Comparison between finite element and experimental evidences of innovative W lattice materials for sacrificial limiter applications

https://doi.org/10.1016/j.fusengdes.2021.112493Get rights and content

Abstract

Power exhaust is a key mission for the realization of fusion electricity. Engineering challenges may arise from the extreme heat fluxes developed during plasma transients, above the limit offered by existing materials. These can reduce the lifetime of plasma-facing components (PFCs), imposing extraordinary maintenance, reactor safety issues and ultimately delayed return to normal operation. Concerning the EU DEMO reactor, discrete sacrificial limiters are being investigated as the last safety resource of the reactor’s wall in case of unmitigated events. Within this context, micro-engineered tungsten (W) lattices are proposed to cope with unmitigated plasma disruptions. Unlike bulk W, lattices can be tailored to meet the operational requirements of the limiter, compromise between steady-state and off-design performances while avoiding overloading of the heat sink and delay the need for extraordinary maintenance. By calibrating an equivalent solid model originally developed and validated for open-cell aluminum (Al) foams, tailored lattices have been modelled and samples fabricated through additive manufacturing for characterization and testing, currently ongoing.

In the present work, the thermal response of lattice samples during thermal shock high heat flux (HHF) tests performed at the linear facility QSPA Kh-50 facility is simulated using ANSYS and compared with available results. Enthalpy changes of W were imposed to simulate phase change. Good agreement with experiments and SDC-IC reference up to melting point was observed. Ultimately, a thermal quench of an unmitigated DEMO disruption was simulated involving an original MAPDL routine that removes mesh elements at the melting or vaporization point.

Introduction

Controlling the heat and particle exhaust inside a tokamak is a key issue towards the realization of nuclear fusion [1]. Inside a tokamak, the lifetime of plasma facing components (PFCs) is influenced by the harsh operating conditions they cope with, in terms of high heat flux, erosion, sputtering as well as additional limitations due to the presence of a neutron flux [2]. Currently, W is the preferred armor material owing to its high melting point, acceptable thermal conductivity, high sputtering threshold and low fuel retention. On the other hand, it is brittle at room temperature and a mismatch exists between its thermal expansion coefficient and that of heat sink materials. In present-day layouts of actively cooled PFCs, the heat sink is a structural part where the coolant flows to exhaust the heat. For this porpose, copper alloy (CuCrZr) and reduced activation ferritic-martensitic steels are to date the reference materials for actively cooled heat sink. Additionally, established manufacturing routes of W represent a limit to the design of advanced layouts [3]. Considering EU DEMO and future reactors, recent studies suggest that the greatest challenges of PFCs may arise from the occurrence of plasma transients, such as disruptions [4]. For instance, during unmitigated disruption thermal quench (TQ) the first wall (FW) will be exposed to wall power density in the order of tens of GW/m2 for few ms [4]. The expected damage factor will be far greater than 50 MJ/s0.5 m2 which is the limit of W [5]. Consequently, no existing material would be able to survive such an extreme amount of energy deposited. Surface vaporization, melting and re-solidification is therefore expected. In severe cases, an eventual failure of conventional PFCs might compromise the reactor safety and delays its return to normal operation. Furthermore, the design of DEMO FW components is currently based on the normal operation, when the heat loads are anticipated to be always below 1.5 MW/m2 [6]. Disruptions should be avoided as much as possible in DEMO. An active field of research aims at developing predictive codes to promptly recognise the precursor of MHD instabilities that may trigger disruptions [7]. When controlled discharge shutdown is not possible, shuttered pellet and massive gas injection are employed as mitigation stratedy to reduce impact on the first wall components at the contact point. However, in the EU DEMO, this could still be not enough, and wall protection strategies are mandatory. These would represent the last resource to prevent the FW from excessive degradation.

Combined with mitigation and prevention strategies, sacrificial limiters are being proposed for the EU DEMO to prevent the direct interaction between the plasma and unshadowed wall [8]. In this context, the aim of this work is to develop a sacrificial limiter to cope with unmitigated upward disruptions. For this purpose, the sacrificial limier will be provided with newly-engineered lattices tailored to match its functional and conflicting requirements. In fact, on one hand adequate heat exhaust capability are needed to withstand the steady-state load in normal operation. On the other hand, when the disruption occurs, vapor shielding formation must be promoted and overloading of the heat sink must be avoided. In this regard, TOKES code predictions have provided encouraging results, with the wall heat flux significantly reduced (up to 8–10 times) when vapor shielding is involved [9]. By interposing a thermal insulator between the heat sink and the sacrificial armor, we might fulfill this goal. However, one can observe that normal and transient design requirements are conflicting with each other. In a previous study [10], we observed that unlike dense W, tailored open-cell lattices can be regarded as a promising sacrificial material: a proper design of their cell and ligaments can be carried out to tailor their thermo-mechanical response. By doing so, they could offer enhanced vapour shielding and perfect insulation of the heat sink against overloading. Moreover the larger design flexibility with respect to bulk W also helps achieve the needed trade-off between normal operation and transient requirements. Pioneeristic literature on W-based micro-engineered armors suggest foams and lattice structures as promising armor materials for transient applications in fusion [11]. Above all, ther peculiar rotation and deformation of ligaments results in less pronounced thermal stresses at high temperatures and more ductile behavior [12]. On the other hand, their geometrical features are hard to reproduce using conventional manufacturing routes of W. Additive manufacturing techniques have been employed to gain the design freedom for the fabrication of complex parts. In fact, recent studies have qualified Laser Powder Bed Fusion (LPBF) for the fabrication of complex W parts for fusion applications [13].

