Sections: Divers: .
|
A missing shim in magnet 1027
29 october 2002,
Manufacturer 1. Warm magnetic measurements of collared coil 1027 (see file)
show a spike in b2 (red alarm, i.e. more than 8 sigma) and a yellow alarm in
b3, a2 (yellow alarm, i.e. between 4 and 8 sigma) in position 8 of aperture 1.
Position 7 is in between - this seems to indicate that the problem is not
related to
the measure, but that something wrong is in the magnet. The error is in
all multipoles, i.e. no symmetry is followed: this means that it should be
in one quadrant only. Moreover, only the low order multipoles are affected:
this means that the source of error should be far from the aperture (i.e.
outer layer). Simulations by E. Todesco on November 1 show that a missing shim
of 0.8 mm fits well the numerical data; A. Devred confirms these results
using a different code (see file)
6 november 2002,
Manufacturer 1. The measurement is repeated, giving similar results.
7 november 2002,
CERN. Meeting of the crisis unit: J. Vlogaert, W. Scandale, M. Cornelis,
J. Rinn, C. Vollinger, E. Todesco, P. Fessia, J. Miles, A. Devred.
The case is illustrated by E. Todesco and A. Devred. Even though the effect
is much weaker than for 2002, the numerical evidence of a missing shim is
strong. Foreseen location of the missing shim: aperture 1, right side, lower
pole, outer layer. The decision of decollaring the magnet is taken.
Wide discussions about the necessity of replacing the mass insulation and
quench heaters, as indicated in the specification. W. Scandale asks for a
double measure of field quality if the cold bore will have to be re-worked due
to deformations.
12 november 2002,
Manufacturer 1. Presents: M. Cornelis, J. Miles, J. Vlogaert. Collared coil
1027 is decollared. A missing shim along 1 m is found where foreseen by the
analysis of magnetic measurements. All the insulation and quench heaters, are
replaced.
18 december 2002,
Manufacturer 1. magnetic field of collared coil 1027 is measured (see file);
the spike in aperture 1 due to the missing shims has disappeared. This second
measurement is performed with the new system. A spike in the main field of
about 8 units in one position of aperture 2 is visible. For this reason the
measurement is repeated (see file);
the spike in main field is disappeared as it was related to a faulty
measurement.
In the positions not affected by the missing shim, field harmonics showed a
variation between the first and second decollaring similar to what
already observed in Firm 2 (see related web page).
It is interesting to note that this collared coils has been refused for
field harmonics variations much smaller than the ones in 1006 (see
file).
In the case of 1006, we had indications that the origin of the field variations
were a coil waviness in all the poles of the order of 0.1 mm. This was
considered not dangerous from the point of view of quench performances. Indeed,
the magnet had no quenches in the straight part. In the case of 1027 the
harmonic variations were traced back to a problem on one side of a pole only,
i.e. a missing shim of 0.8 mm, and therefore it was clear that a decollaring
was necessary. This example shows the impossibility of having a
specification with ranges for multipole variations along the axis
for collared coil acceptance: indeed, each case must be treated separately
and the collared coil is refused only when there is a strong evidence of a
faulty assembly procedure or component.
|