Mykonos Power Flow Physics Research

Power flow physics research focuses on understanding and improving the efficiency of pulsed power current delivery, which is critical for various applications at Sandia National Laboratories (SNL), including stockpile stewardship, dynamic materials research, and fusion science. The TW-class Z machine is capable of delivering up to 20 MA of current to achieve the desired z-pinch load implosion. However, the accelerator’s Magnetically Insulated Transmission Line (MITL) exceeds 240 kV/cm E-fields and 1 MA/cm2 current densities, which subsequently sources 1×1014 – 1×1018 cm-3 plasmas. These plasmas divert a portion of the machine current across the Anode-Cathode (A-K) gap, thus reducing the current delivery to the load region and overall machine efficiency [1].

The findings from both experimental and simulation analyses indicate that the dynamics of transmission line electrode heating, surface contamination, and the subsequent A-K gap plasma formation and transport play a crucial role in the experienced current reduction. Electrode coating and cleaning engineering techniques, which are thought to adjust these dynamics and improve current delivery, are being fielded at the 1 MA Mykonos accelerator. Additional MagnetoHydroDynamic (MHD) and Particle-In-Cell (PIC) computer simulation modeling is being conducted to support the experimental research. By leveraging these advanced computational techniques and new experimental insights, power flow physics research not only addresses present challenges but also lays the groundwork for future innovations in pulsed power technology, ultimately contributing to the broader goals of national security and scientific advancement.

Parallel plate platform

To provide diagnostically accessible lossy power flow conditions on Mykonos that are similar to what is expected in Z’s inner MITL, Derek Lamppa developed the parallel plate platform hardware [2], shown below on the left. Fielded is an elliptical cathode cross section, shown below on the right, to produce an approximately uniform current density. The spacing between the Anode and Cathode can be varied between 0.5 – 3.0 mm.

Right plane sectional view of Mykonos’ entire parallel plate platform hardware [3].
Cross sectional view of Mykonos’ parallel plate platform A-K rod hardware [3].

There is ongoing work to increase the field uniformity between the Anode and Cathode by using electrodes with an Ernst surface profile, rather than elliptical, leading to lower ambiguity in the results from collected data [4].

A COMSOL simulation of the unform field parallel plate platform that utilizes Ernst profiles [5].

Electrode surface coatings

The campaign aimed to assess the impact of a bi-metallic coating design, featuring a 200 µm copper inner conductor and a 10 µm manganese outer conductor, on plasma formation in conditions similar to Z’s inner MITL. Experimental results indicated that the bi-metallic coatings had a non-negligible effect on plasma characteristics, with a notable average delay of 10 ns in self-emission onset for coated electrodes compared to controls, as measured by avalanche photodiodes (APDs). Additionally, the coated electrodes exhibited a substantial reduction in plasma self-emission intensity—32% and 70% less at 60 ns and 90 ns, respectively—suggesting a decrease in plasma density of 19% and 45%. This effect is attributed to the outer conductor staying relatively cool due to the majority of current being carried by the inner conductor; this minimized thermal emission of contaminants and acted as a diffusion barrier for desorbants from the inner conductor.

Coatings show lower plasma self-emission at the same time for the coated electrodes when compared to the control electrodes. (a) Current for the two compared shots in machine time. The plasma self-emission images for the control experiments at (b) 60 ns and (c) 90 ns are shown. This is compared to the plasma self-emission from the coated experiments at (d) 60 ns and (e) 90 ns. The coated experiments show a decrease in self-emission intensity of 32% and 70% at 60 ns and 90 ns, respectively.

In-situ plasma discharge cleaning

Low binding energy contaminants on the MITL electrode surfaces, such as hydrogen and hydrocarbons, are believed to be the main source of desorbed particles and subsequent current loss plasmas. To reduce the quantity of such electrode surface contaminants, pre-shot in-situ plasma cleaning processes are under development at the Mykonos accelerator [2]. In-situ plasma discharge cleaning (i.e. plasma ion surface bombardment) removes contaminants from the electrode surface via physical sputtering or chemically enhanced gettering. The plasma discharge is generated by exciting backfill gas with either a Radio Frequency (RF) or an Audio frequency Square Wave (ASW) supply. Insulating breaks are used to isolate the cleaning plasma to the critically lossy areas.

The resultant in-situ cleaning plasma discharge generated from a 600 mTorr (80% argon, 20% oxygen) gas mixture being excited by a ~ 1 kV ASW frequency [2].

Several diagnostics are fielded on Mykonos to compare the power flow effects of uncleaned vs cleaned hardware, including B-dot probes, Avalanche PhotoDiodes (APDs), self-emission fast frame imaging, and Vacuum UltraViolet (VUV) spectroscopy.

In-situ heating, which uses embedded cartridge heaters for thermal desorption of contaminants, has faced challenges related to fabrication and alignment. However, future in-situ heating work is planned.

MagnetoHydroDynamic (MHD) computer modeling

Sandia’s Magnetohydrodynamic code ALEGRA [6] is used to study the dynamics of electrode material subjected to ohmic heating rates consistent with the inner MITL at the Z Machine. This allows for self-consistent tracking of pressure, density, temperature, and conductivity of the stainless steel electrodes throughout the ~750 kA Mykonos current pulse of the commonly fielded parallel plate platform hardware.

ALEGRA simulation of a 750 kA Mykonos pulse effect on the 1.5 x 1.0 mm diameter cathode rod, 0.5 mm gap spacing, parallel plate platform hardware. Shown is the (left) temperature and (right) density. Simulation run by Derek Lamppa.

In this above modeled Mykonos parallel plate platform configuration, the cathode ellipse (top) and anode (bottom) exceed stainless steel melt temperatures within 50 ns and begin to expand into the anode-cathode gap. Between the expanding density fronts is ostensibly where the lower density power-flow relevant plasma is observed in PIC simulations and experiments.

The MHD simulations provide an understanding of the dense electrode material dynamics without consideration of particle bombardment, which can only be resolved with Particle-In-Cell treatment. Therefore, this represents the “best case” (minimum) temperature and expansion profile for the electrode material.

Particle-In-Cell (PIC) computer modeling

Advanced simulations using the hybrid, relativistic, electromagnetic PIC code CHICAGO® have allowed researchers to develop a more accurate framework for modeling the thermal, mass, and magnetic field diffusion within electrode materials. The results from these simulations provide critical insights into the behavior of electrode plasmas and their impact on current delivery, confirming that these plasmas are a significant contributor to current loss, and inform the design of hardware and diagnostics for the Z accelerator and the proposed Next-Generation Pulsed Power (NGPP) facility.

CHICAGO® PIC, 2D cylindrical, approximated fragmentation ionization model, simulation movie showing free electron density within the MITL A-K gap of Mykonos’ parallel plate platform. Simulation run by Dale Welch. [more info]

The code has recently been enhanced with new algorithms to incorporate the effects of vaporization and melt [7] as well as the effects of resistive walls. CHICAGO® has adopted a hybrid PIC approach that combines kinetic, quasi-neutral single fluid, and multi-fluid treatments [8]. The ability to conduct large-scale simulations in a matter of days represents a transformative step forward in understanding and optimizing the complex phenomena associated with high-power accelerators.

References

[1] Bennett, N. et al., “Electrode plasma formation and melt in Z-pinch accelerators“, Physical Review Accelerators and Beams 26.040401 (2023), doi: https://doi.org/10.1103/PhysRevAccelBeams.26.040401.

[2] Lamppa, Derek et al., “Assessment of Electrode Contamination Mitigation at 0.5 MA Scale“, Technical Report SAND2021-12691 (2021), doi: https://doi.org/10.2172/1825219.

[3] Hines, N. R. et al., “Development of a colinear Second-Harmonic Orthogonal Polarization (SHOP) interferometer for electron areal density measurements in Magnetically Insulated Transmission Lines (MITLs)“, Technical Report SAND2023-10581 (2023), doi: https://doi.org/10.2172/2430189.

[4] Mason, Tyler J. et al., “Design and Preliminary Experimental Results on a Uniform-Field Test Fixture for Power Flow Experiments“, IEEE Transactions on Plasma Science, submitted (December 2024).

[5] Schwarz, Jens et al., “2024 Mykonos Facility Report“, Technical Report SAND2025-01377 (2025), doi: (pending).

[6] Robinson, A. et al., “ALEGRA: An Arbitrary Lagrangian-Eulerian Multimaterial, Multiphysics Code“, 46th AIAA Aerospace Sciences Meeting and Exhibit (2008), doi: https://doi.org/10.2514/6.2008-1235.

[7] Thoma, Carsten, and Welch, Dale, and Bennett, Nichelle Lee, and Cochrane, Kyle, “Improved melt model for power flow“, Technical Report SAND2023-09787R (2023), doi: https://doi.org/10.2172/2430312.

[8] Welch, D. R. et al., “Fast hybrid particle-in-cell technique for pulsed-power accelerators“, Physical Review Accelerators and Beams 23.110401 (2020), doi: https://doi.org/10.1103/PhysRevAccelBeams.23.110401.

Related News