Elsevier

Acta Materialia

Volume 60, Issue 20, December 2012, Pages 6991-7000
Acta Materialia

On the role of pre-existing defects and magnetic flux avalanches in the degradation of YBa2Cu3O7–x coated conductors by quenching

https://doi.org/10.1016/j.actamat.2012.09.003Get rights and content

Abstract

YBa2Cu3O7–x (YBCO) coated conductors are emerging as an important option for magnets for energy systems and experimental science. One of the remaining challenges for YBCO superconducting magnets is quench protection, i.e. ensuring that the YBCO is not damaged due to a fault condition. One key issue is understanding the underlying causes of degradation during a quench. Here, the microstructure of a quenched, degraded sampled is compared to that of an unquenched control sample. To facilitate microstructural analysis of the YBCO surface, the Cu stabilizer and Ag cap layer were removed by etching. Reactions between the Cu etchant and YBCO proved to be a signature of Ag/YBCO delamination. Two types of pre-existing defects were identified as initiation points of degradation. Defects on the conductor edge resulting in delaminated Ag lead to dendritic flux avalanches and high local heating, which cause further Ag delamination. This self-propagating effect results in dendritic Ag delamination, which is seen through etchant–YBCO reactions. Defects within the YBCO layer result in breaches in the protective Ag layer such that Cu etchant penetrates and reacts with the YBCO. Energy-dispersive X-ray spectroscopy analysis showed similar reactions as in the edge degradation but also showed pure Ag particles, which indicates that the local temperature was sufficient to cause localized Ag melting.

Introduction

Significant progress has been made in the technology of high-temperature superconductors (HTS), and technical conductors are now industrially manufactured in sufficient lengths for the development of magnets. One important emerging conductor is YBa2Cu3O7–x (YBCO), which has a very high critical current density (Jc) at elevated temperature or in a very high magnetic field, and very high mechanical strength [1], [2]. As a result, there is commercial interest in YBCO superconducting magnets for nuclear magnetic resonance, high-energy physics, superconducting magnetic energy storage and wind turbine generators [1], [3], [4], [5], [6].

YBCO coated conductors (CC) are manufactured using thin film deposition techniques on Ni alloy substrates [2], [7], [8]. One or more thin buffer layers, which play an integral role in introducing the long-range biaxial texture in the YBCO layer that is necessary for high Jc, are deposited onto the substrate to separate the YBCO layer from the Ni alloy. After encasing the YBCO layer with a thin, protective Ag cap layer, the conductor is encased by stabilizer, typically Cu. The CC is manufactured as a wide, thin tape with very low YBCO volume fraction (1–2%), which is offset by the very high Jc in the YBCO layer. The remarkably high Jc is obtained because the YBCO is highly engineered, with a minimal number of undesirable defects, and because of the inclusion of controlled defects that enhance magnetic flux pinning [9], [10], [11], [12], [13], [14]. With recent advances, YBCO CC are approaching “designability” at multiple length-scales [15], including defects that optimize flux pinning for electrical performance at the desired magnetic field and temperature [16], [17], an optimized YBCO layer thickness for critical current, and varied Ni alloy and Cu stabilizer thicknesses for quench protection [15] and mechanical strength.

One of the remaining challenges for YBCO superconducting magnets is quench protection, i.e. ensuring that the YBCO CC is not damaged due to a sudden superconducting-to-normal transition initiated by a fault condition or transient. Effective quench protection requires reliable, timely quench detection and the ability to respond to a quench in a manner that prevents conductor degradation. It is well established that, although YBCO magnets are quite stable, quench detection is particularly challenging because of very slow quench propagation. Macroscopic measurements of quench propagation velocity consistently show that quench propagation is at least an order of magnitude slower than in NbTi and Nb3Sn magnets, rendering quench detection a significant challenge [15], [18], [19], [20], [21]. Although there are new, emerging options for quench detection [22], [23], it is clear that understanding the YBCO CC response to a quench is essential.

Although the Ni alloy substrate is mechanically strong, YBCO CCs are prone to interlayer delamination failures. Such failures can be caused by interactions between the CC and epoxy during cool-down or other sources of interfacial shear stress [5], [18], [24], [25], [26], [27]. Delamination failure is also a risk factor during quenching. Quenching in YBCO CC is a complicated phenomenon involving time-varying electrical, magnetic, thermal and mechanical behaviors [21], [28], [29]. Slow quench propagation is only problematic if it puts the conductor at risk. Thus, to understand quench protection requirements, one must understand the safe operating limits and the underlying causes of degradation. Recent simulation results have shown that very large thermal gradients can occur within a YBCO CC during a quench [30]. Furthermore, understanding the underlying causes of degradation may facilitate improved conductor engineering such that the limits are increased.

