Pelican Parts Forums

Pelican Parts Forums (
-   911 / 930 Turbo & Super Charging Forum (
-   -   Bell Intercooler (

DonE 02-10-2010 06:14 PM


Originally Posted by Garrison (Post 5177155)
Done, you said above side to side and 3 in core... I am assuming you are talking about the tanks being on the sides, or the rows running side to side vs. top to bottom? Just wondering what you ment and why? Trying to make sure I get it right when I order this. THX

Yes, I am talking about the end tanks being on the sides of the IC rather than the front and back. After hanging around shops that do a lot of high HP turbo stuff, it seems that the side to side works a bit better at cooling the charge. In addition, the volume of air my engine consumes points to a thinner core for best efficiency and lowest lag.

When I spoke to Gunther at Bell, he said the side to side vs front to back wouldn't produce that big of a difference. He might be right, but the big HP (and endurance) guys mostly have side to side. When I ordered my IC, he only had 4in cores in stock and I didn't want to wait the 6 weeks or so for the 3in core.

DonE 02-10-2010 06:18 PM


Originally Posted by 911st (Post 5177207)
Don, I to am wondering about your comment.

In my words your current core runs front to back. If you are saying a 3" side to side, I would not think you could get enough CFM through it.

Your current fabricated IC is a great set up I would think assuming the core size is fit to your HP.

I could see keeping your current lay out and going to a 3" core if it flows enough CFM to support your current HP level.

When I did the math on the IC design, using a 3in inlet and outlet, a 3in core that Bell used was just fine with about 15% buffer. At the time I had it built, they did not have a 3in core in stock.

911st 02-10-2010 09:14 PM

Also the side to side requires a thicker core to flow the same CFM as a front to back.

With a thicker core one needs a longer run to cool the air as the lower part of a thick core will not work as well.

RarlyL8 02-11-2010 05:14 AM

That is not necessarily correct but application/configuration dependant. Side to side is more efficient than front to back. Front to back is easier to package on a 930 due to throttle and turbo position. It is also more costly to do the cross cut side to side than the perpendicular cut front to back.

There are many books on this subject and many proven designs. If you are not using A/C and have room for a full bay intercooler it will obviously have a greater cooling capacity than a half bay intercooler. Front to back side to side makes no matter if you can make up the difference with surface area and size. This is not a race car.

911st 02-11-2010 07:28 AM


Not sure if you are saying if my comment was "not necessarily correct."

If so please help me understand where where I might be incorrect.

My first comment was" "the side to side requires a thicker core to flow the same CFM as a front to back".

What could be incorrect with that?

If you have a say 10" by 22" foot print for a core, if you run it side to side you will need to be about twice as thick to equal the core volume and approach the same CFM as a front to back. Also longer runs can create more back pressure from the slowing that comes with cooling.

My other point was: "With a thicker core one needs a longer run to cool the air as the lower part of a thick core will not work as well."

What might be incorrect with this?

A thicker core is chosen to make up for lack of surface area. Making it longer allows for it to approach, or equal the efficiency of a thinner core.

This dose not mean it needs to be twice as long if it is twice as thick. It also dose not mean a thicker core can not be more efficient than a thin core given a proper sizing and a long enough run. It is just not the best place to start.

Depending on CFM requirement and design, either lay out can out perform the other.

If going to side to side adds another elbow, 18" of tubing, and more fittings it is going to have an effect. As will end tank design and ducting. Over-sizing or unnecessary volume may also have some cost to throttle response.

Packaging and cost is more often the determining factor.

It is very possible that the best possible side to side may very well outperform the best front to back in several ways or applications like drag racing v road racing or throttle response v highest net temp drop irregardless of pressure drop.

I do not mean to infer in anyway that one style is better that the other.

I am far from an expert on intercoolers and have a lot to still learn.

911st 02-11-2010 07:39 AM

Here is some of what Bell says about intercoolers:


Does one style core cool better than another?
No, almost no difference. With three decades of testing intercoolerís, we have found no appreciable difference between any core style or manufacturer. Keep in mind, that the merit of a core is itís efficiency versus its internal drag characteristics. When sized for a tolerable flow loss, virtually all cores will produce essentially the same efficiency results. Perhaps a core with slightly less flow area per linear inch, or one with longer tubes, will need perhaps 5% more tubes to equal the best of intercoolerís with regard to flow loss and efficiency. Not a very important difference.

What are the differences between short tubes and long tubes?
The longer the tube the greater the pressure loss accompanied by a slight increase in efficiency.

Does the length of the tube affect efficiency?
Very little. The most heat comes out of the tube where the temperature difference between the inside and the outside is the greatest. That exists in the first couple inches of the tube. The last inch of the tube, wherein the charge temperature is rapidly approaching the cooling media temperature, will transfer very little heat, thus being of minor use.

