University of Vermont

Vermont Advanced Computing Core

Collaborators

Highlighted Researchers

Adrian Del Maestro Adrian Del Maestro, PhD
Assistant Professor
Department of Physics

To study the exotic, emergent properties of tens of thousands of strongly interacting atoms at low temperature, Prof. Del Maestro's research group exploits the massive parallelization possible on the VACC's three thousand compute cores.

Professor Del Maestro's interests lie within the field of condensed matter physics, which can be broadly generalized as the study of the phases of matter (solids, liquids, ...), and the phase transitions that connect them. Research in condensed matter physics is focused on systems with a large number of constituents (e.g. atoms or electrons) that strongly interact amongst themselves. Even if the basic interactions between the components are understood, their collective behavior can lead to new and interesting emergent phenomena, where the behavior at macroscopic length scales is not seen at the microscopic level. One of the most interesting examples is superconductivity, characterized by the conduction of electricity with zero resistance and the expulsion of any interior magnetic fields below some critical temperature.

For information about Del Maestro's research and more, please see Adrian Del Maestro's homepage.

Return to top

 

Josh BongardJosh Bongard, PhD
Associate Professor
Dept. of Computer Science, Complex Systems Center

The calculations underpinning the full evolutionary transformation of Bongard's "robust" robot required 5000 experiments that took eight hours each on our super computer. On an ordinary PC, it would have required 100 years. The project as a whole took 3 months on the VACC.

If you had to choose the word that best describes the robots Josh Bongard has been working with lately, "robotic" would not be it. "Adaptive" is more like it. Bongard designs robots that can change or evolve — in body and in mind. "Mind," that is, to the extent that a robot has a one, or a brain, in the form of the program that's driving it.

A few years ago he got some attention for his work on a robot dubbed "Starfish" that taught itself to walk. Bongard's recent research, which has also received national notice, looks at a robot with a flexible spine and four legs. Robots with this physique are given a simple task (the scientific term is "phototaxis"): to move from Point A to Point B (a light source) as quickly as possible, without falling.

Rather than providing the robots with a program that prescribes how to walk — as a conventional robotics engineer would do — Bongard feeds them an algororithm that lets them know all the possible ways they might move limbs and spine and then lets them reject the thousands of alternatives that don't work in favor of the few that do. Read the complete article

For information about Bongard's research and more, please see Josh Bongard's homepage.

Return to top

 

Chris Danforth Christopher Danforth, PhD
Associate Professor
Dept. of Mathematics & Statistics, Complex Systems Center

Weather and climate model forecast errors grow with time as a result of chaotic instabilities that amplify uncertainty. Danforth's team is using the cluster to analyze the statistics of millions of past forecasts in order to reduce this growth and improve predictions of the Earth's atmosphere.

Weather and climate model forecast errors grow with time as a result of atmospheric instabilities that amplify uncertainties in the initial conditions, and deficiencies in the numerical model which introduce errors during integration. Weather predictions have benefited dramatically from advances in the science of data assimilation (DA), essentially the attempt to synchronize a computational solution with physical observations of the modeled system by a clever combination of the uncertainties associated with each. However, climate models suffer from errors related to the difficulties associated with resolving multiple temporal scales, a problem DA methods have yet to appropriately address.

The present research project aims to develop and test a new advanced mathematical method for improving predictions made by climate models, focusing on DA and the coupling of slow- and fast-scale phenomena. The techniques are being developed using simple (N ~ 10 variable ODE) and sophisticated (N ~ 10^6 variable CFD) models of a thermal convection loop, a toy climate analogous to the convection system studied by Lorenz, to predict regime changes (flow reversals) in a laboratory fluid dynamics experiment.

We are also testing methods to improve forecasts on the National Center for Environmental Prediction's state-of-the-art Global Forecast System (GFS, N ~ 10^9) to improve short-range forecasts and suggest improvements to the model's physical parameterizations.

For information about his research, please see Chris Danforth's homepage.

