On July 1, 1994, Jefferson Lab's accelerator delivered an electron beam into one of its experimental halls for the first time. It was just 10 years after the lab's founding and the construction of its first-of-a-kind accelerator.
Initial plans called for a machine that could "accelerate" a more-or-less continuous beam of electrons by stuffing each electron with up to four billion electron-volts of energy: 400 billion times the electron's natural energy of one electron-volt.
The electrons in the beam were expected to travel at about the same pace in a tight path through the machine continuously (a duty factor of 100 percent). With each "pass" through the machine, the electrons were to gain more energy. When the electrons reached the appropriate energy, they were to be directed into a target in one, two or all three of the laboratory’s experimental halls.
Construction on what came to be called the Continuous Electron Beam Accelerator Facility began in February of 1987 and wrapped up in December of 1993. Within months, electron beams were circling through CEBAF, recalls Andrew Hutton, associate director of the Accelerator Division.
"We developed a sequential way of commissioning the machine. Single pass into Hall C, do an experiment. Three-pass into Hall C, do another experiment. Five-pass into Hall C. Then start looking at separation, putting it into the other halls," says Hutton.
The machine reached full-design energy, 4 GeV, in November 1995. Shortly thereafter, CEBAF began to meet all of its other design criteria, one checkbox at a time.
"That actually kept our users happy for not very long. They wanted more," Hutton says.
Nor did it satisfy the accelerator builders, who wanted to know just what the new machine was truly capable of.
"There's one difference between CEBAF and many of the other machines: We have, over all of the time we have been running, been pushing the parameters - trying to do something new, trying to invent a new kind of physics, trying to improve something," Hutton says.
Accelerator tunnel (South Linac) construction proceeds in January 1990.
The first production cryomodule was installed in CEBAF in September 1990.
The CEBAF accelerator (1995).
For instance, while the machine was originally designed to pack up to 4 GeV of energy into each electron, the operators realized they could coax more energy out of the machine.
"Because the machine was operating so well, we were able to increase the energy to 5 GeV, and then we pushed it to 5.5. The increased energy provided increased physics reach," Hutton says. Eventually, the machine delivered an electron beam of 6 GeV.
|2 • 10-9 m
|2 • 10-9 m
|2.5 • 10-5
|2.5 • 10-5
In addition to delivering more energetic electrons, the accelerator managed to deliver more of them, as well. The number of electrons delivered in a second is measured in Amps. CEBAF was designed to deliver 200 microamps, corresponding to 2½ million electrons per bunch.
"The 200 microamps was assumed to be 100 into each hall," Hutton says, pointing out that the accelerator has actually managed to deliver up to 140 microamps into a single hall. "To produce 140 microamps to one hall, we actually cheated."
Ordinarily, the electron beam begins with pulses from one laser for each hall that is taking beam. The bunches are then combined into an almost continuous electron beam in the accelerator. To surpass the 100 microAmp-per-hall limit for an electron beam driven by a single laser, operators combined bunches from different lasers and delivered them to one hall. Using this technique, the full CEBAF design current of 200 microAmps could be delivered to either Hall A or Hall C.
Accelerator staff also developed techniques to send extremely small currents to Hall B, as low as 80 picoAmps, corresponding to an average of a single electron per bunch.
Another parameter of the machine concerns how the bunches of electrons are accelerated. Electrons "surf" radiofrequency waves that are pumped into accelerating structures called cavities.
"Each one of our cavities in CEBAF has its own source of radiofrequency power," Hutton says.
Powering each cavity separately allows machine operators to tune the waves poured into each cavity so that all the waves are in sync. Each bunch of electrons can essentially surf on the crest of one continuous wave all the way through the machine.
"If you can do that and maintain it, the front of the bunch and the back of the bunch actually get accelerated the same, and that minimizes the differences in energy across the bunch," Hutton explains.
Experimenters placed other demands on the accelerator when they began asking for the beam to be polarized - to have the electron spins aligned in a single direction. This also was outside the planned specifications of the machine.
The first experiment to use polarized electrons received 35 percent polarization, with about 35 out of every 100 electrons spinning in one direction.
Accelerator physicists, engineers, operators, nuclear physicists and others all worked together over the years to improve this percentage dramatically. Now the machine can produce polarization topping 85 percent.
"Doubling the polarization means you do the experiment in one quarter of the time. That's a huge impact," Hutton explains.
Along with more than doubling the polarization, this group also helped the accelerator achieve a polarized beam with the consistency required for high-precision parity experiments.
"Parity experiments consist of saying: I'm going to flip the polarization and I'm going to scatter off my target, and I'm going to compare the two states. But I don't want the initial position of the two states, or the angle or the intensity or the polarization to be different; they're supposed to be identical," Hutton says.
All In 25 Years of "A Day's Work"
Hutton credits many people for making the accelerator the success it has been throughout its history. He is also looking forward to meeting new challenges.
"Being able to evolve means that the experiments being done now are testing theories that no one knew we could test. I am confident that that will continue in the 12 GeV era."
Author's Note: Thanks to Joe Grames, whose story ideas led to this article.