Chapter 4

Chapter 4, The Point Design

Section 4.3: Enclosure

NOAO Logo    Gemini Logo

The primary purposes of the enclosure for a telescope include: (1) protecting the telescope from weather and physical damage; (2) reducing the wind loading on the telescope structure during observing; (3) thermal control of the telescope and its surroundings; (4) housing various equipment and functions essential to the operation of the telescope; and (5) providing a suitable environment for personnel working on the telescope and its components.

The feasibility of constructing enclosures large enough for GSMT is not in doubt; larger structures of a similar type have already been built, such as sports arenas with movable roofs. The principal challenges are in two areas: (1) to construct an enclosure that meets all the needs listed above for a price that is lower than scaling laws would indicate; and (2) to sufficiently understand the mitigation of telescope wind loading by the enclosure to be able to design an enclosure that allows the telescope to meet its performance requirements a large fraction of the time under varying conditions of wind speed and direction.

The New Initiatives Office (NIO) is approaching these challenges in several ways: evaluating existing enclosures and extrapolating these designs to the size required for the GSMT; pursuing novel ideas that have been tested for small telescopes; and evaluating the designs proposed by other large telescope projects that could be adapted for use with the GSMT. Later studies will model the ability of enclosures to mitigate the effects of wind-buffeting on the telescope structure and key subsystems.

This section describes the requirements for the enclosure in some detail, and then shows several possible conceptual designs based on scaling existing facilities to fit a 30-m telescope. These are examined in terms of: (1) how the telescope and enclosure interact, (2) the shape of the enclosure, and (3) the structural and mechanical systems of the enclosure. Finally, the next steps required for the development of the enclosure concept are discussed.

4.3.1 PROTECTION FROM WEATHER AND PHYSICAL DAMAGE

Typically, the first requirement of the enclosure is to protect the telescope from weather and physical damage. Most telescopes have numerous components that need to be protected from high humidity, rain, snow and ice. These components include mechanisms such as drives, brakes, actuators, and bearings; electronics of various types, including boards, circuits, electrical cabinets, cabling, computers, and detectors; and the optics and optical coatings.

In addition, the telescope and equipment need to be protected from airborne particulates. Dust particles accumulating on mirror surfaces cause the gradual deterioration of the image quality. Dust in coating facilities can reduce the initial quality of mirror coatings. The interiors of instruments, optical windows, etc., need to be kept free of any dust particles. Dust within bogey bearings, hydrostatic bearing oil, and other pieces of equipment can slowly cause damage.

Minimization of problems with airborne particles requires close attention to the sealing of the enclosure, sealing of exposed concrete, and reducing the potential for dust raised by vehicles on nearby roads and parking areas. For example, the Gemini South facility had an ongoing problem with dust, particularly in the coating chamber area, until the seals in the enclosure walls were improved and the entire surrounding site was covered with a non-dusting gravel. The sealing of the enclosure is a detail that needs attention at the initial design stage. Large openings are particularly hard to seal, as are seals at interfaces such as the shutter-to-dome, and rotating- dome-to-fixed-base. Seals are also important for thermal control, which is discussed below.

Protection from weather is also important during work on the telescope. Having an enclosure of sufficient size for full movement of the telescope within the closed enclosure allows a considerable amount of the commissioning, maintenance, and calibration of the telescope and its instruments to be performed during high winds and inclement weather, which results in better utilization of the telescope. This helps to reduce life cycle costs and loss of observing time, but would likely require a higher construction cost than a minimal-volume, co-rotating enclosure.

4.3.2 REDUCING WIND LOADING ON THE TELESCOPE STRUCTURE

Our preliminary studies indicate the GSMT will need some protection under even moderate wind conditions in order to meet the error budget for the final image. As discussed in Appendix 5.5 A, studies of the Nobeyama Radio Observatory (NRO) 45-m telescope lead to the same conclusion.

Section 5.5 discusses studies already under way to improve the characterization of wind loading of large telescopes. Wind measurements taken at the Gemini South facility1 showed that, on average, the enclosure reduced the wind speed at the primary mirror to only 13% of the external ambient wind speed when the ventilation gates were closed.

However, the need to protect the telescope structure and mirrors from the wind must be balanced with the need to provide flushing of air to minimize seeing problems from thermal differences of the air in the optical path. Thermal control of the telescope environment is discussed below. The design of the overall enclosure, the slit, wind screens, and ventilation gates must account for both the buffeting effect of the wind on the telescope components and the thermal flushing of the environment.

4.3.3 THERMAL CONTROL OF THE TELESCOPE ENVIRONMENT

Thermal differences in the air in the optical path of the telescope result in local seeing effects that degrade the final image.2 Minimizing the thermal differences within the telescope chamber requires that the temperature of the telescope structure, the interior of the enclosure, and any equipment in the telescope chamber be kept very close to the ambient external air temperature.

The telescope chamber will need to be maintained during the day at the temperature expected for the next night's observing. This temperature control can be achieved by a combination of active and passive methods. Active methods include air conditioning the telescope chamber. In order to minimize the operating costs, air conditioning the large volume of the telescope chamber required for the GSMT will require close attention to insulation, seals, and the solar loading of the enclosure.

Forced air ventilation may also be required at night, particularly under low wind conditions. Even with ventilation gates, particular orientations of the enclosure with respect to the outside wind direction may require supplemental ventilation to achieve optimum conditions.

Active methods for temperature control will be a significant cost driver during the operation of the telescope, and close attention will be required during the enclosure design to minimize this cost.

The main passive method for thermal control is the use of ventilation gates. This has proved to be very successful where used. As discussed under 4.3.2 above, the design of the ventilation gates must balance ventilation needs with the effect of wind-buffeting on the telescope structure. The passive ventilation system must be capable of operating under a wide range of wind and temperature conditions to provide the ventilation required without adversely affecting the telescope performance.

The air flushing capabilities for a potential enclosure concept can be studied using computational fluid dynamics (CFD) and wind and water tunnel tests. The CFD procedures are analytical and usually require the use of supercomputers to achieve an acceptable level of resolution of the air flow in and around the enclosure. Wind and water tunnel tests provide excellent visual demonstration of large scale flow within and around the enclosure. However, the scale of the models, combined with the fluid properties of air and water flowing at the velocities used in the tests, limits the level of resolution to larger scale distances. NIO's efforts to characterize wind loading of telescopes through direct measurements and computer modeling are presented in subsection 4.3.5.5.

Other effective passive methods include coating the enclosure with low-absorption, low-emissivity paint or reflective film; insulation; and ventilation of the space between the outer skin of the enclosure and the inner skin facing the telescope chamber. When properly designed, these factors can have a significant impact on the amount of air conditioning required during the day. The low-emissivity coating also helps to prevent sub-cooling of the outer surface below the ambient air temperature caused by heat radiation to the night sky.

4.3.4 SUPPORT FUNCTION REQUIREMENTS

General Support Functions. Operation and maintenance of the 30-m GSMT will require a number of support functions to be housed in the enclosure or adjacent support buildings. Anticipated needs include the following:

Mirror recoating will be a major activity for the GSMT, because the current model for the primary mirror consists of a total of 618 segments. The modest size of the segments requires only a fairly small coating chamber and associated equipment. Multiple coating chambers with associated work areas may be required, however, to maintain an adequate pace for the recoating process. The secondary mirror will be the largest optic requiring recoating, and may govern the size for at least one coating chamber setup.

The space for mechanical equipment should be adequate for the equipment itself and should provide working space to maintain the equipment as well as a nearby location for commonly used spares. Consideration should also be given to future upgrades. These upgrades can arise from new equipment that requires increased capacity of certain systems, or completely different equipment not foreseen in the initial installation may be required.

Other support building functions will depend on such factors as facilities that may already be present on site, the actual systems chosen for use, and the method of operating the facility.

Facility Crane. The construction and operation of the telescope will require the use of a large- capacity crane. Although the enclosure structure was erected using a conventional crawler crane, the Gemini telescopes were erected completely with the use of a combination 50 metric ton / 5 metric ton dome-mounted crane, with no external crane required. All normal operations and maintenance continue to be performed using these cranes, as well as handling carts for specific instruments and operations.

