LHC Insertions Upgrade Working Group

Minutes of the 4th LHC Insertions Upgrade Working Group held on 1rst November 2007

Present: V. Baglin, O. Brüning, F. Cerutti, S. Fartoukh, M. Giovannozzi, J.-P. Koutchouk, H. Mainaud Durand, R. Ostojic, E. Todesco, D. Tommasini, E. Wildner, F. Zimmermann

The minutes of the last meeting were approved without any comments but with a reserve issued by E. Todesco concerning the maximum possible inner coil aperture associated to a given cold mass diameter. This topics will be re-discussed between RO and ET. If needed, the result of this discussion will be reported back to the working group.

Ralph summarized the main requirements related to the collimation and machine protection, which are basically to protect the machine against quench and damage. The present collimation scheme is based on a two-stages cleaning system with the primary and secondary collimators set at n1/n2=6/7 sigma's, and additional absorbers and specific devices to protect the experiments and the machine against background and accidental operational failures. More specifically, in the IR region, tertiary collimators, the TCT's located in between D1 and D2, are needed in physics to protect the triplet aperture against the incoming tertiary halo (with, as well, a beneficial impact on the background seen by the experiments). Upstream, in front of Q5, the TCL's are foreseen to catch the possible debris exported from the IP. The transverse setting of these additional equipments ranges in the 1.4 sigma's clearance which exists in between the position of the secondary collimator (n2=7 sigma's) and the current machine aperture, that is 8.4 sigma's given by the triplet and D1 in the layout version V6.5 of the LHC. Any change in the machine aperture needs therefore to be seen in terms of collimation and machine protection, which concerns both cold and normal conducting magnets (keeping in mind that the damage level corresponds to a fraction of a nominal bunch at 7 TeV).

Ralph then listed the different limitations expected with the present collimation scheme, and the possible improvements or additional degradations that would result from an increase of the collimator gaps, even if potentially allowed by an increase of the D1 and triplet aperture.

At 7 TeV (squeezed optics), the main contribution to the transverse impedance of the machine is given by the collimators located in the momentum and betatron cleaning insertions IR3 and IR7. For a given b* and a given machine aperture, the latter may therefore impose severe limitations in terms of beam intensity and therefore machine performance. The machine impedance decreases however very quickly when increasing the gap g of the collimator jaws (approximate scaling in 1/g³). Applying a constant scaling factor of 1.5 to the nominal gap of all moveable elements of the LHC ring is then found to be just sufficient to be compatible with the nominal intensity and filling pattern of the LHC beam (information supplied by Elias Metral). This would lead therefore to a possible working point corresponding to n1/n2=9/10.5 for the primary and secondary collimators, respectively, and then to a minimum allowed aperture of 10.5+1.4=11.9 sigma's for the D1 and inner triplet quadrupole magnets (see above the machine protection related requirements). OB and FZ however mentioned the fact that playing with the transverse damper and the machine chromaticity in physics may also help in practice but to some extent, not fully quantified, due to possible negative impacts on the beam emittance growth and on the dynamic aperture in the presence of the beam-beam effect. SF then insisted on the fact that about 95% of the transverse impedance budget is given by the secondary collimator in physics. Therefore, the possibility of applying the ~1.5 scaling factor to n2 only, and then re-optimizing independently and for other purposes (e.g. collimation inefficiency, see below) the gap of the primary collimators and other absorbers must be kept in mind as a non-negligible degree of freedom which would be allowed by increasing the triplet aperture.

