*Osservatorio Astrofisico di Arcetri ** Steward Observatory - University of Arizona
Proceedings of SPIE conference on Optical Telescopes of Today and Tomorrow, 2871, 1996
Abstract
1. Interfaces with the main telescope structure
2. Finite Element Analyses of the mirror cell structure
3. The mirror cell geometry and instrument interface
4. M1 mirror support system: pneumatic force actuators and positioning
5. M1 mirror thermal control concept and system plant
6. Details on aluminizing requirements and maintenance
7. Conclusion
8 . References
Abstract
keywords: telescope, primary mirror, actuator support design, thermal control.
1. Interfaces with the main telescope structure
The Mirror Cells are connected to the telescope [1] by the main elevation structures having a C-ring shape and
by some lateral trusses to guarantee a stiff lateral supporting as depicted in FIG.1 and FIG.2 . Inside the Mirror
Cell are located two longitudinal stiff beams to realize the basic support of the mirror cell to the telescope structure
(see FIG. 3): they are directly and stiffly connected to the telescope elevation structure either on the front side or
in the back one. Perpendicularly to them other two large beams, and some other smaller, realize the supporting system of the mirror cell floor and the external
cylindrical surfaces. The high stiffness required for the Mirror Cell is basically due to two reasons: first we want
to support the primary mirrors of the LBT by a structure as stiffer as possible to increase their natural frequencies
according to the interferometer requirements of the telescope; then, we need a mirror cell capable to support the
mechanical stresses due to the vacuum load during the aluminizing process.
2. Finite Element Analyses of the mirror cell structure
The LBT structure has been analyzed by several numerical simulation studies carried out by finite element analysis
in some different boundary conditions [2]; to evaluate separately the influence of the Mirror Cell structure some
analyses have been carried out considering it infinitely stiff connected to the telescope frame: a modal analysis,
a static analysis under its dead load, with the telescope horizon and zenith pointing, and a static analysis with the
external pressure load have been the main analyses performed.
In TAB.1 the modal analysis results have been summarized and in FIG.4 the first mode has been depicted. As
shown the first eigenfrequency is relatively high, compared with 9 Hz of the telescope structure, and it seems
sufficient to avoid any influence on the mirror structure. From a static point of view we have few tenth of
millimeter in deformation in both positions (see FIG. 5 and 6), closed to the external diameter, without any
influence on the position of the M1 mirror.
| Mode | Frequency [Hz] |
| 1 | 18.04 |
| 2 | 19.63 |
| 3 | 27.41 |
| Total mass of the Mirror Cell : ~ 28 tons |
3. The mirror cell geometry and instrument interface
The main mirror cell dimension are depicted in FIG.7 and summarized in TAB.2 . Its total weight, without the M1
mirror and others auxiliary devices, is approximately 28 tons including the bearing system and the cable wrap for
the Cassegrain instrumentation. Due to the derotating movement, necessary on the instrument interface, we need
to locate a large bearing system, with four points contact balls on the inner 2800 mm diameter, in the central
volume of the mirror cell structure. A very high stiffness of this bearing is required either axially or radially with a
demanded radial runout less than 0.055 mm.
| external flange diameter | 8880 mm |
| internal diameter | 8740 mm |
| central hole | 1000 mm |
| external height | 2525 mm |
| total volume | ~ 141 m3 |
| weight of the structure only | ~28 tons |
4. M1 mirror support system: pneumatic force actuators and positioning
| System of Force Actuators: | |
|---|---|
| Single Actuators | 46 | Double Actuators | 110 |
| Cross-lateral Actuators | 4 |
| 160 Total | |
| Directionality: | push only or pull only |
| Nominal Axial Force: | -1250 to +1600 N |
| Nominal Lateral Force: | 0 - 1700 N |
| Measurement Accuracy: | 1 / 1200 |
| Measurement Resolution: | 1 / 4000 |
| - Operating Pressure Range: 500 - 600 Torr | |
| - Primary compressed air by four feeding lines | |
| System of Position Actuators: | |
| Positioning Actuators | 6 ( 3 couples ) |
| Axial static stiffness | > 85,000 N/mm |
| Force Measurement range | -300 N to +300 N |
| " " resolution | 0.5 N |
| Force Breakaway | < 3,000 N |
| - Operating Pressure Range: | 3.5 - 4 bar |
At 2000 mm from the cell bottom a support grid is located: as shown in FIG.8 all the force actuators and the other
thermal devices are placed and fixed on the backside of it. As mentioned above the function of the hardpoints
system is to keep the mirror aligned with the other parts of the telescope optical system without applying any force.
