Materials Simulations
Physics 681, Spring 2000
James Sethna
https://sethna.lassp.cornell.edu/Teaching/Simulations/
Monday, Thursday 1:30-4:30, Theory Center
Collaboratory, Rhodes 453
This class is a graduate computer laboratory, focusing on the next generation
of tools for computation, simulation, and research in the science and dynamics
of materials. The course will be pitched at a high level of computational
sophistication.
Scientific Topics
- Molecular Dynamics:
Direct Simulations of Atomic Motion.
Verlet, symplectic algorithms, neighbor lists, boundary conditions;
melting, plastic flow and fracture.
- Lattice Monte Carlo:
Ising
and
lattice-gas
models. Metropolis, heat-bath, Kawasaki, continuous time, and cluster-flip dynamics:
Correlation functions, nucleation, coarsening, and scaling behavior.
- Finite Element Methods:
Crack growth simulations.
- Cellular Automata:
Avalanches and Noise.
Time and space-efficient algorithms for the dynamics of disordered systems:
depinning transitions, self-organized and plain-old criticality,
- Nonlinear PDEs:
Pattern Formation (ripples) and Dendritic Growth (snowflakes).
Spectral methods and phase-field methods: linear stability analysis, wavelength and
pattern selection.
Computational Topics
- Languages.
We will use C++ for implementing the core algorithms, and
Python for
higher-level interpreted interactive analysis. We make extensive use of
design patterns.
- Visualization and GUI Interfaces. A strong emphasis on visual,
interactive control of the simulations will run throughout the course.
- Efficiency. We will introduce the use of BLAS, LAPack,
and other fine-tuned packages for optimized computations.
- Data Structures. We use the standard template libraries
for vector and queue structures to develop flexible, efficient algorithms.
Last modified: December 7, 1999
James P. Sethna,
sethna@lassp.cornell.edu.
Statistical Mechanics: Entropy, Order Parameters, and Complexity,
now available at
Oxford University Press
(USA,
Europe).
