Published papers

Vortices can be considered as one of the most fundamental features of a fluid, and play an indispensable role in the understanding of dynamics of underlying flows. However, it is difficult to capture time-varying, complex vortex structures consistently with traditional trial-and-error visualization methods because flow fields are changing drastically in unsteady fluid datasets. In this paper, we therefore propose a topologically-based approach to visual data mining of unsteady flow field datasets. Since it is well-known that the vortex centers are located at local minima in the 2D associated pressure field, we identify their locations by extracting critical points of the field, segmenting it into sub-regions each of vortices using topological information, and trace the vortices in the unsteady datasets. In addition, we propose a novel transfer function design which employs constant hue assignment and a stepwise brightness setting for timeseries datasets. The design of transfer function allows users to understand the global change in the flow fields from the hue value as well as the behavior of the vortices even if the flow fields are changing drastically. The feasibility of our approach is illustrated with an application to practical datasets for Karman vortex street, which are taken from measurement-integrated simulation of the Hybrid Wind Tunnel.

This paper presents a method of alleviating such visual clutter by contracting multiple axes through the analysis of correlation between every pair of variables. In this method, we first construct a graph by connecting axis nodes with an edge weighted by data correlation between the corresponding pair of dimensions, and then reorder the multiple axes by projecting the nodes onto the primary axis obtained through the spectral graph analysis. This allows us to compose a dendrogram tree by recursively merging a pair of the closest axes one by one. Our visualization platform helps the visual interpretation of such axis contraction by plotting the principal component of each data sample along the composite axis. Smooth animation of the associated axis contraction and expansion has also been implemented to enhance the visual readability of behavior inherent in the given high-dimensional datasets.

In this paper, we propose a novel particle-based Lagrangian method for efficiently simulating viscous fluids by adopting position-based constraints. Our method uses the geometric configuration of particles for the positional constraints to approximate the dynamics of viscous fluids using position-based dynamics; thus the method can plausibly generate their motions while allowing for the use of much larger time steps than those previously adopted in the viscous fluid simulations. We also propose an associated boundary-handling scheme for position-based fluids to precisely specify boundary conditions for the constraints. Additionally, we reproduce elastic deformations of materials by controlling the constraints and incorporate thermal conduction into our framework to simulate resultant changes in particle properties and phase transition in the materials. By adjusting parameters, our method can encompass complex motions of fluids with different properties in a unified framework. Several examples demonstrate the effectiveness as well as versatility of our method.