Minutes of the 16th LHC Insertions Upgrade Working Group held on 5 June 2008

Homepage

Present: A. Ballarino, F. Butin, F. Cerutti, S. Fartoukh, M. Giovannozzi, P. Fessia, J. Kerby, N. Kos, J.-P. Koutchouk, P. Lebrun, A. Mereghetti, K. H. Meb, Y. Muttoni, D. Niesbet, R. Ostojic, A.L. Perrot, H. Prin, S. Roesler, L. Tavian, J.P. Tock, E. Todesco, R. Van Weelderen, E. Wildner, L. Williams

Invited:  D. Macina, T. Taylor, L. Walckiers

    1.    Present situation and integration issues in the MS areas (Y. Muttonni pdf file)

Yvon started his presentation by recalling the proposal to move the Q4-D2 by about 16 m and Q5 by 10 m towards the arc in IR1 and IR5. He presented the current layout in the half cells around Q4, Q5 and Q6, and photos showing the equipment in the beam line (vacuum, TAN, collimators, beam instrumentation and roman pots), and services in the area (DSL, shielding, cable trays, electrical power boxes, etc). He then reported on the impact of the displacing the magnets, other equipment and services. The situation and envisaged solutions are the same in both insertions (except for an additional roman pot in CMS).

Two cases were considered:

·        Keep the present QRL service modules where they are and link the magnets using a cryo-link. This type of solution already exists in IR3 and requires a rotation of the jumper by 90 deg. In addition, vertical and/or horizontal steps may be needed to go around the existing equipment that cannot be displaced. The link has to be supported from the tunnel vault.

·        Adapt the QRL, i.e. move the service modules to the new position of the magnets, which implies manufacturing of new QRL pipe elements. The cable trays also have to be adapted to follow the vault at the new locations of the service modules. Some modifications of the cooling and ventilation may also be necessary.

Tom proposed to connect the QRL to the stand-alone magnets using cryostats extensions. In this case, however, space is lost for the warm equipment. Ranko asked what are the plans for the Roman pots after 2013. Daniela said that the stations are necessary for high-b-optics, and that the station at 220 m in IP5 must be preserved. ATLAS also demands to keep this station as part of the FP420 program.  

Yvon said in conclusion that more studies are necessary to evaluate the modifications in cable routing (powering, cryogenic and beam instrumentation, etc.). He also pointed out the importance of logistics that has to be implemented so that all work can be completed in time, in areas which will be irradiated after machine operation.

2.        Parametric study of energy deposition in the triplets (E. Wildner ppt file)

Elena reported the results of the energy deposition in the triplet as a function of the triplet aperture. A symmetric triplet layout has been assumed with an aperture of the quadrupoles in the range of 90-140 mm (the thicknesses of the cold bore and beam-screen have been kept constant, 3.4 mm and 2 mm respectively, without the additional liner in Q1). All relevant quantities are shown to decrease linearly with the triplet length. For example, the total power decreases by about 16 % for apertures of 90 mm and 130 mm.

The spectrum of particles is in all cases very similar. The peak neutron fluence was estimated to be 1016cm-2 at a luminosity of 2.5×1034. In view of this value, Karl-Hubert insisted that special care should be taken in the choice of insulating material, in particular the cable insulation.

3.        Cryogenic issues for the IR upgrade (R. van Weelderen ppt file)

Rob presented the cryogenic options for the beam screen and magnet cooling for the new triplet. He first recalled the preliminary layout with the triplet length and the estimated loads. For the beam screen cooling, several operating temperature ranges are possible. The number of capillaries required, their size and the required mass flow were compared. Taking into account the regulation margin and the thermal load to the cold mass, Rob concluded that the best option is to use the existing scheme and the temperature range of 5-20 K.

