Minutes of the LRFF Task Force

3rd meeting on Tuesday 03/04/2012 (09:00-11:00 max, 6/R-018)

LRFF members: Alessandro Bertarelli (AlessandroB), Alexej Grudiev (AG), Benoit Salvant (BS), Elias Metral (EM), Fritz Caspers (FC), Giuseppe Bregliozzi (GB), Hugo Alistair Day (HD), Jose Miguel Jimenez (JMJ), Marco Garlasche (MG), Mike Barnes (MB), Olav Ejner Berrig (OB), Oleksiy Kononenko (OK), Oliver Aberle (OA), Ralph Assmann (RA), Raymond Veness (RV), Rhodri Jones (RJ), Roberto Losito (RL), Stefano Redaelli (SR), Vincent Baglin (VB), Vittorio Parma (VP), Wim Weterings (WW).

Present/Excused: AlessandroB, AG, BS, EM, FC, GB, HD, JMJ, MG, MB, OB, OK, OA, RA, RJ, RL, SR, VB, VP, WW, Mauro Taborelli, Sergio Calatroni.

 1) Comments on the last minutes + Actions

- No particular comment on the last minutes and actions ongoing.

- Info from RJ: Raymond Veness will be his alternate => EM already added RV to the members' list.

- EM will add Federico Carra to the list of invited people.

- Sergio Calatroni should replace JMJ => EM already added SC to the members' list.

- EM mentioned past discussions with Rainer Wanzenberg (from DESY) about RF heating and issues with RF fingers. They got similar problems in the past: an RF-spring which dropped down => See photo from PETRA III (RF shielding). Summary from Rainer:

     1) Consequences: total beam current limited to 80 mA, when running with 40 bunches.

     2) Problem with RF spring:  beam loss or at least very bad injection, bad vacuum condition. Had to stop user operation and had to open the vacuum system to replace the bellow.

     3)  Mechanical or impedance effect: not totally clear. Maybe the tension of the rf spring was not strong enough. Not believed to be a geometrical impedance effect. But has to do with mirror currents.

2) News on the VMTSA impedance: CST simulations (BS): pptx

- BS reviewed the past measurements on old and new models. EM summarized all the past measurements made by Jean-Luc Nougaret et al. at  http://emetral.web.cern.ch/emetral/LRFF/InfoOnLHCEquipmentsWithRFfingers/VMTSA/VMTSA.htm. Note that there are also measurements without RF fingers at all (in one bellow only) which would be very interesting to benchmark => See Action 1 below.

- Notes: Model was not imported from CATIA (i.e. it was home made); thin RF fingers cause mesh issues; simulating the loss of contact create small air gaps which are also causing mesh issues; spring not taken into account; loss calculation assumes surrounding in Copper; there are several ways to simulates a bad contact => See 2 here.

- In the case of conforming RF fingers with perfect contact:

- New RF fingers:

- 0.3 mm of RF fingers thickness.

- ~ 1 mW due to resonance at ~ 550 MHz => Very small (in comparison in particular to what will be obtained with bad contact => See later)

- Old RF fingers:

- ~ 10 mW => Still very small but worse than with new RF fingers.

- Simulating bad contacts:

- 1st type => Symmetric gap on top and bottom (in fact only the RF fingers on the left and right are in contact)

- Huge power loss which increases vs. gap (and also the frequency of the mode is increasing with the gap). In particular this would lead to a power loss of ~ 11 kW for a gap of ~ 2.5 mm!

- AG suggested to simulate this again with a displaced beam as it is the case in reality.

- 2nd type => Gap only at the bottom

- Much smaller values in this case => To be checked (in particular: mode 1, gap = 6 mm, Q of 15 seems  small => To be checked).

- MauroT reminded us that even with RF fingers in place there could be a bad contact.

- We should have specifications for the contacts and also for the lateral displacement. Reminder from AlessandroB: the reason for the fifth axis was in case of grazing beam touching the collimators, and it is only for the TCT => Raise again the issue with StefanoR and RalphA to check if this is still needed. See Action 2 below.

