Intercoolers

[dave92tsi]

I was told something from a reputable dsm shop that i would like to debate. They told me that a side to side intercooler was mostly for its looks and a top to bottom was built for performance. I understand that the top to bottom has better flow charteristics which helps spool time but will it cool better than a side to side intercooler? im going with a hybred 50 trim garett for a street/strip application and trying to match it to the best intercooler that would fit my profile. oh sorry its a 1g, 50 trim, s-afc, 660, exhaust, porting, etc.

 

[talonted_one]

side to side intercoolers flow higher amounts of air and tend to be larger.

top to bottom flows less however are more efficent. a large enough core may pose cooling problems of the radiator.

bigger isnt always better, thicker is good to a point, flow doesnt make up for a poor design, efficiency is very hard to gain.

 

[KenTSIII]

Wrong!!!

Side to side normally have slightly higher pressure drop, but also cool the air much better.

Top to bottom have many short rowes & flow better (less pressure drop), but normally don't cool as well as the side to side. Also the way most top/bottom FMICs are configured... about a 3rd of the core isn't even used due to the air bunching up at the end of the endtank. Also the air has to make a 90deg turn to get into the tubes... which also hurts flow.

Regardless of flow/efficiency... Most people use side to side in order to be able to fit a much larger core in a confined space. On our cars we don't have alot of vertical space to fit a large top/bottom core along w/ the additional 6" of endtanks.

 

[DCJ98GST]

The two issues with a intercooler are flow and efficiency.

FLOW:

Low pressure drop is important for intercoolers because your turbo has to work extra hard to overcome the drop and thus heat the air more! Which is exactly opposite of what you are trying to do with the intercooler.

You always want the charged air entering the intercooler on the larger dimension of the intercooler. This will give the air more channels and help maximize flow. The top to bottom are far superior in this aspect. You also want the endtanks to have smooth curves and direct the flow to the air channels. No sharp 90deg bends and no square tanks!

You must be logged in to view this image or video.

You must be logged in to view this image or video.

The air gets caught up in these areas and can reduce pressure drop. Remember this air is traveling at approx. 200 mph in a 2 1/4" pipe and you dont want that air pulse to slam into anything then turn 90deg. You want smooth transitions into the core. A lot of intercooler manufactures talk about how they have low pressure drop across the core, but say nothing about how bad the endtanks are. The pressure drop of a system should be evaluated from the turbo outlet to the throttle body.

Different types of cores flow better than others too.

EFFICIENCY:

As far as which intercooler is more efficient it all depends on the set up and the core. The type of core has a lot to do with efficiency. A big aspect of efficiency is an even distribution of air into all the channels. Side to side intercoolers can do this more easily because there are less channels.

If both the inlet and outlet of the pipe are directly oposite each other the air is going to try to take the shortest route and not use the other cores. Like this:

You must be logged in to view this image or video.

But if the inlet and outlet pipes are on opposite sides of the core all air paths are the same distance and the air will travel more evenly through all channels. Like this:

You must be logged in to view this image or video.

This helps efficiency. This is also why people think the top to bottom IC are less efficient because they usually have the inlet and outlet on the same side. And the air taking the shortest route just uses the first few channels.

You must be logged in to view this image or video.

But just flipping the end tank will make a big difference in efficiency.

Another way to increase efficiency in a top to bott, is to put a baffle in the inlet tank to direct the flow across the core. This will greatly increase the efficiency. Corky Bell talks about this in his book. Basically the air is split up in the inlet tank and half of the air is directed to half the core and the other to the other half. This will give good distribution to the core.

People think that if they have a long side to side core they will get a lot cooler air than with a shorter top to bottom core. This is not true. It has been shown that 90% of the cooling is done within the first 10% of the core. So long cores do not add that much efficiency and just reduce flow.

People say that the air is not in the core as long with the top to bottom cores because the channels are shorter. This is not neccessarily true. As long as you have even distribution of the core the air travels through the top to bottom core at a slower rate because it has more channels to flow through and will be in the core the same amount of time as the side to sides.

So top to bottom cores will flow more than the side to sides. Depending on the set up they can be just as if not more efficient too.

http://www.are.com.au/techtalk/intecoolersMR.htm

 

[KenTSIII]

DCF98GST..

This info was found on that link that you posted.

