Multipactor in Coaxial Transmission Lines

In vacuum systems where conducting or dielectric surfaces are exposed to oscillating electric fields, multipactor discharges carry the potential to significantly disrupt normal and efficient operation. Multipactor breakdown occurs when free electrons are accelerated into transmission line surfaces and undergo secondary electron emission. These secondary electrons can then repeat this process and multiply, allowing the system to develop into a growing electron discharge. Preventing multipactor is essential for ensuring long-term, efficient operation of vacuum electronic systems. At the University of Michigan, we have developed an S-band (3.05 GHz), coaxial multipactor test cell that operates in a new, high-fd regime, where f is the signal frequency and d is the gap spacing. This test cell is used to characterize multipactor discharges and to test methods for suppressing multipactor. Multipactor relies on a resonance between the electron motion and the oscillating electric field. In rectangular geometries, this resonance condition can be solved analytically and then used to generate rough approximations of the breakdown threshold (the minimum power level necessary to sustain multipactor). However, in coaxial systems this resonance is more difficult to represent theoretically. While there are several theoretical models for coaxial multipactor, they are computationally expensive, and their implementation is often impractical. To aid in the design of the multipactor test cell, we have used electromagnetic-particle-in-cell (EM-PIC) simulations to characterize the multipactor discharges and provide predictions of the experimental test cell’s susceptibility to multipactor. Initial experiments explored the multipactor self-conditioning phenomenon. During multipactor, electrons bombard the transmission line surfaces, processing oxide layers and releasing trapped gases. This can reduce secondary electron emission and increase the multipactor breakdown threshold. This process also enabled our experimental hardware to more closely match the SEY model we assumed in our PIC simulations. After fully conditioning the transmission line surfaces, the experimental measurements of the breakdown threshold agreed with our predictions to within 10--15%. We also found that that the multipactor self-conditioning process is relatively rapid when considering the cumulative timescale of the multipactor discharges. Multipactor cannot be allowed to occur in vacuum electronic systems; it will disrupt normal signal transmission and can potentially lead to catastrophic failure. The key to preventing multipactor is to suppress secondary electron emission. One method explored here relies on using textured materials. Textured surfaces can trap secondary electrons and inhibit multipactor. Traditional manufacturing techniques cannot be easily used to produce a textured coaxial transmission line. We explored whether 3D-printing processes---which are inherently textured---can be used to produce multipactor-resistant components. We experimentally tested two partially 3D-printed coaxial transmission lines---each manufactured using either Selective Laser Melting (SLM) or Atomic Diffusion Additive Manufacturing (ADAM)---to demonstrate proof-of-concept. Before our experimental investigation, we used a Monte-Carlo algorithm to predict the modification of the secondary electron yield due to surface roughness; these data were then used to run multipactor PIC simulations. Our experimental measurements consistently outperformed the simulations. The 3D-printed transmission lines were extremely effective for preventing multipactor. In particular, the ADAM-printed outer conductor nearly doubled the breakdown threshold power at our low-fd limit. These experiments suggest that 3D-printed, textured, coaxial transmission lines could act as drop-in replacements in existing devices and provide valuable microwave power margin from multipactor onset.PhDNuclear Engineering & Radiological SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/176485/1/svlangel_1.pd