What factors affect efficiency of an Air-to-Air Intercooler?
Frontal area: This is a rapidly decreasing function. If the proper core size is used, then doubling it will definitely not double the efficiency. More likely, doubling the core would raise the efficiency about 5% and cost twice the necessary amount and add substantially to the weight.

Plate area: Plate Area (the sum of the Core-Plate Area which is exposed to the Atmosphere) is directly proportional to the frontal area and the thickness. Thickness, however, is a double-edged sword. With the greater thickness, the plate area increases but less ambient air can penetrate the thicker core to offer cooling.

Ambient air quantity: It is very important to insure that air coming in the snout of the car will actually go through the intercooler.

Are there any improvements that can be made to the system for improved efficiency?
Yes, several small factors influence the efficiency. A proper duct is probably the single most beneficial thing that can be done to an existing intercooler. Positioning in the main stream of ambient air is crucial. By comparison, a taped up intercooler with no ambient air flow will offer only about 20% efficiency.

David 02-11-2010 07:56 AM

One thing I've done which doesn't show in the previous picture is add weather stripping around the edge of the intercooler. I used the widest and tallest dense weather stripping I could to seal the IC to the whaletail. At the front, I stacked the weather strip to make it tall enough to touch the tail.

drmatera 02-11-2010 08:06 AM

So as long as the intercooler doesn't restrict the air feeding the engine fan it would be wise to have the intercooler core sized exactly the same as the "grill" on the tail and seal them together. But it does seem that on a high speed run the intercooler would heat the air that feeds the engine and cause it's own set of problems...

911st 02-11-2010 09:03 AM

Spearco Intercoolers has some pretty neat design info:

Included is some notes like-- pressure loss in intercooler manfolds (end tanks) can vary wildly and can be in the .1 to .3 psi per side; a mandrel bent longer 90 might create a pressure drop of .1 to .3 psi and that a cast tight elbow might cost .3 to .4 psi.

The more pressure loss from the core, turns, end tank design etc the more the turbo has to work to overcome. The higher the boost the turbo much see, the more pressure in the exhaust system...

Just using this info on a 964 factory intercooler it might only see .1 per each end tank or about .2 or so pressure loss.

An end to end might have say .1 for the S up in to the end tank, .2 for each fabricated end tank. .2 for the 90 out of the tank, and .2 for the 90 to the throttle body for about .7 psi or so pressure loss. That is maybe .2 v .7 psi loss.

To this we add the pressure drop of the core itself.

Just a data point, not a major conclusion.

WinRice 02-11-2010 09:14 AM


Originally Posted by drmatera (Post 5179453)
So as long as the intercooler doesn't restrict the air feeding the engine fan it would be wise to have the intercooler core sized exactly the same as the "grill" on the tail and seal them together. But it does seem that on a high speed run the intercooler would heat the air that feeds the engine and cause it's own set of problems...

It does that anyway, to a certain degree. No matter how big or small the IC is, most of that heated air flow coming off of it will find it's way into the engine. Since our engines are sealed to the body, any air coming in the rear grille only has one way out. If the IC is totally sealed to the grill, forcing all the air through, then all the heated air goes into the engine fan. If we only seal it to half the grill then we get a mix of heated air and fresh air. What's the best mix? Who knows without some testing.

Steve Weiner made the comment in a thread some time back, that it's counter productive to pull all that air through the IC then dump all that heated air into the engine, increasing head and cylinder temps. Although, this works for a street car since we aren't creating heated air until short bursts of boost. A track car is a different story.

The ultimate 930 engine would have pure fresh air to the engine fan and air filter, plenty of air flow through the IC for max efficiency, and all this air flow would dump outside the car as not to effect the engine compartment temps.

Wait a second, that sounds like a flat fan 935 with air/water IC's............;):D

WinRice 02-11-2010 09:25 AM


Originally Posted by 911st (Post 5179391)
If going to side to side adds another elbow, 18" of tubing, and more fittings it is going to have an effect. As will end tank design and ducting. Over-sizing or unnecessary volume may also have some cost to throttle response.

Just an observation, but most large front-to-back IC's make the air flow take short tight turns getting into and out of the core area. This has to have some effect on total flow. The side-to-side, if properly designed, looks like you could make smoother transitions into and out of the core, through the use of longer smoother taper tanks.

JFairman 02-11-2010 09:31 AM

When you ditch the oversize blow off manifold that came in these cars for a smaller C2 vlave and hose setup you could mount a shallow puller fan up to around 10" diameter to the bottom of the intercooler.

You could wire the fan to a thermostat switch that turns it on when the intercooler gets to certain temperature or a manual switch somewhere in tha car. It would keep the intercooler from heat soaking at traffic lights and city driving in the summer.

