expected signal and background per NIRSP resolution element is plotted versus wavelength. A warm spectrograph is adequate for disk observations at wavelengths shortward of about 5.5 µm (dashed lines) while a cold NIRSP-G configuration allows coronal observations shortward of about 4.5

µm. A warm NIRSP-G spectrograph does not satisfy coronal observing requirements.



4.2.4 Visible Tunable Filter

Design Requirements: The Visible Tunable Filter will be used to obtain narrow spectral bandwidth

observations over an extended area of the sun. This capability will provide us with rapid 3D-imaging

spectrometry, Stokes spectro-polarimetry, and accurate surface photometry. It will also deliver

spectroheliograms to measure Doppler velocity, transverse flows, provide a feature tracking capability,

and generally permit the study of evolutionary changes of solar activity. Investigators will use it in

conjunction with the AO system for high spatio-spectral imaging. The filter will operate at four spectral

bandwidths. These four modes and their associated requirements are given in Table 4.7.

Table 4.7. Visible Tunable Filter Modes and Requirements.

Filter Mode Observations Passband FWHM (pm)

Field of View (arcmin)

Typical Spectral Lines (nm)

Desired Peak Transmission

A. Narrow Passband

(Dual/Triple Etalon Configuration)

Spectro-polarimetry using I,Q,U,V Stokes fractional parameters 3D Spectrometry & 3D Tomography & Flow Geometry

2.0 1 FeI: 524.70, 525.02, 525.06, 630.15, 630.28 629.87, 868.8 CaII: 863.5 FeI: 569.1,557.6, 684.27 CI:538.03

>50%

B. Medium Passband (Single or Dual Etalon Configuration)

Filter Vector Magnetograms Filtergrams

12.0 3 FeI: 525.02, 525.06, 630.15, 630.28 CaII: 863.5 CaII: 863.5 MgI : 517.2

>60%

C. Intermediate Passband (Single Etalon Configuration)

Dopplergrams High-Speed Imagery & Flares

20-30 3 HI: 656.3 FeI: 543.45, 557.6, 630.15 HI: 656.3

>70%

D. Broad Passband (Interference Blocking Filters only)

Advective Flows-Transverse Flows Movies & Active Region Evolution

100-1000 3 CN: 430.5 CaI: 399.3 CN: 430.5 Continuum 450.8

>80%

Figure 4.6. Mechanical concept illustrating NIRSP-G. This "transparent aluminum" schematic diagram illustrates

the critical components of NIRSP-G. The overall Dewar length is 1.2 m and its mass is approximately 900 kg. Light enters from a slit wheel assembly on the right. Removable HgCdTe cameras are indicated on the top

surface. Radiation shielding, cold straps and most mechanisms are not indicated.



It will meet the minimum system requirements shown in Table 4.8.

Table 4.8. Minimum System Requirements

Minimum Aperture 200 mm

Spectral Range 450 to 750 nm

Spectral Resolution 200,000

Minimum Peak Transmission 50% with blocking filters

Maximum Ghost Transmission 10-4

Maximum Stray Light 10-3

Drift Stability < 0.1 pm per hr

Design Description: Our design for the Visible Tunable Filter is a triple etalon system based, in part, on

the successful German TESOS system, shown in Figure 4.7. We have selected a multiple Fabry-Perot

(FP) spectral filter for the following reasons:

1. It can provide the required spectral resolution

for high-resolution spectral imaging, Stokes

profile analysis and filter magnetograms

(spectral resolution ~250,000 at 500 nm);

2. It has the high etendue (light throughput) to

obtain a sufficient number of spectral samples

within appropriate solar oscillation periods, and

the required magnetic sensitivity on the

timescale that solar features change;

3. It is mechanically and optically simpler in

design than a Lyot filter;

4. It provides the rapid tuning between

wavelengths that is required for finding the line

center and adjusting the wavelength setting for

Doppler-induced shifts;

5. It is a single system capable of simple

spectroscopy, Stokes line profiles, and filter

magnetograms. The etalons can accommodate

large aperture (~200 mm) filter systems

allowing extended field-of-views without

spectral degradation.

A triple etalon system offers several additional advantages, including superior spectral purity and out-of-

band rejection, excellent throughput, and wider bandpass blocking filters that will remain stable over

timescales of years. It is also a conservative choice in terms of the finesse requirements on the individual

etalons.

Table 4.9 shows the relevant parameters of the triple etalon system. FSR = Free spectral range, FWHM =

full width at half maximum, F = finesse, R = reflectance, D = gap distance, and M = order. This

configuration has the same gap ratios as the TESOS system, however the FWHM of our system is 2 pm

compared to TESOS 3 pm FWHM. TESOS has finesses of 30-40, while ours are just above 50.

Figure 4.7. The TESOS instrument at the German Vacuum

Tower Telescope (Tenerife, Spain). The design of the ATST Visible Tunable Filter will use the experience gained from the

development of this successful instrument.



Table 4.9. Triple etalon system.

Etalon System FSR (nm) FWHM (pm) F R D (μm) M

1 0.106 2.02 52.6 .94 1300 4952

2 0.172 3.27 52.6 .94 802 3055

3 0.242 4.59 52.6 .94 571 2175

Papers by Gary, Balasubramaniam, and Sigwarth (2003), and Gary and Balasubramaniam (2003)

summarize the triple-etalon Fabry-Perot filter of choice for the ATST visible narrow-band filter and

consider the overall instrument requirements. The ATST filter optical design employing a triple Fabry-

Perot etalon system requires that each etalon have a flatness of ~/200 before coating, with the optical

finesse 10-50.

The heritage for the use of etalons and multiple etalon systems in solar physics comes from observatories

in both the United States and Europe, including the NSO/Sacramento Peak, NSO/Kitt Peak, the German

Vacuum Tower Telescope, Big Bear Solar Observatory, and the High Altitude Observatory. The design

of the ATST multiple etalon system relies on the existing experience and expertise from this and other

experience.

4.2.5 Additional Operations Phase Instrumentation

Infrared Tunable Filter: The top-level design requirements for the Infrared Tunable Filter call for a

wide passband range (at least including HeI10830Å, FeI15648.5Å, FeI15652.0 Å, etc), large field of

view, and narrow passband for measurements of solar magnetic field at deeper layers in solar atmosphere.

This filter system can be operated as a spectropolarimeter, a filter-imaging magnetograph, or a high-

resolution imager. The Design Requirements for the IR Tunable Filter are shown in Table 4.10.

The preliminary design combines an interference pre-filter, a tunable Lyot filter and a single FabryPerot

etalon.

Thermal Infrared Polarimeter and Spectrometer: The Thermal Infrared Polarimeter and Spectrometer

(TIPS) will perform vector polarimetry and infrared spectroscopy of the sun’s atmosphere. It will cover

the entire thermal-infrared from 5m out to the long-wave telluric cut-off at 28 m. The TIPS will study

the magnetic structure, dynamics, chemistry, and physical state of sunspots, plages, flares, prominences,

and quiet regions. In the polarimetry mode, vector magnetic fields will be measured using the Mg I

emission lines at 12.3 m. With these lines, the most magnetically sensitive in the solar spectrum, TIPS

will measure the strength and three-dimensional configuration of fields in the upper photosphere. In the

spectroscopy mode TIPS will record spectra at high spectral and spatial resolution.

The primary candidate instrument for the TIPS is a cryogenic grating spectrometer. The instrument will

be placed at one of the coudé stations. A large echelle grating will provide the required high spectral

Table 4.10. IR Tunable Filter Design Requirements.

