Intercooler design - explanation of why this picture shows this?

[onyx1121]

Linked directly from Turbobygarrett.com . Why does it show the one is poor and the other is good? I think all of the ones for the Evo X I have seen are in the category the graphic indicates is "poor." I guess I could theorize why, but I would definitely be interested on hearing input from AMS, ETS, other intercooler makers, and of course all of us regular folks too


[Taken from http://www.turbobygarrett.com]

[►EvolutionX◄]

I would think that they say this because style B (right figures) would create a shorter flow path, creating less resistance, however style a (left figures) creates a better cooling effect. Thats my opinion.

Also in style a, the second photo has endtanks with rough corners, I think that is what they are saying to avoid, basically as many turns and bends as possible.

[seven_227]

I don't see how this will work on the X.

[►EvolutionX◄]

Hes not trying to make it work on the X, hes just wondering why the intercooler design that they do not recommend appears kinda like our intercoolers.

[flyboytb]

i think hes getting at the point that the x intercooler is similar to the "Z flow" listed on the poor side of that diagram.... so the question is does this "Z flow" design really effect performance as compaired to the designs to the right in the Good catagory.

[jcdeng]

the only logical reason I can come up with is the intercooler on the right side has smooth bents and edges, so when the air travels inside of it, there will be less drag/dead spot/fiction.

the z shape is not necessarily bad design, (since the hot air stays up and cold air is more dense and usually travels downward) the z shape will allow the cooled hot air (falling to the bottom of intercooler) be sucked out more rapidly due to outlet pipe is at the bottom. I think the z shape diagram just shows that the oulet piping/connection is not a smooth bent, but rather a sharp/90 degree/cubical bent.

I am no turbo guru, just my logical reasoning from looking at the pics. Maybe should get the Turbobygarrett.com guys to explain to us why the evo X intercooler is "bad design".

[onyx1121]

I was just asking out of curiosity guys - not trying to modify the X to use this design....yet

Agree though, not sure how this would work on the X, but makes you wonder, would the engineers have preferred to use the "good" design but the front frame & bumper assembly prevented it.

[seven_227]

After looking some more I think it is doable but won't work with the stock upper and lower intercooler pipes

[spdracerut]

The Z-flow is a poor design because the majority of the flow will only go through the bottom half of the IC, making the top half basically useless.

This is why: the airflow going into the intercooler is shooting straight down and will want to keep on going straight down unless something makes it turn. What will make it turn? Pressure differential, i.e. there has to be higher pressure at the bottom of the IC to make air entering from the top make a turn at the top. Makes sense right?

Lets look at the 'Good' design. The end tank on the right side of the intercooler is narrow at the bottom and open at the top. Therefore, there's going to be more resistance to flow at the bottom. The greater resistance to flow on the bottom right means more resistance to flow on the bottom left. The resistance to flow on the bottom left means it'll force more flow to turn at the top of the IC as it enters on the left side.

Looking at the 'Bad' design, the bottom of the right end tank is 'open' while the top is narrow. In this case, the narrow top is restricting flow, while the open bottom is encouraging more flow. So in the bad design, you have the flow shooting straight down to the bottom, AND the right side end tank encourages the flow to go through the bottom even more. So very little ends up in the top half of the IC, without some type of internal deflector, or a uniquely designed endtank to force more flow through the top half.

[Evo8Dreams]

I have seen numerous intercooler setups in all kinds of crazy setups that still seem to make power. Pressure is pressure, it may not be optimal, but it will still work.

[ETS Chris]

The flaw in those drawings is that the bad Z design is the same as their good design with just the end tank flipped over. I'd take all this with a grain of salt. Its basically the word of 1 company with no supported data. Just a paintshop drawing........

[spdracerut]

Originally Posted by ETS Chris

The flaw in those drawings is that the bad Z design is the same as their good design with just the end tank flipped over. I'd take all this with a grain of salt. Its basically the word of 1 company with no supported data. Just a paintshop drawing........

Actually, it's fairly basic fluid dynamics. Read my explaination in my previous post. If you want, either get someone to run CFD on an IC for you, or tap an IC with a ton of pressure taps and thermocouples to see what it does.

That '1' company also happens to be the biggest player in the turbocharger market having sold ~9 million units in 2007. Garrett/Honeywell Turbo owns roughly 50% of the global turbo market. Borg Warner owns ~25%. The rest is split up between IHI, MHI, Holset, and a few new players in the game (Seimens/Mahle).

[BigT]

Garrett also arguably makes the best most efficient air to air intercooler cores.

[4G63DSM]

The better designs would have more uniform flow through the whole core, while the bad designs will favor a majority of the air going only through half of the intercooler. Some times the difference is so much you can feel the top and bottom of the intercooler and notice a difference in temp.

However, bigger endtanks can let the flow even out across the core more, even with the "bad" designs.

[spdracerut]

Originally Posted by BigT

Garrett also arguably makes the best most efficient air to air intercooler cores.

I'd be curious to see if anyone has some real data to show that. I don't have any idea who the main manufacters of cores are or how good they are.