"The other design mostly is used in the USA., where tanks are on the top & bottom of the core, causing the charge air to make two 90 deg. bends. By using a second manometer with a long thin probe, we found that 80% of the intake charge flows through the end 25% of tubes on our first test tank, which was of reasonable length. Flow rate was good but cooling rate was terrible. When we doubled the length of the tank, 55% of the intake charge flowed through the end 25% of tubes & overall flow dropped 9%, indicating that this design has too have very long tanks, but should be able to be avoided anyway. We also discovered flow 'syphoning' for the first time during this testing. The intercooler was making a strange noise, & with the aid of the second manometer probe, & then removing the tank & using the old tissue streamer on a stick trick, found that air was actually being sucked back up the tubes next too the inlet end, by the same principal as a spray gun operation. This means that some of the already cooled intake charge was being recirculated through the core again. This creates a double loss as not only does the extra syphoned air create more pressure drop, it has already been cooled so takes away core space needed by the hot air. I didn't waste time measuring the % recirculation as we rarely use this design. Please note that this paragraph refers to intercoolers with the inlet/outlet above each other (on the same side). It is reduced noticeably if one tank is reversed, but still should be avoided if possible. A dual pass intercooler is heaps more efficient where both pipes are on the one side, but they require more height for the core. "

Also I see that you quote corky bell quite a bit. Well alot of what he discusses has to do w/ low boost/low flow V8 motors. I haven't read that book in a while... but alot of the info is outdated compared to what is known today.

 

[DCJ98GST]

I think that quote makes my point very well. 90deg bends are bad. It would be best if the pipe entered on the top of the top-bottom design like their dual pass intercooler. But we dont have room for this. Their dual pass intercooler would be prefered because it would allow better positioning of the end tanks than even the top to bottom types. Take a look at the picture in the link and you will see what I am saying.

You must be logged in to view this image or video.

This intercooler above has some pretty sharp 90deg bends and does not direct the flow through the core. The 90deg turns the air has to make in these endtanks are worse than what happens with the top-bottom intercoolers. Think of being the air and traveling 200 mph and slamming into the back wall of the endtank in that picture. This WILL cause a loss of dynamic pressure. It seems like it would do a poor job of distributing the air through the channels evenly. For the air to reach the bottom channel it would have to make two sharp 90deg turns. The first downward to reach the channel and then a sharp left to get into the channel. Real BAD.

This would be much better:

Notice that there are no sharp bends and it directs the air to the core. These endtanks leave enough room for the air to reach all the channels evenly. Without making any 90deg turns. These tanks are similar to the dual pass intercooler that they had on the link and is always prefered wheather it is a top -bottom or side to side.

I dont think that the top-bottom intercoolers that I posted above are the best set up you can have, I would love to have the tanks shown above on the top-to bottom intercooler (if there was room). But, I think you will have a much better set up than the long side to side types that you typically see on the street dsms especially if you reverse the endtanks.

 

[KenTSIII]

Actually the wall that you mention actually distributes the air across more of the core. We also offer a radius inside the endtank to help flow. We can theorize what would work "best" all day long. But the truth of the matter is that most of what we are discussing would only make a 1% difference vs. the other. We don't have room for the large sweeping endtanks and the average DSMer isn't going to want to pay the price for custom cast endtanks. I still think that a top to bottom core is less efficent than a side to side in our applications. A top to bottom is flipped on it's side is just a really tall short side to side. We have customers that have ran 9.8 @ 146 on our FMIC. And Sheperd ran 9.4 on BR's FMIC which is very similar. There's a reason why you don't see the big dogs using top/bottom in their setups?

 

[draculia]

basing your coice, not that anyone is, on whether its top to bottom or side to side is silly anyways. you can further complicate things when you introduce the different types of cores, bar and plate and tube and fin. on top of that, look into it even further. what type of fins are used, also how dense are the fins. one of the links posted above also shows some nice pics of how some tube and fin cores have the charge side tubes sticking out a few mm, that can't be good for flow.

though we might not have much room, we can maximize a setup that we can fit. instead of poorly designed box type endtanks, you could look for some smooth or smoother tanks with outlets maximizing distrobution and flow, not ease of install.

though a turbo will have to work a little harder for a side to side core, i think the cooling benefits surpass the extra strain on the turbo. how many out there run turbos way past their efficiency anyways? if i'm running a gt30 on the street at 18psi, am i gonna be worried that the extra 2psi, fictitous number, of pressure drop is going to run the turbo out of its efficiencey? nope.

 

[KenTSIII]

The actual difference in pressure drop is less than 1 psi... & in most cases... it's about 0.5 psi or less.

 

[FastRthenU]

Well I know you dont want to much presure drop. But the thing is that the presure drop is caused by the intercooler cooling the air making it more dense. So would it not be the case that, if the intercooler has a large presure drop. It is doing its job, and cooling the air charge.

But with presure drop, comes hotter intake air temps due to the turbo working harder, trying to produce the presure. But what is better? A intercooler that out cools most, but has a large presure drop. Or a intercooler that doesnt cool as much but doenst need to also because of less heat from the turbo?

 

[eyebrowski]

so is the IRC race core a bad design...it has a 90* turn right into the end tanks. why is it such a popular core?

