Uc20b MIT 12 TESLA COIL TEST RESULTS Steeves, MM

Home , Tesla coil

PFC/RR-85-12 DOE/ET-51013-155 UC20b

MIT 12 TESLA COIL TEST RESULTS

Steeves, M.M.; Hoenig, M.O.

July 1985

Plasma Fusion Center Massachusetts Institute of Technology Cambridge, Massachusetts 02139 USA

i MIT 12 TESLA COIL TEST RESULTS

M.M. Steeves and M.O. Hoenig MIT Plasma Fusion Center

ABSTRACT

Test results from the MIT 12 Tesla Coil experiment are presented. The coil was tested in the High Field Test Facility (HFTF) of the Lawrence Livermore National Laboratory in October 1984 and January 1985. The experiment measured the performance of an Internally Cooled, Cabled Superconductor (ICCS) of practical size, intended for use in magnetic fusion experiments. The MIT coil carried 15 kA at 11 T for 5 minutes with no sign of instability. A half turn length in a 10 T field was able to absorb a heat load in 4 msec of more than 200 mJ/cm3 of cable volume while carrying a current of 12 kA. The MIT coil successfully met the performance requirements of the Department of Energy's 12 Tesla Coil Program.

ii P P.-

CONTENTS

page

1.0 Introduction ...... 1

2.0 R esults ...... 2

2.1 Critical Current At 4.2 K ...... 2

2.2 Steady State Operation ...... 4

2.3 Critical Currents Above 4.2 K ...... 4

2.4 Lap Joint Resistances ...... 4

2.5 Transient Stability ...... 6

2.6 Q uench ...... 6

2.7 Q uench Pressure ...... 6

3.0 C onclusions ...... 8

3.1 Steady State Stability Requirement ...... 8

3.2 Transient Stability Requirement ...... 8

4.0 D iscussion ...... 9

5.0 R eferences ...... 10

6.0 A ppendix 1 - Test Plan ...... 11

7.0 Appendix 2 - Interface Specification...... 95

8.0 Appendix 3 - Operation of the HFTF...... 111

iii 1.0 INTRODUCTION

The MIT 12 Tesla Coil was built to test the performance of an Internally Cooled, Cabled Superconductor (ICCS) of practical size. The type of cable-in-conduit conductor investigated is identical to the Westinghouse Large Coil Program conductor and is intended for future use in high magnetic field fusion experiments. It consists of a 486 strand cable of Oxford-Airco bronze-matrix, multifilamentary Nb 3 Sn enclosed in a conduit of JBK-75 superalloy. The test coil contains approximately 120 m of this conductor in the form of three double pancakes. The double pancakes are connected by means of resistive copper lap joints. Fabrication of the coil took place at Everson Electric Company of Allentown, Pennsylvania under MIT supervision. The coil was tested in the High Field Test Facility (HFTF) of Lawrence Livermore National Laboratory in October 1984 and January 1985. It successfully met the requirements defined by the Department of Energy at the outset of the 12 Tesla Program in 1979. Further details of the conductor and coil are listed in Appendix 1 and the References.

1 2.0 RESULTS

Results of the MIT 12 Tesla Coil experiment are summarized below. 2.1 Critical Current At 4.2K

The critical current at 4.2 K of the crossover turn of the central double pancake, called subcoil B, is shown in Figure 1. The length of the crossover turn conductor in high field, used to determine the critical current, was approximately 130 cm. The maximum critical currents measured are shown by the solid circles on the two coil load lines, defined to be curves relating the maximum field at the conductor to its current. These maximum currents were not due to limitations of the conductor. Rather, they were determined by the limitations of the available power supply and by the conservatism of the experimenters. The background magnetic field, due to the six High Field Test Facility magnets, is shown at the intersection of each coil load line with the horizontal axis. The maximum available background field at the crossover turn of subcoil B was 8.93T. The other background field, 8.38 T, was selected arbitrarily.

The critical currents, given as lines that intersect the load lines, are shown at four sensitivities: 0.01, 0.015, 0.03, and 0.11 gv/cm. The most accurate data is at the highest average electric field, that is, at 0.11 uv/cm. Note that data along the load line starting at 8.93 T was measured at 10 times the sensitivity as that along the load line starting at 8.38 T. Thus, there was approximately.10 times the uncertainty in the determination of critical currents along the 8.38 T load line. For this reason, the lines of critical current at sensitivities below 0.11 v/cm have been drawn parallel to that at 0.11 v/cm, starting at points on the 8.93 T load line.

Data were taken at zero helium mass flow and two internal helium pressures: 1 and 3 atmospheres absolute. As shown in the figure, the critical current at 4.2 K is 16,900 amperes at 11.7 T and an average electric field of 0.11 uv/cm.

2 Cal II

......

011

I -I* _I ::1 -

F- -( - - -ia-t r-

-t.

. . . .

Fig. 1. Critical current of the MIT 12 Tesla Coil at 4.2 K. Coil load lines start at HFTF background fields of 8.38 and 8.93 T. Data were taken with no helium flow at internal absolute pressures of 1 and 3 atm. Maximum critical currents are given at a measurement sensitivity of 0.11pv/cm.

3 2.2 Steady State Operation

The coil was held at 15 kA, 10.85 T for 300 s with internal helium at 1 atmosphere absolute pressure, 4.2K and no flow. During this test, the average electric field along the crossover turn of Subcoil B was approximately 0.013 v/cm. The power per unit length dissipated in the crossover turn was therefore approximately 200 pw/cm. Note that since the internal helium was at 1 atmosphere, steady state stability was enhanced somewhat by the available heat of vaporization. This would not have been the case had the internal helium been supercritical (above 2.26 atm at 4.2 K).

2.3 Critical Currents Above 4.2 K

Four measurements of subcoil B critical current were made at elevated tmperatures: two at approximately 5.2 K, and two at approximately 7.5 K. These critical currents are shown in figure 2 as points 1-4, at a sensitivity of 0.015 v/cm. During this portion of the experiment, internal helium pressure was held at 3 atm absolute, and there was no helium mass flow through the subcoil. The uncertainty in helium temperatures was approximately 0.2 K. Note that point 5, taken from figure 1, is a 4.2 K point shown for reference.

2.4 Lap Joint Resistances

Termination lap joint resistances at the coil leads and joints were measured with the HFTF background magnetic field both off and on. All lap joint resistances appeared to be independent of current in both cases. The average magnetic field at the leads and joints when the HFTF was energized was approximately 1.5 T. Lap joint resistances were higher, on the average, when the background magnetic was energized. Note that the largest measured resistance, 74 nfl, was nearly one order of magnitude larger than the value estimated by design calculations. Lap joint resistances are summarized in Table 1 below.

Table 1. MIT 12 Tesla Coil Lap Joint Resistances

Location HFTF Off HFTF On A Lead 25nfl 33nfl C Lead 9nf 9n11 A-B Joint 53ni 74nfl B-C Joint 13nfl 16nn

4 ------11,

1::: 7 7 T :: ~ v. ±

- -- ~ . " ------4

--5 ~ -

i1 i--i-----V~ti -- i .... ~j... .

-T -- -- I 3. -(

Fig. 2. Elevated temperature critical currents of the MIT 12 Tesla Coil. Measurements were taken at 5.2 and 7.5 K at an internal absolute pressure of 3 atm with no helium flow. Data are shown at a measurement sensitivity of 0.015 Av/cm.

5 2.5 Transient Stability

The ICCS transient heat load stability was verified to be in excess of 200 mJ/cm3 of cable volume when operating at 10 T, 4.2K, 3 atm helium pressure, zero flow and a steady state current of 12,000 A. Energy was delivered by an. inductive heating technique that used a pulse coil as a transformer primary and the ICCS as a transformer secondary. The current pulse into the pulse coil had a waveshape that was a quarter sine wave with a duration of approximately 4 ms.

The stability test involved one 4100 ampere current pulse from a 87,000 1tF capacitor bank, charged to 250 V, into the pulse coil of subcoil A. The intended wave shape was a half sine wave. Unfortunately, the pulse coil leads were inadequately supported and broke at the peak of the current pulse, causing the current to collapse in less than one millisecond. This truncated the pulse into a quarter sine wave.

Previous work had shown that a half sine wave current pulse of 4100 A and 8 ms duration would deliver more than 200 mJ/cm3 to the cable under the operating conditions described above. Because the intended half sine wave was truncated into a quarter sine wave, the rapid collapse of current resulted in a higher energy input than expected. This led to the conclusion that the energy delivered was greater than 200 mJ/cm 3 .

2.6 Quench

The crossover turn of subcoil B, operating at 7.5 K temperature, 3 atm helium pressure and zero helium flow was purposely driven normal at 9.5 T by ramping the current to 15,200 A. The resulting quench was followed by a safe discharge of the coil current using the power supply diode rather than the dump resistor. None of the HFTF background coils were adversely affected by the discharge of the MIT coil and did not have to be dumped.

2.7 Quench Pressure

Helium pressure at one end of subcoil B is shown as a function of time in Figure .3 for the quench of the MIT 12 T Coil. Coil current as a function of time is also plotted. The initial conditions were stated in Paragraph 2.6. The boundary conditions were open stainless steel pipes of 1.1 cm inner diameter. Pressure was measured at approximately 71 cm from the connection of the pipes to the conductor. The end pressures rose from 3 atm to approximately 4 atm as a result of the quench.

6 F- I

F- 1:. T I

A 1 Mi- MMMMMMMMEMEMMEMEMMMMEEMEMMk MMMMEMMM...... OM4MMMMEEEM

Fig.3.e_ MMMEMMMMMEMMMMEMMEMMMEMMMM4%_ ssre atstatio 26/26 adM Ie nm ziziz ztzzMMr

MMMMMMMMMEMEMMEM...... EM quench of te 12T . .EAcoil......

MEMEEMMEEMEMMEMMEMMMMMEMtinM

Fi.3.Pessure tain 626aM IT 12 Tes Ci ce ntasfucton o me for quench of the 12 T coil.MMMMM

7 I

3.0 CONCLUSIONS

The MIT Coil satisfied the major requirements of the 12 Tesla Program.

3.1 Steady State Stability Requirement

This program requirement stated that candidate test coils must carry between 10 and 15 kA in a maximum magnetic field of 11-12 Tesla. The MIT coil satisfied the steady state stability requirement. It was tested at 15,000 A and 11 T for 5 minutes and showed no signs of instability.

3.2 Transient Stability Requirement

This program requirement stated that candidate test coils must not quench when subjected to a transient heat load of 100 mJ/cm3 over a half turn length in the high field region, with the time of energy deposition not exceeding 50 msec. As mentioned above, conductor stability was verified to be above 200 mJ/cm3 at 10 T for a 4 msec heat pulse.

8 4.0 DISCUSSION

The MIT 12 Tesla Coil experiment demonstrated that an ICCS of practical size works well in a realistic coil environment. The conductor met the steady state and transient stability requirements stated at the program inception in 1979. The major objectives of the program were thus satisfied.

However, the opportunity to thoroughly study the stability of a full size ICCS was lost. The pulse coil inductive heater leads were inadequately supported and failed when the capacitor bank was discharged. Although more than enough energy was delivered to the ICCS to satisfy the transient stability requirement, the lead failure brought an immediate halt to stability testing.

Had the leads not failed, stability testing would have been limited by two other considerations. First, the energy delivered by the pulse coil power supply did not cover a large enough range of fields and currents at 4.2 K. Although data for ratios of transport- to-critical current near 1.0 could have been obtained, this would have been an incomplete set. Second, and perhaps more economically important, the heat load on the HFTF dewar was larger than the available refrigeration capacity. This meant that liquid level could be maintained only by using liquid from external storage dewars. The helium recovery system was not sized to capture all the boiloff from the experiment, resulting in substantial helium losses to the atmosphere. Stability testing was begun when the storage dewars were perhaps 30% full. At best, the available liquid helium would have permitted several hours of stability testing.- Had the test been able to continue, it would have been necessary to purchase at least another 4000 liters of liquid helium.

Note that none of the reasons cited above preclude future stability testing of the MIT 12 Tesla Coil. For example, the pulse coil leads can be brazed back together, wet-wrapped with fiberglass tape and epoxy, and then perhaps placed inside stainless steel tubes for support. To repeat, the pulse coil leads are repairable.

The amount of delivered energy could be increased in the future by using a double current pulse, a technique that requires two separate capacitor banks. These banks and associated electronics now exist at MIT and have been used many times in similar stability experiments.

Efforts have been made to upgrade the refrigeration capacity of the HFTF. Future testing should no longer be determined by the storage capacity of external dewars. If interest in stability testing of the MIT 12 Tesla Coil revives, then the HFTF may be operational for extended periods of time.

In summary, the MIT 12 Tesla Coil experiment satisfied the program requirements outlined in 1979. However, the MIT 12 Tesla Coil still has the potential to provide useful information about ICCS conductor stability to the fusion community. Should the need arise, the experiment can be revived with relatively modest effort.

9 5.0 REFERENCES

1. Hoenig, M.O., et al.,"Progress in the ICCS-HFTF 12 Tesla Coil Program," IEEE Trans. on Magnetics, Vol. Mag-17, No.1, p.638, Jan. 1981.

2. Hoenig, M.O., et al.,"Cryogenic Aspects of the Internally Cooled, Cabled Supercon- ductor (ICCS) for the 12 Tesla Program," Adv. in Cryo. Eng., Vol. 27, p.217, 1982.

3. Steeves, M.M. and Hoenig, M.O., "Lap Joint Resistance of Nb 3 Sn Cable Terminations for the ICCS-HFTF 12 Tesla Coil Program," IEEE Trans. on Magnetics, Vol. Mag-19, No.3, p.378, May 1983.

4. Hoenig, M.O. and Steeves, M.M., "MIT 12 Tesla Test Coil Experiment," IEEE Trans. on Magnetics, Vol. Mag-21, No.2, p.1052, March 1985.

10 6.0 APPENDIX 1 - TEST PLAN

. The test plan for the MIT 12 Tesla Coil is presented here as an archival document. It covers the intended steps of the experiment from the room temperature check of the helium supply system to final stability tests with an AC ripple superimposed on the transport current. It also gives information about the conductor, coil, helium supply system and instrumentation used in the experiment. TABLE OF CONTENTS

PAGE

1.0 INTRODUCTION ...... 1

2.0 TEST PLAN ...... 2

2. 1 Room Temperature Check of Helium Supply System ...... 2 2. 2 Room Temperature Check of Instrumentation ...... 2 2. 3 Cool-down to 4.2 K ...... 3 2. 4 Zero Background Field - Flow, Pressure and Ter- perature Calibrations ...... 0...... 3 2. 5 Zero Background Field - 12 T Coil Terminations Tests ... 3 2. 6 Zero Background Field - Heating Tests ...... 5 2. 7 Zero Background Field - Protection Circuit Test ..... 10 2. 8 Calibration of Pressure Transducers and CGR's in Background Magnetic Field ...... 10 2. 9 Critical Current Tests ...... 10 2.10 Transient Stability Tests ...... 14 2.10.1 Zero Flow, 2.5 atm, 4.2 K and 10 T ...... 14 2.10.2 Zero Flow, 4.2 K, 10 T and 1-10 atm ...... 16 2.10.3 Mass Flow, 4.2 K, 2.5 atm and 10 T ...... 17 2.10.4 Zero Flow, 2.5 atm, 10 T and T > 4.2 K ...... 19 2.11 Critical Temperature Tests ...... 23 2.12 Steady-State Current Sharing Tests ...... 23 2.13 Quench Propagation and Helium Expulsion Tests ...... 23 2.14 AC Ripple Tests ...... 28

APPENDIX 1 - Conductor and Coil Parameters ...... 30

APPENDIX 2 - Helium Supply System ...... 35

APPENDIX 3 - Instrumentation ...... 41

APPENDIX 4 - Termination Lap Joints ...... 56

APPENDIX 5 - Transient Stability Model ...... 60

12 - " - ,, 7

-I-

TEST PLAN - MIT ICCS 12 TESLA COIL

1.0 INTRODUCTION

The MIT ICCS 12 Tesla Coil has been built to test the performance of an Internally Cooled, Cabled Superconductor (ICCS) of practical size. This cable-in-conduit conductor is similar to the Westinghouse LCP conductor and is intended for use in fusion experiments such as the proposed Alcator DCT. It consists of a 486 strand cable of Airco bronze matrix multifilamentary Nb 3 Sn enclosed in a conduit of JBK-75 superalloy. The test coil contains approximately 120 m of conductor in the form of the three double pancakes. The double pancakes, called subcoils, are connected by means of resistive copper lap joints. Details of the conductor and coil are listed in Appendix 1.

