Introduction
During the initial testing of the Renascence cryomodule, thermal issues with the end groups were discovered. Cavity performance was limited by heating in the end-groups. There were no heat straps incorporated into the original design of the Renascence HOM feedthroughs. The HOM cables were the flexible type and heat stationed to the 2 K helium circuit. The cables were the same type used for the SNS cryomodules. The SNS cables had a static heat load of ~250 mW. At the conclusion of initial testing, the cryomodule was warmed up and several end-groups were outfitted with various diagnostics (heaters, extra temperature diodes) and hardware (heat straps) to clearly identify 1) the cause of the thermal instabilities and 2) possible solutions. This technical note describes the follow-up test configuration and compares analysis with test data for cavity #1, HOM can “D” and cavity #3, HOM can “B”.
Test Setup
The following paragraphs detail the modifications made to cavity #1, HOM can “D” (on field probe end of cavity) and Cavity #3, HOM can “B” (on FPC end of cavity). To summarize the modifications, heaters were added to both feedthroughs along with additional temperature diodes and one of the feedthroughs was heat stationed to the 2 K helium circuit.
Cavity #1, HOM “D”
The modifications to cavity #1, HOM can “D” were 1) remove the HOM cable, 2) add a heater, Fig. 1, to the end of the feedthrough and 3) install additional temperature diodes, Fig. 2.
Cavity #3, HOM “B”
The modifications to cavity #3, HOM can “B” were to 1) clamp a copper heat strap, Fig. 3, to the copper sleeve on the feedthrough, 2) add a heater to the heat strap clamp, Fig. 4, and 3) reapply the temperature diode that was located on the copper sleeve to the copper clamp, Fig. 4.
Analysis
The geometry from K. Wilson’s model was imported and material properties were copied and input into ANSYS. New meshes were created for each feedthrough and analyses were run duplicating the test conditions for Renascence. The models for both feedthrough configurations include a variety of materials, Fig. 6, instrumentation labels, Fig.7 & 9, and two distinct sets of boundary conditions, Fig. 8 & 10.
Cavity #1, HOM “D”
FIGURE 6: HOM Can Materials
FIGURE 7: Cavity #1, HOM “D” Instrumentation
FIGURE 8: Cavity #1, HOM “D” Boundary Conditions
Cavity #3, HOM “B”
FIGURE 9: Cavity #3, HOM "B" Instrumentation
FIGURE 10: Cavity #3, HOM "B" Boundary Conditions
Results
Cavity #1, HOM “D”
The results for the analysis on the cavity #1, HOM can “D” are shown in Table 1. For all the load steps in Table #1 the port tube is fixed at 6.0 K (this is based on test data). The applied heater heat is input as a constant heat load. For the baseline static case, analysis was done to determine the heat conducted down the heater leads. FEA indicated a static load of 20 mW for the heater leads. This is in good agreement with the thermal conductivity integral for OFHC copper, which indicates a static load of ~17 mW for the same boundary conditions. For the static case, no applied heater heat and 20 mW conducted down the heater leads, the tip temperature is 8.14 K. With an applied heat load of 9.8 mW (+ 20 mW static load) the tip temperature is 9 K. Any increase in applied heater heat raises the tip temperature above the critical temperature of Nb (9.2 K). The temperature profiles for 0 mW, 9.8 mW and 50.9 mW of applied heater heat are shown in Fig. 11, 12 and 13. Using the temperature data from the analysis, the thermal impedances were calculated. The thermal impedance across the flange pair was calculated to be 25-34 mK/mW. The thermal impedance across the NbTi flange to the port tube was calculated to be 42-48 mK/mW. The thermal impedance across the copper collar to SST flange was calculated to be 15-24 mK/mW. These values can be compared to the measured values of 24, 35 and 9mK/ mW, respectively. The calculated sum of thermal impedances is 82-107 mK/mW, compared to the measured value of 68 mK/mW.
