Skip navigation

Gravity Probe B

Testing Einstein's Universe

The GP-B Cryopump: Sintered Titanium "Flypaper"

In August 2004, towards the end of the Initialization and Orbit Checkout (IOC) phase of the mission, the four gyros were spun up, one-by-one, to their final science spin rates, averaging 70 Hz (4,200 rpm). The spin-up of each gyro was accomplished by injecting a stream of ultra-pure helium gas, moving at the speed of sound, into a channel in the gyro's housing and blowing the helium over the gyro rotor for several hours. An exhaust system vented much of this spin-up gas out into space, but a significant amount of helium gas still remained in the probe. In order to maximize the performance of the gyros and maintain their extraordinarily low characteristic spin-down periods (time required to slow down to ~37% of their initial rotation speeds) of greater than 10,000 years, it was essential to draw as many of the remaining helium molecules as possible away from the gyro rotors and housings. This is the main purpose of the cryopump.

Typically in a vacuum system, a pump draws gas from inside the system and forces it to the outside. The GP-B cryopump is not a “pump” in this traditional sense. Rather, it is more of a scavenger or collector that adsorbs gas molecules--that is, it collects and holds gas molecules in a thin film on its surface. It works by maintaining a large surface area at a temperature very near absolute zero. As long as the pump surface is kept at this cryogenic temperature, it will attract and adsorb stray molecules of any gas floating around in the probe. However, if the surface is heated up by just 5-6 kelvins, it will release the adsorbed gas molecules.

Designing a cryopump that would fit inside the probe and meet all the constraints required by the GP-B experiment proved to be yet another formidable engineering challenge for lead probe designer, Gary Reynolds of Lockheed Martin, GP-B Co-Principal Investigator John Turneaure of Stanford and their collaborating teams. The main constraints for the cryopump included it having to:

  • Sit above the telescope inside the probe, and thus it had to have a 6” round opening through its center to allow light from the guide star to pass down into the telescope.
  • Have a huge surface area.
  • Be made of ultra-clean material that would not introduce any particulate matter into the probe chamber around the gyros.
  • Be non-magnetic; it could not impart any magnetic field that could affect the spin axis orientation of the gyros.
  • Be stable and inert at cryogenic temperatures.

Reynolds et. al. quickly realized that the cryopump for the GP-B probe needed to have a shape and structure something like the cylindrical air filters used in older model automobiles. These filters are shaped like doughnuts, with a large hole in the middle. The body of the doughnut is composed of a long strip of filtering material folded accordion-style and wrapped radially around its circumference. This compresses a very large surface area into a very small package. An automobile air filter is a closed system, in which air is drawn through the filtering material to clean it. The cryopump designed by Reynolds et. al. only resembles an automobile air filter in shape. It is not a closed filtering system, and helium gas is not drawn or forced through it. Rather, it is a passive device with a large surface area for trapping stray helium molecules bouncing around inside the probe. Instead of a folded strip of filter material, the cryopump has a series of individual blades or fins, stacked radially next to each other around its circumference. Furthermore, the cryopump blades could not be made of traditional filter materials such as activated charcoal because such materials would impart particulate matter into the probe. The question was: “What non-magnetic material could be used for the cryopump blades, and how could the surface area of each blade be increased without introducing particulate matter into the probe?”

The solution was sintered titanium. Sintering is a process of powdering a material into tiny granules and then heating the granules until they coalesce into a solid or porous mass. The Mott Corporation in Connecticut, which specializes in manufacturing porous metals, had already produced sintered titanium filters used to ensure the purity of the helium gas entering the gyro housing spin-up channels. Thus, Reynolds and Turneaure once again turned to the Mott Corporation to manufacture sintered titanium blades for the cryopump.

The titanium for the blades was produced in very tiny balls--like microscopic cake sprinkles--which were then sintered into flat, thin porous sheets, having the texture of a very fine sieve. The individual blades were then cut out from these sheets. To ensure that the blades and other parts of the cryopump were completely free of any residue, they were subjected to a precision cleaning process, which involves blasting them with high pressure Freon gas and collecting any residue on a sheet of filter paper marked with grid lines. After each cleaning, sample squares of the filter paper were placed under a microscope to count trapped particles of various sizes. These particle counts were then compared with a standard, and if the count exceeded the standard, the cleaning process was repeated.

In all, the cryopump has 360 blades, each about the size of a tongue depressor, welded 1/16” apart around its circumference. In aggregate, these blades have a huge surface area of 230 square meters (2,475 square feet--the size of a 4-bedroom house!). Each blade is shaped like a bow tie, 6” long and 3/4” wide at the top and bottom, but narrower in the middle to accommodate a metal ring or “waistband” surrounding the cryopump. The ring contains three brackets, attached to short arms for mounting the cryopump to the inside of the probe. In addition to providing mechanical support, the waistband ring contains heaters that can be turned on to warm up the blades, freeing the helium molecules adsorbed to them. The ring and mounting arms also serve to thermally isolate the cryopump from the probe wall. The photos and drawings to the right show the GP-B cryopump.

The three mounting brackets secure the cryopump to the probe wall, above the telescope, in the coldest part of the probe. During the gyro spin-up process, the excess helium gas that does not escape through the spin-up exhaust channels bounces around inside the probe and gets attracted to the extremely cold cryopump blades. By the time all the gyros were spun up to full speed last August, the cryopump had adsorbed a considerable number of helium molecules. Thus, to ensure that the blades did not get saturated and that the cryopump would continue to function throughout the mission, the team executed a procedure called Low Temperature Bakeout. During this procedure, the heaters on the cryopump and elsewhere in the probe were turned on, raising the temperature in the lower probe by 6-8 kelvins. At the same time, two large exhaust valves were opened in the cross flange at the very top of the probe. The increased temperature in the lower probe caused the cryopump blades to release their adsorbed helium, which was then driven out into space through the exhaust valves. At the end of the bakeout procedure, the heaters were turned off, the exhaust valves closed, and gradually, the lower probe returned to its normal cryogenic temperature. At this point, the gyro spin axes were aligned with the guide star, and the science phase of the mission then commenced on 27 August 2004.

Throughout the science phase of the mission, helium molecules have continued to leach out of various surfaces inside the probe, and they have been adsorbed by the cryopump. This process has kept the pressure inside the probe at an extremely low level of less than 10 e-11 (0.00000000001) Torr. This is several hundred times less pressure than the vacuum of space at GP-B's orbital altitude of 642 km (400 miles) above the Earth. However, when the liquid helium is depleted from the Dewar sometime in early September, the probe will warm up, the cryopump will release its adsorbed helium molecules, the niobium coatings on the gyros will cease to be superconductive, the SQUID readouts will go silent, and the GP-B relativity experiment will come to an end.

Drawings & Photos: The photos of the gyros, micro thrusters, and the Mission Operations Center (MOC) during spin-up of a gyro, as well as the drawings of a gyro and housing and the cryopump drawings are all from the GP-B Image Archive here at Stanford. The photo of the cryopump is courtesy of Lockheed Martin Corporation. Click on the thumbnails to view these images at full size.