13 -2-

2.0 TEST PLAN

2.1 Room Temperature Check of Helium Supply System

2.1.1 Description

Appendix 2 gives a detailed description of the helium supply system for the 12 Tesla Coil. Briefly, it consiists of a room temperature helium source capable of supplying up to 9 g/s at a maximum inlet pressure of 100 psig. The inlet helium travels through an external liquid nitrogen heat exchanger before entering the helium flow assembly built by MIT. The assembly contains counterflow and liquid helium heat exchangers and has hand operated cryogenic valves in the flow circuit of subcoil B.

2.1.2 Helium Purge and Leak Check

The helium flow assembly will be leak tight per helium mass spectrometer leak tests upon departure from MIT. At Livermore, piping subassemblies containing transducers will be welded or brazed V~ C U.ALw to the 12 Test Coil helium piping. On completion, the entire cool- I- ing system will be pumped down to vacuum and then purged with helium. The system will then be pressurized to 100 psig and pressure decay monitored for at least a 2 day period. This test will be implemented -s / / ~ by helium mass spectrometer leak testing, if necessary. Leaks will be patched by welding or brazing. While undergoing vacuum pumpout I- and purge, the system shall be heated, if possible, by heat lamps.

2.1.3 Valve Check

All valves will be checked out at this time. After check out, all hand operated valves except cryogenic valve VC1 shall be left closed. VC1 shall be left open. See Appendix 2.

2.2 Room Temperature Check of Instrumentation

2.2.1 Description

Instrumentation sensors and heaters are detailed in Appendix 2. Briefly, the 12 Tesla Coil has voltage taps, pulse heating coils, pulse pick up coils, a heater wire in subcoil B, carbon glass resistor thermometers, pressure transducers and a Hall probe.

14 -3-

2.2.2 Continuity Tests

The voltage taps, pulse heating coils, pulse pick coils, and heater wire shall be tested for continuity.

2.2.3 Excitation Tests

The carbon glass resistors, pressure transducers and Hall probe shall be excited according to their individual specifications. Output signals will be measured while individual sensors are under excitation.

2.3 Cooldown to 4.2 K

When the coil assembly is positioned in the HFTF 2 meter diameter cryostat, and cooldown is about to begin, all hand operated valves associated with the helium supply system shall remain closed except cryogenic valve VC1. When cooldown begins, inlet valve VR1 will be opened and the inlet pressure set at 22 psig (2.5 atm absolute). This allows helium to condense in the 12 Tesla Coil as the entire assembly cools to

2.4 Zero Background Field - Flow, Pressure, and Temperature Calibrations d-

Assuming no cold leaks, zero background magnetic field testing will begin. The first test will be to establish the mass flow capability of the blnq.-doni nystem through mubcol B !hunT valv.R Vil4--rnd- TT4 are - opened from an initial condition of 2.5 atmospheres at zero mass flow. Maximum mass flow available will be established by increasing the inlet LTh(, pressure to its peak value and reading mass flow on SAroob tetperature flow meter, apstree.-.FV4. Flow will also be mesured on venturi meter VM2. VS e Once'mass flow capabilities are established at 2.5 atm, recovery valve will be closed, returning mass flow to zero. Then all pressure transducers and CGR's will be excited and their outputs recorded. Pressure will then be raised in steps to 100 psig, allowing inlet flow to drop to zero after each step, to calibrate the pressure transducers in zero magnetic field.

2.5 Zero Background Field - 12 Tesla Coil Terminations Tests

With the HFTF solenoids at zero current, the 12 Tesla Coil will be energized in steps of 5,10,14,16,17,18,19 and 20 kA. The initial condition will be zero mass flow at 2.5 atm pressure with valve# VB- and VB5 VF closed. Voltage drops across the two lap joints between subcoils and across the coil leads will be monitored to check the performance of the coil terminations and joints.

15 Faa - c 1 Q.~c~

PT:a - ;-~

7E VMv%

T ~ k Co-L

V(0 ol) - 1X.. 1 b&-4 ~Ic-~'

P, -VT4),YE1.

VVM1

Sj FL Y- -4- V(

r- "--. V ~~ V-~AL~CP7

£4p

16 .0- f0

17e

16~~~~- Srae tte nf 0

1.4L

*Aw

q-4 -4

4 I- -~~~c I .J -'. .5 .1 0j

-I -CZ

c~jj IN

4%-

-1

18= ......

-4 -

S La -To n - Lc ,J Volky~ Tp5

VT -1g E ke h FTF T.R Ck A

I VT 13-.LpZToivt A-B

NFT-rR Ct- ( V(qfC

VT 13 -:5 k 1. i - B 1%-4 9-t Natt

VT 13- 4 etb

NFTPER C3 -Lag oldC-- L

Figure 1. Termination lap joint voltage drop tests at zero background magnetic field.

19 -5-

The expected voltage drop across each lap joint at 20,000 A is less than 100 pV, resulting in a total heat load of not more than 2 watts per joint. Details of the lap joints are given in Appendix 4. The expected volt-ampere characteristic of the lap joints in a 1 Tesla magnetic field is shown in Fig. 2. The zero field curve should lie below this line.

The termination lap joint tests provide a convenient opportunity to check the 12 Tesla Coil load line, since a Hall probe is installed in the coil case near the high field point. The 12 Tesla Coil is expected to produce 1.77 G/A. a .

B 4--7x 10 i [T - (1)

At 20,000 A, the maximum output of the 12 Tesla Coil power supply, the expected self field will be 3.54 Tesla , as shown in Fig. 3.

2.6 Zero Background Field - Heating Tests

Tests of the heater wire in subcoil B and the three crossover turn pulse coils will take place after completion of the termination lap joint tests. The initial condition will be zero mass flow at 2.5 atm and 4.2 K as in previous tests.

Depending on the voltage and power output of the available heater wire power supply, the heater will be supplied with currents ranging from 0.2 to 2 amperes. These currents will produce steady-state temperatures ranging from 4.5 to more than 20 degrees kelvin as shown in Fig. 4. The heater will be energized in steps of 0.2 A according to the schedule of Table 1. Thermal equilibrium should be reached in approximately 3 to 5 minutes at each current level.

Steady-state temperatures will be measured by carbon glass resistors stationed at the 900 and 2700 positions on the crossover turn of subcoil B. It is estimated that these thermometers will read within 5% of the helium temperature, and therefore within 5% of the superconductor temperature. After the highest equilibrium temperature is recorded, subcoil B will be flushed at approximately 1 gram/second and the time to return to 4.2 K will be recorded.

20 N--1S- -6-

-4 ... # --.

- - - -

-4-.+ +- + ------t - -in * -4

!Hii-

Figure 2. Estimated A-B and B-C lap joint voltage drops at 4.2 K. The joints are assumed to be in an average magnetic field of I Tenla.

=I I 1, rv ( )!ey:-j V

21 -7- z,~-11* 4

47------A- 4---

- I I A __

...... U

i m~

7 -7

J- 4 J _1 ------

-4 __ V Ike~

_NV _ ± ~ ------r*~-~ -- ~------

4--4-

laa

.. .-......

.4 4~ ---- 4.-4-.. - - -- -

-- i-4 I. I ~ ... P-F-T 7_i: 7 :7

Lft 0 0 LA #A 40 V5 N S 22 -8- IA

1-*~~

T!T

.- Ril Ili. -H I. 1 +44-e-4

, J # 4 U

4. 4.

14 0.

C v 0 0

414 40 0 *. o 0

c 00 SM 0S4I g &o q .0 S41 v Oa I 4J

I 23 -9-

TABLE 1. Subcoil B Heater Wire Test at Zero Current, Zero Magnetic Field and Zero Flow

Vheater iheater ESTIMATED STEADY-STATE T OV OA 4.2 K 24 .2 4.5 48 .4 5.3 72 .6 6.7 =Z 2" - 96 .8 8.6 AA 120 1.0 11.2 144 1.2 14.3 168 1.4 17.9

Rh = 134 a (4.2 K)

Upon completion of the heater wire tests, the three pulse coils will be fired. starting with pulse coil B. The initial condition will be zero flow at 2.5 atm and 4.2 K. Prior to firing, the series resistance of the power supply leads and pulse coil will be measured. Pulse coil B will then be fired in steps of 100 volts according to the schedule of Table 2. Temperatures at the 900 and 2700 positions will be monitored during the sequence of discharges. Voltages from the pulse pick-up coil will also be recorded.

TABLE 2. Subcoil B Pulse Coil Test at Zero Current, Zero Magnetic Field and Zero Flow

B- 0 Vcap PEAK V I3-0 Ishunt

0V OA OV 100 2,100 4.5 200 4,200 9 300 6,300 13.5 400 8,400 18 ._50 --tO50 22.5

R-k -( Cc \

24 -10-

2.7 Zero Background Field - Protection Circuit Test

The 12 Tesla Coil will be protected by a 2l dump resistor. The circuit will be interrupted by a single-pole circuit breaker.

20 V _ 2Q 12 T coil

With the coil at zero flow and 4.2 K and the background field at zero, the coil will be charged to 5,000 amperes and interrupted by the circuit breaker to test the protection system. The protection circuit test will be repeated at higher currents if deemed necessary.

2.8 Calibration of Pressure Transducers and CGR's in Background Magnetic Field

Critical current tests of the 12 Tesla Coil will start by first energizing the HFTF background field coils. These coils will be energized in steps of approximately 1 Tesla with pauses between steps to record the outputs of the pressure transducers, CGR's and Hall probe. The initial condition will be zero mass flow at 2.5 atm. Figure 5 shows the response of CEC-1000 pressure transducers in magnetic field. Figure 6 shows the apparent thermal error of a typical CGR in magnetic fields ranging from 0 to 12 T.

2.9 Critical Current Tests

Once the peak field at the 12 Tesla Coil midplane reaches 8 T, critical current testing can begin. The initial condition will be supercritical helium at 4.2 K, 2.5 atm and zero mass flow. Current will be slowly ramped until the voltage drop across the crossover turn of subcoil B (point of highest magnetic field) reaches 2pv (0.015pV/cm). Cukrent will then be slowly ramped until the voltage drop reaches 20pV (0.151iV/cm), and then cautiously ramped to 200iiV (1.5pjV/cm), the Airco electric field cri- terion of the cable wire. Similar critical current tests will be conducted with background fields of 7 T and-, if permissible, 9 T. Figure 7 gives estimated critical current characteristics of the 12 Tesla Coil.

E I VI_/c'V..,

0 0-. A5V_/

25 Z 7 q

4 --. Y- Ax,- IV' A -7 -5-V \IT i fV7-1 A. 0171Vi~

2.-> IC .

1- A- to jo4 /c-

26 - if-

3.0 - I 2.51 GOULD B / 2.01 B0.8 T 2-8T c 1.5

E CEC 1.0 B-

OT NA T 0. 5 T -2-8 T ~7 0.51 ~-'3- T= 4.2K 0 I -

-0.5

0 50 100 150 200 Vi~ ~. 5 P (PSIG) l~ RESPONSE OF TWO PRESSURE TRANSDUCERS IN MAGNETIC FIFI DS OF 0-8 T AT 4.2 K

27 -12-

0.3 .I I I I I I I I

- CARBON GLASS RESISTOR LAKESHORE CRYOTRONICS -- 4.2KK 0.2 ul 7Ai k

0 0.1

LUJ

I 0 Ii1-i 0 2 4 6 8 .9i 10 12 MAGNETIC FIELD (T)

Fi .6G APPARENT THERMAL ERROR DUE TO APPLIED MAGNETIC FIELD

28 -13- SA.A4EtA&-P

- -.-- -

-- -.

4 -4

d 0 0 g 0 z

U. 4.1 4- 0' til HE tt -_ So g = B + .7 .4. a.WE I~~Il I U z -~ 0 -i

I- U

-~~ - I U 0 S

z I i ! i --- g .

-A a z 0

a 0 0

S 0

0 U

'.4

z4, 0

C C I !!_rf kp , I 1 -,, N U' z0

Figure 7. Estimated critical current characteristics of 12 Tesla Coil at three electric field criteria. The maximum available background field is 9 T. The maximum available power supply currernt is 20 kA.

R . 2T c. vT-r-

E,- .- 015I~l /4g\c:-A V 29 -14-

2.10 Transient Stability Tests

Upon completion of the critical current tests, transient stability tests will begin. The HFTF solenoids will be charged so that the peak field at the 12 Tesla Coil reaches some preassigned value. Then the 12 Tesla Coil current will be raised to an appropriate fraction of critical current to begin testing. Testing will focus on subcoil B. After the capacitor bank has been discharged into pulse coil B, -the voltage versus time plot is expected to behave in one of the following ways:

(1) No voltage is observed across the crossover turn. (Imposed energy is less than the critical energy).

(2) Voltage appears, recedes to zero, and remains at zero for more than 20 seconds. (Imposed energy is less than the critical energy.)

(3) Voltage appears, recedes to zero, then reappears after 2 or 3 seconds. (Imposed energy is less than the critical energy. Heat diffusion from the pulse coil has caused quench.)

(4) Voltage appears, recedes to zero or near zero, then starts increasing. (Imposed energy equals the critical energy. This is the case of interest).

(5) Voltage appears and remains above zero for 0.1 to 1.0 seconds. (Imposed energy exceeds the critical energy.)

The idea behind transient stability testing is to measure the energy margin of the conductor. Energy margin is defined as the maximum imposed energy from which the conductor will recover. This type of testing is somewhat complicated by the poor cooling of the pulse coil, since heat diffusion from the pulse coil can possibly lead to a superconductor quench. The range of expected outcomes tabulated above reflects the possibility of an unwanted quench due to heat diffusion (outcome No. 4). Figure 8 plots estimated energy delivered to the cable as a function of initial capacitor voltage (See Appendix 5).

2.10.1 Zero Flow, 2.5 Atmospheres, 4.2 K and 10 Tesla

The initial condition will be zero mass flow at 2.5 atmo- sheres and 4.2 K. Table 3 lists a possible scenario for stability tests at 10 Tesla . These tests are relevant to the proposed Alcator DCT.

30 ...... - -- ...... I ...... I ...... I...... 1- ...... "...... H±H f ...... - ...... I ...... - -- _...... --...... II II ...... I...... - --I.I ...... I...... I ...... 1 1...... - 1.I ...... 1. 1 I . I - II ...... I...... I...... - ...... ;. - ...... 1 ..- .1- .....1.1- 1.1...... I...... - - ...... 11 ...... i ...... I...... 1...... ; ...... I ...... I...... I ...... - . I...... I ...... - I. I ...... I ...... I- ..... ' ...... 1- 1...... 11 . 1 11 ...... " ...... I ...... - I...... - I..... I1 1. . 1...... - ... I -.. -.. .I...... I ... I...... - ...... - ...... I ...... I...... I. I I .....1- .1 ...... I... I ...... --.. - ...... - ...... --...... _ ...... log ...... I ...... I - I...... ±hH ...... O dD ...... - ...... -Sop ...... - 11 ...-- : ...... I I ...... 1- ...... - -- ......

qO ......

