Superconducting Undulator (SCU) R&D for Free-Electron Laser (FEL) Systems

(Argonne)

Highlights– September 01-30, 2015

The cryostat was partially disassembledand cooled down twice in September (cool-downs #2 and #3) to enable improvements to be tested.Both of the additional cool-downs were extremely successful. A lot of experience was gained with the device and with the magnetic measurement system. The phase error was quite improved after cool-down #2, but degraded after subsequent changes prior to cool-down #3. Plans are underway to improve the clamping uniformity before the next cool-down.

SCUMagnet and Test CryostatFabrication:

The cryostat, shown in Figure 1, was partially disassembled to enable some improvements to be made, including addition of a heater to the helium circuit, replacement of a bent sensor guide tube, and addition of an external magnet structure clamp to improve the gap at the upstream end. After completion of these modifications on September 4th, the device was cooled down for the second time and measurements were made.

Figure 1: First partial disassembly of the cryostat is shown.

After the aforementioned modifications were tested, the device was again warmed up for another set of improvements. Another magnet clamp, shown in Figure 2, was added (making a total of 5 clamps). All clamps were then re-torqued.

Figure 2: The new intermediate clamp is shown.

Second-integral correctors were added upstream and downstream of the magnet. They are located on the beam chamber between the bi-metallic transition and the outer thermal shield intercept, as shown in Figures 4 and 5. The 3rdcool-down began on September 18th.

Figure 3: One of the second-field-integral correctors is shown.

Figure 5: A side view of the cold mass, showing magnet cores with clamps and corrector magnets.

SCU - NbTi Undulator Horizontal Tests and Measurements - 2nd and 3rd cool-downs:

Figure 6 shows the training and quench history of the top and bottom cores in cool-downs 2 and 3. Training was uneventful as can be seen in the figure, and there was a clear decrease in the number of quenches required after each successive cool-down. Figure 7 shows the RMS phase errors after the 2nd and 3rd cool-downs. The improvement after cool-down #2 was maintained after operations carried out prior to cool-down #3 as shown.Figure 8 shows normalized peak fields along the device at the design field of 1.67T at 600A for all three cool-downs.

Figure 6: Training and quench history of the top (orange) and bottom (blue) cores. There is a clear decrease in the number of required quenches over the three cool-downs.

Figure 7: Phase errors for the 2nd and 3rd cool-downs are shown. Improvement was seen after the 2nd cool-down with one new clamp, but either the addition of the 5th clamp or the re-torqueing exercise was not as successful as can be seen in the figure.

Figure 8: Normalized peak fields are shown for all three cool-downs.

Figure 9 shows the vertical field with 2nd-integral correctors energized, and Figure 10 shows the 2nd integral, with and without the correctors energized, as measured with the Hall probe. A full set of field integral measurements will be made with the rotating coil.

Figure 9: Vertical Field along the device is shown.

Figure 10: The 2nd integrals, measured with the Hall probe, are shown with correctors not energized (left plot) and energized (right plot).

Table 1 shows the main design parameters of the NbTi undulator, and Table 2 shows achieved tolerances and performance parameters.

Table 1: Main NbTi undulator design parameters.

Table 2: Achieved tolerances and performance parameters.

All three cool-downs to-date were extremely successful.A lot of experience was gained with the magnet and with the magnetic measurement system. The cryostat waswarmed up twice during this month, for a total of three cool-downs. Quench performance improved each time. Several modifications were made;2nd integral correctors were added and performed as designed. The phase error was improvedafter the 2nd cool-down, but degraded after subsequent changes. Plans are underway to improve the clamping uniformity before the next cool-down.

Issues and Actions

•None at this time.

Schedule Status

•Initial results successfully obtained, delayed by about 8 weeks from the original plan.

90-day Forecast

•Field integral measurements will be made with the rotating coil.

•LBNL magnet arrives at Argonne – now forecast for mid-October.

•The test of the Argonne SCU magnet will be complete; all performance specifications should have been met.

•The Argonne team will be ready for acceptance and start of assembly of the LBNL SCU magnet in mid-October.