Computational Modeling of Turbine-Wake Effects^*.

K. Gruchalla, M.J. Churchfield, P.J. Moriarty, S. Lee, S. Li, J.K. Lundquist, J. Michalakes, A. Purkayastha, M.A. Sprague. SciDAC 2011, Denver, CO, 2011.

As the United States moves toward utilizing more of its wind and water resources for electrical power generation, computational modeling will play an increasingly important role in improving the performance, decreasing costs, and accelerating deployment of wind and water power technologies. We are developing computational models to better understand the wake effects of wind and marine hydrokinetic turbines, which operate on the same principles. Large wind plants are consistently found to perform below expectations. Inadequate accounting various turbulent-wake effects is believed to be in part responsible for the under-performance.

Particle Dynamics in a Fluidized Bed Reactor^*.

K. Gruchalla, P. Pepiot, O. Desjardins SciDAC 2010, Chattanooga, TN, 2010.

^*2010 DOE OASCR Award -- Outstanding Achievement in Scientific Visualization

Fluidized bed reactors are a promising technology for the thermo-chemical conversion of biomass in biofuel production. However, the current understanding of the behavior of the materials in a fluidized bed is limited. We are using high-fidelity simulations to better understand the mechanics of the conversion processes. This video visualizes a simulation of a periodic bed of sand fluidized by a gas stream injected at the bottom. A Lagrangian approach describes the solid phase, in which particles with a soft-sphere collision model are individually tracked along the reactor. A large-scale point-particle direct numerical simulation involving 12 millions particles and 4 million mesh points has been performed, requiring 512 cores for 4 days. The onset of fluidization is characterized by the formation of several large bubbles that rise and burst at the surface and is followed by a pseudo-steady turbulent motion showing intense bubble activity.

Numerical Simulation of a Turbulent Atomizing Liquid Jet.

K. Gruchalla, O. Desjardins, P. Pepiot, A. Purkayastha Super Computing 2010 (SC10), New Orleans, LA, 2010.

A detailed numerical simulation of a turbulent liquid jet. Atomization of liquid fuel is the process by which a coherent liquid flow disintegrates in droplets. Understanding atomization will have far reaching repercussions on many aspects of the combustion process. This was a large-scale scaling study run on Red Mesa, , involving 1.36 billion cells, run on 12,228 cores, rendered on 1,248.

Simulation Characterization and Optimization of Metabolic Models with the High-Performance Systems Biology Toolkit.

M. Lunacek, A. Nag, D. Alber, K. Gruchalla, C.H. Chang, P.A. Graf.
SIAM Journal on Scientific Computing (vol. 33), 2011.

paper code bibtex

The High-Performance Systems Biology Toolkit (HiPer SBTK) is a collection of simulation and optimization components for metabolic modeling and the means to assemble them into large parallel processing hierarchies suiting a particular simulation optimization need. The components come in a variety of different categories: model translation, model simulation, parameter sampling, sensitivity analysis, parameter estimation, and optimization. They can be configured at runtime into hierarchically parallel arrangements to perform nested combinations of simulation characterization tasks with excellent parallel scaling to thousands of processors. We describe the observations that led to the system, the components, and how one can arrange them. We show nearly 90% efficient scaling to over 13,000 processors, and we demonstrate three complex yet typical examples that have run on ∼1000 processors and accomplished billions of stiff ordinary differential equation simulations.

Segmentation and Visualization of Multivariate Features using Feature-Local Distributions.

K. Gruchalla, M. Rast, E. Bradley, P. Mininni.
In Advances in Visual Computing, Lecture Notes in Computer Science (vol. 6938), 2011.

pre-print bibtex (The final publication is available from www.springer.com)

We introduce an iterative feature-based transfer function design that extracts and systematically incorporates multivariate feature-local statistics into a texture-based volume rendering process. We argue that an interactive multivariate feature-local approach is advantageous when investigating ill-defined features, because it provides a physically meaningful, quantitatively rich environment within which to examine the sensitivity of the structure properties to the identification parameters. We demonstrate the efficacy of this approach by applying it to vortical structures in Taylor-Green turbulence. Our approach identified the existence of two distinct structure populations in these data, which cannot be isolated or distinguished via traditional transfer functions based on global distributions.

Computational Modeling of Wind-Plant Aerodynamics.

M.A. Sprague, P.J. Moriarty, M.J. Churchfield, K. Gruchalla, S. Lee, J.K. Lundquist, J. Michalakes, A. Purkayastha.
Invited paper in Proceedings of SciDAC 2011.

paper video

As the US moves toward 20% wind power by 2030, computational modeling will play an increasingly important role in determining wind-plant siting, designing more efficient and reliable wind turbines, and understanding the interaction between large wind plants and regional weather. From a computing perspective, however, adequately resolving the relevant scales of wind-energy production is a petascale problem verging on exascale. In this paper we discuss the challenges associated with computational simulation of the multiscale wind-plant system, which includes turbine-scale turbulence, atmospheric-boundary-layer turbulence, and regional-weather variation. An overview of computational modeling approaches is presented, and our particular modeling strategy is described, which involves modification and coupling of three open-source codes—FAST, OpenFOAM, and WRF, for structure aeroelasticity, local fluid dynamics, and mesoscale fluid dynamics, respectively.

