Cool Down Two

We developed a pre-cooling procedure for the second cool down that has become the standard method of cooling the cryostat. We partially fill the tank with LN2 and then rotate the entire cryostat by 180 degrees to allow LN2 to flow down the fill tube and come in direct contact with the heat exchanger surfaces, providing rapid cooling of the VCSs. To stop LN2 from flowing out of the cryostat, we place a cap on the vent line that feeds a 1/4 inch rod through an ultra-torr to KF fitting that reaches the bottom of the tank allowing the boil off exhaust to exit the tank. This process brings VCS1 down to 77 K, slightly above the operational temperature of 50-60 K, and VCS2 down to its operational temperature of ∼ 160 K within two days. The

Figure 7.2: A simplified view of all components of the pump pot 1 K system starting at the helium tank and working its way out to the manifold outside the cryostat. Liquid helium is pulled into the pot from the main helium tank reservoir from a pipe with a 10 µm filter on it. The pipe leads to a valve that is controlled by a stepper motor mounted to the lid of the cryostat and is opened to fill the pump pot through a capillary that controls the flow rate. The pot is then pumped on through an exhaust pipe that leads out the top of the cryostat. During ground tests the pipe leads to the manifold and a vacuum is pulled on it to reduce the vapor pressure to create the 1 K stage. During flight a motorized valve opens the exhaust valve to the near vacuum of the stratosphere which performs the same function as the pump.

Figure 7.3: To ensure there are no blockages in the 1K pumped pot system we have a valve manifold setup to ensure that only ultra-pure helium is present in the system during the cool down. Initially Valve 1 and 2 are closed and Valve 3 is opened to evacuate the manifold. We perform 3 or 4 purges of the manifold by closing 2 then opening 3 and repeating. Once we are sure the manifold is filled with helium we close 3 and open 2 to provide positive pressure before opening 1 to force pure helium through the system. Once we are cold with liquid helium we close 2 and open 3 to pull liquid helium into the pump pot and lower the vapor pressure to bring the stage down to 1 K.

procedure speeds up the cool down process and significantly reduces the amount of liquid helium needed to cool the cryostat. Using this process to cool the cryostat requires at least 300 L of LN2. After pre-cooling, the initial liquid helium cool down requires at least 300 L to bring the cryostat to its equilibrium cold state with liquid in the tank for one to three days before additional helium is required.

During the second cool down the pump pot system became clogged as observed by a drop in exhaust gas flow and a temperature rise at the 1 K stage. The clog could temporarily be cleared by warming the cryostat above LN2 temperatures suggesting a contaminate gas in the pump pot system was freezing out at lower temperatures and causing the blockage. Thermal cycling of the cryostat was unable to clear the blockage and so the cool down was terminated. However, the cool down confirmed the cryostat and fridge systems were leak tight and the 270 mK fridge and 1 K pumped

pot systems worked with only minor modifications needed on the pump pot to ensure it would not become clogged again. We also found the loading on the 4 K was larger than predicted by the thermal model though we were not able to pinpoint the cause necessitating further investigation in subsequent cool downs.

7.5.2

1 K System Modifications

The pump pot system had worked well inside a small test dewar prior to instal- lation within the flight cryostat. We closely examined what had changed when the system was transferred and came up with two possible explanations for the pump pot malfunction. Either trapped impurities such as oxygen or nitrogen had clogged the thin capillary tube that runs between the valve and the pumped pot reservoir or steps in the aperture size through which the liquid helium flows produced cavitation that reduced the flow rate.

To address both possible causes of blockage in the 1 K system we decided to change our filtering strategy. The filters are made from metal foam with calibrated gaps to prevent particulates from entering the system from the helium tank. Filters were placed both before and after the valve as well as at the entrance to the helium tank which we decided was overly aggressive and could be causing the issues. If a filter was too fine it could either be trapping the impurities that we want to drive out with the helium gas purge or creating the change in flow impedance that could cause cavitation. To address these concerns the filters before and after the valve were removed and the filter at the intake from the helium tank was increased in gap size from 10 to 15 µm.

Once the changes were made the cryostat was cooled down for the third time as shown in Figure 7.4. The 1 K system operated well during this cool down though we eventually had trouble again with the 1 K pot system plugging. However, during

Figure 7.4: An image of the cryostat during the third cooldown showing a liquid Helium transfer in progress as well as two vacuum pumps attached.

the cool down we were able to perform tests on the 270 mK fridge, pumped pot 1 K system, HWPR, and the 4 K loading levels.

The next set of alterations to the 1 K system attempted to eliminate the last few possible sources of clogging from the above mentioned sources. First, the 15 micron filter at the intake from the helium tank was replaced and offset from the bottom of the tank with a stainless steel pipe that placed the intake proud of the tank curvature. The standoff was added due to concern that contaminants were building up at the bottom of the tank. A second change that was made to the 1 K system was to replace the pipe that led from the Helium tank to the valve. In previous tests the pipe had been made of convoluted stainless steel tubing with a 1/4 inch inner diameter. We were concerned that contaminant gasses were being trapped in the convolutions and could have dislodged and blocked flow through the capillary. We replaced the

convoluted tubing with straight copper tubing that was designed so the liquid helium in the pipe would always be flowing down to eliminate potential traps in the system. During the next cool down the 1 K system performed well and did not clog suggesting these modifications were effective.