The Multiple Mirror Telescope (MMT) Observatory uses an external crane to install the "bell jar" on the combination mirror cell/coating chamber base. This requires an expensive external crane to be brought to the summit each time the coating chamber is used. With the external crane, weather conditions may delay work until wind conditions are mild enough to open the enclosure and to allow the safe use of the crane. Although the work is performed in this manner, the convenience of a permanent, dome-mounted crane should not be underestimated, and probably can be demonstrated to be the most cost-effective approach.

Because the GSMT will require weekly removal and reinstallation of primary mirror rafts to maintain an adequate recoating cycle, a crane would be required on site almost full-time in any case.

The height of a GSMT-sized enclosure (see Figures 1-10) is excessive for almost all of the portable cranes used in construction. Mounting the crane on the enclosure structure may be the only feasible solution. The load capacity of the enclosure crane can probably be limited to 10 tons or less, if larger cranes are available during construction.

4.3.5 DESIGN LOADINGS FOR THE ENCLOSURE

The design loadings for an enclosure structure are typically defined by at least the following three requirements: (1) survival of extreme events such as high winds, major earthquakes, and heavy snow and ice loadings; (2) full operation for observations under moderate wind conditions; and (3) handling case-specific loadings, such as enclosure mounted cranes, and instrument and mirror handling carts.

4.3.5.1 Wind

Survival conditions. The enclosure should survive the highest wind conditions reasonably expected at the site without damage to the enclosure itself, the telescope, or any of the other equipment and functions housed by the enclosure.

Observing is usually not performed under extreme wind conditions; however, consideration should be given during design to conditions in which power failures, broken mechanisms, etc., could result in open shutters, vent gates, etc., during high wind conditions.

Operating conditions. The enclosure should be able to function in all normal observing and operating modes such as opening and closing shutters, rotating the enclosure, and deploying and stowing wind screens under an arbitrary limiting wind speed established for the observatory.

The magnitude of the wind loading is determined by the site wind characteristics, the physical shape of the enclosure, any projecting structures that will catch the wind, and the porosity of the building. These loadings affect bogie sizes, drive mechanisms, and brake mechanisms, in addition to the structure proper.

Another operational consideration is the transmission of wind-induced vibration from the enclosure to the telescope structure through their foundations. Enclosures are typically designed with foundations that are separated from the foundations of the telescope pier, to minimize the transmission of vibration, including vibration caused by rotation of the enclosure in addition to wind shake. Even so, measurements at existing facilities indicate that a significant amount of wind-induced vibration is transmitted from the enclosure to the telescope. This emphasizes the need to make sure the enclosure design is streamlined to minimize drag.

4.3.5.2 Snow and Ice

Typically enclosures are fully closed during snow and ice storms, and the design is based on the maximum reasonable ice and snow loadings expected for the site. Certain configurations, particularly those with flat roofs, tend to collect snow and ice. Structurally, the loadings from these conditions can be anticipated in the design, with no concern for structural failure. However, these situations can present operational problems and may require hand clearing to remove the accumulations in a timely manner. Enclosure configurations that are self-clearing are desired.

4.3.5.3 Seismic Considerations

Many of the premier telescope sites are located in active seismic zones, so earthquake loadings are a critical part of the design of the enclosure and should be considered in the design of the telescope structure and components. Because of the magnitudes of design earthquake ground motions, the loadings on the enclosure may be a significant factor in the design.

Enclosures typically require a true dynamic analysis and design because the design procedures listed in model building codes are based on assumptions about the structural behavior of the building that do not apply to most telescope enclosures. The simplified design methods permitted by model building codes are based on some level of inelastic action, and the resulting damage is both allowed and expected. Enclosures use mechanisms for rotation and provisions for prevention of sliding and overturning, which can be very expensive to repair or replace should they be damaged.

4.3.5.4 Cranes and Other Internal Loads

Enclosure mounted cranes typically impose large, concentrated loads on the roof beams or arch girders of the enclosure. If the cranes are bridge cranes, the enclosure structure must be able to handle the variation in the load path as the crane is translated. The design of the enclosure bogies or other support mechanisms must either be able to transfer the additional loads from the crane to the supporting structure, or a secondary structure must be included to bypass the bogies in the load path.

Handling carts for heavy loads typically travel on rails or specially reinforced areas on the floor. With the anticipated sizes of instrument components, secondary mirrors, and other components that are proportionally sized for a 30-m telescope, these floor loadings will be correspondingly large, but will probably be smaller than the weight of mirror cell assemblies on existing 8-m telescopes. The supporting structure must be able to handle these loads.