Properties Requirements Comments

Spectral Coverage 1.0~1.7µm (scanning)

Resolving Power > 150,000

FOV 1~3 arcmin

Bandpass 0.1Å@15650 Å

Spatial Resolution < 0.1 arcsec AO – 0.09 arcsec at 1.6µm

Operation Mode Narrow/Medium/Broad band Consideration of flexibility to serve different observation purposes. This shall be realized by taking in/out individual filter from the system.

Aperture > 36 mm (Lyot), > 150 mm (FPI) Compared to the currently designed similar filter system at BBSO, ATST shall have better performance.

Throughput > 40% (Lyot), > 80% (FPI) Instrumental Principles regarding filters. Polarizer will be a major drag in this design.

Scattered Light < 10-3

Stability ~0.05 Å/hour (FPI & Lyot) Due to the properties of components of Lyot and FPI.



resolution. One spatial dimension will be imaged along an input slit and the second spatial dimension will

be mapped by stepping the slit across the field. The spectrometer will use a large format (10241024)

As:Si detector array with operation optimized for rapid cadence.

5. HIGH LEVEL CONTROLS & SOFTWARE

The ATST software provides the means to control and coordinate observations performed with the

telescope and instruments. Numerous types of software will be in use on ATST, ranging from the lowest

level servo or logic controller to the highest-level queue and scheduling processes. Each of these software

components fulfills some part of the science requirements for the ATST mission.

The ATST software system is designed to operate the ATST through all stages of observational detail,

from science program submission into observation scheduling, on to instrument configuration and data

collection, through data reduction and archiving, and finally to data retrieval. This chapter describes how

the software performs each of these tasks and how the requirements for each task drive the design in a

particular direction.

5.1 SOFTWARE DESIGN REQUIREMENTS

The software requirements have been derived from several sources. First, the SRD defines a number of

functional and performance requirements that may be traced through the whole ATST software design.

Chief among these are the requirements for flexibility, adaptability, and availability. Second, additional

requirements have been discovered at ATST workshops or design reviews. And third, the technical

engineering requirements further constrain the software design in behavior.

The software design is based upon the scientific and operational requirements. These requirements may

be categorized into four principal areas.

Instruments: The software system must operate all instruments through their complete range of

functionality. Instruments must coordinate observations with other instruments and with other

components of the ATST. The instrument configuration must be flexible and dynamic to support a variety

of experiment setups. Future instruments should not be constrained in their design by ATST software.

Telescope: The telescope must be capable of software control for acquiring, tracking, guiding, and

offsetting on and around the sun. The required accuracies for each function should be met by both the

mechanical and software system. The telescope software must provide the science image to the requested

location with the required image quality.

Observations: Observers must be able to operate the telescope and instruments in a variety of ways.

Observations may be taken at the telescope or remotely, they may be performed in real-time or scheduled,

and they may be synchronized with other local or remote observations.

Data Handling: The instruments may generate a large volume of data. The data must be transferred from

the instruments to a permanent storage facility. The data must contain pertinent state information about

the observation.

5.2 SOFTWARE DESIGN DESCRIPTION

The four categories above are the major systems responsible for each aspect of operation of the software.

The Observatory Control System (OCS) coordinates programs, experiments, and observations. The Data

Handling System (DHS) manages the flow, storage, and processing of image data. The Telescope Control

System (TCS) operates the mechanical and optical structures. The Instrument Control System (ICS)

coordinates the instruments and their associated calibration systems. Together the principal systems

provide a lifecycle structure within which an observing program may reside.



Communications: The principal systems need to communicate with each other to coordinate telescope,

instrument, data, and observer activities. The communication channels are well defined and simple, since

the majority of activities occur within each system and not between them. There are two types of

communications activities, commands and events. Commands are synchronized activities between a

client, who requests an operation, and a server, who performs the operation. An example of a command is

a request to configure an instrument's mechanical assemblies. The requesting client, usually the OCS's

instrument user interface, issues a configuration command to the instrument's server process. The

command is parsed for accuracy and completeness, the readiness of the instrument is checked, and the

result of the configuration operation is returned to the client. Note that the returned result indicates only

that the configuration was accepted and the required movements or reconfigurations were begun.

Command completion is returned though the event channel. By performing actions in this three-step

process—called command-action-response—the server remains available to execute other commands

while the last command is underway. This is extremely useful if the next command is an abort.

Observations: An observing program is the basis for all that goes on in the control system. The origin of

the observing program is in the science proposal, where an observer selects one of the many possible

operational scenarios. For instance, the observer may choose to use the ATST in its diffraction-limited

mode. The observer further selects the instruments and targets. Finally, the operational strategy, or

sequence of operations, is determined. All of this information is incorporated in an observing program.

Sometimes an observing program may be a richly complex series of instructions and system interactions,

possibly describing a synoptic, or regularly scheduled, observation. Sometimes it may be a simple set of

operations to release telescope and instrument control to the observer for setup or serendipitous

observations. Regardless of the complexity or lack thereof, the observing program encapsulates

everything necessary to execute the planned observations.

5.2.1 Common Services

The Common Services architecture provides the infrastructure used by all ATST software, from the

lowest level communications protocols to the control mechanisms between components. By using a

unified infrastructure, components can take advantage of both design and development of other

components and access to common utilities (such as events, logging, and databases). The control flow of

the ATST software is enforced through the Common Services architecture, allowing new components to

be easily integrated into the software framework.

Services: The Common Services are responsible for the fundamental services provided to all ATST

software. These services are best described by examining the information flows supported by these

services.

Command-action-response directives: Direct control of one system component by another component

is accomplished using the command-action-response model pioneered by Gemini. In ATST, commands

are implemented as state-change directives. To effect a change in behavior of a target component the

controlling component describes the conditions necessary to accomplish the state-change by providing the

target component with a set of attributes (name, value pairs) that characterize the difference between the

existing state and the desired state. This set of attributes, along with a unique identifier, is a

configuration. A configuration may be simple, consisting of only a few attributes, or it may be quite large

with hundreds of attributes.

Connections: System components operating in a distributed environment must be able to locate those

other components that they need to communicate with. In ATST, a connection service tracks the

locations and status of all system components and provides a name service that is consulted when a

connection needs to be established. If necessary, the connection service is capable of directing that a non-

running component be started in order to satisfy a connection request.



Figure 5.1. CS-1: The classic container/Component Model

Event Notifications: Status information is communicated throughout ATST using the event service.

Events are messages that are broadcast from some source. System components that are interested in

particular events must subscribe to the appropriate event channel. Events are published by name and

contains sets of attributes as values. Subscription to events is also by name but wildcards may be used by

a subscriber to receive classes of events through a single channel. The ATST event service is reliable and

high-performance. Events from a given publisher are also delivered to subscribers in the same order in

which they are published. All events are time stamped and identify their source – both the generating

component and the configuration that was active in that component when the event was generated.

Alarms: System alarms have the same structure and distribution properties as events but are functionally

distinct. Alarms denote abnormal conditions that require operator intervention. Alarms are not

considered an integral part of the ATST safety system, however. Ensuring safety is solely the

responsibility of the Global Interlock System. Alarms are useful for monitoring safety status as well as

other abnormal conditions and software systems may be implemented to refuse many commands when

unchecked alarms exist. Alarms are tagged in the same manner as events.

Log Messages: Log messages are simple string messages that record system activity. As with events and

alarms, log messages are transmitted using a publish/subscribe mechanism and are time stamped and

source tagged. All log messages are categorized as one of debug, note, warning, and alarm (the alarm log

message category is not isomorphic to system alarms – all system alarms are logged using an alarm log

message but the handling of system alarms does not depend upon this logging). In addition, log messages

in the debug category have an associated level, and debug messages are only published if their level is

less than or equal to the current debug level of the originating component.

