Shopping: Real Estate |  Costumes  |  Guitars
This Issue Archived Articles Blog About Us Contact Us

Getting Air In & Out

Looking at atmospheric flows through intercoolers.

by Julian Edgar

Click on pics to view larger images

Like other radiators, an intercooler works by exchanging the heat of the fluid passing through it (in this case air) with the ambient air being forced through the core by the forward movement of the car. Like many apparently simple things, this is a lot more complex in the reality than the brief description implies. For example, there needs to be turbulence within the intercooler tubes if all the air is to come in contact with the walls of the tubes - that's why many intercooler tubes have swirl-inducing spiral metal strips placed within them. The flow of air through the tubes also needs to be as even as possible, the reason that some OE intercoolers have interesting shapes to their tanks and entry/exit pipes.

However, in this story we won't be looking at any of that. Instead, we'll concentrate on getting outside air through the core.

It's an obvious but frequently overlooked point: if outside air isn't flowing through the intercooler core at as high a rate as possible, intercooler efficiency will be lower than it could be. This means that locating the intercooler without giving any thought to directing the maximum flow through the core (and also considering how all of that heated air is then removed) will often result in poor intercooler efficiency.

Intercooler Flows

Click for larger image

It is a good idea to look closely at how manufacturers have sited their intercoolers, and the way in which it is intended that outside air both arrives at the core and then departs from it. Remember that we're not talking about the combustion air on its way from the turbo or supercharger to the throttle body; instead it's the outside air that we're concerned with.

A few points need to be kept in mind.

  1. In front of the core generally a higher than ambient air pressure needs to be generated. In other words, most often the intercooler needs to be sited so that air is rammed into its front face. This is usually achieved by placing the core at the stagnation point of the frontal flow - ie where the airflow must separate, with some passing over the roof of the car and the rest moving under the car. It is at the stagnation point that the air pressure generated by the forward movement of the car is greatest.
  1. The pressure behind the intercooler core must be kept as low as possible. In fact, in some cars it is a pressure lower than ambient that causes the flow to occur through the core. (eg in a turbo Porsche where the intercooler is placed at the back of the car.) In any case, there must be a free exit path for the major amount of air that has been pushed into the front face of the core, allowing it to escape out the back. This is a vital but often overlooked point, but it's made easy to understand if you consider how poorly the airflow through the core would occur if you taped a piece of cardboard over the rear of the intercooler.
Click for larger image

While many manufacturers place their intercoolers in front of the radiator at the stagnation point, others (for packaging or other reasons) cannot do this. Instead they site the intercooler on top or behind the engine (eg Pulsar GtiR, Impreza WRX), or to one side of the radiator or forward of one of the front wheels (eg Daihatsu Charade GTti, Skyline R33 2.5, Audi S4).

Click for larger image
In the case of the top-mount intercooler, air is directed to the core by a bonnet scoop that seals against the intercooler when the bonnet is shut. (In some cases, air is picked up at the front of the car and directed by means of an underbonnet duct to the engine-mounted core.) Behind the core, air flows out past the engine and into the low-pressure under-car area, dispersing in the wake behind the car.

Click for larger image
Where the intercooler core is placed to one side of the radiator, invariably the core is fed by a purpose-built duct that picks up air from a frontal opening. Behind the core, the air disperses into the wheel-well, a low pressure area on most cars. However, in some cases the exit path for the air isn't at all obvious - a guard liner is usually present to separate the tyre from the in-guard area and this can block much of the flow leaving the intercooler.

Assessing Intercooler Flow

There are three ways of assessing the airflow that is passing through an intercooler.

  1. Physical examination

Look carefully at how air is reaching the core. Are there easy escape routes? Air will take the least obstructive path, so if the intercooler is sited so that it is easy for the airflow to make its way around the core - rather than through it - that's just what will happen. For example, a large front-mount core should have either end (and the top and bottom) blocked off to prevent the bypass flow of air. This blockage can be achieved in some cars with foam rubber strips (spray-paint them black) which can be easily inserted to fill escape gaps. Flow director plates made from thin sheet aluminium or steel can also be used to train the flow to head in the right direction.

Click for larger image

Once you've looked at how air arrives at the core, check how it departs. For front-mount cores, does the escaping air have to pass through the radiator? Is there an easier exit route that can be opened up? (Obviously you still want air to be available to pass through the engine cooling radiator, so what you do here depends on how big the intercooler core is, and how else pressurised air can reach the rad.) One very effective way of promoting airflow through a front-mount core is to provide an opening on the bonnet following the radiator and intercooler. The flow of air over the top of the bonnet will help draw air out of the paired heat exchangers.

With intercoolers that are mounted to one side of the rad, look in particular at the how the air can depart the scene. Is the guard liner sealed up tight? If so, the exit path of the air is often blocked in a major way - it has to squeeze out through any openings around plumbing and bolts. We'll come back to how to improve this in a moment.

  1. Wool Tufting

If you don't really have any idea where the air is moving at the front of the car (or at the entrance to a bonnet intake), stick lots of small tufts of brightly-coloured wool over the area of interest. Get someone else to drive the car at speed and inspect the pattern of tufts from a nearby car. (Doing it on a multi-lane road makes this easy.) What you want to see is laminar (non-whirling) tufts disappearing into the mouth of the opening. If there is turbulence at this point, or the direction of the tufts show that lots of air is bypassing the intake to the intercooler, you at least know where to place the flow-straightening and directing panels.

Wool tufting the air exit of the intercooler is more difficult. However, while we haven't tried it, the use of a remote miniature TV camera (now available cheaply from electronics stores) together with a 12V TV or even video camera with 'video in' facilities, should still let you view what's going on under actual operating conditions.

