These days, most people know that placing an exposed airfilter under the bonnet is just asking for the engine to take big breaths of hot air. To avoid that outcome, it's now common to fit a cold air intake - often abbreviated to 'CAI' - and seal it to the airbox.
However, what's generally not realised is that if a positive pressure can be generated in the cold air intake, the engine will perform better again. Actual on-road measurements that we have made ("Eliminating Negative Boost - Part 5") have shown that at speed, a good cold air intake can easily have sufficient air pressure within it to completely cancel the flow restriction of the aircleaner. So, even though the effect of ram air on performance isn't huge, there's a worthwhile gain that can be made.
But getting that pressure build-up in the duct requires that the mouth of the cold air intake be sited in an area of high pressure.
And where's that, then?
As the car moves forward, air is deflected above, below and to each side of the car. As an example, take the flows occurring along the centreline of a sedan that has a very good drag coefficient (ie it's slippery) like the pictured early Lexus LS400. Some of the air is deflected over the top of the bonnet, passing along the bonnet, flowing up the windscreen, along the roof, down the rear window and then leaving the car's body at the trailing edge of the boot. (To keep the flow attached to the body in the described manner, the car must have gentle changes of angle only - especially from the roof to the rear glass to the bootlid.)
Other air is deflected beneath the car, flowing along an engine under-tray and then getting mixed with the turbulent flows exiting the engine bay (the air going thorough the radiator has to get out somewhere!) and the other turbulence created by the exhaust and suspension, not to mention the spinning wheels.
But back to the front of the car...
At the front there must be a point above which the air goes over the car, and below which the air goes under the car. This is called the stagnation point, and it is here that there will be the greatest pressure developed on the front of the car as it moves forward. This graphic of a Mercedes shows the high-pressure stagnation area in red, lesser high-pressure areas in green, and the low-pressure areas in blue. (They're low pressure areas cos the airflow is being accelerated around a curved surface at each of those spots.)
And hey, that's all very well and good - but how do you find where the stagnation point is on your own car? And what if it's impossible to site the intake for a cold air duct there anyway? How do you rank the quality of other potential sites?
Finding the pattern of the positive pressures acting on a car is easy and can be carried out very cheaply. The testing can be done in a number of ways, including using a manometer, Dwyer Magnehelic gauge, or as we've done in this particular story - using a very sensitive pressure switch.
The last approach has some advantages - it's very cheap, very easy, and there's no likelihood of manometer fluid going all over the carpet!
The switch that is used is the Micro Pneumatic Logic Pressure switch available in the AutoSpeed Shop for $11.95. An extraordinarily sensitive switch, it is factory set to trigger at about 1 kPa (0.145 psi) and is perfect for this application.
The test procedure is simple:
- Buy some small diameter plastic hose that fits tightly over the pressure switch nipple. I bought 5 metres of clear plastic hose from the local hardware store for 70 cents a metre.
- Place the open end of the hose at the location that you are investigating, then run the tube back in the cabin, holding it in place with pieces of good quality masking tape. (Good quality so you don't harm your car's paint!)
- Push the other end of the hose over the pressure switch.
- Connect the pressure switch to a buzzer, power and earth so that it sounds when the switch is subjected to pressure. The pictured AutoSpeed Variable Volume Buzzer costs $4.95 and is fine for the application. Alternatively, if you have a multimeter with an audible continuity function, you can connect this to the pressure switch, so that the meter beeps whenever the switch is subjected to pressure.
Incidentally, if you're wondering how to go about connecting the wires to the pressure switch (it uses odd, small spade terminals) all that you need to do is pull off the back mounting section, revealing the terminals which can then be easily soldered to.
When you have that all set up, go for a drive, with someone else in the car to write down the results.
Because the pressure at which the switch clicks over is preset, you can't directly measure the different pressures acting on the front of the car. But you can still work out really well what's going on.... How?
What you need to do is monitor the speed that you're travelling at when the buzzer sounds.
So let's take the example of the car used here. The '91 Lexus LS400 has a section of bumper beneath the large headlights where there is the appearance of bars forming a grille, but in fact the bumper is solid. It was my guess that this would prove to be a high-pressure area, and so it proved. What happened was that with the open end of the hose positioned in this spot, the buzzer sounded at about 85 km/h. (Note that you want to do the testing on a fairly windless day, and be wary about following cars too closely as this noticeably upsets the results!)
So at around 85 km/h there was enough pressure being generated at this location by the force of the airflow to trigger the switch.
Next I moved the pressure tap to just behind the radiator grille, between the headlights. At 85 km/h the buzzer didn't sound, nor did it at 90 or even 100 km/h. In fact the car had to be moving at 110 km/h before the buzzer came alive. That same result was also gained with the probe placed against the front of a headlight.
So, since the Lexus had to be moving much faster to generate the same pressure, in general the pressure across the upper grille and headlights was lower than at the first probe position.
The testing therefore showed that an air intake positioned beneath the main part of the bumper would work well - although of course with the intake in this location, the possible ingestion of water lying on the road should also be considered. (Air-borne water normally causes no dramas.)
But what about at the base of the windscreen? Since time immemorial the intake to the cabin ventilation system has been sited here, because this is generally a high-pressure area. It's also common to place the mouth of cold air intakes through into this plenum volume - in fact we've covered one such approach at "Free-Flowing a Miata MX5"
But how did this location actually compare for pressure with the best spot already found? In the Lexus it is way poorer: with the probe placed a little off-centre at the base of the windscreen, the car needed to be travelling at 150 km/h before the switch triggered. However, the pressure was a little higher at the mid-point of the windscreen - 145 km/h triggered the switch.
So the pressure measured at the base of the windscreen was actually the lowest yet recorded.
||Speed Needed to Trigger Switch
(higher = lower pressure being found)
|Blocked grille in lower front bumper
|Radiator grille between headlights
|Surface of headlight
|Base of windscreen, centre
|Base of windscreen, off centre
And what about the standard Lexus cold air intake? The LS400 uses an over-radiator intake, common in many cars. Forward-facing, it is a preferred factory location because it places the intake well above water on the road, conceals its mouth from the direct pick-up of dust, and has the potential to gain some positive pressure. But was the LS400's standard intake in fact developing any positive pressure?
With the probe placed in the intake, some further testing was undertaken. At 160 km/h, full throttle, there wasn't sufficient pressure build-up to trigger the switch, but immediately the throttle was lifted at these speeds the buzzer sounded. What was happening was that the massive amount of air being demanded by the 4-litre V8 meant that air passing into the duct was being drawn straight into the engine - and not building up a pressure. However, with the throttle closed, the pressure build-up was obvious. In fact, with the engine drawing only idle air consumption, the buzzer sounded all the way down to 110 km/h.
This showed that the duct, which draws air from just behind the radiator grille, was being subjected to the same positive pressure recorded earlier at that location.
So yes, the standard duct is in a pretty good position - not the best possible for ram air pressure, but taking into account those aforementioned compromises about dust and water, yep, pretty good.
Spending under twenty bucks and an hour of your time can give you a clear picture of some of the aerodynamic pressures acting on the front of the car. That'll mean that you can locate the engine intake to make the most of the free power often going to waste. You'll certainly avoid making some of the glaring errors that we occasionally see... like drawing air straight from the wheel arch - a low-pressure area in many cars!
But keep in mind that once you've found the best position for the mouth of the intake, seal it all the way back to the filter. You don't want that precious pressure just dissipating itself in the engine bay....
Heat Exchanger Flows
Of course, the pressure switch technique can be used for a lot more than just finding the best place to put a cold air intake. One other use is to assess the flow of air through a radiator - or intercooler.
As mentioned above, the pressure switch clicked over at 110 km/h when the probe was placed in front of the radiator of the Lexus. But what was the pressure like behind the radiator? You see, if enough pressure builds up there, no flow though the heat exchanger can occur at all. Of course, the designers make sure that there are sufficient air exits behind the radiator that the pressure is always lower than in front of the core - but the same isn't always the case with aftermarket intercooler installations!
Out of interest, I placed the probe behind the radiator and found the speed at which the switch clicked over - 180 km/h. That compared with 110 km/h in front of the radiator. So, the pressure build-up behind the radiator was well below that occurring in front of it...