Minutes of the 13th LHC Insertions Upgrade Working Group held on 2 April 2008
Present: F. Cerutti, S. Fartoukh, A. Ferrari, P. Fessia, J. Kerby, N. Kos, J.-P. Koutchouk, D. Nisbet, R. Ostojic, H. Prin, E. Todesco, R. Tomas, E. Wildner
Excused: F. Zimmermann
Invited: R. Calaga, R. De Maria, M. Fuerstner, J. Johnstone, P. Limon, M. Mauri, A. Mereghetti, N. Mokhov
1. News and approval of the minutes of the last meeting
The minutes of the last meeting have been approved with a comment by Reiner Denz concerning the presentation by Y.Muttoni on equipment integration around IP1 and IP5.
2. Nb3Sn substitute Quads in Phase-I Upgrade Optics (J. Johnstone, Fermilab, pdf file)
John presented first results concerning the possibility of inter-changing NbTi quadrupoles with Nb3Sn magnets in the new LHC inner triplets foreseen for phase I. "Interchangeable magnets" means magnets
a) with the same slot length (i.e. shorter or equal length for Nb3Sn magnets),
b) equal integrated transfer function in a given current range to minimize the impact in terms of additional electrical equipments.
c) minimizing the retuning of the LSS magnets and of the triplet itself after the magnet exchange.
The magnets under consideration are either Q1 or Q3. This idea was then tested on low-beta-max (LBM) and symmetric triplet (Sym) layouts (see minutes of LIUWG#3 for more details). For both layouts, the Q1 was replaced by a 90 mm or a 110 mm aperture Nb3Sn quadrupole (shorter and with higher gradient). Then a possible replacement of Q3 by two Nb3Sn 110 mm aperture 3m long modules was also tested. In all cases, the magnet exchange does not change much the machine performance in terms of aperture or peak beta functions, except perhaps in the case of Q3 for which a 110 mm aperture might be a bit too small compared to the 130 mm aperture NbTi Q2 (see slides for more details). The re-trimming of the LSS magnets for rematching the optics is also found to be marginal, being said that only the collision optics was tested.
P. Limon commented that, even with the same transfer function at high current, Nb3Sn quadrupoles will have quite different behavior at injection (e.g. hysteresis effects, decay, snap-back).
J.-P. Koutchouk commented that, in terms of machine performance, the best candidate for replacement is definitely Q1 which is the most exposed to heat deposition and for which increasing the gradient helps in optimizing the overall triplet performance (which is the basic idea behind the LBM layout). John and Ranko replied however that, from the integration point of view, the Q1 magnet is no longer accessible after Q2 and Q3 have been positioned in the tunnel. Therefore replacing the Q1 (NbTi or Nb3Sn) will require a major intervention in the tunnel.
3. Energy Deposition Results for Hybrid NbTi/Nb3Sn Triplet Configurations of Phase I Upgrade (N. Mokhov, Fermilab, ppt file)
Nikolai presented simulation results concerning heat deposition studies for
the different possible hybrid triplet configurations discussed by John (see
slides for more details). In all cases, a 3
mm W or SS liner running all along the triplet is included to
protect the low-beta quadrupoles, except in the case of LBM hybrid
layout with a 110 mm aperture Nb3Sn Q3 for which a liner is found not to be
strictly needed (see slides for more details).
Generally speaking all the numbers obtained (peak losses in the magnets,
longitudinal density of losses, i.e. ~10 W/m, or integral per magnet, i.e. 400 W
in total per triplet out which 25-35% are intercepted in the b.s. and absorbers)
are quite comparable to the numbers obtained at CERN for the NbTi
symmetric triplet.
4. Update on energy deposition
studies at CERN (F. Cerutti, pdf file)
Francesco presented an update of the heat deposition studies performed on the new LHC inner triplet, with new results concerning D1 (see minutes of the LIUWG#8 for reference). The update concerns first a modification of the layout (induced by a revision of the best achievable gradient for 130 mm aperture NbTi magnets see minutes of the LIUWG#12). This modification is found to have a marginal impact on the expected losses. Then, for simple geometrical reasons, a thicker liner in Q1 (SS instead of W, and with a thickness increased up to 8-13 mm with no impact on the clearance of the primary beam) solves the problem of peak loss density at the entry of Q2a. The highest peak is now at the exit of Q3, but its magnitude, quite independent of the presence of the liner in Q1, is below the 4.3 mW/cm3 criterion. Several other simulation results were also reported (e.g. impact of the X-angle and of the MCBX corrector field on the non-IP side of Q1) to demonstrate the robustness of the present magnet shielding scheme, namely: no TASB or specific shielding in the triplet but a ~10 mm liner running along Q1 and its attached orbit corrector.
The model is now extended up to D1 (normal conducting, super-ferric or cold). In particular, for the preferred option of a 130 mm aperture superconducting D1 (see minutes of the LIUWG#10), simulations demonstrate that the expected losses are still within specifications without any specific protection in D1 except a 3.5 mm thick cold bore and a 2 mm thick beam-screen oriented horizontally. The results presented were obtained for the case of a vertical crossing but others, still preliminary, seems to indicate that the horizontal crossing configuration will not change the above conclusions for D1.
5. A.O.B.
S. Fartoukh, R. Ostojic and H. Prin