DIY top mounted intercooler

[scottiescottiescottie]

So trying to mimik the advantages of the tube and fin core...

I am keen to cut tubes into half round and weld them to the leading edges internally and externally.

This should provide a smoother entry to the air in the per core end tank. On the "post core" tank I want to create a 'V' shape to emulate a tear drop and decrease air separation and vortices as the high speed air exits the tube.

(seperation and vortices on the back of a round let alone a flat!)

Cutting the tubes accross the 90mm face is easy but the front 280mm face may need to be machined to ensure accuracy and good contact for heat transfer. The two wall thicknesses combined create a 48% contact with the external facing which should easily be enough to retain good heat transfer between the face and half round.

Bear in mind that the front face of a bar and plate core is not like a tube and fin in that its actually quite thick by comparison... The front "leading edge" of the core is 5.5mm thick so will actually be primarily acting as a heat sink than a heat exchange surface as seen below...

End tanks are coming along nicely (but cheaply!) so will post progress soon!

[scottiescottiescottie]

Well here are my end tank patterns...

I made them out of platicard but this should transfer nicely to sheet alloy

more to come...

[soopersubaru]

your mk 1 unit looking good.

Aero plane wing flow from hi temp to lo temp side a top idea!

Nice pre fab work!

[scottiescottiescottie]

Well I have been busy but here are the inlet and outlet tubes I turned up on the lathe today...

Also the outlet end tank rough shape in plasticard...

Just the mounts and BOV templates to make, final touches to existing templates... then all the templates are off to get made of sheet alloy then ill tip them into submission and tack it all up! ;D

Thinking of using pems for the BOV bolts.

On an important note has anyone got a good link OR experience with fast acting temp sensors that I could utilize for my testing?

[ADIKT]

Subscribed. Am very interested to see how this turns out. Top shit

[scottiescottiescottie]

Been a long time sorry but been busy as always with life and children...

Plasticard patterns are being transferred to sheet alloy and should be done the week before Christmas.

Will attempt to have one assembled and weilded before new years.

Pics to come

[scottiescottiescottie]

Well its taking ALOT longer than I wanted but I have had a morning to get something done.

So today the parts have gone from this...

To this...

I have made the eliptical leading edges and pressed some trailing surfaces but the trailing tip is still very round unfortunately as you can see below...

Will be getting into the wielding as soon as possible as I hope to use this cooler 1st of Jan 2012...

[scottiescottiescottie]

Its coming together alright now and will be finished tomorrow :~)

I plan to do some extensive testing post Christmas but simple tests as soon as its finished. Will upload pics in the next two days.

[Stoffa]

you clever little bugger

[scottiescottiescottie]

 Stoffa said:

you clever little bugger

Thanks mate

Anywho here it is! after 8 hours of wielding, drilling, linishing, pressing, forming, sanding, testing, fitting, flowing... I am now -smiling ;D

first the pics then the test results... ill split it into two posts to make it a little easier to digest.

Here is one of the internal wields...

Now here are my controversial inlet and outlet "guides/flow enhancers/ect"

The idea behind them is to reduce the pressure drop, creating a very high flowing core for its size...

My custom-wood jig in action! still makes me laugh looking at it... this little beauty worked a charm, the final fitment was PERFECT

flip side WIP shot...

Finished product comparison to original...

Used hank bushes pressed into the 3mm alloy flange for the BOV mount. THANKS TO BLACKS FASTENERS FOR THE FREEBE ;D

Mounted on the car. Like I said the fitment was perfect. popped into place and sealed up just like a factory item. Glad I spent the time making the jig the stunner she is

Engine bay...

Close up...

Test results in the next post...

[scottiescottiescottie]

So I did a bench comparison with the standard core using a hair dryer and some cardboard

For the first (and IMO one of the most important) test... How well does it act as a Heatsink?

The slower to heat soak was a fore gone conclusion simply judging by the much larger mass of alloy. but still interesting to note how much slower.

Notes...

-I used K type thermocouples for the testing and they were very reliable for me.

-I sealed off the cores to limit any external heat transfer ...

-Both cores were cleaned and internally spotless.

-Heat input was at a tested constant of 91 degrees and the ambient temp at a constant 24.

in intervals of 30 seconds... My core on the left STD core on the right...

Time My core... STD core

0.00 -24 C -24 C

0.30 -26 C -31 C

1.00 -28 C -37 C

1.30 -33 C -49 C (12 C leap on the stock core!)

2.00 -40 C -64 C (My core jumps 7 C now, stock jumps a whopper 15C)

2.30 -48 C -77 C

3.00 -56 C -88 C (Heat soak reached on STD core)

3.30 -63 C 1st hair dryer dies

Predicting that my core would have heat soaked at around 6mins

So as you can see the STD core heat soaks quickly...

