1 FTP/P7-23

SST-1 Magnet system progress towards device assembly

S. Pradhan1, A. N. Sharma1, U. Prasad1, K. Doshi1, Y. Khristi1, P. Varmora1, D. Patel1, S. J. Jadeja1, P. Gupta1

1Institute for Plasma Research, Bhat, Gandhinagar, India

E-mail contact of main author: ansharma@ipr.res.in

Abstract. Steady State Superconducting (SST-1) Magnet system comprises of sixteen superconducting Toroidal Field (TF) Magnets, nine superconducting Poloidal Field (PF) Magnets, resistive central solenoid (CS) and compensating magnets, resistive Vertical Field (VF) equilibrium magnets, resistive in-vessel feedback coils and radial control coils apart from the support structures against the in-plane and out-of-plane forces. The magnet system could only be partially commissioned in 2006 as a result of distributed leaks appearing at cold in the helium circuit including the dissimilar metal joints at the superconducting magnet winding packs terminations and joints. Since then, under the SST-1 Mission mandate, all the sixteen TF magnets have been completely refurbished with modified leak-tight joints, enhanced insulation resistances in cold between the winding pack and the ground potential surfaces. Each of the TF magnets have been equipped with a supercritical helium cooled bubble type shield on the bore of the TF magnets facing the plasma with an intention to prevent disruption induced eddy current heating of the edge double pancakes. Each of these modifications have been validated at several stages prior to their implementations on the magnet winding packs following a strict QA/QC and process control mechanisms. As a final verification, each of the TF magnets have been tested in their self field in representative supercritical helium flow conditions (4 bar, 4.5 K, 16 g/s at the inlet) with full transport currents of 10 kA (2.2 T self field). In these extensive test campaigns, the modifications carried out on each of the TF magnets including the integrated performance of the potential breakers have been validated. The normal zone propagation characteristics of the winding packs, various off-normal scenarios leading to the quench of the magnets, electronics leading to the accurate fail-proof detection of the normal zone and the magnet protection aspects have also been experimentally established. Following the successful test validations, the TF magnets have been fully assembled together with the vacuum vessel modules and intervening thermal shields on the SST-1 machine shell. All superconducting PF magnets have also been suitably modified and have been assembled. Refurbishment of a damaged divertor coil incorporating bridge type joints and modifications of each of the joints in the outer PF magnets are some of the important experimental and process modifications carried out on the PF magnet systems. The central solenoid has also been tested for representative operational scenarios of SST-1 successfully. A pair of single turn radial control coils meant for controlling the null during the plasma break-down as well as for providing radial control to the plasma has also been designed and would be installed shortly inside the SST-1 vacuum vessel. These coils employ high temperature insulation and will be inertia cooled.

The integrated testing and engineering commissioning of the superconducting magnet system of SST-1 and testing of the resistive magnet systems are envisaged during the `engineering validation phase’ of the SST-1 towards the middle of 2012.

1. Introduction

The steady state superconducting Tokamak, SST-1, at Institute for Plasma Research (IPR) has Superconducting Magnet System (SCMS) for the magnetic confinement and plasma shaping. SCMS comprises of sixteen superconducting (SC), D-shaped, Toroidal Field (TF) coils and nine superconducting Poloidal Field (PF) coils. Central solenoid, compensating magnets vertical field equilibrium coils, radial control coils and in vessel feedback coils are copper coils [1], [2]. SST-1 commissioning was attempted in 2006 with the TF magnets being cooled down to 4.5 K at the inlet and 6 to 8.8 K at the outlet. The helium flow was in two phase flow regime with helium inlet and outlet pressure at 1.8 bar and 1.3 bar respectively. Overall cryostat pressure was 10-4 mbar. In these service conditions, SST-1 TF magnets could be charged up only to 1000 A as against the nominal value of 10000 A. From these attempts, it

2 FTP/P7-23

was evident that understanding and rectifying the problems was essential for successful commissioning of SST-1 magnet system.

Main problem in the magnet system was of helium leak in the joints and isolators which restricted the operation to two phase flow as against the nominal supercritical operation at 4.5 K and 4 bar. The leaks in the joints were present in the (a) welding of the SS 304 L conduit with that of the ETP copper terminations on its square section interfaces (b) brazing of the SS 304 L helium exit tubes on to the termination (c) the end copper plates and the copper solder tubes used for solder filling. Attempts with in-situ repair of the leaks with cold curing epoxies or with local repairs were not found to be feasible. Also the insulation resistance of magnet winding pack with respect to ground was found to be low. After a thorough review of the issues and problems associated with SST-1, a decision of dismantling SST-1 was taken. It was decided to make all required design level changes to solve these problems. It was also decided to cold test all TF magnets at its nominal current to ensure its performance as per its design values.

Figure 1 : New Joint Box design

2. Design Changes in TF Magnet system:

New joints, new inlet and outlet manifolds were installed on all the 16 TF magnets. Case cooling panels cooled with 5K helium were installed on all TF magnets. Insulation of exposed leads and ground insulation at few locations was also strengthened.

2.1 New Joint Design:

Almost all leaks in magnet system were in SS-Cu dissimilar metal joint location. In order to avoid these, a box type joint was designed; prototype tested and finally implemented on SST-1 TF magnets [3]. These joints are shaking hand type lap joints. After installing end caps and copper block over the un-conduited CICC, copper block was filled by silver solder 95Sn-5Ag to increase contact area between the cable and inside surface of the copper blocks. Solder

3 FTP/P7-23

filling quality was ensured by ultrasonic method. After proper cleaning, two copper blocks were soldered with each other using 60Pb-40Sn solder to make a lap joint. Overlap length of joints is 300 mm, equal to the last twist pitch length of the cable. This ensures contact of all strands with copper septum and minimum current transfer length between two joining copper blocks to obtain low joint resistance.

To verify quality of the joint, its room temperature resistance was measured. Iteratively after various room temperature and subsequent low temperature test, a figure of merit had been obtained for this measurement. It was found when the room temperature joint resistance is ~ 6 to 8 µΩ, it gives sub-nano ohm joint resistance at low temperature. So after ensuring room temperature joint resistance, it is covered by SS pipe of length 310 mm. This pipe is made in two parts, to allow its insertion after lap joint is made. This pipe is welded with end caps and end caps are welded to CICC. This design avoided earlier problem of Cu to SS welding. Cross section of this joint and actual TF coil joints are shown in figure 1a and 1b respectively.

2.2 Improvement of Insulation resistance of Magnet winding pack

SST-1 TF magnet CICC conductor has been insulated with two half lap layers of 0.09 mm thick dry glass tape treated with accelerator solution to react with Bisphenol A impregnating Resin + Hardener. Interlayer insulation of chopped glass mat & glass cloth composite with accelerator treatment has been placed between each DP layer as well as between two layers of DP itself. Ground insulation is approx. 3mm thick provided with glass chopped mat + glass cloth composite & glass tape and finally VPI impregnated. The conductor insulation is electrically & mechanically quite sufficient for withstanding normal and fault condition related surge voltages. However some locations have been identified in the coil that are prone to insulation damages due to coil handling and assembly related activities. These locations were coil outboard side middle section (figure 2a), leads coming out from casing top and bottom side (figure 2b), and Helium inlet stub (figure 2c). Insulations of all these locations have been suitably improved. Voltage taps installation method was also changed from earlier bolted concept to brazed voltage taps. This new design of voltage taps without sharp projections and new joint with circular shape are expected to give better performance in Paschen like conditions.

2.3 New manifold design:

In 2006 campaign leaks have been found in GFRP isolators at coil inlet and outlet manifolds. These leaks were in some cases due to defects in the material used for isolator and in some cases due to higher stresses. So all isolator were tested at Liquid nitrogen temperature prior to installation in the manifold. Welding of isolators with manifold tubes was done in steps and with cooling arrangements to limit the temperature at the GFRP and SS tube interface locations. Bonding at this location is prone to damage at temperature > 120 ° C. Also bending radius of the tubes was increased in the manifold and S-bends were provided in long length tubes. New manifold design is also shown in figure 2d.

4 FTP/P7-23

2 a

2 b 2 d

2 c

2 e

Figure 2: Insulation improvement, new manifold and case cooling panel of TF magnet

2.4 TF coil case cooling:

In 2006 campaign, SST-1 TF coil cases were left un-cooled due to leaks in the cooling tubes soldered on the coil casing. Upon dismantling of SST-1, it was realized that the un-cooled TF cases should be cooled to give additional temperature margin to superconducting cable. In events like plasma current disruptions and vertical displacement events, coil case temperature rises due to eddy current heating. In SST-1 TF coil this temperature rise was calculated to be less than 5 K.

All TF coils were equipped with bubble type 5K helium cooled panels. This additional cooling is expected to increase the temperature margin of edge-most pancakes inside the TF winding packs as these pancakes have highest contact with the coil casing. This in turn will increase the reliability and availability of magnets during plasma operation. These panels are shown in figure 2e.

2.5 Improved QA/QC during magnet preparation:

A holistic approach was adopted in Quality assurance (QA) and Quality Control (QC) of processes and components used in the preparation of magnet joints, manifolds, insulation and coil handling during test and machine assembly. Pre-fabrication inspection comprised of visual inspection, dimensional measurements of copper blocks and jacket end caps and Destructive Tests (DT) of all materials used for magnet components like joints boxes, manifolds, sleeves and tubes etc were done. Fabrication inspection involved visual inspection, argon gas flow parameters control during welding and temperature monitoring during welding. Post fabrication inspection comprised of helium leak tests of joints and manifolds at 12.5 bar pressure. Mass flow measurements were performed at room temperature in all magnets to ensure absence of flow reduction or choking due to solders, resin or dust particle.

5 FTP/P7-23

2.6 Design changes in associated sub systems

After analyzing the performance of quench detection system and signal conditioning cards in 2006 these systems have been modified to improve their performance. Major changes in Quench detection system include modifying voltage tap installation methods, change in quench logics and quench detection card level changes. New voltage taps give brazed and soldered contact as compared to brazed and bolted contact. This has drastically improved the signal level specially in the coil charging and discharging phase. Quench logic of central average has been modified to simple average. Redundancy has been improved by increasing the comparison of all magnet sections. Each section of magnet is compared at four channels using two redundant cards. Each of these cards uses separate voltage taps to increase redundancy. Quench card level changes include increase in the time and voltage threshold stability, introduction of isolation between analog and digital section of cards. CPLD based cards are designed for ORing different quench signals [4].

Figure 3. Block diagrams of quench detection and signal conditioning cards

Improved versions of signal conditioning cards have been developed for all sensors [5]. Signal conditioning cards developed for Temperature sensors and hall probes are identical, except for current source section, which is piggy back type interchangeable card. Other special features

Input RFIFilter

3 port

Isolation

Amplifier

+ C -

Filter

and

Driver

+ C

Output

Signal to

DAQ

Precision SensorExcitation

Input Raw

DC Supply

External

AmplifierG = 100

O/P Linear DC

Supply Inside

Chassis

O/P OffsetAdjustment

+

-

Digital

V1

PrecisionVoltageReference

DriverV2

+

V2-V3

V1-V2-

DAQDAQ

VoltageTapProtection

Relay

+

NO/NC

-

Amplifier

V3

Isolation

-

AnalogFilter

+DP-2

DifferenceAmplifier

Delay

Quench

Signal

- 1

Analog

DigitalIsolation

Window

DP-6

+

Six Channel Card - 2

Time

DP-1

Drivers

Redundant

&

Quench

Signal

- 2

-

FinalQuenchSignal

comparator

TFMagnet

Six Channel Card -1

Latch

Quench detection card

Signal conditioning card

6 FTP/P7-23

of these cards are 2 channels per card configuration, careful PCB layout for small signal handling, output anti-aliasing filter and driver, offset Adjustment circuitry, separate linear output power supply for each channel to eliminate driving problems, modular piggy back type current source for sensor excitation, power entry module with EMI filter etc. Cards developed for displacement sensor and strain gages have similar architecture, with difference in the sensor excitation section. Block diagrams for quench detection system and signal conditioning cards is shown in figure 3.

3. Design validation tests on TF Coils

After validating design changes in laboratory experiments, all TF coils have been refurbished and cold tested [6]. Each TF magnet was tested in its self field in representative supercritical helium flow conditions (4 bar, 4.5 K, 16 g/s at the inlet) with full transport currents of 10 kA (2.2 T self field). The major objectives of TF coil were demonstration of acceptable joint resistances, helium leak tightness at actual operating temperature, pressure and current and acceptable ground insulation resistance. During these tests integrated behaviour of all design changes have been verified.

Helium partial pressure in the cryostat during the entire campaigns of coil cool down and charging experiment was better than 10-8 mbar. This validated the leak tightness of all joints, inlet and outlet manifolds, 5 K panel and inlet/outlet connections and TF coil winding packs. Inter pancake joint resistance for all TF coils were measured individually, at 10 kA, using nano-volt meter and in-house developed precision signal conditioning cards. The measured joint resistance values are less than 1 nΩ. In different coil test campaigns, about 25 quench events have occurred. Quenches were mostly due to in-sufficient cooling in Busbar-current lead section, or in edge pancakes due to short coil support leading to hot coil casing. Additional heat load from short coil hanging supports was unavoidable due to test chamber geometry, unless cool down period was extended.

Quench detection and protection system worked well in all these events. Not a single quench event was missed by this system. Performance of signal conditioning cards have also been validated in more than 30 coil test and other test campaigns. Typical results from quench detection system, hall probes, mass flow and pressure measurements are shown in figure 4.

