Project: Peltier Beer cooler / Tea warmer

[Mattt]

Design and Introduction

Project Index-

Making the copper round and making the Hot/Cold Plate
Heat sink construction Part 1
Making the insulating layer between the hot side and the cold side of the product, this contains the Peltier.
Top Heat sink construction
Fixing the top heat sink to the hot/cold plate
Test fitting of componets so far
Heat sink pins and Heat sink construction Part 2
Fan Attachment
Assembly of the components I have manufactured so far
Electronics Design
Electronics Prototyping

General Project Info-

Inspired by the (imo) brilliant Peltier powered beer cooler here I felt that I needed a beer cooler. However I also felt beer cooling hadent been taken as far as it could go and what about keeping my tea warm?? So this project was born.

Origionaly the project was my A-level tech project, however due to one of the tech teachers at my school being stupid I was forced to change my design, which left me with a beer cooler that I hated because it was not what I had wanted.

So now that i have left school forever I have striped the cooler back down to the core where my design had not been changed, and this project is the rebuild. so technicaly this is "Peltier Beer cooler / Tea warmer" Mk2.

Anyway thats the history of the project so far, onto the design...

Overall Design-

My intention is to build a cool looking beer cooler, that is in the main self contained, so it wont need a beefy computer PSU to be able to be used. I also want the cooler to give me full control over my beer/tea, so my aim is to have full temperature control built into the cooler.

CAD pics

Heatsink-

Core of design-

Possible Stand/Support-

What Peltier is required?-

To be able to work out how powerful I need the Peltier to be I need to work out the amount of energy that is needed to cool or heat a certain amount of drink by a certain number of degrees.

Calculating the amount of heat energy (q) required to change half a litre of water (about a pint) by 1 ^oC.

mass of substance= 500 g
specific heat capacity= 4.18 J ^oC^-1 g^-1
initial Temperature= 20 ^oC
final Temperature= 19 ^oC
Calculation details= (500 x 4.18 x -1)
Q= 2090 J

This means that 2090j’s of energy is needed to change 500ml of water by 1 ^oC

Now that I have worked out how much energy in joules is needed to heat or cool the 500ml of water by 1 ^oC I can work out how different electrical powers of a Peltier will affect the cooling/heating rate.

With the use of a 30Watt Peltier the following is true-

1j/s = 1watt

2090 / 30

(Number of joules divided by the number of watts)

= 70 seconds

(Number of seconds to heat/cool 500ml of water by 1 ^oC)

This means that a Peltier with a power rating of 30watts would be able to cool or heat 500ml of water by 1 ^oC approximately every one minute. This also means that a 60watt Peltier device would be able to cool or heat 500ml of water by 2 ^oC approximately every one minute.

This would mean that the beer cooler would benefit from a high power Peltier. However since I have to take into consideration other requirements such as keeping the size of the heat sink to a minimum, the amount of energy that will have to be removed by the heat sink and the cost. I have decided that a 33watt Peltier would be suitable for this product.

Heatsink Design Calculations-

To test the effectiveness of my heatsink pin design I used the following test rig. allong with a prototype heatsink made on a manule lathe.

The only way the energy could be dissipated would be through the heat sink pin. The test was carried out by placing warmed water in the container at a certain temperature (40 ^oC), leaving the container for a set amount of time (5 minutes) and then re taking the temperature. This will then allow me to calculate the energy disapated by the pin and allow me to work out how many pins are required.

Time taken (minutes) = 5
Starting temperature (^oC) = 40
End temperature(^oC) = 30
Total temperature change (^oC) = 10

Ambient temperature = 27^oC
Amount of water in test rig = 20ml = 20g

Formula= Mass x SHC x (Final temperature – Initial temperature)
Mass of substance= 20 g
Specific heat capacity of water= 4.18 J ^oC^-1 g^-1
Initial Temperature= 40 ^oC
Final Temperature= 30 ^oC

Calculation details (20 x 4.18 x -10)

Joules dissipated in five minutes= 836 J
Joules dissipated in one minute= 836 / 5 = 167.2 J
Joules dissipated in one Second= 83.6 / 60 = 2.786 J

Joules per second moved by Peltier Watts = Joules per second 33Watts = 33 J

Number of pins needed to dissipate energy moved by Peltier= 33 / 2.786 = 11.84

As you can see from the results 12 pins are needed to maintain the heat sink at an equilibrium between energy input and energy output.

Now seeing as were going for extream cooling I decided that go for more pins then the number required, also i feel more pins will look better. So i am going to try and fit 45 pins on the heatsink.

Electronics Design-

I have decided to include all of the electronics designs for the beer cooler allong with the electronics prototyping and building later on in the project. it will probrably make more sense that why because it will then be in context.

Let the project commence....

Discuss

[Mattt]

Making the copper round and making the Hot/Cold Plate

To start, I needed to make two 100mm Diameter Round Copper plates. One, for the heat sink for the Peltier and one as a hot/cold plate for the drink. They will be attached to each of the sides of the Peltier, to allow for good heat transfer.

