Using Heatsink to cool large volume of water

Question

Hi everybody,

I'm currently trying to upgrade a cold water tank at my work.

We use this to cool down hot wax parts quickly.

Currently, its a plastic box with a large household radiator thrown in and copper pipes connecting to a constant flowing cold water source.

It works, however, a household radiator wasn't meant to be submerged and is quickly turning into a pile of rust.

So I'm looking into a variety of things, including making a whole new one. (new one will have insulated sides and top to try and keep the temp)

- one suggestion I had was to build a radiator into the skin of the tank, and then put some insulation on the outside, all of which would be a massive faf to make, and I'm not a radiator designer/engineer so I currently have no idea if that would work ether.

This tank is ~800 x 600 x 300mm and should hold ~140 liters of water, the current radiator is around 2200 BTU.

The room is kept at a constant 21 degrees, and the water comes in at about ~10 degrees iirc.

I've found a heatsink on farnell and wondered if it or something similar could be used (considering its just ally and copper, I could probably get a custom sized one made), but I have no idea if it would be suitable for this application.

I was originally thinking of just having copper pipe run around the inside of the tank, but a colleague mentioned that they tried to do a similar thing for a project at home and found that the copper pipes didn't do too much.

I'm not sure if there is something with the aluminium block along with the copper pipe that might work better, or if I can even get enough copper pipe in there for that kind of volume.

Is there anyone with a better understanding of thermal conductivity or any experience that could give me some advice?

Or maybe someone who has a spare one of these who might be able to do a simple/rough test?

Thanks in advance.

jw0752 · Answer

Hi Henri, Here are the physics of your system. Thermal energy (heat) moves from warmer to colder. It can do this is three ways (1) by conduction where a hot area makes physical contact with a colder area. (2) by radiation where the Infrared radiation from a warm area escapes and is reabsorbed by a colder area. (3) Convection where differences in density caused by the relative temperatures causes fluids to move relative to each other in the Earth's gravitational field. From what you describe I think your best bet will be for the inner tank and outer tank to be made of the same metal. This will help to prevent any "galvanic" activity which can cause corrosion and deterioration of the metal of one of the tanks. The water between the tank should be kept moving so that the metal of the inner tank is constantly in contact with water that is colder than the water of the inner tank or at the desired equilibrium temperature. Heat will always flow from warmer to cooler. How quickly this happens depends on the difference in temperature and the thermal resistance between the two areas. Left long enough the temperature of the warm area will nearly match the temperature of the cooler area. If there is no removal of heat from the outer layer they will meet at a temperature between the warm area temperature and the cold area temperature. Our replenishment and flowing of the water in the outer layer ensures that the final temperature of the water in the inner tank will approach the temperature of the water in the outer layer. The specific heat of wax is fairly low so it doesn't carry too much energy into the water. There is no way to reflect cold or to improve the flow of heat by introducing layers of metal between the warm area and the cold area. This would only likely introduce layers of still water which would act as an insulator and defeat your intention of moving heat energy from inside the water of the inner tank to the water of the outer tank. If there isn't a lower limit on the temperature of the water that you can use (other than the temperature of freezing) the colder that you can make the outer water relative to the temperature of the wax parts the quicker they will loose their thermal energy to the water. Many of the ideas that you mentioned are aimed at creating a greater temperature difference between the outer layer of water and the inner tank water. Don't forget that two different metals and an electrolyte (water and contaminates) can produce electricity and corrode one of the metals (galvanic effect). I learned this the hard way once and had a system where parts were miraculously dissolving on a regular schedule until I made all the parts of the same metal. One more thought, I understand that the wax forms are delicate but even a minuscule movement of the water in the inner tank will have a serious effect on the efficiency of the heat transfer. John

dougw · Answer

Those radiators are designed for liquid-to-air heat transfer where you need a large surface area on the air side. For liquid-to-liquid heat transfer I would expect simple copper tubing provides about the same heat transfer performance on both sides without fins. You will improve performance by circulating the water in the tank or by having more tubing in the tank. Circulation can be done with a pump or a propeller, you could even run a turbine off the water line to run the circulation pump.

