Jefferson Lab
On Target (September 1998)
Jefferson Lab Home Search Contact JLab

    [ Previous Page | Table of Contents | Next Page ]


    Physics well done
    Part One: Mastering the menu (or how to get an experiment cooking!)

    by Jim Schultz

    Great chefs know there are two basic secrets to culinary success: preparation and execution. No chef would agree to cook a meal if its components and recipe were largely unknown or likely to change without warning. Yet that is the challenge faced by every nuclear physics researcher at Jefferson Lab probing matter at the limits of current understanding.

    Although investigators presume a certain mix of basic ingredients in their experimental "meals," the list is subject to radical change. Researchers sometimes discover that, in the cooking, or the actual experiment itself, the outcome is not a souffle, but instead the physics equivalent of a thick sauce.

    "You can have a theory that fits beautifully with existing data," says physics theorist Warren Buck, a Jefferson Lab user and director of Hampton University's Nuclear and High-Energy Physics Research Center. "So, after reproducing existing data, you predict the behavior of yet another observable (something that can be measured experimentally). If that behavior holds, it puts the entire puzzle together. But then an experimentalist comes along who says, 'I want to prove or disprove your theory.' It's at that point the proposal procedure begins."

    No fine meal can take place without an imaginative, well-thought-out menu and the dedicated support of a trained staff. Likewise, for an investigator to arrange the physics equivalent of a banquet at the Lab-an experiment that can take years to arrange and weeks, or months, to conduct-certain key steps must be taken. Not the least of these is settling on what outstanding question in nuclear physics should, and can, be addressed.

    A Complete Bill of Goods

    In arranging an experiment at the Lab, the first step is a response to the Laboratory's call for proposals. Proposal calls occur twice a year, usually six weeks prior to a meeting of the Lab's Program Advisory Committee, or PAC, which judges proposals on scientific merit, technical feasibility and manpower requirements. Each proposal is given a few "principal readers" from the Committee. The proposal readers contact the experiment spokesperson prior to the formal PAC meeting to discuss any outstanding issues or answer questions about the proposal.

    The PAC makes a recommendation for the disposition of every proposal reviewed: approval, conditional approval, deferral or rejection.

    The proposal must answer major questions of: the nature of the experiment; its uniqueness; the ways the experiment will advance physics understanding; identification of required resources; the amount of accelerator time or "beam days" needed; and the composition of the collaborative group submitting the proposal and contact information on the individual(s) leading the collaboration.

    "Experimenters have to go through and justify all the science and the reasons attention should be given to the proposal," says user liaison manager Karen Hokansson. "Once the proposal has been approved, the experimenters have to make preparations; they may need to build new equipment. Typically, it takes 1-6 years to get an experiment done. The proposal is just the first step."

    In the case of approved experiments, the experiment must be run or scheduled within three years from its date of approval. Otherwise, a new proposal must be written, defending the original research goals.

    If an experiment is conditionally approved, the proposal authors must update their submission to address the issues raised and undergo additional PAC review. Unresolved technical matters must also be laid to rest. Deferred proposals are those the PAC believes have scientific merit, but contain difficulties that must be overcome for the Committee to issue approval.

    As of July, 197 proposals had been reviewed since the Laboratory began accepting proposals. Of those, 105 experiments have been approved by the PAC, with an additional 15 conditionally approved. 715 scientists and researchers from 138 institutions in 21 countries are involved with the collaborations behind these proposals.

    Distilling the Recipe

    The difference between a good meal and a great one lies in execution. The individuals or groups conducting experiments must ensure the right mix of theoreticians, experimentalists, technicians, equipment, supplies and funding. Typically, each experiment team appoints one or more spokespeople, whose job it is to champion and lead the physics effort including presenting the experiment to the PAC, leading collaborative efforts and interfacing with Jefferson Lab for execution.

    And there must be the right tools. Among those with long experience in nuclear physics, Jefferson Lab's accelerator is among the world's best for cutting-edge physics research. "We are talking about extremely small things," theorist Buck explains. "You can't touch them with your finger. You can't see them with the naked eye. You can't taste them... What's so nice is that after decades of wanting to look at certain observables, you finally have a machine that enables you to see at a very fine resolution-the finest available. This is the chance to do new physics, to see if we can verify or improve theories with critical experiments. Alternately, we may need new theories to explain the experimental results.

    "It's when you get down to the end, when you collect data and begin to analyze it-that's very exciting. That's when you find out what you know and what you don't. That's when you begin to learn," Buck said.

    Pushing back the boundaries of knowledge is difficult, requiring years, sometimes decades, of unremitting work. At Jefferson Lab, that exertion is often distilled into several weeks of intense effort. It is during this time that an experimental team can discover Nature's own particular, hidden recipes.

    Coming next month: Up and Cooking: Part Two of Physics Well Done.

    maintained by webmaster@jlab.org