Persistent Stores Access: A great deal of information in ATST needs to be recorded for arbitrary periods

that are independent of the lifetimes of specific system components. In addition, system components

need access to initialization parameters on startup and reinitializations. Finally, information specific to an

experiment (virtual instrument details, science programs, configurations, and science header data) is

preserved. ATST uses various persistent stores for these types of information. System components have

access to these stores either directly or through database proxy services.

Alarms and log messages are always recorded in persistent stores. Events are not normally recorded but a

high-performance engineering archive is available for recording events upon demand. These persistent

stores are searchable using a general query mechanism under program control.

Application Framework: The Common

Services also provide the application

framework supporting consistent operation

in a distributed environment. The ATST

application framework is based on the

Container/Component Model made popular

by EJB (Enterprise Java Beans), Microsoft's

.NET initiative, and CORBA's CCM and

patterned after the model developed as part

of the ALMA project (Figure 5.1). The

OCS provides containers that wrap system

components in a common environment

providing uniform access to services.

Component developers focus on

implementing the functionality required of

each component and rely on access to a

container for basic services.



Figure 5.2. OCS-1: Functional Categories of the OCS

5.2.2 Observatory Control System

The OCS has the following responsibilities:

Management of system resources

Management of experiments, science programs, observations, and configurations

Coordination of the TCS, ICS, and the DHS

Management of ATST systems during coordinated observing with other observatories

Essential services for software operations

User interfaces for observatory operations

In general the OCS assumes managerial responsibilities for the ATST system and directs the activities of

the remaining principal systems. Services that are central to the operation of ATST software are provided

by the OCS. The OCS acts as the interface between users and the ATST systems during normal

operation, allowing users to construct science programs and virtual instruments for use in an experiment,

monitor and control the experiment, and obtain science data from the experiment.

The OCS also provides basic services to support system maintenance and general system engineering

operations. This includes tools to examine system diagnostic information, handle alarm conditions,

monitor safety systems, and perform routine engineering tasks.

Functional Organization of the OCS:

The OCS can also be viewed as

organized hierarchically into broad

functional categories: application

support, experiment support, and

resource management. The top levels

of these categories are shown in Figure

5.2.

The application services provided by

the OCS include the event, alarm, log,

and persistent store services. The

application framework includes APIs

and libraries as well as a general

framework for building and deploying

ATST applications. The TCS, ICS and

DHS (as well as the OCS itself) are

resources that are managed by the OCS. The OCS provides for direct operator control of these resources

as needed. However, the normal operational model is to allow experiments as much resource control as

practical over the resources that are allocated to that experiment.

Performing Experiments with the OCS: Experiments are the heart of ATST operations, and the control

system is designed with this in mind. A laboratory-style environment provides flexible support to carry

out experiments that are likely not understood or defined at the time the laboratory itself is designed. An

experiment undertaken at the ATST requires a Virtual Instrument and a Science Program of

Observations. The OCS interacts with the ICS to create and manage virtual instruments. Science

program management is the sole responsibility of the OCS.



Figure 5.3: DHS functional architecture

5.2.3 Data Handling System

The DHS is responsible for:

Bulk data transport

Quick look channels

Data storage, retrieval, and distribution

Data reduction pipelines support

The top-level functional architecture appears in figure 5.3.

The DHS manages the flow of scientific data collected by ATST instruments. The data reduction

pipelines support is a potential upgrade that is supported by the initial DHS design.

Because of the performance requirements placed on the DHS, parts of its functionality are distributed

across other system components. For example, instrument camera systems perform any data processing

required to reduce data output to meet bandwidth restrictions imposed by the implementation of the bulk

data transport. Similarly, instrument component developers are responsible for providing the processing

steps required to convert raw quick-look data into meaningful quality-control information.

Bulk Data Transport: The role of the bulk data transport is to reliably transfer science data from

scientific cameras sources to data store targets (Figure 5.4). Physically, the bulk data transport uses

multiple data channels on a high-performance switched network. The use of a switched network allows

for increased flexibility – data channels can be established between any data source and target.



Figure 5.4. DHS-2: DHS 'pipeline' showing distributed functionality

Figure 5.5. DHS-2: Routing of Quick look data using publish/subscribe

The ATST scientific cameras are individually capable of generating large amounts of data quite rapidly.

This is compounded by the fact that multiple experiments may run simultaneously, each using multiple

cameras. The bulk data transport is to be implemented using the latest stable technology for high speed

data transfers and operates using data channels that are physically distinct from other system

communications. This ensures that system control and monitoring activities may continue unaffected by

bulk data transport loads.

Quick Look Channels: ATST cameras can generate quick-look images and post them onto quick look

data channel streams. The quick look facility is a publish/subscribe mechanism allowing applications to

accept, process, and display quick look data from any source (Figure 5.5). This allows, for example, an

operator’s GUI to display quick look data while a separate process performs automatic analysis of quick

look data with feedback into the ATST image quality control system. At the same time, a third process

may be sub sampling the same quick look data channel and recording selected images. The publish/

subscribe mechanism also simplifies display of quick look data at multiple stations in the observatory.

Data Storage, Retrieval and Distribution: ATST provides temporary storage for all scientific data

products and permanent storage for calibration data products. The temporary storage acts as cache

between the high-speed bulk data transport and slower distribution media (DVD, tapes, removable hard

drives, etc.). All science data, whether located in temporary or permanent storage, maintains associations

via a relational database with the configurations, observations, and experiments that were involved in the

creation of the data. Header information is also archived and associated with data products. Any header

attribute may be used as a key to retrieve one or more data products from the store.

When an experiment completes, the data products and all related ancillary products associated with that

experiment are retrieved from the store and made available on distribution media for that experiment’s

investigators.



Figure 5.6. Telescope Control System.

Support for Data Reduction Pipelines: Support for on-line data reduction processing is a future upgrade

to the ATST system. The use of a switched network for bulk data transport and database access to all

data products simplifies the process of integrating this upgrade into ATST.

5.2.4 Telescope Control System

The TCS is the central coordination facility for the delivery of the solar image to the instrument. It is

responsible for the precise pointing and tracking calculations necessary to observe the sun. The TCS does

not itself operate any mechanical components;

rather it delegates this responsibility to the

various ATST telescope subsystems and

manages them according to the observation

requests. The TCS does interact with the other

principal systems, most notably the OCS and

ICS. Observation configurations generated by the

OCS are sent to the TCS for proper telescope

positioning and configuration. Coordinated

events are returned by the TCS so the OCS (and

associated observer) is informed about the

telescope's status. If an instrument uses the

telescope in any coordinated fashion, such as

scanning or calibration, the ICS and TCS

synchronize these activities through the TCS

interface (Figure 5.6).

The TCS also manages the wavefront image reconstruction process. The high-speed adaptive optics

corrections take place in the Adaptive Optics Control System (AOCS), a TCS subsystem. The TCS

manages the state of the AO system, including the offload of accumulated errors to other subsystems.

The TCS delegates the high-speed control loops of the telescope components to its subsystems. The green

elements in Figure 5.6 show the major TCS subsystems and their relationships. The scope of a

subsystem's functionality is limited both by construction and control. Generally, if both the error

detection and error correction of a simple control loop is handled locally, that loop is part of a subsystem.

For instance, the M1 mirror assembly has a number of axial force actuators that detect applied forces and

apply corrective forces to position the actuator at the correct position. The force map required to figure

the primary mirror is downloaded from the TCS to the M1 Control System (M1CS), but the M1CS is

responsible for positioning and maintaining the actuators in their proper positions. The TCS coordinates

the control loop by downloading “set-points”; the M1CS provides the actual control loop function.

Similar control methods are used for other subsystems like the AOCS and the Mount Control System

(MCS).

The description of the TCS subsystems can be found in their appropriate mechanical or optical system

description. These systems are developed and delivered by the subsystem vendor and follow an interface

to the TCS. A brief description of how they interact with the TCS is given here.

