Last week we discovered that hiding inside the Audi's intake system were some heinous, hairy negative boosts. These critters are masters of disguise, masquerading as semi-normal shapes and features when in fact they're depraved mutants that deserve only to be exterminated in an utterly horrible way. In the past I've dealt out death by hacksaw, death by pliers, death by sanding belt - you can be sure of one thing, when I find them, I give them no mercy. Not because I like to see them in pain, you understand, but because it usually takes heavy-duty tools to do the extermination job. Or that's what I tell the RSPCA anyway....
The Audi
A total full-load pressure drop through the Audi's intake system of 32 inches of water is pretty fearsome. It's not as bad as a standard Subaru Liberty RS (which records a pressure drop of 31 inches of water at only 5000 rpm - still with 2000 rpm to go!), but it's still pretty woeful. Much worse, for example, than a 2-litre Holden Camira that records a max pressure drop of only 9.6 inches of water through its standard intake system. (See the breakout box at end of last week's article.)
| RPM |
Total Pressure Drop
(inches of water) |
| 2000 |
2 |
| 3000 |
10 |
| 4000 |
16 |
| 5000 |
26 |
| 6000 |
32 |
| 7000 |
29.6 |
So the Audi's Total Pressure Drop table (above) tells a sorry tale of restriction to flow. In there are those varmint negative pressures, clawing away at the turbo 5-potter's power in an underhand and usually hidden way. The poor engine is gasping for air - the synoptic Low that the Audi's permanently in is literally as intense as the eye of a tropical cyclone.
But what's causing the restriction? Last week's testing showed that most of the negative pressures were between the airfilter and the turbo - so the culprits could be any (or all) of the following: the exit duct from the airbox, the airflow meter, or the hose that connects the airflow meter to the turbo compressor. Of these possibles, I guessed the airflow meter.
The Airflow Meter
The Audi's airflow meter (MAF) is a typical Bosch hotwire design. But it's a pretty big one, being 75mm (3 inches) in diameter. At each end is a dense wire screen, with the individual openings in the screens quite small. I'm not sure why the screens exist at both ends; on the intake side it's there to stop damage occurring to the hot wire if the airfilter element fails. However, if the filter gets a hole in it and lets a rock through, the damage to the turbo is going to be pretty horrible, so I guess the airflow meter would then be the least of my worries...
Closely woven mesh causes a major flow restriction. This is not just because of the blockage factor of the wire itself, but because of turbulence which is generated as the air squeezes through the small openings. In the Bosch meter used in the Audi, the screens are held in place with wire circlips which require only fingernail pressure to release. Once you've done that, the screens fall out - totally undamaged.
Mmmmmm, where're my fingernails?
Thirty seconds later, the unscreened airflow meter was back in the car. Time to hit the road and do some more pressure testing.
| RPM |
Total Pressure Drop
(inches of water) |
Total Pressure Drop
Screens Off
(inches of water) |
Improvement |
| 3000 |
10 |
10 |
0% |
| 4000 |
16 |
14 |
13% |
| 5000 |
26 |
20 |
23% |
| 6000 |
32 |
26 |
19% |
| 7000 |
29.6 |
24.8 |
16% |
Well, there sure was an improvement - I'd found and killed one (two in fact) of those bloody negative boosts. But the improvement wasn't as much as I'd hoped for - so how much was it? Up until 4000 rpm there was no change - which is to be expected because the flow restrictions at low rpm aren't much anyway. But from there upwards, an improvement could be seen. The peak pressure drop was reduced (and that means the flow improved!) from 32 to 26 inches of water, a 19 per cent reduction. But unfortunately 26 inches of water is still a helluva restriction - I hadn't found and killed the chief honcho negative boost.
But - hell - I'd still made a measurable improvement. Slowly, slowly catchee monk-, er negative pressure. And what I didn't tell you is that last week I measured the actual on-road acceleration of the Audi through second gear. Using an accelerometer, you can measure how hard the car is accelerating every 1000 rpm, giving you a really good picture of what's going on with the actual on-road performance. And did it go better with the airflow meter screens removed, as you'd expect given the reduced pressure drop? Yes it did. At 7000 rpm the car could previously pull 0.25g - with the screens removed that increased 28 per cent to 0.32g. And that new-found top-end grunt was obvious on the road.
See, doubters, sceptics, ye of little faith? Find and reduce negative boosts and your car will go harder!
But there were still some other deviant negative boosts hiding in there, no doubt plotting their revenge at the loss of two of their brethren. Best to find and exterminate the others as well!
Airbox Exit
I looked and looked at the system between the airbox and the turbo. Apart from the airflow meter (now as good as it was gonna be) all I could see that might possibly be causing a pressure drop was the exit duct in the airbox. I thought this pretty bloody doubtful, though. The 70mm (2? inch) duct has a full bellmouth and also a X-shaped flow straightener inside it. Extending 80mm into the airbox, all that I could remotely see as a negative boost nasty was the fact that its mouth came fairly close to the opposite wall of the airbox. It seemed awfully unlikely that Audi engineers had made it just a bit too long, but that was literally all I could think of that could be still causing over 15 inches of pressure drop between the airbox and the turbo. Everything else looked as sweet as.
Rather than chop up this factory piece, I made a new bellmouth from plastic pipe about 80mm in diameter. Using a heat gun to soften the plastic, I formed a bellmouth at one end and expanded the other end slightly so that it became a firm push fit into the airbox. It didn't meet up as nicely with the airflow meter as the factory item (there was a slight lip present), but at least the intake mouth was now well clear of the opposite wall of the airbox. More of a worry was that I didn't have much room for a flow straightener before the airflow meter - so on this trial one I left it out. That could result in rich mixtures as the airflow meter read turbulence as extra flow - but at least the pressure drop measurements would indicate whether the factory bellmouth exit duct was the problem. If in fact testing proved that I was leaving a pressure drop nasty behind on the workbench, well, I could do a better job of its replacement when I knew that for sure.
But it was not to be. Whereas with the airflow meter screens removed, the peak pressure drop was 26 inches of water, with my new bellmouth in place (and still with the screens off) this rose to 30 inches of water! Hmmm, the different bellmouth was making an extra 4 inches of water pressure drop - much better to go back to the factory item.
So what could be causing the restriction?
The Rubber Duct?
I took off the rubber duct that connects the airflow meter to the turbo. Was the smelly, hairy negative pressure hiding inside here? The duct has an 80mm opening at the airflow meter end, while the turbo end is 55mmm in internal diameter. Along the way there are entrance points for the blow-of valve, and the PCV return (which I'd been using as the pressure sensing point). In addition a smaller hose from the wastegate control valve also joins this duct.
From a flow point of view, the worse aspects I could see were: (1) the PCV and blow-off valve connections were right on the bend in front of the turbo, where the duct was at its smallest diameter, and (2) about one-third of the duct was corrugated with large internal ribs, designed to allow the duct to flex when the engine moved. Could these corrugations also be causing a pressure drop?
Rather than make a completely new steel or alloy intake duct (which would take quite a lot of time and work) I quickly made one from PVC plastic elbows, blending smoothly from an 80mm bend down to a 55mm bend. This assembly was held together with electrical tape - it was just durable enough to allow one burst of on-road testing. However, the duct didn't have any pressure tap points and so I used the accelerometer as a quick guide as to whether the car was going harder. And was it? Well, kind of. Working without an assistant this time, all I could do was shoot very quick looks at the accelerometer as I used full throttle in second gear through the Saturday afternoon traffic! And it looked like it was holding 0.4g a little longer than in standard trim.
There was nothing for it but to make a new, proper, intake duct. Or was there an alternative? What if in fact it was the airflow meter that was still causing all of this restriction? What about placing another pressure tap directly after the airflow meter - rather than at the end of the duct close up against the turbo? That way I could narrow down what was going on.
Re-Visiting the Airflow Meter
Rather than put a hole in the rubber duct straight after the airflow meter, I used a short section of plastic pipe to space the end of the duct a little away from the meter. The pressure tap point was then put in this spacer ring, which was held in place at one end by the rubber duct's clamp and at the other, by electrical tape. So what were the readings were made here? To keep things simple, I'll just list the peak pressure drops that were now occurring in this screens-removed form.
| Pressure Tap Location |
Peak Pressure Drop |
| Total at end of turbo duct |
26 |
| After airflow meter |
20 |
| Before airflow meter and airbox exit bellmouth |
10 |
So the rubber duct between the airflow meter and the turbo was causing 6 inches of water pressure drop, and the airflow meter (with screens removed) and airbox exit bellmouth were still causing 10 inches of water pressure drop. In other words, of the 16 inches of water still being lost between the airbox and the turbo (and remember, that's over half of the flow restriction of the complete intake!), 37 per cent was from that convoluted rubber duct, and 63 per cent was being caused by the airflow meter and airbox exit duct. I knew I couldn't improve the airbox exit bellmouth - remember, I'd tried unsuccessfully - and replacing the airflow meter wasn't going to be cheap.
If I made a new intake duct and then reduced the pressure drop in front of the airbox as much as was possible, could I pick up enough gains so as to not need a replacement airflow meter? Hmmmmm, if I halve the airflow-meter-to-turbo rubber duct restriction and reduce by three-quarters the flow restriction into the airbox - that'll give me a total pressure drop of around 15 inches of water, versus the starting point's 32 inches. Hmmmmm. Better do everything else first before playing with different airflow meters.....
Key Points:
- The greatest flow restriction of the intake system was occurring between the exit of the airbox and the turbo.
- The airflow meter was causing the majority of this restriction.
- Removing the airflow meter screens made a major improvement.
- The airbox exit bellmouth could not be easily improved upon.
- There still remains an unacceptable pressure drop between the airbox and the turbo.
Next week: making a new airbox intake duct.
Eliminating Negative Boost - Part 1
Eliminating Negative Boost - Part 2
Eliminating Negative Boost - Part 3
Eliminating Negative Boost - Part 5
|