4. PF Magnet and other Resistive magnet refurbishment:

PF4 and PF5 top and bottom coils have 3 joints each. All joints were refabricated with TF like box type leak tight joints. During the 2006 commissioning and charging experiments PF3 top coil was damaged by electrical breakdown between its two turns. It was repaired by removing the damaged CICC and replacing it by superconducting shunt of SST-1 CICC with bridge type joints. After repair it was cold tested in modified TF test stand. It was charged up to 6 kA to ensure its operational feasibility during actual SST-1 operation phase.

During the ground level pressure drop testing of PF2 T coil it was found it has severe blockage. It was realized that some resin had entered through busbar end region during coil VPI. Blockage was found to be in about 13 m CICC (busbar + 2.5 outer turns). This length of CICC was removed and two turns were further opened and profiled like busbar. Blank turns were filled using Hylem cutouts and insulation of entire magnet was reinforced with GRFP tapes and resin. After RT insulation test and pressure drop clearance the coil was released for assembly. PF3 coil repair and test and PF2 coil repair are shown in figure 5. All old helium

7 FTP/P7-23

isolators and hydraulic connections of the PF coils were replaced with new isolators cold tested at liquid helium temperature.

672.4 672.8 673.2 673.6 674.0 674.4

2000

4000

6000

8000 Current DP1 DP2 DP3 DP4 DP5 DP6

time (s)

Co

il C

urr

ent

(A)

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

DP

VOltag

e (V

)

QuenchTrigger

22/09/201 0 08:39:34.277203 PM 22/0 9/2010 08:59:34 .517203 PM

0

2

time

Mag

net

ic fi

eld

(T)

0

200 0

400 0

600 0

800 0

100 00

120 00

Curre

nt (A)

Figure 4: Different results from TF Coil tests

The Ohmic transformer system of SST-1 comprises of an air core central solenoid (TR1) and three pairs of compensating coils (TR2, TR3 and TR4). It will be used for plasma break down, start up and initial current ramp up before auxiliary heating and current drive system take over. These coils are water cooled copper coils made with OFHC copper conductor of 19 mm x 22.7 mm with a central hole of 8 mm. Electrical tests have been done on the Central solenoid to ensure its safe operation till 20kV to ensure its suitability for plasma breakdown operation. A pair of vertical field (VF) coil is also installed to keep the plasma in equilibrium during the initial phase.

Radial control coils & Active feedback control coils are a pair of copper coils placed inside the vacuum vessel. These coils are used to control plasma movements in radial direction and in vertical direction. Radial control coil is a pair of single turn coils meant for controlling the null during the plasma break-down as well as for providing radial control to the plasma. It has been designed and would be installed shortly inside the SST-1 vacuum vessel. These coils will employ high temperature insulation and will be inertia cooled.

8 FTP/P7-23

PF3 Coil

repair and

test

PF2 Coil

repair

Figure 5. PF3 Top coil repair and cold test and PF2 Top coil repair

5. Conclusion:

Leak tight sub nano-ohm joints, new cold tested isolators, manifolds and helium cooled case cooling panels have been implemented on TF and PF magnets,. Ground Insulation of the coils was reinforced and tested at room temperature and cold conditions. Cold testing of coils with nominal operating currents have confirmed suitability of these coils for actual SST-1 operation phase. Associated subsystems like quench detection system, signal conditioning and data acquisition systems have also been validated during different coil test campaign. Ohmic transformer coils has also been tested for 20 kV operational requirements. SST-1 TF, PF and Resistive magnets have been fully refurbished and assembled on SST-1 machine shell. Presently SST-1 is under cool down and engineering validation phase.

6. References

[1] Pradhan S. & SST-1 Mission Team, “Status of SST-1 Refurbishment,” Journal of Fusion Research Series, vol. 9, 2010, pp. 650-657

[2] S. PRADHAN et. al. “SST-1 Status and Plans” IEEE Transactions on Plasma Science, 40, 614-621, 2012

[3] U. Prasad, A. N. Sharma et al “Sub Nano-Ohm Joints in SST-1 TF Magnet Winding Packs” 22nd International Conference on Magnet Technology (MT-22), Marseille, France, 12- 16, September 2011

[4] Yohan Khristi, A.N.Sharma et. al. “Quench Detection System for TF Coil-test Campaigns of SST-1” IEEE Transactions on Applied superconductivity, Volume 22, Issue 2, April 2012

[5] K. DOSHI et. Al. “Precision Signal Conditioning Electronics for Cryogenic Temperature and Magnetic Field Measurements in SST-1”, IEEE Transactions on Plasma Science, 40, 641-645, 2012

[6] S. PRADHAN and SST-1 Mission team “SST-1 Toroidal field magnet tests: some results and and lessons learnt” IEEE Transactions on Applied superconductivity, vol 22, No. 3 June 2012, 9501804