28 Feb. 1996 G. Bunce g-2 Ring News ------------- The ring is warm, we have surveyed it, and analyzed the data we have. We believe we have a consistent understanding of our observations and calculations, although this is a work in progress. This is the work of many. To follow the following, recall that the outer coil hangs on 16 G10 straps. The straps are held on the outer diameter of the cryostat (room temperature), and attach to the coil below the midplane (liquid helium temperature). The straps swing radially up as the coil is cooled, to an angle of about 9 degrees when the coil is cold. When the coil is cooled it shrinks 3 cm in radius, and onto 4 radial stops which push from the inside radius. The stops are set to be just engaged when the coil is cold and at full current. When the coil is cold and unpowered, the stops push 2 mm into the coil, and the coil takes a scallop shape. Half-way between the stops, the coil radius will be about 4 mm less cold and unpowered. From radial field measurements, which are sensitive to the coil position vertically and radially, the coil is above the midplane by an average of 1 mm at low current, and one region (the kicker location) moves about 5 mm above the midplane at 3/4 of full current. It is at this time that we see the large voltage difference between the two outer coils (Delta V), which triggers energy extraction. The radial field measurements, when compared with a computer model of the magnet, also show that the coil moves radially inward at the kicker location by 2 cm at this time. This is a huge amount and we tended to dismiss this when Sergei Redin first presented it. A strain gauge was placed below the outer coil cryostat to see if the coil indeed levitates. It does, beginning just above half current. There is a large step up when we reach 3/4 current. This step, in Delta V, integrated over the time, corresponds to a movement of about .1 mm up. (The vertical force from the field is unstable about the midplane. Our design was for the coil to be set slightly below the midplane, so that the force from the field would always be in the same direction as gravity. We included side plates on the straps that hold the coil for an upward force.) With the coil warm, we confirmed that the coil was placed as planned, about 2.5 mm below the midplane. We also observed that the radial stops rubbed/scraped on the mandrel horizontally 1-2 cm. The marks at the 4 stop locations indicate that the coil took an oval shape with the coil in the kicker location moving inward. (A 1 cm horizontal movement at the straps would represent a very large radial movement at the center. It is doubtful that we are correct about the horizontal scraping.) When the coil is cold and unpowered, the straps will lift the coil between the stops where the coil radius is smaller due to the scallop shape. There is also an effect from the cryostat loading which lifts the coil at these locations. These effects were not included in the vertical offset of the coil. We believe that the coil would be roughly 1 mm high cold and unpowered between the stops. This is the Prigl effect. It is consistent with the low current radial field measurements. There is an unstable horizontal force from the field, with the force larger outward for a larger radius. This will take the coil into an oval shape (with 4 stops) as soon as the magnetic force is greater than the restoring force from the bending of the mandrel. This will occur for any small difference in initial radius of the coil around the azimuth. Francis Farley has made some initial calculations of this effect, and, ignoring the friction from the straps and stops, the calculation shows that this would occur at roughly 1/4 to 1/2 full current. The friction seems sufficient to raise the onset to above 1/2 current. We believe that this is occuring. The sounds and small Delta Vs that we hear and observe beginning above half current are from the coil shifting to oval from the scallop shape. At 3/4 current the coil takes the oval shape. At the smaller radius locations (the kicker position, and opposite) the straps raise the coil several mm, and we observe a large Delta V, and we observe the levitation from the larger vertical force seen by the coil several mm above the midplane. (At the large radius locations the coil is pushed further below the midplane by the straps. However, the net effect is a higher average coil radius due to the angle of the straps. The net effect corresponds to only a .1 mm rise, but the local effect is several mm.) We did not anticipate this effect--the coil changing from scallop shape to oval. We are developing a plan to handle this effect. We are now evaluating whether 8 radial stops will be sufficient. It is clear to us that the coil should be lowered about 2 mm, since the coil has been observed to be high by 1 mm at low current, and our evaluation of the coupling between the straps and stops indicate that the coil will be above the midplane by about 1 mm between the stops. We are also considering adding windows to the cryostat to monitor the coil position optically. In order to lower the coil, we will need to remove the top yoke. This will give us access to all the straps and we will test their integrity (we believe that they can undergo a 2 cm azimuthal rotation without damage; we need to verify this). We need to quantitatively compare this model--for instance, can we make a reliable measurement of the horizontal scraping on the mandrel? We need to study the inner coils. We have an inconsistency at 1000 amps. At low current (1000 amps), the radial field indicates that the coil is about 2 mm high at the kicker location. The strain gauge placed under the cryostat at that location increased slightly in strain until about 2600 amps, then the strain reduced, implying that the coil was below the midplane at 1000 amps and only went above the midplane at 2600 amps. This is a work in progress.