Recently, an existing solid model originally developed for Al open-cell foams [14] was calibrated to design optimized lattice geometries having 49.6 % and 53.3 % relative density but relative thermal conductivity respectively of 39.9 % and 25.2 % with respect to the base material. As part of the collaboration established between Tuscia and “Tor Vergata” University in the framework of the EUROfusion WPDIV activity, W lattice samples have been successfully designed and fabricated by means of LPBF [15]. Material characterization is currently ongoing as well as thermal shocks high heat flux (HHF) testing on linear plasma devices, such as at the QSPA-Kh50 facility.

In the present work, preliminary evidences from the experimental campaign at QSPA Kh-50 [16] are compared with results from finite element (FE) analyses carried out in ANSYS. Enthalpy changes of W as well as an original MAPDL routine that removes the mesh elements at the melting or vaporization point were implementd to simulate phase change. Ultimately, the thermal quench of an unmitigated DEMO disruption thermal quench (TQ) was simulated. Conservative assumptions were imposed to the loads reported in [6]. After an initial stationary operation under 1.5 MW/m2, TQ was imposed consisting of 5 G W/m2 deposited in 2 ms.

Section snippets

Cell geometry and optimized mesh

As shown in Fig. 1, two sample layouts, called A and B model, were tested on QSPA Kh-50 [16]. Each sample is a 10 × 10 × 10 mm3 cube obtained from the linear repetition of the regular elementary cell. The geometrical features of each cell variant are listed in Table 1 together with the correponsing relative density and relative thermal conductivity. A refers to the cell anisotropy, an index corresponding to the ratio between vertical and horizontal cell dimensions. L is the ligament length and

Modelling phase change

Aiming at reproducing the degradation of lattices during severe plasma transients in the EU DEMO, melting and vaporization of W were included in the simulation through enthalpy changes. We did not account for erosion, vapour shielding and melt motion since the major goal of the work was centered on the possibility to model the physics of vaporization and melting in finite element analysis. Enthalpy was defined in J/m3, as a non-linear function of temperature valid from room temperature to

FE simulation of pulses at QSPA-Kh50 – model validation

QSPA-Kh50 is a linear plasma gun hosted at the KIPT Institute of Karchov, Ukraine. It is able to deliver up to 30 MJ/m2 plasma stream energy in 0.25 ms, with initial target preheating up to 500 °C [19]. Despite the very short pulse duration, plasma energy is compatible with the TQ in EU DEMO. Both plasma stream energy density Q and target (absorbed) energy density q are measured by calorimetry: the plasma stream energy by a movable calorimeters that can be protruded beyond the target whereas

EU DEMO scenario simulation

The thermal behavior of W lattices as sacrificial armor material for the PFCs of the EU DEMO upper limiter (UL) was assessed in ANSYS under preliminary assumptions. The considered layout of the PFC is based on the flat tile configuration in Fig. 6 and is the result of a previous design optimization [10]. It consists of a water cooled heat sink in CuCrZr, provided with a 10 mm thick armour consisting of W lattices (B278 layout), for a global size of 23 × 26.5 mm. Even if a rectangular cooling

Conclusion

In the present work, we have compared thermal FE analyses and preliminary evidences from the preliminary experiments at the QSPA-Kh50 facility [16], carried out by CCFE on W lattice samples. Non-linear material properties including enthalpy changes of W were implemented to simulate melting and vaporization during fast transients. Model validation in the vaporization regime is still ongoing, but good agreement was suggested by replicationg existing studies and experiments [30]. Similarity with

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work has been carried out within the framework of the EUROfusion Consortium and has received funding from the Euratom research and training program 2014-2018 and 2019-2020 under grant agreement No 633053. The views and opinions expressed herein do not necessarily reflect those of the European Commission.

References (31)

  • T.R. Barrett

    Fus. Eng Des.

    (2016)
  • Av. Müller

    Nuc. Mat En.

    (2019)
  • F. Maviglia

    Fus. Eng. and Des.

    (2018)
  • S. Pestchanyi

    Fus. Eng. Des.

    (2016)
  • R. De Luca

    Fus. Eng. Des.

    (2020)
  • M. Andersen

    Fus. Eng. Des.

    (2006)
  • P. Fanelli

    Mat. Des.

    (2017)
  • P. Tolias

    Nuc. Mat. En.

    (2017)
  • I.E. Garkusha

    J. Nuc. Mat.

    (2009)
  • J.H. You

    Nuc. Mat. En.

    (2018)
  • R. De Luca

    Fus. Eng. Des.

    (2019)
  • J.H. You

    Nuc. Mat. En.

    (2015)
  • A. Cardella

    J. Nuc. Mat.

    (2000)
  • F. Maviglia

    Nuc. Mat. En.

    (2021)
  • F.A.J.H. Donnè

    Philosophical transactions of the royal society a: mathematical

    Phys. Eng. Sci.

    (2019)
  • Cited by (0)

    View full text