28th April 2023
This month there was a focus on prototyping of the fold mirror system at LJMU; including testing the gluing process. We also visited our partner (Arup)'s offices in Liverpool where Helen Jermak, NRT project scientist, gave a talk for staff.
Fold mirror system: post CDR steps towards mirror manufacture
After successfully completing the Critical Design Review for the fold mirror system, the team at LJMU got to work on how to develop the procedure for bonding the pads onto the back surface of the mirror.
Planning discussions with the Istituto Nazionale di Astrofisica (INAF; who are responsible for manufacturing the mirror) have now taken place. It is very important to polish and measure the mirror on its supports, so the need to have the mirror support structure ready for the final phase of polishing (planned mid 2024) was agreed.
The three pads on the mirror back surface are made from Invar 36. This is a typical material used to bond metal to mirrors as it has a low Coefficient of Thermal Expansion (CTE). This means it changes size a lot less than typical metals with changes in temperature. The use of invar reduces distortion in the optical surface which, for this mirror, needs to be extremely flat (~16nm RMS).
It is well documented in large telescope and general optics literature that '3M 2216' is a good choice of epoxy for bonding metal to glass for this kind of application, so the NRT design used this well-tested approach documented well by the TMT (Thirty Metre Telescope).
Glue application prototypes and testing
The team applied the glue to the surfaces with syringes to carefully control the quantiy used. Too much glue would impact the optical surface in response to thermal changes and too little would risk joint failure if the bond were not strong enough.
The thickness of the glue layer between pad and mirror is a key characteristic for bond strength. 250um is the optimum thickness for this design. Shims (a thin piece of material used to fill in space between things) are used to keep the pad from touching the glass directly and ensure the thickness is controlled.
Automation and control engineer Adam Garner carefully applies glue to the support surface
The team had small glass samples (float glass) made by a local supplier and used these to provide a cheap but realistic substitute for the actual mirror. The parts could then be tested to destruction.
As part of the prototyping, shim material was used to test the different ways the glue spread when the pad was applied. Aluminum pads were used for the testing as they are much less expensive than invar 36.
Gluing of the mock support structures onto test pieces of glass
A range of methods were tested to find the most repeatable approach. The team then load tested the bond over long periods ranging from days to weeks to verify it would support the weight of the mirror over its lifetime (with a significant safety factor of 2 to 3 which is typical for these kinds of bonds). With a high test load between 5kg and 10kg it is possible perform an accelerated life test in the order of weeks or days which gives a representative test for the expected life of the bond (which should be around 25 years of telescope operation).
Invited talk at Arup
Helen introduces the background science case for the NRT
The Arup Liverpool office host a regular informal event each month with guest speakers on a range of topics and Helen Jermak, NRT Project Scientist, was invited to deliver a talk on the LT and NRT. Arup was selected as a design partner for the enclosure and site, foundations work package early in 2023 and have a history of working with the Liverpool Telescope in the early days of its design and construction.
Helen covered the proud history of LT and the evolving science case that means we need to build a bigger and faster telescope. Helen also presented Arup’s planned involvement in the NRT enclosure planning and design work which will begin soon.
The talk was well attended by Arup employees and generated a range of interesting discussion topics including the use of cloud infrastructure to help futureproof NRT's software capabilities.
