------------------------------------------------------------
Minutes of the ABP-RLC team meeting of 12.09.2006
present: OB, FC, UD, AG, WH, Tom Kroyer, EM, TP, FZ, Bruno Zotter
excused: FR
web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/
------------------------------------------------------------
(1) Reports from meetings & announcements
------------------------------------------
OB reported from meetings and commented on other issues concerning the ABP group:
A general group meeting is planned for 28.9., where he will present new group members and the result of the staff review and medium term planning.
Section leaders are asked to give a 5 to 10 minutes presentation on the activities of the sections.
In the absence of FR he asked WH to present the LHC part of the RLC section and HB to report on the CLIC activities.
Section members are invited to provide relevant material for these presentations.
A discussion on the presentations will follow in the next section meeting.
=> ACTION: All
In a special LTC the LARP collaboration is discussed and feedback on the experience with the present activities should be sent to OB.
=> ACTION: All involved in LARP collaboration
A new rule for the participation in conferences for members of the AB department was defined.
The candidates for attending the PAC 2007 should be nominated and forwarded to the group leader.
Please note that abstracts MUST NOT be submitted unless the participation is decided.
This differs from the practice in previous years.
=> ACTION: section leaders
OB mentioned that it is possible to attend the ICAP in Chamonix (early October) on a single day instead of the entire week.
However it is required to register for the conference.
=> ACTION: people interested
OB briefly mentioned the restructuring of the AT department, in particular the possible merging of the MAS+ACR and MTM+MEL groups.
Caused by the problems to deliver the necessary number of collimators, the LHC will run in 2007 with a minimal configuration.
To install the necessary collimators for 2008, the schedule for the planned 2007/2008 shutdown will be changed.
J. Poole will soon retire and some of his present activities are redistributed.
It is suggested that the editing of the Chamonix proceedings is taken over by ABP.
OB proposed C. Carli as a candidate.
(2) Wire measurements on the LHC TDI absorber (Tom Kroyer)
----------------------------------------------------------
TK first described the layout and geometry of the TDI. It consists of 8 50-cm long Al cells, each containing 3 blocks of collimator jaws.
The longitudinal spacing between cells is 5 mm, and the depth of the separating slots about 9 cm.
Most of the jaws (namely those in 6 of the 8 Al cells) are made from Boron Nitride with 4 micron Ti coating on the surface, and with electrical contact at the sideways shoulders.
The total jaw length is 4.2 m (2.85 m BN) and the nominal aperture at injection is 8 mm. The elements were recuperated from LEP.
There are no well defined electrical contacts between the blocks. The spacing between blocks is much less than 1 mm.
In simulations, AG had earlier found a large number of trapped modes starting from 30 MHz with peak impedances below or near 1 MOhm/m;
see the minutes from the RLC meeting of 13.01.2006. As a result the installation of ferrites had strongly been recommended.
The ferrites should damp the modes leaking into the surrounding tank. Recently it turned out that the ferrites were omitted from
the first TDI units to be installed in the LHC without informing us.
In the beginning of 2008 the installation of a new TDI with ferrites is foreseen.
For the bench impedance measurements a special CuBe wire is used, since the yield strength of Cu alone would be insufficient, for
a 50 N force needed to keep the sag below 1.2 mm.
OB asked whether the measurements could not be done in a vertical configuration. The answer was that the TDI is too heavy and too big,
OB pointed out that vertical measurements are possible for the s.c. LHC magnets.
Depending on the jaw position the characteristic impedance of the TEM line changes between 340 and 100 Ohm. The line was matched
for a nominal symmetric aperture of 8 mm. Impedance measurements were performed for open and closed configurations as well as for intermediate cases.
The data were corrected for the wire attenuation. The remaining mismatch causes some ripple as a function of frequency, but its magnitude is
small compared with the overall losses seen for 8 mm aperture above 300 MHz. Deep dips above 800 MHz are explained as probably due to lambda/4 resonances due to the 9-cm gaps between
Al units. Measurements for transverse sweeps give information about the transverse impedance as a function of distance to the jaw.
In particular shunt impedance and Q values of resonators can be inferred as a function of the transverse offset.
A problem is that the position of resonances shifts as a function of the jaw position, since the internal
geometry is varied. Nevertheless, it is clear that as the wire is moved off center additional and stronger resonances appear.
The sum of the driving and detuning impedance is found to be about 8 MOhm/m for a very strong absorption dip at 84 MHz.
The transverse broadband impedance is inferred from a parabolic fit. With 8-mm gap, it is 2-4 MOhm/m in the range between 0 and 200 MHz.
For 15 mm gap the BB impedance is 0.65-0.9 MOhm/m. FC and TK remarked that the longitudinal impedance appeared significant as well.