4.3.5.5 Defining the Structural Loads

The dead and live loads arising from the structure self-weight, floor loads, equipment handling carts, etc., are specific to the design of the particular facility. The self-weight of the facility is determined from the structural weight, plus any nonstructural items such as interior partitions, ceilings, flooring, insulation, air conditioning equipment, plumbing, and any other parts of the enclosure that are incorporated into the structure, and will remain there for the entire life of the structure. These loads are determined fairly easily as the design progresses. Equipment handling cart loads can be determined from the weight of the cart, plus the weight of the equipment being handled, plus weights of items such as cooling liquids, cabling, etc., that may not be considered a part of the equipment but may be carried with the cart. These loads also vary with the piece of equipment, and may increase during the life of the facility as new instruments and other pieces of equipment are commissioned.

Uniform floor design live loads are available from several sources, particularly model building codes. Because of the unusual nature of a telescope facility, the anticipated uniform floor live loads should be evaluated closely.

Cranes, lifts, and other pieces of equipment typically have a rated capacity. However, overloading of the equipment is common, and dynamic effects during movement result in equivalent static loads that exceed the rated capacities of the pieces of equipment. For these loads, the structure typically is designed on the basis of a load factor applied to the rated load capacity of the equipment. Most model building codes set minimum values for multipliers to be applied to the rated capacities in determining the design loading.

Wind, snow, and ice loads tend to be very site specific. With the telescope facilities typically placed on mountain tops, the severity of the wind, snow, and ice loadings tends to be much higher than for nearby towns at lower elevations. Most of the established sites, such as Mauna Kea, several sites in Chile, the Canary Islands, and a number of mainland sites in the United States have established design levels, backed up by performance of existing facilities and a log of the weather conditions.

New sites require the gathering of information from various sources such as measurements made by other types of facilities at nearby locations, satellite information, forest service data, and even word-of-mouth information from long-time residents. Some of these data can be obtained during site testing. But while this information may be sufficient to judge the scientific quality of the site, the amount of time may not be sufficient to make an adequate statistical prediction for the design loads.

Figure 1 For full independent movement of the telescope within the enclosure, the interior
diameter should allow for the fully swept motion of the telescope, with some clearance as a
safety measure for personnel and equipment. A spherical enclosure is shown here. Rectangular,
faceted, and other enclosure shapes can be used provided the minimum clearance still allows
the full telescope motion.

4.3.6 ENCLOSURE CONFIGURATIONS

Enclosure configurations can be categorized in several ways, including: (1) how the telescope and enclosure interact, (2) the shape of the enclosure, and (3) the structural and mechanical system of the enclosure. Enclosure configurations are discussed according to these categories in the following.

4.3.6.1 Interaction Between the Telescope and Enclosure

Full independent movement of the telescope inside the enclosure. This approach to the enclosure is represented by the Gemini and Keck enclosures at the 8- and 10-m class. The telescope can move independently of the enclosure in both azimuth and altitude. A significant advantage of this approach is that a number of commissioning and testing functions can be performed with the enclosure fully closed.

In order to have this fully independent motion, the interior diameter of the enclosure must be sufficient for the fully swept motion of the telescope, plus some margin to avoid pinch points, special equipment, and as a simple safety measure for personnel and equipment (see Figure 1).

Co-rotating enclosure and telescope. This approach to the enclosure is represented by the Subaru and MMT enclosures. By restricting the telescope to rotate in azimuth with the enclosure, the size of the enclosure can be significantly reduced. The sides of the enclosure may be brought close to the telescope, to the point of allowing only enough space for the free movement of the telescope in altitude, along with some safety margin for personnel and equipment (see Figure 2).

Figure 2 A co-rotating enclosure may be smaller in size, compared to an enclosure
that



 allows fully swept motion of the telescope. The width of the telescope chamber needs to be only
large



 enough to allow movement of the telescope in elevation along with a minimum clearance for safety.
The



 telescope chamber volume may be minimized if sized for the stowed telescope only, and allowing the
telescope



 to project outside the enclosure at other elevation angles. Approximate dimensions: 50 m
wide



 by 50 m deep by 68.6 m high.

