Doctor of Philosophy (PhD), Wright State University, 2010, Computer Science and Engineering PhD

Three dimensional integration-based geometric visualization is a very powerful tool for analyzing flow phenomena in time dependent vector fields. Streamlines in particular have many perceptual benefits due to their ability to provide a snapshot of the vectors near key features of complex 3D flows at any instant in time. However, streamlines do not lend themselves well to animation. Subtle changes in the vector field at each time step lead to increasingly large changes between streamlines with the same seed point the longer they are integrated. Path lines, which show particle trajectories over time suffer from similar problems when attempting to animate them. Dynamic deformable objects in the flow domain also complicate the use of integration-based visualization. Current methods such as streamlines, path lines, streak lines, particle advection and their many conceptual and higher dimensional variants produce undesirable results for this kind of data when the most important flow phenomena occurs near and moves with the objects. In this work I present methods to handle both of these problems. First, the flowing seed point algorithm is introduced, which visually captures the perceptual benefits of smoothly animated streamlines and path lines by generating a series of seed points that travel through space and time on streak lines and timelines. Next, a novel dynamic seeding strategy for both streamlines and generalized streak lines is introduced to handle deformable moving objects in the flow domain in situations where static seeding objects fail for most time steps. These two methods are then combined in order to visualize the instantaneous direction and orientation of a flow which results from flapping objects in a fluid. Initial tests are performed with a single rigid flapping disk. Further tests were performed on a more complex biologically inspired CFD simulation of the deformable flapping wings of a dragonfly as it takes off and begins to maneuver. For this test (open full item for complete abstract)

Committee: Thomas Wischgoll PhD (Advisor); Yong Pei PhD (Committee Member); Arthur Goshtasby PhD (Committee Member); Haibo Dong PhD (Committee Member); Joerg Meyer PhD (Committee Member) Subjects: Computer Engineering; Computer Science

Doctor of Philosophy, The Ohio State University, 2012, Computer Science and Engineering

Many scientific and engineering fields employ computer simulations of specific phenomena to help solve complex problems. Supercomputers and other high performance computing machines are regularly used to perform these scientific simulations. The resulting data then needs to be analyzed and visualized, which is difficult when the data is large. One approach to producing visualizations faster is to generate them in parallel. Many challenges remain, though, when attempting to analyze and visualize large data in parallel, while maintaining good performance and scalability. The size of the data is one challenge. When data size becomes very large, the I/O overhead from loading the data becomes a bottleneck, which could hinder performance. In addition, some visualization algorithms have unknown communication and computational load, which results in poor workload distribution and load balancing. This load imbalance hinders overall scalability. Another possible reason for poor parallel performance is that the method does not take advantage of the specific hardware architecture of the host machine. In order to meet these challenges, we present methods to parallelize several visualization techniques. First, a scalable shared memory rendering technique was found by adapting established parallel rendering methods to a shared memory architecture. Three rasterization methods, including sort-first, sort-last, and a hybrid method, were tested on a large shared-memory machine. Next, parallel streamline generation in static flow fields, due to the nature of the problem, suffers from high load imbalance. To make the computation more load balanced, we analyzed the flow field and estimated the workload of each block in the flow field. A load balanced partitioning of data blocks was then computed from this workload estimation. In our tests, we were able to scale up to thousands of processes while using hundreds of thousands of seeds. For time-varying flow fields, the Finite-Time Lyapunov (open full item for complete abstract)

Committee: Han-Wei Shen PhD (Advisor); Yusu Wang PhD (Committee Member); Gagan Agrawal PhD (Committee Member); Kate Calder PhD (Committee Member) Subjects: Computer Science

Master of Science, The Ohio State University, 2011, Computer Science and Engineering

We present a framework to interactively explore features present in a vector field through three geometric measurements: curl, curvature, and torsion. Measuring each streamline results in a distribution of measurements. The key assertion to our framework is that these distributions reflect characteristics of the streamline effectively. Therefore, from these distributions we may begin to compare each streamline. We perform dimensionality reduction using two methods, principal component analysis and multidimensional scaling. The MDS is performed by calculating the pair wise distance between cumulative distribution functions and the principal component analysis is performed on a set of descriptive vectors built from summary statistics. The statistics in the vector are built from one or more distributions of geometric measurements for every streamline. The idea is to place streamlines in feature space relative to how different they are with respect to each streamline's geometric measurement distribution. If the distributions are the same then the streamlines will occupy the same spot in feature space. Finally, our framework mitigates feature occlusion by using a feature space from which to select streamlines. To explore the vector field, streamline selections of various sizes are made from the feature space. This renders a subset of streamlines ideally revealing a feature of the underlying vector field.

Committee: Han-Wei Shen (Advisor); Rick Parent (Committee Member) Subjects: Computer Science; Statistics