Ranko noted that the curves defining the beam screen capillary diameter as a function of the heat load are flat and asked if the conclusion was still valid if the nominal power was less than 4.4 W/m taken in the calculations. Rob replied that a slightly different pressure range would then have to be envisaged, and pointed out that there is no real argument to operate at 60 K from the cryogenic point of view. The question was raised concerning the vacuum requirements. Nicolaas answered that operating at 5-20 K reduces the gas density by a factor of two compared to 40-60 K, and that temperatures in between are of no interest. Nicolaas asked if the heat load was homogeneous, and if not what influence does it have on the cooling circuit. Rob answered that the cooling flow has to be optimized so that Q1 is at the end of the beam screen cooling circuit.

Rob then presented the two possibilities for the 1.9 K heat exchanger, an internal one, similar to the bayonet heat exchanger in the main arcs, and an external, like the one in the present triplets. The assumption is that the local heat load can be up to 30 W/m and the total power 400 W. The heat exchanger size is such to limit the vapor velocity to bellow 5 m/s. The solution is then to have one ID 100 mm tube or two ID 71 mm tubes, with a preference for the first case as it gives better flow control and easier interconnections.

The presentation then turned to the issue of the required conduction area necessary for exchanger control, which is about 220 cm2 with local widening to 400 cm2. An additional free path of about 50 to 150 cm2 has to be provided in magnet yoke in case of a parallel heat exchanger. Philippe questioned the approach adopted, where a significant length of the exchanger remains wetted. Laurent suggested that localized cold source at the interconnects should be checked as an alternative. Ranko asked what are the temperature drops in the two cases and Rob confirmed that an internal exchanger gives a higher temperature margin. It was also remarked that a large external exchanger may have an influence on the off-center of the cold mass in the cryostat, which may not be compatible with the beam position.

In conclusion, a single heat exchanger seems to be the most appropriate solution for the new triplets. Control issues will be further discussed.

4.        Cold Powering of the Triplets (A. Ballarino pdf file)

Before reporting on the status of the conceptual design, Amalia commented on the possible powering schemes and recalled the function of the cold powering, which consists of a feedbox, superconducting link and an interface box.

Amalia then presented a schematic diagram of the interface box between the cryo-magnets and the SC link. It consists of a 1.5 m long cylinder, OD ~350 mm, and serves to connect the Nb-Ti bus bars to the MgB2 cables. The interface box is also used to provide the cold helium to the link and the feed box. Rob suggested that the splices, foreseen on the schematic diagram at 4.5 K, could be made in the 1.9 K volume to save an extra helium supply and an additional plug.

The cold link consists of four concentric corrugated tubes, the inner one of about Ø60 mm containing eight 13 kA cables made out of MgB2 for the quadrupoles, as well as the appropriate number of corrector cables, and the outer one of about Ø200 mm. The link also contains an intermediate shield at 50-75 K. The corrugated structure is commercially produced by Nexans and is sufficiently flexible to be transported on a drum and installed as a complete unit. A preliminary estimate is that a length of about 50 m would be reasonable.

Amalia then presented the design of the SC cable using MgB2 strands. Test samples of the strands have been produced by Columbus and tested at CERN. This material offers high current density with good mechanical properties at a reasonable cost, and is better suited than some other HTS materials. Using MgB2 for the cable means that it can be cooled in a temperature range from 5-15 K, thus avoiding liquid helium cryogenics at distant locations. The cable is fully transposed and is composed of 36 strands, grouped in six assemblies each containing six twisted strands with a Cu or Fe core. The core of the cable is a helical tube, carrying the instrumentation wires.

The feedbox will also be simplified in this case as the current leads will be fed with 20 K helium gas directly from the link. The temperature of the link and of the bottom of the current lead is controlled by helium flow through each lead.

Philippe remarked that electrical connections in helium gas have to be very well designed and characterized. Laurent remarked that the flow calculations in the link have to be checked. Rob suggested that the return of the 50-75 K shield cooling is made back to the QRL via the link, rather than to the WRL as suggested to minimize issues with high pressures. Ranko asked how are the thermal contractions managed. Amalia gave answers to these questions and pointed out that a number of issues need to be solved and validated in tests, which are planned in the coming months. A final test of a link of about 10 m could be made towards the end of 2009.

S. Fartoukh, R. Ostojic and H. Prin