TECH-X CORPORATION — Department of Energy SBIR Phase I: 21

Description

Statement of the problem or situation that is being addressed Magnetic fusion experiments, which aim to dependably and reliably produce electric power by confining superheated plasma in magnetic fields, are subject to a host of complex kinetic physical processes which influence the stability and power production of the plasma discharge. Analogous kinetic processes occur in low-temperature plasma discharges, and influence the e ectiveness and reliability of techniques such as etching or thin-film deposition. Kinetic processes in fusion or lowtemperature plasmas can be numerically modeled using particle-in-cell (PIC) echniques, but such models are subject to numerical instabilities unless the highest-frequency, fastest behaviors present in the model are resolved in time. The ensuing time-resolution constraints raise the computational cost of modeling low-frequency kinetic processes, sometimes unfeasibly. Further, although these high-frequency, fast kinetic processes dictate the primary computational constraints on the model, they are often of reduced importance relative to the lower-frequency kinetic physics of interest. Statement of how this problem is being addressed A new numerical technique, called ”speed-limited particle-in-cell” (SLPIC) modeling, has been developed within the past year. This approach formulates new ‘slowed-down’ equations of motion for the fastest particles in conventional PIC models, such that timesteps of increased size can be used in the particle-in-cell simulation without triggering numerical instability. Simulations of slow kinetic processes of interest can thus be run more quickly. Further, simulations which were previously considered intractable are now brought into the realm of feasibility. What is to be done in the Phase I SLPIC methods will be implemented in Tech-X’s VSim code, and tested against benchmark problems in fusion and low-temperature plasma modeling. We will demonstrate that SLPIC models can quickly and accurately compute the benchmark solutions, and will quantify the speedup enabled by these techniques. Commercial applications and other benefits Plasma processes such as etching or thin-film deposition are key manufacturing techniques in the modern electronics industry. Computer modeling plays a critical role in the design of plasma processing equipment that reliably and dependably carries out these manufacturing processes. SLPIC will enable these equipment-design e orts to proceed more quickly and e ciently, and also allow more complex design studies (which are too computationally costly using present-day PIC models) to be explored. Such e orts ultimately permit quality increases and price reductions in commercial and consumer electronic devices (phones, tablets, laptops, etc.). Key words: speed-limited particle-in-cell, kinetic processes, timestep constraints, plasma processing Summary for members of Congress Computer models of plasma discharges, used in the manufacture of modern electronic devices and also in fusion energy experiments, must take many small timesteps to be accurate. We are developing a new technique, SLPIC, which enables these models to take larger timesteps and run faster while still retaining good accuracy.