  1. Pressure Measurement

As indicated earlier, if there is a positive pressure differential across the core, air will flow through it. The higher that pressure difference, the more air that will flow. When referenced against the atmospheric pressure, you can have:

In Front of the Core with... Behind the Core equals... Flow
High pressure   Atmospheric pressure   Good Flow
High pressure   Below atmospheric pressure   Excellent flow
Atmospheric pressure   Below atmospheric pressure   Good flow
High pressure   High pressure   Bad flow
Atmospheric pressure   Atmospheric pressure   Bad flow

There are other combinations as well, but these are the most likely.

Click for larger image

Measuring the actual pressures is quite easily done if you have a Dwyer Magnehelic Differential Pressure Gauge, an instrument that we have already covered in detail at "Eliminating Negative Boost" All that you need to do is to run a tube from the gauge to the area where you wish to measure the pressure (either above or below atmospheric - just select the right port on the gauge) and then drive the car. The pressures will build more quickly as you go faster; usually you'll need top be doing at least 100 km/h before they become easily measurable. (That doesn't mean that nothing's happening at slower speeds, but the bigger the pressure variation, the easier it is to measure.)

Instead of the Magnehelic gauge you can build a U-shaped manometer that will measure both positive and negative pressures - basically, just a U-shaped tube part-filled with water and with the top of one arm connected to the area of interest. When held vertically, the pressure difference will either push or pull the fluid up (or down) the arms.

Doing It

My 1993 Audi S4 uses an intercooler core placed to one side of the engine, in the forward portion of the wheel-arch. It is fed air by a dedicated forward-facing duct that is sealed to it. The pressure in front of the intercooler was measured at 0.35 kPa at 120 km/h. When the probe was placed behind the intercooler (ie in the space between the intercooler core and the inner guard lining), a pressure of 0.25kPa was measured at the same speed. To try to lower the pressure build-up behind the intercooler, the following modification was undertaken.

Click for larger image
As standard, the inner guard lining behind the intercooler core appears to be completely sealed.

Click for larger image
The guard lining was removed (much safer than trying to do it on the car!) and a 2.5-inch hole saw used to cut a hole that would form one curved end of the opening.

Click for larger image
The other end was cut in the same way and then a knife used to join the two holes, forming a rectangular opening with rounded ends.

Click for larger image
A stainless steel grille was obtained from a boating store. Steer clear of plastic and painted steel grilles - these are sure to become brittle and eventually break or rust, respectively.

Click for larger image
The grille was bent a little to match the contours of the slightly curved guard liner and then pop-riveted into place.

Click for larger image
While very obvious in this photo, after a few weeks a coverage of road grime makes the previously shiny grille blend in with the black guard liner.

Unfortunately, after all of this had been done, the measured pressure behind the intercooler remained exactly the same at 0.25 kPA at 120 km/h! Hmmmm. There were a number of possible reasons for this:

  • The grille being used didn't have sufficient capacity to get rid of the air flowing through the intercooler.
  • The pressure inside the wheel-arch was higher than atmospheric.

Measurement was then made of the pressure inside the wheel-arch... something that perhaps should have been done before any of the mods were done! But this showed a zero pressure build-up, so the reason that a positive pressure could still be measured after the intercooler was because the flow capacity of the grille was insufficient. In retrospect, this makes sense - the grille openings are collectively far smaller than the intercooler feed opening at the front of the car.

At this stage I have not replaced the grille with a larger one - the difficulty is in finding a grille that will flow huge quantities while still preventing mud being thrown from the wheel onto the back of the intercooler.


The atmospheric airflow into and out of an intercooler core will help determine its efficiency. Escape routes that air can take around the core should be sealed off, and attention should be paid to ensuring that good exit routes are present. A Dwyer Magnehelic Differential Pressure Gauge or U-shaped manometer can be used to measure actual pressures occurring at the front and back of the intercooler. If modifying the exit route for air, keep in mind that you are dealing with major quantities of air.

Did you enjoy this article?

Please consider supporting AutoSpeed with a small contribution. More Info...

Share this Article: 

More of our most popular articles.
Assessing the performance of an amazing human powered vehicle

DIY Tech Features - 15 January, 2008

Air 150 Recumbent Trike, Part 2

Do hybrid cars make sense?

Technical Features - 16 October, 2007

Alternative Cars, Part 6 - Hybrid

Got an old cordless drill around the place? Here are the parts you can salvage from it!

DIY Tech Features - 8 May, 2008

A Heap of Parts for Nothing!

Cheaper than a half-cut and lots more bits!

DIY Tech Features - 17 April, 2012

Buying at Salvage Auctions

Buying and using a lathe

DIY Tech Features - 29 November, 2007

Making Things, Part 8

Want to build your own home workshop? Here's how to begin.

DIY Tech Features - 12 August, 2008

Building a Home Workshop, Part 1

Designing structures so they won't fail

DIY Tech Features - 21 February, 2006

Making Things, Part 1

Form beads in aluminium, steel and stainless steel intercooler plumbing

DIY Tech Features - 11 February, 2014

Make your own tube beader

...but it never turned a wheel in anger

Special Features - 8 April, 2014

The fastest Mercedes of them all?

The 1100hp Porsche 917

Special Features - 18 April, 2003

The Early Days of Turbo Part 3

Copyright © 1996-2020 Web Publications Pty Limited. All Rights ReservedRSS|Privacy policy|Advertise
Consulting Services: Magento Experts|Technologies : Magento Extensions|ReadytoShip