Enclosure Control System: The ECS operates the enclosure carousel and shutter drives to properly

position the entrance aperture. The enclosure needs to be moved to an accuracy of better than one-half

degree, and must avoid collisions with the telescope mount assembly due to the shape and dimensions of

the enclosure volume. The TCS provides a stream of altitude and azimuth trajectory data to the ECS.



Mount Control System: The MCS operates the telescope altitude and azimuth drives to properly

position the telescope mount assembly. It also controls position of the Nasmyth and coudé rotators. The

TCS provides trajectory information at 20 Hz to the MCS.

M1 Control System: The M1CS controls the axial support actuators used to shape the figure of the

primary mirror. The TCS provides the default shape based upon the current position on the sky.

Additional shape information from the active optics may be delivered by the TCS.

M2 Control System: The M2CS operates the tip-tilt-focus actuators on the secondary mirror. The TCS

provides the default positions based upon the current position on the sky. Additional tip-tilt-focus

information from the adaptive optics may be delivered by the TCS.

Feed Optics Control System: The FOCS controls the smaller mirrors delivering the image to the coudé

or Nasmyth instruments, and it controls the calibration equipment located at the Gregorian Optical

Station. The TCS provides the required commands needed to position the optical elements.

Wavefront Correction Control System: The WCCS controls the wavefront correction hardware,

including the adaptive optics real-time controller and the active optics wavefront sensors. The TCS

manages the distribution of image correction data to the appropriate subsystems.

Acquisition Control System: The ACS operates the external acquisition telescope. The full-disk image

is used by the operator for target selection and positioning. The ACS provides 1 arc-second resolution in

four different imaging wavelength bands at rates up to 10 Hz.

Polarization, Analysis, and Calibration Package: The PAC operates the optical elements located at the

Gregorian Optical Station (GOS). The TCS configures the PAC for the appropriate observation type by

selecting suitable calibration optics and/or polarizers.

Heat Stop Assembly: The HSA absorbs the rejected solar light near the prime focus. The TCS controls

the associated Lyot stop and occulter mechanisms. It also monitors the thermal performance of the HSA.

Pointing and Tracking: The TCS has the responsibility for target acquisition and tracking. This requires

a number of simple yet important steps. First, the principal target for the ATST is the sun. This requires a

calculation of the solar position and rate. The second step is the conversion to an appropriate coordinate

system. Solar observations are usually carried out in either heliocentric or heliographic coordinate

systems.

Closing the loop on the TCS pointing is achieved by guide error signals from the adaptive optics system.

Figure 5.7 shows the TCS servo loops required to track features on the sun. Although the AO system has

a very good resolution of the positional error, the features tracked by the AO system may be moving with

relation to the solar center. The AO signal is useful to the TCS when the tracking of the telescope is to be

unlocked from the solar disk, thus allowing the telescope to guide on this moving feature.

Active and Adaptive Optics: The TCS is responsible for controlling the image quality parameters of the

telescope optics. The adaptive and active optics control systems provide residual wavefront error data in

the form of Zernike coefficients that is propagated to other telescope subsystems. Figure 5.7 shows the

schematic flow of Zernike coefficients from the adaptive optics system to the other TCS subsystems. Tip-

tilt-focus offloads at about 100 Hz are sent to the M2 mirror controller for removal of moderate

bandwidth errors (on the order of 10 Hz) such as wind shake. Any accumulated M2 bias needs to be

removed to prevent pupil wander; this error is further offloaded to the mount controller at about 1 Hz.



5.2.5 Instrument Control System

The ICS is responsible for managing virtual instruments. The ICS provides the mechanisms for

associating components into virtual instruments, determining the availability of components, and holding

the representations of the virtual instruments. The ICS also maintains a set of active instruments: those

virtual instruments that are physically realized and actively sharing the light beam. Thus the ICS enables

multiple experiments to take place simultaneously.

The ICS assumes no active role during the carrying out of an experiment. The OCS directs control of a

virtual instrument during an experiment.

Requirements: The ATST is required to provide the flexibility inherent in a laboratory environment.

This is a key science requirement, and has a significant impact on the system design both in mechanical

systems and in software. The DST at Sunspot, NM, is specifically mentioned as a model that well

illustrates the desired flexibility. On the DST a series of optical benches on a protected rotating platform

provide the principal support for observing. Scientists can construct instruments specific to their

experimental needs from existing components. While a few instruments are “facility” and consist of a

fixed set of components, even these instruments may be combined with other components using

dichroics, beam splitters and slit-jaw reflections. The ATST mechanical systems provide a similar

flexibility through optical benches on a two level rotating coudé platform in the telescope pier.

The Experiment: Observers at ATST are

interested in performing Experiments (Figure

5.8). A central tenant of the ATST control

system model is that the system should be

adapted to the requirements of the

experiment. A laboratory environment

provides flexible support to carry out

experiments that are likely not understood or

defined at the time the laboratory is

designed. Consequently, experiments are a

formal concept within the model.

Figure 5.8. Information flow in an Experiment

Figure 5.7. Tracking and adaptive optics control loops controlled by the TCS.



Figure 5.9. A possible instrument configuration

In ATST, an experiment includes a science

program of Observations and a Virtual

Instrument capable of performing those

observations. (The experiment also

ultimately includes Results but that aspect is

not relevant to this discussion.)

Observations contain sequences of

operational steps describing the behavior of

the instrument. Each operational step

consists of a set of configuration parameters

and a simple command describing a state

change within the instrument, collectively

referred to as a Configuration (Figure 5.9).

This use of science programs is typical of

modern observatory operations and matches

similar functionality provided at SOLIS,

Gemini, VLT, ALMA, and other

observatories.

What differentiates the ATST approach from these other observatories is that, instead of adapting

experiments to fit within the bounds imposed by instruments consisting of fixed components, the ATST

observer can construct a virtual instrument from available components to meet the needs of the particular

experiment. This provides a great deal of flexibility in the nature of experiments that can be performed at

ATST.

The Virtual Instrument: Instruments consist of one or more Components. Some components may be

purely mechanical with no associated software (e.g., a dichroic filter). Others may be purely software (a

sequencer). Most, however, include both mechanical and software aspects (cameras, scanners, etc.).

These last components are called Devices.

In a conventional instrument the set of components that comprise the instrument are fixed and

permanently associated with each other. Nevertheless, there is some software that understands these

associations. Thus the primary difference between a virtual instrument and a conventional instrument is

merely that the associations within a virtual instrument are not fixed but rather managed by software. A

subtle difference that is implemented in the ATST virtual instrument model is that telescope components

can also be associated as part of a virtual instrument. For example, an experiment that needs to perform

drift scanning across the solar disk can include the telescope mount as a component, while an experiment

performing coronal observations is likely to include the occulter as a component.

Scientists assemble the requisite components and combine them into the virtual instrument. Virtual

instruments are then named and saved for potential future use. Some virtual instruments are used so often

that the physical component associations are also maintained. This would be the case for ATST facility

instruments.

From a control perspective, once the component associations have been made the control of a virtual

instrument is identical to that of a conventional instrument. This simplifies the integration of instrument

operation within the otherwise conventional ATST control system.

Components are hierarchical and may be composed from other components. In particular, instruments

themselves are components. Composition of instruments is a common feature of operation of the DST

and is expected to be a key operational characteristic of ATST as well. With a few additional components

several facility instruments may be associated into a new virtual instrument.



Figure 5.10. The lifecycle of a virtual instrument

Some experiments require cooperation between ATST and off-site observatories (including off-planet).

A virtual instrument can include special proxy components for coordination with off site facilities.

The Life Cycle of a Virtual Instrument: The first step when performing an

experiment with ATST is to construct a

virtual instrument (Figure 5.10). This can be

done either by browsing and selecting an

existing virtual instrument or constructing a

new virtual instrument from a catalog of

system components. The components are

then configured as needed for this

experiment. While many components can be

configured by setting a few parameters,

others – such as sequencing components –

may take more effort to configure, depending

on the requirements of the experiment. Once

the virtual instrument is defined, it is

registered with the ICS, which records the

instrument.

A science program that uses that instrument can now be constructed and added to the experiment.

Observations within the program may be controlled through sequences of configurations or interactively.

In either case the observations are scheduled with the OCS for execution.

As the time for observing approaches, the scientist lays out the physical systems associated with the

virtual instruments components and prepares for observing. The OCS also notifies the ICS that a

particular virtual instrument is needed for an upcoming observation and the ICS confirms that it is

available and enabled. When enabled, the virtual instrument assumes control over its components, and

observations proceed according to a prescribed sequence.

Once the observations in the science program have been completed, the OCS notifies the ICS that the

virtual instrument is no longer needed. The ICS then deactivates the instrument. Once deactivated, the

virtual instrument’s physical layout may be preserved for future use in other experiments or the

instrument's physical systems may be made available for use in other virtual instruments.

The Role of the OCS: The OCS acts as the interface between the scientist and the ATST control system

as a whole. The OCS maintains the science programs and sequences the observations of experiments by

providing configurations to the associated virtual instruments. The actual sequencing may be

accomplished using scripts or through graphical user interfaces for interactive observing.

Implementation of the Virtual Instrument Model: Virtual instrument component software is

constructed using a Container/Component Model as the basic framework. Containers provide access to

services required by the components and are responsible for managing component life cycles. This

allows component developers to concentrate on the functionality required in the component and provides

a common implementation for standard features. This approach also enhances the distributability of

components. Software only components can be instanced on arbitrary computer systems easily as long as

a container is available to hold the component.



Figure 6.1. The Enclosure.

6. ENCLOSURE

The enclosure is comprised of five major components: (1) the

carousel; (2) the carousel drive system; (3) the lower

enclosure; (4) the enclosure thermal system, (5) the ancillary

mechanical systems; and (6) the enclosure control system.

These five items are described in detail, below. The external

enclosure elements appear in Figure 6.1.

6.1 ENCLOSURE DESIGN REQUIREMENTS

The ATST enclosure is comprised of the structural

components, drives, and thermal equipment that are used to

protect and track with the Telescope. The enclosure will meet

the following requirements:

Provide complete protection for the telescope and

optics under all weather conditions expected at the

Haleakalā Observatory site (survival & operations);

Point, track and slew along with the telescope over its

full required range of travel, while providing full

shading of the telescope structure;

Provide an unobstructed optical path from the sun to

M1, with acceptable seeing characteristics in and

around the Enclosure;

Provide a light-tight display to the surroundings when

closed at night; and

Provide the telescope with protection from wind-induced vibration and mirror buffeting, while

still allowing good flushing characteristics in and around the Telescope.

In addition to these top-level requirements, there are a number of second-level functions that the

enclosure provides. For example, the enclosure has a variety of safety systems and features included to

protect personnel and the telescope from damage (e.g., failsafe brakes; GIS interface, etc.). The complete

specifications and design for the enclosure, including all the top-level and second-level requirements, are

outlined in the Enclosure Design Requirements Document (ATST Document #SPEC-0010).

6.2 ENCLOSURE DESIGN DESCRIPTION

The enclosure design is critical to the performance of ATST. It must protect the telescope from wind,

weather, and direct sunlight (except on the primary mirror), and must do so without contributing

significantly to local seeing. Analyses have shown that even a white-painted enclosure requires active

skin cooling systems to keep from generating self-induced seeing. The design described below

incorporates active cooling of all critical insolated areas. Careful attention has also been paid to the

geometry of the structure for optimum performance and minimum operational cost.

6.2.1 Carousel

The carousel is the large structure that forms the basic envelope for telescope protection. It rotates about

an azimuth axis that is coincident with the telescope azimuth axis. It is a highly ventilated, but includes

cooled sun shades to prevent direct sunlight from entering the enclosure through the vent opening. The

enclosure and telescope can rotate independently. The design also features steeply sloped sides on either

side of the carousel entrance aperture that minimizes surface area normal or near-normal to the sun

(Figure 6.2). A 5° taper from front to back along the sides also contributes to maximizing surface area

that remains shaded through tracking operations.



The carousel structure design is based on modern enclosure construction methods incorporating a large

steel truss ring for the base, supporting dual arch girders, along with intermediate framing and supports.

The plate coil panels (see section 6.2.4, Enclosure Thermal Control) will be used directly as the cladding

system in a standing seam or other weather-tight configuration. In areas that don’t require plate coil

coverage, panels designed for the same size and expansion properties will be used. This greatly simplifies

the detail work necessary to create a leak-free exterior envelope. All exterior surfaces are finished using

the thermal coating system detailed below. Foam insulation is applied to the interior surface to minimize

surface temperature variation. The structure is designed as an ‘exoskeleton’ with a weather-tight skin

installed on the inside of the structure. Fire-rated insulated steel panels, such as the DSL-FR model of the

Metecno-API line, are used as shown in an industrial application to the right.

Shutter Assembly: The shutter assembly is comprised of a primary and a secondary shutter (Figure 6.3).

As in the carousel, plate coil panels comprise all portions of the shutter skin that receive solar insolation

during tracking operations, with ‘dummy’ panels making up the rest. They are mounted on the exterior

side of a powder coated tubular steel frame which also serves as a mounting structure for the shutter

support and guide roller assemblies. The primary shutter is actually made in two pieces and bolted

together in a gasketed, flange-like, weather-tight arrangement so that when separated, starting from the

zenith-pointing position, half is left in place. The remaining, driven portion is then moved to the horizon-

pointing position, leaving a large opening to facilitate the staging and assembly of the telescope mount.

Carousel Entrance Aperture

Carousel Aperture Stop

Primary ShutterSecondary Shutters

Primary Shutter Flange

Secondary Shutter

6” diameterTelescopeAlignmentHole w/Plug

Shutter Seal

Shutter Stop

Figure 6.3. Shutter Assembly – on the left is a view from the front with the Shutter in the zenith-pointing position; on

the right is a view from the back with the Shutter in the horizon-pointing position.

Figure 6.2. Carousel geometry.



The primary shutter requires an inner skin to provide a sealing surface between the various shutter

segments. Since the thermal coating on the outward facing surface of the plate coil on the shutters won’t

hold up to the friction of a seal scraping over it, the seal is made from below. Since there will be

conditions that cause the formation of condensation, the interior slopes to a drain. All drains are routed to

a sump where the condensate is pumped through the utility transfer system for disposal.

There is a 5.4-meter diameter carousel entrance aperture in the primary shutter is oversized with respect

to the ultimate requirements. This is to accommodate a separate carousel entrance aperture stop. A roll

up door, is provided in the shutter.

The secondary shutter segments are constructed similarly to the primary shutter. Plate coil panels

comprise all portions of the shutter skin that receive solar insolation during tracking operations, with

‘dummy’ panels making up the rest. They are mounted on the exterior side of a powder coated tubular

steel frame which also serves as a mounting structure for the shutter support and guide roller assemblies.

Unlike the primary shutter however, the secondary shutter segments are not driven; they are passively

lifted by the primary shutter using a grab bar arrangement.

Sun Shades: The sunshades over the vent gates are plate coil panels mounted on tubular steel frames

attached to the underlying carousel structure. When viewed from above, they provide shade to the entire

carousel surface below it down to the next lower sunshade. They also serve to channel the wind into the

interior of the enclosure.

