The solenoid that will provide the magnetic field for the AT-TPC is a used medical MRI device that was most recently utilized in the TWIST experiment. TWIST has now been decommissioned and an agreement was reached between NSCL and TRIUMF to allow the solenoid to be moved to NSCL for the AT-TPC project. The purchase cost of an equivalent system is estimated to be in excess of $600,000. The properties of the TWIST solenoid are superior to those required by the AT-TPC project. Detailed characteristics of the device are shown in the table below. NSCL is currently in the processes of incorporating the superconducting solenoid into the laboratory cryogenics system. The TWIST solenoid is shown in the photos below in the NSCL assembly hall where it is being tested.

The TWIST solenoid and yoke were loaded for shipment to NSCL on September 16, 2008. The delivery arrived at NSCL the morning of September 22nd.

It was observed that the Shockwatch sensors had been tripped prior to unloading the solenoid. All items were unloaded the morning of September 22nd and the solenoid was kept in its original packaging. The shipping company was notified of the status of the Shockwatch sensors on September 23rd. On October 13th the shipping company approved the unpacking and testing of the solenoid.

Visual inspection of the solenoid after unpackaging revealed no damage evident to the eye. The most likely form of damage is to the internal structure, namely dislodgement of the rods. Without opening the cryostat this would not be visible. The best way to determine if misalignment or dislodgement has occurred is to map the field alignment. This requires operation of the solenoid.

Connections have been established between the cryo teststand and the TWIST solenoid in the East Highbay area. The cooldown is being using the 2500L white dewar. LN2 is used to cool the outer chamber of the magnet while LHe is used to cool the inner chamber. The LN2 cooldown can be monitored via the level meters and gauge in the power supply rack. The LHe cooldown can be monitored via a voltmeter connected to a carbon resistor that is recessed within the magnet. At temperatures in excess of 4K the resistance is maintained at ~80 Ohm, but will jump to 1000 Ohms when cooldown is complete. The transfer of LHe from the dewar to the magnet is conducted via a pressure differential. Due to the low pressure tolerance of the He chamber (5.5psi) the rupture valve must be maintained at 5psi. This results in a narrow pressure differential window for transfer between the dewar and solenoid making the fill process slow.

November 10, 2008: (McCartney)
We have pumped and purged the helium vessel 3 times. I backfilled the vessel with helium this morning. I started the liquid nitrogen to the shield around 7:30 am this morning Monday 10NOV08. We installed the helium supply U-tube to the “white” Dewar this morning. I pressurized the Twist helium container and started to open the helium supply control valve on the Dewar. The Dewar pressure was set to 7.5 psig and the valve opened 20%. The pressure started to slowly rise from 2 to 3.0 psig on the Twist. I set the valve to zero and let the pressure drop back below 2psig. I then set the feed valve to 10% and the pressure stayed low indicating the feed valve, although set to 10%, was still closed. I set the valve to 18% and the Twist pressure started to rise again slowly. When it got to around 3 psig I reduced the valve to 17%. I think I was a little late and the pressure spiked to a reading of ~4.5 psig and the carbon rupture disk burst. I replaced the disk and re-pressurized the vessel to the return pressure of about 1.2 psig. I am lowering the supply Dewar pressure to a range between 3.7 psig and 4.85 psig. I will open the feed again today and try to establish a steady flow of helium. I restarted the cool down at 1 pm, with the supply pressure in the Dewar set to a maximum of 4.8 psig. The pressure in the TWIST solenoid has stabilized at just over 3 psig with the supply valve open 16.3%. I am going to leave it in this state overnight. The Cyclotron Operators are monitoring the pressure and they will notify me if the disk bursts. I have been feeding LN2 to the shield for about 8 hours now and I will shut that off for the night and restart it when I come to work tomorrow. We are monitoring a carbon resistor in the helium cryostat that will increase resistance as it cools. It has been steady 83 ohms all day. It will read closer to 1,000 ohms at LHe temperature.

November 11, 2008: (McCartney)
reached LN2 temperature stabilization

November 21, 2008: (McCartney)
reached LHe temperature stabilization

November 25, 2008: (DeLauter)
field ramped to 200G

November 26, 2008: (DeLauter)
map symmetry of field in horizontal and vertical directions

December 1, 2008: (DeLauter)
warm up begun

Cooldown Results:
The field symmetry data taken on November 26 was placed in two graphs. This was done to visualize any deviations in the magnetic field. If one of the brackets that hold the coils had been damaged a non uniform magnetic field would result. This would present itself as slope in the magnetic field strength. Fortunatly the graphs show very little deviation in the field as you move horizontally or vertically within the central field of the magnet. The total deviation was less than 1%, which matched previous magnetic field data from TRIUMF. The deviation present in the lower half of the vertical field plot is not a reflection on the magnet itself. To ensure accuracy in the measurment of the position, a aluminum guide was placed inside of the magnet. This was attached to a track used to move equiptment within the magnet by steel fasteners. This steel within the magnet caused fringe effects, which were measured as deviations in the field.


Vertical Field Strength


Horizontal Field Strength



Cryogenic Fluid Use
The level of liquid nitrogen and liquid helium were measured throughout the cryogenics testing. This information was useful for determining the rate at which coolant will be used by the magnet during operation. The two peaks in the liquid nitrogen graph indicate refilling to 100%. These fillings occured on the 10th and 20th of november. Fittings of the first filling slope and the second filling slope were done to determine the rate of coolant loss with and without liquid helium. These values were found to be 1.6L/h without liquid helium, and 1.11L/h with liquid helium present. The level of liquid helium shows two distinct slopes. The initial 5 points were taken under stable conditions which yielded one slope. The second slope is much larger and is due to heat from the preparation of the magnet. It was the stable slope that was fitted, and the boil off rate of liquid helium was determined to be .243L/h.

The field uniformity of the magnet is maintained with the use of a steel yoke surrounding the magnet. Shown in the photo below is the TWIST detector mounted in the experimental hall at TRIUMF. The yoke is shown in purple surrounding the magnet. This same yoke will be used in the AT-TPC project. The physical parameters of the yoke are shown in the table below.

The mapping of these fields is a important step that both feeds into our calculations as well as providing a safety boundary for magnetically sensitive items. Below is a presentation describing the steps we have taken to analyze the magnetic field, and their results.

:solenoid_magnetic_variability.ppt

The final magnetic field data write up.

:final_magnetic_field_analysis.doc

The AT-TPC solenoid and yoke were moved into the ReA3 hall on June 22, 2009. Because the floor plan for the building has not yet been finalized, the steel & magnet have been stacked on the East side of the ReA3 hall pending assembly in their final resting place. To complete the move of the items from the east highbay where they were stored to the ReA3 hall required two cranes and a flatbed truck. These were rented from Connelley Crane Rental on an hourly basis. A quote for their services is posted here. Four hours were required for the move. Photographs of the move are posted at this link. One of the key features to observe is that due to the low ceiling height the cranes cannot extend to their full height and thus cannot lift as much weight as their capacity allows (~30k lbs). For this reason two are required to move each item. The photographs document the floor space needed by the cranes to manipulate the solenoid into place. This needs to be taken into consideration in the ReA3 floorplan to allow the yoke to be disassembled and moved out of the ReA3 hall in the future.