Guide: DIY charge-air water cooler build (intercooling)

Hi, here is a quick guide on how to make your own charge-air cooler on the cheap.


What is charge-air-cooling (aka air to liquid intercooler)?

Charge-air cooling is different to normal (FM) inter-cooling, inter-cooling uses air to cool the charge-air (air to air). Charge-air cooling uses water (or a different fluid/ice) to cool the charge air, the water is then cooled separately in a larger heat radiator positioned in the airflow of the car. (air to water to air). The charge-air cooler jacket surrounds the core so that it is permanently immersed giving a huge surface area to allow for greater cooling. You can get away with a smaller core than a FMIC or the same cooling effect.


Why charge-air cooling?

Charge-air cooling allows you to rapidly cool fast moving air more efficiently than air to air intercoolers. The air to water charge-air cooler is a better way of transferring the heat from the charged air, to the water in the system, the water in the system is pumped through a overall greatter surface area heat exchanger (radiator core) in the airflow, often at the front of the vehicle. Water cooling through aluminium has a 14 to 1 greater ratio of heat transfer than air through aluminium according to Corky Bells Maximum Boost book. Link to PDF here.

Choosing a suitable core.

I just took a guess at the size of the core I used in the thread and in theory, it should be up to the BHP I want to run. However, it is wise to do some claculations and make sure your core is up to the job of flowing enough air for your engine application. Firstly you need to know your expected BHP. Below are some very useful extracts from books written on the subject and after discussing with Russ, it was clear you can run the boosted air through the core in two different ways. First is the way I have done and is the std OEM route for the boosted air, it runs through the extruded tubes, the water runs across the tighter packed cooling fins. The second way is to run in reverse, so that the boosted air goes through the cooling fins and the water runs through the extruded tubes. But when do you know which way is best to run? Below are the calculations from C. Bell book, forced induction performance tuning provided by Russ.

Packaging and the hardware.

First of, all you need to work out where you are going to locate your charge-air cooler, the closer to the inlet manifold the better to keep temps down and short pipe work for throttle response. Space and packaging are the obvious issues here. I mounted mine on a MK2 Golf straight off the inlet plenum and part under the slam panel, similar to the position of some G60 intercoolers. For the core, I used an intercooler from a VAG turbo diesel, new of the Bay, this was 35.00!

Next you need to decide where you are going to mount your radiator (heat exchanger), again packaging will be the limiting issue here, I run a large 19 row oil cooler in front of the std 8v radiator, I have little space with the small bumpers but just managed to squeeze an older adapted Mini radiator into the gap to the front drivers side of the engine radiator.

The pump needs to be mounted at the lowest point in the system ideally, I used a Davies Craig pump for about 70 comes with some fittings, it's design to be used in this application. I did think about the Audi/VW secondary water pump some of the 1.8Ts and 3.2L use, but decided after seeing many reports of them leaking it was best avoided.

Mounting of the header tank needs to be at the highest point in the system, I fabricated my own, the larger it is, the more water/coolant it can hold, the bigger the system the more thermal mass it has, this means it will take longer to heat up if it is very large volume (good for traffic and track) but it also means when it has heated up it takes longer to cool down! And it's heavy. For my system I decided to run 19mm pipes bores.


How to go about fabricating your own charge cooler Pt1

The intercooler needs the plastic ends removing. This is simply achieved by prising a screw driver under metal tabs and levering them back so that you can remove the plastic and the rubber seal.














Once you removed the plastic parts you need to remove the excess material that was used to clamp them in place. I used a bandsaw and then band facer to get a flat smooth edge, I also duck taped the core up to keep any metal swarf or filings out - you don't want that one biting you in the behind.

I then made a card templat for the ends and traced this onto 2.5mm sheet. Once cut out I used a coping saw to make the pipe holes and then set about welding them up.

 

How to go about fabricating your own charge cooler Pt2

To be continued...

 

Oh my word. You are awesome.

This inspires me to get my finger out and actually learn to use my AC/DC TiG to work with alloys. I've only ever done steel with it so far!

 

Thanks Trev, get yourself over here to talk welding!

 

Do you realise that you are supposed to reverse the flow when using air-air intercoolers as charge coolers? Ie the charge air should be passing through where the cooling air used to flow.

 

Any reason why? Do you have any links to more info on this? All the ones I have seen have been this way.

 

No link, but its taken from corky bells maximum boost. The ones i have seen are like that too. But it probably why they are cheap as they as they are using air to air cores. There are expensive ones and either use the right cores or weld them in the correct way which would make welding/design more involved.

