Minutes of the 12th LHC Insertions Upgrade Working Group held on 20th March 2008
Present: S. Chemli, S. Fartoukh, P. Fessia, M. Karppinen, J. Kerby, N. Kos, J.-P. Koutchouk, Y. Muttoni, D. Nisbet, R. Ostojic, V. Parma, H. Prin, V. Parma, J.-P. Tock, R. Tomas, E. Wildner
Excused: M. Giovannozzi, H. Mainaud-Durand
Invited: F. Butin, S. Roesler, R. Veness
1. Approval of the minutes of the last meeting
The minutes of the last meeting have been approved with a minor comment by F. Butin.
2. Update of the parametric study of the triplets (E. Todesco, ppt file)
Ezio presented an update of the parametric studies which was performed to characterize a possible inner triplet for the LHC upgrade phase I. This study was already published in the LHC design Report 1000 and presented in 2006 at the CARE-HHH Valencia Workshop. The main change w.r.t. the reports mentioned above is a 10% reduction of the maximum achievable gradient for a given aperture of the triplet.
Ezio started his presentation by reminding the basic assumptions which have been used to define the triplet topology. The triplet is symmetric, which means the same gradient in Q1, Q2 and Q3, the same length for Q1 and Q3 but a substantial difference of length for Q2. The interconnect distance between each quadrupole magnet is assumed to be the same, set to 1.3 m magnetic to magnetic (see also minutes of LIUWG#3). The design of the new triplet is based on standard cos-theta two-layers quadrupoles using the current inner and outer cables of the LHC main dipoles. All the relevant triplet and beam physics parameters can then be expressed as a function of the overall triplet length, assuming different kind of constraints: the same peak beta function in the triplet, the matchability of the optics to the arcs (roughly quantified by the beta function at Q4), a sufficient beam clearance in the triplet with, when applicable, a 3 sigma's margin to facilitate the collimation, the possibility to correct the additional contribution of the inner triplet to the machine chromaticity and, if possible, to allow for a full compensation of the chromatic aberrations induced by the low-beta quadrupoles. Last but not least, the length of the inner and outer cables which will equip Q1 and Q3 (the longer quadrupoles) shall obviously fit with the specifications of the LHC main dipoles (about 450 m for inner cables and 750 m for the outer cables, see also comment by Ranko below).
The conclusions obtained can then be summarized as follows.
Maximum possible triplet aperture (140 mm for a gradient of 104 T/m)
Given the maximum length of the existing LHC cables, the maximum triplet aperture cannot exceed 140 mm. Then, assuming the quadrupoles to operate with a 20% margin w.r.t. the critical surface, the gradient could not be pushed beyond 104 T/m. This triplet quadrupole would then be compatible with a minimum beta* of 17 cm without margin for collimation (10 sigma beam clearance in the triplet) and just at the limit for linear chromaticity correction. Requesting an additional 3 sigma's aperture margin, the beta* should then be increased to 25 cm. On the other hand, in the second case (and a fortiori in the first one), the full correction of the induced off-momentum beta-beat is not be possible (strength limitation in the defocusing sextupoles).
Intermediate triplet aperture (130 mm for a gradient of 112 T/m)
In order to warrant both a 3 sigma's aperture margin in the inner triplet and the possibility to fully correct the chromatic aberrations induced by the low-beta quadrupoles (linear and non-linear chromaticity, and off-momentum beta-beat), the triplet aperture shall exceed 130 mm and beta* cannot be lower than 27 cm (with a marginal loss of performance w.r.t. 25 cm). Relaxing the aperture requirements to a 10 sigma beam clearance inside the triplet, the beta* could then be further squeezed to 19 cm but preventing a full correction of the induced off-momentum beta-beating.
Minimum triplet aperture (110 mm for a gradient of 130 T/m)
The minimum triplet aperture compatible with a 10 sigma beam clearance (no margin) is found to be equal to 110 mm for a beta* of 25 cm. Strictly speaking a beta* not lower than 25.5 cm will then be needed to make possible the full compensation of the induced off-momentum beta-beating by the lattice sextupoles. Then, a 3 sigma's aperture margin will be recovered only if beta* is increased up to 36.5 cm inducing a loss of performance of about 20% w.r.t. to a beta* of 25 cm
Ranko suggested that cable measurements are repeated on a few units before the model construction begins. After the meeting, he then strongly encouraged to do a detailed survey of the actual stock to verify the assumptions made above, in particular in terms of length of the available cables (possibly 430/720 m for the inner/outer cables instead of 460/780 m assumed above, due to uncertainty and after a 10 m systematic cut performed at CERN for cable testing and archive). Therefore, a relevant range for the aperture of the new inner triplets may be in between 110 and 130 mm, instead of 110-140 mm (to be confirmed).
