The purpose of this research was to improve the smoothing operation in smoothed particle hydrodynamics, SPH, when the flow of matter is not smooth. Our main focuses are on the kernel selection, identifying the discontinuities in the sequences to be smoothed, and use of the Laplacian as opposed to artificial viscosity for improved physical accuracy. The results show that alternative kernels result in differences in how matter flows. These effects are explained by the kernels' gradient and Laplacian properties. Five alternative kernels were included in our analysis and our SPH-based simulation cases. Further, the sequences to be smoothed by the kernel function were found to contain numerous discontinuities. As it is well known from multiple areas of science, such discontinuities lead to degraded accuracy if the smoothing is performed without taking discontinuities into consideration. Several methods are introduced to detect discontinuities and perform smoothing by individually and independently smoothing the segments between discontinuities. We analyzed results from sloshing tank SPH simulations and found such segmentwise smoothing impacts the flow. Discontinuities were identified by first-generation wavelets. We found that in about 24 to 27 percent of the fluid particles have sequences containing discontinuities, independent of time step. A second-generation wavelet analysis showed coherent vorticity structure in the flow, and the fluid particles with discontinuous sequences combined with coherent vorticity were the focus of our quantification of effects on particle movement. The research work presented here serves as a tool for further improvement of the SPH method, and is substantiated by the results obtained herein.
| Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:etd2020-1833 |
| Date | 01 December 2021 |
| Creators | Brun, Oddny |
| Publisher | STARS |
| Source Sets | University of Central Florida |
| Language | English |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Electronic Theses and Dissertations, 2020- |
Page generated in 0.0019 seconds