Air-to-air split intercooler sizing

[Zach34]

I've been researching intercooler sizing, and I think I have a viable plan but I'm hoping to bounce it off of some of you more experienced turbo guys.

My planned setup:

-EJ25 (2005 STI motor)
-Garrett GTX3071R turbo with external wastegate
-Target HP = 280-320
-818R intended for track-only use.

The turbo is large for my power goals, but my engine builder recommended it as an option and the LB/MIN flow rates I've calculated would put me exactly in the middle of its optimal efficiency range, which should mean a cooler intake charge at sustained WOT on the track and lower turbo RPM for less stress on the turbo and oil system, among other benefits.

I want to use the side vents/scoops for the intercooler(s). Since the side vent openings are very small (I plan to enlarge them like Hobby Racer did), I'm considering splitting the airflow from the turbo outlet to two identical intercoolers, one per side vent. If my calculations are correct (and that's a BIG IF), my turbo's CFM output will be very comparable to the stock turbo's max output - mid-400's CFM, 460 to be exact. That makes sense because the stock engine/turbo produces about 300 hp, and I've seen someone report from an unknown source that the stock turbo produces 435 cfm. Therefore, I'm thinking the total volume/flow capacity of my dual-intercooler setup should be roughly the same as the OEM TMIC.

So here's a comparison between the OEM IC and some custom-size cores I found from Bell Intercoolers:

The OEM STI intercooler's core measures 18.5 x 6.625 x 3" for a total volume of 368ci. I have found no flow data on it.

The Bell A300051120 (https://bellintercoolers.com/i-30005...ficiency.html#!) measures 12 x 5.1 x 3" for a volume of 183.6ci. Multiply that by 2 since I'm using 2 of them for a total volume of 367.2ci. Bell claims it flows 261 CFM (again, 2x = 522 CFM total).

My volume calculations do not account for internal construction differences between the OEM and Bell IC's that might mean more or less air capacity in the IC core, but hopefully that is minimal.

Will this work? Am I out to lunch on this idea?

Any suggestions on Intake pipe sizing?

[frankc5r]

What kind of track car? autocross where you are doing 30 sec runs at slow speeds or real tracks where races can last a hour.
You only need about 30 mph to get full flow to an air ic. Hard to do in autocross.
But air-air is best for long track events. Not sure I would split things up. An old formula from my factory racing days as applied to radiators was you need about 1/5 to 1/7 the area of the
radiator for air inlet size. If you opened drv side inlet to max size, that would appear to be enough air and simplify plumbing.
Been engineer on 24 hr LeMans cars and if u look at them with way more hp than your target, they have radiator on one side and air-air on other side.

Get a copy of Corkey Bells book, Maximum Boost. He is a good friend a very sharp. He has math formulas u can use for air flow and tubing sizes. You dont want sonic speeds.
I used the formulas for calculating losses over length of pipe and response times. Not as much as u would think

[Zach34]

Road courses almost exclusively. 20+ minutes on track or more at a time.

I did come across some excerpts from Maximum Boost in my online research. Going to pick up a copy. You know the author? That's really cool!

I think it would be difficult to open up one of the side scoops sufficiently to accomodate something the size of the stock TMIC. 12" is about the max height. Hobby Racer widened his to what looks like about 4.5 inches. The IC could be angled to expose the full width to that narrow opening, but I think a small fan or two might be needed to pull air through the IC regardless. I don't think much airflow gets pulled into those side vents.

I agree, I'm not sure if splitting the intake charge between 2 IC's is going to cost me more fabrication and flow problems than it's worth. Here's a list of what I know will need cool air:

-Engine air intake
-Intercooler
-Oil cooler
-Transmission cooler

As it happens, we have 2 side scoops and two large scoops on the engine cover. Maybe this is a better setup:

-Engine air intake - engine cover hump #1
-Intercooler - engine cover hump #2
-Oil cooler - side scoop #1
-Transmission cooler - side scoop #2

That would package the air intake and IC in the same vicinity and potentially makes plumbing them much simpler. The downside is the fabrication project for the underside of the engine cover to accomodate all of that, and potential heat soak with a TMIC setup.