This configuration can result in a significant reduction in the initial cost of the enclosure, plus some reduction in operating costs from having a reduced mass of enclosure to place into and maintain in motion, and a reduced volume of enclosed space to air condition for active control of temperatures.

Two major limitations are: (1) both the enclosure and telescope drives and control systems must be operational for rotation of either about the azimuth axis; and (2) working space within the enclosure is more limited.

The size of the building can be reduced even further if the slit must be open in order to allow movement in altitude as well (see Figure 2). This configuration would expose the telescope top end to increased wind-buffeting, however, which would be undesirable under many wind conditions and telescope orientations with respect to the wind. This option further has the risk that the telescope cannot be protected in case of a failure in the telescope altitude motion when the telescope is not in the stowed position. It also limits the commissioning and maintenance work that can be performed on the telescope without full exposure of the telescope and personnel to the weather. For these reasons, we don't believe this configuration is satisfactory for GSMT.

In order to provide protection of the telescope top end, the size of the enclosure could be increased so that the top end is contained within the enclosure space for all orientations of the telescope, as shown in Figure 3. The configuration as shown will be subjected to torques with wind loading from the side. If the back side of the enclosure is extended to balance the wind forces, the size of the enclosure becomes comparable to, or exceeds, that of a spherical enclosure.

Figure 3  The co-rotating enclosure can be increased in one dimension to
completely



 contain the telescope top end within the enclosure space. The wind loading for
this



 configuration, however, is eccentric to the axis of rotation of the enclosure, requiring
continuous



 torques or braking to hold the enclosure in position.

Roll-away and Retractable Enclosures. The roll-away enclosure has been commonly used by amateurs for the protection of their telescopes and is in use at some professional facilities, such as the 2.5-m Sloan Digital Sky Survey Telescope. In this approach, the building is mounted on wheels rolling on rails or restrained by guides and is simply moved away from the telescope, leaving the telescope fully exposed to the weather. This approach has the major advantage of eliminating dome-induced seeing. The major disadvantage is that a failure of the rolling mechanism while open could leave the telescope fully exposed to adverse conditions.

The Starfire Optical Range 3.5-m Telescope and the 3.6-m AEOS Telescope on Haleakala have enclosures that separate into segments and move down below the lowest possible angle for the telescopes.

For the NIO GSMT point design, vibrations induced by wind will be excessive, except at the lowest wind speeds. Having an enclosure to reduce the wind forces acting on the telescope structure is an essential requirement, so these types of enclosures are not being considered for the GSMT point design

Partial or No Enclosure. If all of the components of the telescope that can be damaged by the weather and other physical conditions are adequately protected, the telescope itself can be left exposed to the weather, with no enclosure required. The OWL project is pursuing this approach for their 100-m telescope concept. The telescope structure is exposed to the weather, with covers to be placed over each of the four quadrants of the segmented primary mirror. As was mentioned above for the roll-away and retractable approaches, the NIO GSMT point design will require wind protection to be provided by an enclosure under most wind conditions.

Figure 4 The WIYN enclosure on Kitt Peak is an example of the faceted shape.  (Photo
courtesy of the WIWN Observatory.) 4.3.6.2 Shape of the Enclosure

Spherical Shape. Most of the older enclosures, as well as the recent Keck and Gemini enclosures, have the classical spherical (or "dome") shape. Typically, these enclosures were designed to have full independent movement of the telescope within the enclosure, and the spherical shape provided this capability at the required minimum volume. Figure 1 shows a cross- sectional view of a conceptual spherical enclosure with the GSMT.

Faceted Shape. The WIYN enclosure on Kitt Peak (see Figure 4) and the Magellan enclosures use straight segments and planes instead of the sphere of the conventional dome. This shape has the advantage of being able to use straight steel sections in the structural framing, and flat faces for the mounting of louvers and vents. The fully independent motion of the telescope within the enclosure is still achieved with only a slight amount of extra volume when compared with the spherical enclosure.

Figure 5 The Large Binocular Telescope (LBT) uses a rectangular configuration. As
with the LBT, the rectangular configurations usually are co-rotating with the telescope.
(Aerial photo by Stephen Criswell, courtesy of the LBT Project.)

