Water Cooled Shirt

I visit a camp every year in Denmark where it can get quite warm. While not an insurmountable situation, I wanted to try and see if I could make things more comfortable with a water cooled cooling shirt. The idea is not exactly new, most famously, astronauts wear a water cooled garment in the spacesuit to keep cool, but I wanted to try and see if it helped.

This project consists of 2 pieces. The shirt, and the source of coolness. The shirt is 2 T-shirts sewn together with around 10m of flexible PVC tube in it. For the source of cool I carry an insulated water container with freezer packs in it. A small pump cycles the cool water through the shirt. Freezer packs were chosen because they were calculated to have a few hours of cooling each, and can be frozen easily in any freezer. A few hours of cooling seemed like enough.

Things like refrigerant loops were considered, but rejected for complexity. Peltier elements might work, but would be so inefficient that I would need to carry big batteries and produce a lot of waste heat. If I am going to carry a lot of weight, I might as well carry it in ice, which does not produce heat. The freezer can freeze the packs overnight when it is cooler, or somewhere where I am not, keeping that heat away from me.

This project is not a refined and finished garment and cooling solution, it was a proof of concept, thrown together quickly for the camp to see if it would work. I will share more thoughts at the end, but I learned enough that I stopped further refining this project, and want to do a followup.

For this project I used:

1. 2 T-shirts my size
2. 10m of flexible PVC tube, 6mm OD, and 1mm wall thickness
3. Tube connectors fitting the tube (I used a type for gardening irrigation)
4. a 12V water pump (the type used for mobile homes to pump water from the tank)
5. flexible tape
6. An insulated water vessel (a Coleman "Performance Jug", with about 4QT or 3,8L of capacity)
7. Freezer packs (I used 400ml ones, but any that fit will work)
8. Sewing thread
9. A portable power supply (I used a Makita battery and a small adjustable power supply, but more on that later)

For tools I used:

1. A sewing machine
2. A needle
3. Pins
4. A soldering Iron
5. Scissors
6. A way to freeze the freezer packs

First I placed the tubes on the bottom shirt. My goal was to get pretty even coverage of my entire torso. I wanted the tubes to enter the shirt on the back, since that is where the source of coolness would go.

I taped the tubes together on most of the shirt, with one tube being the supply, and one the return. A small loop on the back is where the flow switches direction. In theory the counter flow should provide more even cooling, as the loops exchange heat while flowing, evening out the temperature between the lines. If I would simply zig-zag the tube on the shirt the entry point would be very cold, and the exit would be pretty warm, with the risk of parts of the shirt not cooling at all. I got this idea from water based floor heating, where you can either spiral or zig-zag the lines, but spiraling is preferred since it evens out the heating (for the same reasons).

The tube with the outgoing flow heating the tube with the cooler water is not actually a loss of efficiency. If my body was not there, the lines would not absorb any heat, and the inflow and outflow would be the same temperature. Any heat flowing from the hot tube to the cool tube might cost some cooling power, but would not consume the ice any faster.

I place the top shirt over the tubed shirt, and use pins to hold the top shirt in the correct position. The tube is guided by the pins on either side, so the tubes should not shift. At this point is important not to puncture the tube.

A sewing machine with blue thread was used to sew through the 2 shirts, on either side of the tube. I did this in small sections because sewing in the middle of 2 shirts is hard. Surprisingly, the tubes do provide some good guidance to the foot of the sewing machine. Also, do NOT sew through the tube.

Did I say do not sew through the tube? I sewed through the tube. I noticed a spot where the fabric was not sliding over the tube. After making sure it was not just my imagination, I cut through the fabric on the back to see that I did in fact, sew through the tube. I cut out the bad section of tube, connect the 2 pieces of tube with some barb fittings, and close the shirt again. On the front it is hard to see I even went through.

Next up is actually filling the shirt with water and testing for leaks. I did this in my kitchen with a bench-top power supply and a tub of water.

The tube was made to fit the pump with tape. It does not need to be a perfect seal on a submersible pump, since any leak leaks into the container anyway. With the power supply I can test if the pump is strong enough, how much power it uses, and how low I can go. With my 12V water pump, at 12V I use around 1A, and there is a good flow, enough to prime the shirt, and enough for a good stream from the return line.

With lower power, the water flow and power usage goes down. at around 5.5V the water flow becomes only a trickle, and if there is no water in the lines, the pump cannot purge the air. below that the water basically stops.