A load cell is provided in every leg to measure the instant load on it and consequently to correct the axial actuators
forces in order to unload the hardpoints. The intrinsic mirror stiffness and the relatively large bandwidth of the
actuators response (several Hz) allows for maintaining the mirror figure within specs under most wind conditions
encountered on Mt.Graham. A detail study has been carried out on the geometric configuration of the hardpoint
legs [4] in order to optimize the support system stiffness and a prototype has been constructed. In our study we
basically considered three parameters: the orientation of the legs (V or inverted V assembled), the angle
between the legs and the lateral angle of the hardpoints 
. In the mirror cell layout we have employed a
geometric configuration with =100 ° and =15 ° in the external direction where 3 points on the backplate of the
mirror (120 ° spaced) are connected to 6 points on the cell floor.
5. M1 mirror thermal control concept and system plant
The M1 borosilicate honeycomb mirror requires a fine temperature conditioning to obtain the required optical
performances. The temperature control system has to achieve a mirror mass temperature stability within 0.1 ° C
when the external temperature changes 0.25 ° C per hour [5]. The temperature control system adopted in the LBT
project is based on static devices, like an ejector, capable to blow control air inside the mirror cell, without any
rotating mechanism.
Because of the mirror cell works as a plenum overpressured, the air flows from the mirror cell volume right away
to the honeycomb mirror cells through some nozzles. To realize this overpressure we employed about 260 air-air
ejectors downward oriented (about one ejector for eight mirror nozzles). About 90% of the total air is then
recirculated and the 10% is controlled to achieve the required air conditions driving the mirror temperature as the
specifications request. Each ejector supplies about 60 g/s by 10,000 Pa difference pressure between the primary
and the secondary inlet flow boundary conditions and with about 80 Pa back pressure in the plenum volume.
Many experimental tests on the ejector have been carried out in the past to check the global response in different
boundary conditions; a more extended discussion on its working performances is in [6]. We can summarize the
main qualities of this temperature control system as following:
Static air conditioning system based on 260 ejectors feeding 1668 single nozzles
| Total required air flow rate for the mirror : | |
| ~ 14,000 l/s : | |
| ~ 12,600 l/s ( 90 % ) ricirculated flow | |
| ~ 1,400 l/s ( 10 % ) on line controlled flow | |
| Main thermal control specifications: | |
| Temp. variations in the glass < 0.1 ° C | |
|
Temp. differences between mirror and external ambient air < 0.15 ° C |
6. Details on aluminizing requirements and maintenance
One of the innovative concepts adopted in the LBT design is to foreseen the washing and aluminizing processes
of the primary mirror in-situ, without remove the M1 mirror off the mirror cell. A vacuum bell-jar is moved to the
board of the telescope, by a trolley, and applied on the front of the mirror cell while the telescope is horizon
pointing. The F.E.A. of the mirror cell in the aluminizing position, under internal vacuum and external atmospheric
pressure submitted, shows that the deformation of the actuator support plane is limited to few millimeters, a value
that can easily be accommodated in the range of travel of the actuators. Before the aluminizing phase the
pneumatic actuators are evacuated and the mirror rests on the soft pads that limit the actuator range of motion.
External seals have to be installed, before aluminizing, to close the cassegrain instrument interface, the cell
access door and the ventilation air inlet. On the external edge of the mirror cell a circumferential seal is present
to connect it to the bell-jar and a vacuum seal system divides the high vacuum volume (~ 10-6 Torr) from the low
vacuum in the mirror cell (~ 10-3 Torr) [7]. An accurate choice of the employed materials in the mirror cell is then
required to avoid the outgassing phenomena during the aluminizing phase.
7. Conclusion
The M1 Mirror Cell drawing has been an exciting exercise of mechanical and system design to take into account
all the aspects involved in this important substructure of the LBT. As show in TAB. 5 and in FIG.8 there are some
crucial components, as the top plate, requiring hi-tech solutions during the manufacturing and assembling
processes. After the design phase we are now request to solve the manufactory and testing problems getting
the best result with the lowest cost.
All the functions requested in the mirror cell specification [8] have been solved by the LBT Project Office and ADS
Italia s.r.l. that we wish to thank for their high quality work.
Actuators: 274 holes  100 mm + 114 elliptic slots ( 640 fixing holes 20 mm ) |
|
Air ejectors: 260 holes 127 mm ( 780 fixing holes M5 ) |
|
Exhausted air sucking pipe: 8 holes 135 mm ( 32 fixing holes M5 ) |
|
Shock Absorbers: 327 holes 25 mm ( 1308 fixing holes M5 ) |
| Air Nozzles: 1668 holes f 36 mm |
| · Total machined slots: 3291 |
| · Total fixing holes M5: 2120 |
