129
reads

Jong-Shinn Wu

2013-03-29
10:45:00 - 11:10:00

101 , Mathematics Research Center Building (ori. New Math. Bldg.)

Low-temperature non-equilibrium plasmas (or gas discharges) have found wide applications in modern science and technology, including semiconductor related materials processing, display technology, energy source, gas laser, surface cleaning, surface modification, analytical chemistry, electric propulsion and, especially recently, biomedical field. Understanding of these complex plasma physics and chemistry would lead to a better design of plasma sources in practical application. In addition to modern electrical and optical diagnostics of plasma phenomena, computer modeling becomes a cost-effective and important alternative method in unveiling complex physics and chemistry of gas discharges. Among these, fluid modeling that is a velocity moment of the Boltzmann equation represents one of the most powerful numerical methods for modeling most gas discharge problems. However, it is still very time-consuming, especially for multidimensional problems with complex plasma chemistry, and with electromagnetic wave effect, to name a few. Thus, it is very important to reduce the computational time dramatically to make fluid modeling an effective design and research tool. In this paper we would like to present our recent progress in developing a set of parallel fluid modeling tools for simulating general low-temperature plasma physics and chemistry, and their possible applications. In the proposed simulation tools, both the compressible flow equations (continuity, Navier-Stokes, energy and species equations) and the fluid modeling equations (charged and neutral species continuity, drift-diffusion approximation for momentum, electron energy density and Poissons equations) are discretized using collocated cellcentered finite-volume method via semi-implicit scheme. The former is solved based on an extended SIMPLE scheme which makes the code valuable at all speeds [1]. The latter is solved using Scharfetter-Gummel scheme that is implemented to resolve thin sheath near the solid walls. In addition, either the frequency- or time-domain Maxwells equation solver using finite difference (FD) or finite difference time domain method (FDTD), respectively, is coupled to consider electromagnetic wave effect in plasma when necessary. All the codes are parallelized based on domain decomposition using MPI (message passing interface) for large-scale computation [2, 3]. A practical multi-scale concept is also proposed to speed up the coupling between gas flow and fluid modeling solvers [4]. To demonstrate the accuracy of our fluid modeling codes, simulations are benchmarked against previous references or experimental data obtained in our laboratory. High-level quantum chemistry calculations have being used to obtain reliable rate constants for those reaction channels either without proper experimental data or very difficult to obtain. Several realistic examples, including 100 MHz driven argon plasma enhanced chemical vapor deposition (PECVD) process and helium DBD atmospheric-pressure plasma jet, are demonstrated. Finally, outlook of the future research is also presented. [1] M.-H. Hu, J.-S. Wu and Y.-S. Chen, Computers & Fluids, 45, 241 (2011). [2] C.-T. Hung, M.-H. Hu, J.-S. Wu, and F.-N. Hwang, Computer Physics Communications, 177, 138 (2007). [3] K.-M. Lin, C.-T. Hung, F.-N. Hwang, M. R. Smith,Y.-W. Yang and J.-S. Wu, Computer Physics Communications, 183, 1225 (2012). [4] K.-M. Lin, M.-H. Hu, C.-T. Hung, J.-S. Wu, F.-N. Hwang, Y.-S. Chen and G. Cheng, Computer Physics Communications, 183, 2550 (2012).