IMO The most interesting and helpful result is the first 30 seconds as this is the spot we all use when overtaking/giving our cars heaps!

It is worth noting that the temp only JUST changed to 2 C as it hit the 30 second spot. Very helpful real world info here...

In real life? At a highway cruise you hit WOT as you pass a line of cars and trucks (not that anyone would ever do that right?). The core can soak up a bucket load more heat before the temp even begins to move. Then as per real life road driving conditions you have plenty of time to cool off between WOT moments.

The big question now... How does it stack up as a heat EXCHANGER?

We assumed that it would act well as a heat sink since it had so much more alloy than the STD core. So can we also assume that it will be slower/more difficult to cool?

Can it shed the heat as quickly as the stock core even though it has so much more material to cool down?

Ill save you the pain, the short answer is no... BUT ONLY JUST!!!

More info and solution in the next post...

[scottiescottiescottie]

For the next test I used the same bench setup as previous (hair dryer at 91 degrees) but then removed the covers and added a ducted cooling fan blowing an ambient temp of 18 degrees.

time my core STD core

0.30 -19 -22

1.00 -20 -26

1.30 -24 -30

2.00 -28 -34

2.30 -31 -37

3.00 -35 -39

3.30 -38 -40

4.00 -40 -40

4.30 -42 -41

5.00 -43 -41

5.30 -44 -42

6.00 -44 -42

6.30 -45 -43

7.00 -45 -43

8- -45 -43

9- -45 -43

10- -45 -43

Soooooooooooooooooooooooooooooooo... As you can see the smaller STD core transfers/sheds the heat well and even ends up at a constant 2 degrees cooler than my bigger core...

What does this equate to in real life situations? if you have your foot to the floor without lifting off for 4 minutes? my core will give cooler air.

If you are boosting on WOT for over 4 mins the stock core will provide 2 degrees of cooler air BUT I believe it will do it with a higher pressure drop (TESTING TO COME POST CHRISTMAS)

After a load of research and now some testing under my belt I really see this intercooler core upgrade as only one part of the cooling package.

I am in the throws of designing and assembling a spray system and SUBTLY larger intake scoop for the 03-08 LGT.

Japanese manufacturers have made many scoops for this car and the best ones are only 7mm higher than the std scoop. The scoop doesnt look -WRX try hard- on our cars and yet it provides and increase of 15-20% extra air flow over the core (very desirable)

On top of this though a good atomized spray setup with a 10 liter tank will provide the extra edge to make this a very competitive cooling alternative instead of just going for the infamous "big front mount" as it will be keeping the snappy turbo response. Also its subtle enough that it shouldn't interfere with your insurance unlike the big front donk.

I will also be heat shielding the hot side of the alloy cooler to reduce radiant heat transfer from the close by turbo. The stock core is aided by end tanks that are made of plastic here (they also pop open if you push the boost past 20psi)

So the three tiered upgraded to cooling my intake is this.

-bigger and free flowing core

-High quality atomized water spray setup

-15% more air coming in via a subtle but 7mm higher scoop

For now the car is running sweet, I hope to do a pressure drop comparison by end of January but after that I will be fitting the new scoop and spray system.

Ill post results as soon as I have finished the tests ;D

thanks for watching some DIY...

[BT]

Awesome work man, and good on ya for sharing the info!

[kwikaus]

wow there is some good info there, hanks for sharing

[soopersubaru]

looks like a good result for your considerable attention to detail and your efforts.

I'm still a water cooler fan though ...the over taking 30 sec temperature increase is still more than the soopercooler.

and with water spray coming on at 20 psi the soopercooler IAT does not get above 58 deg....despite the turbo providing air at 145 deg c.

Not bagging your cooler scott.........Your system is lighter and provides a top result and is obviously far better than the front mount system for the "road car"....TOP STUFF and well done!!

[subout]

amazing skills and big thumbs up for sharing your date, very interesting indeed!!!!

[ADIKT]

Make a shit load of them and sell them world wide. I'm sure you'l get alot of Overseas attention?

[ryn]

Great read, amazing work with great attention to detail. Thanks for sharing!

[reaperblitz]

Nice build! Lots of reading material lol.

Quite keen on having one for the forester over a front mount/ Sti upgrade top mount.

Would like to understand/ see an 'oven' test or temp dissipation simulating temperatures coming from/ off an engine to temperature lost or gained in the Intercooler.

Just based on the last test done and coming to the conclusion that the bigger unit can store more heat to whether it can loose that heat just as fast with the engine/ bay temperature variable

[scottiescottiescottie]

 reaperblitz said:

Nice build! Lots of reading material lol.

Quite keen on having one for the forester over a front mount/ Sti upgrade top mount.

Would like to understand/ see an 'oven' test or temp dissipation simulating temperatures coming from/ off an engine to temperature lost or gained in the Intercooler.

Just based on the last test done and coming to the conclusion that the bigger unit can store more heat to whether it can loose that heat just as fast with the engine/ bay temperature variable

Thanks mate,

At the moment I will just see how these two I have built will sell before making any steps in another direction ie developing them for anything other than 03-08 GT legacy's.

The "oven test" idea is not a bad one. I will be testing underbonnet temps at differing locations especially under the core in the coming months so should have some conclusive data on that issue soon enough. On that though radiant heat from near the turbo seems to me to be the biggest issue so will be heat shielding my 3 intercooler units once the temp sensor test shows me where I need it.

thanks for the feedback

[scottiescottiescottie]

 soopersubaru']

looks like a good result for your considerable attention to detail and your efforts.

I'm still a water cooler fan though ...the over taking 30 sec temperature increase is still more than the soopercooler.

and with water spray coming on at 20 psi the soopercooler IAT does not get above 58 deg....despite the turbo providing air at 145 deg c.

Not bagging your cooler scott.........Your system is lighter and provides a top result and is obviously far better than the front mount system for the "road car"....TOP STUFF and well done!!

[quote name='soopersubaru said:

looks like a good result for your considerable attention to detail and your efforts.

I'm still a water cooler fan though ...the over taking 30 sec temperature increase is still more than the soopercooler.

and with water spray coming on at 20 psi the soopercooler IAT does not get above 58 deg....despite the turbo providing air at 145 deg c.

Not bagging your cooler scott.........Your system is lighter and provides a top result and is obviously far better than the front mount system for the "road car"....TOP STUFF and well done!!

The temp increase on this core in the first 30 seconds is 1 degree. What is yours? that is impressive if its 0-1

Good work with your cooler info mate, I would think though by the internally restrictive design of the standard intercooler (that you have modified to work with your setup) that the pressure drop on your unit will be higher and not flow as well.

I understand that a lower temp output is>flow/pressure drop but still the fact remains. I will hopefully get some comparable data soon to see if my core is measurably different/better than the std core.

Once I have my spray setup going I will also do some data logging to compare.

It will be interesting to know if the sprayer will make a big or small difference to output temps...

Thanks for the comments mate

[soopersubaru]

Must agree with you on the flow rate.... Yours should be lower...and.. Yes!!... less than one degree on the average overtake manoeuvre..unless its a B train or a few cars at once.Then its 1.5C to 2C from 90Kph to *$??? on completion and return to correct lane.... Some considerable distance down the road !..Then 10 seconds to recover to 8C above ambient.....Bearing in mind there is 140 C + temp from the turbo on WOT. (20 psi)(20 C ambient)

Will be interesting to see your pressure drop across results.

Am not sure what the std intercooler Pressure drop curve is......Also..Remember that turbulence within does play a significant role in the heat transfer effect and is very important too!

Spray makes a Huge difference!... Latent heat of vaporisation at work!

[scottiescottiescottie]

 soopersubaru said:

...Also..Remember that turbulence within does play a significant role in the heat transfer effect and is very important too!

I think this is where most people get waylaid. There is more than enough turbulence in the maze of internal fins to promote heat transfer.

[reaperblitz]

Pressure drop shouldn't be too bad considering and comparing the front mount designs requiring which require more pipe and a significant distance from turbo to manifold.

As for turbulence within the I/c? I doubt there would be much that would affect temperature as air is forced from one end to the other and by design hot air rises and heat gets transferred -> out.

[soopersubaru]

 scottiescottiescottie']

[quote name='soopersubaru said:

...Also..Remember that turbulence within does play a significant role in the heat transfer effect and is very important too!/quote]

I think this is where most people get waylaid. There is more than enough turbulence in the maze of internal fins to promote heat transfer.

There is good turbulence in your design mate!..Proper finning = higher pressure drop. ( unavoidable)

However...without turbulence...i.e. using a straight pipe design...which will provide a very small pressure drop...virtually no heat transfer occurs.

This is true for the air within(charge air) and the cooling air passing through as well.

An important balance is required for the cooler to work well.

Compromise on pressure drop and required turbulence for effective heat transfer is needed.

Also,

The most effective way to transfer heat in any heat exchanger is to provide cooling flow from Hot to cold.(both flows.....charge and cooling air).

I.e. top to bottom.. for cooling air flow through a TMIC

That is to say, opposite to heat rise characteristics.

( hope this answers Reaperblitz)