Summary of Visit to CERN on 2nd March 2012

Participant Institute Responsibility Email
Marcin Wojciechowski NCBJ project coordination
Marek Marczenko NCBJ production
Frank Gerigk CERN project coordination
Gilles Favre CERN project coordination
Maurice Favrel CERN machining
Alessandro Dallocchio CERN mechanical design
Luca Gentini CERN mechanical design;
Ahmed Cherif CERN metrology
Dominique Pugnat CERN metrology
Rolf Wegner CERN project coordination

1st presentation: Machining of standard discs on TOS, short summary:

  • presentation:
  • precision machining of outside diameter with Kaiser Tool, 50 rpm, feed 2 mm/min
  • milling of slots for feet,
  • drilling of cooling channels,
  • making repulsed threads,
  • drilling other holes as pinholes for targets, etc.
  • total time needed per disc on TOS: 70 hours
    corresponding to 525 shifts of 6.4 hours for total of 48 standard discs
    additionally, 24 central discs and 24 end discs are needed
  • additional information:
    • the temperature of discs has been measured and is compensated during machining
      final machining is done on Tuesdays to Fridays because temperature changes during weekend and on Mondays (heating on/off)
    • machine calibration: metrology measurements have been performed after various machining steps, deviations are compensated
    • machine calibration repeated several times with good reproducibility
    • the groove for placing the support feet is used as reference and for clamping
    • drilling of cooling channels takes a considerable amount of time (~ 8 h) due to a slow drilling speed to avoid vibrations of the long drilling spindle

  • discussion after presentation:
    • clamping methods are essential because copper deforms easily
    • clamping force controlled by measuring deformation of outer diameter; force is adjusted so that no deformation is seen
    • the usage of shims has been discussed in detail to avoid deformation of copper pieces on the machine. No deformation shall be caused by the clamping, neither in horizontal position (diameter) nor in vertical position. This is crucial to reach the required tolerances.
    • cooling channels have been drilled after final machining; that was a misunderstanding, only the pinholes and reference surfaces for metrology shall be drilled after final machining, compare document, page 3. Cooling channels shall be drilled BEFORE final machining as described in detail by Gilles Favre in the CERN machining procedure:

2nd presentation: Machining of standard discs on DMU 70, short summary:

  • presentation:
  • machining of inner disc parts: coupling slots, tuning islands, nose cone, beam pipe and inside-outer radius (R20) in several steps
  • total time needed per disc on DMU 70: 40 hours (for both sides)
  • additional information:
    • clamping on DMU needs to be stronger because the table and disc will rotate during machining. Care is taken to avoid deformations
    • critical surfaces are protected by plastics to avoid scratches
    • discussion with a new subcontractor are ongoing to use a bigger machine that allows to do all machining steps on the same machine, including the drilling of cooling channels and outside final machining to avoid re-positioning and to reduce production time

  • discussion during presentation:
    • inner disc parts are rotational symmetric (apart from tuning islands) and dedicated to be machined by turning. The tight tolerances are easier to be reached by turning while good surface quality is provided and a considerable amount of time can be gained
    • at CERN the 1st rough machining step (+2 mm) was done by milling, the 2nd rough machining step (+1 mm), the 3rd (+0.2 mm) and the final machining steps were done by turning (compare Gille Favre's description It took about 16 hours from the 2nd rough machining step to the finished machined disc on the lathe.
    • a strong recommendation was given by the CERN workshop to finish discs on a lathe because of the achievable precision and machining speed
    • NCBJ was recommended to search for subcontractors with a big lathe

3rd presentation: Machining experience and Outlook, short summary:

  • presentation:
  • the copper pieces can be deformed with little force, great care is needed to achieve the required tolerances. Appropriate clamping and shimming is essential.
  • A surface quality better than the specified Ra of 0.8 has been achieved by milling.
  • Machining of discs (and rings) takes more time than expected. In total it took 126 hour for the disc G_2-3 (rough machining 16 hours + inside final 40 hours + outside final 70 hours). Particularly the final machining takes long (~ 80% of the total time)
  • Investigations how machining can be speeded up
    • using new tools
    • using a 2nd cutting head instead of the Kaiser tool's counterweight
    • organising a 2nd shift at NCBJ, hiring people
    • searching for subcontractors for rough and semi-final machining

Discussion of machining results of discs G_2-3:

  • machined to an overlength of +0.5 mm in radius, one side was finished
  • metrology report:
  • corresponding drawing:
  • non-conformity report:
  • profile measurements: and
  • temperature compensation was switched off although the temperature of the disc was 20.7C. A temperature compensation is recommended by the CERN metrology experts.
  • profile of the inside disc looks promising, the specification of 20 μm has nearly been met, the deviations seen can be accepted - no problems for the RF tuning are expected
  • all non-conformities have been discussed:
    • planarity on side of reference B: the clamping might have lead to small deformations of the disc. Remedy: Shims as discussed above will be used in the future to avoid deformations in the micrometer level during machining of the 1st side. When turning the disc for machining the 2nd side, the final machined 1st side shall be flat.
    • perpendicularity to Ref A and coaxialities: Surprise as beampipe (Ref A) and concerned phase of the disc are machined on the same machine without unclamping. This might also be linked to a clamping problem.
    • outside diameter already machined to final dimension: Outside diameter with 538.757 mm, 27 μm out of the specified dimension 538.7200.010. Difficulties to achieve specified value, outside of this disc could be re-machined but this non-conformity seems to be acceptable for the assembly and beam dynamics requirements. The decision was taken not to re-machine.
    • nose cone length 44.806 instead of 44.7750.005: Clamping deformations are also assumed of having caused this deviation.
      The nose cone length has an enormous influence on the resonance frequency: the deviation of 31 μm (26 μm too much) will increase the cell's frequency by +40 kHz (34 kHz too much) while all inside tolerances of a disc side together are specified to limit the frequency deviation to 68 kHz.
  • The following tolerances are in particular difficult to reach. A discussion with the CERN workshop and design office will take place to see if they can be relaxed:
    • inside-outer diameter 518.7200.020
    • coaxiality 0.020A at diameter 518.7200.020
    • outside diameter 538.7200.010

Visit of CERN workshop:

  • the CERN lathe, claming jaws and cutting tools have been shown and discussed
  • the CERN milling machine has been shown, clamping tools, machining procedures, tools, drills and drilling techniques have been discussed in detail

Presentation on X-ray analysis of electron beam welds done by Julich:

Summary discussions:

  • The claming of elements on the machine is crucial for achieving the tight tolerances required. Principles have been discussed in detail during the meeting.
  • A strong recommendation was given by the CERN workshop to finish discs on a lathe because of the precision and machining speed.
  • Cooling channels shall be drilled into discs before final machining.
  • A short weekly report from NCBJ shall be sent to CERN which describes the following topics:
    1. the activities,
    2. the results achieved,
    3. the problems met,
    4. the decisions taken.

Strategy for series production:

  • At the moment it is important to reach the required tolerances and to identify the most difficult ones. From the CERN side we will see where we can accept bigger deviations than specified in the agreed drawings.
  • In the next phase, the machining procedures shall be optimised to reduce the machining time significantly.
  • Subcontractors shall be found that can support the machining, in particular the final machining which requires the biggest amount of machining time. In particular subcontractors with big lathes shall be contacted.
  • In the collaboration contract it is foreseen to deliver the last of the 12 cavities by latest June 2012. Seeing the machining difficulties met, a delay is unavoidable. However in order not to delay the Linac4 project as a whole, all cavity parts need to be finished by June 2013 at latest.

Further points:

  • A disc supplied to CPL for machining has disappeared. The cost of purchasing a new disc is estimated to be 15 to 20 kCHF.

Actions to be taken:

  • NCBJ: finish the 2nd side of disc G_2-3 with an overlength of +0.5 mm
  • NCBJ: if the specified tolerances are achieved, finish the disc G_2-3 to final dimensions
  • NCBJ: start rough machining of elements of module M for qualification. The elements needed for the short module are rings M_1-2 and M_13-14, end discs M_1 and M_14 and standard disc M_2-3
  • CERN: see if bigger deviations of the discussed tolerances can be accepted concerning RF, beam dynamics and assembly considerations
  • CERN: propose a format for coherent quality assurance / metrology reports for all 6 basic PIMS elements with details on important tests / measurements to be performed
  • NCBJ: send the brazing procedure of the waveguide ring with drawings that show all details of assembly and clamping methods to CERN for discussion
  • CERN: see if Molybdenum wires a) 3 mm, length 4 m and b) 0.8 mm, length 1 m can be sent to NCBJ

Rolf Wegner, 28. March 2012

-- RolfWegner - 29-Mar-2012

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r2 - 2012-03-29 - RolfWegner
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    SPL All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright