|
Документ взят из кэша поисковой машины. Адрес
оригинального документа
: http://www.apo.nmsu.edu/Telescopes/eng.papers/milt/enclosure.html
Дата изменения: Mon Dec 13 18:51:08 1993 Дата индексирования: Sun Apr 10 00:30:42 2016 Кодировка: |
The wind speed at Sacramento Peak can be very high. The load on the enclosure corresponding to the highest velocity expected wind gust is the largest single load which the design must withstand.
Figure 5.1 is a histogram of the wind gust frequency vs. velocity. The data for this graph come from weather records kept by Sacramento Peak Observatory and from DeMastus (1976). The straight line is the envelope to the data and provides a means of predicting the maximum wind gust in 100 years. The wind speed adopted for design purposes is 60 m/s (118 knots). The maximum recorded gust 1954-1974 was 50 m/s (97 knots) in 1960.
Sacramento Peak gets much more snow than the major optical observatories in Arizona. The maximum accumulation depth of snow on the ground 1954-1974 was 47 inches in 1960 (DeMastus 1976). In one storm in December 1982 24 inches of snow fell. In normal operation we expect to remove snow as it falls by heating the enclosure roof; the design load for the enclosure corresponds to 1.2 meter of snow however. This is justified because wind loading still dominates the design of the structure and it is risky to depend on an active snow removal system for enclosure survival.
Lightning protection for the enclosure and electronics is required. Lightning apparently is not as serious a problem as at KPNO and MMTO, based on the experience of SPO. Effective lightning protection is not expensive however and the consequences of neglecting the problem are serious.
The enclosure must support a 5 ton bridge crane which will be used to handle the primary mirror, the secondary assembly, and instruments.
The enclosure must provide an effective barrier to wind, dust, and precipitation.
The roof, upper surfaces of shutter panels, and shutter tracks will be heated during snow storms in order to remove snow.
The enclosure will rotate 540 degrees with 110 degrees at the midpoint. Maximum rotation rate will be 9 degrees/second or 1.5 rpm. The acceleration time will be 3 seconds.
The record high and low temperatures observed at SPO for the 1954-1974 time period were 34 C (93 F) and -31 C (-23 F) respectively. The next most extreme temperatures were 33 C (91 F) and -24 C (-12 F). The enclosure will be designed to operate between 30 and -20 degrees. This will cause negligible loss of time due to temperature extremes.
The shutters will close in less than 30 seconds.
The enclosure will produce minimal local degradation of the seeing.
Ample clearances between the telescope and the enclosure will be provided. The crane will be mounted high enough so that the secondary mirror assembly and the primary mirror can be lifted clear of the telescope tube. A path for the primary mirror will be provided to be used while the mirror is being transported between the telescope and the aluminizing chamber. Hatches and doors will be sized to accomodate the primary.
There will be no occultation of telescope optics above 10 degrees elevation. The opening will allow complete clearance of the optical beam between 10 and 90 degrees elevation.
Flooring will be durable to rubber wheeled carts and to foot traffic and will require low maintenance. The floor will have a design load of 5000 N/m^2 (100 lb/ft^2). Special load bearing points and their loads are TBD. They will be used as staging stations during assembly and telescope reconfiguration.
The enclosure is rectangular with a nearly square base. The shutters are two L-shaped one piece panels which part laterally. Each is supported at its four corners by wheels which ride on tracks mounted just in front of the slit opening and at the rear of the roof.
The enclosure rotates via four wheels which are mounted under the four corners of the floor and which ride on a circular track mounted on the enclosure base.
The roof slopes toward the enclosure rear so that water (and cold air) drains away from the slit opening. This makes the weatherproofing problem more tractable and provides moderate wind shielding for the telescope.
The enclosure base (Figures 5.4 and 5.5) raises the enclosure so that the telescope can be mounted above the mean height of the turbulent boundary layer (TBL) of the atmosphere at night. We are relying on microthermal measurements at the site to determine this parameter.
The telescope pier is concentric with the enclosure base and is mounted on afoundation separate from the that of the enclosure base. The telescope and telescope pier are mechanically isolated from the enclosure and enclosure base so that wind loads on the enclosure are coupled to the telescope only through the earth.
Our design for the telescope enclosure owes much to the MMT and to current well-developed pre-engineered metal building technology. This technology is widely disseminated and is used for warehouses, offices, equipment shelters, etc. as well as the familiar farm buildings. It is inexpensive and mass efficient and results in a weather tight structure which can be made to withstand mountain top wind loads.
