Eric Weeks - personal pages - past research Past Projects This includes research done at Harvard University, the University of Pennsylvania, and the University of Texas at Austin. Click here to see publications list. Click here for current research.

erweeks / emory.edu

Colloidal Glasses Small colloidal particles, when densely packed, can act like a glass. I use microscopy to study the motion of particles as the packing is increased towards the glassy state. This reveals possible microscopic origins to the glass transition. (Harvard/U Penn)

Microrheology How can you study the motion of small particles to determine the viscosity of a fluid? That's easy, if you know the Stokes-Einstein formula. How can you use small particles to determine the viscoelastic moduli of a complex fluid? That's what this web page will tell you about. There are some nice animated GIF movies showing Brownian motion. (University of Pennsylvania/Harvard)

Atmospheric Blocking Certain global weather patterns can lead to phenomena such as snowfall in Texas. A long-standing conjecture is that these patterns are directly caused by mountain ranges: in particular, the Rockies and the Alps. We did an experiment simulating the global jet stream, with two mountain ranges, and found strikingly similar flow patterns. (University of Texas at Austin)

Anomalous Diffusion and Levy Flights Levy flights are a strange sort of mathematical random walk, for which it is rare to take long jumps, but not as rare as you'd expect. This has some interesting implications, although for the most part it's just a mathematical curiousity. However, we observed these random walks in experiments, providing some sense of what fluid flows might lead to Levy flights. I did some theoretical work and also some experiments. (University of Texas at Austin)

Controlling chaos with neural networks and genetic algorithms. I heard several talks claiming that neural networks could solve almost any problem, no matter how intractable. I took a class on neural networks to learn if this was true. While controlling chaos isn't an intractable problem, it turns out that neural networks do indeed have some nice properties compared to conventional methods for controlling chaos. You can download the program that we used, with some slight documentation. (University of Texas at Austin)

Fun with chaotic time series analysis This is a side project analyzing time series, both standard ones and also experimental data. I haven't added to this since 1997, although I may add more later. (University of Texas at Austin)

If you have any questions, send me email, address below.