sporkeh · Answer

Thanks dougw and jw0752 for your replies; These wax parts get cast in metal, and can have very delicate sections, so they are easily damaged. Unfortunately, the water in the tank needs to be isolated, for a similar reason to having circulation, as there is a worry that too strong a blast of water could push the wax parts around the tank and damage them (some of them float, some are in metal setters which make them sink). This also could be an issue if any bits get sucked out causing a blockage, even with a filter on there I can guarantee someone won't notice and we'll end up flooding the shop-floor. I'm not too worried about keeping the tank circulated, as when in use, the water will be disturbed every couple of minutes from someone putting/removing something. -I've got a hot water tank as well without circulation, and that seems to be okay. I can always add circulation at a later date if its not enough though. I was thinking of pretty much a box inside a box to allow water to run around the outside, but I wasn't sure if it would work. My idea was to have the tank made from sheet metal, with the walls made up of 3 layers with a gap in between to allow the cold water to flow around; It would have an aluminium inner skin which contacts the water, thinking this will act like a heat storage, and then a copper skin outside of that which will be in contact with the aluminium and the cold water, thinking the copper will transfer the cold from the cold water and into the ally skin which will then disperse into the tank water. Finally a stainless steel outer skin which will protect it all. Hopefully the image below gives a decent idea of what I was thinking. We may also put some insulation on the outside of the tank to help reflect the cold into the tank. The next idea would be to use copper piping to go around the sides of the tank, but actually in the water, and then have a 2-layer wall with a gap for insulation in the walls. The last idea would be to have some kind of heat sink or cooling device actually in the water to chill it that way, and still having an insulated wall.

jw0752 · Answer

One other idea is to put a metal box within an outer box. Cold water is pumped into the outer box and surrounds the inner box. If you can elongate the shape of the box you will increase outer surface area for a given volume thus improving heat transfer. Water itself is a fairly good insulator of heat with a thermal conductivity of 0.6 W/m dot K. This means that if you put your heated wax in the center of the pool of water the heat will not quickly move to the cold sink. It will be a good idea to use a stirring mechanism as recommended by dougw as this will help a lot with the transfer of heated water to the colder sink. If you must put a radiator or heat sink in the water tank place it towards the top of the water surface so that the cold water produced will sink to the bottom and produce a convection current. Placing the radiator at the bottom will produce a stable cold to warm gradient from the bottom of the tank to the top. John

sporkeh · Answer

I've had a meeting with others involved and I think what we're going to do is try the copper piping first, as we can get some from across the road (which saves a trip to a nuclear facility, although I would still like to go...), and if that isn't sufficient enough, we'll try an aluminium radiator in there. The problem with the current steel radiator is that yes it rusts, but its also quite thick and so takes up a lot of room, whereas we can get a reasonably thin aluminium one rather than going through the effort of sealing this one off. I like the sound of the wine chillers, not sure how useful the concept is for this project, but I still like it I will try and get some testing done over the next couple of weeks, I'll post updates if I remember, but I've also got quite a lot of actual work to get on with this week Thanks for everyones input so far! its all been very interesting and helpful. I don't suppose anyone knows of a formula or online calculator I can use that looks at the heat transfer of copper piping? I will google this myself when I get a chance, but on the off chance that someone knows of a decent one, it would save me some googletime

jw0752 · Answer

Hi Henri. Is there a reason why the water in the tank has to be isolated from the cooling water? Why not just flush cold water into the tank and drain it from the other side? John

colporteur · Answer

Great coffee break topic for discussion.

At first I thought of the wine chillers used at the liquor store. The bottle is submersed and five minutes later the wine is cooled before you take it to the cash. Then my thoughts jumped to thermonuclear reactor coolers. I don't suppose you have the budget for that? I then settled on the swimming pool solar panels, except instead of using them to capture heat your are using them to release heat.

