Title: Effects of Frit Expansion on Mirror Optical Surfaces
1Effects of Frit Expansion on Mirror Optical
Surfaces
- Robert Besuner
- 16 September 2004
2Background
- SNAP PM, SM, and TM are currently envisioned to
be closed-back, frit-bonded ULE. - Frit is essentially a bonding agent between face
sheets and core. - Frit forms a fillet between parts.
- Exact properties of frit are proprietary, but it
is known that in air, the frit itself expands
irreversibly. - Magnitude of this expansion is being sought...for
this analysis a uniform bulk expansion of 1ppm on
linear dimensions is applied. - Elastic properties are assumed the same as ULE.
- Expansion of frit imposes strain on core and face
sheets, causing some distortion of optical
surface. - If frit-imposed distortions are too high, may
need to seek alternatives.
3Finite Element Analysis on SNAP Mirror Cells
- One ppm strain imposed on frit only.
- Frit assumed to have same elastic properties as
ULE. - Frit shape assumed a 45 degree fillet, 1mm x 1mm
or 2mm x 2mm. - Both of these conservatively have more frit matl
than Corning predicts/observes, see following
slide. - Visual inspection of parts _at_ ITT (formerly EK)
showed frit size and shape is pretty irregular,
even on a continuous bead. - In no instance was there as much as 2mm of fillet
leg. - Simpler to model 45 degree fillet.
- Assumption is a mirror cell in an infinite array
of cells. - No edge effects.
- Symmetries of hex and isogrid cells are taken
advantage of to simplify models. - Configuration variations
- 1mm or 2mm fillets
- All cells have identical fillets vs. One cell has
finite fillet and adjacent cell has no fillet.
4Frit Profiles
5Primary Mirror Unit Cell
6Secondary Mirror Unit Cell
7Tertiary Mirror Unit Cell
8PM Distortion Result (All cells 1mm x 1mm frit
fillets)
9PM Distortion Result (All cells 1mm x 1mm frit
fillets)
10PM Distortion Result (1mm x 1mm fillet next to no
fillet)
11PM Distortion Result (1mm x 1mm fillet next to no
fillet)
12TM Distortion Result (All cells 1mm x 1mm frit
fillets)
13TM Distortion Result (All cells 1mm x 1mm frit
fillets)
14TM Distortion Result (1mm x 1mm fillet next to no
fillet)
15TM Distortion Result (1mm x 1mm fillet next to no
fillet)
16Tabular Distortion Results
17Conclusion
- It all depends on what the actual expected frit
expansion is. - There are definite distortions at mid-spatial
frequencies. - Distortions will vary linearly with frit
expansion value. - Alternatives
- Low-Temperature Fusion (no frit).
- Bonded Zerodur (new process, no heritage).
- Open-back Zerodur (mass is higher).
- Open-back Silicon Carbide (bulk thermal
expansion).