Pop Goes the Photomultiplier Tube

How do scientists make glass stronger? Break it.

Brookhaven Lab physicists and engineers take this hands-on approach a step further. In order to strengthen the design of glass bulbs known as photomultiplier tubes, the researchers submerge the devices in 500,000 gallons of pressurized water, punch a small hole through their sides, and watch as the glass cracks, crunches, and, just milliseconds later, implodes (see videos below).

The implosion of a photomultiplier tube in a tank at BNL, as seen through combined high-speed camera images. Time scale: 6,000 frames per second.

Fifty thousand of these tubes are planned to coat each of two massive detectors in the proposed Long Baseline Neutrino Experiment (LBNE). LBNE will send the world’s most intense beam of neutrinos — extremely elusive and lightweight elementary particles — about 800 miles underground from Fermi National Accelerator Laboratory in Illinois to a mine in South Dakota.

i-2206ddfcd87c6c73e639aadcb1eb06f8-D0290211_sm-thumb-250x376-61155.jpg

A photomultiplier tube

Brookhaven is leading the development of the experiment’s two water Cherenkov detectors, cylindrical tanks large enough to each hold a 20-story building or nearly 37 million gallons of ultra-pure water. When a neutrino interacts with these water molecules, it will create light that gets detected by the array of tubes coating the detector’s walls.

The problem is that the tubes, which are shaped like enlarged, flattened light bulbs, are expensive and somewhat fragile. In a similar detector made for a Japanese experiment, the implosion of just one faulty phototube caused a chain reaction that destroyed almost 7,000 of its 11,000 tubes, bringing the experiment to a halt.

By studying the tubes’ demise, BNL researchers hope LBNE will avoid this fate.

The group started their investigations with an item much smaller, and cheaper, than a photomultiplier tube — a light bulb. In a 60-gallon, pressurized tank at Brookhaven, the researchers used a high-speed camera to watch the bulb implode.

Combined high-speed camera images of an imploding light bulb. Time scale: 3,800 frames per second.

They then moved on to the real thing, first watching the implosion of photomultiplier tubes in house, then switching to a much larger, mothballed Navy facility in Rhode Island.

At the Naval Undersea Warfare Center (NUWC), the researchers reactivated a 17-year-old torpedo test bed called the Propulsion Noise Test System.

Within the facility’s 50-foot sphere, the photomultiplier tubes are submerged in water pressurized to 88 pounds per square inch (psi) — the pressure that will exist in the LBNE detectors. A small hole is knocked through the glass bulb with a hydraulic metal poker, breaking its vacuum, and the resulting implosion and shock wave is recorded by high-speed cameras and sensors.

Combined high-speed camera images of a photomultiplier tube implosion in the Naval Undersea Warfare Center’s Propulsion Noise Test System. Time scale: 6,000 frames per second.

At the end of the implosion, which takes just 4.5 milliseconds, the researchers found that the pressure of the water closest to the tube soars to 800 psi, a 10-fold increase. Although intense, the shock wave starts to weaken after just a few microseconds. Is it powerful enough to cause a wave of implosions throughout the detector? That’s what the team hopes to find out.

PMT test.jpg

BNL and Naval Undersea Warfare Center researchers monitor the shock wave produced during one of the photomultiplier tube implosions at the Propulsion Noise Test System in Rhode Island.

The researchers are now exploring two ways to prevent a catastrophic implosion in the detector: 1). Make the glass itself strong enough to withstand the impact of a shock wave, and 2). design phototube enclosures that, in the case of an implosion, deflect or absorb the shock wave, protecting its intact neighbors.

Once they have a new design in hand, the researchers will return to NUWC to test an array of the beefed-up tubes in the tank. The current LBNE schedule calls for a final design of the photomultiplier tube array by 2014.

    Current ye@r *