By Grant Bunker, Chair and Professor, Physics
Physicists have long used computers in research. The modern stored program computer architecture was conceived and implemented by mathematician and physicist John von Neumann in 1945, inspired by Alan Turing’s theoretical ideas. In physics, computers are used to test theoretical models against observations, to analyze experimental data, and to serve as virtual experiments in themselves.
Initially “computers” were human individuals. Later, computations were done by carefully choreographed groups of humans operating mechanical calculators; later yet, by electronic computers that were based on vacuum tubes, then transistors, then CPUs; and nowadays by carefully choreographed clusters of CPUs and GPUs.
Another very different computational approach that was used by physicists employed analog computers, which (among other things) solved differential equations by devising electronic (or fluidic) circuits whose behavior was a mathematical analogue of the behavior of the physical system of interest.
Computation is an essential, pervasive, educational, and fun aspect of doing physics at Illinois Tech. We couldn’t function without it, and we plan to do much more.
Our undergraduate programs are unusual in requiring not just an introductory programming course in C/C++, which is offered by the computer science department, but also PHYS 240 Computational Science and PHYS 440 Computational Physics, which give practical, hands-on experience in numerical methods and programming. At professors’ discretion, computational methods also are integrated throughout courses. For example, students in PHYS 308/309 Classical Mechanics I and II do symbolic, numerical, and graphical computation using Mathematica.
The computational resources used in the department are varied, ranging in scale from tiny Arduinos and Raspberry Pi for mechatronics; micro-controllers for experimental data acquisition; desktop computers for data analysis and 3D stereo molecular graphics; high-powered multi-GPU machines and parallel clusters with hundreds of processors for number crunching; and some of the largest supercomputers available for modeling huge biomolecular complexes and novel materials.
Some of our faculty members have research programs that are focused on computation, and others use computation to model and simulate experimental data. Assistant Professor Jeff Wereszczynski calculates molecular dynamics of very large biomolecular systems at the atomic level to understand how they function, and to determine their statistical mechanical properties and energetics. One system of particular interest is chromatin, which in cells is the complex of DNA, RNA, and proteins that stabilizes the genome and dynamically controls gene expression.
Assistant Professor Pavel Snopok’s research also is primarily computational: modeling the complicated nonlinear dynamics of muon beams in accelerators. This research is key to the design of “muon cooling” technology, which is an essential part of Illinois Tech’s muon and neutrino physics research program that involves a diverse international group of collaborators at Illinois Tech, Fermilab, Oxford University, and other institutions.
Zack Sullivan, associate professor, is a theoretical particle physicist who uses computers to interpret data from the Large Hadron Collider. Associate Professor Liam Coffey is a theoretical physicist who computes solutions to complex integral equations of superconductivity (with John Zasadzinski, Paul and Suzi Schutt Endowed Chair in Science and professor of physics) and other systems in condensed matter physics. A number of us—Karoly Nemeth, adjunct professor of physics; Jeff Terry, professor of physics; Carlo Segre, Duchossois Leadership Professor of Physics; and I—routinely do elaborate computations at the molecular level to calculate X-ray spectra, to understand structure/function relationships of novel materials and biophysical systems, and to interpret experimental data.
[above] The interactions of the membrane-targeting motif of Protein kinase C (red) with a lipid bilayer (stick representation) as modeled with molecular dynamics simulations