3) Past development work on RF contacts (Sergio Calatroni): pptx

- SC summarized the work of many people in the past (MauroT, etc.).

- SC reminded us that good electric contacts require (referring also to the book "Electrical Contacts: theory and applications" by R. Holm):

- Low surface roughness,

- Soft metals (at least one), => Surfaces are never flat => Only few points are in contact => Better to have soft metals..

- No oxide layer at the surface,

- Theoretically:

- In the elastic regime: R proportional to P^-1/3

- In the plastic regime: R proportional to P^-1/2

- Experimentally: R proportional to P^{-1+- 1/2}

- SC will concentrate on the work on RF fingers for 2 equipments: PIMs and collimators.

- RF fingers for PIMs:

- Constraint of low contact resistance < 0.1 mOhm from FC (et al.) at low frequency.

- No cold welding, low friction.

- Good formability properties.

- LHC Beam Vacuum Interconnection Study Group for the design of the PIMs:

- N. Hilleret as chairman.

- J. Knaster and R. Veness as technical secretaries.

- WG minutes: http://at-div-vac.web.cern.ch/at-div-vac/VACPAGES/MINUTES/LHC_Beam_Vacuum_Interconnection/wel.html.

- Topics:

- Good formability => Was covered by Stefano Sgobba and Karine Couturier.

- No cold welding => Was covered by Paolo Chiggiato.

- Low contact resistance => Was covered by SC.

- Low friction (to avoid peel-off etc.) => Was covered by Mauro Taborelli.

 - CuBe alloy with grade 17410 was chosen as material for RF fingers:

- Good adhesion of coatings.

- Weldability by e-beam.

- Formability w/o need of further annealing for hardening. It is hardened by precipitation but it can still be a formed after. The usual one has to be formed before and treated after.

- Low magnetic permeability => Lower content of Ni. But contains Co (amount small enough for RP). Note that it is called CuBe whereas it should be called CuCo as we have more Cobalt than Be...

- CuBe is better than Cu alone as it has a higher elasticity (which is better for RF fingers).

- Electrical conductivity is higher than standard 17420 and better for mechanical reasons.

- Reminder: IACS = International Annealed Copper Standard. The IACS establishes a standard for the conductivity of commercially pure annealed copper. The standard was established in 1913 by the International Electrotechnical Commission. The Commission established that, at 20C,commercially pure, annealed copper has a resistivity of 1.7241E-8 Ohm-meter or 5.8001E7 Siemens/meter when expressed in terms of conductivity. For convenience, conductivity is frequently expressed in terms of percent IACS.

=> All this was also applied for the RF fingers in the collimators.

- WW mentioned that for MKI they considered stainless steel (to be confirmed) => See Action 3 below.

- From the beginning it was decided to have CuBe because we needed a good conductor. But the RRR is very low (2.3) => This is still too high for a reasonable surface impedance:

1) Coating is needed in order to increase the surface conductivity.

2) Coating is needed for a low contact resistance (+ to avoid cold welding).

- No cold welding => 2 solutions: 1) put a diffusion barrier between the 2 metals (oxide layer) or choose metals with which the solubility of one metal is negligible in the other.

 - The 2nd solution was chosen (as we don't want to have an oxide layer between to have a good contact, see above).

- From the Phase Diagrams, the best materials are Au-Rh as they are the best enemies that we can find (almost no solubility). Ag-Rh are more or less the same. Finally, Au was chosen instead of Ag because it was the best for cold welding.

- Contact resistance => Measurements on samples have been performed:

- With a plating of the CuBe RF finger in Au and a plating of the base material (Cu) in Rh, the resistance was measured to be 3 mOhm for 1 RF finger. The use of Ag instead of Au led to 2 mOhm but Au was chosen for the cold welding reason.

- Contact surface on the insert => Is it electro-polished before putting the Rh coating.

- Rh on the insert surface instead of putting it on the RF fingers themselves => It was to avoid affecting the spring ratio because of the big Rh stiffness (due to its Young module).