......

0 ...... - , ...... 11...... 1 .I ' l l ...... -- ....11 ...I.- ..... I.-.. - I ...... I . ....I ...... I ...... I ...... - . b. I......

0 ......

i 1111 i H 111 i 4 4j i + W11111iiiIiIIIII W611iIIIIIII IIII Ill ------IWI

Figure 8. Estinated energy per unit volume of cable vire delivered to cable as a function of initial capacitor voltage at 10 Teala and 4.2 K...... ill I IIll Li U-SuIllub"11 ...... ------

31 -16-

TABLE 3. Proposed Transient Stability Tests at Zero Flow, 2.5 Atmo- spheres, 4.2 K and 10 Tela With I/Ic as the Independent Variable (Ic a 24,000 A at 1.5 VV/cm).

CASE BHFF ICOIL BTOT I UV EXPECTED ESTIMATED NO. - (1.5 -) Vcap Q Ic - cm

I 9 T 5,600 A 10 T 0.23 550 V 1750 mJ/cc

2 8.5 8,500 10 0.35 Soo 14SO

3 8 11,300 10 0.47 H4D ci>o

4 7.5 14,100 10 0.59

5 7 16,900 10 0.71 Li9o

6 6.5 19,800 10 0.82 i(GO

--- - 2~-~~- * -

i A , _ - _ _

i44 T_ T; 1 Tj I 43C

-: 7 T

Fw =4- 4 - 1 7! * j

i -P

J~~4 t -4 41 4

defined 'in Tkble 3. Fig Lre 9. Energy margin versus IC for the tests

32 - -- - r EL ------r ------

-L't

------LID cl-i - 7 ------r ------11 Icl I CL I

r ---T-: ------r ------4

Cl ----- C\i I ----- r T ------

M OD 0 co LD C,J C6 . m (@J''n JO 33/rw) N19 eW O AINI

33 ~., ~- I f

.. 7 1 t_7II 7 1:c JSt _ _ ------.....4 ------...... 40 7r...... _ _ i77- uJFW ~ F j ~ I ~

------_ _ __......

...... _ FI-7 --- --*~- ___ ------_ _

477=_ -~

-- 7

:7,

______+ ___

i i

_ I _ _- M/_

- 4 _ _ 777=

_i A =#

------

34 / -____ y__ 0f -2 J a ,,2 COl' C , ((

Aill e 4 B = o T70 -L

' _-0/0 c k.~-~* iL C K) 1 1h. (J /cc.) L / .) VIa C iC~ AST a--,-

1. 9 o .3 y -4 5- ~ t62l o S2- 002 1 06 o.35 logo

-o 0-35 I090 -/30

0-60 r- 5

/4 5

V ? /60 Y5-8 ILI /00 c -6o 5'.? * ~s.z v-/y,39' as- 7'0

1-f, 100 c .6 0 1o .0 6.f.S Y ~3.5 V-3

,'utoo o-62 R4 P7 00 - -C/'

00 0 .Fr.C -. 30c 7.

19 ~4 .5 .IV c2a -B

- -3

9.9 -0 sc I3.

23'! 2y

2 z /66

.90 (L-2L /o3 16.3

21 11 vo to IV -C, .CIO ) ~L /,&po 35 -'7-

2.10.2 Zero Flow, 4.2 K, 10 Tesla and 1-10 Atmospheres

This set of tests varies the helium pressure from subcrit- ical at approximately 1 atmosphere to supercritical at 10 atmospheres while holding I/IC and B constant. Table 4 lists proposed tests with pressure as the independent variable.

TABLE 4. Proposed Transient Stability Tests at Zero Flow, 4.2 K, 10 Tesla and I/IC M 0.6 With Pressure as the Independent Variable.

CASE P BHFTF ICOIL BTOTAL EXPECTED ESTIMATED NO. Vcap 9

S c. 7 5 atm 7.5 T 14, 100 A 10 T '3ioV'

8 10 7.5 14,100 10

9 2 7.5 14,100 10

10 1 7.5 14,100 10

C- -~-iv~1 14,ov 14 u /cc

igure 10. Energy mgrin versus pressure for tests defined in' Table 4.

36 -18-

2.10.3 Mass Flow, 4.2 K, 2.5 Atmospheres and 10 T

This set of tests is similar to those in paragraph 2.10.1 except the effects of mass flow will be studied. The expectation is that energy margin will be independent of mass flow. Table 5 outlines the proposed energy margin tests. Figure 11 illustrates the expected dependency on mass flow.

TABLE 5. Proposed Transient Stability Tests at 2.5 Atmospheres, 4.2 K, 10 T and I/Ic - 0.6 With Mass Flow as the Independent Variable.

CASE m BHFTF ICOIL BTOTAL EXPECTED ESTIMATED NO. Vcap 9

11 1 g/s 7.5 T 14,100 A 10 T

12 2 7.5 14,100 10

13 3 7.5 14,100 10

14 4 7.5 14,100 10

15 5 7.5 14,100 10

Figure 11. Energy margin versus 1 for tests defined in Table 5. -19-

2.10.4 Zero Flow, 2.5 Atmospheres, 10 T and T > 4.2 K

In this set of tests the helium temperature will be raised in steps above 4.2 K until the energy margin of the conductor reaches zero. Helium temperature will be controlled via the heater wire in subcoil B. Table 6 outlines the proposed set of tests. Figure 12 illustrates the expected behavior. Figure 13 plots expected noncopper critical current density as a function of temperature. Fig- ure 14 is essentially the same plot with critical current plotted as a function of temperature.

TABLE 6. Proposed Transient Stability Tests at Zero Flow, 2.5 Atmospheres, 10 Tesla -and Ic - 2_4,100 A With Temper- ature as the Independent Variable.

CASE T BHFTF BTOTAL I MV EXPECTED ESTIMATED NO. - (1.5 -) VcapQ Ic cm

) SaD V 11% "t/, 11 -4.5 K 7.5 T 10 T .62

12 ~5 7.5 10 .68 ~L.t %.

13 5.5 7.5 10 .76 a&ov :~2c) %/r~

14 ~6 7.5 10 .86 [A7l':x IJ7 1 ,c I V _15 _ 6.5 7.5 10 .99

,(A) At- -r I

C4 4- ~ ~ '-C

N I C.-t.~ i. ~ I f- - ---i ' -1------*

9.2ka

38 -20-

Figure 12. Energy margin versus temperature for the tests defined in Table 6.

39 -21-

0 0

0 N --- 1 4 0 - t 4 4- 0 z -4~z:.~;z:z -~ Z-E= -:7-A--. M --- 4--.-- I------4 --- 4 zt7T t - -, 4~fj- ~ ±zcz~:fr-- o Uo 4-AzztiFI~ 4-.--- 77:

- 0 0U, u 0

K. .4

U e __4' Kie .. 1-- ~J~I 2 L --- I * ----- 4--

----- * .4 -1V

-- I :t__ 7L-- 4- W V4

0 ~1?--- * 4

IC * 00 6 V * U 1 0 4J V 0 a S -- t- 4j . i i 4.

it: -4--

.94

40 -22-

-~ ~rrri- - - -~i I ... - - --

t=- 7 --

t..". . , I

4 S0

-4-

-4

-z ; i -4-z- - 0

- 0

D 4 .4

'4J

-- 40 =. -! il i i I

41 J1-

2.11 Critical Temperature Tests

These tests are designed to determine where current sharing begins at temperatures above 4.2 K. The initial condition will be zero mass flow at 2.5 atmospheres. -The heater will be energized and the temperature of the crossover turn of subcoil B will be allowed to stabilize. This temperature will be measured by CGR's labeled BT-900 and BT-270 0 (See Appen- dix 3). The HTFT will be raised to a specified background field and the 12 Tesla Coil current slowly ramped until the critical voltage is reached. Figure 15 shows load lines plotted with estimated critical current characteristics for temperatures from 4.2 to 10 K.

2.12 Steady-State Current Sharing Tests

These tests are designed to study the stability of the coil in pro- longed current sharing. The idea is to raise the background field while the test coil is at zero mass flow, 2.5 atmospheres, and 4.2 K, and then slowly ramp the test coil current until voltage is developed in the crossover turn. Current will then be held at this level from 1-5 minutes.

2.13 Quench Propagation and Helium Expulsion Tests

These tests are designed to measure quench propagation, velocity, pressure at the ends of the subcoil, temperature at the ends of the subcoil, and mass flow of helium out the ends. All quench tests will be done on subcoil B.

2.13.1 Quench Propagation

The initial condition will be zero mass flow at 2.5 atmospheres and 4.2 K. Figure 16 gives an estimate of the time required for the normal zone to reach the ends of subcoil B when coil current is varied from 0 to 20 kA. In general, either raising the joule heating of the conductor or raising the ambient helium temperature will increase the velocity of the normal front. Raising pressure above 2.5 atmospheres will lower the velocity of the normal front.

42 R

i -+ 4- o 1 4.-

0Ua i 44 W m q4

4J 0

$4 .4

W044.8g

gq.=

'.4 8

4i4. -410 -A c4 S 0. wf 1 .C $4 4104 u0 Vtz 0S

r 0 .5.

43 ...... M.". , . ..- V". I I ... II ... - - .''.'',.. . I I . .. I ..... - ...... I I . I I I .. I . I I . I .. .. I .- I ...... I ...... I . I I ...... I ...... I I ...... I . I . I ...... I ...... I ...... I ......

...... I...... I.' ' ..'...... '...... I-...... - ...11 ......

......

...... - ......

...... IIIIIIHIIIiiiHNN NH

Figure 16. Estimated time f or normal zone to propagate 20 m as a function of coil current. Assumes helium at 4.2 K and 2.5 atmospheres. The half length of each subcoil is approximately 20 a.

44 -26-

This series of tests will be approached with great caution to prevent a possible quench of the HFTF. When the 12 Tesla Coil quenches and is dumped through its 2mQ dump resistor, the flux link- ages with the six background coils of the HFTF will change. The transport currents in the background coils will increase somewhat depending on the' nature of their power supply controllers. The maximum possible induced currents can be estimated by assuming the background coils to behave as superconducting magnets in a persis- tent mode. Then the increase in current in a given background coil is simply

Mi,7 Ali - 17 Li where Ali - maximum induced current in coil i

Mi,7 - mutual inductance between coil i and coil 7 (12 T Coil)

Li M self inductance of coil i

17 - prequench current in coil 7

Table 7 lists parameters of the coil system for estimates of maximum induced currents.

-t - -27-

TABLE 7. Estimates of Maximum Induced Currents in HFTF Coils Due to a Quench and Dump of the 12 Tesla Coil (Maximum 17 - 20,000 A).

CASE TURNS Li M 7 Max I, M7 NO. Max 17 - CRITICAL Li CURRENT

1 2,366 8.48 H 0.04 H 1,200 A 94 A 2,000 A

2 1,793 5.16 0.017 1,200 66 2,000

3 2,366 8.48 0.04 1,200 94 2,000

4 1,793 5.16 0.017 1,200 66 2,000

5 380 0.073 0.0044 5,000 1,200 7,500

6 380 0.073 0.0044 5,000 1,200 7,500

7 57 0.002 - 20,000

LEGEND:

Li = self inductance of coil i

M7 W mutual inductance between coil i and coil 7 (12 T Coil)

Max I, = maximum operating current of coil i

M7 M17 Max 17 - W (20,000 A) - Li Li

46 -28-

In addition to inducing transport currents, the collapsing flux of the 12 Tesla Coil will lead to AC losses in the six HFTF coils. Assuming the 12 .Tesla Coil current to decay with an L/R time constant of about 1 sec (decoupled from the HFTF), maximum d is less than 3.5 Tesla per second at Nb 3 Sn coils No. 5 and No. 6. Since these coils are cryogenically stabilized and will operate at no more than 67% of critical current, the AC losses due to the collapse of 12 Tesla Coil current should be tolerable.

2.13.2 Helium Expulsion

Helium expelled from subcoil B during a quench will pass through venturi meters VM1 and VM2 (See Appendix 2). The estimated pressure drop as a function of mass flow through these meters is shown in Fig. 17.

2.14 AC Ripple Tests

The intent of these tests is to superimpose an AC ripple current on the DC transport current of the 12 Tesla Coil. The ripple current will impose a steady-state heat load on the coil.

4-7 -29-

a Figure 17. Pressure drop versus helium mass flow for Venturi No. 611995A-5. See Appendix 1.

T-- 45 2S S2. S WT

00-- Lr@4

A= 25q, k-YI I_

w 6 rn44/ is)

; 4 7-- C.~I&~..o~~cc~

'0 .A1rtJi+ (~ t 45%.O ~

G- -- b I SL 0.0 ------j -A -

to 14 n Z2 749 Z8

C.I I s z ,z . s ir 2 I2 ( -42

' s) 4 PFC/RR-85-12 DOE/ET-51013-155 UC20b

MIT 12 TESLA COIL TEST RESULTS

Steeves, M.M.; Hoenig, M.O.

Duplicate

July 1985

Plasma Fusion Center Massachusetts Institute of Technology Cambridge, Massachusetts 02139 USA

i -30-

APPENDIX 1 - Conductor and Coil Parameters

CONDUCTOR

The 12 Tesla Coil conductor is identical to that of the Westinghouse LCP coil. Tables A.1.1 - A.1.3 list relevant conductor parameters. Fig- ure A.1.1 shows a cross section of the conductor.

COIL

The 12 Tesla Coil consists of three (3) series-connected, double- pancake subcoils of nineteen (19) turns each. Each potted subcoil has overall dimensions of 14.7 in ID x 33 in OD x 1.86 in H. The subcoils are separated axially by trapezoidal G-10 spacers that cover 50% of the load bearing surface area. These spacers provide for secondary bath cooling of the 12 Tesla Coil. Figure A.1.2 shows the relative distribution of spacers. Table A.1.4 lists coil parameters.

TABLE A.1.1 - 12 Tesla Program ICCS.

Outside Dimensions ...... 2.08 x 2.08 cm (.818 in x .818 in)

Outside Corner Radius ...... 0.46 cm (.18 in)

Final Wall Thickness ...... 0.17 cm (.068 in)

Sheath Material ...... JBK-75 Super Alloy SS

Void Fraction ...... 32% -

Cable Configuration ...... 6 x 34 (486 strands)

2 Conductor Area (Including Sheath) .... 4.157 cm

Cable Space Area (Metal Plus 2 Helium) ...... 0...... 2.943 cm

Helix Factor ...... 1.04 Cable Area (Perpendicular to 2 Sheath Axis) ...... 1.945 cm

2 Helium Area ...... 0.973 cm

2 Steel Foil Area ...... 0.025 cm

Copper Area (Perpendicular to 2 Wire Axis) ...... 1.202 cm

Noncopper Area (Perpendicular to 2 Wire Axis)...... 0.668 cm

Hydraulic Diameter ...... 0.40 mm

49 -31-

TABLE A.1.2 - Parameters of Airco Multifilamentary Nb 3 Sn Single Strand Wire

Diameter ...... 0.7 mm

Copper-to-Noncopper Ratio ...... 1.8/1

Matrix ...... Bronze

Niobium Filaments Per Strand ...... 2,869

Filament Diameter ...... 3.5jim

Weight Percent Tin ...... 13%

Resistivity Ratio ...... 50*

1 0 Resistivity (RRR- 50, B) ...... (3.2 + 0.48 B)10 Q'm

Surface Coating - 12 Tesla Coil Wire ...... Oil ** + Soap

Average of 8 samples, in batches of 1 and 7, fired separately for 30 hours at 750 0 C. Three wires from batch 7 were oil coated.

