Why does exhaust leak cause lean condition?

[MistaScott]

I think this is an important question that I was having trouble with theoretically.

If you are leaking exhaust through a crack, that wouldn't introduce new air into the system. You would just end up passing less exhaust past the o2.

If you are pressurizing the exhaust system, that would still push out exhaust. Only under vacuum would you suck in air and exhaust is clearly a pressurized system.

So, when does air get in?

 

[JDSfastGN]

I am under the impression that exhaust leaks usually make you run rich, as most lose gas mileage when you get one. I know an intake leak or vacuum leak can cause a lean condition.

 

[MistaScott]

Right, you RUN rich in the motor, but the o2 reads a lean condition because of the air getting in and thus you get a high BLM...

But does air really get in is the question...????

 

[JDSfastGN]

oh my bad i misread your question, i guess the air enters in the short time between the pulses in the exhaust, as long as its not a lot of rpms or exhaust flow

 

[85HOT-T]

Jds has it. The air is pulled into the exhaust between powerstokes. Between each powerstoke there is a void in the exhaust. That is how flow masters work, they amplify the voids to create stonger scavageing in the combustion chamber.

 

[MistaScott]

So, the exhaust valve closes and causes a brief period of vacuum in the header which sucks the air in?

 

[85HOT-T]

No. Think of like an explosion. First there is an explosion, then an implosion inediately after. That void is the implosion. I don't know if that explain enogh, if not I can look up the exact science and exlanation for it.

 

[MistaScott]

As far as I know, the explosion occurs in the combustion chamber, the valve opens and lets the exhaust out. The exhaust flows through the header and to the turbo where it builds pressure and eventually spins the turbine.

I don't see where in that process air would be sucked in through a header crack...To suck in air there has to be vacuum. When is there vacuum?

For example, take a straw (or even your fist) and put a tiny hole in it. Then blow through the straw. Is air getting sucked into the straw through the hole? I don't think so...

 

[85HOT-T]

When the exhuast vavle opens the intake vavle is open momentarely ( sury for the spelling) too. With the low pressure following the ( explosion shall we say) helps suck all the air + fumes out the combustion chamber. I will give you the exact details tgomorow, I am kida drunk right now but the explnation is right on.

 

[Ormand]

I agree with MistaScott. If you're running at idle, or at cruise, when there is an intake vacuum, then and only then can there be a vacuum in the exhaust system. I can understand that under those conditions, there might be some air sucked in throug a crack in the exhaust. But under boost, there is a significant pressure in the exhaust headers. NOT a series of pressure pulses, with vacuum in between, but a pulsating pressure, that goes from high to higher, instead of from moderate to vacuum. The "void" that 85Hot mentions just can't exist, (well, I guess it is a relative void) when there are several cylinders all filling the same pipe, and the turbo backing up pressure in that pipe. When the lowest pressure in the system, at intake/exhaust overlap, is 10 to 20 psi above atmospheric, there ain't going to be no vacuum in the exhaust headers.

 

[ThikStik]

Dont forget Bernouli's theorem of pressure differentials. That is how venturies work in carbs, pitot tubes, etc. So depending on the shape of the crack, the rushing of gas past the inside of the crack could entrain gas from the outside. In this case, the shape of the crack could really cause variencesin the entrance of air.

 

[Ormand]

I thought about the possibility of a Bernoulli effect, but it would require something like a venturi, as you mentioned, to pull a vacuum. When you consider that the intake/exhaust system has maybe 15 or 20 psig on one end, at the intake valve, and 30 or more on the other end, at the turbo, it's hard to imagine that there is just the right geometry for the crack to pull in air. If you've ever looked at a something like an air jet ejector, or a steam jet ejector, they have a pretty pronounced venturi. I guess it may be possible for the crack to work that way, but it sure doesn't look like it would happen. Especially when you look at the crack, and you can see that there has exhaust gas leaking out. I can see having air leak in at idle, maybe, or at cruise, but not under boost.

 

[fubar2.5]

I only know that on my car after I fixed exhaust leaks the BL went higher at idle and O2s went lower at wot.

 

[MistaScott]

The question is whether exhaust valves closing would create vacuum in a header. Since they close at different times, I don't think it is possible to draw in outside air even if they did.

Therefore, I think this myth is busted. With a header crack, exhaust gets out, but air doesn't get it (at least not in significant quantities). If you think about it, a significant amount of air would have to get in to raise the BLM, let alone peg it.

So, can we conclude that a header leak will NOT raise BLM (at least not much)? If so, that would change the whole process for diagnosing BLMs. An exhaust leak cannot be to blame...

 

[ThikStik]

It does seem counter intuitive that exhaust could actually draw in ambient air. And throw in the fact that we are turbos makes it even harder to conceive. But the fact is, an exh crack gives us rich mixtures and hi blms. I say it is analogous as to why a bad o2 sensor effects WOT . A crack throws off o2s during pre wot conditions , the ecm takes this into account and adjust wot to the richer side. So, even at boost, an exh leak matters because the ecm calculates wot fueling from info during low boos/no boost conditions. Gurus, is this called a scaler? Heck if I know.

 

[Ormand]

ThikStik's idea makes sense to me. If the leak causes fueling errors at conditions other than WOT, but that affects WOT, then that would explain how the cracks affect WOT performance. From somewhere (Thrasher site, maybe?) I seem to remember that the chip uses info from the last block before WOT, to change fueling at WOT. That would also explain why some engines run good when you gradually ease into the gas, while not if you "stomp on it". (That's a Dixie expression)

 

[85HOT-T]

I did a quick look in my books and couldent find were it talks about this, but I can guarentee you from experience that it does happen. I go to University of Northwestern Ohio, and we do everything possible to a car there. Now On a vehicle tuner, The hudge computers with all the probes you see at a dealership, You can get a reading of 18 to 1 out the exhuast while the engine is really running 14.7 to 1. If you do not belive me, make a call to any shop that deals with fuel injection performance. I will keep looking in my book and get you guys word for word expaination.

 

[MistaScott]

Here is what I think happens:

Exhaust escapes from the header. The o2 sensor expects a certain quantity of oxygen to declare the mixture stoichometric (per amount of fuel injected into the engine). The o2 sensor only measures oxygen content of the exhaust. Since there is an exhaust leak, the o2 sensor does not receive as much oxygen as it otherwise would (relative to the amount of fuel injected). Therefore, the o2 sensor tells the computer that the mixture is rich (and lowers the BLM). As a result, the ECM subtracts fuel from the mixture. (This would explain fubar's experience).

This is one intuitive explanation I came up with. The explanation for an exhaust leak shouldn't take rocket science; it should make sense intuitively. I am glad people are giving this some thought...

 

[ThikStik]

I believe that the crack passes oxygen pass the o2sen and fools it (really ecm) into thinking system is lean, so ecm trims blm lower. Then at wot, the correction factor takes into account the "artificial leaner mixture" and gives too much fuel then too. I have read where if a mixture is unburned, it will further fool the o2sen by cooling it and thusly causing it to read even leaner. This is known to be the case if a cylinder is not firing and unburnt fuel is passed by the sensor. But I dont know how much it is relevent in our hypothetical case here where there is not an unburned mixture, but just a rich one. How the he$# outside air gets in is beyond me too. It seems way more plausible on n/a cars where you work hard to have a scavenging effect.

 

[ThikStik]

OK, Im confused as Ive contradicted myself on my posts re blms. A percieved oxygen content in exh should drive blms higher , but were seeing them go lower... so I just dont know.