Answer 1. Following is our public statement about the first two thin shells:
Richard F. Green Director, LBTO February 3, 2006The first two shells are damaged and not suitable for scientific use on the LBT. The first shell was fractured in final processing during production in the Steward Observatory Mirror Lab on 2 May, 2005. That result is attributable to a failure in process control in the use of a tool for finishing the edge of the fully thinned shell. The shell was still attached to the granite blocking body. The cylindrical tool was used to cut down to make a "shelf" in the granite exterior to the thin shell. It was then used to cut inward in radius, with the intention of providing a "clean up cut" to the edge of the shell (as well as cutting the supporting granite). A combination of factors apparently led to regions of edge fractures. The choice of tool and diamond grit resulted in a somewhat slow removal rate with higher force than typical. The rate of bringing the tool inward in radius may have been too rapid, given the higher force. Finally, the cooling water or force application debonded the edge of the shell from the pitch in spots, leading to vibration and subsequent fracturing. Changes in procedure led to a successful clean cut during production of the second thin shell. For remediation, a propagating radial crack was drill stopped then sawed to relieve any further stress. The fractured perimeter was cut down (also successfully) to leave a constant but undersized radius. The shell was shipped to ADS in Lecco, Italy, and arrived in the same condition as before packing. This shell is serving as a pathfinder for processing and optical testing, before a science-quality shell is delivered.
The second shell was successfully completed and accepted by LBTO from the Mirror Lab. It was transported to the Sunnyside aluminizing facility, where it was cleaned, masked and aluminized. Design and fabrication errors in the shipping container led to an accident during packing in which the shell sustained an impact fracture to the edge. Damage was not noticed during visual inspection. The combination of compression and accelerations during subsequent shipment caused propagation of a radial crack to near the central hole of the shell. The shell is currently in its packing material at ADS, awaiting stress relief and possible "stitching" of the crack. The shell is unlikely to be suitable for optical testing.
The shell is packed for shipping in a specially made transport container. The shell is first placed in a naugahyde and fiberglass bowl which is actually formed on the shell during the production process. Foam blocks are placed on top. These blocks are constrained in a metal frame with triangle base and top, with vertical metal posts. When bolts are tightened, the frame puts the foam into compression to restrain any movement of the shell. The machined depression was cut to plan, but the engineer's drawings erroneously specified the foam to match the exact radius of the shell, requiring centration accurate to a millimeter. The engineer's original intention was to have the top and bottom foam extend several millimeters beyond the actual radius of the shell. The engineer was called in to inspect this situation but felt with proper centering the shell would be safe. In addition, the vertical posts were erroneously fabricated too short. Aluminum spacer blocks were fabricated to go between the upper triangle and the vertical posts, attached with the clamping screws. As the upper frame was being attached, the packing crew noticed that the shell was decentered by several millimeters. When they raised the upper frame with the crane, one of the spacer blocks became detached from its screw and struck the exposed few millimeters of the edge of the shell, producing a fracture.
It is the conclusion of LBTO that there is nothing fundamental in the manufacture, handling, or shipping of thin shells that makes their ultimate deployment unlikely because of cumulative risk of damage. The thin shell for the MMT adaptive secondary system was successfully shipped to Italy, processed, and deployed at the telescope in Arizona without damage. Strict adherence to carefully developed protocols, handling equipment with large safety margins, and work areas appropriately configured and access controlled should lead to successful integration and use of thin shells in the LBT adaptive secondary systems.
Answer 2. We confirm that there are two adaptive secondary units in LBT. However, this RFP is for a single thin shell. The Steward Observatory Mirror Lab is already producing replacements for the two shells that were broken (see previous question). The LBT Board of Directors felt that it was prudent to explore a second source of thin shells, given the schedule pressure to commission the telescope and their concern about the risk to that schedule inherent in this innovative approach to adaptive optics. We have no objection to your proposal including an option for a second shell, but our expectation is that the contract issued as a result of this RFP will be for a single shell.
Answer 3. After the thin shell is delivered, the next operation is to glue the 672 magnets to the back surface. This operation changes the nature of the handling of the shell. We already have the equipment in place for the magnet gluing and the subsequent handling of the shell when it is installed into the adaptive secondary unit.
So the direct answer to the question is: Any handling tools needed for the direct packing and unpacking of the shell into and out of the shipping container should be included with the delivery of the shipping container and shell. Handling tools needed for earlier steps in the polishing, thinning and coating of the shell are not required as deliverables.
Answer 4. We confirm that the optical concave surface is to be delivered uncoated. The reason is that it is quite useful to be able to see through the front surface while the 672 magnets are being glued onto the back surface. The optical surface will be coated later before the adaptive secondary unit is deployed at the telescope.
5b) Our main concern about the polishing is about the section "POLISHING SPECIFICATION" of the document "POLISHING SPECIFICATION for LBT672 F/15 adaptive secondaries":
5c) We did not find any specification concerning the low frequencies (except tolerance on radius and conci constant) so errors for all first 91 ZERNIKE. Could you send us these values ?
Answer 5.
5a) You are free to propose whichever type of optical fabrication method you prefer for the production of the shell. We expect that your proposal will demonstrate the superiority of the method you have selected.
5b)The point of notes 1 and 2 is to remove the low frequency errors which are easily removeable by bending the shell on its adaptive support. It is the high frequency errors left in the shell that influence the delivered wavefront in the telescope. The requirement for high frequency errors is specified by notes 1 and 2 and by related specifications on micro-roughness.
5c)The limits on the low frequency errors to be removed or ignored are set by the amplitude of the forces required to bend the thin shell. You may subtract off low frequency errors, but only up to the point where the actuator forces to bend out these errors reach the limits described in the polishing specification section. These forces vary significantly depending on which low-order bending modes of the shell are involved in the correction. It is our intention to supply a computer program to the successful bidder which will calculate the forces and surface residuals for bending a particular shape error out of the shell. As an example, a surface error of a few microns peak-valley with a reasonably low order content can be corrected effectively with the actuator geometry while respecting the force limits. Note that it is important to get both the front surface and the thickness correct -- the force limits would allow correction of tens of microns of astigmatism, but this much bending would risk a mismatch on the rear surface.
Answer 6. "unstressed spherical rear surface" means the state of the shell's spherical rear surface when the front surface matches the "POLISHING SPECIFICATION" including any bending forces required to remove low-order errors.