Rectangular Shape. The rectangular shape was first used for a large telescope with the original MMT, with this concept being carried even further with the Large Binocular Telescope (LBT) (see Figure 5). This shape works well when co-rotating the enclosure with the telescope, and the volume and surface area of the enclosure can be reduced below that of the conventional dome. The reduction in volume and surface area results in lower capital costs for the enclosure, and lower operating costs if air conditioning of the telescope chamber is employed. Additionally, the moving mass of the enclosure is reduced, with less power required to start, maintain, and stop motion of the enclosure.

Cylindrical Shape. The cylindrical shape has been used for the James Clerk Maxwell telescope on Mauna Kea (see Figure 6). This configuration has the advantages of using straight structural members in its construction. It also has the advantage that the track upon which the building rotates matches the shape of the enclosure. This is in contrast to the rectangular shape, which requires special attention to the load path of the structural weight to the bogies for a uniform loading. This configuration typically has had problems with snow and ice accumulating on the roof and top shutter.

Figure 6  The James Clerk Maxwell telescope enclosure on Mauna Kea is an example of the cylindrical
enclosure.  (Photo courtesy of the Joint Astronomy Centre.)

4.3.6.3 Structural and Mechanical Systems of the Enclosure

Arch Girder Construction. The arch girder construction has been a common structural and mechanical system for spherical-shaped enclosure designs, with the Keck and Gemini enclosures as examples for 8-10-m class telescopes. A variation on the arch girder design uses straight segments instead of a continuously curved beam. The twin enclosures for the Magellan telescopes use this concept.

The primary structural elements for the arch girder concept are shown in Figure 7. The two arch girders are located on either side of the slit and are framed onto a ring girder that moves over a fixed base. The enclosure rotates on wheels, either mounted on the ring girder or mounted on the fixed base and supporting the ring girder.

Arches work well when the loads are distributed, such as from the arch girder self-weight, wind loads, and snow and ice loads. Arches offer little advantage over straight beams when subjected to concentrated loads. Depending on the design, concentrated loads may come from dome cranes, sections of the shutter, wind screens, or other sources.

Figure 7 The primary structural elements of the arch girder type construction consist of the
two arch girders on either side of the slit, framing into a ring girder. The ring girder rides
on wheels over a fixed base, allowing the enclosure to rotate. The shell sections provide bracing
to the arch girders and close off the telescope chamber. The arch girders typically have deeper
sections than the shell. As shown here, the arch girders are flush with the interior diameter of
the enclosure and project beyond the shells on the exterior.  Approximate dimensions: 43.4 m
inside radius, springline 15.3 m above fixed base, 6 m high fixed base.

The existence of the slit greatly reduces the structural benefits gained from having a full, complete dome. For many enclosures using the arch girder construction, the true structure is closer to that obtained from two half-shells. As the width of the slit increases in comparison to the diameter of the dome, even that structure advantage is lost.

The arch girders structurally need to be much deeper in section than the adjacent curved shell. If the arch girders project beyond the shell on the exterior (see Figure 7), the arch girders will catch the wind and increase the wind loadings on the enclosure, compared to the corresponding loads on a spherical shape without the projection. These wind loadings will vary as a function of the angle of the enclosure with respect to the wind. If the arch girders project to the interior (see Figure 8), the shell has a larger diameter than necessary, along with the associated costs.

The load paths with arch girders may require a non-uniform arrangement of support wheels beneath the ring girder. The weight of the shutters, crane loads, and the self-weight of the girders go directly through the arch girders to the ring beam. More wheels at a closer spacing are required beneath the arch girder landing points than at points away from the arch girder.

The arch girder configuration has some advantages during construction. With the arch girders erected and shored off, the remaining structural members can be erected off the arch girders, reducing the amount of falsework required.

Figure 8 In contrast to the exterior projection of the arch girders shown in Figure 7, the arch
girders shown here are flush with the exterior and project into the telescope chamber. With this
configuration, the arch girders do not catch the wind, but the shell sections and the telescope
chamber must be larger.  Approximate dimensions:  48.4 m inside radius, springline 15.3 m above
fixed base, 6 m high fixed base.

Rectangular Configuration. From a practical construction point, this configuration uses straight members, allowing common rolled steel shapes to be used instead of specially fabricated, built- up steel shapes. The structure can be framed with regular rolled steel shapes, trusses, or space trusses, which are discussed below. The size of the enclosure required for the GSMT will probably require trusses for at least the major horizontal structural members.