[frankc5r]

Tests have shown that typical electric fans used on typical oil coolers , ic heat exchangers are good for about 12 mph.
Not what u would get from side scoop or roof vents.
Good luck.

[RM1SepEx]

I got in a 40% larger than stock AAIC on one side and used a scoop, seems to work fine on the street and autocross, no road race experience. I rotated the turbo 135 degrees and used 2 1/2 IC piping. 12 inch fan on the IC don't need to use it very often.

[Zach34]

That's good to hear positive results from a side scoop position.

I admit I've started to become highly skeptical of the AAIC as an overall better solution to the AWIC in our cars. Mid/rear-engine cars with AAIC's seem to always have ram-air aerodynamically engineered into the body in some fashion. The 911 turbo's side scoops sit proud of the body panels in front of them on much more pronounced rear fender flares. The Lotus Exige has a dedicated scoop on the hood that no doubt gets clean air.

Looking at the wind tunnel videos of the 818R, the front hood and windscreen on the R completely spoils any air coming into the engine cover humps - plus the passenger seat would have to be removed to get a clear path to begin with. The airflow seems to mostly skip over the side scoops.

Although I would love the simplicity of plumbing a top-mount AAIC using the engine cover humps as the cooling air source, it would still be a big fabrication project to get all the ducting right without significantly hindering access to the engine, and I haven't found any info suggesting it would significantly outperform and AWIC aside from the weight penalty of the water and additional radiator. I want consistency and reliability on the track above all else, and the AWIC probably provides that with the least installation hassle assuming I'm diligent about avoiding leaks in the plumbing.

Probably going AWIC....

[phil1734]

Most of your interaction along the side of the car is driven by the wheel wake. As the smoke-wand tunnel tests were done with a stationary car at 40 or 45mph, it's impossible to interpret what is going on there.

A properly designed end fence on the front splitter would probably direct whatever air flow you wanted directly into the scoop (particularly the enlarged ones,) but since FFR didn't do it and I doubt any single individual is going to pony up and run their own rolling-road tunnel test, you'd have to spend copious amounts of time driving around with yarn tufts or similar to figure out what works. Probably not the most efficient use of time and effort.

And as always, remember that getting air to your heat exchanger of choice is only half the battle. You also need to give it a low pressure zone to go to after it has passed through.

[frankc5r]

Good comment Phil. On my awic heat exchanger ( in back),
I am going to duct air from roof to top and
put fan on bottom where it works better pulling.

[Mitch Wright]

As a side note I am at 285whp using a STi TMIC taking air from the right hump on my track only car. Intake temps are at or a few degrees above ambient temps taken from the IC close to the throttlebody so it does work. front finished 818.jpg 818 duct underside.jpg The fabrication isn't as tough as you might think, I have more photo's if you decide to use the hump.

[Zach34]

Thanks Mitch. I'd love to see more photos! The ones above were pretty small. That's great to hear that it works!

I'd like to have a passenger seat installed occasionally, which would block most of the airflow.

[Mitch Wright]

A passengers seat will be block the majority of the opening, I have my fuel cell in the passengers area so a seat was not part of my planning. TMIC Duct 1.jpgTMIC Duct 2.jpgTMIC Duct 3.jpg818 duct underside 2.jpg818 duct underside 3.jpgTMIC underside 4.jpg
PM me your e-mail and I will send you the photos that are larger.

[lpmagruder]

I got in a 40% larger than stock AAIC on one side and used a scoop, seems to work fine on the street and autocross, no road race experience. I rotated the turbo 135 degrees and used 2 1/2 IC piping. 12 inch fan on the IC don't need to use it very often.

