Summary of Visit to NCBJ on 23-24 April 2012 and follow-up

Participant	Institute	Responsibility	Email
Pawel Krawczyk	NCBJ	general manager	Pawel.Krawczyk@ncbjNOSPAMPLEASE.gov.pl
Marek Marczenko	NCBJ	production coordination	Marek.Marczenko@ncbjNOSPAMPLEASE.gov.pl
Marcin Wojciechowski	NCBJ	project coordination	marcin.wojciechowski@ncbjNOSPAMPLEASE.gov.pl
Lukasz Kujawinski	NCBJ	brazing, leak testing	Lukasz.Kujawinski@ncbjNOSPAMPLEASE.gov.pl
Grzegorz Brzezinski	NCBJ	metrology	Grzegorz.Brzezinski@ncbjNOSPAMPLEASE.gov.pl
Maurizio Vretenar	CERN	Linac4 project leader	maurizio.vretenar@cernNOSPAMPLEASE.ch
Frank Gerigk	CERN	project coordination	Frank.Gerigk@cernNOSPAMPLEASE.ch
Rolf Wegner	CERN	project coordination	Rolf.Wegner@cernNOSPAMPLEASE.ch

1st presentation, R. Wegner: PIMS production results since last meeting and qualification status, short summary:

• presentation: https://edms.cern.ch/file/1215545/1/Presentation_01_Rolf_-_Status.pptx
• list of acceptable deviations from tight specification (EDMS 1210268)
• 1st PIMS element machined: disc G_2-3 (EDMS 1209676)
• machining of rings by turning without "straightening"
• postponing machining of module N to end of production, qualification with module M
• improvements in EB-welding procedure by cosmetic pass, qualification ongoing
• valuable weekly reports
• qualification status: discs (standard, central, end) to overlength +1 mm, rings (standard, pick-up) by turning / milling to overlength +2 mm, waiting results of brazing test to take decision on strategy (horizontal or vertical brazing)

2nd presentation, R. Wegner: Points of discussion / sensitisation, short summary:

• presentation: https://edms.cern.ch/file/1215545/1/Presentation_02_Rolf_-_Points_of_discussion.pptx
• assembly of disc and rings: outside alignment => tight tolerances,
• critical dimensions for assembly and their tolerances
• proposal of expanding groove of all rings from 1.5 mm to 1.7 mm to avoid assembly problems
• proposal for changing tolerances of reference surfaces => tighter local perpendicularity, relaxed global positioning
• raising awareness to threaded holes for intertank flanges and intertank tables that are located close to the cooling channels (remaining copper wall thickness less than 6 mm)

Discussion after presentation:

• agreement to change groove of all rings for 1.5 mm to 1.7 mm (details EDMS 1215638)

Discussions at CERN after meeting:

• reference surfaces do not need to be rectangular, they can also be circular to ease machining. However, the tool initially though to be used for machining cannot achieve the tolerance 8H7. Hence, the present machining principle is kept.

3rd presentation, M. Vretenar: Linac4 Masterplan, short summary:

• presentation of Linac4 / LHC injector upgrade Masterplan: https://edms.cern.ch/file/1215545/1/Presentation_03_Maurizio_-_Linac4_Masterplan_12_15_v2.pdf
• Linac4 needs to be ready from end of 2014 to be able to replace CERN's only proton supplier, Linac2, as a backup.
• beam commissioning of the full Linac4 need to start not later than 1st August 2014.
• the full PIMS section needs to be installed in spring 2014.

4th presentation, F. Gerigk: PIMS installation planning, short summary:

• presentation: https://edms.cern.ch/file/1215545/1/Presentation_04_Frank_-_PIMS_installation_planning.pdf
• from reception of the 15, well machined, cavity elements to the final welded cavity will take minimum 2.5 month (for the prototype it took 3.5 month)
• after the welding, 2 month are needed minimum until a cavity is ready for installation (installation in test stand, conditioning, metrology, alignment)
• maximum 1 cavity can be welded per month
• the current installation schedule foresees the first cavity to be at CERN in May 2012, which is unrealistic.
• shifting the installation of 3 PIMS cavities to installation slots of other cavities (! difficult installation scheduling !) allows to postpone the arrival date of the 1st PIMS cavity from the Soltan Institute to August 2012.
• from September 2012 onwards, 1 cavity per month needs to arrive at CERN in order to get the last cavity in September 2013 (including 2 month for vacation + Christmas + unexpected problems).
• In summary: In order to follow the Linac4 masterplan it is necessary to machine 1 cavity (consisting of 15 elements: 2 end discs, 6 discs, 6 rings and 1 waveguide ring) per month.