3. Present situation and plans for TAS in IP5, part II (F. Butin, ppt file)
In order to complete the presentation he gave two week ago concerning the TAS in ATLAS (see minutes of the LIUWG#11), François reported on the situation around CMS. The layout of the machine in the two insertions is exactly the same from the TAS to the triplet including the VAX regions. However the TAS supporting structure is quite different. During his presentation, François remarked that in point 5 the access is possible from the cavern to the tunnel with better working conditions due to a flat floor in front of Q1. In CMS, the TAS is supported inside the conical FIN steel shielding (60 tons) which is fixed to the blockhaus and surrounded by the very massive rotating shielding. The TAS is adjusted and aligned transversally with tie rods and support jacks sitting inside the FIN. The position is monitored with fiducials outside the rotating shielding.
François gave details of the TAS assembly procedure that occurred mainly on surface. The TAS were lowered inside the FIN on wooden cradles and then translated along beam axis by 30 cm. The upper surface at the nose of the FIN was completed with a steel plug (7 tons) that closes the shielding conical shape. The support rods were pulled through the FIN to reach the TAS and the jacks installed to lift it and the wooden cradles removed. An additional plug (2 t) requested by the vacuum group is added to close the FIN nose. The complete assembly was then lowered into the UX55 cavern. Both TAS were roughly aligned in the tunnel and a final adjustment has to be redone before the beam injection. The question was raised about the alignment system and François answered that there is no request from the survey group at the moment. He also mentioned that the longitudinal alignment is related to the vacuum components that does not require a very high accuracy.
At the moment, no strategy is defined on how to remove the TAS in CMS. Reversing the mounting procedure would imply too many manual interventions in high activated areas. François presented a possible scenario and he warned that a level of activation has to be assessed to confirm the feasibility. The idea is to benefit from the access around the FIN once the rotating shield is open and to remove the TAS with the FIN plug as a single unit. This would avoid heavy operations on the absorber in the cavern. The whole set could then be transferred with an appropriate lifting structure in the cavern and brought up to the surface with the crane. A mechanical study has to be done in order to guarantee that the positioning rods are stiff enough to lift the TAS through the plug. The plug could eventually be reused to put back a new absorber. As in ATLAS, remote handling for some bolting operations on the IP side could be envisaged. The estimated schedule is about one week in the same order of magnitude as in point 1.
4. The first study of equipment integration in IP1 and IP5 (Y. Muttoni, ppt file)
The presentation is the result of studies worked out with David Nisbet and is the first estimate of space available for the equipment in the tunnel and neighboring underground areas that could be used to install powering equipments and related racks.
Yvon first gave a global overview of the geographical locations with the underground works named around IP 1 and 5. He mentioned that the implantation study presented concerns the cavern UJ14. It is transposable to the UJ16 which is symmetrical around IP1 and almost to the UJ56 at the right of point 5. The particular case of L5 has to be treated separately due to the configuration singularity.
Compared to the existing equipment situation, an energy extraction system with discharge resistors, controllers and switches has to be included for each triplet. A new smaller and lighter type of DQR using semiconductors was discussed with K. Dahlerup-Petersen and R. Denz. A structural analysis shall confirm that these resistances together with the DQS and DJPC could be installed at the UJ14 first floor where free space could be used. Yvon remarked that in point 5, there is no additional level to put this equipment. Their integration is kept in mind.
Among other changes the electrical distribution feed box is shifted away from the beam line; either it could be placed along side the triplet similar to DFBMs in the matching sections, or further away in an alcove like DFBLs implying superconducting link to power the magnets and cryogenic line extension to the QRL. In the case of a DFBL-like feed box located near the racks in UJ14, the routing of the electrical and cryogenic links has to pass through the shielding wall where additional holes have to be drilled. The diameters of the ones presently used for ventilation shafts, piping and powering cables can not afford the DSL and QRLX to go through. David commented a table that resumes the circuits naming with their associated power converters and racks compared to the present situation. The table was drawn with the assumption that the triplet quadrupoles are powered in series with a 13 kA circuit associated with four 600 A trim circuits, and the corrector circuits as the present triplet. The integration of the UJ14 floor takes into account this powering scheme to redistribute the racks in the available space which is delimited by the access control gate and the reserved space for transport. During the machine run, this area stays accessible while a removable fencing prevents entrance to the machine without passing through the controlled access. Taking into consideration the option of a cold D1, another layout was discussed: due to the additional convertors and the larger size of the feed-box, it sits in the UJ13 inside a shielded alcove. This configuration gives more space for the racks in UJ14 and facilitates the routing of the electrical and cryogenic links. Stefan ROESLER confirmed the possibility to use the space in UJ13.
Although the studies are not done yet for point 5, Yvon gave an overview of the space constraints particularly on the left side where RZ54 is probably not spacious enough for the feed box. On the other hand a junction to UCS55 may imply either civil engineering or obstructions in UP53. The typical cross section in R542 shows that the integration will be a challenge.
5. News and A.O.B.
Ranko announced the creation of the LIU project website and presented a first proposal for the logo of the project
S. Fartoukh, R. Ostojic and H. Prin