First, I found the centre of the plates by scribing two lines diagonally to opposite corners. The point at which they crossed was the centre. I then used dividers to mark a perfect circle on each plate.

Once I had marked out the circle I began by cutting off the corners of the square plate using a hacksaw.

I then used various files to cut away at the copper to get nearer to the line marked.

Once the copper was finished as near to my markings as possible using hand tools, I decided that the finish was not adequate.

At this point I decided to use machines (and some unorthodox methods ) to gain a perfect finish.

Since I could not drill through the centre of the copper plates I could not use the lathe in the standard way. So I decided to araldite a circular piece of wood onto the plates and then place the wood (and so the plate) into a lathe, this however did not give me a totaly perfect circle. (seeing as this was a extreamly dangerous procedure due to having a hefty lump of metal spinning around at a couple of thousand RPM which is only prevented from launching itself across the room by some glue i only have a befor pic. )

At this point i decided to glue a small piece of metal rod into the hole in the center of the wood which i had used for centering the wood for glueing earlyer, I then switched on the power drill and used emery cloth on the edge of the copper, slowly this gave me a perfect circle. To make this unorthodox procedure safer I clamped the power drill to a work bench, and also made sure that I worked from behind the piece of work as well as wearing standard safety goggles and overalls.

Now that my copper plates were circular I had to remove the wood that was fixed onto it, this was also done using an unorthodox method since I did not want to cause any damage to the copper. The method I created is show below-

• 1. First put your copper with plywood attached into the freezer.
  
  2. While your waiting for the copper to cool down (about 15mins)
  
  3. Preheat oven to gas mark 9.
  
  4. Boil one kettle full of water.
  
  5. Boil one saucepan full of water.
  
  6. Get a bucket.
  
  7. When you feel your item is as cold as it’s going to get, pour the 2 lots of boiling water into the bucket and as fast as possible put the item into it.
  
  8. Wait until the water has cooled down sufficiently to remove the item from the bucket without burning your hands (about 15mins) and place it on a baking tray into the oven.
  
  9. Refill the bucket with cold water
  
  10. Wait for 30 minutes until the plywood is a nice golden brown, get some BBQ tongs at the ready.
  
  11. Now turn off and open the oven, remove the baking tray from the oven and then lift the item with the tongs by the plywood. The plywood should come away easily, if it does not hit it a bit.
  
  12. Put the plywood in the bucket of cold water to cool down.
  
  13. Wait 30 minutes for the copper to cool down slowly.
  
  14. Use some light grained wet dry to remove all of the oxidation.
  
  15. Clear up the kitchen so that mother is none the wiser.
  
  16. Your copper is now ready to serve with a side dish of toasted plywood.
  

This method was successful due to the multiple expansion and contraction of both the wood and the copper.

To increase thermal conduction, and to make it look cool I gave the copper a mirror finish.

This sort of finish was achieved by the use of decreasing grades of wet dry paper that was affixed onto a flat surface. Then finally using t-cut, car polish and then just to give the final finish some metal polish was used.

The completed hot / cold plate

Discuss

[Mattt]

Talk about this project log here.

all comments and advice welcome.

[Mattt]

Heat sink construction Part 1

I now repeated the entire process show above to produce the copper plate for my heat sink, however once those processes were complete I proceeded to mark out the holes for my heat sink pins.

This was done by the use of dividers to make circles at certain intervals, and then by marking lines using a protractor at a certain numbers of degrees around the circle, where the lines crossed I used a centre punch to mark the points ready for drilling.

I then drilled the holes using a pillar drill. I used the depth gage to make sure that all of the holes were of the same depth and that they did not go all the way through the copper.

The completed copper plate



Rob (Fatboy) in reflection of plate, probrably eyeing it up ready to steal when my back is turned

Making the insulating layer between the hot side and the cold side of the product, this contains the Peltier.

I first cut out a circle the same size as the copper plates, using the circle cutter shown below. I then used a knife to cut out a gap for the Peltier to sit in and I also made some channels for the wires to be pressed into.

Discuss

[Mattt]

Top Heat sink construction

To make the top heat sink I first had to make two aluminium rings for all of the heat sink fins to be mounted in-between, this was achieved by the use of two different sizes of hole saws to cut the inner and outer diameters of the ring. I then used wet dry paper to smooth the rings and to give them a nice finish. Unfortunately I could not take pictures of the cutting process used to make the rings since it was a high risk process and had to have my concentration fully on the work at hand .

I then began the construction of the heat sink fins, I first started with extruded aluminium with a square cross section on to which I marked a line down its centre.

I then cut down the middle of this box section with the use of a hack saw. At this point I removed and burrs and imperfections from the cut surface with a file.

I then began to cut across the two halves of the box section at regular intervals using a hack saw to leave me with the fins show below.

Once they had been cut they were then filed to remove any imperfections from the cut surface. I also used wet dry paper to remove the marks left by the file and give an over all clean finish.

