Re: Can Cooler 4000
Holy Cow! The Can Cooler 4000 works. It started out at about 4 minutes but now looks like we might be able to go under 2 minutes after some changes. The following is some detail of how the first tests performed and how it is evolving.
Starting out during our initial 4 minute runs, the machine cooled the center of the diet coke to 41 degrees with about a ¼ inch of ice caked all the way around the inside sides of the can. If you then take the can out of the machine and swirl it by hand for a few seconds it quickly cools all the liquid down to 33 degrees F (0 degrees C). But we do not want ice.
The initial power supply settings:
Voltage: 12V
Current: 9A
Starting conditions (time zero):
Copper heat sinks : 68 degrees F
Water or coke: 68 degrees F
End conditions (time 4 minutes):
Copper heat sinks : 84 degrees F
Water or coke: 33 degrees F
Ice is a better thermal conductor than water but it hurts our cause. The copper/aluminum are much better at transferring the heat away from the water directly. So we tried another experiment. We opened the top of the soda can filled it with water and stirred the liquid with a chopstick while it was cooling. This gets all of the water to 34 degrees in 2 minutes 30 seconds and almost no ice!
Starting conditions (time zero):
Copper heat sinks: 67 degrees F
Water or coke: 67 degrees F
End conditions (time 2:00):
Copper heat sinks: 75 degrees F
Water or coke: 40 degrees F
End conditions (time 2:30):
Copper heat sinks: 78 degrees F
Water or coke: 34 to 32 degrees F (minimal ice)
In these tests we are not running anywhere near 4000W, we do not know the optimal settings yet and are playing it safe. In other previous tests we had the current turned up to 12A (17V). But back then we did not know it was creating ice in the can so quickly. When the ice forms the cold plate temperature goes way down and the whole system slows down. There is a balance point we need to find with the current applied to the Peltier chips based on the delta- T of the system. Too much power in and the efficiency goes down.
Agitating the liquid is essential. But, we do not want to mix the liquid by hand so the machine needs to agitate the liquid itself. I did realize this beforehand and that is why we have the option for powered mecanum wheels. Unfortunately, I now know they will not be enough. Especially after we turn up the power. So we are off redesigning the can cooler. This is a major redesign but the cooling engines are not being changed, only their support structure.
My hope is to cool a can to 35 degrees in under 120 seconds. To achieve this, we are going to cut about 33% of the copper mass out of the cold plates. That will reduce 2lb of copper that need not be cooled. We are going to turn up the power to find the optimum value. And last we are going to agitate the liquid.
It looks like the best way to mix an unopened can is to roll it or rotate it around its long axis with that axis aligned horizontal. So I will be rotating the cooling engines around an axis that is horizontal. This sketch shows the structure. A four jaw scroll chuck will be inside at the center to actuate the four arms simultaneously. A rotational armature will connect the power to the engines.
Cheers
Dave