Neither voltage nor temperature measurements provide sufficient information regarding conductor degradation induced by quenching. Jc measurements indicate whether the sample has been degraded, providing spatially averaged information over the voltage-tap length-scale, but do not provide localized information [31], [32]. Typically, quench protection requirements are based upon a surface temperature threshold, which is used as an operational limit. Since the surface temperature is influenced by an extrinsic thermal source, and a time-varying temperature gradient exists between the CC surface, where a sensor may be located, and the YBCO layer, where degradation occurs, temperature limits vary significantly and depend on the specific characteristics of the sample and experimental approach used to determine the limit. As a result, there is no established, reliable temperature limit for YBCO CC upon which a quench protection system can be designed that is both sufficiently conservative to not risk failure yet sufficiently informed to not result in an over-engineered protection system.

During a quench, the time-varying thermal gradient between the Ag and YBCO layers may initiate delamination. If the thermal gradient is sufficiently large, if multiple quenches result in thermal fatigue or if there are large Lorentz forces on the conductor after steady-state operation has resumed, a locally delaminated region of the Ag layer may separate further, resulting in YBCO that is locally unstabilized. In fundamental studies of unstabilized HTS thin films, quench behavior is explained by the flux-flow instability theory of Larkin and Ovchinikov (L&O) [33], later modified by Bezuglij and Shklovskij [34]. They found that, besides the critical current density Jc at which dissipation begins (low-dissipative regime), a highly dissipative regime occurs abruptly (voltage jump) at the supercritical current density which is several times Jc. The central question of whether it is the high current density, temperature or magnetic field that drives the voltage jump, however, remains unanswered. While current–voltage measurements are a primary tool for studying flux-flow instabilities [12], [35], [36], they do not provide localized spatial information.

Magneto-optical imaging (MOI) is a tool for investigating magnetic domains via optical microscopy [37], [38], [39] in the presence of non-magnetic layers situated between the magnetic layer (YBCO) and the microscope. Using time-resolved MOI, Song et al. [32] investigated dynamic current redistribution during quenching in stabilized YBCO CC. In this work, a time-varying triangularly shaped propagating normal zone, consistent with the Critical State Model for current and magnetic field penetration, was visualized. In contrast, using MOI on unstabilized MgB2 thin films, Johansen et al. [40] observed abrupt penetration of magnetic flux dendrites. Several theories have been proposed to explain why magnetic flux avalanches develop dendritic patterns [41], [42], [43], [44]. By solving thermal diffusion and Maxwell equations, taking into account nonlocal electromagnetic fields in the film and thermal coupling to the substrate, one analysis showed that thermomagnetic instability results in a dendritic distribution of temperature, electric field and magnetic field. The dendritic distribution occurs when the background electric field is larger than a threshold, E > Ec, where Ec is the critical electric field [45], [46], [47]. Such dendritic growth is expected to occur only in the absence of a metallic layer in contact with the superconductor, as the presence of a conducting metallic layer reduces the electric field. Bobyl et al. [48], [49] found that a pulsed transport current also created dendritic patterns. In this case, the pulsed current induces the electric field related to the dendritic penetration. There are no reports of dendritic flux penetration in stabilized superconducting films in the absence of such a driver.

Experimental and theoretical studies of L&O theory and avalanche-like flux dynamics observed using MOI are helpful in understanding quench behavior in YBCO CC but do not give any insight into conductor degradation induced by quenching. Numerical modeling and simulation provides dynamic three-dimensional mapping of temperature and other parameters [21], but simulation results depend on unknown physical parameters, including interfacial resistivities between layers of the conductor. To fully understand degradation in the YBCO layer, one must observe the degraded YBCO microstructure directly. This requires removal of the surrounding Cu stabilizer and Ag cap layer in such a manner that does not affect the YBCO layer.

Here we investigate degradation in YBCO CC due to quenching. This study reveals the quench degradation mechanisms in YBCO CC and provides evidence of unstable flux motion in two dimensions as predicted by L&O theory. The results extend the dendritic avalanche of flux motion in the superconducting state as observed by MOI to dendritic avalanches during quenching as a cause of Ag/YBCO of delamination.

Section snippets

Experimental approach

The experimental approach involves inducing controlled quenches in YBCO CC such that the conductor is partially degraded. The microstructure of the degraded conductor is then compared to that of an unquenched control sample.