Triplet and LSS: In the early versions of the LHC collimation system, the local collimation inefficiency was found to be too high in the region of the inner triplet. Therefore, the tertiary collimators TCT's were implemented in between D1 and D2 to solve the problem, with a gap adjusted to 8.3 sigma's, i.e. just below the nominal aperture of the triplet itself. For a nominal aperture of 70 mm in the current triplet, these tertiary collimators were designed with a maximum possible gap of 58 mm. This will be far from being sufficient to be compatible with the ~130 mm aperture of the new inner triplet quadrupoles. Said differently, these tertiary collimators, if not removed nor modified would become secondary if not primary collimators for b* values, let say below 35-40 cm (estimate made by SF after the meeting). As soon as the layout and optics solutions will be finalized, it is therefore highly recommended to check whether the TCT's will still be needed for the IR upgrade phase I, but also, as noted by SF, whether additional similar devices might also be requested to protect Q5, Q4 and D2: n1~15 in the LSS for the nominal LHC which will hardly exceed 9-10 for phase I, even for an as short as possible triplet and assuming that the LSS magnets are not changed (see presentation by SF at the LIUWG#2). If such limitations are demonstrated, which is very likely, and if no budget is allocated for (re-)building these specific equipments, SF then concluded that the enlarged aperture of the new inner triplet and the associated 25 cm b* optics might on the contrary not fulfill the collimation requirements, regardless of the working point of primary and secondary collimators.

LHC arcs and DS of IR3 and IR7: A priori no specific degradation of the collimation efficiency at 7 TeV is expected in the LHC arcs in general when increasing the gap of the primary and secondary collimator jaws. On the contrary, the losses of off-momentum particles in the dispersion suppressors of IR3 and IR7 remain a open issue. For the LHC version V6.5, assuming a beam life-time of 0.2 hour, this already limits the intensity of the LHC beam to about 40% of its nominal value. Then increasing by 3 sigma's the gap of all moveable equipments of IR7 (i.e. n1/n2=9/10 and n3=13 for the absorbers), the local collimation inefficiency will be further degraded in this zone by about 30%, limiting therefore the maximum allowed beam intensity to about 30% of its nominal value. SF argued that this degradation might only be due to the retraction of the primary collimator (and/or absorber) jaws which, strictly speaking, is not mandatory to reduce within acceptable bounds the machine impedance (see above). He therefore suggested to reevaluate this degradation assuming n2 fixed to10, but possibly using a more convenient value for n1, deeper than n1=9 (while not affecting, of course, the collimation efficiency in the other parts of the rings). To summarize, it is clear that this collimation related limitation will not be solved by the increased aperture of the new inner triplet. On the contrary, it can only become more severe if one naively rescales the collimator gap with the triplet aperture.

Finally Ralph presented a non-exhaustive list of potential issues which could result from an increase of the collimator gaps and from the reduction of b* in IR1 and IR5:

a) opening the collimator jaws increases the delay between the detection of any failure mode and the trigger of the dump. SF and OB however argued that most of these modes are expected to be relatively slow at 7 TeV (e.g. drift of the closed orbit) and SF added again that the need for increasing n1 if n2 is increased has first to be demonstrated (e.g. for collimation inefficiency) before drawing any such conclusion.

b) the huge chromatic aberrations resulting from the decrease of b* (in particular the the non-matchable spurious vertical dispersion induced in the HV crossing scheme or the off-momentum beta-beating, see presentation by RdM at the LIUWG#3) remains currently an open issue. Indeed, depending on the momentum deviation of the particles, primary collimators become secondary and vice versa. A possible remedy, not yet tested, is to increase the clearance n2-n1 with, as mentioned above, a marginal impact on the transverse impedance of the machine provided the gap of the secondary collimators remains n2 around n2=10 sigma's.

c) an increase of the collimator gap in the betatron cleaning insertion IR7 will de facto impose to review the setting of the collimator jaw in the momentum cleaning insertion IR3 in order to minimize the possible cross-talk between the two insertions. The current setting of the IR3 collimators is mainly guided by the abort gap cleaning requirements at 7 TeV (truncation at dp=1.5 10^-3 for particles with zero betatron amplitude). It is not clear if increasing n2 (and possibly n1) in IR3 will remain compatible with these requirements.

Due to lack of time, the presentation of Ezio (maximum aperture of a two-in-one quadrupole as a replacement for Q4 in a second phase, see minutes of the second LIUWG meeting) was postponed to the next meeting.

Next meeting scheduled for 15 November 2007:

- Limits of the present cryogenics in IP1/5 and possible "Phase I" upgrades (L. Tavian)

- Possibilities of improving the heat transfer from cable to HeII in coil cross-sections (D. Tommasini)