Integration and Dissemination of Citizen Reported and Seismically Derived Earthquake Information via Social Network Technologies.

M. Guy, P. Earle, C. Ostrum, K. Gruchalla, S. Horvath.
In Advances in Intelligent Data Analysis, Lecture Notes in Computer Science (vol. 6065), 2010.

pre-print bibtex (The final publication is available from www.springer.com)

People in the locality of earthquakes are publishing anecdotal information about the shaking within seconds of their occurrences via social network technologies, such as Twitter. In contrast, depending on the size and location of the earthquake, scientific alerts can take between two to twenty minutes to publish. We describe TED (Twitter Earthquake Detector) a system that adopts social network technologies to augment earthquake response products and the delivery of hazard information. The TED system analyzes data from these social networks for multiple purposes: 1) to integrate citizen reports of earthquakes with corresponding scientific reports 2) to infer the public level of interest in an earthquake for tailoring outputs disseminated via social network technologies and 3) to explore the possibility of rapid detection of a probable earthquake, within seconds of its occurrence, helping to fill the gap between the earthquake origin time and the presence of quantitative scientific data.

VAPOR: Visual, Statistical, and Structural Analysis of Astrophysical Flows.

J. Clyne, K. Gruchalla, M. Rast.
In Proceedings of Numerical Modeling of Space Plasma Flows: Astronum-2009

paper bibtex code

In this paper we discuss recent developments in the capabilities of VAPOR (open source, available at http://www.vapor.ucar.edu): a desktop application that leverages today’s powerful CPUs and GPUs to enable visualization and analysis of terascale data sets using only a commodity PC or laptop. We review VAPORs current capabilities, highlighting support for Adaptive Mesh Refinement (AMR) grids, and present new developments in interactive feature-based visualization and statistical analysis.

Visualization-Driven Structural and Statistical Analysis of Turbulent Flows.

K. Gruchalla, M. Rast. E. Bradley, J. Clyne, P. Mininni.
In Advances in Intelligent Data Analysis, Lecture Notes in Computer Science (vol. 5772), 2009.

pre-print bibtex (The final publication is available from www.springer.com)

Knowledge extraction from data volumes of ever increasing size requires ever more flexible tools to facilitate interactive query. Interactivity enables real-time hypothesis testing and scientific discovery, but can generally not be achieved without some level of data reduction. The approach described in this paper combines multi-resolution access, region-of-interest extraction, and structure identification in order to provide interactive spatial and statistical analysis of a terascale data volume. Unique aspects of our approach include the incorporation of both local and global statistics of the flow structures, and iterative refinement facilities, which combine geometry, topology, and statistics to allow the user to effectively tailor the analysis and visualization to the science. Working together, these facilities allow a user to focus the spatial scale and domain of the analysis and perform an appropriately tailored multivariate visualization of the corresponding data. All of these ideas and algorithms are instantiated in a deployed visualization and analysis tool called VAPOR, which is in routine use by scientists internationally. In data from a 1024x1024x1024 simulation of a forced turbulent flow, VAPOR allowed us to perform a visual data exploration of the flow properties at interactive speeds, leading to the discovery of novel scientific properties of the flow. This kind of intelligent, focused analysis/refinement approach will become even more important as computational science moves towards petascale applications.

Progressive Visualization-Driven Multivariate Feature Definition and Analysis.

K. Gruchalla. Ph.D. Thesis, University of Colorado at Boulder, 2009.

thesis bibtex code

One of the barriers to visualization-enabled scientific discovery is the difficulty in clearly and quantitatively articulating the meaning of a visualization, particularly in the exploration of relationships between multiple variables in large-scale data sets. This issue becomes more complicated in the visualization of three-dimensional turbulence, since geometry, topology, and statistics play complicated, intertwined roles in the definitions of the features of interest, making them difficult or impossible to precisely describe. This dissertation develops and evaluates a novel interactive multivariate volume visualization framework that allows features to be progressively isolated and defined using a combination of global and feature-local properties. I argue that a progressive and interactive multivariate feature-local approach is advantageous when investigating ill-defined features because it provides a physically meaningful, quantitatively rich environment within which to examine the sensitivity of the structure properties to the identification parameters. The efficacy of this approach is demonstrated in the analysis of vortical structures in Taylor-Green turbulence. Through this analysis, two distinct structure populations have been discovered in these data: structures with minimal and maximal local absolute helicity distributions. These populations cannot be distinguished via global distributions; however, they were readily identified by this approach, since their feature-local statistics are distinctive.