Onset of degradation in transport behavior due to quenching

The peak temperature profiles for quench cycles 3, 5, 6, 7 and 8 are seen in Fig. 2. Thermocouple TC34, which is adjacent to the heater, became loose and detached from the conductor after quench 5, so the peak temperature appears to have shifted to TC35. For quench cycle 7, the peak surface temperature measured at TC35 is 290 K and the peak voltage is 0.57 V; no reduction in Ic occurred. In quench cycle 8, the measured peak surface temperature increases to 335 K, with a peak voltage of 0.80 V. Due

Discussion

The microanalysis shows that there are two distinct sources of degradation in YBCO CCs during quenching: Ag delamination or defects at the conductor edge and pre-existing defects within the YBCO layer, resulting in localized hot-spots.

Conclusions

This study has investigated the underlying causes of degradation in YBCO CC during quenching. The microstructure of a quenched sampled, with a 6.3% reduction in Ic, was compared to that of an unquenched control sample. To facilitate microstructural analysis of the YBCO surface, the Cu stabilizer and Ag cap layer were removed through a two-step etching process. Chemical reactions between the Cu etchant and YBCO, identified primarily via the presence of S, proved to be a signature of Ag/YBCO

Acknowledgements

The authors thank Yan Xin, Robert Goddard, Fumitake Kametani, Jianyi Jiang and Aixia Xu for assistance with SEM and EDS, Hubertus Weijers and W. Denis Markiewicz for providing the YBCO sample, Liyang Ye for assistance with the quench experiments and David C. Larbalestier, Ulf Trociewitz and Jun Lu for helpful discussions.

References (59)

  • T. Takematsu et al.

    Phys C – Superconduct Appl

    (2010)
  • Y. Yanagisawa et al.

    Phys C – Superconduct Appl

    (2011)
  • Y. Yanagisawa et al.

    Phys C – Superconduct Appl

    (2012)
  • T. Izumi et al.

    Phys C – Superconduct Appl

    (2009)
  • J. Matsuda et al.

    Phys C – Superconduct Appl

    (2008)
  • J. Schwartz et al.

    IEEE Trans Appl Superconduct

    (2008)
  • V. Selvamanickam et al.

    IEEE Trans Appl Superconduct

    (2009)
  • Y. Iwasa et al.

    IEEE Trans Appl Superconduct

    (2010)
  • C.L.H. Thieme et al.

    IEEE Trans Appl Superconduct

    (2009)
  • U.P. Trociewitz et al.

    Appl Phys Lett

    (2011)
  • A.B. Abrahamsen et al.

    Superconduct Sci Technol

    (2010)
  • V. Selvamanickam et al.

    MRS Bull

    (2004)
  • Y. Shiohara et al.

    Superconduct Sci Technol

    (2008)
  • T. Aytug et al.

    J Appl Phys

    (2008)
  • Z.J. Chen et al.

    Appl Phys Lett

    (2007)
  • D.M. Feldmann et al.

    IEEE Trans Appl Superconduct

    (2001)
  • J. Gutierrez et al.

    Nat Mater

    (2007)
  • J.L. Macmanus-Driscoll et al.

    Nat Mater

    (2004)
  • S.R. Foltyn et al.

    Nat Mater

    (2007)
  • W.K. Chan et al.

    IEEE Trans Appl Superconduct

    (2011)
  • T.G. Holesinger et al.

    Adv Mater

    (2008)
  • T. Haugan et al.

    Nature

    (2004)
  • A. Ishiyama et al.

    IEEE Trans Appl Superconduct

    (2009)
  • X. Wang

    Superconduct Sci Technol

    (2011)
  • R.C. Duckworth et al.

    AIP Conf Proc

    (2002)
  • W.K. Chan et al.

    IEEE Trans Appl Superconduct

    (2010)
  • J. Schwartz et al.
  • Turenne M, Johnson R, Hunte F, Schwartz J, Song, H. Multi-purpose fiber optic sensors for high temperature...
  • Y. Shiroyanagi et al.

    IEEE Trans Appl Superconduct

    (2011)
  • Cited by (55)

    • Gaussian process modeling of doped-lanthanum manganites Curie temperature from compositions and ionic radii

      2023, Solid State Communications
      Citation Excerpt :

      There are several advantages of the MR technology over the traditional CGC, including much improved cooling efficiency, complete solid-state operations, much smaller footprints, and being benign to the environment [2–6]. In addition, high-temperature superconductors that can generate high magnetic fields are promising magnetic field sources for high-efficiency MR devices [7–19]. Principal component analysis is a quantitative method for generating a new set of variables called principal components, each of which is a linear combination of the original variables.

    • Modeling of lattice parameters of cubic perovskite oxides and halides

      2021, Heliyon
      Citation Excerpt :

      With different constituting elements, compounds will have varying properties, which could include magnetoresistance, ferromagnetism, ferroelectricity, superconductivity, and piezoelectricity [2, 3, 4, 5].

    View all citing articles on Scopus
    View full text