CM 2 Helium Circuit Vacuum Leak Repair Tug

CM 2 Helium Circuit Vacuum Leak & Repair Tug Arkan, 6/18/2012 1

Leak Location From the in-situ leak checks that we were able to do at NML cave, the leak is located within the close vicinity of the Ti bellows between the magnet and cavity #8 on the 2 -phase pipe 2

Welds Weld V Weld III Weld I: Ameriflex Weld III: Ameriflex Weld IV: FNAL with the automatic welding machine Weld V: FNAL, manual From the in-situ leak checks, we believe the leak is on the bellows or close proximity to the Weld II and III 3

Probable Causes • Without disassembling, we cannot pinpoint the exact location of the leak: it could be cracked weld or a damaged bellows • Ti Bellows compensating misalignment of the 2 -phase pipes? • Pressure Test? • Cavity to Cavity Alignment Process? • Bad design? 4

Design Ti 2 -phase pipe of a cavity is welded to the Ti helium vessel and has also a sliding support bracket Ti 2 -phase of the magnet is connected with a 2 -3/4 inch Ti Conflat flange to the SS flange of the magnet SS helium vessel. Magnet and its 2 phase pipe are DESY XFEL design. There is not enough space to put another sliding support bracket from the 2 phase pipe to the magnet helium vessel Stresses on the welds during assembling the downstream end Ti CF 100 flange with mating flange? 5

Magnet 2 -phase pipe Leak check records done at CAF during assembly Weld II & III Weld I Fabricated & Leak checked by Ameriflex. (Not pressure tested) After welding to the CM 2 assembly, all the welds were bagged and fine leak checked at CAF-MP 9 6

Assembly & Welding at CAF Weld IV: Automatic welding machine to weld the Ti bellows which was already welded to the magnet 2 -phase pipe to the cavity #8 2 -phase pipe Weld V Weld IV Weld V: Manual weld to weld an adapter tube to the cavity #8 2 -phase pipe. Adapter tube: 3 mm wall thickness for the manual weld and 2 mm wall thickness in order to perform the Weld IV 7

Ti bellows between Cavity #8 and magnet 2 -phase pipe 8

Pressure Tests-I • 2 -phase pipe pressure test was done at CAF-MP 9 during assembly: – A vacuum leak check was done after the 2 -phase pipe was welded with Ti bellows. – A pressure test to 2. 3 bar followed the vacuum leak check. – Post pressure test, another leak check was not done. Ti bellows were clamped for the pressure test in order to restrict the axial forces 9

Pressure Tests-II • Ti bellows were procured from Ameriflex. The vendor provided calculation and certification that the bellows are compatible with the specified pressure ratings. • None of the 2 -phase Ti bellows including the bellows already welded to the magnet 2 -phase pipe were pressure tested at the vendor. • After the assembly of the Ti bellows, it was observed that some of the bellows had to compensate significant misalignment between the 2 -phase pipes of adjacent cavities due to out of specs parallelism of the 2 -phase pipe wrt. helium vessel. 10

Pressure Tests-III • Another pressure test will be done at NML once the module is installed on the test stands. During this pressure test, cavities are held with the invar rod and the ends of the 2 -phase pipe are assemble to the feed and end caps therefore the axial forces are balanced anchored. 11

How big is the leak? • The 2 -phase circuit was pumped down to lower 10^-10 Torr x liter / second sensitivity leak rate • When it is sprayed with small amount of helium, the leak rate rose to mid 10^-9 Torr x liter / second • When it is sprayed with a hefty amount of helium, the leak rate rose to mid 10^-7 Torr x liter / second 12

Repair Options Evaluated 1. In-situ in the NML cave – not feasible due to tight space and reachability limit (~1 meter inside from downstream end) to the area to locate the exact location of the leak 2. In-situ at NML floor – does not seem feasible due to reachability limit to the area for any kind of repair even we can locate the exact location of the leak 3. At CAF-ICB (partial cold mass disassembly) 4. At CAF-ICB and CAF-MP( (full cold mass disassembly) 13

Repair Strategy • Partial Disassembly at CAF-ICB • Order spare parts (such as magnet current leads components) that are risky to break during the disassembly • Disassemble components (warm end couplers, instrumentation flanges etc. ) assembled to the vacuum vessel so it can be slid off the cold mass • Disassemble cold mass downstream end MLI and heat shields to reach the leak region and pinpoint the leak • Repair the leak • Reassemble • Detailed repair plan and a tentative schedule are being developed this week • Estimated duration: 7~9 weeks, most important is to complete the repair work safely and understand the root cause of the problem. 14

Partial Disassembly – Repair – Reassembly Steps Estimated duration: 7~9 weeks, requires all CAF technicians involvement • Workflow: – Backfill the cavity string beamline vacuum to atmosphere – Backfill the warm coupler vacuum to atmosphere, disassemble ion pump and titanium sublimation pump – Move CM 2 to NML floor – Remove big sliding bellows – Remove the tack welded GRHP adapter flanges – Transport CM 2 to ICB – Remove the cold mass support top hats 0. 5 day – Grind the magnet current leads flange from the vacuum vessel 1 day – Grind the magnet instrumentation pipe flange from the vacuum vessel 0. 5 day – Disassemble instrumentation feedthrough flanges 2 days – Disassemble warm coupler vacuum header pipe 2 days – Disassemble coupler tuner motors, waveguides and warm couplers 8 days 15

Steps-II – – – – – Move CM 2 to the Big Bertha Fixture 0. 5 day Disassemble cold mass supports posts Slide Vacuum Vessel off the cold mass for 6~8 feet from the downstream end 0. 5 day Cut 80 K MLI, Grind 80 K heat shield finger welds for the last downstream end aluminum heat shield 2 days Cut 5 K MLI, Grind 5 K heat shield finger welds for the last downstream end aluminum heat shield 2 days Leak check 2 -phase circuit and pinpoint the leak location 3 days Repair leak (probably with manual cutting & welding procedures) 3 days Leak check, pressure test 2 days Reassemble 5 K shield & MLI 1 day Reassemble 80 K shield & MLI 1 day Vacuum vessel assembly 0. 5 day Welding of the magnet current leads & leak check 2 days Alignment 2 days Warm end couplers assembly 8 days Instrumentation flanges assembly and QA 3 days Final QA checks 2 days Insulation vacuum check 3 days 16

Summary • CM 2 is planned to be transported to CAF-ICB on June 26, 2012. • Partial disassembly, repair of the leak and reassembly will be done at CAFICB • Currently we have 2 CAF technicians in the shutdown, they need to be back to the CAF group for the duration of the repair. Also, CAF group will not be able to send technicians to the shutdown until CM 2 is transported back to NML. This is being arranged. • Estimated duration of the repair for Option 3: Seven ~ Nine weeks after CM 2 arrives to CAF-ICB. • A detailed disassembly procedures, schedule documents are being prepared. We do not have disassembly travelers for 1. 3 GHz modules therefore we will start drafting them with CM 2 work and complete them with CM 1 work. • As for the lessons learned, until we can find the exact location of the leak, we cannot do a root cause analysis. It is clear that more vacuum leak check steps need to be added for the helium circuit during the module assembly. 17