slcsubaru.com

Well to Actually know you need to be as software engineer, and if you knew that you could be working for Cobb ;)

So the MAP does have two ports? One for atmospheric, and one for manifold?

Depends on the car.
Early JDM cars had 2 ports. I believe the newer cars only have one. But they sample the air right before the key is turned over to start the car.

Hi guys, this is my first post on here, I'll make an intro post in a bit. I do a fair amount of tuning and one of my friends recently moved out to SLC and asked if I had any info that might be relevant to this thread.


I think it's easier to understand the impact of altitude on turbocharged Subarus if you start basic and move up to more complex concepts. As has been mentioned in the thread already, atmospheric pressure decreases with altitude. The result of this is that the density of any given volume of air, and subsequently the amount of oxygen that volume of air contains decreases. Atmospheric pressure at 4,200ft is roughly around 12.6 psi as opposed to 14.7psi at sea level. This is obviously part of the reason that all cars make less power at altitude.

Turbocharged cars can compensate for some of this by compressing the intake charge to above atmospheric pressure, however there are 2 major things to consider when thinking about why they make less power than an equivalent car at sea level. First of all, the ecu contains atmospheric pressure boost control tables which decrease target boost pressure as atmospheric pressure decreases. As RallySportDirect mentioned (nice post by the way) the MAP sensor either has 2 ports or it samples atmospheric pressure at startup. As a result, when the ECU detects decreased atmospheric pressure the ecu automatically dials back the target boost.

Once you know that the ecu is decreasing the target boost level in response to ambient atmospheric pressure it's easier to process what is actually happening to the engine in terms of power output.

1.) Atmospheric pressure is decreased by about 2 psi
2.) Target boost levels are decreased due to atmospheric pressure correction tables (amount depends on tune)
3.) The result of 1 and 2 is that the Absolute manifold pressure in the manifold is decreased due to the lower atmospheric pressure and also by the lower boost pressure (atmospheric pressure + relative manifold pressure = Absolute manifold pressure)

Lets take 2 identical boosted Subarus with 16bit ecus, identical tunes, both with a non-zeroed atmospheric pressure correction table. The first is at sea level and the second is in Salt Lake City. Since the Atmospheric Boost Compensation table values will depend on the tune, we'll use the value that is currently mapped on my DD. My car decreases target boost by 12% (so a car that starts at 26psi target boost ends up with a target boost of 22.8 psi) at an ambient atmospheric pressure of 12.6 PSI.

Sea Level: Atmospheric pressure (14.7) + Target boost pressure (26 psi) = Absolute manifold pressure (40.7 psi)

SLC: Atmospheric pressure (12.6) + Target boost pressure (22.8 psi) = Absolute manifold pressure (35.4 psi)

From the above illustration it's easy to see that turbocharged cars take a hit on both ends with respect to power output at altitude. The one saving grace here, is that if you have the fuel or methanol injection to provide a bit more knock resistance in the face of a hot intake charge you can zero out some of the altitude correction in the boost correction table and eliminate some of the decrease in target boost associated with altitude.

There's been some discussion of variations in % oxygen concentration (irrespective of air density) due to variations in gravitational field strength with altitude. From a theoretical standpoint this is absolutely correct, however at 4,200 feet the effect is fairly minimal, especially when you consider the fact that 99% of the molecules in the atmosphere have a molecular mass of ~30g/mole (28g/mole for nitrogen and 32g/mole for oxygen). The fact that the two primary consituents are fairly close in mass and that SLC in absolute terms is still quite low in the atmosphere the effect of gravitation on atmospheric composition is relatively minor. This could be determined mathematically, but I'm not quite that bored right now :) On top of that, the air mass surrounding SLC is not static and wind currents will generate enough mixing of air from lower altitudes to further negate the gravitational effect on composition.

So I think the main goal of this thread was to determine the ECU's strategy for determining target boost with respect to ambient atmospheric pressure. To that end, the below equation is what the 32bit 2004 STI ecu uses to correct target boost pressure for variations in atmospheric pressure:

(Atmospheric pressure) x (Multiplier determinate) = X

X + (Mult. Offset) = Y

(Initial target boost) x (Y) = Final adjusted boost target

The Multiplier determinate and Multiplier offset are both mapped values and thus the final change in boost target is totally dependent on the tune. Additionally the way the car calculates the alteration in final target boost due to atmospheric pressure varies significantly from Subaru to Subaru depending on which ecu is used.