The pump is a bit noisy at 12V, but below 8V it becomes quiet. You can hear in in a quiet room, but among people it is not noticeable.

In other news, no leaks.

The cooling loop was further built at the camp itself. The jug got 3 holes in the top, 2 for the water tubes, and i for the lead on the pump. I made these holes with a soldering iron, though a drill would have been nicer.

I hacked together a power supply using a Makita 14.4V battery with a special holder, and a portable adjustable power supply. The power supply was directly wired to the pump. I will not provide much details since it was a hackjob that should not really be replicated. I probably damaged one of my batteries with this setup, and yanked the wires out on more than one occasion.

The setup allows me to adjust the voltage to adjust the flow. More flow equals more cool.

I sadly have little photo's of me using it, because I was also busy with the camp.

The jug was stuffed in a backpack with the battery and power supply. The jug was filled with a frozen freezer pack through the lid and the pump powered. Through adjusting the voltage to the pump, the cooling is increased and decreased. Every 400ml freezer pack lasted for about 1,5 hours, providing a nice (and surprisingly even) amount of cooling. The shirt never really felt too cool, but cooled enough that even in 30C heat in the shade I was not warm. I was hot without the cooling on (with the double layer cooling shirt, and with a normal shirt).

Throwing ice from an ice maker did make the shirt almost uncomfortably cool. The ice has a much higher surface area and mixes with the water, instantly cooling it. This means that the freezer packs transfer the heat slow enough not to make the water too cool.

If I just carry the jug in my hand, the cooling is not that great. Wearing the jug in a backpack helps immensely by pressing the back of the shirt to my back. The backpack was quite an improvement to the cooling.

Having worked on the first day of camp while it was hot, and then having used this shirt on the hottest days of the camp, I am pleased to say the shirt worked great. While other people were warm or hot, I was not. 4 freezer packs lasted me long enough for a day, and I could freeze the packs overnight.

I was actually planning to make a fancy controller for this shirt, with a flow meter, a few temperature sensors and a motor driver. The though was that I would be able to tell the controller how hard to cool me, and the controller would adjust the pump speed to match the desired cooling based on the sensors. A lack of the right parts at the camp hindered me in this, but the main reason I did not make this controller is that this solution actually worked good enough for the camp. With all the ideas I already had, I did not feel like spending much more time on this version and just use it as is.

It is not something I'd wear outside the camp much. It is a bit too hacky of a job to wear in public, though if there is another heatwave here, I might actually use the shirt in the house.

Having worn the shirt for a few days, I have a lot of ideas for a future version:

1. Sewing tubes into an intact shirt is actually quite hard. If I were to make another shirt, I would probably open one of the seams on the side to get easier access to the entire shirt. I'm also not yet sure whether sewing 2 shirts is easier, or sewing strips of fabric over the tubes is.
2. I do not actually think I need a full shirt. With the tubes in the shirt it is a bit of a hassle to put on, and you are then tied to the jug and the battery. Given that most of the cooling of the shirt was done through the back with the backpack on, I think making just a cooling backpack that has tubes contacting the back, and maybe hips and shoulders, should work fine. It would also make it easier to put on and take off, and be more of a self contained unit.
3. Make the backpack smaller. The jug is way too big for what it is carrying, has a lot of lost space. The container carrying the freezer packs does not need great insulation, since it is warming up within the span of 2 hours anyway. A few centimeters of expanding foam should be fine. Any container that can handle 1 or 2 freezer packs and is watertight should be fine.
4. Make any backpack watertight in any orientation. I had to take my backpack off when I needed to work at angles. It was mostly watertight, but not perfectly.
5. Add a second loop of coolant. This shirt uses a jug with the ice packs in the same water that runs through the shirt. This means I cannot control the temperature of the water. It worked (by chance) for this shirt, but in the future I would like a bit more control of the exact temperature. A separate loop with the coolant, and a loop with the freezer packs, with a pump between them, gives me a way to control the temperature of the loop in the shirt (or on the backpack)
6. More tube(?). This point is with a question mark, because 8m of tube in the shirt (and 2m outside of it) was actually enough to cool me, but I wonder if I can be cooled with a slightly longer and warmer loop of tube.
7. More freezer packs. I do want the backpack to remain compact, but a bit longer cooling would be nice.

I already have ideas on how to do all of these, and plan on making a better cooling backpack next year, with all of the improvements. For now I hope this project can give some people some inspiration on how to stay cool.