The grand descent has begun for CMS

Until recently, the CMS experimental cavern looked relatively empty; its detector was assembled entirely at ground level, to be lowered underground in 15 sections. On 2 November, the first hadronic forward calorimeter led the way with a grand descent.

The first section of the CMS detector (centre of photo) arriving from the vertical shaft, viewed from the cavern floor.

There is something unusual about the construction of the CMS detector. Instead of being built in the experimental cavern, like all the other detectors in the LHC experiments, it was constructed at ground level. This was to allow for easy access during the assembly of the detector and to minimise the size of the excavated cavern. The slightly nerve-wracking task of lowering it safely into the cavern in separate sections came after the complete detector was successfully tested with a magnetic field at ground level.
In the early morning of 2 November, the first section of the CMS detector began its eagerly awaited descent into the underground cavern. This was a long day for the team from VSL, the main contractor of the lowering operation. 'Rendez-vous was given for 6am this morning and it will take twelve hours to complete the operation,' said Hubert Gerwig, CMS project leader for the lowering. 'The first piece being lowered is the Hadronic Forward Calorimeter (HF+). There are two of them; the second one (HF-) will be lowered next week.' Each HF section is approximately 7x5x4 metres in size and weighs 240 tonnes.

You may imagine the CMS detector as a loaf of sliced bread, cut into 15 slices of different sizes. The two HF sections are the end pieces; the slices in between will be lowered sequentially according to their positions in the 'loaf', starting from the HF+ section at the far end of the cavern, towards the access shaft at the opposite end.

During the lowering process, each corner of a detector section is attached to a bundle of 55 cables, which is in turn connected to a piston fed by a large reel of steel cables. The lifting machinery is enclosed in two small buildings on the roof of the main building, 24 metres above the shaft opening. Four pistons work together to transport each detector section into the cavern in a step-by-step descent. There are two gripper plates, one above and one below each piston, which alternately open and close to hold the bundle of cables in position while supporting the weight of the detector. During the descent phase, the gripper plate above a piston fastens to lock in the cables, and the plate below opens to release them. The plug of the piston moves vertically from its highest position downwards into the cylinder, thereby lowering the detector suspended below. After the plug has descended 50 cm into its lowest position, the gripper plate below the piston fastens to hold the load in a stationary position. The plate above then releases its grip, before the plug moves upwards to repeat the cycle. There are 20 cycles in an hour, which corresponds to an average descent speed of 9 metres per hour in practice. A third building on the roof houses the control room for the lowering operation. Inside, a controller sits in front of a monitor, constantly checking the tilt of the load to maintain an even weight distribution.

The second HF was safely lowered into the cavern on 9 November (see cover photo). After the two HF sections, the end cap sections will enter the cavern once the authorisation for lowering heavier loads (>300 tonnes) is obtained. The team hopes to lower one section a week after Christmas, with the largest and heaviest central section, YB0 (yoke barrel zero), weighing 1920 tonnes, to descend in mid-February 2007. Hubert is looking forward to the operation: 'every time a piece is lowered, its landing will be a special moment, just like a moon landing!'

CMS magnet: Final current ramp-up before the descent into the cavern

Just as the first part of the CMS detector was being lowered into the experiment cavern, the magnet was undergoing its last magnetic ramp-up, bringing the test campaign to a highly exemplary conclusion. A 19,120 amp current was injected into the superconducting coil for the last time (see screen-shot of test monitor).

Throughout the past three months of tests, a maximum magnetic field of 4.13 tesla has been obtained on the inner periphery of the coil, thereby allowing a 4-tesla field to be obtained along its central axis. This field was maintained for three days. 'The behaviour of the magnet was very stable,' notes with satisfaction Domenico Campi, speaking on behalf of all the CMS Magnet and Integration Group. 'All the auxiliary components, which were the most sensitive to the stray field, also proved to be very stable.'

During this test campaign, magnetic field mapping was performed at various magnetic intensities (2.0, 3.0, 3.5, 3.8 and 4.0 tesla) to a precision of 10-5. Once the field mapping was completed a quench was induced in the coil to bring the temperature up from -269°C to -200°C (approximately 70 K). The magnet is now being gradually heated back up to room temperature to be prepared for its descent into the experiment cavern. Every night a 300 amp current is injected in order to generate heat inside the coil. The coil and the central part of the CMS detector will be lowered into the cavern in the early part of next year.