Getting Your Quarks in a Row

"Getting Your Quarks in a Row" by Brian Hayes. American Scientist, November-December, 2008.

This article deals with the challenges of using QCD (quantum chromodynamics) instead of QED (quantum electrodynamics). In a nutshell, QED is about understanding the interactions between quarks as interactions along axes of "charge" (I use the term charge here loosely, metaphorically, and not altogether correctly). QCD is a way of understanding quantum behavior in which the mechanism by which quarks interact is through the emission of various photons of different wavelengths.

QED has been more useful for making predictions in the past, due in part to the fact that it's easier to handle mathemetically. Basically, the Feynman sum-over-paths method (a way of "averaging" all the different possible states of the subatomic particles to determine where they are "likely" to be) works naturally for QED because the infinitely long series it produces converge, and so can be treated like perturbation models. QCD is more problematic, because the series it creates do not converge and are made more complex by the fact that gluons can "emit" other gluons, not just photons.

Lattice computing methods are used to calculate these things. The lattics is typically 4D (4th dimension is time). Every conceivable path via which things can "move" (even just in time) is mapped out on a big grid, and the paths are "summed." As the size of the grid squares over which the paths are described shrinks to 0, the lattice-based calculations should be useful predictions. However, the computing power for these models is still massive (and higher for QCD than for QED).

New techniques in QCD as well as new computing power are starting to make the QCD predictive models more feasible. These methods and computing systems will be more useful once the new Large Hadron Collider is available for experimentation.