Intercooler setups

[brett_celicacoupe]

volume of air entering engine per minute = displacement * rpm * VE/2 ??

where VE is a number between 0-1 (percentage)...either as a constant or a function of rpm

not sure on how to calculate the velocity of the air in a pipe of given diameter and temp.....(yet )

[kingmick]

You can get a cooler, caculate it and then ill show you on a flow bench how far out you are! hahahahahahaha

[BlackSupra]

No no 70mm will be fine as long as the run of piping is less than 2 metres

I have about 4m of 70mm piping.

[Skip]

mmm Generally you can get away with 3m on a 7mgte if your intercooler has delta fins, but 4m is too much you had better install a 2JZ before the HG blows or, buy 2 ratchet straps from bunnings, wrap them around your engine and tension to 200kg, this will keep your HG in one piece. But only use the bunnings ones, the supercheap ones will squeeze the block too tight and collapse your oil galleries.

[Jorrs]

www.photobucket.com

[Howieau]

Haha funny Skip.

After a few read overs, I finally understand the delta fin concept haha. Im having troubles passing out rep cos it says I need to spread it around first, but big cudos to Veep, mic, all who have made an input into this thread.

I also am beginning to understand the balance between turbulence/flow etc, that I need to achieve for an effective intercooler piping setup. But I still dont understand how to reach a final pipe size (no fkin idea how to do any of those equations haha).

So what do you guys think would be a good sized pipe for my setup?

Once again I'm running a 3SGTE with stock internals, pod filter, cold air intake, 3" dump pipe/exhaust, boost controller, (running about 13psi but might turn it down to 11psi).

Thanks again for everyones input, ive heard so much hearsay bullshit, so it's good to finally get some decent info.

[ed]

SKIP - im not sure if youre so full of shit, or youre trying to be sarcastic and funny, or what. either way, neg rep for you... so far everything youve posted in this thread has furrowed my brow in grave concern as to what technical fundament any of your assertions ae based on. back em up with facts or shush...

[BlackSupra]

I wouldn't use pipe with 3mm wall for it. It is just a bit too heavy in the wall section.
As for how much power can be made through it, take a look at Richard Anderson's 3RZ turbo 180SX - it uses 60mm or 63mm intercooler pipe for about 780rwhp.
As for how restrictive it is; check the size of the compressor outlet.Toyota 7M GTE CT26 is about 44mm (using my sketchy memory and sketchier vernier eyeball).

THis is the best advice i can see:

60mm pipe and 780rwhp.

As jase suggested, make the pipe slightly larger than the compressor outlet on the turbo.

Skip - WTF?

[Skip]

Yeah okay sorry I was originally having a dig as Howieau asked a fairly technical question but didn't tell us what setup he was running, so yeah it was obviously all rubbish except for what I wrote about Bernoulli Equation, guess I got a bit carried away...

So at 6000 rpm the 3SGTE would consume 6000x2L / (60x2( for 2 stroke) = 100 L/s = 0.1 m^3/s

Therefore the air velocity in a 60mm pipe with 3mm wall would be 0.1x4/(Pix0.054^2) = 43.66 m/s, therefore it is okay to model the flow of the air through a pipe based on it being incompressible.

Firstly i must determine the reynold's number to see if the flow is laminar, wholey turbulent or in transition:

Re = rho x V x D / mu

where rho = density of air = 1.23 kg/m^2
V = velocity = 43.66 m/s
D = pipe diameter = 0.054m
mu = dynamic viscosity of air = 1.79 x 10^-5

therefore Re = 1.62x10^5.

Ill now use the Darcy-Weisbach equation which works for all fully developed flow to find the headloss in the pipework between using 60mm and say 75mm. The equation:

head loss = fx(l/D)x(V^2/2g)

where f = friction factor
l = length of pipe (m) = say 1.5m for Howieau's car
D = diameter = 0.054 m
V = 43.66 m/s
9 = acceleration due to gravity = 9.81 m/s^2

f the friction factor must be taken from the moody diagram using an equivalent roughness of drawn tubing = 0.0015mm

from the diagram f = 0.057 and is in fact wholy turbulent flow.

therefore the friction loss in the 60mm pipe at 6000 rpm is = 154 m

pressure = rhoxgxh

= 1.23 x 9.81 x 154 = 1858 pa = 0.27 psi

now ill base it on 70mm pipe and cut all the crap above we get

Re = 1.4 x 10^5
f = 0.054 (nearly identical)

head loss = 63 m = 760 pa = 0.11 psi

so even though the pressure loss is less than half if we use the 70mm pipe it is still really small compared to the pressure loss across an intercooler, usually in the range of 1 - 1.5 psi from memory.

Note the values for density of air and dynamic viscosity i used above are based on air at 15 degrees celcius. What i did above could be redone with the air at post turbo temp around 90 degrees c (lower density) but i think i proved the point that the 60mm 3mm wall pipe will be fine

[Skip]

Just did a search to find the Moody diagram for you kids playing at home

http://biosystems.okstate.edu/darcy/...odyDiagram.htm

And also found a site that, at a quick glance calculates everything doh!

http://www.lmnoeng.com/moody.htm

[Skip]

Just did a search to find the Moody diagram for you kids playing at home

http://biosystems.okstate.edu/darcy/...odyDiagram.htm

And also found a site that, at a quick glance calculates everything doh!

http://www.lmnoeng.com/moody.htm

Appears to be quite a few of those sites, ohwell at least you can get an understanding of how they spit out an answer

[Draven]

I don't think the point was 3mm wall pipe owuld be too small... just too heavy

[Skip]

Yeah I realise your point Draven, it's not required to be that thick.

Also i relaised when calculating air speed to the engine I used a VE of 100%, obviously would differ for shitter old NA engines to modern race turbo engines.

[brett_celicacoupe]

interesting......

[mic*]

Skip hang on,

You said air can be modeled as incompressible if flow (velocity) is greater than 102ms-1. Then you calculated V to = 43.66ms-1 and said its OK to model as incompressible. Contradiction.

Your "volume of air consumed" calculation does not account for boost either, and this in turn will affect the velocity you calculated.

Also, you have used a STANDARD VALUE for density as you stated, but...
Density should = STANDARD VALUE X pressure ratio (boost in bar + 1) X Compressor efficiency.
Compressor efficiency is a measure of density ratio.


These calculations are suitable for the pipework, but the "sum of frictional coefficients" will make it difficult to work for an IC. You need a stated value from the maker or else you need to break down the cooler into every tube and start crunching numbers. You end up where Mick said;

Go calculate away, and then ill show you how far off you were on a bench.

Mick, the graph i posted shows the compromise you speak of. More turbulators means more charge air face required so that the velocity drops more and subsequently friction remains about the same (minimal). Ill post up a pic that shows what i meant by "longer charge air channels, for the same frontal area tends to reduce internal flow area".


My school of thought is getting the IC and all bends etc right is FAR more important than the size of pipework, as this is where frictional loss lies. Pipe work size = compressor outlet or slightly greater (as said).