The rectangular configuration has advantages when bridge cranes mounted to the roof structure of the enclosure are desired. The rectangular arrangement may allow bi-parting shutters to be utilized, without having an excessive projection exposed to the wind.

Variations on the rectangular shape can be made to improve air flow in and around the enclosure. For example, the Subaru configuration starts with the basic structural concepts, and then has been greatly modified to maximize the performance with respect to air flushing of the telescope and protection of the telescope from wind shake.

Two-Axis Rotation System. The two-axis rotation configuration consists of a base section formed by a spherical dome rotating on a horizontal plane, with a subsection of the dome rotating on the base section at an angle to the vertical axis (see Figure 9). An opening in the subsection, eccentric to the axis of rotation of the subsection, is used for the optical path. By simultaneously rotating both the base section and the subsection, any area above a minimum angle of about 30 degrees above the horizon can be observed. The important feature of the two-axis rotation design is that no arch girders are required, with the two-axis rotation configuration acting more like a true dome.

Figure 9 The two-axis rotation configuration functions by two independent motions. The overall
enclosure rotates about a vertical axis similar to the arch girder construction. The second, smaller
section rotates on the main section, but about an axis at an angle with the vertical axis. By placing
the opening for the telescope eccentric to the axis of rotation of the smaller section, and
simultaneously rotating both sections, full sky coverage above a minimum horizon angle can be
achieved.  Approximate dimensions: 43.4 m inside radius, springline 15.3 m above fixed base, 6 m
high fixed base.

The elimination of the arch girder with the two-axis rotation configuration eliminates two problems with the arch girder configuration: (1) There are no exterior projections, resulting in reduced wind loadings; and (2) the weight of the dome is more uniformly distributed to the ring girder, with a corresponding more nearly uniform distribution of loads to the wheels.

The primary concerns with the two-axis rotation configuration are: (1) the method for ventilation of the telescope chamber; (2) the complexity of the supporting mechanism for the rotation of the subsection, which is at a significant angle to the horizontal; and (3) the method for sealing the opening for the optical path when the enclosure is closed. Some smaller enclosures of the two- axis rotation configuration have been built and are performing satisfactorily, but no large enclosure using the two-axis rotation configuration has been used to date.

Passive ventilation is possible with the use of ventilation gates on the rotating base section of the enclosure, and perhaps on the subsection. Forced air ventilation may also be required. Although there has been significant experience with arch girder arrangements in terms of passive ventilation, the two-axis rotation configuration will require significant analytical studies and wind or water tunnel tests to establish effective methods for suitable ventilation.

Space Frame/Space Truss Construction. Space frame and space truss construction may offer a cost breakthrough for the enclosure structure. These structures are three-dimensional trusses constructed with relatively short, lightweight individual members. With geodesic domes, which are specific types of space frame construction, the structure is lighter in weight from both the use of the space trusses and from the overall structural behavior as dome and arch shapes. Reduced construction costs may be possible, as lighter structural members can be erected with a smaller construction crane, and the erection can proceed without the need for elaborate falsework.

Figure 10 The Hobby-Eberly Telescope enclosure is an example of the space frame type of construction.
(Photo courtesy of the McDonald Observatory.)

The space truss can be used with most configurations, including the spherical and rectangular shapes. Because the individual structural members used in framing the space trusses typically are fairly small, and the space frame is not well-suited for large concentrated loads, difficulties are encountered in providing the slit opening, the supports for the shutter, and special structure at points where large concentrated loads are imposed, such as crane support points. The Hobby- Eberly Telescope dome is an example of a large space frame enclosure (see Figure 10).

4.3.7 DESIGN CONSIDERATIONS IN CHOOSING AN ENCLOSURE CONFIGURATION

The choice of an enclosure requires the balancing of a number of factors. The enclosed volume is a major factor in capital cost, as the amount of structural steel, roofing, siding, insulation, and number and size of bogeys, for example, are all functions of the size. Simpler mechanisms with less weight are usually preferable to complex mechanisms and heavier masses. If the decision considers operating cost, the power required for active control of temperature of the telescope chamber will depend on the exposed surface area and solar loading, insulation, and passive methods incorporated into the design, in addition to the volume of air being conditioned. Enclosure shapes with linear faces mean that standard milled steel shapes that cost less can be used instead of built-up shapes that are required for the more expensive curved and complex faces.