If you live in the south and already have a half dozen fans mounted on existing and added on AC condensors than the car already sounds like a hovercraft in the summer so who cares... other than the little alternator in the fan housing driving them.. then again it's surrounded with a fan.

David 02-11-2010 11:39 AM

Even with fabricated end tanks, you can do some little things to reduce pressure loss and turbulence.

Where the inlet tube enters the inlet end tank, I put extra weld metal so the inside could be radius-ed to improve flow out of the inlet tube and toward the right side of the IC:

On the IC outlet, I machined a radius-ed bell mouth piece to improve air flow out of the outlet end tank, although this could only be done effectively with the 4.5" core:

911st 02-11-2010 01:09 PM


Nice work. I like how you offset the IC a bit so the air cleaner would have a clean shot at air flow through the grill.

About the only way we are going to get to increased IC capacity is to go with fabricated tanks. Still, we might think of paying as much attention to back pressure or pressure loss in our intakes as we seem to want to on the exhausts.

As noted earlier, when thinking about preheating air that is a big no no.

There is a direct relationship between intake air temp and discharge temp. If we suck in 100 deg air instead of 80 deg air our temps out of the IC and into the intake ports are going to be about 20 deg hotter.

We might consider what Porsche did with the 993TT's. They ran dedicated ducts up to the grill to ensure the intake got fresh air that was not pre-heated.

Looking at the C2 turbo's the IC is ducted to the grill and there are openings of limited size just above the intake. I suspect this ensures things are ballanced so intake air is not pre heated.

Surface area of the end tanks and plumbing can account for a surprising amount of the interecooling effect.

Again, I really like with this CIS turbo owner did below.

911st 02-11-2010 02:04 PM

Found this here:ARE Cooling (Aluminium Radiators & Engineering P/L)


Most importantly is that the square edged tanks hurt air flow & increase "static" pressure drop. The intake charge does not want to make a sharp turn with the majority of the air taking a radiused path around the edge, causing some of the air to eddy back around through an elongated 360 degree tumble. This causes friction with both the wall surface & the following air, really hurting flow figures.


...Asking charge air to open out at 360 degrees, as in a 50mm pipe into the middle of a 90mm flat rectangular tank, causes a big pressure drop...

Lots of good info for those interested.

JFairman 02-11-2010 02:17 PM

Interesting stuff. Thanks for posting all these things Keith.

I wouldn't want to get caught in a heavy rainstorm with that cool air duct on the origonal air cleaner housing unless he drilled some drain holes in the bottom.

911st 02-11-2010 02:32 PM


I wonder with Porsche did on the 993TT to keep water out?

Rich Buckner 02-11-2010 07:31 PM

Here is my take on the intercooler topic.
I spoke with Bob Holcomb of Mode on several occasions about the 965 intercooler. He regards the 965 intercooler as a highly efficient and desirable addition to any 930 application. Based on his recommendations, I decided to go the Porsche intercooler rout with my EFI conversion. The only problem was my choice of a Carrera intake manifold.

My solution was to fabricate my own end tank on the existing 965 intercooler. I used a buck of my own design and hand formed the tank. It is not 100% finished, but it is close.

The following photos are the fruit of this insane passion for efficiency, short distance, low pressure loss, stock appearance and ultimately, horse power!

The silicon elbow will dump directly into my Carrera throttle body. I am running a 3.2 Carrera Motronic ECU modified by Todd Knighton (Protomotive) for pressure sensing.

My engine is currently at Ollies being machined. I will post photos of the assembly.

911st 02-12-2010 07:39 AM

It looks like a 965 IC is tube and fin with cool air end at about 60 sq in. If it were a bar and plate that could flow up to 1100 CFM interpolating from if it were a Bell Bar & Plate core. T&F is some percentage of that.

Overall a great intercooler that should be able to support the HP of most radical CIS and modest EFI builds.

911st 02-12-2010 07:52 AM


It is not hard to test an intercooler. Pressure drop and temps and in and out are what mostly tells us how good an intercooler is functioning.

There is something called a differential pressure gauge. It shows the difference between two connections. Or just use two MAP sensors.

I would think we would want to measure at the turbo discharge and just before the throttle body. After the throttle body would be good to but then any pressure loss from the TB would be included. This would include the intercooler plumbing and tanks in the measurements.

Temps are usually measured at the in and out of the intercooler.

Same with turbos and exhausts. Good tuners measure the pressure reached in the area before the turbine and sometimes before and after a turbine to see if what they have built is working as it should. Mufflers can be tested to pressure loss to.

Ambient compared to turbo output temp lets us know if the turbo is operating in an efficient range.

All times are GMT -8. The time now is 04:24 AM.

Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.
Search Engine Optimization by vBSEO 3.6.0
Copyright 2018 Pelican Parts, LLC - Posts may be archived for display on the Pelican Parts Website

DTO Garage Plus vBulletin Plugins by Drive Thru Online, Inc.