Azimuth Track and Bogies: The azimuth track is a large hardened steel ring that is mounted to the top

of the stationary enclosure base. The rail is shaped and finished so the lateral guide rollers have stable,

smooth mating surfaces; there is provision for the seismic restraint system; and, the convex bogie wheels

are provided a stiff and smooth surface to traverse on the top. Each of the rail sections is individually

machined flat and subsequently bolted to the other sections with an integral shimming system. Once

assembled, and trued with the shims, a final in-place machining operation is performed to achieve the

required flatness if necessary.

To keep the carousel on the azimuth track in the event of an earthquake, seismic restraint brackets are

utilized. These brackets hang down from the underside of the carousel structure and extend underneath

the edge of the azimuth track. There is a small clearance between the extension portion of the bracket and

the underside of the azimuth track. If the enclosure experiences upward movement relative to the azimuth

track, the bracket minimizes the motion, thereby keeping the carousel on top of the track.

There are 16 support bogies arrayed on the underside of the carousel, four of which are drive units; the

other twelve bogies provide vertical support only. A series of guide rollers provide lateral definition. The

bogies are designed with a suspension system to minimize vibration transfer from the enclosure into the

foundation. Alignment of the bogie wheel is adjustable with respect to the center of rotation as well as

inclination with respect to vertical. Load cells are incorporated for bogie loading adjustments.

6.2.2 Carousel Drive System

The carousel drive system is comprised of the drive assemblies, encoders, and controllers that allow the

carousel and shutter to move in synch with the telescope during operations.

The altitude and azimuth axes of the enclosure feature closed loop three-phase permanent magnet AC

servomotors. All motors are NEMA premium efficiency motors. Using a modified flux vector control

algorithm and feedback from a motor mounted incremental encoder, the motor will behave very much

like a DC Motor but without the maintenance. Full torque at zero speed, ability to directly command

motor torque, tight speed regulation, quick acceleration, and fast positioning are all available. On the

azimuth axis, the 20-hp vector drives are the motive power for the cable-driven bogie assembly.



The shutter drives are based on the Keck design, though there are some significant differences. Instead of

being mounted within the shutter proper, the ATST system mounts the drives at the azimuth mechanical

floor level at each end of the arch girders. This puts a significant potential source of heat further away

from the optic path and in a place where access for maintenance is less challenging.

The cables are attached to the primary shutter and are guided along each arch girder. There is a drive

motor on each end of each cable, keeping the cables in tension. Load cells on the cables along with torque

levels allow the servo control systems to maintain this balance. The Power Pac Wire Rope by Wire Rope

Industries consists of oval wire cables with an expected life of 30 years. Annual tensioning and

inspection will be required.

Fail-safe, redundant brakes are used on both the altitude and azimuth axes. In both cases, they are used as

parking brakes or in case of power outage or emergency stop only. In normal operation, the motors are

used for deceleration. As illustrated above for the shutter case, these units are spring set with the

activation of the electromagnet in the solenoid commercial units. The drive bogies have a similar

arrangement and in both cases, the discs are mounted directly on the drive shaft of the drive motors.

The encoders for the altitude and azimuth axes are high-resolution absolute linear non-contact encoders

(e.g., Stegmann KH 53 Pomux® Type B) that are matched to Omega Profile sections containing

permanent magnets as a unique position marker. In addition to these absolute encoders and the

incremental encoders installed on each drive motor, limit switches are used in discrete travel positions at

the end of motion range to act as a back up to the system and to supply position information to the ECS.

The altitude encoder Omega Profile sections are mounted with a non-ferrous OEM system 10-cm away

from the steel surface along each arch girder. The non-contacting read heads are attached to adjustable

brackets affixed to the primary shutter. The azimuth encoder Omega Profile sections are similarly

mounted to the inner diameter of the Lower Enclosure ring beam. The read heads are affixed to adjustable

brackets extending downward from the bottom of the azimuth mechanical level floor.

6.2.3 Lower Enclosure



The lower enclosure is comprised of the stationary structure that supports the carousel and transfers its

loads to the ground (Figure 6.4). The lower enclosure includes foundations and interface to the soil. The

lower enclosure also includes the stationary floors, stairs, ladders, access doors, catwalk, and mounting

surface for the carousel azimuth track.

The foundations for the lower enclosure will be designed as a system along with the foundations for the

telescope pier and the support and operations building. They will be designed to minimize any vibration

transmission from the lower enclosure into the telescope structure. The lower enclosure structure is

conventional steel construction.

Half of the skin panels of the lower enclosure are actually integral to the lower enclosure cooling system.

More fully described below, they are perforated plate coated on the outside with the same thermal coating

system that covers all exterior components. The remaining panels are light gauge steel plate welded in

place to form full-size panels. Again, they are coated on the outside with the same thermal coating

system that covers all exterior components.

6.2.4 Enclosure Thermal System

The enclosure thermal system includes a passive ventilation system, an active ventilation system, the

carousel cooling system, the lower enclosure cooling system, and a thermal coating system.

Azimuth Track

Stationary side ofCarousel Seal

Ring Beam

Catwalk

Stationary side ofCable Chain carrier

Catwalk EmergencyEgress Ladder

Structural members(typical, non-optimized)

Area without flooring, left open for airplenums (see Sec. 7.4)

Personnel access

Support columnsand footers

Maintenance andequipment access

Figure 6.4. The Lower Enclosure.



Passive Ventilation System: Passive flushing is provided by a combination of twenty independently

controllable vent gates along with the carousel rear access door and the carousel entrance aperture.

Outside wind blows through the vent gates and over the telescope structure, removing thermal turbulence

and assisting the thermal control of the mount and optics (Figure 6.5). When ambient conditions are

excessively windy, the vent gates can be partially closed to throttle the interior wind speed using

commercial roll-up shutters. They are opened prior to the observing day in order to equalize the inside

and ambient conditions. Assistance from the active ventilation system will be required during mornings

with no wind.

Active Ventilation System: Eighteen propeller fans, nine on each side of the enclosure, provide air

movement across the mirrors under no wind or low wind conditions. The fans are sized to provide

roughly 0.5m/s air flow across the mirrors’ travel paths. The system is designed to be used with the

ventilation doors open so that the fans draw in ambient air.

Carousel Cooling System: All surfaces of the carousel which receive sunlight are comprised of or

covered with plate coil heat exchangers containing a propylene glycol solution circulating through to

remove the resulting thermal load. The conventional chiller and circulation pump are remotely located in

the utility building to the west of the support and operations building. The chilled water is supplied at

4°C below the highest ambient temperature of any of the approximately 60 zones to all of the plate coil

heat exchangers. The volume/velocity is controlled for each zone so that the return does not exceed the

local ambient temperature. The control valves are all located at the azimuth mechanical level for

accessibility. The circulation can be shut down during periods of high ambient wind conditions when the

resulting fully developed forced convection is effective in removing the thermal load.

Lower Enclosure Cooling System: The lower enclosure cooling system has been optimized for times of

excellent seeing, which typically occur in the early morning. Data collected during the site survey

indicates that while there is a significant amount of excellent seeing during times of no or low wind, so

the lower enclosure cooling system is limited to those surfaces subject to insolation during the morning

hours. During times of moderate to high wind, the heat fluxes developed are removed by the wind.

The lower enclosure is cooled to within 1.5°C of ambient temperature by drawing ambient air through the

perforated metal skin and exhausting it west of the utility building. The 0.5-mm thick perforated plate has

a 2-3% open area using 1-mm diameter holes. The vane axial fans for this system are located in the

utility building.

Figure 6.5. Passive flows through the Enclosure.



Thermal Coating System: The Thermal Coating System is based on AZW/11-LA Inorganic Low Alpha

White, non-specular thermal control coating manufactured by AZ Technology or IIT Research Institute.

