General Information

Top of Page

Applied Mathematics

Headed by Roman Samulak and James Glimm of Stony Brook University.

James Glimm works on projects associated with quantum dots, short term weather forecasting, turbulent mixing and quantitative finance.

Roman Samulyak works on simulations associated with High Power Targets for Neutrino Factory / Muon Collider, New Smart Grid: Algorithms for Real Time Optimization of Power Systems , Development of Scalable Lagrangian Mesh and Particle Methods for Multiphase Flows, Evaluation of Plasma Liner Driven Magneto-Inertial Fusion via Advanced Computing, and  Mesoscale Models for Brittle Failure of Solids.

Research Projects

High Power Targets for Neutrino Factory / Muon Collider

Development of mathematical models and software for free surface / multiphase megnetohydrodynamics in the low magnetic Reynolds number approximation and their application to the simulation of high power liquid mercury targets interacting with proton pulses in strong magnetic fields. Simulations made predictions for the targetry experiment at CERN called MERIT and influenced its design. Future designs will increasingly rely on simulations. Two codes are being developed: front tracking code FronTier-MHD and MHD code based on Lagrangian particles. More information

New Smart Grid: Algorithms for Real Time Optimization of Power Systems

The main goal of this collaborative project with NYPA is the development and deployment of numerical algorithms and software to be used by the Smart Grid control equipment.  New software will dramatically improve the real-time control of the electric power grid by solving power control problems on orders of magnitude faster than what is currently available. Emerging Smart Grid technologies, including the deployment of telecommunication networks and new generation of sensors, such as Phase Measurement Units (PMUs), will provide large amounts of useful data for such control decisions.  This project develops mathematical algorithms and software to provide robust state solution for dynamic visibility of the power system behavior for the systems with more than ten thousand nodes.

Applied Mathematics: Development of Scalable Lagrangian Mesh and Particle Methods for Multiphase Flows

The goal of the project is to develop new and enhanced Lagrangian particle and moving mesh mathematical models and scalable algorithms extending and improving ideas of smooth particle hydrodynamics and front tracking. The developed methods will improve the mathematical rigor and accuracy of the current particle-based PDE methods and extend them to new systems. The software, that implements modern programming paradigms for multicore supercomputers and GPU clusters, will be used for a variety of fundamental science and applied problems.

Evaluation of Plasma Liner Driven Magneto-Inertial Fusion via Advanced Computing

In the plasma jet deriven magneto-inertial fusion concept, a plasma liner, formed by merging of a large number of radial, highly supersonic plasma jets, implodes on a plasma target, and compresses it to conditions of the fusion ignition. Our goal is to evaluate this method via high fidelity numerical simulations and to provide guidance for the Plasma Liner Experiment at Los Alamos. Recent results include verification of theoretical scaling laws, development of new physics models, and investigation of factors affecting the liner quality and fusion energy gain. Work is supported by the DOE Program in High Energy Density Laboratory Plasmas.

Highly Scalable Electromagnetic Solvers Coupled to Particles

The goal of the project is the development of highly scalable code for the simulation of electromagnetic fields coupled to particles. Using a hybrid parallelization method based on MPI with multithreads for multicore supercomputers such as BlueGene-Q, the software will be applicable for a variety of fundamental science and applied problems. In particular, it will be used by BNL scientists working on advanced laser wakefield acceleration methods.

Mesoscale Models for Brittle Failure of Solids

Development of new mass-conserving brittle fracture algorithms based on the energy minimization of elasto-plastic networks. The main research goals are to improve the method of Extended Finite Elements, in particular its ability to simulate complex fracture regimes and flows of disintegrated materials, develop highly scalable software, and apply it to problems relevant to Army Research Lab and DOE Nuclear Energy research.