With 8-mm gap, it is about 600 Ohm at 400 MHz. TK also added that rather large uncertainties for the broad band Z_Y arise from
the dependence of the line impedance on the wire position etc. They could easily be a factor of 2.
A single Al cell housing 3 BN blocks was examined more closely. The DC longitudinal resistance of the Ti coasting was found to be about
twice the value calculated from the ideal material parameters. Both transmission and resonator measurements were performed. The measurements suggest that block-block transitions
contribute strongly to the overall longitudinal impedance. Extrapolating the transverse sweep measurement for the single cell to the full BN
length, a moderate 0.2 MOhm/m broadband impedance is obtained. The 10-20 times larger impedance found for the full assembly is attributed to
the effect of untapered transitions and the large uncertainty in the transverse measurement on the full structure.
Summarizing the measurements, the resonant modes have a transverse impedance of order 5 MOhm/m, some higher.
The broadband impedance is 2-4 MOhm/m, with a factor 2 uncertainty.
AG pointed out that the transverse impedance of the TDI had been expected to be about 16% of the total LHC broadband impedance.
(namely 320 kOhm/m corresponding to 16% of the total LHC BB impedance budget of 2 MOhm/m; see the ABP-RLC team meeting of 13.01.2006).
The measured value of the impedance equals or exceeds the total LHC broadband impedance budget.
This result partly confirms the earlier recommendation to install ferrites.
AG added that the ferrites only reduce the resonator part of the impedance, while to reduce the broadband impedance,
we should remove the discontinuities. FC quoted as a rule of thumb that 1 Ohm longitudinal impedance corresponds to 20 W of power.
It was discussed whether 8-mm gap was needed for machine protection or for avoiding quenches in a "sweeping failure mode".
OB remarked that the aperture of the TDI should be larger than that of the secondary collimators.
=> ACTION: Follow-up with Brennan the TDI gap requirements (EM)
=> ACTION: Check that beam-induced heating of several kW is acceptable (EM)
It was pointed out by FC that the true Q value of the resonances was not measured, but that the measured value is
dominated by the wire.
It was unclear whether the small aperture of the TDI would be seen for a single turn, a few seconds, or 20 minutes.
The weight of the movable parts is 100-200 kG.
=> ACTION: Clarify TDI operation mode with Brennan (EM)
FC noticed that at the very least we should document the problem and be aware of it. OB advised before taking any action to demonstrate
that there is a problem. FC recommended the ALARA principle. OB commented that we still have the chromaticity knob to combat the instability.
EM recalled the instability calculations at injection and at top energy. The stability is marginal also at injection.
There are three TDIs, one for each ring and one spare. It was also unclear whether the TDIs for both rings will be installed without ferrites
or only one of them. In the first case, how could both TDIs be replaced in 2008?
=> ACTION: Follow up no. of TDIs w/o ferrite and spares (EM)
(3) Transverse resistive wall impedance with finite length (EM)
---------------------------------------------------------------
EM computed the finite-length resistive-wall impedance in mathematica, implementing formulae derived by
Bruno Zotter and Bob Gluckstern in a note from August 2006. Choosing a matrix truncation order N=10 and assuming
typical parameters, the resulting impedance agrees within a factor of 2 with the infinite resistive-wall impedance,
and the shapes of the impedance-vs-frequency curves are identical.
FZ pointed to an earlier paper by Gennady Stupakov on a similar topic. EM replied that Gennady assumed
the Leontovich boundary conditions, which are not valid in the regime of the inductive bypass.
(4) New GSI RW-impedance result for the LHC collimator (EM)
-----------------------------------------------------------
New results were sent by Rainer Hasse,. The earlier ones were affected by numerical problems. The shape of the GSI
impedance vs. frequency now looks consistent with the shape calculated by EM. However, while before the GSI result was
100 times lower than the CERN result, it now is 4 orders of magnitude larger. The discrepancy appears to be
likely due to a missing numerical factor.
The discrepancy also persists in the limits of the thick-wall and thin-wall regimes.
(5) Comparison of CERN and GSI RW impedance results for SIS100 (EM)
-------------------------------------------------------------------
Remarkably, for SIS100 there is a good agreement between CERN and GSI results for the real part of the RW impedance.
EM commented that several thin-wall theories exist, and that the one used by GSI goes back to Sacherer (1977),
whereas the one he is using can be found in the Zotter-Kheifets textbook.
EM mentioned that Federico Roncarolo will perform low-frequency transverse impedance measurements for FP420.
FC said that in a previous contact with Daresbury they planned to pursue high-frequency longitudinal measurements.
(6) AOB
-------
The next meeting will be on 19.09.2006. It will include a discussion of LARP and preparation for the group meeting.
Posted on the web: Slides by TK and EM
Web site: http://ab-abp-rlc.web.cern.ch/ab-abp-rlc/