Questions after presentation:

• What is possible ?
• Can CPL deliver the required quality and quantity ?
• Where and how can CERN support ?
• Can CERN underline the urgency of the PIMS production with a letter to the Soltan Institute? Yes
• Can CERN support the purchase of a new welding/milling machine with an advance payment? In principle yes, but CERN needs to know the needed amount first before discussing with the CERN finance department about possible ways of advance payments.

5th presentation, M. Marczenko: PIMS production strategy, short summary:

• presentation: https://edms.cern.ch/file/1215545/1/Presentation_05_Marek_-_Strategy.ppt
• 3 machines at NCBJ used for machining at the moment: Haas for rough machining, DMU for drilling of cooling channels, threaded holes, milling the inside disc shape and final machining and TOS for drilling of long cooling channels, milling of coupling slots and outside diameter.
• with this re-distribution, the operation time of the TOS could be reduces from 70 hours to 40 hours per disc. The machining time on the TOS is the limiting factor for the production speed at the moment.
• considerations to purchase an additional machine, dedicated for the PIMS production, DMU 80 (guaranteed accuracy ±8 µm) - a combined turning-milling machine. The machine if purchased quickly could be installed until December 2012 and could double the production capacity at NCBJ.
• further information: a seasoning time of 24 hours for relaxing stresses in the material is respected.
• 2 new people have been hired to complete a 2nd shift on TOS and DMU
• considerations to subcontract the drilling of cooling channels
• potential sub-contractors with lathes have been contacted, no one replied positively due to the tight requirements. CPL stays the only option for the moment. However, for rough machining operations (less time consuming) sub-contractors could be employed if needed.
• at CPL 4 machines are working for the PIMS production (DMC 100V, DMU 70, PARPAS and DMC 1035) but without temperature control. Metrology can only be done at NCBJ.
• the contract with CPL has been made for all 12 cavities (end discs, central discs and final machining of rings)
• a detailed list of the production status of all PIMS parts is given at the end of the presentation

Discussions during visit with P. Krawczyk and director general G. Wrochna:

• The tight time constraint for the production of the PIMS cavities of Linac4 has been discussed with the management. The Soltan Institute considers the purchase of a new machine specially to keep the schedule for the PIMS production. This would be a big investment and the strong desire for further collaborations with CERN on future projects has been expressed.

Discussion after presentation:

• to raise awareness: some cooling channels are close to holes and the vacuum section. Furthermore, they shall meet such that the channel cross-section of 12 mm diameter is not reduced more than 1 mm maximum.
• pick-up rings M_1-2 and M_13-14 to be used for the short module will be shipped quickly to Juelich for welding (the welder will go on holidays for 3 weeks in mid may).

Discussions at CERN after meeting:

• the CERN lathe has a guaranteed tolerance of ± 10 μm, and is therefore regularly calibrated - including the temperature - with the help of the metrology lab.

Discussions of qualification procedure:

• a short review:
  • technical specifications: 1 complete (pre-series) module sent to CERN, 2 month for inspection (including low and high power RF tests), after approval the series production can start. Option to approve certain production procedures during construction of the pre-series module.
  • to advance quicker with the series production, decision taken in April 2011 (EDMS 1141661) to qualify series production with a short module (2 end discs N, 2 pick-up rings G and 1 disc G) by metrology, vacuum leak and low power RF checks.
  • the machined end discs of module N are out of tolerance, adjacent rings N_1-2 and N_13-14 need to be adapted particularly. Decision taken to re-start short module for qualification with the 5 elements of module M (EDMS 1207335)
  • disc G_2-3 and end disc M_1 have been finished. End disc M_14 will be machined soon, rings M_1-2 and M_13-14 will be sent for welding of ports soon and will be finally machined afterwards.
• the rings M will be the last elements to arrive for completing the short module. Disc G_2-3 and end disc M_1 will be sent to CERN directly after completing the quality assurance programme. After a successful the metrology check at CERN, the green light will be given for the series production of discs in order to advance quickly. A vacuum test will be performed at CERN to ensure that no material impurities connect the water channels with the vacuum volume.
• the RF check will be performed once the rings of the short module arrive at CERN. In case of unforeseen design issues, the discs produced by then will be modified if the available tuning range is not sufficient (to lower the frequency the cell length can be reduced by re-machining the (delicate) assembly diameter of discs or even rings, while the gap can be opened to increase the frequency by re-machining the (delicate) nose cone tip of discs).

6th presentation, L. Kujawinsk and M. Wojciechowski: Brazing of Waveguide Ring, short summary:

• presentation: https://edms.cern.ch/file/1215545/1/Presentation_06_Marcin_Lukasz_-_Brazing.ppt
• details of horizontal brazing as discussed with A. Vacca – who brazed the waveguide ring of the CERN PIMS cavity.
• additional brazing wires for ports and waveguide flange as discussed with A. Vacca.
• to position the waveguide flange, a spacer will be used for the 1st ring. If successful, a geometrical change is proposed to integrate the spacer directly into the ring for the remaining 11 PIMS modules.
• a brazing test is foreseen to validate the chosen procedure
• 2 questions to be answered before the test:
  1. Shall the copper test piece be shaped like the waveguide ring or can it be cubic?
  2. How shall the brazed piece be analysed?
• in order not to loose time, the waveguide ring of module M will be machined with the integrated spacer – in case the decision is taken to braze the ring vertically, the spacer will be machined off.
• a list with points to be discussed with the CERN brazing expert before the test will be sent to CERN.
• change of drawings: CERN will adapt all functional drawings, NCBJ will prepare all executional drawings as agreed in the technical specifications.

Discussions with A. Vacca at CERN:

• the shape of the copper piece for the brazing test shall be similar to the one of the waveguide ring for a representative test. Thin copper regions expand more than thicker ones and particularly the thin copper area in the middle of the long side of the flange is critical for the brazing.
• for the evaluation of the brazings it is mandatory to install a thermocouple close to the brazing wire to measure the temperature profile of the copper piece. Details have been given.
• a remark from A. Vacca: the furnace will behave differently for the test piece and a complete waveguide ring due to different masses and shapes. Different calibrations might be needed.
• after the brazing test, the copper piece can be cut in a plane parallel to the flange so that a vacuum test can easily be performed.
• the following analyses are foreseen to be repeated/made at CERN: vacuum leak test, ultrasound test and metallurgy. Please do NOT cut the test piece after the brazing. After the metallurgical cuts done at CERN the test piece will be sent back to NCBJ if needed to do further cuts.
• for the Quality Assurance the following documentation is needed:
  • drawings that show the locations/positions of the brazing wires
  • diameter and specification (material, etc.) of wires
  • details of all machining steps and all stabilisation steps of copper parts :
    1. material overlength xxx, temperature profile during stabilisation, location of thermo sensors, type of thermo sensors
    2. material overlength xxx, temperature profile during stabilisation, location of thermo sensors, type of thermo sensors
    3. material overlength xxx, temperature profile during stabilisation, location of thermo sensors, type of thermo sensors
    4. etc.
  • details of all machining steps and all stabilisation steps of stainless steel flanges :
    
    1. material overlength xxx, temperature profile during stabilisation, location of thermo sensors, type of thermo sensors
    2. material overlength xxx, temperature profile during stabilisation, location of thermo sensors, type of thermo sensors
    3. material overlength xxx, temperature profile during stabilisation, location of thermo sensors, type of thermo sensors
    4. etc.
  • temperature profile during the brazing of the following parts:
    1. the copper piece close to the brazing wire for the waveguide flange
    2. the stainless steel waveguide flange
    3. the furnace
  • analysis of all brazings
  • a report from A. Vacca as an example for part of the documentation can be found here: https://edms.cern.ch/file/1215545/1/Agostino_Vacca_-_Brasage_PIMS-English.pptx
  • the documentation does not need to be perfectly formatted, the pieces of information are important.

Discussion of Quality Assurance Reports:

• the Quality Assurance Report format for all 6 element classes has been discussed. The agreed changes have been implemented: https://edms.cern.ch/file/1215545/1/Quality_Assurance_Report_format.pdf and can also be found in EDMS 1210268.
• a strong recommendation from the last visit to CERN was to use temperature compensation for metrology measurements. In case the calibration is unacceptably time consuming, a test with disc G_2-3 is proposed: the outside diameter shall be measured just before shipment to CERN without temperature compensation either on 4 defined points or as profile. The same measurement will be repeated at CERN. A remark: The absolute temperature of the copper piece is very important to be measured precisely. To cross check the metrology machine sensors, a CALIBRATED contact thermometer ("Testo 720" with a head "PT 100") is regularly used at CERN.

EDMS-link:

• https://edms.cern.ch/document/1207335/

Actions to be taken:

• NCBJ: Finding answers to the 3 essential questions:
  • Is it possible to produce 1 cavity per month at NCBJ + CPL ?
  • Can CPL deliver the required quality and quantity ?
  • Where and how can CERN support the production ?
• NCBJ: full Quality Assurance Report of disc G_2-3 according to discussed format
• NCBJ: full Quality Assurance Report of end disc M_1 according to discussed format
• CERN: Shipment of remaining material to NCBJ
• CERN: design and machining of tooling for vacuum leak test
• NCBJ: brazing test and analysis

We would like to thank our Polish colleagues for the good visit and the extraordinary hospitality!

---

-- RolfWegner - 16-May-2012