As an example, Table 1 shows a comparison of the approximate enclosed volumes for representative enclosure configurations, with an approximately 2-m clear distance between the top end of the point design telescope and the interior of the enclosure. This comparison illustrates that the enclosed volume required for fully independent motion of telescope is greater than for a co-rotating, minimum volume enclosure. The volume of air to condition, if active temperature control is used, is approximately 75% greater. However, when one looks at the exposed surface area, which is an indication of the solar loading, and the surface through which heat loads pass, the differences are only 6%.

Table 1 Comparison of enclosure areas and volumes for selected enclosure configurations. The geometries are based on a 2-m clear distance between the telescope structure and the interior of the enclosure.
COMPARISON OF ENCLOSURE AREAS AND VOLUMES
Enclosure ConfigurationEnclosed Volume
(cubic meters)
Surface Area
(square meters)
Dome, fully independent motion of telescope, arch girders toward exterior (Figures 1 and 7)298,00017,000
Dome, fully independent motion of telescope, arch girders toward interior (Figure 8)397,00019,000
Dome, two-axis rotation (Figure 9)298,00017,000
Co-rotating, rectangular, minimum size (Figure 2)172,00016,000
Roll-away, rectangular234,00019,000

The value of lost observing time should not be ignored. The capability to perform maintenance, engineering, commissioning, and calibration procedures on the telescope when the weather is such that the enclosure cannot be opened will result in overall better utilization of the telescope. This drives the enclosure design toward a larger enclosure that allows fully independent motion of the telescope within the enclosure, even though the capital costs considered alone would be lower with a smaller, co-rotating enclosure.

4.3.8 FEASIBILITY STUDY

NIO has consulted with AMEC Dynamic Structures, Ltd., (ADS) to prepare a feasibility study for a classical arch girder enclosure and a two-axis rotation enclosure, each with a size sufficient to allow independent motion of the telescope within the enclosure.  The complete ADS report is attached as appendix 4.3.A .  The complete report includes a first-order cost estimate and estimates for the sizes of the major structural members, the total and moving masses, and power and torque requirements

ADS concluded in their report that either approach is feasible.  The calculated masses for the two approaches are nearly the same: 3,150 metric tons for the arch girder approach and 3,000 tons for the two-axis rotation approach. There are significant differences in the power and torque requirements.

The two approaches have comparable torque and power requirements to drive the entire enclosure about the azimuth axis.  However, there is a factor-of-three difference in the power and torque requirements required to operate the shutters for the classical approach and the two-axis approach to operate the "cap."  While the power and torque requirements are high, further studies and optimization may allow these to be reduced.  Further studies into the rotation and travel acceleration and velocity requirements may lead to a relaxation in requirements, with a corresponding reduction in power and torque requirements.

ADS have noted that, regardless, the power and torque requirements are high, and considerable heat will be generated in the area of the telescope.  Thermal studies will be required to find solutions to mitigate the heat problem.

4.3.9 THE NEXT STEP

NIO will build on the AMEC Dynamic Structures feasibility study for the classical arch girder approach and the two-axis rotation approach for the telescope enclosure, particularly with respect to the ability of those enclosures to (1) mitigate the effects of wind-buffeting on the telescope and (2) provide air flushing sufficient to minimize dome seeing.

Co-rotating enclosures will be studied, particularly to understand the associated construction and operational life cycle costs.

NIO will continue to evaluate other enclosure approaches, including a space frame type enclosure.  The space frame approach has the potential of providing the required enclosure functions at a cost significantly below the more conventional arch girder and rectangular configurations.  Further studies are required to determine if this type of construction can be satisfactorily adapted to the expected enclosure requirements.  In any case, NIO will continue to examine novel approaches  as possible enclosure designs. 

REFERENCES

  1. M. K. Cho, L. Stepp, and S. Kim, "Wind buffeting effects on the Gemini 8m primary mirrors". To appear in Proc. SPIE 4444 (2001).

  2. Lorenzo Zago, "An engineering handbook for local and dome seeing", Proc. SPIE 2871, 726 (1996).


October 2002