The enclosure is rectangular with a nearly square base. The overall dimensions are 12 X 13 X 10 meters (40 X 42 X 34 feet). Figures 5.6 and 5.7 show the enclosure floor framework and the utilization of the enclosure floor space. A large portion of the floor is taken up by the telescope itself and the Nasmyth platforms. Since the enclosure corotates with the telescope and since the mount is alt-az, the floor can be quite near the altitude axis; the main constraint is that it not obstruct the tube while observing near the horizon. A high floor minimizes the moving mass of the enclosure, the wind loading on the enclosure, and makes the telescope and instruments very accessible for maintenance and instrument changes.
The primary mirror hatch is located near the shutter opening. The bridge crane will be used to remove the hatch cover and to set it to the side.
One of the rear corners will contain a small work area with tool boxes and awork bench. This area will be used for minor repairs, debugging and for storage of instrument changing tools. This space will be unheated consistent with our philosophy of keeping heat sources away from the telescope. Major work will be done in the shops in the main laboratory.
No elevator is included in the enclosure design. Instead a 1 ton fixed hoist with variable speed capability will located above the equipment hatch and will be used to transfer equipment between the ground and the enclosure. The equipment hatch will be opened manually.
A catwalk (not shown) will extend around the outside of the enclosure below the level of the shutters. Part of its function will be to protect the enclosure track and to complete the seal between the enclosure and the enclosure base.
Figure 5.8 is a finite element model of half the enclosure minus the floor. The main members are steel wide flange I beams, the diagonal members are tie rods, and the purlins are standard cold rolled shapes used commonly in pre-engineered metal building construction (Kirby Building Systems, Houston Texas is an example of a manufacturer of such structures). Additional purlins will be required in the actual structure. The mass of the half model is 5700 kg.
Figure 5.9 is a view of the same structure showing the deflections due to the survival wind load applied in the X direction. The deflections of the members are greatly exaggerated for clarity. The largest deflection in this picture is 33 mm. The other wind loading and snow loading cases produce smaller deflections. A deflection criterion of 1/240 of member lengths was used to determine member sizes. This is the standard architecture criterion for buildings without interior plaster finish.
The enclosure floor structure (Figure 5.6) consists of steel wide flange I beams. The wheels which support the enclosure will be mounted at the center of the short diagonal members near the four corners of the floor. A detailed analysis of this structure has not been performed yet; its mass is estimated to be 9000 kg based on the analysis of a similar structure. The mass of floor joists is not included in this number. The enclosure will be covered with conventional steel roofing and siding panels and will be insulated. The framing will be insulated on all sides also. The interior finish has not yet been decided but will be light colored and easy to clean. Little IR emissivity control will be needed; during observations most of the roof will be covered by the shutters. Emissivity control for the shutters is discussed below.
The floor will be supported by steel floor joists which in turn will be supported by the floor frame. The flooring will be 1/2 inch plywood. Antistatic carpet will cover the plywood flooring. The roofing, siding and flooring and their supports account for 60% of the mass of the enclosure. The values used to estimate the enclosure mass are given below.
Roof including purlins 40 kg/m^2 8 lbs/ft^2 Walls including purlins 30 kg/m^2 6 lbs/ft^2 Floor including joists 74 kg/m^2 15 lbs/ft^2
The enclosure has a mass of less than 60000 Kg. Moving the telescope support functions out of the enclosure has helped keep this value down. The importance of minimizing the enclosure mass is discussed in section 5-8.
The roof, upper surfaces of shutter panels, and shutter tracks will be heated during snow storms in order to remove snow. Based on the experience at the MMT, 110 watts/m^2 (10 watts/foot^2) is adequate to melt snow as it falls. Once the snow has built up on the roof it is much harder to melt it.
The lightning protection system for the enclosure will be based on a low resistance mountain ground established by placing conductors in holes drilled for soil testing and in the site drain field. Lightning arresters will be mounted on the perimeter of the roof and connected to the mountain ground via external conductors. The connection between the moving and stationary portions of the structure will be made by brushes to prevent high currents from damaging the wheels, track and bearings of the enclosure rotation mechanism. Power entering the enclosure base will be protected by zener diodes and gas tubes. All cables between the enclosure and the laboratory will be fiber optics or opto-isolated at each end.
Safety rails will be provided for the primary mirror hatch. Warning lights will be installed below the hatch covers of both the primary mirror hatch and the equipment hatch. The lights will be permanently on and will have very long lifetimes since they will be low wattage lamps operating at reduced voltage. In normal operation the light from these lamps will be baffled by the hatch covers and kept out of the enclosure. If a hatch cover is not in place the warning lights will indicate this condition to personnel working in the dark.
The proximity of the telescope and its instruments to the floor is an important safety feature of this design. The altitude axis will be about 2 meters above the floor. We expect that most maintenance can be done by personnel standing below the level of the altitude axis thus enormously reducing the risk of a serious fall in comparison to a traditional large telescope. We expect an increase in the productivity of personnel as a result of this feature as well.
The enclosure will be rotated in a manner similar to the MMT. It is supported by four wheel mounted near the corners of the floor which ride on acircular track supported in turn by the enclosure base. Two diagonally opposite wheels will be driven by DC motors coupled via a harmonic gear reducer. An error signal provided by a linear differential variable transformer (LVDT) mounted between the telescope and the enclosure will control the building via a simple nested position and velocity control system. Commercial motor controllers will supply power to drive the motors. DC offsets will be added to the error signal before it goes to the controllers so that the motors keep backlash out of the gear trains. In this way the enclosure rotation is slaved to the telescope azimuth rotation though there is no mechanical connection.
The motors are approximately 20 hp each. Based on experience at the MMT we expect the friction of our rotation system to be 0.003. Thus the force required to rotate the building is 2000 N. This value is dominated by the force necessary to accelerate the building. The design ramp up time is 3 seconds and the force required is 12000 N. Wind loading is expected to be small under operation conditions due to the symmetry of the enclosure.
The peak acceleration near the walls of the enclosure will be roughly 0.05 g. Since this could cause a person who was not expecting it to have an accident the acceleration function will be smooth and will not contain step discontinuities.
The support wheels will be steel. The design of the radial definition system has not been selected. It will probably consist of separate wheels rolling on either the inside of the main track or on a separate track similar to the MMT.
If the building is obstructed and prevented from rotating the drive system will limit the current supplied to the motors to a safe value.
The enclosure is not heavy enough to prevent the design wind gust from blowing it off the track due to the mass efficiency of the design. For this reason combination wind and earthquake clips provide a noncontacting but positive attachment of the enclosure to the track. Under operating conditions no contact will occur. In addition in the parking position of the enclosure locking clamps will prevent any vertical motion or enclosure rotation.
Limits to rotation are necessary to avoid stressing or damaging the cabling between the enclosure and the nonrotating base. Limits will be implemented with limit switches which will cut drive motor power and apply braking to provide soft stops. The ultimate limit will be mechanical shock absorbing bumpers which will not sustain damage as a result of a crash. The kinetic energy of the rotating building at its maximum rotation rate is 30000 Joules. The kinetic energy of the MMT building at its 1.5 deg/sec maximum rotation rate is estimated to be about half this value suggesting that what we are proposing is not completely outrageous.
The slit width is set by the primary mirror diameter plus allowances for field of view, diffraction at millimeter wavelengths, fabrication tolerances and enclosure control system error.
Primary 3.5m Field size 0.096 Diffraction TBD Fab tolerance 0.10 Control error 0.05 Total 3.75m
The slit width adopted for design purposes is 4.0 m.
The shutter panels are 2.7 m (9 feet) wide. This width includes overlap for labyrinth type weather seals. When open they do not protrude beyond the sides of the building or the edges of the slit. This reduces much of the wind loading on the shutters and operating mechanism while the shutters are open. The shutters extend to the rear of the roof. The rear shutter track is mounted behind the rear wall of the enclosure. In this way the shutter track does not interfere with roof drainage or disrupt the integrity of the roof. Moving the track closer to the slit rear would reduce panel length by less than 2 meters. The shutters have a mass of about 4000 kg each.
The shutter framework is steel I beams (W16x26) with diagonal cross bracing (members have not yet been decided). The outer covering will be steel roofing covered with aluminum foil tape. The structural members near the slit (when the shutter is opened) will be insulated but there will be no general insulation of structural members since they will be "outside" during operation. The structural members will be insulated from the roofing.
The shutter roof panels will be heated in a manner similar to the stationary portion of the enclosure roof. Power is supplied to each shutter panel via a cable running diagonally from the lower corner of the slit opening to the panel midline at the front of the roof panel.
The panels roll on wheels mounted on the extreme four corners of each shutter. The wheels are V-groove and roll on a track which is steel angle mounted with the vertex at the top. The track will be heated during snow storms.
The weight of the panels is inadequate to insure that they will not be blown off during the design wind gust. Clips mounted on each shutter near the wheels will engage a slot which is part of the track assembly and will provide a positive attachment of the shutter to the enclosure. In normal operation there will be clearances so that no contact occurs.
Shutter actuation will be similar to the MMT system; a gearhead motor drives cable rigging via a chain and sprocket moving both shutters simultaneously. We expect the friction of the rollers on the track to be 1% or less. The actuating force required to overcome friction is 200 N. Lateral wind loads at the maximum shutter operating wind speed (35 m/s or about 70 mph) are 3000 N assuming an effective lateral cross-section of 26 square meters. Wind loading clearly dominates the shutter actuator design. The shutters will normally be closed when the wind speed reaches 15 m/s, the telescope operation maximum. It is important that the shutter mechanism be reliable since closing the shutters is by far the most graceful way of protecting the telescope from a sudden rain storm. The system described above satisfies this requirement assuming an appropriate preventive maintenance schedule. The system requires a torque limiting clutch in the drive train to prevent catastrophic damage in the case of an obstruction or limit switch failure. The fall back position in the case of failure of the system is to manually close the shutters using a block and tackle, for example. Power failure would be handled by starting a backup power generator. The power requirement is estimated to be 1500 watts per shutter assuming an efficiency of 0.2 and a closing speed of 0.1 m/s.