Near-A-Lard #250-H Oil (Drawing Lubricant). Wire was treated with methanol (CH 4 0H) vapor during reaction furnace heat-up. Methanol fraction was controlled by CO monitor at coil outlet.

TABLE A.1.3 - Twist Schedule of 12 Tesla Coil Cable (Identical to LCP Westinghouse)

CABLE ELEMENT TWIST PITCH

1 (Single Strand) 2.5 cm* ( 1 in)

31 (Triplet) 2.5 cm ( 1 in)

32 (Triplet of Triplets) 3.8 cm (1.5 in)

33 7.6 cm ( 3 in)

34 15.2 cm ( 6 in)

6 x 34 30.5 cm ( 1 ft)

When triplets are made, individual strands are internally twisted on the same pitch as the triplet.

50 -32-

SUPERCONDUCTING STRAND

HELIUM FLOW CHANNEL

STAINLESS STEEL SHEATH

Figure A.1.1 Cross-section of the 12 Tesla Coil conductor.

51 AJdd.~4d.V ##r~tz~ ?8'

4'-'SFce

C g

~Lsje 4 (~~J wJ~ 'J~AhAJe4-~ r~ C.A.~~- hJ'

je,

A~V~-' __ *)~%L AA)~I44A~ ~L*~.- -- - -

52-- -34-

TABLE A.1.4 - 12 Tesla Coil Parameters

Number of Pancakes ...... '...... 6

Number of Subcoils ...... 3

Number of Turns Per Subcoil ...... 19

Number of Turns Total ...--...... 57

Potted subcoil ID ...... 14.7 in

Potted Subcoil OD...... -...... 33 in

Potted Subcoil Height ...... 1.86 in

Inner 'Compression Ring ID x Thickness ...... 14.00 in x 0.34 in

Outer Compression Ring OD x Thickness ...... 39.25 in x 0.625 in Case Height* 6:v(...0 t17 S -, 04r, C H ...... o 6.34 in

Self Inductance ... o ...... -...... 2 mH

Charge Rate at 0.1 Volt -...... 50 A/S

Dump Resistance (Per LLNL) ...... 2 mR

Discharge Time Constant (Isolated Coil) ...... 1 sec

Maximum Terminal Voltage for 20 kA Discharge ...... 40 Volts

C.0

rt ILL Olt

53 -35-

APPENDIX 2 - Helium Supply System

OVERALL FLOW CIRCUIT

The helium supply system is shown in Fig. A.2.1. Table A.2.1 defines the symbols shown in the figure. It is assumed that a 100 psig, 300 K helium source is available. Flow from this source will be regulated by valve VR1 and will pass through two heat exchangers supplied by LLNL. The 77 K helium will then pass through station STO7 and into the 2 meter diameter cryostat. There is a manifold at the inlet of the counter flow heat exchanger that divides flow into two separate paths.

PRIMARY FLOW CIRCUIT

The principal path, and one of highest interest, is through a liquid helium heat exchanger and into subcoil B. Flow into B is controlled by two cryogenic valves VC1 and VC2. In normal operation, VC1 is open and VC2 is closed, by-passing venturi meter VM1. The outlet flow from subcoil B passes through venturi meter VM2, back through the counter flow heat exchanger, and out station STO4 to be collected by the helium recovery system. Figure A.2.2 illustrates the primary flow circuit containing subcoil B. When quench tests of B are carried out, valve VC1 will be closed and valve VC2 opened. Thus all helium expelled from the coil will pass through the venturi meters and through the room temperature flow meter. Note that when steady-state mass flow tests of B are carried out, the parallel flow path through subcoils A and C will be closed by closing ball valve VB5.

54 Mf m V'2Z v 3eis

0 111

MVASS FALO WZ I FPRE~SSRES(.C FM MkS7 'If vmz 'VPM~ c1 Ic~'; c~H Rtl dl,:xl vcz.close4 VCZ op'pi I ---- I

______I -

-i lI L__ 1- -41 -r

55 -36-

t;*

al . ' 0

l~! m IM

4-~ A I J

00 Ii U - 'S.

I-Y-. 0

0 L) z 1.1' 0

-4e-

44 ~4j C 0 *0I 4

042 - . - 0.0 - s AU. a -

I td I a SI I -iy I I Nr -4 -4 I UrS I - -i- h-b - 0-.0M-,ySW w~-aa 4 1-i .

1- -i

N -4 t-- i-- 56 i iv= im i ~I I 1j I

N S .Th

* p

"0 - ~* Ali 0 0 ji 0 -- 0 0

1 5. 0 f~4~ - u-i

04.-i ~-~--

~' ~

7 0..

U MI I

2 ihV ~ K

I' 0- Zr:

* ~I ~ ~ *1 huh K5 ~ F ~ ~'0 '*~' ~ .,~ 6~ 40 .0 14 *o~o 0~ S 4 N- 6

~: ;~

57 -37-

TABLE A.2.1 - Definitions of Symbols in Helium Supply System Schematic

CF HX Counter flow heat exchanger ...... 0 DB Dielectric break ..... 0... 0...... 6...... F Filter 0.. ... 0...... 0. . ... 0...... 0 FM Flow meter ... .. 0.... 0... 0. .. 0. .. 0.....0..0...... 0 LHe HX Liquid helium heat exchanger 0...... 0. .. 0. .. 0...... 0...... P Pressure measurement point .0...... 0. . .. 0...... 0 ST Station (point of piping connection) 0...... 0...... 0...... 0. .. 0...... 0 T Temperature measurement point 0..0..0..0.. ... 0. . . .. 0...... 0..0...... 0...... VB Ball valve 0...... 0. . .. 0. .. 0. .. 0...... 0...... VC Cryogenic hand operated valve ...... 0. .. 0...... 0. .. .. 0...0. .. 0...... VCH Check valve .... 0..0..0...... 0...... 0 VR Regulator valve ...... 0..0...... 0...... 0..0. ... 0..0.... VS Safety relief value 0... 0..0. ... 0...... 0..0..0.....0. . .. 0...... VTH Throttle valve

58 -38-

SECONDARY FLOW CIRCUIT

Figure A.2.3 shows the secondary flow path through subcoils A and C. This path will not be used in flow tests unless there is a failure of pulse coil B.

OVERPRESSURE PROTECTION

The flow inlets and outlets of each subcoil are protected by pressure relief valves VS1-VS6. These valves will relieve at overpressures of 400 psia.

VENTURI METERS

Figure A.2.4 shows a cross-section of one of the two venturi meters at the outlet of subcoil B. Both meters have a throat diameter of 0.080 in (2.03 mm).

59 I I- - -1. AL I -39- I" I f. 4-11 I'M 4

%/ci

14~~

A~. Z.,-,?- CZ, tce4. C.L,. wt.. Wkd

±L~. k 4"Lit , f Cxl--) tk--4 4, rl-o ' P-4,

wl~vL-.f V"c. d

-AA A

Cl 'V.- )ib.wL

F F- (2 1*~ C~( I (-z C%4-

- 60 -40- w w z CL a @2 0 F".4 '4L* - a cl 3 ge P4 0 00 -. :14 0 OD'09 0 :C 0. S -4 - 0 U, U I0 c 0 C 0 -4 S I- M IL a a 4. U'

01. U 14 I a i 0 9 ~ - . -- S SI' 0.4 zc 0 I .4&I .4 S. R E .4 >~i .4 a Id 14 S C, ~ii

S 0 to 4

0 'I zw 0

LU j W4 P4 WJ .11

a I F.

w /3 d lii 0 40 -J d LL ,0 S z -ii N he.

0 61 C.rr ab-A ____

BULLETIN NUMBER 76-02 /V(I - r-..- C-e.-l ,/

CRYOGENIC VALVE - BRONZE k't e r ?

ES8 VALVE FEATURES

These extended stem cryogenic globe valves provide absolute shut-off for cryogenic fluids. Their self-aligning, resilient Kel-Fo seat seal assembly permits tight shut-off in adverse oper- 'I ating conditions and backseating permits replac- ing of the stem seal without reducing system pressure , should it ever become necessary. Extension tube and stem are fabricated from 304 stainless steel for strength and low conductivity which insures that the handle and stem packing remain at ambient temperature during continuous operation with any cryogenic fluid. It should be noted that these valves are equally effective at isolating the packing and handle from the high temperatures of steam or other hot fluids . The body and bonnet are pressure tested and fabricated from corrosion resistant bronze per ASTM 62-1. * Trademark-DuPont

ES8 VALVE SERIES 6 *Low conductive stainless steel extension *L body *Proven Kel-F seals and 300 tapered seat ES8 SERIES BRONZE VALVE *Economic, reliable service for most cryogenic fluids

62 4 ITEM: Extended-stem Cryogenic Valve SERVICE: Liquids; nitrogen, natural gas czzczzz(zZJiD (methane), oxygen (with special cleaning and packaging) RANGES: Operating temperature ± 400 F Operating pressure 300 PSIG (150 max.PSIG when above 2200 Proof pressure 450 PSIG - CONNECTIONS: IPS threaded ends per ANSI B2.1 E Socket ends per ANSI B16.11 3 OPEN Other ends available on request 4 MATERIALS:

6.0 ITEM PART NAME MATERIAL (Specification) TYP. ALL SIZES 1. Packing Gland Bronze (ASTM B-62) (N.A.) 2. Packing Teflon Impregnated Asbestos 3. Ext. Tube Stainless Steel (ASTM A-269) 4. Ext. Stem Stainless Steel (MIL-S-25043) 5. Seat Seal Kel-F S GAT x k( - F Remainder of parts: Pred. pem (ASTM B-62)

DIMENSIONS: (NOM.) WEIGHT (APPROX.) SIZE "A" "B1" SIZE WT. LBS. 1/2" 2 11/16 10 3/8 1/2" 1.7 3/4" 3 3/16 11 1/8 3/4" 2.5 ------1" 3 3/4 11 3/4 1" - 3.9 1 1/2" 4 3/4 13 11/2" 8.3 2" 5 3/4 14 1/16 2" 12.9 A Oro -t 't4 & r) N G7s w ORDERING INFORMATION:

SIZE LINE CONNECTIONS PART NUMBER 1/2" Female Pipe Thread ES8-84-2T1 3/4" " " " ES8-86-2T1 1" " " " ES8-88-2T1 1 1/2" " " " ES8-812-2T1 2" " " " ES8-816-2T1 1/2'' Socket Solder, Pipe Size (IPS) ES8-84-2WP1 3/4" " " " ES8-86-2WP1 1" " " " ES8-88-2WP1 1 1/2" " " " ES8-812-2WP1 2"1""1 ESd-816-2WP1

Cryolab BOX 6008 - LOS OSOS, CA. 93402 (a S*-mdI5

63 -41-

APPENDIX 3 - Instrumentation

Instrumentation for the 12 Tesla Coil can be arbitrarily divided into four groups: voltage measurement and heating systems; temperature, pressure, and flow measurement systems; coil current and magnetic field measurement systems; and acoustic measurement systems.

VOLTAGE MEASUREMENT AND HEATING SYSTEMS

Figure A.3.1 is a schematic of the voltage measurement and heating systems. Shown are voltage taps, the pulse coil heating system, pulse pick up coils, and the heater wire heating system. Each of the three subcoils has twelve (12) pairs of voltage taps as shown. Note that each subcoil has nine (9) turns per pancake and one (1) crossover turn to yield a total number of turns per subcoil equal to 2 x 9 + 1 - 19 turns per subcoil. Thus the entire coil has 3 x 19 - 57 turns. In addition to the 3 x 12 - 36 pairs of voltage taps on the coil, there is one (1) pair of voltage taps across the entire coil (VT 14) and four (4) pairs across the termination lap joints. Therefore, the sum total of voltage tap pairs is 36 + 1 + 4 - 41 pairs total. Figure .A.3.2 shows the voltage tap placement on a typical subcoil. Wire color codes are also given. Note that the reference view in A.3.2 is in the direction of gravity. Figure A.3.3 illustrates the voltage tap placement on the termination lap joints. Table A.3.1 lists the numbering of all voltage taps. 11e -42- F,'1 At.-f 7,

(CO-11 J) LC Ct4) - 0 vtt4 Y__ ftlt,-3w4,v AK-4 qe4i* A f4,-a

?ick up

Vnres VM A Y-rJA

r n,-e vfr Vr VITe, VU e A 4-~Z

YT, 4.),2 k.

rt.pjo., ~

*.,L tjoav

-ie , ct

C.~~~ 14 4L4 - W r

c. w -p Figre A.;i3

Figure A.3.1 Schematic of 12 Tesla Coil voltage tape, pulse coil Ubating sytem, pulse pick-up coils, and heater wire heating syutem. 65 "1- V j1 80 -43- I

00 Ref. ,- VT< - VT 172.' VT7 -e

?- Level 2, 4 and 6

Level 1, 3 and 5

G TS VT Io -

Ref.' View

VT4 VT1 -p,,.

k,- VT3 VTa -

- Vt 7-VT. CT - Tr

ukA/ JZI S'irj - V Tl10 Crossover Turn

Figure A.3.2 Schematic of voltage tap placement on typical subcoi:

66 -C - A --- e , 4

I IM 6 13

- &g: R4-Iv .

A - A.at{

-- -3 Vt 3

- ~ ~

-~-C. t-

Figure A.3.3 Voltage tap placement on termination lap joints. Also shown are VT 14, the voltage taps across the entire coil.

67 -45-

- TABLE A.3.1 - Voltage Tap Numbering. Taps are Stranded No. 28 Wire Insulated With Type - E Teflon.

SUBCOIL POSITION TAP COLOR NUMBER CODE A B C White/Red CROSSOVER VT 1 Stripe No. 11 No. 23 No. 35 VT 2 Blue No. 12 No. 24 No. 36 VT 3 Orange 13 25 37 TOP VT 4 Purple 14 26 38 VT 5 Gray 15 27 39* VT 6 Brown 16 28 Yr Il-b1 40

VT 7 Blue No. 17 No. 29 W0 1 No. 41 VT 8 Orange 18 30 42 BOTTOM VT 9 Purple 19 31 43 VT 10 Gray 20 32 44 VT 11 Brown 21 33 45

ENTIRE White/Black SUBCOIL VT 12 Stripe No. 22 No. 34 No. 46

Lead A A-B B-C Lead C Lap Joint Lap Joint Lap Joint Lap Joint

LAP JOINTS VT 13 Blue No. 1 No. 2 No. 3 No. 4

ENTIRE COIL VT 14 Orange No. 1

* Wire broken -- will not work -46-

Tfie pulse coil heating system consists of three (3) individual pulse coils, one on the crossover turn of each subcoil. The pulse coils consist of approximately 300 turns of .075" x .150" copper wire insulated with double dacron glass insulation to a build of .005" per side. Pulse coil layers are insulated by epoxy-glass and Kapton. Pulse coil B is of primary interest, since it surrounds the crossover turn at the highest magnetic field point in axial center of the 12 Tesla Coil. The three pulse coils have a common lead as shown in Fig. A.3.1, so that four (4) cables are needed to connect them to the room temperature power supply. Pulse coils A and C provide double redundancy in the event of an insulation failure in pulse coil B. The power supply consists of a nominal 87,000aF capacitor bank behind an ignitron switch. The ignitron is trig- gered by a voltage pulse from a pulse generator that is part of the power supply package. A shunt of 200uV/A is connected in series with the capacitor bank and will be used to manitor the current into a given pulse coil. In Fig. A.3.1, R.1 represents the resistance of the series shunt and V.1 the voltage across it due to a current pulse. The capacitor bank voltage will be measured by a voltmeter. The expected range of capacitor voltage is from 0-AK volts. The expected current pulse width is 10 ms. Table A.3.2 lists parameters associated with the pulse coil heating system. Figure A.3.4 shows the geometry of a typical pulse coil.

TABLE A.3.2 - Parameters Associated With the Pulse Coil Heating System

PULSE COIL PULSE COIL PULSE COIL A B C

TURNS 287 304 304

INDUCTANCE 72j H 81 UH 81iiH

R.T. RESISTANCE

CAPACITANCE ...... 87,00O0F

PULSE WIDTH ...... 10 ms 1-100 VOLTAGE RANGE ...... 0-'se V

SHUNT VOLTAGE 200.pV/A

PEAK CURRENT AT 5-06 V ...... > 10,000 A quo~

69 -47-

VT-1 0 0 Re

Ref. View

"dir.'/6vu-

CGR CGR (back side)

270

Pulse a 131 Pickup. Coil

Pulse Coil

1 '

-p _ G-10 Filler

Surface Mounted CGR Ref. View Section at ~850

Figure A.3.4 Sketch showing placement of carbon glass resistors, pulse coil and pulse pick-up coil.

70 -48-

The pulse pick-up coils consist of three (3) 20 turn coils of No. 28 stranded wire insulated with teflon. There is one pick-up coil per pulse coil. These coils are magnetically .coupled to the pulse coils and give voltage signals proportional to the time derivative of flux generated by the pulse coils. Figure A.3.4 shows the location of a typical pulse pick-up coil. Table A.3.3 lists the color code and numbers associated with the pick-up coils.

TABLE A.3.3 Pulse Pick-up Coil Color Codes and Numbers

SUBCOIL COLOR NUMBERS A RED 1 B RED C RED 3

Subcoil B contains an imbedded heater wire made by ARi Industries Inc. The wire consists of an Inconel heater insulated by magnesium oxide and encapsulated in a 347 stainless steel sheath. The heater wire has an RRR = 1.05 and a resistance at 4 K of 31.2 mg/cm, yielding a total resistance at 4 K of approximately 120. Table A.3.4 summarizes heater wire parameters. A forcing voltage of 300 volts. will be sufficient to supply heater currents for this experiment.

TABLE A.3.4 Heater Wire Parameters

OUTER DIAMETER ...... 0.063 in

HEATING ELEMENT ...... INCONEL

INSULATION ...... o...... MgO

JACKET ...... o...... *...... o...... 347 ss

RRR ...... o...... 1.05 ± .02

0 RESISTANCE PER FOOT AT 20 C ...... 10/ft

RESISTANCE PER CM AT 4.2 K ...... 31.2 mD/c

~ViJ~ RESISTANCE AT 4.2 K ...... - 120 0 ( ) <7

MAXIMUM FORCING VOLTAGE ...... 300 V

71 -49-

TEMPERATURE, PRESSURE AND FLOW MEASUREMENT SYSTEMS

Figure A.3.5 shows simplified schematics of the primary and secondary flow circuits. Most of the 12 Tesla Coil experiment will be done with zero helium mass flow. However, when flow experiments are under- taken, the present helium supply system allows for flow in subcoil B, while subcoils A and C remain at zero flow. Thus, flow experiments will be focused on subcoil B.

The arrangement of temperature, pressure and flow instrumentation is shown in Fig. A.3.5. Temperatures are measured by carbon glass resistors located at the inlet, center and outlet of each subcoil. The CGR's at the inlet and outlet measure helium temperature, while those at the center measure sheath temperature. Figure A.3.4 shows the locations of CGR's at subcoil centers. Table A.3.5 lists parameters associated with the CGR's. There are a total of eleven (11) CGR's yielding 44 lead wires.

TABLE A.3.5 Carbon Glass Resistor Parameters (Lake Shore Cryotronics Model CGR-1-2000)

WIRE CODE FOUR MAX. LEAD SUPPLY FIG. A.3.5 SERIAL LOCATION RESISTANCE VOLTAGE NUMBER NUMBER YELLOW GREEN WHITE BLACK (4.2 K)

T1 C4678 Inlet -B No. 7 No. 7 No. 7 No. 7 1.91 kQ -' 10 mV

0 BT-270 C4676 Center -B 4 4 4 4 2.07 v- 10

0 BT-90 C4674 Center -B 3 3 3 3 1.83 v" 9

T2 C4681 Outlet -B 8 8 8 8 2.21 / 10

T3 C4719 Inlet -A 9 9 9 9 2.17 10

0 AT-270 C4670 Center -A 2 2 2 2 1.76- 9

AT- 900 C4072 Center -A 1 1 1 1 1.79 9

T4 C4722 Outlet -A 10 10 10 10 1.76 9

-/ T5 C4750 Inlet -C 11 11 11 11 1.94 - 10

0 CT-270 C4703 Center -C 6 6 6 6 1.78 v 9

CT-900* C4677 Center -C 5 5 5 5 1.88 V 9

T6 C4751 Outlet -C 12 12 12 12 2.01 / 10 * Wire broken

72 -A _ -I

I~,EJd~.A ~ 2L I ~ - I (I I.

______(77, 11K) t.dc

Cl q678 0 o-f2- a. SL2 j'1iOXL

aS-r-2Z70 0 CL4 47(o I qQ 0- -

AqT7O0 C 44 10 Cl--A 147q5.9 1o

~T- l 0 C io'7-L cefc- -I L79 I-T. 1T9 0

TT c 47so ILt - C. 9,07 174 Wo

C.- 270' C'?o73 1-IN 7.(o '3z. o 78o

0 CTr- 0 C47 7 /5.4 bAC s"P o8

73 SERIES CGR-1 CARBON GLASS RESISTANCE TEMPERATURE SENSING ELEMENTS Technical Specification: CGR-1

SERIES CGR-1 CARBON GLASS RESISTANCE CRYOGENIC TEMPERATURE SENSING ELEMENTS

e Stable * Excellent in Magnetic Fields * Monotonic R vs T and d 1nR/d 1nT 9 1 to 300 Kelvin Range 9 Rugged Construction

The CGR-1 series Carbon Glass Resistance Temperature Sensors represent the best choice for a highly reproducible temperature sensing element for the range 1-300 K in high magnetic fields. The full temperature range capability extends the usefulness of the CGR-1 series (over Germanium Sensors) in applications not involving high magnetic fields. The configuration of the carbon glass sensing element is the result of an extensive computer aided design analysis based on a primary requirement that the element be of the classical four lead potentiometric configuration. The result is an element typically 4 millimeters long with half millimeter voltage ears. Each of the models listed in table I are designed-nd optimized for the specific recommended temperature range with emphasis on achieving specific sensitivities (d 1nR/d 1nT) over the recommended range. Each model will exhibit monotonic response to 300 K regardless of the recommended range. Every completed CGR- 1 sensor is quality tested to assure that the encapsulation is leak tight and the sensor is stable when cycled from room temperature to liquid helium temperature. Typical response to magnetic fields is shown in Table 11. TYPICAL SENSOR SPECIFICATIONS: Temperature Reproducibility: =0.00075 K Maximum (:0.005 K Typical) when thermally shocked between room temperature and liquid helium Size: 3mm (0.120") diameter by 8.5mm (0.335") long cylinder, four (4) each 6" long phosphor- bronze insulated leads having epoxy strain relief at sensor. See figure on reverse side. Special packages available on request. Internal Atmosphere: Helium 4 is Standard. Helium 3 or Nitrogen are available as options. Materials of Construction: See construction details on reverse side. Data Supplied: All testing and calibration data supplied is measured with D.C. power. Included with each uncalibrated sensor are a 2-wire (continuity) resistance at ambient, liquid nitrogen, and liquid helium temperatures; and 4-wire (potentiometric) resisiance at ambient, liquid nitrogen, and liquid helium temperatures. These values should be considered as nominal rather than precision calibration data points. TABLEI Model Typical Resistance at 4.2 K (ohms) Suggested Useful Range (Kelvin)

CGR-1-50C 350-750 1 to 77 (300) CGR-1-100C 750-1300 1.5 to 100 (300) CGR-1-1500 1300-1750 2 to 100 (300) CGR-1-200C 1750-2400 2.5 to 100 (300)

Consult factory for availability of other resistance values

TABLE I Typical Magnetic Field-Dependent Temperature Errors' for Carbon Glass Sensors T(K) 2.5 tesla 8 tesla 14 tesla 2.1 0.5 1.5 4 4.2 0.5 3 6 15 <0.1 0.5 1.5 35 < 0.1 0.5 1 1.5 77 < 0.1 0.5 Maanitude of raIntiva tarnnaratetra arrnr ATIT I.1 #^r R - 1 A SERIES CGR-1 CONSTRUCTION DETAIL 3mm -(0. 120") Lead Identity for Carbon Glass Resistor A

Key B - A. Index Mark Lead Color C 0 D 0 B. V Yellow 0 C. V Green C D. White D QD +I E. I Black 8.5mm (0.335") A E Limit R

VEMF oCurrentr C

YV+ v- G A. Gold-plated copper enclosure W~ L------,-Bk B. Current contact zone C. Phosphorus-bronze leads 0.20 mm (0.008" dia.) CGR-1 Resistor 0. Sensing element E. Gold leads 0.05mm (0.002" dia.) Basic Measuring Circuit F. Epoxy heat sink (voltage drop across sensor should be kept to 10mV or less) G. Beryllium oxide base

CGR-1 CARBON GLASS RESISTANCE THERMOMETER TEMPERATURE VS. RESISTANCE TYPICAL CURVES Recommended Current (Maximum) 0.1 uA

------.

- ... L ------. -

10,000 1 uA

- I - - - -. - - -- -

Resistance (ohms) 1,001 10 uA

ie-. - ~J. - -. - - --- N

100 I I i I X 1XXI I I I I I I I IlUUUA N

10L I 1L 1 mA 1 10 : 100 300

PRIN%7E '._ -:A Temperature (Kelvin) 75 _114r, Z _u f I4

LAKE SHORE CRYOTRONICSINC. 64 EAST WALNUT STREET WESTERVILLE, OHIO 43081

DIP TEST DATA FOR:

SENSOR MODEL: CGR-1-2000 SERIAL NO.: C4750 DATE: 16-NOV-83

TWO LEAD RESISTANCE(I+ TO I-):

ROOM TEMPERATURE 0.16E+02 OHMS LIQUID NITROGEN(77.4K) 0.32E+02 OHMS LIQUID HFLIUM(4.2K) 0.36E+04 OHMS

FOUR LEAD RESISTANCE

ROOM TEMPERATURE 0.887E+01 OHMS LIQUID NITROGEN(77.4K) 0.174E+02 OHMS LIQUID HELIUM(4.2K) 0.194E+04 OHMS .4-

SELF-HEATING CONSIDERATIONS:

TO MINIMIZE SELF-HEATING DURING CALIBRATION LAKE SHORE ADJUSTS THF CURRENT TO THE SENSOR TO MAINTAIN SENSOR OUTPUT VOLTAGE AT THE LISTED VALUES:

ABOVE 1 KELVIN 1 TO 3 MILLIVOLTS 0.1 TO 1 KELVIN 0.1 MILLIVOLT BELOW 0.1 KELVIN 0.03 MILLIVOLT

MAXIKUM ALLOWABLE CURRENT TO AVOID DAMAGE TO SENSOR IS 200 MILLIAMPS

LEAD COLOR CODE:

WHITE: I+ BLACK: I- YELLOW: V+ GREEN: V-

76 4 -50- F: A_A -I-,i o 4,3-11

vCi

P2. DPZ

LO C- S '

S 3

Se 7 Cow- Circ ;+ - S,,Ls A C

Figure A.3.5 Simplified schematics of primary and secondary flow circuits. 77 -51-

Absolute pressures are measured by sputtered strain gauge transducers (CEC 1000). These devices have a pressure range from 0 to 1000 psia. Table A.3.6 lists parameters associated with the pressure transducers. There are a total of five (5) PT's yielding 20 lead wires.

TABLE A.3.6 Absolute Pressure Transducer Parameters

MODEL ...... CEC 1000 EXCITATION ...... 10 V dc rated; 15 V dc max. PRESSURE RANGE ...... 0 - 1000 psia

WIRE CODE INPUT FIG. A.3.5 SERIAL LOCATION RESISTANCE SENSITIVI NUMBER NUMBER RED PURPLE GRAY BROWN (200 C)

P1 .62a4- Inlet -B No. 5 No. 5 No. 5 No. 5 -4O3. 2973-mUi 2 9 P2 4.6242 Outlet -B 6 6 6 6 -42

P3 4+2t5- Inlet -A 7 7 7 7 992

P4 g 0 16046- Outlet -A 8 8 8 8 464

P5 2t(L4'4

P6 0cj Od_ U 4 - (-

It is worth noting that the PT's contain magnetic materials and will therefore require good mechanical support. They will be located in the an- nular gap between the HFTF coil stack OD and the cryostat ID.

Supercritical helium mass flow will be measured using two venturi meters with differential pressure transducers. The differential pressure transducers are made by Validyne Engineering Corporation. The vendors wiring diagram shows each device to have three (3) wires. Therefore, there

V'_i-V' o'tt. s (:1~

78 -52-

will be a total of six (6) wires required for the differential pressure measurements. Table A.3.7 lists parameters associated with the differential pressure transducers.

TABLE A.3.7 Differential Pressure Transducer Parameters

MODEL Validyne P10 EXCITATION (Supplied by MIT 5 V rms, 3-5 kHz OUTPUT 30 mV/V full scale nominal PRESSURE RANGE ...... ± 80 paid

WIRE CODE FIG. A.3.5 SERIAL LOCATION NUMBER NUMBER WHITE/ WHITE/ RED STRIPE BLACK STRIPE ORANGE

DP- 1 Inlet -B 2 2 2

DP-2 Outlet -B 3 3 3

COIL CURRENT AND MAGNETIC FIELD MEASUREMENT SYSTEMS

Coil current will be measured using a shunt that is part of the LLNL power supply. See Fig. A.3.6. Magnetic field at the 12 Tesla Coil midplane will be measured by a Hall probe located in a hole in the HFTF inner compression ring. Parameters of the Hall probe are given in Table A.3.8. This device has four (4) wires. See Fig. A.3.6

TABLE A.3.8 Hall Probe Parameters

MODEL ...... * ...... BELL BHA-921

EXCITATION ...... 100 mA Constant dc Current

MAGNETIC SENSITIVITY ...... 6.52 mV/Tesla

SERIAL NO...... 2083

ZERO FIELD INPUT RESISTANCE ...... O.6989

ZERO FIELD OUTPUT RESISTANCE o...... ,...... O.6262

WIRE CODE:

CONTROL CURRENT ...... RED-50, BLACK-50

HALL VOLTAGE ...... o.*...... BLUE-50, ORANGE-50

79 a .~ 1TpI.a-( *Vpuul --a-3-~:, ~- 1 -'-a PTP%"g 'VIIIP'7

P.-4c

4-A u ~ f A 44A

Figure A.3.6 Schematic of Hall probe location and current shunt. Current into the lead of subcoil A produces a downward magnetic field (in direction of gravity).

80 -54-

ACOUSTIC MEASUREMENT SYSTEMS

Six differential style acoustic emission sensors were installed on the 12 Tesla Test Coil on Wednesday, April 18, 1984. The coil is constructed of three subcoils. A sensor is attached to both leads of each subcoil (See Table A.3.8). Five of the six sensors are attached to the lead wires inside the coil case. They are well protected. Only sensor #10, on lead A Level 1, is exposed.

The sensors are held in place by G-10 strips which act as springs. Mylar tape electrically insulates the sensors from the coil. The cable running from each sensor is white teflon coated, shielded, twisted twin- ax (Belden #83316).

TABLE A.3.8 - Acoustic Sensor Layout

LEVEL PANCAKE SENSOR RELATIVE SENSITIVITY 1 A 10 1 (Defined)

2 A 16 2.7

3 B 13 2.5

4 B 15 2.2

5 C 14 3.0

6 C 11 1.5

Sensors numbered on end of cable

81 -55-

1. TJ~} L~L~

Lw~ I

Le~4 4.f

~mm~;~z

82 Figure A.3.7 Acoustic sensor locations. -II

kz1, ~V D.2 rf

0k3

It-

iLI-4o VC10- -zLufuj a______41r 0 al- a 10 I

e? rl 's-/ Lji

u-j- C- kz'

. It a

L 83 -56-

APPENDIX 4 - Termination Lap Joints

DESIGN CONSTRAINT

The 12 Tesla Coil lap joints are constrained by the physical size of the HFTF cryostat. The principal restriction involves lap joint length as shown in Fig. A.4.1.

DETAILS

There are four (4) termination lap joints associated with the six (6) terminations of the 12 Tesla Coil. Each termination has been swaged to a final diameter of 0.775 inches (2 cm) and has a cross-section similar to that shown in Fig. A.4.2. The terminations are approximately 9.75 inches long. The ends of the terminations are plugged by 1/8 inch copper caps that have been welded on. Leaks in the welds, found after reaction of the coil, have been plugged by brazing.

Lap joints have been made with pairs of OFHC copper blocks of geometry shown in Fig. A.4.3. Lap joints have been made by soft soldering pairs of these shunt blocks to the terminations. The solder composition is 60/40 (Sn/Pb). Estimated lap joint resistance is approximately 1-2 n.

84 -57- -57-7,A.d& A. (z4 1 J1~y

95 cm 1R

/e I I

L. 13c.;i- k 7f I-I-CI

3 S.i 4

04 * 7o 0o too

R (cm)

Figure A.f.l Relative position of the 12 Tesla Coil termination lap joints in the Livermore High Field Test Facility (HFTF). Coil lap joints are constrained to lie in a 5 cm gap between two stainless steel flanges. The approximate field profile at the lap joints is also shown.

85 U, I

roee

*.-~Oso

I4 fis v.

Figure A.4.2 Typical termirfation 'crcas-section (diameter - 0.775 in).

86 ..------.-- -3 R -

1 .34. Rys

1 MAT'Ll. CDA IOZ COPPER (OFMC) TO T .2ST

LEAD SHUN*T iJO.1 DMI

S .3qet

PAIL CD0A IOZ COPPIER COPwH) JbiNT SHUNT DMIL

Figure A.4.3 Copper shunts used in 12 Tesla Coil termination lap joints. -60-

APPENDIX 5 - Transient Stability Model

MAGNETIC DIFFUSION TIME CONSTANT

The approximate magnetic diffusion time constant of the Airco 0.7 mm diameter wire at 10 Tesla is

2 Tm = 0 a (cuAr). (4 x 10 -7)(1.25 x 109)(1.5 x 10-4)2 - 35 ys

where ah e = copper stabilizer conductivity [Q_ *m. 3

Ar W copper stabilizer radial thickness [m]

Thus for AC fields at approximately 50 Hertz (T - 20 me), one can assume complete penetration of the magnetic field into individual strands.

ENERGY DELIVERY MODEL

The energy delivery model treats the ICCS inside the pulse coil as the secondary of a transformer, a technique commonly used to analyze non- destructive eddy current testing. At 50 Hertz, the frequency of interest, hysteresis losses in the superconductor are small compared to eddy current losses in the wire. * The model, therefore, is solely one of eddy current heating. The cable and sheath are modeled as resistors that have been reflected from the secondary to the primary circuit. The energy extracted from the capacitor bank in the first half cycle divides in the primary circuit according to the relative strengths of the primary resistors. Hence knowledge of the reflected resistors leads to an estimate of the energy delivered to the ICCS cable. Figure A.5.1 shows the lumped parameter energy delivery circuit model. It is worth noting that the inductance L represents that of the pulse coil with a lossy core. That is, the pulse coil inductance, as calculated in free space, is lowered slightly (~ 3%) by the presence of the ICCS. '-'I-

IrniteIn Le&c1 & IC. Skedt+I

R R RL

t V0

Figure A.S.1 Lumped parameter energy delivery circuit model.

09 -o

ALGORITHM

An algorithm to calculate specific energy delivered to cable is presented below in Table A.5.1. It states that the energy delivered to the cable in- one half cycle equals the ratio of the reflected cable resistance to the total series resistance multiplied by the total energy delivered in one half cycle. It is assumed here that the switch is a rectifier that allows a half sine wave current pulse. Parameters used in the energy delivery model are summarized in Table A.5.2.

TABLE A.5.1 - Energy Delivery Algorithm

1. Energy delivered in one half cycle.

E 1/ '(1/2 02) E1/2o CV [J] initial stored fraction delivered stored energy in 1/2 cycle

where C - capacitance [F]

Vo W initial voltage on caps [V]

T - period of damped sine wave [s]

T - time constant []

2. Period of damped sine wave.

1 2w LC T - - =- - = 27 [s] f W l-d

where f - frequency [s-1 ]

L - equivalent inductance (pulse coil with lossy core plus stray inductance) [H]

L - Lpc + Ls

2C d - -- damping constant 0 [-] 4L

RT - Rr X Ri + Rpc + Rs + Rc - total series resistance

X - leads r - ignitron

s - sheath pc - pulse coil c - cable

90 -0.3-

3. Time constant

2L T R [s]

4. Sheath resistance

1 24

Rs E ( ) asro2r2E f 2

where Ve - volume of sheath [M3

N p - - pulse coil turns per unit length tp

as - sheath conductivity

ro - eequivalent outer radius [Mi]

ri -- M ratio of inner to outer radius [- J ro

5. Cable resistance

Rc = R1 , + RI

where

R1 1 - - a1R Cs 2 2 R 2 )2 RR

Ri = Vc( N) Lf 20. + acuR2W2 sin2 2p 1 -64-

where

Vc M volume of cable metal Em[3 3J

ac - conductivity of copper stabilizer [ - e M-1 I

R - strand outer radius [m]

R1 - bronze matrix outer radius [m]

p - angle between pulsed field and strand axis [deg]

Lf - filament twist picth [m]

a1 = -a matrix conductivity [Am ]

A - fraction of superconductor in matrix [-]

am - bronze conductivity [A *m ]

6. Energy delivered to cable in one half cycle.

Rc Ec -- * E1/2 (J] RT

7. Energy delivered to pulse coil in one half cycle

Rpc E[A RT

8. Current

V0 i(t) e-t/T sin wt [A] WL

V0 Ip -- e-t/T (A] to f T/4 WL

92 9. Specific energy delivered to cable

Ec 1 Rc By volume: ecV - - - - - * E1/2 Vc Vc RT

Ec 1 Rc By mass: ecm - - - - E1/2 mc VCYC RT

where

Vc - cable volume [M3

tmc - cable mass [kg)

10. Capacitor voltage after discharge [ VJ

Vc M Vo e-T/2T

93 TABLE A.5.2 - Parameters Used in Energy Delivery Model

C - 87,000PF Lpc - 81pH

Rr = 5 mQ Ls 19yH

3 Vc = 1.26 x 10-4 m L - 100H pcu(RRR = 50, B) - (3.2 + 0.48 B) x 10 y1m p __ (RRR - 100, B) - (1.6 + 0.48 B) x 10 Sem

304 x L 0.649 m N8p 486 ds = 7 x 10-4 m Ns =

1.04 R - 3.5 x 10-4 m fh

390 R 1- 2.15 x 10-4 m

2.5 x 10- 2 m X - 0.21 LV - 3 m - 1.1 x 10 a 7.826 x 10-5 m Vs =

67 x 10-8S2m a i - 1.68 x 107 a 1m-1 ps - 0.012 m - m( ri - R = 45 r = 0.0103 m Rp - 5.65 x 10-4 (1.6 + 0.48B)

94 7.0 APPENDIX 2 - HFTF INTERFACE SPECIFICATION

The interface specification for the 12 Tesla Coil Program was issued as a reference document to the four program participants in 1979. It describes the test coil design requirements and other factors associated with the Livermore High Field Test Facility interface.

95 INTERFACE SPECIFICATION FOR 12 T INSERT

TEST COILS (PRELIMINARY)

University of California

Lawrence Livermore Laboratory

Livermore, California

November 1979

96 INTERFACE SPECIFICATION FOR 12 T INSERT TEST COILS (PRELIMINARY)

Contents

1. Design Requirements

2. Coil Orientation

3. Space Constraints

4. Mechanical Interface

5. Electrical Interface

6. Diagnostic Equipment

7. Thermal Interface

8. Magnetic Interface

9. Quench Protection

Sketch SK-1000: Coil Data

Sketch Sk-1001: High Field Test Facility--4 m Diameter Cryostat

Figures 2A, B, C, & D: Typical Quench Performance

Dwg. AAA-79-104494-00: Preliminary Layout of HFTF Coils

Dwg. AAA-79-112148-00: 12 Tesla Coil Assembly, Test Coil Lead Space

97 INTERFACE SPECIFICATION FOR 12 T INSERT TEST COILS (PRELIMINARY)

1. DESIGN REQUIREMENTS

1.1 The conductors used in the test coils are to be candidate prototypes for Tokamak E.T.F. toroidal field coils. The conductors are, therefore, to be designed to meet the following "modified E.T.F." specification:

1.1.1 Plasma major radius, - 5 m

1.1.2 Coil major madius, = 5.8 m

1.1.3 Coil bore size, 6 m horizontal x 10 m vertical

1.1.4 Coil shape, modified D-shape

1.1.5 Number of coils, 12

1.1.6 Field on plasma axis, - 5.8 T

1.1.7 Peak field at the winding, 12 T

1.1.8 Conductor, Nb3Sn or alloyed Nb-Ti*

1.1.9 Field profile, maximum on the center line

1.1.10 Amp turns/coil - 12 MA

1.1.11 Current density over winding pack, ie1700 A/cm2 ; over conductor, i-tte- A/cm2

1.1.12 Operating current, 10 to 15 kA

1.1.13 Stored energy/coil, - 1500 MJ

1.1.14 Stability margin: one-half turn length at the high-field region can withstand 100 mJ/cm3 without undergoing a quench.

1.1.15 Tolerance for pulsed fields*

A. Normal operation

1. B : 0.15 T/sec for one second up, one second down 2. B 11: 0.15 T/sec for one second up, one second down 3. Repetition rate: one pulse every five minutes

B. Upset condition

Simulation of a plasma disruption by a downpulse in By of 0.5 T in 0.10 second with a repetition rate of once a week.

*As revised at 12 T meeting. 98 1.1.16 Tolerance for radiation, 1 x 109 rads 1.1.17 Vacuum topology: bell jar with reentrant holes

It should be noted that the specifications are the minimum to be achieved. If, for example, the Seller can obtain a viable design while achieving a higher overall current density, then such an approach is acceptable. Grading of conductor and structural design is left up to the preference of the Seller.

1.2 The test coil is to be designed to meet the following criteria when tested in the HFTF:

1.2.1 Carry design current (between 10 and 15 kA) in a maximum field of 12 T.

2.2.2 Shall not quench when an energy of 100 mJ/cm 3 is deposited in the conductor in the high-field region over the length of half a turn and in a time not exceeding 50 msec. The thermal environment of this portion of the conductor shall not be significantly different from that in the remainder of the coil. 2. COIL ORIENTATION

The toroidal field coils of a Tokamak operate with the axis horizontal. Since the heat transfer properties of a pool-boiling conductor may be affected by its orientation, arrangements will be made for the test coils to be tested with the axis horizontal. Sketch Sk-1001 shows this arrangement diagrammatically. 3. SPACE CONSTRAINTS

The test coil will be inserted in the HFTF between the two Nb3Sn coils. The overall dimensions of the coil, complete with any case or cryostat in which it may be mounted, but excluding its leads and any coolant connections to the coil., shall not exceed:

Outside Radius !! 48.34 cm (to be confirmed) 11 Inside Radius =_ 18.49 cm (to be confirmed) -7't Axial Length t- 20.32 cm (to be confirmed)

4. MECHANICAL INTERFACE

The arrangement of the HFTF coils with and without the test coil is shown on Drawing AAA-79-104494-00. All dimensions are nominal at this stage of the design. When the design of HFTF is frozen, final dimensions together with tolerances will be specified.

4.1 The HFTF is designed so that the magnetic forces on its coils are self-contained and are not transmitted to the test coil. There will be a small deflection of the thick end plates bounding the space for the test coil resulting from the axial forces on the HFTF Nb3Sn coils. This deflection will be less than 0.050 inch.

99 4.2 The test coil is located radially by being a sliding fit over the inner compression ring.

4.3 The test coil will be positioned axially by the thick end plates which are attracted towards each other by the magnetic forces on the Nb3Sn coils and which are restrained by the inner and outer compression rings. Some adjustment, by shimming either the test coil or the compression rings, can be accommodated.

5. ELECTRICAL INTERFACE

5.1 Coil Insulation Test. The coil shall be fully insulated from its case and shall withstand a test of 500 V dc for one minute. During this test, any diagnostics attached to the conductor, such as heaters, strain gages, etc., shall be connected to the coil; any diagnostics attached to the case, such as helium level and temperature sensors, etc., shall be connected to the case during this test.

5.2 Main Current Leads. Two 15 kA vapor-cooled leads will be provided by the Facility to carry the current to the test coil. The interface will take place at the low-temperature end of the vapor-cooled lead, where a bolted connection will be made to the leads from the test coil.

Details of the interface to be specified later.

The leads from the coil to the interface will be such that, when carrying full current, the temperature at the interface shall not exceed 10 K.

5.3 Auxiliary Leads

5.3.1 The Facility will provide the following diagnostic leads, terminating at panels in the region of the liquid helium level, which will be the interface with the diagnostic leads from the coil:

5.3.1.1 Thirty heater channels, each two leads, untwisted, unshielded. Rating, - 25 A pulse.

5.3.1.2 Ninety shielded, twisted pairs for voltage taps; 600 volt rating (1000 V test) between wires and to ground.

5.3.1.3 Twelve triple shielded leads for strain gages.

5.3.1.4 Six four-conductor leads for thermometers, etc.

5.3.2 Diagnostic leads from the test coil shall be securely anchored and suitably labelled by the fabricator 'and be not less than the length to be specified.

6. DIAGNOSTIC EQUIPMENT

6.1 One power supply is available for energizing heaters or small pulse coils. The output voltage can be controlled from 0 to 120 V an~d te maximum current is 50 A. The output is square wave with the time

100-

IZ~T pza.. '2tcv A controllable from 1 mec to 1 sec. The unit can be switched to any number ofV30 parallel channels.

6.2 Twenty B&F signal conditioners capable of operating in either constant current or constant voltage mode.

Maximum Current = 150 mA

Maximum Voltage = 300 V

6.3 A modcomp 1,300 wide range A/D System is available for data acquisition and analysis.

Total number of channels (including those for HFTF) = 104.

Maximum Sampling Rate = 300/channel/sec.

Voltage Range =-5 mV to L10 V (automatic range gains)

Up to 48 channels can be routed via Preston amplifiers-- approximately 30 channels available for test coil.

Maximum Common Mode Voltage = 1000 V

Maximum Input = 10 V

A further 56 channels are available, not through amplifiers, but through relays, to isolate the diagnostics in the event of a quench in any of the coils, ie., either the test coil or HFTF--approximately 30 channels available for test coil.

Honeywell 18-Channel U/V Recorder

Biomation 4-Channel Analogue Recorder; Variable Rate, 1000 Samples/ Channel

7. THERMAL INTERFACE

The HFTF operates submerged in liquid helium at nominal atmospheric pressure, approximately 4.3 K. Pool-boiling test coils, designed to operate in these conditions, may be supplied as open-type coils.

7.1 Coils designed to operate at pressures or temperatures other than these must supply the coil, sealed in its own case, complete with any thermal insulation that may be required. In these instances, the supplier must also provide all equipment and coolant interconnections to enable the coil to operate under the designed conditions.

8.0 MAGNETIC INTERFACE

The coil data given on SK-1000 can be used for calculating the magnetic field on the test coil. The dimensions and average current density given in this table for the test coil (number seven) are nominal value and should be replaced by actual values.

101 9.0 QUENCH PROTECTION

The HFTF coils are energized and discharged in pairs to maintain symmetry about the center line. The inner Nb-Ti coils (Nos. 1 and 3), the outer Nb-Ti coils (Nos. 2 and 4), and the Nb3 Sn coils (Nos. 5 and 6) are discharged through center-point-grounded resistors.

9.1 The inductance matrix for the coil system is given in Table I.

9.2 The values of the protective resistors are given in Table II.

9.3 A value of 0.002 ohms is proposed for the protective resistor for the test coil; the maximum voltage generated at 20 kA is, therefore, 40 volts. In this case, one end of the resistor will be earthed and the circuit will be interrupted by a single-pole circuit breaker.

9.4 LLL has a QUENCH computer program, which calculates the approximate discharge characteristics of the complete coil system during a quench or a trip (circuit breakers opened, but no quench). Figures 2A, B, C, etc. indicate typical performance; other cases can be computed on request.

102 71 A\A ci'. go-

TABLE I

Inductance Matrix for Coil System

M(1, 1) = 8.48 M(1,2) = 3.87 M(2,2) = 5.16 M( 1, 3) , 3.46 M(2,3) * 1.43 M(3,3) = 8.48 M(l, 4) * 1.43 M(2,4) = 6.39 M(3,4) = 3. 87 M(4,4) = 5.16 M(1,5) = 0.31 M(2,5) = 0.20 M(3,5) =0. 15 M(4,5) = 0. 06 M(5,5) = 0.073 M(1,6) = 0.147 M(2,6) * 0.06 M(3,6) =0.31 M(4,6) = 0.20 M(5,6) = 0.0098 M(6,6) = 0.073 M(1,7) = 0.06 M(2,7) = 0.026 M(3,7) = 0.06 M(4,7) = 0. 026 M(5,7) = 0.0067. M(6,7) = 0.0067 M(7,7) = 0.0042

Note: Coil No. 7 is assumed to have 86 turns, operating at 10 kA. Inductance values are in Henrys.

TABLE II

Resistance of Protective Resistors (Ohms)

Coil Number Resistance

1 + 3 0.4 + 0.4 2 + 4 0.27 + 0.27 5 + 6 0.02 + 0.02 7 0.002

103 -NO. 4 NO. 3 NO. 1 NO. 2

-T

NO. 6 NO. 5 LNO. 7

-Z . L +Z

7 N -. XV 4 1XI-1 -i

IX1 x

-- 5 CM

COIL COIL FORM (CM) WINDING (CM) AMPS NO. I.D. X O.D. LENGTH I.D. X O.D. LENGTH Z SPACE Cmz N

100.0 x 177.8 34.29 105.0 x 172.6 26.25 + 8.35, +34.60 3200.0 7 2 100.0 x 177.8 27.29 105.0 x 170.0 19.90 +38.70, +58.60 3200.0 3 100.0 x 177.8 34.29 105.0 x 172.6 26 .25 -34.60, - 8.35 3200 .0 L4 100.0 x 177.8 27.29 105.0 x 170.0 19.90 -58.60, -38.70 3200.0 5 37.0 x 100.0 34.96 40.0 x 91.0 24.25 +16.01, +40.26 3079.0 6 37.0 x 100.0 34.96 40.0 x 91.0 24.25 -40.26, -16.01 3079.0 T 37-0x- -965---p-0632- E

(AL COIL DATA

104 SK-1000 TELEPHONE CONFERENCE MEMORANDUM WCS 'NGHOUSE FORM 22822 A DATE Jan. 3, 1980

INCOMING F OUTGOING c.O. -G.0. 90P0735

MR. Don Cornish OF THE LLL

WITH MR. OF THE

COP IES TO: A. .. JIrbak

A. Mn-ntnmaru (MTT) - ~ -

SUBJECT: Interface specs for 12 T coil with HFTF TIME

COST

FILE CHARGE

DETAIL OF CONFERENCE

" The current density has been reduced in the HFTF coils as a result of more detailed

analysis.

" The nominal current and turns for each coil are as follows:

Outer NbTi (2 and 4) I = 1200 amp N = 1793 turns

Inner NbTi ( 1 and 3) I = 1200 amp N = 2366 turns

Nb3Sn Coils (5 and 6) I = 5000 amp N = 380 turns

" Symmetrically opposite coils are connected in series. The series lead is brought out

of the dewar.

" There is a separate power supply for each symmetric coil pair and an additional

supply for the 12 T coil itself (4 supplies total).

" Dump resistors are permanently connected across each power supply.

.C1

DEPT. Advanced Programs EXT. NO. 3077 105 (SI GNATURE) 0~

I1, I ;

* 8 0- &,C,..

0. .1 *~ .6 P

6 m -0o 4.' 0 p ~ &T-HIGH FIELD TE5T FACILITY 4 METER DIA.~CRYOSTAT-

106 SK-1001 -(V) 'WIejjnouS qN

0 -1O 0

oN o o)'dwe 0ua 0 N 0 coW N

z qr- m

1cn

CL 0

E~~ z ~ii C

4-f

D C4 0 f - N I: (V)'uejno0 0q

107 ED (V) 'iusaun3 u~ qN

o 0 0 0O o(I0 0 0 o 0 0 0

LU 00 0 0 0 (V)Iujjn~j-N

3 (V) IUs~:qN wU8Jifl 0 0 0 0 0 Ln0 0 0 0 0 qom 0 0 0 (N I-- 0 0

CN c; 0O

C 4- C,,

E 4-0 (N S

9 z OOL X) V~suI (Vluejn: w

(N 0 o e N 0 0 z (QOL X) !I-qN (V)iuwen3

109 ZU ~ (V) 'iuann uS q

4-0

1- 10

0 z (OOLX)I-qN (V)IaI I - - I NwWMWWRq- I-,.-.-

8.0 APPENDIX 3 - OPERATION OF THE HFTF

On March 28, 1985, J.R. Miller of LLNL issued report UCID-20394 that describes the assembly and operation of the High Field Test Facility during the MIT 12 Tesla Coil experiment. It is included here as an archival reference.

111 UCID- 20394

THE OPERATION OF HFTF TO TEST THE MIT 12 T COIL

JOHN R. MILLER

MARCH 28, 1985

V \z~ V

This is an informal report intended primarily for internal or limited external distribution. The opinions and conclusions stated are those of the author and may or may not be those of the Laboratory. Work performed under the auspices of the U.S. Department of Energy by the Lawrence Livermore National Laboratory under Contract W-7405-Eng-48.

112 THE OPERATION OF HFTF TO TEST THE MIT 12 T COIL

The MIT coil was tested in HFTF to a peak field of very nearly 12 T. The accomplishment is note worthy as a demonstration of "reactor relevant" superconductors to high field in a realistic coil environment. Just the operation of the High Field Test Facility (HFTF) in providing the necessary test conditions is also no trivial feat. This note reports the work involved in testing the MIT 12T coil, with emphasis given to the test facility itself. Data specific to the MIT coil are relegated to appendices.

ASSEMBLY OF THE MIT COIL IN HFTF

For inclusion in the HFTF coil set for test, the MIT coil was placed in a steel cannister especially fabricated by LLNL for that purpose. Dimensions of the cannister are detailed in Fig. 1. The material was 316 stainless steel. Its primary purpose was to support the two NbaSn sub-coils of HFTF against the tremendous axial centering forces present during operation to full field. Figure 2 shows the MIT coil situated in its test position in HFTF. The outer cylindrical wall of the can was sized to support the total load with adequate safety margin.

The MIT coil is constituted of three double pancakes (denoted A,B, and C). The inter-pancake splices and the two current terminals projected radially outward from the coil, essentially on the midplane of HFTF. Space for these splices was provided between the NbTi sub-coils of HFTF by separating them with solid, 50-mm thick, steel blocks. Electrical insulation around the splices and terminals was provided by shimming with strips of NEMA G-10 of various appropriate thicknesses. The shims were also wedged in tightly enough to provide mechnical support against lateral deflection under Lorentz forces.

Vapor cooled current leads rated at 20 kA were purchased from American Magnetics, Inc. to deliver current to the MIT coil. Buswork between the bottoms of these leads and the coil terminals was constructed of 13x51 mm 2 copper bar'stock, traced with MF- Nb3Sn composite conductor (surplus from winding the HFTF NbaSn sub-coils). Joints between the MIT coil and the buswork ends were soft soldered using 60Sn-4OPb solder and bolted.

Helium at supercritical pressures was delivered to the MIT coil through a heat exchanger system provided by MIT. The system consisted of a counterflow heat exchanger located in the dead space above the HFTF coil set and a final heat exchanger located in the bore of the coil (Fig. 2). Cryogenic valves to

113 select alternate flow paths (cf. the schematic of Fig. 3) were also a part of the package provided by MIT. Control of the pressure and flow of the helium was accomplished in a system external to the 2 m cryostat consisting of a valve panel, a LN 2 temperature precoole'r, a VJ delivery/return line, a counter flow heat exchanger to warm the gas before flow measurement, a mass flowmeter, and temperature sensors and pressure gages at various points on the flow paths (cf. the schematic of Fig. 4).

COOLDOWN

Cooldown of the HFTF/MIT 12T ensemble was accomplished using the Airco refrigerator--at first using only the LN2 precooler for cooling power, then later starting the turbine when return gas from the cryostat had dropped sufficiently in temperature. Figure 5 displays the cooldown record of the October run and Figs. 6 and 7 show that of the January run. Both records emphasize that the most dramatic results were achieved in the first three days of cooling before the turbine was turned on. Without the turbine, the coil temperature was lowered to -100 K. The temperature was dropped an additional 50 K using the turbine; but great care, constant attention to the refrigerator controls, and three to four more days were required. In the January run, due to delays caused by the data acquisition computer, the coil temperature was allowed to warm back to about 100 K before transferring LHe. The amount of liquid required to complete cooldown and fill was not noticebly greater than in the October run where liquid transfer was begun with the coil temperature around 60-70 K. Several days of effort by a 3-5 man crew could have been saved by dispensing with the turbine for further cooldown and beginning immediately to transfer liquid when the coil temperature reached -100 K.

During the January run, liquid transfer was begun initially from a 4 kl dewar borrowed from MIT. The transfer from this dewar was relatively slow compared to transfer from our own 10 kl dewar. Using the 4 kl dewar in that run, five hours elapsed before liquid was measureable in the 2 m cryostat. In another two hours, the level reached 50 X on the lower probe (see Fig. 8 for LHe volume in the 2 m cryostat vs. indicated level). All of the initial 3 kl in the 4 kl dewar had been transfered in those seven hours.

The rate of dumping liquid from the 4 kl dewar was unnecessarily slow (-400 1-h-1). The reason was discovered when we switched over to transferring from the 10 kl dewar. The section of transfer line that remains in the 2 m cryostat for either connection was pushed against the bottom of the cryostat leaving only a narrow slit for liquid to escape. When this situation was corrected the dump rate of liquid from the storage

114 dewar was increased more than twofold to 980 1-h-1.

Some interesting insights to the cryogenic situation in HFTF can be gleaned from the cryostat fill data. For example, with the 980 1-h-1 dump rate, the accumulation rate in the 2 m cryostat was only about 600 1-h-1 even when liquid had covered the coil. The boiloff rate at high liquid level was later measured to be no greater than 180 1-h-1. This implies that losses in the transfer line from the 10 kl dewar consume at least 200 1-h-1. To put this in perspective, it should be realized that if we tried to maintain constant level in the 2 m cryostat by throttling the flow through this line, it would be necessary to use about 380 1-h-1 from the storage dewar just to maintain level in the 2 m cryostat. The transfer line losses should be reduced; if they cannot be reduced, only very rapid transfers should be considered.

The rate of liquid consumption is not the same at all levels. in the 2 m cryostat. As much quantitative information as possible was extracted from strip chart records of the LHe level during these runs. The essence of these records as they relate to variation in boil off rate is displayed in Figs. 9 and 10. There are elements of chronology that influece these records (for example, time since the previous fill, the proximity in time to an event that drastically changed the boil off rate, etc.) but on the whole they are relatively consistent. It can be said at least that when the level is very low on the coil (after having been full) the rate of liquid consumption drops to -40 1-h-1. When the cryostat is very full (in.the range needed for safe operation), the boiloff rate rises to as high as 180 1-h-1. The magnitude and variation of the rate of consumption can be explained in part (but not entirely) in terms of conduction down the inner wall of the 2 m cryostat, which is nearly 5-mm thick. I suspect the foam plug as a virtual heat leak that contributes a significant heat load for any practical test period. Careful thought should be given to redesign of the portion- of the cryostat above the coil, because a boiloff rate of 70-80 1-h- 1 , limited mainly by the vapor cooled lead losses, should be achieveable.

PRELIMINARY ELECTRICAL TESTS

In the October run, coil testing was delayed initially due to an error in electrically connecting the seven individual coils constituting HFTF and the MIT 12 T test coil so that all contributed in the same field direction. The error occurred because of a simple labeling mistake on the power supply leads to the six HFTF subcoils. The mistake was not discovered previously in tests of the HFTF coil set alone because it was, of course, not important until another coil was included. However, the

115 discovery of the error, and the seriousness of the consequences had it notbeen discovered, emphasize the importance of seemingly mundane and redundant checking before operating so complex a system.

Because of this delay, the first period of testing during the October run was confined to check-out of the electrical equipment (power supplies, breakers, etc.) and the cryogenic electrical joints. MIT coil joint resistances are reported in a memo from M.O. Hoenig attached as Appendix A. Pertinent comments to the Hoenig memo are included in a letter from myself attached as Appendix B. In addition to these measurements, we determined the total resistance of the buswork (including all joints) connecting the MIT coil to the 20 kA vapor cooled leads. The resistances, constant vs. current and time at all current levels up to the 15 kA used for measurement, were 0.13 micro-ohm for the negative bus and 0.083 micro-ohm for the positive bus. At full current (20 kA) a heat load of 85 W (117 1-h-1) would result. This should not be considered good performance, but since the load was not large compared to other loads (180 l-h-1 standby at safe operating level) and temporary (only present at full current), it was acceptable.

COIL TESTING

Actual testing of the MIT 12 T coil at full current and field was carried out during only three days of the October and January runs (October 30 and January 28 and 29). The October test was foreshortened by leaks in the MIT cryogenic package. The January run was halted by failures of two of the three induction heating pulse coils required for stability testing in the MIT coil. In spite of these obstacles, which prevented full completion of the proposed test plan, a great deal of useful information and experience was obtained--information that will be invaluable in designing and building future coils with conductors of this type.

RESULTS

Results of the October test are included in Appendices A and B as already mentioned. The results of the January tests are being prepared for presentation at the 1985 CEC/ICMC, Cambridge, MA. The following qualitative conclusions can be drawn from the results: (1) The cable-in-conduit conductor provides a potential means of obtaining high current densities at high fields. (2) The cable-in-conduit conductor provides a means of taking advantage of the high critical temperature of NbsSn in the removal of steady heat loads to the conductor. (3) The mechanical interaction of the conductor cable and the sheath in a cable-in- conduit conductor is a subtle phenomenon that must be eventually sorted out in order to take advantage of the sheath as a

116 structural component in a coil design. (4) The stability of the cable-in-conduit conductor against external perturbations in a realistic coil environment was demonstrated to be quite high.

PROBLEMS EXPERIENCED IN THE OPERATION OF THE 2 M CRYOSTAT

Too often, only those things that went right in an experiment find their way into a report, and posterity is left with a rosy picture that may be pleasant but unreal. Since tests of the type discussed in this report must be repeated many times in our laboratory in the near future, an account of the problems experienced in the execution of these tests will be very useful. Some of the operational difficulties have been mentioned already in the course of reporting the test results. The following is a more complete list of problems--the kind that would ordinarily be forgotten, at least until the next time they occurred. In most cases I try to suggest a possible prevention for future runs.

Loss of vacuum in cryostat acket

During cooldown for the October run, vacuum, in the 2 m cryostat jacket deteriorated badly (to about 10-3 torr). The primary cause was eventually determined to be the feedthru into the vacuum space that delivers LN2 to the intermediate temperature radiation shield. A soft solder joint to a copper tube constituting the feedthru cracked under thermally induced stresses. The crack was repaired using Wood's metal so as not to make the leak larger while trying to repair it (the system was left under vacuum during the repair). The repair was sufficient for completion of the first test run, but after the test was completed and the line was allowed to warm up, a nearby heater (located there to prevent the freezing of an 0-ring seal) melted the Wood's metal and recreated the vacuum leak. Another repair was made, this time with Sn-Pb solder, but it too failed during the January run. This time the line was warmed and the leak repaired with a filled epoxy.

The feedthru design is bad. It should be replaced with a design that utilizes only welded or brazed joints and careful attention should be paid to the relief of thermally induced stresses.

Leak in the LN2 baffle of the cryostat vacuum system diffusion pump

In the search for the leak just mentioned, a less severe leak was found in the pumping system on the 2 m cryostat vacuum jacket. The LN2 baffle on the diffusion pump was identified as the source of the problem. The LN2 reservoir in the baffle had cracked loose from its mechanical supports, either due to

117 vibration or rough treatment during storage or installation, leaving a. small leak. The baffle was replaced for these runs with a straight spool piece. The baffle leak was apparently present before the more serious, feedthru leak occurred. For when the system was put back on line after both repairs, the vacuum was an order of magnitude better than before the vacuum failure occurred. However, this "fix" should only be considered temporary and a proper baffle should be replaced in the system.

Failure of shaft seals and belts on cryostat vacuum system fore

As is often the case, when a system failure occurs there are more than one or two contributing factors, all masquerading as the major problem. In the search for the problem, it is also unusual to discover all these factors the first time through. The above mentioned vacuum failure is a prime example. In addition to the two leaks already described, it also happened that the foreline vacuum pump failed. Apparently a shaft bearing failed causing a shaft seal to leak and reduce the performance of the pump. The problem was discovered only because the wobbling shaft also caused the drive belts to wear and become loose. All future runs in HFTF should be preceded by a thorough inspection of the various components of the vacuum system. This system. is typically run continuously whether the cryostat is in use or not, and components can be on the verge of failure without the operators' knowledge.

Remote JT valve operator malfunction

The Airco refrigerator has remote JT valves for both the 10 kl dewar and for the 2 m cryostat. These have quite different valve operators attached. The positioner on the 10 kl dewar is air operated and that on the 2 m cryostat delivery tube is operated by an electric motor. The motor turns a nut that engages a screw afixed to a shaft. The shaft in turn transmits linear motion to the JT valve needle. To do this, the screw must be constrained from rotating. In the past, this constraint had been provided by means of a projection that also indicated valve position by moving a potentiometer. However, this projection had been removed when the potentiometer had failed in some manner. Since its removal, only the combination of a freely moving nut and a relatively tight shaft seal had provided some measure of rotary/longitudinal motion conversion. The position of the JT valve, recorded as turns of the drive nut, must have been a very imprecisely known quantity under such conditions.

I recommend that the motor driven valve positioner be replaced with an air operated positioner like the one on the 10 kl dewar and that the controls for the two positioners be

118 consolidated.

Malfunction of LN2 autofill system for Airco precooler/absorber

The autofill system for the precooler/absorber in the Airco refrigerator cold box failed to operate during the October run. We were able to continue by manually filling or stopping when external signjs (absence or presence of frost on particular pipes) indicated that the heat exchanger reservoir was either empty or overflowing. However, this procedure resulted in extremely excessive LN2 consumption, and probably suboptimal cooling at times also. The problem has since been attributed to faulty contacts in a cable connector associated with the level sensors and corrected. The autofill system was operational for the January run.

Freeze-ups of the Airco refrigerator

The 2 m cryostat is an inherently "dirty" system that cannot be completely "cleaned up" no matter how great the effort (the foam plug above the coil stack must contribute greatly to this problem). Thus some "freeze-ups" of the refrigerator cold box, when operating directly into the 2 m cryostat, should be expected. During the January run, however, three turbine failures caused by icing occurred within a period of four days. All the failures occurred after coil temperature was below 100 K, two occurred when the coil temperature was below 60 K. The typical impurities causing freeze-up (air and water) should all have been frozen out in the 2 m cryostat at these temperatures. Some other evidence indicates that the purifiers and not a dirty 2 m cryostat may have been the main problem by allowing impurities (mostly Na and 02) through. Details are sketchy at this time but the performance of these purifiers bears careful investigation before future runs. Frequent freeze-ups may also be avoided by using the Airco only for precooling to around 100 K as suggested earlier, without starting the turbine at all.

Deficiency and failure of He gas recovery compressors

Several times during the October run, and once during the January run, the helium gas recovery bags had to be vented to atmosphere because they overfilled. The recovery compressors had been incapable of keeping up with the boiloff from the 2 m cryostat. In the October run only one compressor was available, the 125 hp Worthington, and it failed during the run. For the January run, the 75 hp Worthington was operational. Both compressors were run continuously to maintain adequate recovery speed. The two together are sufficient, but care must constantly be taken to keep one bag nearly empty, especially during initial filling of the 2 m cryostat when huge quantities of gas are being

119 generated. It is essential that both be operable before a run. Obviously,- reducing the cryostat losses would also ease the burden of reliability and of operating at peak performance from these compressors.

Difficulty in rapidly dumping He xas when recovery bacs fill

For the October run, the recovery bags could only be dumped manually in the event of and overfill. This resulted in a dangerous situation. In the present configuration, when the recovery bags fill completely, an inlet valve shuts and all gas flow stops, including flow out of the vapor cooled leads, which could easily result in lead burn-out and magnet failure. Before the Jan.uary run, emergency, electrical operators were installed on the dump valves near the lead flow-meters to allow the operator to act quickly to dump the bags instead of having to run outside to the vent valves in the event of an overfill. In the future, this operation should also be alarmed and automated.

"Malfunction" of JT valve positioner on 10 kl dewar

The pneumatic positioner on the JT valve for the 10 kl dewar is an ancient unit, but still sophisticated and stone- reliable. However, two associated components, a human operator and an instrument-air supply hose, caused minor problems during the October run. Once, the mechanical lever that closes the feedback loop in the pneumatic positioner circuitry was tripped (literally), and once a leak was found in an air supply hose. In both cases the valve and positioner were wrongfully accused of giving deficient service. My advice is, "If the 10 kl JT valve is not working properly for some reason, look elsewhere first."

Loss of vacuum in VJ return from 2 m cryostat to Airco

The vacuum jacketed return line from the 2 m cryostat to the Airco cold box frosted during the October run. No leak was found. The line was pumped overnight with a mechanical pump and put back into service. It would be wise to routinely pump all VJ lines that carry LHe before any extended run.

Unexpectedly high boil off rate in 2 m cryostat

The unduly high consumption of LHe during both the October and January runs has been mentioned already several times. However, I feel the attention is fully warranted. Cryogenic inefficiency is currently the single most serious problem with HFTF because it presents operational difficulties in so many different ways: the losses far exceed our current refrigeration capacity or reasonable expectations of future refrigeration capacity; they deplete our present 14 kl storage capacity in

120 intolerab-ly short periods; the rapid evolution of gas qvertaxes our recovery capabilities (possibly also our capablities to maintain gas purity); and since none of the cooling power of this excess boiloff is being utilized, handling of the extremely cold effluent causes problems with maintaining leak tightness of the entire cryostat and recovery system.

The list could go on, but it is already long enough to convince the reasonable person that solutions must be found if HFTF is to be a truly useful tool. I suspect two possible sources of the heat: conduction down the inner wall of the 2 m cryostat has already been alluded to, but the virtual leak of heat -slowly being conducted out of the huge foam plug above the coil stack must also be considered. My personal opinion is that the foam plug offers no advantages over a system of baffle plates, and several disadvantages. There is some support in the literature for my opinions, but they are mostly couched in personal experience and the dislike of such a huge source of contamination in a cryogenic system. Metal baffle plates result in a much cleaner system and offer the opportunity of actively controlled cooling by having them traced with a tube, through which a metered flow of cold gas can be extracted (the performance of the plates as a radiation baffle is improved by having a portion of the heat they intercept transmitted to exitting gas rather than liquid in the bath). The boiloff is thereby reduced and the gas that is evolved is warmed in the process to a temperature that makes it easier to handle.

The baffle plates could also be put in good thermal contact with the upper portion of the vessel inner wall to allow much of that conduction heat load to be intercepted also. The opportunity exists to reduce the wall conduction heat load by a factor of 1/35, making it negligible in comparison to the vapor cooled lead losses. The plates could be made economically of aluminum with a tracer tube welded on. Good contact to the vessel wall could be obtained by sectioning the plates to allow them to be cammed out against the cryostat wall after the magnet is lowered in place.

Consideration should also be given to opening the vacuum jacket of the 2 m cryostat and modifying the LN2 tracer tubing on the shield by providing a.section that encircles the inner wall and makes good thermal contact to it at the appropriate distance from the top.

Auxillary fill line impropery positioned in 2 m cryostat

An auxillary fill line was used in the October run to allow LHe transfer initially from several 500 1 dewars. The thought was to precool the cryostat with the 500 1 dewars and save all

121 the liquid in the 10 kl dewar for a single uninterupted run. The auxilaary -fill line did not, however, extend deeply enough into the cryostat to guide the flow underneath the coil stack where it was needed for efficient cooling. In fact it was later learned that it did not even extend below the foam plug. In spite of this, some cooling was obtained from the five or so small dewars used, but nothing like their potential cooling capacity was obtained. The importance of knowing the routing of all the various lines penetrating the top plate before lowering the magnet system into the cryostat cannot be over emphasized.

Gas leaks in top of 2 m cryostat

In the January run alone, the gas equivalent of 6000 1 of liquid was lost. Much of this loss was due to leaks in the top plate of the 2 m cryostat (predominantly around the vapor cooled leads and in the VCL boxes). The leaks were tolerable before the run, but got much worse during it because of the excessive flow of cold gas associated with the unexpectedly high boiloff rate. Nevertheless, the sources of these leaks must be eliminated by redesign of the lead and lead-box seals before future runs. A full blown leak test of the low pressure portion of the facility (2 m cryostat and all recovery lines) is also in order. The test could be accomplished while the system is warm by carrying out a pressure .ecay test extending over several days. Reduction of the liquid consumption in the cryostat will ensure the reliabilty of whatever seal modifications are made by eliminating undue thermal stresses on them.

CONCLUSIONS

The MIT/LLNL 12 T test was extremely useful. The HFTF coil set itself operated faultlessly, unperturbed by the variations of field, temperature, etc. necessary to provide the test environment for the MIT coil inside it. It will be an extremely useful tool for future such testing, unduplicated in many ways, anywhere else in the world. The operations crew also gained experience that will be invaluable in future testing. In general, the entire facility, including the cryogenic system, is excellent; but there are many details that can be, and for efficient operation must be, corrected.

122 FIGURE CAPTIONS

Figure 1. Cannister for containing the MIT 12 T coil inside HFTF.

Figure 2. Elevation view showing the MIT 12 T coil and the cryogenic package in HFTF.

Figure 3. Schematic of the flow circuit of the MIT 12 T coil and the cryogenic package inside the 2 m cryostat.

Figure 4. Schematic of the flow measurement/control panel located outside the 2 m cryostat.

Figure 5. Record of the cooldown of the HFTF/MIT coil assembly for the October run.

Figure 6. Record of the cooldown for the January run ( first six days).

Figure 7. Record of the cooldown for the January run (continuation).

Figure 8. Estimated volume of LHe in the 2 m cryostat vs. indicated level.

Figure 9. Measured standby boil-off rate vs. level in the upper range. The cryostat was previously full.

Figure 10. Measured standby boil-off rate vs. level in the lower range. The cryostat was previously full.

123 Upper cover plate

Inner compression ring

Test coil assembly

f f Outer compression ring

-Lower cover plate

Exploded assembly of test coil and coil case.

124 jf.oo 3 _____

*400W -I

11 I. I J~l ~ 4 'r~ ~1~

L3L 1A. ~

-- f" d 7- ~IIi{ I I-

I, £500 i

I I.

2.37 L77 1

§ 171

1 1 , .a fl"7 "07 58.7~

50.50 I lx -X

U , /4 /-7-71 7 7- I -l 77/ -707 71// / / Z~ X/ /I/ Z 00A Z / X / x777 / 7?1 /

iLy~teZ. 125 I g!

.b. 9"C

HZ r L n -J T 0 1ri U H I-' U) - u LiJ H

H c'J rvj

U C.' In N I-. -I' r ~1' r -II r yr r ,~i r 1 r 'I

'.4 *-6 S.' hi U a * ~ N

* a a a a a

SNOL tfNIu L 11mo lu SWIM±~ L J f's' .I

126 474~

Z145

1,41-es 1--rw V7W : vr//W2l t'r/3 xllr r Aill

i wr 11.1vsro sri' S7 Sr6 4 e '4~l

127 October cooldown--MIT 12 T Test 300

250

200

150 E

0 100

0 0 1 23 4 56 7 day 0 vacuum leak X start turbine + temporary shutdown

128 January Cooldown--MIT 12 T Test 300-

250

200

d- 150- E

100-

50-

o 0 1 2 3 4 5 day

129 January Cooldown--MIT 12 T Test 120

100-

80 -

60- E

40-

20 -

0 I - ____ 5 6 7 89 10 11 day 0 turbine failure X restart turbine

130 2 m cryostat liquid fill vs. level 3000

2500

2000

75 1500

D .2

1000

500

0 U ZD 50 75 100 125 150 175 200 liquid level (z of probe height) It' Standby boiloff va level (upper) 180

150

120

130 60 0 0 0

0 0 20 40 60 80 IC10 level (z upper probe) 0 after coil dump 1/28/85 A before test 1/28/85 0 during run 10/25/84

132 Standby boiloff vs level 100- (lower)

80.1 I I I

60 0 -c N 9TM 0 o N 0 0 0 -v 40- 10 03 n0E ]

0. 0 20 40 60 80 100 level (z of lower pr obe)

I /Le

133