The coating provides superior thermal protection by allowing only 8-12% of the solar radiation impinging

on the surface to be absorbed through to the interior systems while emitting 90-92% of the internal heat

generated. AZW-11LA incorporates a stabilized pigment system with a silicate binder.

6.2.5 Ancillary Mechanical Equipment

The ancillary mechanical systems include minor mechanical elements of the enclosure such as cranes,

sensor arrays, access doors, utility transfer systems, carousel entrance aperture cover, and other

miscellaneous elements of the enclosure.

The rear access door has two functions: access from the equipment lift for the primary mirror when

mounted on the mirror cart and additional passive ventilation capacity. The door is a standard dock-type,

wind-rated roll-up door. It is located at the back of the carousel. There is a matching door in the primary

shutter that coincides with it when the shutter is in the zenith pointing position. The door in the Carousel

is slightly larger than that in the shutter to provide access for maintenance to the shutter door.

The carousel entrance aperture provides a larger-than-necessary hole (5.4-meter diameter) in the shutter.

A smaller (movable) aperture stop is affixed to the shutter. Its cover is a smaller version of the

commercial roll-up shutters used for the vent gates.

A utility transfer system provides coolant, power, signal, and other utilities to each shutter segment. The

altitude wrap is provided by a series of hanging cable drapes. A guide trough is provided. A COTS-type

cable chain is used to support and manage the cables and utility lines. The azimuth cable wrap is

mounted just below the azimuth utility floor level. The wrap is non-powered; its movement is caused

directly by the movement of the carousel. Another COTS-type cable chain is utilized here as well.

7. SITE INFRASTRUCTURE AND SUPPORT FACILITIES

Site Infrastructure is the compilation of technical requirements, specific site characteristics and a

conceptual layout of the ATST facilities at the Haleakalā site. This includes the support and operations

building and certain equipment within that building, including a mirror cleaning and coating facility; and

mirror handling equipment. Each of these items is described in more detail below.

7.1 SITE SELECTION TECHNICAL DESCRIPTION AND IMPACT

During the D&D phase the ATST tested six candidate sites. We determined that it would be feasible to

build ATST at all six sites, but significant logistical and cost differences exist between them. These were

brought to the attention of project management and the Site Selection Working Group, and became part of

the information used to down-select to three sites. These are Big Bear Lake, CA; Haleakalā, HI; and La

Palma in the Spanish Canary Islands. The eventual recommendation of the Site Selection Working Group

was the Haleakalā site, based on its excellent seeing and low sky brightness.

7.1.1 Technical Site Requirements

Accessibility - A range of vehicles, from standard passenger cars to large construction cranes and flatbed

trucks must be able to reach the facility, both during construction and in long-term operation. This

applies to the roads leading to the site as well as to the local access in the immediate vicinity of the

telescope.

Dimensions – To accommodate the observatory structures a minimum horizontal area of approximately

200 ft. by 200 ft. is necessary, and topography that will allow the creation of a suitable platform. A site



larger than the minimum would allow a more flexible site layout, and would also facilitate construction

staging.

Structural Characteristics – The soil/rock of the site must have sufficient bearing capacity to support

the loads imposed by the telescope pier and the building foundations while also allowing adequate

isolation between the two. Stiffer natural substrates that increase the lowest resonant frequency of the

telescope support system are considered very advantageous. The lateral force factors (seismic and wind)

inherent to the site must be of a magnitude that can be safely designed for without prohibitively expensive

structural measures.

Manageable permitting process – The environmental issues inherent to the site must be such that the

construction of ATST would not likely be precluded based on the applicable environmental protection

statutes. Any necessary construction permits issued by the regional authorities must also be obtainable.

Utility infrastructure – Sufficient electrical power, data/telephone connection, and domestic water/sewer

service must be achievable at the site. Existing infrastructure that can be extended to ATST and a low

cost connection to local utility company lines are considered very advantageous.

The Haleakalā site meets all of these requirements.

7.1.2 The Haleakalā site

This site is at Haleakala Observatory on the island of Maui, within two hours of coastal cites and less than

a one-hour drive from the observatory’s base lab facility (Figure 7.1). The entire compound, including a

large Air Force telescope complex, is

owned by the University of Hawaii and

managed by the Institute for Astronomy

(IfA). Two potential sites within this

compound have been identified for

ATST. The primary site is close to the

existing Mees Telescope. A rendering of

the ATST facility at this site appears in

Figure 7.2.

No local building permits are required for

construction at Haleakalā, however,

environmental permitting is a significant

cost, schedule, and public relations

factor. An Environmental Impact

Statement and a Conservation District

Use Permit are required, and the effort to

obtain these is underway.

Relatively little excavation will be required to create a suitable level platform for the ATST structure.

The volcanic gravel and cinder on the site has inherently low bearing capacity, so the pier and building

foundations will be wider than normal or extend down to more solid rock layers well below the surface.

The utility infrastructure at Haleakalā, especially electrical power and data connection, is well-developed

and has sufficient reserve capacity to serve ATST.

Figure 7.1. The location of ATST on Maui.



7.2 SUPPORT AND OPERATIONS BUILDING

These two building elements (identified in Figure 7.4) are spatially and structurally contiguous and have

similar functional requirements. They are treated as a single entity for this analysis.

Structural Requirements: The structure of the Support and Operations Building carries the weight of

the roof and exterior walls; the interior floor loads of all levels; lateral seismic and wind loads; and the

dynamic loading of the rotating enclosure above. The building structure must be isolated from the

telescope pier to prevent unacceptable levels of vibration from reaching any critical optical elements. The

foundations must be sufficient to safely transmit all these loads to the bearing capacity of the soil.

Thermal Requirements: The design of the Support and Operations Building must minimize any

contribution to thermal turbulence in critical optical paths. This impacts the height of the support

building and its proximity to the enclosure, the appropriate location for heat generating mechanical

equipment, the selection of exterior materials and finishes, the potential need for active cooling of exterior

building surfaces, and the appropriate thermal separation of interior spaces. The CFD analysis used to

model the air flow performance of the enclosure allowed the project to thermally optimize the design and

orientation of the building once the site dependent variables of wind speed and direction, topography, and

available site space were established. The prevailing winds tend to come from the northeast at times of

excellent seeing. Seeing is also best in the early morning hours before heating disturbs the ground layer.

This led to the decision to build the support and operations building to the west of the telescope enclosure.

Functional Space Requirements: Table 7.1 lists the spaces defined within the support and operations

building and associated utility building. If remodeled space within the Mees building is included, the

ATST facility provides approximately 13,170 square feet of space.

Figure 7.2. ATST at the Mees site.



Building Design: The building and layout is depicted in Figures 7.3 and 7.4. The basic construction of

the Support and Operations Building will be a steel-framed structure on concrete foundations with

standard metal panel roofing and siding. This is a conventional building system that is economical,

flexible, and adapts well to a variety of site conditions and lateral load cases. The exterior finish will be

determined by the best thermal performance (probably high-titanium white) with consideration given to

environmental impact issues as required. The interior build-out will be similar to standard office and light

commercial construction.

7.3 FACILITY EQUIPMENT

Facility equipment includes the outfitting and furnishing of the control room, shops, labs, offices and

other ancillary spaces. It also includes the special observatory-related mechanical and electrical

equipment that will be permanently installed in the facility to serve the utility needs of the telescope and

instruments, and to address the extensive cooling requirements of ATST. This does not include normal

building utility items such as lighting, domestic plumbing and general air conditioning, which are

incorporated into the budget and design of the building itself.

The following is a list of the necessary special utility equipment identified to date:

Electrical Generator – Capacity of ~200 KVA.

Uninterruptible Power Supplies – Two units serving a total load of ~50 KVA.

Chillers – Two units of ~30 ton capacity with appropriate temperature range to serve the cooling

requirements of optical components, the heat stop, the telescope mount, the enclosure, and other

special heat sources.