Sorry for reviving a super old thread here. I've been doing some autocrosses and on days with more runs (especially practice days where we can get 10+ runs basically back-to-back) I'm getting detonation and FBKC/FLKC in the later runs. My fueling is already conservatively rich and my timing is already pretty conservative too, so I don't think I'm going to be tuning this out. I'm 95% sure this is due to intake temps. The intercooler is HOT enough to sizzle water off when I squirt it on the right (turbo) side. Left side is noticeably cooler, can almost hold a hand on it.

So I'm on the fence between the zero decibel AWIC setup ($$$ and probably a lot more time investment), and fabbing up my own side-mount intercooler. I'm pretty sure I can do my own side-mount air-to-air for about half the cost, and quicker since it doesn't involve taking the side sails off to route heat exchanger lines. I'm pretty convinced that I want it OFF the top of the motor, however I do it.


Does anyone have photos of the setup I quoted above? I'm going to start on my cardboard-aided-design Thursday but would be nice to copy someone else's homework if possible.


Side note- not sure if something changed later in production, but my side scoop ducting is almost completely blocked by the rool hoop support. It's like the guy working on engine/body integration wasn't talking to the guy working on the frame. I can't even get a hand through it, it's definitely not flowing much air.

[driveslikejehu]

I have a question related to this. I also hope to stick with air-to-air for now. I enlarged the IC scoops a bit as shown on my thread. Around town I got pretty good temp drop thru the IC but at the track I still got about 30* temp delta, but IC inlet temp and consequent outlet temps went way up. As far as I know, I saw no reduction in power or warning about knock.
Folks running AWIC say they get intake temps just above ambient. Cool, but with a stock engine and controller, does that relate to more power (meaning sensible difference, not a couple of HP)? Does the engine management respond that fast? I suspect with a stock setup like mine, you get normal power up to when you don't.
So, I think my big question is, what is the "do not exceed" intake temp? Assuming you stay below this and avoid knock, no matter how you do it, you get all the power a stock system is going give.
Love to hear more explanation

[lpmagruder]

From my (limited) tuning experience, you don't by default get a ton more power with just cooling intake charge, but it will allow you to run either leaner or with more advanced timing (or some combination of the two), and that's where you get power.

What I do is get out there on about the hottest day I think my car will see, get it good and heat soaked and running in the 10.2's ish AFR, and advance timing until I just start to hear the sparkle of detonation in the det cans (factory knock sensor usually isn't even picking up anything at this point) and then pull back timing like 2deg from there. I'm currently sitting at like 10-11deg timing at WOT on boost which is way retarded vs what I've seen in an actual WRX. I think if I can get the air charge cooler I could get it up near 15-20deg.

Obviously didn't have det cans on when racing but my ECU was definitely seeing some knock and yanking timing down below 9deg which is crazy late for where it would have been on even a fully stock WRX.

[Dave 53]

My understanding, correct me if I'm wrong... The general goal is the stoichiometric air / fuel ratio of 14.7:1. Hot air is less dense than cold air. If the air is hot / less dense, to maintain the 14.7:1 ratio, the amount of fuel needs to be reduced making less power. It's neither richer or leaner - the ratio stays the same. So, on a hot day and/or lots of WOT causing the turbo to heat up the intake air temp, the ECU won't pull timing. It will reduce fuel delivery.

I did have high temp tuning issues on track 30 track days ago and they were related to the OEM MAF tune. The MAF tune tables maxed out with the intake temps I was hitting on track. A Speed Density tune fixed it and I've had no issues since then. I flog the **** out of it on tracks in our beautiful sunny California weather including a 107 degree day at Buttonwillow making oil temp the weak link on a hot day.

Note that a MAF tune measures intake air temp at the air filter and a Speed Density tune measures it at the throttle body. With a MAF tune and temp measured at the air filter, I'm not sure how you can measure the efficiency of an intercooler setup because you're measuring air temp before it goes through the (turbo and) intercooler.