These steps were then repeated until I had 41 identical heat sink pins. The number 41 was worked out by working out the inner circumference of the aluminium rings I made earlier and then dividing that number by the width of the fins (which was half the width of the original box cross-section)

I then marked out one of the two aluminium rings buy the use of a jig that I made out of a centre square and a rule, I used this to mark at intervals which were the width of the pins.

Marking out completed

Once the marking out was completed I glued the fins onto the disk that I had marked out with super glue.

This would now be where I would fix the other disk onto the other side of the fins, however at this point I decided that the heat sink would be better off without this second ring because it would allow for better heat transfer and it would still be able to be adequately affixed to the copper cold plate which was the main reason for having the second ring in the first place.
Now that the heats sink was completed I cleaned it up using metal polish, to give it a clean finish.

View from top of completed heat sink

View from side of completed heat sink

Discuss

[Techno-Dann]

Looks great! Your cooling pins remind me a lot of the Swiftec air-coolers... I suppose that was intentional.

[Mattt]

Fixing the top heat sink to the hot/cold plate

The aluminium heat sink was fixed to the hot/cold plate using my own thermally conductive glue. The final composition of the glue was 1 part epoxy glue, 1 part epoxy hardener, 1 part thermal compound. All of these measured by volume. I found these values by a series of experiments on scrap pieces of copper and aluminium, until I found a mix that was thermally conductive and with which the glue was able to set.

The impossible task of taking a picture of the surface without getting me in it.

Close Up of the join between the copper and aluminium

Test fitting of componets so far

With pint glass

Discuss

[Tulatin]

Sweet, but wherefore the hell are you getting this machinery?

[Firehed]

Very cool, very different. Definately bares a resemblance to the swiftech coolers

[Fatboy]

Originally Posted by Tulatin

Sweet, but wherefore the hell are you getting this machinery?

CNC Mill is only small, but its at our old school.

[Mattt]

Heat sink pins and Heat sink construction Part 2

I then created the pins for the heat sink, since I needed each to be identical and I needed 45 pins in total I used the computer controlled lathe to make them. To start I first learnt how to use the lathe and the software, I then proceeded to make some artwork for the machine to cut out based on the design of my final product. I then proceeded to make these pins. I had to repeat this process several times until I had a design that the lathe could manufacture, this was due to the machine having certain limitations which some of designs were exceeding.

This is the design profile that the lathe cut out for me on each of the pins

This is the simulation of the profile

I had to cut 45 lengths of aluminium rod that were long enough for the stick out where the profile would be cut as well as enough to hold the work piece in the chuck



I then set about making all 45 pins

And here is the finished result

These pieces were the waste from each heat sink pin I manufactured

A few of the pins I manufactured were not of a satisfactory standard, usually due to aluminium bending when the first lathe tool comes into contact with the work piece. So I decided to make them again.
Once I had the 45 pins required I constructed the heat sink. I first filled the holes with a small amount of thermal compound, to allow good conduction of heat from the copper heat sink plate to the heat sink pins, this good conduction is essential to allow the large surface area on the pins to dissipate the heat.

The pins then had a small amount of super glue applied around the outer bottom edge, they were then pushed into the holes.

While they were drying I used an engineers square to make sure that each of the pins was exactly perpendicular to the copper plate.



The four holes that you can see that do not have pins in are the pins that allow the attachment of the fan.

Discuss

[TTmodder]

i like the cooler

[Shadowed_fury]

Interesting as hell, and what a good idea.

[ozstrike]

Oohh, me wants

[Mattt]

glad everyone likes it.

much more to come soon. possibly tonite if i can bare to sit there croping and resizing anymore images .

[Fatboy]

Teh beer cooler was gona be a damn sight better, but it got 'Crouchified' by a teacher so had to be made so it wasnt as amazing.

[Bandit_zoraK]

What kind of power source are you planning to use with this, just a psu or a power brick of some sort? Looks really good btw.

[Mattt]

Fan Attachment

I then moved on to the attachment of the fan to the heat sink. To do this I first modified the four heat sink pins that had not been used. I used the standard lathe to bore a hole into the centre of the heat sink pin that was the correct size for a M3 thread.

I then proceeded to thread the hole.

Threaded pin with a M3 bolt next to it

I then repeated this process for the other 3 pins. To the fan (low profile zalman 80mm) I had purchased for this project had mountings that were not in a suitable position for me, the fan was also slightly too large and so I cut off the corners and then drilled my own mounting holes.

Discuss

[Mattt]

12 volt 60watt power brick.

its pritty meefy i can tell you that. .

as for what fatboy says, this is true, however this is a near total rebuild, all of the rubbish stuff will be gone, you shall only see the good.

once the project is finished I mite put some pics up of the previous "finished product " just so you can see how stupid my ex tech teachers who forced me to change from my origional design were.

[Timmee]

I was curious, is there a reason you're using a mix of copper and aluminum in your construction? Why not stick with just copper instead? Other than that one thing, I look forward to seeing how this project finishes up.