I have ideas for various projects in theoretical solid-state physics. They involve crystal structures, interatomic potentials, computer simulations, geometry (mathematical physics!), lattice vibrations, and/or solid state chemistry. They are related to my Quasicrystals, biological physics, or frusrated antiferromagnets research (see "Research supported by the DOE", below). These projects are challenging enough to be quick starter projects for grads, too.
Undergrad projects don't require quantum mechanics, so they can be started after you complete the introductory sequence. (Usually begin with a reading course.) My projects do demand good organization and motivation on the part of the researchers. Firstly, I usually don't have a grad student on your subject to help you, since I'm spread between too many topics. Secondly, this is not make-work: all my undergrad projects are things on my list to get done, but that I don't have a grad student working on (or the grad student is on a project entailing more advanced knowledge). It means I'm more invested in the projects, but you may also feel pressure sometimes. Click here for two undergrad project descriptions.
Note -- I do not have funds to pay students from other countries for summer internships. Undergrad research is (1) through one of the REU ("research experience for undergraduates") programs at Cornell, (2) for Cornell course credit, or (3) occasionally, paid summer assistantships for students I already know.
As with other professors, the way to get started is to do a research project in the spring semester (if you're on a fellowship) or a summer GRA after your first year. It's valuable to overlap with, and learn from, the preceding student -- but often there won't be funds to support you fulltime. I rarely support students during the academic year till their third year.
Topics are a mix of analytic and computational -- depending on student's aptitudes, but I insist that primarily computational students have an analytic projects and vice versa. I'm not a professional computational physicist (i.e. I don't focus on parallel processing and optimized algorithms); rather I treat simulations as artificial experiments, which inspire much more interesting analytic questions than you concoct in a vacuum. My taste is to open up a new model, perhaps to invent new (simple) techniques for it, rather than to apply well understood techniques or to do further work on well-studied problems in competition with many other groups.
Comments on my research areas: