02/22/08
A 2-Cell Temperature Mapping System for ILC Cavities
G. Ciovati, R. Flood, P. Kneisel, D. Machie, M. Morrone.
Introduction
High power RF tests of 9-cell cavities for the International Linear Collider (ILC) at Jefferson Lab showed significant variations in the quench field above Eacc = 25 MV/m by repeating the same surface preparation procedure in some cavities, and reproducible low quench field (Eacc ~ 17 MV/m) values in others [1]. The first step to understand the cause of the quench is to identify the location on the cavity surface where the quench occurs. Temperature mapping of the cavity surface has been a useful tools to characterize anomalous losses in superconducting RF cavities for more than 15 years and thermometry systems have been successfully designed and implemented at Cornell, KEK, DESY, Orsay and most recently at Jefferson Lab. Most of these are fixed systems applied to single-cell cavities and consists of about 600-700 Allen-Bradley 100 W RTDs distributed over the cavity surface. DESY has developed a rotary system to map 9-cell cavities with a limited number of thermometers [2]. The disadvantages of the rotary system are: complications due to a rotary frame at cryogenic temperatures, poor thermal contact of the thermometers to the cavity surface in superfluid helium, relatively long time required to obtain a full map of the cavity. These complications reduce the reliability of the system, compared to a fixed one.
A compromise between the need of a thermometry system for 9-cell cavities and limited resources lead to the decision to build a 2-cell system which would cover the equatorial region of two of the nine cells of an ILC cavity, which had been identified in a prior high power RF test by measuring all TM010 pass-band modes up to the quench limit.
Details of the system
In order to limit the number of thermometers and to share the data acquisition system already in place at JLab for a single-cell temperature mapping system [3], we proposed to build a system which would cover the equator (high-magnetic field region) of the two cells identified during the first RF test. Each equator is covered with 160 thermometers, divided in 10 semi-circles. The thermometer spacing is about 2.2 cm along the azimuth and about 1.5 cm along the meridian. The thermometers are built using Allen-Bradley 100 W carbon resistors, as they were developed at Cornell and successfully used in most thermometry systems [4, 5].
An aluminum frame bolted to the stiffening rings between the cells holds the thermometry cards on the cell. Each thermometry card covers half of the cavity circumference and consists of a 1/8” thick printed circuit board (PCB) with copper traces on one side. 16 thermometers, equally spaced, are connected to the board in the following way: for the boards which will be installed on the equatorial weld of the cavity, the thermometers are inserted in receptacle holes directly drilled in the PCB board, while for the ones above and below the equator line, they are inserted in G10 rings, bolted to the PCB board, with holes drilled at an angle to maintain the thermometer normal to the cavity surface. Set screws at the edges of the side of the aluminum frame push the thermometry cards against the cavity surface. Figure 1 shows a mechanical drawing of the assembly for one cell.
Figure 1: Mechanical drawings of the assembly of the thermometry system on one of the 9-cells of an ILC cavity.
Flat ribbon cables connect the thermometers to a feedthrough box on the top-plate of the vertical test stand which holds the cavity for the cryogenic test at 2 K. Round shielded cables connect the feedthrough box with the National Instruments data acquisition system used for the single-cell thermometry system [3].
Figure 2 shows a picture of a thermometry card while pictures of the thermometry system assembled on a 9-cell ILC cavity are shown in Fig. 3.
Figure 2: Thermometry card covering half circumference of a cell of an ILC cavity.
Figure 3: 2-cell thermometry system assembled on an ILC cavity.
Commissioning
For the first test, the thermometry system was assembled on cells #2 (top one in Fig. 3) and #8 (bottom one in Fig. 3) of an ILC cavity built at JLab from fine-grain RRR>300 Nb. The thermometers were calibrated while lowering the temperature between 4.3 K and 2 K. The result from the high power RF test in the TM010-p mode is shown in Fig. 4 and the temperature map of the two cells at the highest field before the quench is shown in Fig. 5. Unfortunately, the quench in the p-mode did not occur in either cell with thermometry system attached.
Figure 4: Q0 vs. Eacc measured for the ILC cavity built at JLab with the thermometry system assembled on two cells.
Figure 5: Temperature map at the highest field before the quench, showing a “hot-spot” in cell #2. Thermometers number 3 and number 8 are on the equatorial weld of the respective cell while the azimuthal position 0 to 31 covers the whole circumference of the cell. As it is explained in the text, it was found after the test that the field profile was “tilted” with higher field in cell #2. The peak surface magnetic field, Bp, for this T-map is about 118 mT in cell #2, while it is about 93 mT in cell #8.
High power RF measurements, along with temperature maps, were done on the other modes of the TM010 pass-band. Quenches at different locations in cell #8 were found in the 3p/9 and 4p/9 modes, as shown in Fig. 6.
Figure 6: Temperature maps at the highest field before quench in the 3p/9 mode (a) and 4p/9 mode (c) and heat pulse captured during quenches in the 3p/9 mode (b) and 4p/9 mode (d). The Bp value for the T-map in (a) is ~128 mT in cell #8 and ~120 mT in cell #2, while it is ~ 90 mT in cell #8 and ~ 95 mT in cell #2 for the T-map in (c).
In a cavity with a field flatness of 100 %, the field amplitudes in cell #2 and #8 should be the same, but the fact that the T-maps showed much higher number of hot-spots in one cell than the other, made us suspect that the field profile was not flat. In fact, after the cavity was warmed up to room temperature and disassembled, a bead-pull measurement showed a field flatness of only about 72 % in the p-mode as shown in Fig. 7.
Figure 7: Stored energy in each cell for the p mode measured at room temperature, after the RF test at 2 K.
The temperature increase DT above the He bath temperature as a function of the cell’s Bp in the 3p/9 mode is shown in Fig. 8 for a number of hot-spots. The heating observed at these locations is consistent with the strong heating frequently observed in bulk niobium cavities above about 100 mT, which cause the so-called “Q-drop” and whose origin is not yet understood. The Bp values in the cells were obtained by multiplying the square-root of the measured cavity stored energy, U, for a certain mode by the coefficient Bp,cell/ÖU calculated in [6]. Figure 9 shows DT vs. Bp for hot-spot 30-2 which is consistent for the various modes.
Figure 8: DT vs. Bp measured at 2 K for a number of hot-spots in the 3p/9 mode. The hot-spot locations are identified by two numbers, the first one being the azimuthal position and the second one being the thermometer number.
Figure 9: DT vs. Bp measured at 2 K for hot-spot 30-2 (in cell #2) for different modes of the TM010 pass-band. Note the logarithmic scale.
Conclusions
A thermometry system, which covers the equator area of two cells of a ILC 9-cell cavities has been designed, built and tested at Jefferson Lab. The system was designed with a limited number of thermometers in order to share the data acquisition system used for the single-cell thermometry system already in place at Jefferson Lab, while still providing sufficient diagnostic capabilities. The system worked as expected during its first test and the results pointed to a non-flat field profile, which was verified after warm-up. Quenches at 3 different locations of one cell were detected in two modes of the TM010 pass-band.
Acknowledgements
We would like to thank P. Kushnick for cryogenic support during the cavity test and D. Forehand for the bead-pull measurements.
References
[1] R. Geng, “Update of JLab results on ILC 9-cell high-gradient R&D”, TTC Meeting, DESY, January 14th-17th 2008, https://indico.desy.de/conferenceOtherViews.py?view=standard&confId=401
[2] Q. S. Shu et al., “An advanced rotating T-R mapping & its diagnosis of TESLA 9-cell superconducting cavities”, Proc. of the 1995 PAC, p. 1693.
[3] G. Ciovati et al., “Temperature mapping system for single cell cavities”, JLab Tech Note TN-05-59 (2005).
[4] G. Müller and P. Kneisel, Cornell University Report No. SRF 851291 EX, 1985.
[5] J. Knobloch, Ph.D. Thesis, Cornell University, 1997.
[6] H. Wang, “TM010 Pass band modes of TESLA 9-cell cavity”, JLab Tech Note TN-07-052 (2007).