Large surface area and temperature difference are the two factors I'm considering. The solar panels for pools are designed to take the feed pipe water into capillary like lines for the water to run and collect heat then back to a larger line for distribution back to the pool. If you roll the panel up, ensuring that there is separation between the layers in in the roll and submerse it in a water reservoir that has water moving through it, you have the start of a design.

The solar panel won't give you a corrosion issue it is a form of rubber or plastic. It is designed to be under pressure from a pool pump. You would need to move the water through the panel and also move the water around in the tank.

If you ever watch the moonshine television shows, you see the good old boys setting their copper pipe in a flowing stream. Got a stream to run over your solar panel. Now I don't know if it transfers heat in a liquid to the same level it does from the sun in air.

dougw · Answer

Your first idea is okay, but you don't need the extra copper layer - it actually reduces heat transfer. If you can't have water circulation, make the tank as skinny as possible so the heat has a minimum distance to travel to get to a cold wall. Also put feet on the inner tank so cold water can flow under it.

dougw · Answer

jw0752 is right, stainless steel will eventually galvanically erode the aluminum.

sporkeh · Answer

Thanks Everybody for your input;

For the moment, lets forget about circulation in the tank, and lets focus on the cool stuff, I can always add circulation later on if its needed.

For reference, I've included a picture of our current tank (this has been through multiple modifications since the initial design, originally we were going to have a freezer until we found that water was much more efficient)

You can see the current radiator thats going rusty, and on top is one of the most used setters which will hold a wax component.

One thing to note, is that this particular one (and others) uses a gantry crane to assist lifting in and out the tank (H&S rules :/ )

The plan is to replace this with a 2 hydraulic lifts for each half (we have all the bits, just waiting for our maintenance team to catch up, its only been on their list for about a year )

A bit more background; one of our most produced parts only takes up about half a tank, we inject the component, put it in its setter, then put it in the hot water tank to soften it up for 2-3 minutes(HWT to the left on the image under the lid, cant see much), and then take it out, shake off the excess water, and then put it in cold water for another 2-3 minutes to solidify the shape of the setter, and then leave it to dry.

The CWT is checked before anything is put in to ensure the temp is below 16 degrees.

When we are running this particular part, we would be doing 4 of these at once, 1 in each half of each tank, meaning the platforms would be going up and down every couple of minutes massively disturbing the water quite often.

Currently, using the gantry crane, I wouldn't like piping go around the inside of the tank as it would probably get damaged unless I put a fence in between to protect them. However, once we have the lifts and platforms, it would make it much safer to have piping around. so it may be that this is this best option, just need to work out how much I'd need, I'm sure there is a formula somewhere for surface area per water volume... at least this would be a fairly cheap option we can test out right away rather than going all out on a new tank...

Got it, I'll keep aluminium and stainless separate.

Sean.... ermmm... what? wine chillers in a liquor store? I think I live in the wrong area....we have to wait until we get home and use our own fridge/freezer like peasants...please share with me what I'm missing out on!

As for a thermonuclear reactor, I mean, how much does a black mask, some rubber gloves and a large van cost?

However, your swimming pool solar panel idea intrigues me, my mum used to have a pool, and we used one of these solar covers, kinda looked like bubblewrap, I believe the idea was that the sun heated up these bubbles and released the energy into the water, this was really good at heating up the top inch of water, however, the rest of the pool was still freezing, and I'm pretty sure its not quite what you're suggesting...

Ether way, as much as I like this idea, I think something rolled up as you suggest would probably take up too much room and require quite a lot of movement in a specific direction to work efficiently.

Although I'll think about it over the weekend and see if it inspires anything.

I think I may have seen the odd episode of the moonshine boys, I'll have to watch a few more over the weekend, maybe I can have two uses for this tank...

Ultimately, I think the best starting point would be to have copper piping running around the inside, its cheap and quick, and less likely to get me arrested...