- Thickness of gold coating => 30 microm needed => Was reduced in reality to ~ 10 microm as it was measured to be OK and then it was much cheaper (to be checked).

- Friction => Performed at the Ecole Centrale de Lyon:

- Mauro made this contact with Lyon (linked to a review...).

- Engineering of PIMs:

- Giovanna Vandoni made a test of a full PIM at the cryolab, before final validation in the String2.

- RF fingers for collimators:

- The 3 constraints above with  0.1 mOhm replaced by 1 mOhm. This was given by Francesco Ruggiero (for each side of a collimator jaw).

- Resistance to bake out at 250 deg C for 1000 h.

- Resistance to RF heating: good thermal conductivity.

- Collimator design WG led by Alessandro Bertarelli and Manfred Mayer: http://lhc-collimation-project.web.cern.ch/lhc-collimation-project/coll-design-meeting.htm.

- NIMA paper: http://www.sciencedirect.com/science/article/pii/S0168900206012095.

- AG should present his past studies on the RF fingers for the collimators => See Action 4 below.

- The RF fingers are sliding longitudinally and transversally.

- Gold cannot be used for coating here due to bake-out => Ag was chosen in this case => We have CuBe RF fingers Ag plated (instead of Au) on Rh.

- Measurements have been done.

• Heating effect including particle bombardment and RF:
  • Rf fingers are heated, assumed uniform.
  • 0.23 W / finger and 0.16 W / finger if Ag plated => DeltaT = 100 deg of this RF finger. Could be 150 deg if no contact on the flanges (as there is no evacuation through the flange).
  • 0.33 mm thickness of the RF fingers.
• Bending stress (as we know the temperature):
  • The fingers are bent, so they are under stressed.
  • These were first estimates and then AlessandroB made more sophisticated estimates.
• Reminder: 1000 h needed for bakeout in the LHC.
  • There is no data and it is probably 0.
  • The longer the bake-out time the lower the residual elasticity you have.
• AlessandroB said that for the TCTVB (see http://emetral.web.cern.ch/emetral/LRFF/InfoOnLHCEquipmentsWithRFfingers/TCTVB/Visite%20TCTVB4L2.pptx) it was seen that some RF fingers (longitudinal) were not in contact and it could be a consequence of bake-out and they lost their elasticity (consequence of the choice of CuBe17200 imposed by the supplier) => Valid for RF contacts. Reminder for AlessandroB => In case of bake-out the grade of CuBe is very important.
• TCTVB => There is a plate also replacing some RF fingers.

- Comment from AlessandroB: How can we cool the ferrite in case it becomes too hot? => See Action 5 below.

- The bulk material for collimator is SS.

- Comment from MikeB after the meeting => Report written in 2007 by a summer student (New Design for the MKI RF Finger Contacts in the LHC (Jason Rabinovitch, 2007)), looking at alternatives for the MKI RF fingers. It includes a summary of why certain materials were chosen.

4) Actions to be taken for the next meeting

- Old actions.

- New actions:

- Action 1 (Impedance team): Try and benchmark some past measurements without RF fingers at all in one bellow of the VMTSA => http://emetral.web.cern.ch/emetral/LRFF/InfoOnLHCEquipmentsWithRFfingers/VMTSA/VMTSA-02-02-2012-2.pptx.

- Action 2 (EM): Check with StefanoR and RalphA if we still need the fifth axis on the TCT collimators (as it complicates considerably the situation with the RF fingers).

- Action 3 (WW): Check if for the MKI stainless steel was chosen for the RF fingers (instead of CuBe) and if yes, why (maybe a good conductivity is not needed...).

- Action 4 (AG): Review of past impedances studies with the RF fingers for the collimators.

- Action 5 (?): How can we cool the ferrite if it becomes too hot?

5)  Miscellaneous

- The next (4th) meeting will take place on 10/04/2012 between 09:00 and 11:00 (max.) in room 6/R-018 => Agenda:

1) Equipment review: some equipments from TE/VSC => Vacuum modules in the LHC septa (VB).

- See preliminary agendas for the next meetings.

Minutes by E. Metral, 13/04/2012.