Winding Pack Height Management during Fabrication of the ITER CS Module
Explore the source record for details and available documents.
Engineering topics
Publications and source records attributed to Martovetsky, N N.
Explore the source record for details and available documents.
Abstract not provided
Abstract not provided
Abstract not provided
Abstract not provided
Abstract not provided
Since the introduction of the cable-in-conduit conductor (CICC) concept, a variety of alloys have been proposed for fabricating the conduit in high field magnets. The conduit provides containment of the liquid helium coolant and is typically also the primary structural component for the magnet coils. These functions create requirements for strength, toughness, fatigue crack resistance, and fabricability. When the CICC uses Nb{sub 3}Sn superconductor, the conduit alloy must retain good mechanical properties after exposure to the superconductor's reaction heat treatment. Here we present data from cryogenic tensile, fracture toughness, fatigue crack growth rate, and axial fatigue tests for a modified heat of JK2LB, before and after the exposure to the reaction heat treatment. The alloy is presently being considered as a candidate for use in ITER Central Solenoid (CS) Coils. The direct comparison of the data from the comprehensive test program with earlier versions of JK2LB and another CICC candidate alloy (modified 316LN) is intended to assist design engineers with material selection for CICC applications.
An insulation system for ITER Central Solenoid must have sufficiently high electrical and structural strength. Design efforts to bring stresses in the turn and layer insulation within allowables failed. It turned out to be impossible to eliminate high local tensile stresses in the winding pack. When high local stresses can not be designed out, the qualification procedure requires verification of the acceptable structural and electrical strength by testing. We built two 4 x 4 arrays of the conductor jacket with two options of the CS insulation and subjected the arrays to 1.2 million compressive cycles at 60 MPa and at 76 K. Such conditions simulated stresses in the CS insulation. We performed voltage withstand tests and after end of cycling we measured the breakdown voltages between in the arrays. After that we dissectioned the arrays and studied micro cracks in the insulation. We report details of the specimens preparation, test procedures and test results.
The US Domestic Agency is one of six parties supplying TF cable-in-conduit conductors (CICCs) for ITER. Previous tests have shown that measured performance of the TF CICCs can be much lower than expected from the strand properties at the projected uniaxial strain and that the cabling pattern may also be an important factor. Worst of all, voltage signals well below the expected critical surface could not be reliably interpreted or canceled, making test results very suspect. The TFUS1 sample was prepared to achieve multiple goals: (1) to ensure uniform current distribution and to eliminate parasitic voltage signals by improving joints, (2) to explore the potential benefits of a different cabling pattern for better support of strain-sensitive strands, and (3) to explore the source of voltage development in the cable through the use of innovative penetrating diagnostics. Test results of the first US-made samples are presented and discussed.
ITER TF coils will see a significant longitudinal magnetic field in the event of the plasma disruption. This abrupt change of magnetic fields results in the appearance of an additional electrical field in the strands. The mechanism of this electrical field is the induced currents that expel the flux from the strands. This effect was known since the late 1970's [1-3] and most of the details necessary for the analyses given in this report are presented in [4]. Let's assume for simplicity a zero transport current in the strand. When a longitudinal pulsed field is applied, the outer filaments will carry an induced current repelling the change of flux. The current density of this current is 'critical' in the simplification of Bean's critical state model, where superconducting transition is represented as j=j{sub c} at any non-zero electrical field and zero where the electrical field has not penetrated. In reality, since the current density is roughly logarithmic with the electrical field, E=E{sub c}*exp[(j-j{sub c})/jo], Bean's model is just a simplification, and current density is slightly nonuniform in the outer filament and more so for the interior strands. The inner portion of the filaments will carry a current of the opposite sign. Even in the Bean's model it is not uniform, but the assumption that it is uniform and less than critical simplifies mathematics significantly and does not deviate far from the real current density distribution. In certain circumstances, the average electrical field in the strands will be high enough to exceed the take-off electrical field averaged across the cross section. In this case, the multifilamentary strand will become unstable and will experience transition to the normal state. With zero transport current, it will eventually recover, of course. This phenomenon is analogous to the flux jump. If the strand carries a transport current, the situation becomes more complicated. If it goes unstable and the transport current is higher than the cryostability limit (by Stekly), or if there are enough losses to bring the temperature above the current sharing temperature taking into account limited heat capacity of the CICC, the strand will not recover, and the CICC will go normal. Conservatively, we will consider that if we find an instantaneous unstable situation, it is not acceptable. In presence of a transport current, the situation is sensitive to the direction of the strand twist, direction of the pulsed field and direction of the transport current. Recently, ITER decided to increase the twist pitch of the TF strands from 15 mm to 30 mm to improve the stability of the strands against the longitudinal field. In this report we will quantify the effects of this proposed change and perform a trade off study. The issue is that by increasing the twist pitch of the strands we not only increase the coupling losses in the transverse magnetic field, as expected in classical multifilamentary composite superconductors, but also increase the hysteresis losses in the strands with internal tin. In classical multifilamentary superconductors, twist pitch change should not cause an increase of the hysteresis losses in the transverse field. However the high Nb3Sn content internal tin strands develop transverse links, which couple the filaments into clusters. These links turn out to contribute a significant fraction to hysteresis losses [5]. If we project the results of [5] onto the ITER proposal to increase the twist pitch from 15 to 30 mm, we should expect the hysteresis losses to increase by a factor of two, which will likely disqualify strands with 30 mm twist pitch. This very strand twisted to 15 mm twist pitch would likely pass the ITER criteria. So, increasing the twist pitch has a very negative consequence and we need to make sure that it is absolutely necessary. Recently, A. Vostner (private communication) reported preliminary results on the losses in candidate TF strands. In agreement with what was reported in [5]; he found that TF strands with 15 mm twist pitch have hysteresis losses about half of what the strands with 30 mm twist pitch have.
ITER Central Solenoid uses butt joints for connecting the pancakes in the CS module. The principles of the butt joining of the CICC were developed by the JAPT during CSMC project. The difference between the CSMC butt joint and the CS butt joint is that the CS butt joint is an in-line joint, while the CSMC is a double joint through a hairpin jumper. The CS butt joint has to carry the hoop load. The straight length of the joint is only 320 mm, and the vacuum chamber around the joint has to have a split in the clamp shell. These requirements are challenging. Fig.1 presents a CSMC joint, and Fig.2 shows a CS butt joint. The butt joint procedure was verified and demonstrated. The tool is capable of achieving all specified parameters. The vacuum in the end was a little higher than the target, which is not critical and readily correctable. We consider, tentatively that the procedure is established. Unexpectedly, we discover significant temperature nonuniformity in the joint cross section, which is not formally a violation of the specs, but is a point of concern. All testing parameters are recorded for QA purposes. We plan to modify the butt joining tool to improve its convenience of operation and provide all features necessary for production of butt joints by qualified personnel.