Space Description

ft2

m2 ft m

Control Room 650 60 9 2.7

Computer Room 280 26 9 2.7

Instrument Prep Lab 650 60 10 3.1

Site Manager's Office 130 12 8 2.4

Visiting Observer's Office 130 12 8 2.4

Shared Office Space (~4 people) 400 37 8 2.4

Kitchen/Break Area 250 23 8 2.4

Restrooms (3 - one on each level) 150 14 8 2.4

High-bay Receiving/ Mirror Prep 1,400 130 20 6.1

Mirror Coating Area 800 74 20 6.1

Entry vestibule to enclosure (4 levels) 1,380 128 varies

Platform Lift (4 major levels @460 sq.ft.) 1,840 171 76 23.2

Elevator (4 levels @ 80 sq.ft.) 320 30 65 19.8

Stairs (3 levels @ 160 sq.ft.) 480 45 varies

Machine and Service Rooms 250 23 9 2.7

Mechanical Equip. Space 700 65 N/A N/A

Total Net S&O Building Space 9,810 912

Utility Building 2,560 238 16 4.9

Remodeled Space in Mees Building 800 74 16 4.9

ATST Support Facility Space Requirements

Area Height

Summit Support Facilities

Table 7.1. Space Requirements



Glycol (or other liquid coolant) – Supply and distribution piping as required.

Special Fans – For active ventilation of the telescope enclosure and adjacent spaces.

HEPA Filtration – Air handlers and filters as required for particulate control in the coudé labs,

mirror coating area and possibly in specified stationary lab areas.

Cryogens – Supply, distribution and processing for liquid nitrogen, compressed helium, or other

coolants as required for special instrument related systems.

Air compressor(s), vacuum pumps, and other equipment required for general utility use and

special telescope-related applications.

7.4 MIRROR COATING AND CLEANING FACILITIES

An appropriate area and the necessary equipment are included in the ATST facility for the handling,

cleaning and recoating of M1, M2, and the Nasmyth mirrors. This equipment includes the M1 assembly

handling cart, M1 lifter and coating plant. The availability of an appropriate existing coating facility or

the potential shared use and co-development of this facility with neighboring observatories is being

studied.

The other mirrors will be coated with protected silver to improve throughput at the coudé observing

station. Those relatively small mirrors will be sent out for coating.

Description: The coating plant utilizes an evaporative system to deposit a coating of pure reflective

aluminum. The coating plant itself is a large clam-shell stainless steel vacuum chamber used to apply the

coating to the mirror surface. The coating plant assembly includes vacuum pumping systems, chilled

water delivery, the magnetron assembly, and the vacuum tank itself. Additional associated equipment

fan

fan

tunnel

utility shaft

vehic

ula

r access?

ice

tanksgenerator

Existing

Cistern

support & operations building

Receiving &

Mirror Prep

offices

existing Mees solar observatory

kitchen

Expanded Shop

lab

utility building

existing main observatory road

service &

parking area

ple

num

Equipment

Area

Platform

Lift

Base of

Piers

concrete

pier

Mirror

Coating

Facility

ventilation

lower

enclosure

ups

ups

ups

cond.

n

ew

X-f

orm

er

hatch

chillers

1050

0 5

30 ft.

10 m

north

Notes:

- Building orientation and layout based on CFD analysis and site space restrictions.

- Elevation of ground floor level is 9983' (~4 ft. higher than Mees building floor level)

waste

treatment

plant

50' turning

radius

for trucks

34 m

11.8

m

control

dimensions

elev.

exterior

utility area

thermal

ground shield

(3 m high)

mirror box

20

Figure 7.3. ATST site plan.



includes a drainage system and holding tank for stripping fluids, compressed air delivery, and gas

cylinder racks.

Operation: The upper cover of the chamber is lifted clear of the lower half on jacking screws so that the

lower half can be moved on floor rails to the mirror stripping/cleaning area. After stripping, the mirror is

lifted by a crane and held suspended while the mirror cart is moved out of the way and the lower half of

the coating chamber moved under the mirror. The mirror is lowered onto a turntable support frame in the

coating chamber lower half and the chamber is moved back under its cover. During the coating process

the mirror must be positioned central to the axis of rotation of the turntable and be retained in that

position. The front surface of the mirror must remain in a normal plane to the axis of rotation.

Additional Parameters:

The chamber interface to the coolant lines is by manually operated valves; there will be a flow

requirement and outflow temperature requirement.

For the installation of the plant the mobile section of the vessel will use air bearings and drives to

position the vessel within the coating plant chamber. Compressed air couplings normally used to

support the mirror handling cart will provide compressed air to the chamber air skates when the

chamber is installed in each observatory. Handling trolleys will be used for the maintenance and

installation of the magnetron systems.

A rack for 8 gas cylinders will be required for the Argon supply used in coating the mirror.

7.5 HANDLING EQUIPMENT

The most challenging requirement for material handling is transporting the primary mirror from the

telescope to the coating facility and back, which may occur as often as every six months. A platform lift

is provided for that purpose. The mirror in its cell and cart is approximately 5 m in diameter by 2 m high

Control

Room

Computer

Room

DN

Access Balco

ny

Coudé

Platform

WC

utility

closet

DN

UP

utility shaft

C1

C2

C3

C4

C5 C6

C7

C8

C9

C11

C10

C12

C13C14

8'-

2"

[248

9]

cle

artma

tma

enc

future

light feed

north

Advanced Technology

Solar Telescope

at

Haleakala Observatory

S&O BUILDING

COUDÉ LEVEL PLAN

sheet 12 of 20

scale: 1"= 10'-0"

drawn by Jeff Barr Jan 6, 2006

Notes:

For door & opening dimensions

refer to door schedule (sht 20)

For finish materials & ceiling heights

refer to room finish schedule (sht.20)

1050

0

20 ft.

5 m

roof over coating area

Platform Lift

maximum instrument

dimension

2.4

m

5 m

5' x 12' optical bench

when parked

lift serves as additional

lab floor space

8'-0

" window

chase

Mees

Bldg.

vestibule

non

-rotating zone

Instrument

Prep Lab

elevator

Figure 7.4. Coudé Level plan view.



and weighs close to 15 tons. The capacity and dimensions of the lift are based on the size and weight of

the mirror. The design of the lift also accommodates the maximum defined volume and weight for

instruments that will be moved to and from the coudé stations.

For personnel and smaller equipment a standard building elevator is also provided that serves all five

levels of the Lower Enclosure and the two levels of the Support Facility. For disassembling the primary

mirror from its cell and for loading and unloading large instruments and other equipment, a bridge type

crane with ~20 ton capacity is provided in the high-bay receiving and mirror prep area. A smaller

capacity monorail crane is provided in the instrument lab. Appropriate hatches and removable flooring

are designed into the upper levels of the Lower Enclosure to allow the jib cranes on the Enclosure

(described elsewhere) to be used for material handling in that area. Appropriate additional equipment

(portable scissors lifts, fork lifts, special purpose hoists, etc.) will be provided as well.

All handling equipment, especially the lifts and cranes that are integral to the structural design of the

building, will have the highest affordable capacity and be configured for maximum flexibility as future

requirements are difficult to predict.

7.6 REMOTE OPERATIONS BUILDING

To augment the support and operations building, there is an identified need for a facility that would serve

ATST functions that do not require direct proximity to the observatory. This would allow for a smaller

structure and less heat generation adjacent to the telescope. The functional requirements for this facility

would be mostly for administrative offices with some auxiliary lab/shop space geared toward long-term

maintenance/storage of instruments and equipment. A high-speed data connection, for effective

teleconferencing, for real time communication with the observatory, and for transmission of data to home

institutions will be provided.

iii. design of the atst

Documents