1. Dom mechanics.

2. Earlier design; Reflector-ring mounted to the Pmt before the assembly of the complete Pmt foam-support. The support was made of milled foam. In 2012 Dom in Antares was produced in this way. 30-01-20132 KM3NeT collaboration meeting DOM-mechanics Edward Berbee

3. 30-01-20133 KM3NeT collaboration meeting DOM-mechanics Edward Berbee The used mold (with Pmt and ring) for mounting the Reflector-ring to the Pmt. The Pmt is kept in place together with the aluminum reflector ring, Silgel is poured in and after some hours the set can be removed. Works ok, but takes time, need of a lot of them, if 100 pmt’s a day are to be molded; 100 molds are necessary. PMT pre-assembly

4. 30-01-20134 KM3NeT collaboration meeting DOM-mechanics Edward Berbee -Pre-assembly of the reflector to each PMT is very awkward, a lot of tooling, space and hours are needed. -The PMT’s must each one of them be mounted liquid tight to the foamcore to prevent the Silgel from leaking in to the center of the hemisphere during the pouring in of the gel. -The milled foamcore produces dust and keeps it in it’s surface. Dust is hard to remove completely. -Cheaper compared to SLS, if produced with a mould they could even be much cheaper. -Surface tends to initiate bubbles. -The narrow gap between reflector ring and foam is a source for bubbles. SLS versus foamcore; The foamcore

5. 30-01-20135 KM3NeT collaboration meeting DOM-mechanics Edward Berbee -Air evacuation tubes expel the air surrounding the PMT’s. -Even without degassing after pouring the gel, there where no bubbles (the shell is not porous and there is no source for bubbles between ring and foam) -The freedom of design is extreme, each new core can be designed separately, each core can be adapted to different shapes of PMT’s, special parts can be implemented for guiding of fibers, nano-beacon etc. -Fast and easy assembly possible, all PMT’s mounted is gel in just one step. -Disadvantage; it is more expensive, the latest 4 cores ordered where 300,- euro’s each, this price can reduce drastically in future (150,- euro??) -In future they can be delivered in black, the foam has to be painted (and also made fluid tight). SLS versus foamcore; The SLS-core

6. 30-01-20136 KM3NeT collaboration meeting DOM-mechanics Edward Berbee First Rapid manufactured PMT- core with O-ring grooves dummy lenses and air-evacuation tubing. To reduce assembly hours a PMT- core was designed to use rapid- prototyping as a manufacturing- method (in our case; SLS / selective laser sintering) was used. “Lens” (only a ring for the PPM) PMT O-ring SLS-core

7. 30-01-20137 KM3NeT collaboration meeting DOM-mechanics Edward Berbee Top- and bottom core as ordered for the PPM Lenses with aluminum coating ready for the PPM. costs for acrylic-like lenses with aluminum coating should be evaluated. A pressed aluminum shell glued in likewise could be a cheaper alternative. Top support with nano- led aperture, led sticking out embedded in the gel and pointing up.

8. 30-01-20138 KM3NeT collaboration meeting DOM-mechanics Edward Berbee Calculation for losses in the path have to be done to compare these situations. Transmittance measurements are prepared to check the absorption of the now used Silgel mixture (the Antares mixture, 100 A : 65 B) Reason; as you can see in the next slides there seems to be some inconsistency between different measurements and there are no data for the used mixture. Cross-section with PMMA “lens”, possibly exchangeable with a pressed aluminum reflector ring or maybe even a coated foil. New reflectance tests then are necessary PMMA-lens aluminum coating deposited by vacuum evaporation. Coated “lenses” made for the PPM.

9. 9 Absorption length in cm Wavelength in nm

10. 10 Absorption length in cm Wavelength in nm Antares from publication 03-06-1998, probably 100:100 Silgel 612 acc. Spec > 88% at 400/700 nm (10mm) INFN measurements emanuele leonora in 2010 - 100/40, 50, 60 and 70 GE; measured at Chiba for amanda / icecube, probably including approx. 8% reflectance In 1999 SEMICOSIL 912: 100 A / 40 B used in antares? Semicosil 912 in baikal en nestor (but differend absorption lengths?)

11. 30-01-2013 KM3NeT collaboration meeting DOM-mechanics Edward Berbee 11 -accommodating reduced diameter (0.41 mm / 1000 m dept) of the DOM due to the deep sea pressure to be measured, first test looked very promising, the collar has quite a “springeffect”. Aluminum (only for the test deployments, later on it will be made of titanium) laser- or water-cut belt, the ends connected and tensioned with two bolts. For a clamping force equal to 0.2 bar the force should be approximately 3700 N. For two well greased m6 bolts this means a tightening torque of 1.4 N/m. In titanium the collar could be reduced in cross section by 80 %.

12. 30-01-201312 KM3NeT collaboration meeting DOM-mechanics Edward Berbee Shadowing effects is being looked at, future alternative might be like these pictures. In reality the two parallel strips will diverge. An exploded view of the “boulders”, existing of PE and rubber anti-slip parts. The Dyneema ropes are blocked solid by two titanium pins in the rope.

13. 30-01-201313 KM3NeT collaboration meeting DOM-mechanics Edward Berbee At this moment the VEOC has an extra 10% length compared to the ropes, this will be “stored” in crossings at the “ropeconnectors”. The VEOC is fixed in the middle of these crossings, but is free for rotation at this point. The VEOC, extra length used (left) and not used (right)

14. 30-01-201314 KM3NeT collaboration meeting DOM-mechanics Edward Berbee Bubbles sometimes appear, this seems to be influenced by; -materials (for instance; paint) -Being pressure tested, temperature changes. -Time elapsing. -Air closed-in adjacent to the gel. -Having under-pressure in the DOM. -Air bubbles being present in the uncured gel. To be solved! Bubbles in this case also seem to originate at the edge of the aluminum ring, there was some air closed in between the ring and the foam. A lot of these problems seem to reduce using the SLS core. The next idea is to leave the path of having under-pressure in the DOM, the clamping force can be replaced using the collar.

15. 30-01-201315 KM3NeT collaboration meeting DOM-mechanics Edward Berbee Penetrator / exploded view BOB (break out box) Penetrator Dyneema ropes DOM-rope connection part VEOC

16. 30-01-201316 KM3NeT collaboration meeting DOM-mechanics Edward Berbee -All the parts for the Lom deployment test ready in 3 weeks from now. -All internal (cooling) parts ready for the PPM (small adaption's during assembly) -Ongoing design this year for production. -Parallel design and production of cores and lenses ongoing in Erlangen. Desperately needed; More knowledge on the forming of bubbles in the gel.