Dorfan Today: Case Closed... Successfully
I want to congratulate the Accelerator Division, BABAR, MFD, the SLAC shops, and the many others who contributed to getting PEP-II running again at high luminosity. This team worked incredibly long and hard to discover the origins of an extremely elusive problem in an inaccessible region of the collider, was able to replace the damaged parts with a carefully orchestrated set of surgical steps, and then got everything up and running smoothly again within a matter of days.
Resolving the problem involved some considerable detective work. PEP-II had been delivering record amounts of luminosity to the BaBar detector last October but, after a planned downtime, an unexpected problem surfaced in mid-December and quickly became chronic. When the current in the lower-energy ring (LER) topped roughly 1600 milliamps, just over half of the current needed to produce October's peak luminosity, short intense vacuum pressure spikes lasting a few seconds began to occur near the detector with increasing frequency, many resulting in a beam abort due to background conditions. As a result, PEP-II had to limp along with only half the desired current, and thus only half the luminosity. While still well above the luminosity the collider was designed to deliver, this fell short of what Babar needs to pin down new physics and other measurements of rare events.
The accelerator team devised and conducted a wide variety of experiments to isolate the source of the problem or eliminate possible causes. Suspect vacuum components were removed or replaced in the incoming LER to eliminate these as candidates. Controlled amounts of gas were introduced at various locations in the vacuum system near the detector to find a pattern of gas pressure readings in pumps and vacuum gauges that reproduced the actual pressure profile of the problem bursts.
Babar physicists analyzed the pressure pattern and timing for burst events, and studied measurements from the detector radiation protection system, to triangulate the source of the vacuum bursts. This allowed them to pinpoint the source to a region at the front end of the detector, about 2.5 meters from the point where the electrons and positrons collide. This area is normally very difficult to access. As a result the accelerator team turned to using a surgeons' trick to actually see what was going wrong there: a camera at the end of an optical fiber line (called a borescope). But the "smoking gun" was hidden from view.
Having eliminated all other sources, the decision was taken in early March to replace the two vacuum parts at the front of the detector. The completion of a $0.5 million vacuum chamber was accelerated. New electrical seals were designed and fabricated to address a known weakness of the installed system. In late March, with all these preparations in place, crews spent two days removing utilities, Babar detector cables, and magnet elements in the very constricted space at the front of the detector. When they reached the problem spot, it was immediately clear: electrical burn marks and lost material in two ceramic tiles in the center of a crucial vacuum system bellows near the detector. Electrical arcing had chipped away cubic millimeters of ceramic material over time. It would only take cubic microns of ceramic material, vaporized in the passing positron beam, to create the kind of observed burst of gas that has limited LER current.
Knowing in advance that the bellows was the likely source of the problem, a clever strategy was devised for dealing with replacement. Only two of the bellows parts exist, both already installed at the two ends of the detector with the expectation that the forward one likely was the culprit. Each part contains two rows of 46 tiles that are designed to absorb radio frequency electromagnetic energy that gets trapped in this complicated region of the vacuum pipe. To get around this problem, crews performed "surgery," removing the bellows at the rear end of the detector and moving it to the front end instead. To replace the commandeered part, they used a spare older version with only one row of tiles, which should work in the rear of the detector where the energy deposited is only half as large.
To top off this story, anticipating the type of damage that actually turned out to be seen, the electrical seals were swapped and older vacuum bellows were also replaced with parts of the new design to prevent future arcing.
After restoring the vacuum system and detector week ago Saturday, PEP-II was already able to inject 2000 milliamps into the lower-energy ring during Monday's owl shift, well above the threshold where vacuum bursts were previously triggered. An extremely challenging detective case successfully closed.
I am delighted that BaBar is back on its high luminosity track, and looking forward to the wonderful results this new set of data is sure to bring.
Jonathan Dorfan SLAC Today, April 10, 2006