.

The Field Quality Crisis Unit of the LHC main dipoles production

Inner radial movement of a conductor block in collared coil 2032

21 may 2003, Manufacturer 2. Warm magnetic measurements of collared coil 2032 (see file) show a spike in b8, b13, a6, a7, a9, a11 (red alarm, i.e. more than 8 sigma) in five measuring positions (3,4, 8 12 and 16) and yellow alarms on most multipoles of aperture 1. Measurement is repeated, confirming the anomaly. First analysis shows that the defect is not following any symmetry, i.e. it is likely to be in one quadrant only. Since high order multipoles are affected, one can deduce that the defect is very close to the cold bore. The order of magnitude of the movement necessary to create this effect is of several tenths of mm. A inner radial movement of block 6 of 0.6 mm could explain the field anomaly in positions 12 and 16.

30th May 2003, CERN. The crisis unit is activated. Two meetings are organized: one on May 30 (presents: W. Scandale, E. Todesco, C. Vollinger, M. Bajko, F. Savary, E. Wildner), and one on June 2 (presents: W. Scandale, L. Rossi, J. Vlogaert, M. Modena, F. Savary, M. Bajko, E. Todesco, E. Wildner). The evidence of a cable mispositioning of 0.5 to 1 mm is given, and it is decided to ask for a decollaring.

18th November 2003, Firm2. After some discussions between CERN and Firm2, the collared coil 2032 undergoes the procedure of disassembling in presence of M. Bajko, F. Savary, A. Musso, E. Todesco and C. Vollinger from CERN. Locking rods are removed at the pressure of 550 bar without problems. Then, collars and coil protection sheets are removed in the zone showing the worst field anomaly. At around 18.00 also the insulation and part of the coil interlayer is removed, allowing to see the pole of the inner layer where the two conductors of block six are detached from inner layer and shifted towards the cold bore by around 1 mm, exactly as foreseen by the multipole analysis.

On november 19th, all collars are removed, the pole are separated and the cold bore is removed. One pole of one aperture shows visible defects along several sections. In most cases the cables of blocks six are detached from the inner layer, and sometimes also the copper wedges is not glued to the inner layer.In all cases the pattern of the defect agrees with multipoels anomalies. A complete report can be found here (doc file). This case also provides a further check of the consistency of signs of the measuring system.