Minutes of the 18th LHC Insertions Upgrade Working Group held on 24 July 2008

Homepage


Present: F. Cerutti, P. Fessia, M. Fürstner, J. Kerby, G. Kirby, M. Mauri, A. Mereghetti, K. H. Meb, D. Niesbet, R. Ostojic, S. Peggs, H. Prin, S. Roesler, SL. Williams, E. Wildner.


1.      Status of the corrector design (M. Karppinen ppt)

Mikko first recalled that in the present layout for the Phase-I the correctors are grouped together in a single assembly named "corrector package" (CP), located between Q3 and D1, which is 5 to 7 m long. CP has two orbit correctors MCBX1, a skew quadrupole MQSX, a sextupole MCSX and eventually other higher order correctors. The design of the correctors will be done by CIEMAT, STFC/RAL and CERN, who will also develop the first models. In the frame of the CERN “White Paper”, CEA will follow the series production. The requirements as known today were given for all magnets. Mikko also pointed out that the initial assumption was to use the existing power converters, current leads and circuit protection.

Mikko then presented the fabrication process of the present MCBX. The magnet consists of two concentric impregnated coils, giving horizontal and vertical fields, surrounded by a shrinking cylinder and off-centered laminations, which give the pre-stress. The layout envisaged for the Phase-I upgrade reduces the number of orbit correctors to 2, while at present there are 3 assemblies (6 modules) associated to each quadrupole in the triplet.

Two options are considered for MCBX1 (6 Tm at 1.9 K or 4.5 K): (1) using present wires 3 or 4, combined in 8-way ribbon cable, or (2) winding the coil with MQY cable. The first option was taken as the baseline, since it uses the existing 600 A power converters and standard quench protection system. Mikko pointed out that the available wire quantity at CERN is not sufficient for the coil production, and that new wire has to be ordered. The estimated procurement time is about 6 months. The second option would lead to higher field and shorter magnets with easier fabrication process. The drawback is that a new power converter is needed (about 8.5 kA for the design presented), with a new quench protection system based on heaters and/or dump resistors. Karl Hubert commented that the existing quench protection has also to be adjusted for faster current changes.

Mikko focused on the solution with ribbon cable, and presented the results of quench protection studies, showing quench propagation and temperature distribution under different conditions of initial quench in the magnet. The study showed that the copper ratio in wire 3 is not high enough and the protection is risky. Wire 4 gives a safe design both at 1.9 K and 4.5 K, since the magnet operates at 60 % margin. A mechanical design is being studied to optimize the stress distribution in the collars. It was also shown that the heat exchanger tube does not increase iron saturation significantly. Mikko commented on the tight milestones for the model magnet (test in September 2009), and pointed out that decisions concerning the aperture, the nominal current and the operating temperature are required to finalize the detailed design. Glyn expressed concern about the stability of epoxy impregnated coils and recommended using the MQY-type cable. Karl Hubert remarked that the experience with the MCB(X) manufacture was not positive and noted that heat extraction from impregnated coils could also be a concern. He also pointed out that using epoxy could be risky in the high radiation environment.

The design of the MQSX skew quadrupole has started in collaboration with Rutherford. The presently available requirements for the field gradient and field quality are very demanding, and a sensitivity study will show if the present specifications lead to a realistic design in terms of tolerance margins and fabrication costs. Mikko stressed again the need to specify the basic parameters as soon as possible, as the available manpower in RAL is limited.

CIEMAT developed a very promising design of a super-ferric MCSX using race track coils with small wires. The presently available specs can be easily met and the design seems straight forward. A similar type of magnet could also be considered for the MCTX corrector, if it is needed.

In conclusion, Ranko said that the field error tables for the quadrupoles and D1 dipole must be clarified as soon as possible and used as basis for the multipole correctors. He also pointed out that in the new layout there is less redundancy for the orbit correctors, which means that they must have the same level of reliability as the low-b quadrupoles. Having also in mind that the corrector package will be cooled at 1.9 K, the preferable solution is to use a small Rutherford-type cable and coil insulation similar to the other main magnets in the LHC. This approach requires a close collaboration with PO to define the circuit parameters appropriate for the magnets and power converters.

2.      Update of the quadrupole design (F. Borgnolutti ppt)

Frank first gave an update of the quadrupoles studies with an aperture of 110, 120 and 130 mm. He recalled that the use of dipole cable 01 and 02 is the basic constraint for the coil design and showed the measured cable performance. The short sample gradient obtained for each design is 138 T/m for 130 mm, 148 T/m for 120 mm and 157 T/m for 110 mm. There seems to be no significant increase using special grading for winding the external layer of the coil. As the collar thickness is 37 mm, the iron yoke increases the gradient by 3-5 %. Several studies are ongoing to refine the coil cross-section.