Table 1:Cavity #1, HOM “D” Results Summary
FIGURE 11: Cavity #1, HOM “D” Static Results
FIGURE 12: Cavity #1, HOM “D” 9.8 mW Applied Heat Results
FIGURE 13:Cavity #1, HOM “D” 50.9 mW Applied Heat Results
Cavity #3, HOM “B”
Similar analysis was done on cavity #3, HOM “B”. Based on Cavity #1, HOM “D” results, the static heat load of 20 mW for the heater leads was repeated for these cases. The results are shown in Table 2. The contour temperature plots for 0 mW and 80 mW of applied heater heat are shown in Fig. 14 and 15. Also, several analyses were run with up to 500 mW of applied heater heat with the probe tip remaining below 9.2 K (contour plots not shown).
Table 2: Cavity #3, HOM "B" Results Summary
FIGURE 14: Cavity #3, HOM "B" Static Results
FIGURE 15: Cavity #3, HOM "B", 80 mW Applied Heat Results
Comparison
Cavity #1, HOM “D”
Fig. 16 shows the results from the analysis compared with the test data. The analysis results are within a .25 K of the raw test data up to applied heats of ~50 mW.
FIGURE 16: Cavity #1, HOM "D" Temperature Comparison
Fig. 17 shows a comparison of the calculated and measured thermal impedances. The percent difference between calculated and measured impedances is ~14% for the NbTi flange/HOM body impedance and ~4% for the feedthrough flange pair.
FIGURE 17: Cavity #1, HOM "D" Thermal Impedance Comparison
The potential sources of error in temperatures could be due to the widely varying thermal conductivities of the various materials. Also the static heat load down the heater leads is assumed to be constant, not a function of temperature as expected. Also, there could be a geometric discrepancy between the model and the actual feedthrough. The feedthrough contact areas were modeled based on assumptions about the weld penetration and braze alloy flow length.
Cavity #3, HOM “B”
The results from analysis and testing are summarized in Fig. 18 and 19, respectively. The analysis indicates that the feedthrough temperatures are lower than the raw test data. Material properties and the static heat load from the heater leads were determined from the analysis on the cavity #1, HOM “D” feedthrough which had good agreement between test data and analysis. Fig. 19 shows the stainless steel flange as the warmest part of the feedthru. ANSYS runs were made with up to 1 W of static heat conduction on HOM cable and the test data could not be duplicated. Possible causes for the discrepancy in temperatures could be the heat stationing of the diode lead wires and poor thermal contact between the copper heat strap clamp and copper sleeve on the feedthrough. There are also variations between end-groups with respect to the flow of braze alloy in the feedthroughs and weld penetrations between various parts of the end-group assemblies. Fig. 20 and 21 compare the measured and calculated values for thermal impedance for cavity #3, HOM “B”.
Summary
Static analyses were done for two feedthrough configurations. The analyses on cavity #1, HOM “D” are in good agreement with the test data. These results verify that good heat strapping is required to keep the Nb probe tip superconducting. The analysis on cavity #3, HOM “B” is complicated by the additional variables of the HOM cable (static heat load) and heat stationing (thermal contact area). The analysis is further complicated by the opposing influence of these two variables on the thermal performance of the feedthrough. The conclusion drawn from both analyses are: 1) the thermal conduction path to the 2 K circuit must be improved and, 2) the thermal performance of the HOM cable must be optimized and 3) better quality control is needed in the fabrication of the feedthroughs. A robust thermal strap that has good thermal contact with the feedthrough copper collar and the 2 K header pipe will satisfy requirement #1. Proper selection and heat stationing (if needed) of the HOM cable is needed to satisfy requirement #2. The variation in thermal performance of the feedthroughs could be caused by inadequate flow of the braze alloy between the components of the assembly. Better quality control during fabrication of the feedthroughs could improve the thermal contact and performance.
References
Wilson, K., Thermal Analysis of HOM Feedthroughs, JLAB-TN-04-022
Reece, C., Renascence Issues and Thermal Testing Analysis, JLAB-TN-06-003