If this isn't enough, then I think I'll go towards a new tank with cold water circulating around the walls and base.

Would it be worth, going back to the idea of water in the walls, to have copper bars in the wall, going through the inner wall, and poking out into the actual tank, so that the cold water has direct contact with both the cold and warmer water.

It might have to be a lot of bars all around, but if it works, then its not silly...

If all of this fails to be enough, I'll be interviewing for getaway drivers.

colporteur · Answer

I found whiteboard exercises are so enlightening and so much cheaper than prototypes.

The liquor store wine chiller is a drum rather like the old wringer washers that had the roller assembly across the top. The chiller is full of water. The attendant puts the bottle in the middle of the opening and closes the clear lid to start a rotation of water creating a whorl pool effect. It was explained that the water on the outside is constantly moving taking away heat from the bottle. I have used them and they work. Now I only drank the wine in the car once, on the way home just to test the temperature.

I understand there might be some used piping real cheap at these locations Fukushima Daiichi, Chernobyl or Three Mile Island. Carrying it away might be a bit of a issue.

The solar mats come in about 4 X 4 foot lengths. This picture shows a ten foot roll. Rolled up in the box the mat is not that big. Laying it on its side would require a pool slightly larger than four feet. Rolling it up would give you a fairly small circumference, so the depth of the pool of water could be less than 3ft. The key is the water moving across the mat. The need to move the water become important or it will not cool in the middle.

Good luck on your project. I would be interested to see you final version.

sporkeh · Answer

Thanks for your input dougw , jw0752 & rstone Interesting about removing the extra copper layer, but what you say makes sense. Also interesting to hear about your galvanic experiences, there are electrics on the tank frame and so everything is grounded, however, I would imagine this would make that effect worse? in which case I will probably stick to Stainless Steel. I was planning on having the cold water flowing around the 4 sides + base and then there would be a lid for when its not in use. I can't make the sizing much smaller, as some of the parts we have only just fit, so I know I may struggle with a hot spot in the middle of the tank when the water is undisturbed, but if it becomes an issue, I can look into putting an eductor to add some movement. As for using parts from an AC or fridge, its been a while since I've looked at the back of my fridge, but I would imagine a fair few of the parts would rust, whether or not I could strip enough away whilst leaving the useful stuff is something I'd have to look at. The cold water around the tank will be constantly moving, it feeds straight from our water line, currently goes around the radiator, and goes back out, so its always fresh cold water, however I don't have any control over making it colder. I'll read this again in the morning after I've some time to think about it.

rstone · Answer

If it was me, I'd try to use the parts from an old air conditioner or refrigerator. Put the condenser coil in your water tank to cool the water. I don't think you're going to get enough thermal conductivity with just a solid block, think automotive radiator, the fins are there too increase the surface area for heat to be pulled away from. I would think that even a small 5000 btu window unit would be enough. Rick

michaelkellett · Answer

I think you would do much better and simpler with a coil of copper pipe in the water. Run the cold water through the copper pipe. It will be well worth using some gentle pumping action to stir up the water in the tank, otherise you will tend to get a cold layer at the bottom (where I assume the copper coil would be.) Pretty much every domestic hot water system uses an indirectly heated water tank with a copper coil inside, through which hot water flows to heat up the water in the tank. The tank water comes out of the taps and the boiler water never mixes with it. Copper tube is easy to get from plumbing merchants, 25m of 8mm tube costs less than £50. MK

beacon_dave · Answer

"... It works, however, a household radiator wasn't meant to be submerged and is quickly turning into a pile of rust. ..." If it works as is and the only real problem is the exterior rust, then perhaps sandblast the original enamel paint off the radiator and replace it with an epoxy-based marine paint ?

dougw · Answer

If you want an easy way to keep seals and still have copper bars, you could hang some U shaped bars of the edge of the inner tank with one leg in each tank.

Using Heatsink to cool large volume of water

Top Replies