Frank then reported on the studies of the influence of the heat exchanger size and position on the magnetic field. Several cases were considered (four-fold symmetry assumed in all cases): (1) two 80 mm HX located at the pole angles, (ii) one 110 mm HX at the pole, and (iii) one 110 mm HX at the vertical mid-plane. In all cases, the transfer function is reduced by 1- 1.5%, and the short sample gradient by about 0.2-0.5%. In the third case, there is a small decrease of b6 at the nominal current, while Δb10 and Δb14 are affected by less than 0.1 units.

 

 

3.      Radiation protection considerations for the IR upgrade (M. Fuerstner ppt)

Markus introduced the topic by recalling the legal dose limits and the associated area classification. A first assessment of the dose equivalent rate in the LSS during the 2012 shut down leads to a categorization between "limited stay" and "high radiation" areas. In both cases, all interventions have to be planned according to the ALARA principle, which means that a detailed work schedule must be established to determine the individual dose for all workers. The aim of the planning is to document and demonstrate the job feasibility within acceptable dose levels, and to optimize the activities to reduce the dose.

Markus then showed the results of FLUKA simulations of the residual dose rates in the tunnel, depending on the location and cool down time, from 1 hour to 4 months. The conditions assumed for the simulation are the nominal luminosity (2012 run) achieved with the present triplet. At 30 cm of the TAS axis, the residual dose rate decreases from 70 to 4 mSv/h after 4 months of cooling. For the same cool down times, it reduces from 2 to 0.25 mSv/h at the end of the Q1 cryostat, and from 1 to 0.2 mSv/h. at the front face of the TAN. Monitors are installed at these locations to measure the actual dose rate.

Ranko asked if the dose rate will be less than 50 mSv once the present triplet is taken out of the tunnel. Markus showed on the example of TAN area that the tunnel activation depends on short lived isotopes in the tunnel walls. Paolo remarked that some equipment will remain in the tunnel, for example the QRL, and suggested calculating dose maps for several configurations of equipment in the tunnel. Lloyd pointed out that tooling improvements have to be studied for the work on the interconnections. Ray suggested learning from the LEP dismantling experience.

4.       Estimate of particle fluence in electronics locations (F. Cerutti ppt)

Francesco presented the evaluation of the radiation levels and their effects in the machine tunnel around IR1 and IR5, in particular the total ionizing dose, the 1 MeV equivalent neutron fluence and the high energy (E > 20 MeV) hadron fluence. Following the CNGS run in 2007, when many failures of the electronic devices occurred due to single event upsets provoked by high energy hadrons, the "Radiation-To-Electronics Taskforce (R2E)" was formed to establish, extend and evaluate the inventory of information needed for assessing radiation-induced electronics failures. This dedicated working group started to review the situation for LHC.

UJ56 was modeled with the present layout and FLUKA simulations have shown that this area is one of the hottest in the machine, comparable to the cleaning insertion IR7. Indeed, behind 2 m of concrete shielding in UJ56, the hadron fluence at the beam height is from 1.3 109 up to 1.3 1010 cm-2 y-1. As the detector is not taken into account, it could be thought that the model is conservative. However, Francesco showed that even assuming that the cavern walls are totally absorbing, the result does not change. In fact, high energy hadrons in UJ56 come from interactions located between the TAS and D1. In the upper floor of UJ56, 3 m above the beam axis where the electronics is installed, the hadron fluence is 1.2 - 3.6 109 cm-2 y-1, which is one order of magnitude higher than the radiation level in CNGS for 2 days.

Calculations performed in the past for RR15 and 17, including the shielding made of concrete blocks and thick steel pipe around the beam tube were also presented. The results are equivalent to those for UJ56. On the other hand, there are no simulations for UJ14 and 16, which are quite different as the shielding is removable and the TAS is a few meters closer. He also showed that the holes in the shielding must be filled.

Francesco concluded that the high energy hadron fluence in UJ56 is a serious concern already for nominal luminosity. The topic will be addressed in the R2E Taskforce to plan further studies for the critical areas that are now well identified. These studies will determine the compatibility of future equipment with the hadron fluence in the first years of LHC running. Installation of additional shielding, compatible with the transport zone, may be needed and could be installed during the shut-downs. Moving the equipment in the US15 cavern, which is much better shielded, could be a serious option for sensitive electronics.


R. Ostojic and H. Prin