Message

2) Not surprisingly, in hindsight, no one I have approached at Lockheed Martin or NASA wants to talk about Space Shuttle external tank (ET) insulation! I guess these people are so tired of reporters and other groups pinging them about the Shuttle disaster that it easier to simply talk to no one rather than assess whether someone (me) is on a legitimate quest.

3) North Carolina Foam Industries (NCFI) is the supplier to Lockheed Martin for the ET insulation. They have a confidentiality clause in their contract that also prevents them from discussing these methods. Tough sledding so far.

4) The NCFI fellow, however, threw me a major cautionary bone. NCFI produces an array of closed cell (important) urethane foams for doing roof jobs. This is apparently big business. They also supply to boat manufacturers that use these products to spray the foam in hulls to serve as both structure and flotation. NCFI had an instance, when the product components were not properly proportioned, in which the foam bonded to the boat hull (normal), contracted more the normal, and collapsed a 30-foot F/G hull inwards! When he saw the insulation supports on the Manufacturing website, he immediately recalled the event as similar in configuration. He feels that the CTE of closed cell urethane foam is so bad (he had no numbers) that a bonded system would fail our fiberglass supports. I guess that spherical or cylindrical tanks can take the inward forces as compressive hoop or membrane stresses.

5) Closed cell foam (previous entry) has one particularly desirable feature. It cannot load up with frozen water vapor which might serve as a thermally-conductive path in the insulation. I have found Dow's Trymex 2000 to be a QC-traceable rigid closed cell polyisocyanurate product that seems OK for our use. We actually have experience with it as the successful cold trap elbow on the planar racetrack cryostat.

6) You certainly recall our desire to use Inspec Solimide polyimide open cell foam to insulate between the cryostat nozzles and port extensions and ducts. Ignoring our side discussion about long-term degradation of Solimide, the beauty of this product is that it remains flexible (and resilient?) at 77K. The relative (or complete) lack of shrink is due in part to the open cell nature of this product. The expected downside is the need for GN2 bleeds into these spaces to prevent the ingress of atmopheric constituents (especially water and oxygen).

7) The rigid Trymex 2000 is sold in large "buns" measuring slightly better than 2'x4'x8'. The buns are purchased by Dow-authorized "fabricators" (typically mom and pop operations) that cut-to-order. As a pricing benchmarch, 6 each of 1-inch x 4'x8' panel will cost $136. This is cost-competitive with the spun-glass duct board (open cell-like) we used on the planar racetrack cryostat. Very tangential: The Solimide is sold in similar buns to Inspec-authorized fabricators.

8) Voids within the tubs are to be avoided at all reasonable costs. These can result in convective cells than may transfer heat vigorously. As the bonded insulation approach may have a difficult CTE to contend with (above), it seems that a system of rigid board wth semi-resilient spun glass (duct board) compliant elements may offer a best path forward. A target composition for a cryostat panel sub-assembly might be 5-10% open cell-like compliant insulation with the balance as high CTE rigid insulation. GN2 bleed bleed gas will be necessary, but modest.