In the 16bit WRX ecu the final target boost is calculated by the below equation:

(Initial target boost) x (Target boost compensation %) = Final adjusted boost target

In the 16bit ecu the Target boost compensation is a mapped value which is a % change value based determined by what the ambient atmospheric pressure detected by the MAP sensor at startup. So, as you recall from the prior calculation for the 16 bit wrx in SLC at atmospheric pressure of 0.85 bar the target boost compensation value is 12%.

I hope that helps!

That's a helluva first post.

I agree! I like it!

Thanks guys, I hope it didn't come off as preachy! I've been following the thread for over a week now, but I wasn't able to make a new account because of the Ride of the Year voting! Anyway, I really like the forum so I hope someone finds the post useful.

I did. It is nice that someone is willing to share that kind of info. Where are you located? I want to learn more!

I live in the same town that Geof (Rectangular) just moved to SLC from. Marquette, MI to be specific. I'll be moving for a new job in June, but I won't know exactly where until March :)

I'll try to post up whenever I have useful information to share on a topic.

So, I guess there's a little more to describe as far as the physics that underlies the function of a turbocharger. One of the other initial questions posed in the thread was that a turbocharger could compensate totally for the difference in atmospheric pressure.

"I maintain that once a turbo has reached target boost, and it is not out of it's compressor efficiency map, altitude doesn't have much of an effect. Once the wastegate opens, the turbo has more than enough air to reach it's goal."

While this seems plausible and is sound logic when you think about it at first this isn't the whole story.

If you've ever spent some time digesting or trying to understand turbo compressor maps you'll be familiar with the concept of pressure ratios. Essentially compressor maps describe the correlation between turbocharger efficiency and the pressure ratio they are producing and various air flow levels. Generally speaking most factory subarus are pretty close to leaving the efficient ranges of their compressor maps if they're pushed even to stage 2 at sea level.

Pressure ratio essentially describes the ratio between Ambient Atmospheric Pressure and the Absolute Manifold Pressure. So if a turbocharged car is at sea level (14.7psi) and is running 14.7 psi of boost the turbo is producing a pressure ratio of 2 (14.7 + 14.7 = 29.4 and thus 29.4/14.7 = 2).

So, as the pressure ratio increases (which happens as boost is turned up) the turbo will begin to leave it's compressor map (the boost pressure at which that happens is dependent on which turbo and what flow rate is being discussed). At any rate, this has important implications with respect to altitude. As altitude increases and ambient atmospheric pressure subsequently decreases the boost pressure at which any specific pressure ratio is achieved decreases.

That's a mouthful so here it is in mathematic form:

Sea level: (Atmospheric pressure 14.7 psi) + (Boost pressure 14.7 psi) = 29.4psi Absolute Manifold Pressure
Thus: 29.4/14.7 = 2 (pressure ratio)


SLC: (Atmospheric pressure 12.6 psi) + (Boost pressure 14.7 psi) = 27.3psi Absolute Manifold Pressure
Thus: 27.3/12.6 = 2.16667 (pressure ratio)



Now, the pressure ratio at which the car really starts to fall off the high efficiency islands in the compressor map is going to be variable depending on the turbocharger, but as a general rule, for the boost pressures we most often discuss when tuning Subarus the higher the pressure ratio the more inefficient the turbo becomes.

This has two implications for cars at altitude.

1.) The compressor functions more inefficiently and thus the compressed air is hotter.

2.) The turbine must extract more energy from the exhaust gases to produce any given boost level.

The results of this are that the car will have a less dense intake charge at any given boost pressure (due to the heat), it will be more prone to knock due to the hotter intake charge and the turbine will generate more of a flow obstruction in the exhaust and increased back pressure due to the fact that it is being forced to extract more energy to produce the same boost level.

The decreased charge density and hotter (more knock prone) intake charge make it impossible, or at least more difficult to tune a car to make the same power as at sea level when the altitude increases and the increased exhaust flow obstruction from the turbine means that even if the intake charge was identical to the intake charge at sea level the car would make less power due to the decrease in exhaust flow.


Anyway, I'm posting this from Africa and it's time for me to go to bed now... If any of this doesn't make sense I'm going to blame it on how sleepy I am. I'll check back on here tomorrow and answer any questions (if I'm capable of answering them), or hopefully revise any ridiculous typos that I've made tonight!