Return to top

 

Yves Dubief Yves Dubief, PhD
Associate Professor
School of Engineering, Mechanical Engineering

We use massively parallel computing to access physical phenomena that cannot be captured by current experimental techniques. The VACC is our test bed for the development of multiscale, predictive algorithms with outcomes ranging from the study of new approaches to knee lubrication and its relation to osteoarthritis, to insights into the complexity of space reentry ablation mechanisms.

Using an array of algorithms, we investigate physical, biophysical and engineering problems that involve a wide range of scales. The current research projects include:

  • Turbulence Control: The core expertise of our group is the simulation of turbulent flows of Newtonian fluids (water, air) and non-Newtonian fluids (polymer solutions). The objective are (i) to use controlled manipulation of turbulent flows to further the understanding and predictive modeling of turbulence and (ii) to understand and develop new strategies for turbulence control. Our current work focus on the fundamental study of the Kolmogorov flow, dynamic wall control and polymer drag reduction.
  • Flow-Surface Interactions: We are developing predictive simulation algorithms for the ablation of a surface by a fluid flow. The interplay between an ablated surface and an ablative flow is complex, multi-physics and, for fast erosion rate, non-equilibrium and therefore very challenging to simulate. The simulation and prediction of ablation is critical to the safety of reentry space vehicles, yet current predictive algorithms do not solve capture accurately the coupling between flow and receding surface. Using low-temperature ablation, we seek to establish the fundamentals of ablative turbulent flow/ablated surface interactions. In the near future, we will translate the acquired knowledge and simulation techniques to high temperature ablation with the simulation of UVM Inductively Coupled Plasma torch, which reproduce realistic conditions of earth and other planets' atmospheric reentry.
  • Biophysical Fluids: Two bio-fluids have captured our interest: blood and synovial fluid. The effect of flow on blood coagulation is still poorly understood, yet its understanding is critical to the development of drugs than may trigger and prevent blood coagulation. We are developing a range of approaches to address specific mechanisms of activations of proteins in the coagulation cascade. Synovial fluid is the lubricant in our major articular joints (hip, knee, shoulder...). Using molecular dynamics, we are testing the hypothesis that the amazing lubrication in our joints is driven by intermolecular forces. The anticipated outcomes are new insights of the synovial fluid/cartilage interactions with applications to osteoarthritis and orthopedic prostheses, and the development of a new approach to lubrication in mechanical systems.

For information about his research, please see Yves Dubief's homepage.

Return to top

 

More Collaborators

Faculty Researchers
  • Brian Beckage
    Brian Beckage
    Associate Professor
    Plant Biology
  • Peter Dodds
    Peter Dodds
    Associate Professor
    Dept. of Mathematics and Statistics, Complex Systems Center
  • Julie Dumas
    Julie Dumas
    Research Assistant Professor
    Clinical Neuroscience Research Unit, Brain Imaging Program
  • Maggie Eppstein
    Margaret Eppstein
    Chair
    Dept. of Computer Science
    Associate Professor
    Dept. of Computer Science, Complex Systems Center
  • Paul Hines
    Paul Hines
    Assistant Professor
    College of Engineering, Complex Systems Center
  • Darren L. Hitt
    Darren L. Hitt
    Associate Professor
    College of Engineering and Mathematical Sciences
  • Jarlath O'Neil-Dunne
    Jarlath O'Neil-Dunne
    Researcher/Analyst, Geospatial Specialist
    Spatial Analysis Lab
  • Frederic Sansoz
    Frederic Sansoz
    Associate Professor
    College of Engineering and Mathematical Sciences
  • Neil Sarkar
    Indra Neil Sarkar
    Director of Biomedical Informatics
    Center for Clinical and Translational Science
    Assistant Professor
    Dept. of Computer Science, Dept. of Microbiology and Molecular Genetics
  • Jim Vincent
    Jim Vincent
    Director
    Bioinformatics Core for the Vermont Genetics Network (VGN)
    Research Assistant Professor
    Dept. of Biology
  • Rory Waterman
    Rory Waterman
    Assistant Professor
    Department of Chemistry
  • Beverley Wemple
    Beverley Wemple
    Associate Professor
    Geography, Geology, and Environmental Sciences