Immersive Examination of the Qualitative Structure of Biomolecules.

K. Gruchalla, M. Dubin, J. Marbach, E. Bradley.
In Proceedings of the International Workshop on Qualitative Reasoning about Physical Systems, 2008.

paper bibtex

We studied the added value in using immersive visualization as a molecular research tool. We present our results in the context of "embodied cognition", as a way to understand situations in which immersive virtual visualization may be particularly useful. PYMOL, a non-immersive application used by biochemistry researchers, was ported to an immersive virtual environment (IVE) to run on a four-PC cluster. Three research groups were invited to extend their current research on a molecule of interest to include an investigation of that molecule inside the IVE. The groups each had a similar experience of visualizing a feature of their molecule they had not previously appreciated from workstation viewing; large-scale spatial features, such as pockets and ridges, were readily identified when walking around the molecule displayed at human scale. We suggest that this added value arises because an IVE affords the opportunity to visualize the molecule using normal, everyday-world perceptual abilities that have been tuned and practiced from birth. This work also suggests that short sessions of IVE viewing can valuably augment extensive, non-IVE based visualizations.

Porting Legacy Applications to Immersive Virtual Environments: A Case Study.

K. Gruchalla, J. Marbach, M. Dubin.
In Proceedings of the International Conference on Computer Graphics Theory and Applications, 2007.

paper bibtex code

Immersive virtual environments are becoming increasingly common, driving the need to develop new software or adapt existing software to these environments. We discuss some of the issues and limitations of porting an existing molecular graphics system, PyMOL, into an immersive virtual environment. Presenting macromolecules inside an interactive immersive virtual environment may provide unique insights into molecular structure and improve the rational design of drugs that target a specific molecule. PyMOL was successfully extended to render molecular structures immersively; however, elements of the legacy interactive design did not scale well into three-dimensions. Achieving an interactive frame rate for large macromolecules was also an issue. The immersive system was developed and evaluated on both a shared-memory parallel machine and a commodity cluster.

Immersive Visualization of the Hurricane Isabel Dataset^*.

K. Gruchalla, J. Marbach.
IEEE Visualization 2004

paper video

In this paper, we describe an immersive prototype application, AtmosV, developed to interactively visualize the large multivariate atmospheric dataset provided by the IEEE Visualization 2004 Contest committee. The visualization approach is a combination of volume and polygonal rendering. The immersive application was developed and evaluated on both a shared-memory parallel machine and a commodity cluster. Using the cluster we were able to visualize multiple variables at interactive frame rates.

Immersive Well-Path Editing: Investigating the added value of immersion.

K. Gruchalla.
In Proceedings of IEEE Virtual Reality 2004.

paper bibtex video

The benefits of immersive visualization are primarily anecdotal; there have been few controlled user studies that have attempted to quantify the added value of immersion for problems requiring the manipulation of virtual objects. This research quantifies the added value of immersion for a real-world industrial problem: oil well-path planning. An experiment was designed to compare human performance between an immersive virtual environment (IVE) and a desktop workstation. This work presents the results of sixteen participants who planned the paths of four oil wells. Each participant planned two well-paths on a desktop workstation with a stereoscopic display and two well-paths in a CAVE-like IVE. Fifteen of the participants completed well-path editing tasks faster in the IVE than in the desktop environment. The increased speed was complimented by a statistically significant increase in correct solutions in the IVE. The results suggest that an IVE can allow for faster and more accurate problem solving in a complex three-dimensional domain.

Immersive Well-Path Planning: Investigating the added value of immersion.

K. Gruchalla. Master's Thesis, University of Colorado at Boulder, 2003.

thesis bibtex video

The benefits of immersive visualization are primarily anecdotal; there have been few controlled users studies that have attempted to quantify the added value of immersion for problems requiring the manipulation of virtual objects. This research quantifies the added value of immersion for a real-world industrial problem: oil well path planning. An experiment was designed to compare human performance between an immersive virtual environment (IVE) and a desktop workstation with stereoscopic display. This work consisted of building a cross-environment application, capable of visualizing and editing a planned well path within an existing oilfield, and conducting an user study on that application. This work presents the results of sixteen participants who planned the paths of four oil wells. Each participant planned two well paths on a desktop workstation with a stereoscopic display and two well paths in a CAVE-like IVE. Fifteen of the participants completed well path editing tasks faster in the IVE than in the desktop environment, which is statistically significant (p < 0.001). The increased speed in the IVE was complimented by an increase correct solutions. There was a statistically significant (p < 0.05) increase in correct solutions in the IVE. The results suggest that an IVE allows for faster and more accurate problem solving in a complex interactive three-dimensional domain.