Argonne National Laboratory

Visualizing Multiscale, Multiphysics Simulation Data: Brain Blood Flow

TitleVisualizing Multiscale, Multiphysics Simulation Data: Brain Blood Flow
Publication TypeConference Paper
Year of Publication2011
AuthorsInsley, JA, Grinberg, L, Papka, ME
Conference NameProc. VisWeek 2011
Date Published10/2011
Conference LocationProvidence, RI
Other NumbersANL/MCS-P1930-0911

Accurately predicting many physical and biological systems requires modeling interactions of macroscopic and microscopic events. This results in large and heterogeneous data sets on vastly differing scales, both physical and temporal. The ability to use a single integrated tool for the visualization of multiscale simulation data is important to understanding the effects that events at one scale have on events in the other. In the case of blood flow, we examine how the large-scale flow patterns influence blood cell behavior. In this paper we describe the visualization tools that were developed for data from coupled continuum - atomistic simulations. Specifically, we overview a) a custom ParaView reader plug-in that processes macro-scale continuum data computed by a highorder spectral element solver; and b) an adaptive proper orthogonal decomposition-based technique for the visualization of nonstationary velocity fields from atomistic simulations. We also discuss how the ParaView parallel processing and rendering infrastructure was leveraged in the new tools. We apply our methods to visualize multiscale data from coupled continuum-atomistic simulations of blood flow in a patient-specific cerebrovasculature with a brain aneurysm.