Building a Gifford-McMahon Cryocooler With 3d-Printed Parts
Hyperspace Pirate 17:17
150,434 views · 10,108 likes Watch on YouTube ↗
Go to https://protonvpn.com/hyperspacepirate to sign up and get up to a 70% discount on Proton VPN 2-year plan!
In this video I'll demonstrate a two prototypes for a DIY "Gifford-McMahon" or "GM" type cryocooler. The GM cryocooler uses the same concept of a displacer piston that the stirling cycle uses, but instead of being directly coupled to a piston creating pressure fluctuations, the displacer chamber is connected to the high and low sides of a pneumatic circuit with a conventional compressor.
Typically this is done on a closed loop using Helium as a refrigerant (though Hydrogen could theoretically be used but with the risk of material embrittlement). In my prototype, I'll be using compressed air on an open loop for the sake of simplicity since the models in this video are just for demonstrating the concept. My air compressor is connected to an aftercooler with a liquid trap to collect moisture condensed out of the air, and the air is further dried by a dessicant canister downstream of the liquid trap. However, the temperatures reached by my cooler were far below the dewpoint of the compressed air, even with drying, so ice still formed internally, which ultimately caused the displacer piston to seize.
The lowest temperature reached was -84C using a 1kW compressor running at approximately 27% duty cycle, meaning the average power consumption was 270W. This was done with a model built from PVC and a 3d-printed displacer piston with rubber O-rings and silicone grease to seal it against the cylinder wall. The regenerator mesh was made of lead bird shot. I also tried a regenerator mesh using stainless steel beads, but the cooldown rate was slower and the system seized from ice before I could reach a lower temperature.
My first model used 1" PVC for the cylinder, but I also built a model with 1.5" PVC to see if the larger consumption of compressed air - and therefore higher cooling power - would lead to lower temperatures. The 1.5" model ran my 1kW air compressor at 63% duty cycle (630W average input). The temperature dropped significantly faster on the 1.5" model, indicating higher cooling power, but it only reached -52C before seizing from ice relatively quickly, due to the larger airflow rate.
So it's clear that the next phase of this project will require a closed loop system using helium as refrigerant. The cylinder should also be constructed from thin-walled stainless steel (to minimize axial heat conduction) and the cold and hot heat exchanger areas should be made with copper.
Believe it or not, though, in the 1960's there was a commercial GM cryocooler that ran off open-cycle air from an ordinary shop air compressor and reached about -175C. This was the Welch Scientific Model 3150. Here's a video of one in use:
https://www.youtube.com/watch?v=IsLBSoLWR4s&t=1s
And here's a link to a very old Air Force maintenance manual that has more detail about the device:
https://apps.dtic.mil/sti/tr/pdf/AD0740638.pdf
In this video I'll demonstrate a two prototypes for a DIY "Gifford-McMahon" or "GM" type cryocooler. The GM cryocooler uses the same concept of a displacer piston that the stirling cycle uses, but instead of being directly coupled to a piston creating pressure fluctuations, the displacer chamber is connected to the high and low sides of a pneumatic circuit with a conventional compressor.
Typically this is done on a closed loop using Helium as a refrigerant (though Hydrogen could theoretically be used but with the risk of material embrittlement). In my prototype, I'll be using compressed air on an open loop for the sake of simplicity since the models in this video are just for demonstrating the concept. My air compressor is connected to an aftercooler with a liquid trap to collect moisture condensed out of the air, and the air is further dried by a dessicant canister downstream of the liquid trap. However, the temperatures reached by my cooler were far below the dewpoint of the compressed air, even with drying, so ice still formed internally, which ultimately caused the displacer piston to seize.
The lowest temperature reached was -84C using a 1kW compressor running at approximately 27% duty cycle, meaning the average power consumption was 270W. This was done with a model built from PVC and a 3d-printed displacer piston with rubber O-rings and silicone grease to seal it against the cylinder wall. The regenerator mesh was made of lead bird shot. I also tried a regenerator mesh using stainless steel beads, but the cooldown rate was slower and the system seized from ice before I could reach a lower temperature.
My first model used 1" PVC for the cylinder, but I also built a model with 1.5" PVC to see if the larger consumption of compressed air - and therefore higher cooling power - would lead to lower temperatures. The 1.5" model ran my 1kW air compressor at 63% duty cycle (630W average input). The temperature dropped significantly faster on the 1.5" model, indicating higher cooling power, but it only reached -52C before seizing from ice relatively quickly, due to the larger airflow rate.
So it's clear that the next phase of this project will require a closed loop system using helium as refrigerant. The cylinder should also be constructed from thin-walled stainless steel (to minimize axial heat conduction) and the cold and hot heat exchanger areas should be made with copper.
Believe it or not, though, in the 1960's there was a commercial GM cryocooler that ran off open-cycle air from an ordinary shop air compressor and reached about -175C. This was the Welch Scientific Model 3150. Here's a video of one in use:
https://www.youtube.com/watch?v=IsLBSoLWR4s&t=1s
And here's a link to a very old Air Force maintenance manual that has more detail about the device:
https://apps.dtic.mil/sti/tr/pdf/AD0740638.pdf
Category (YouTube): People & Blogs
Playback is via YouTube's official embedded player. Data from YouTube; Exumo is not affiliated with YouTube.