 

A good piece of information there, thanks. When reading up I found this from AVT here.

As you can see from the bold text I saw that there was no good reason in my mind to reverse the flow. If the core is designed to flow boost in one direction, I thought it best to keep it flowing that way, I don't want to loose efficiency (as the above would have you believe it could be proper gander, however, it does seem grounded and logical when you look at a core) but I would really like to know why Corky Bells recommends to reverse the flow? Anyone know?

 

Its because it gives a larger flow area and less restriction to the flow.

You may find you wont get a good flow of water going round the system if it goes through the air side of the intercooler. Reducing the ability of the heated water to pass through the rad and be cooled.

 

Just had a quick read of Corky Bells book and cannot find any reasoning why he recommends that. But I will keep looking.



Interestingley he recommends the use of bilge pumps and oil coolers for radiators. I have read up about bilge pumps not being up to the job and them breaking (they are designed to be run at short times when emptying a boat and over heat when constantly in use - how true this is I am unsure but will use a pump designed to run continuously and at a temperature of the charge cooler water, it may effect the plastic seals being hotter than sea water etc), oil coolers I like the idea of and was going to use one, but the hose fittings on them are only AN10 at best, I want to use AN12 which has a 19mm bore to match the hose I will run.

 

The page from AVT is basically saying use proper cores(or buy theirs). But corky bell also says if you are going to use air to air, use the largest flow area possible and they should ideally be thicker to aid cooling. (bottom of the page you post re thickness)

 

Internal flow area p49

 

That basically says what I was saying and what the AVT quote says, keep the boosted core flow efficient to reduce boost pressure drop as per his Rule! But changing the way the boost is passed through the core, ie reversed, you have more restriction and less efficiently due to the structure of the core (according to AVT, do you agree with this?) Water does not need to flow through that way anywhere near a fast as air (in an air-to-air) intercooler so I would think it should not cuase to much (if any) issue for the flow of water. But, I will find out when I get the charge-air cooler complete, and will report my findings here to let use all know.

He goes on further about streamlining but still does not give a reason why to reverse direction in a water to air cooler. Infact, he writes very little about charge-air water coolers. I do like what he says about inlet runner length, backs up what I was saying about keeping them long even in a boosted application.

 

Anyway, a quick reflection on this, if you follow his rule in the above pic:

That begs the question, which way of flowing the boosted air would have the least internal pressure loss? And why would the normal root the boost is designed to flow not be efficient enough?

 

I think you basically treat them like a radiator. You don't weld a rad up and put air through the tubes, and water across the front.
I cant find the reason but in a rad, air passes through the fins. Must work better that way. Plus the larger area will have less flow restriction although ideally should be thicker.
Water is 4 times more effective at soaking up heat. So doesn't need the large area to do its job.

That's how I take it anyway and corky must know his stuff.

 

I found the same mention of a reversed system in the "forced induction performance tuning" by A Graham bell. Will try a scan

 

With a radiator, just like an air to air intercooler, you want the biggest suraface area presented to the airflow, just as Bells says with a diagram on page 56 in the above.

I have just been looking at the my intercooler core, I have not measured it yet, but the gaps, either way round look pretty much the same size and the fins are also pretty much the same either way round you run the boost or water. Thinking about the surface area, it will be equal if you run either way round as well, so that rules that out, there are two differences I can see that reversibg will effect. One the thickness, running my core reversed will use the core at its thinnest, does this matter when trying to cool air? Possibly, the longer the air is in the core the more it can be cooled, so a longer/thicker core would be best.

However, I did think I understood why C. Bell would run it reversed, but I'm not sure now re reading his book. I thought it was purely down to air flow. The stock core and boost air path has the extruded tube core which may hinder flow looking at his diagrams below, where as running it reversed, you dont have the extruded tubes, it's plate and shell style which looks better for flow in his diagrams.




Then i read his notes next to the diagrams, it says the extruded core is more efficient, even though the diagrams make it look worse for flow with the vortexes it creates! Maybe it is better for flow? Wish C. Bell had given his reasoning as I to am sure he knows his stuff.

 

Another explantion from A.G Bell. Compares to two

 

An excellent article, even talks about header tank position, pumps and heat exchangers backing up what I already said. There is lots of detail in there too, much more in depth than C. Bell and possibly more up to date, I like the emphasis on track/sprint cars, very useful.

 

But, Depends on the size of that intercooler which seems to point to tube size. and turbulator density. Would need some calcs to be sure what would work best.