Intake air temp just above ambient... only driving down the freeway. My last track day, minimum intake air temp was 85 (just above ambient) under yellow flag and maximum was 136, but mostly it was around 120 (again, measured at the throttle body, not the air cleaner). I've seen it in the 160's with no DAM, Feedback Knock or Fine Knock Learn issues. I have asked my tuner (Travis, Snail Performance, Auburn CA) for a "go into limp mode" intake air temp and he won't give me a hard number and doesn't seem particularly concerned about 150+ intake temps. The Speed Density tune deals with it by cutting power, I imagine by restricting fuel. I can't imagine an autocross car seeing these temperatures. I don't watch intake air temp real time - oil temp is the gauge focus. Until Travis tells me to stop sending it, I'll keep sending it!

I have an AWIC. AWIC isn't as efficient as AAIC. I'm trying to figure out how to do both, but I have an 818S, so no serious solution found. If you have an R and can do a big AAIC, seems like that would be better than an AWIC. An oil cooler also needs to be on the list and maybe a Speed Density tune.

[driveslikejehu]

Thanks for the temp info. I was up to ~150* at full boost with no apparent deleterious effects.
I think with a fully stock set up like mine there's to air temp sensing after the MAF sensor- I might be wrong. If not, then the ECU can only control per ~ambient. And no ECU changes till knock is sensed.
Even so, I think more air over the IC is better, to give the system some overhead.... I'm not sure where I'll get it as I have a street windshield for now. Side vents and some ducting?

[Dave 53]

I'm not an expert tuner. These are just my opinions after studying up on it when I was having issues.

We want cool intake air because the cooler air has more density/mass so more fuel can be metered to maintain proper air/fuel ratio which equals more power. Cold air = more power, so get the biggest intercooler you can!

The ECU needs to know and tunes the engine based on the mass of the intake air. The MAF sensor measures air mass directly. With Speed Density, air mass is calculated by manifold intake air temp and pressure. So, the ECU doesn't wait for a knock before adjusting - it's looking at air mass. I'll bet you a steak dinner there is no intake air temp measurement after the MAF sensor.

Per my tuner, the first step is to be sucking in the coolest air possible. I moved my air intake from the top of the engine to the side scoop and even insulated the piping. Where is you air filter? If your air filter is on top of the engine, you are sucking in 150* air and throttle body temp might be over 200*! When I hear people say their intake temp is only a few degrees above ambient, you have to understand that that has nothing to do with the air temp after it passes the turbo and intercooler because it is measured at the MAF sensor and has nothing to do with the effectiveness of your intercooler. Only knowing throttle body temp (Speed Density tune) can you make an assessment of the effectiveness of the intercooler. The exception being I guess some measure intercooler air temp in and out.

In any case, all this stuff only matters with a track car running long duty cycles of WOT. I would love more intercooler for more power. But, my AWIC seems effective enough to make oil temp the weak link, not intake air temp. At this point, the effort to increase intercooler capacity for more power isn't worth it to me. No data log data indicating damaging knock. Send it!

[lpmagruder]

Typed out a longer reply but the browser crashed when attaching images :-(

Your ECU will only target 14.7ish) on cruise. If it did that under boost/high load it'd detonate, probably pretty bad. Generally 11.2 or so is "safe" on pump gas with 15-20deg of timing but YMMV. Heat will increase knock risk so you either have to retard timing or richen it up, or both, and leave power on the table. I tuned my car when it was cooler outside and it wasn't knocking but the heat soak in the paddock and slower racing on a hot day was enough to push it over the edge apparently. Hopefully no permanent damage done. Lesson learned there, tune on a hot day or leave some more wiggle room.

Anyway, I grabbed the largest intercooler that would (barely) fit in the side pod and rough-plumbed it in. I'll test it out this weekend. Did a bit of driving yesterday and it was still a safe temp (can hold my hand on it no problem, the old setup would quickly get too hot to touch at the area directly above the turbo)


IC Side.pngIC Top.jpg