The shutters will have a labyrinth type non-contacting rain seal. Neoprene Pcross-section gaskets which have been used successfully at the MMT will provide air seals at interfaces where the shutter motion is perpendicular to the enclosure. A sliding type air seal will be made at the front and rear of each shutter. The requirements for the sliding type seal are similar to those for the building rotation seal and will be implemented in the same way.
The final component in the enclosure sealing system consists of positive pressurizing this space with filtered air when not observing. This will insure a positive pressure gradient from the inside to the outside as long as the dynamic pressure of the wind against the enclosure walls is less than the pressure maintained inside. It should be possible to maintain this condition during all but the strongest winds. The cleanliness of the enclosure will approximate a laboratory environment except during these rare occurances.
A separate track may be required to define the center of rotation and to provide radial constraints for the enclosure. This system is yet to be determined. Removable panels attached to the underside of the enclosure floor will separate the track space from the rest of the enclosure base interior. The track space will be positive pressurized with clean air in order to preserve cleanliness.
Sealing between the enclosure and the enclosure base will be accomplished in three stages. First a skirt attached to the under side of the enclosure and concentric with the enclosure base will extend well below the top of the enclosure base providing a non-contacting weather baffle. Secondly a contacting dry seal will provide a physical barrier to dust, air and insects. Such a seal is expected to require less maintenance than the MMT type liquid seal. Thirdly the track space interior to this seal will be positive pressurized with clean air as described above.
The base will support the enclosure track and will limit the deflections of the track due to the applied loads to be within the slope specification for the track.
Within the enclosure base one level below the enclosure floor is an intermediate stationary floor (Figure 5.10). This floor is a convenience during enclosure track installation, provides access to the drive system for maintenance, and provides part of the interface between the rotating and stationary portions of the structure. A fixed staircase provides access to this level near its inner radius. A stairway attached to the underside of the enclosure floor provides access from the intermediate level near its periphery to the enclosure. The radial separation of the two staircases prevents interference during rotation. The wall on the moving stairway which faces the center of the building is designed to gently push obstacles and people out of the way.
Also shown is the primary mirror hatch which is opened using the bridge crane. The equipment hatch is opened manually. Both hatches are interlocked electrically with the enclosure rotation mechanism to prevent collisions. In addition the equipment hatch will activate a mechanical interlock preventing enclosure rotation when opened.
A door sized to accomodate the primary mirror will be provided at the ground level. It will be located beneath the intermediate level primary mirror hatch.
The pier itself is protected from wind loading by the enclosure base; the telescope is only partially shielded from wind by the enclosure. Wind loading on the telescope causes deformation of the pier and the pier foundation. The deformation of the pier causes negligable pointing errors. An analysis of the effect of the deformation of the ground will be done when soil test results become available.
Cool air is produced by contact with surfaces which are cooling radiatively. It is important that surfaces exposed to the night sky have low emissivity at 10 microns; surfaces which are black in the IR cool as much as 10 degrees C.
Sources of heat include structures with large thermal inertia, heated spaces, and electrical dissipation in motors and electronics. The enclosure is constructed primarily of steel and has a mass of about 60000 kg. A cooling rate of one degree C per hour would release 6600 watts. This can be minimized by insulating the major structural members thus greatly reducing the thermal coupling of the structure to the air.
A wall insulated to R-21 (150 mm of fiberglass insulation, dry wall, and air boundary layers) conducts heat at a rate of 1 watt per square foot from aroom heated to 20 C to a telescope chamber at -20 C. An allowance for doors and windows could easily double this figure. We propose to avoid this problem entirely by heating none of the space in the telescope enclosure or base.
The enclosure drive motors and electronics are a localized source of heat as opposed to the general sources discussed above. The motors and electronics will be enclosed and the surrounding air will be withdrawn from the enclosure and vented down wind from the telescope.
It is rated at 5 tons and will have sufficient range to move equipment between the ground and the enclosure. It will have a speed between 0.75 and 3.5 cm/sec.
The crane trolley plus its rated load weighs approximately 60000 N and must be treated as a live load in the design of the enclosure since it will be able to move essentially over the entire area of the enclosure floor. Horizontal loads on the enclosure due to the swinging of a large load being lifted by the crane are small compared to wind loading. Also self resonant effects in this situation do not exist since the lowest natural frequency of the enclosure will be above 5 Hz.
This procedure is reversed in order to re-install the primary after re-coating.