 

[KenTSIII]

It's not a bad design.. it's just not considered "ideal". The IRC race core has ran 9.8 @ 146 & helped make 611 HP on another car. We can add the round radiused endtanks for the people that feel the "box" design is inherently bad.

 

[eyebrowski]

Originally posted by KenTSIII
It's not a bad design.. it's just not considered "ideal". The IRC race core has ran 9.8 @ 146 & helped make 611 HP on another car. We can add the round radiused endtanks for the people that feel the "box" design is inherently bad.

a word from the man himself
thanks, you will be getting a call from me soon.
-Greg

 

[RED_AWD_BLUR]

So from what I have been reading I think it is time to take a good HARD look at the Hahn racecraft FMIC. It has all the benifits and they actually have enough balls to post there performance w/standardized numbers on their site.

http://www.turbosystem.com/hahn/parts/intercool.htm

 

[eyebrowski]

idealy you wouldnt want the in and out on the same side would you? the air will take the path of least restriction which may for some air be right out the outlet bypassing the core.

 

[Tevenor]

SO just clarify: which one of these would be ideal for replacing the stock sidemount? And which one of these has the 'autocrosser' in all of us in mind when designed?


:laugh:

 

[KenTSIII]

Sean.. I can make you the baddest sidemount know to man... of course it would be air/water.. but you get the idea )

 

[Tevenor]

Originally posted by KenTSIII
Sean.. I can make you the baddest sidemount know to man... of course it would be air/water.. but you get the idea )

Schweet! Are we talking self replenishing (i.e. streetable) or 'just add ice water' type? I had a crazy idea once for a setup but never had the balls, time, or money to try it.

 

[KenTSIII]

Sean...

It could be done either way. Email me privately & I'll tell you what I had in mind.

 

[Tevenor]

Originally posted by KenTSIII
Sean...

It could be done either way. Email me privately & I'll tell you what I had in mind.

On its way.

 

[jayntguru]

Originally posted by DCJ98GST
The 90deg turns the air has to make in these endtanks are worse than what happens with the top-bottom intercoolers. Think of being the air and traveling 200 mph and slamming into the back wall of the endtank in that picture.

You said earlier that the 200 MPH figure was for 2.25" pipes. Ken's IC uses 2.5" pipes. Point is the same, but the air speed is less.

As far as the Hahn one, take a look at the installed pic:

http://www.turbosystem.com/hahn/parts/intercool.htm

It's an NT specific intercooler it looks like. Knowing HRC's products from past experience, it will look great, be expensive as hell, and not fit worth a crap.

 

[Gadget Man]

So what is the general consensus on the tube and fin type of air coil used as an FMIC. I can't really afford a real FMIC at the moment, but I have acces to a 5 row thick (roughly 4") or a 2 row (1"1/2 thick) water to air coil from a heat pump. I'm looking at cutting it down to size and trying it out just to see. I figure it can't be any worse than the stock SMIC. It has 3/8" tubing in it so I think it will have pretty decent pressure drop. I intend to route the tubes into a 2.5" piece of copper as a header, so there would not be any end tanks as such and the flow should be pretty evenly distributed I would think. The fins certainly aren't as tough as a real FMIC so itwill get beat up easily, but I think a piece of "gutter guard" with a screen on the back should protect it quite well without cutting air flow off much. Just a way for me to get a temporary free FMIC, and permanent if it actually works well. Has anyone used one of these before?

 

[P8ntBalla]

Everyone here is making great points and all, but I haven't seen anyone discuss how a bar-and-plate would flow/cool compared to the tube-and-fin. I have heard alot of great things about people who have Apex-i Sylvia and GT-R cores on their cars, and I was thinking about getting one myself. However when I compare the pics (I don't have a bunch of I/C cores sitting around so the net is the next best thing) of bar-plate and tube-fin, it looks like the bar-plate has alot of thick metal going across the core. Now I'm 17 and in high school so I'm certainly no genious, but I do know that more metal = slower heat exchange. Are these people that are telling me that their Apex cores are "the next best thing since fly-cut pistons" wrong? I mean, i can see how they might flow more but if they're not cooling the air that they're flowing then how would they be advantageous over the tube-fin?

I think im going to start another bar-plate / tube-fin debate, sorry

 

[KenTSIII]

The japanese cores are actually quite a bit less efficient than spearco core. They have more pressure drop & don't cool as well.

 

[MaineDSM]

Originally posted by KenTSIII
The japanese cores are actually quite a bit less efficient than spearco core. They have more pressure drop & don't cool as well.

Like the Greddy and Apex'I for the 2G? I'd love to see some numbers to back this statement up.

Just about every performance part I've seen from the Japanese companies has been very high quality. (eg. HKS cams, Greddy gauges, etc. etc.) I find it hard to believe that the intercooler kits they make are crap.