Boost Leaks and Speed Density

[LiquidX]

Please don't mention just do a boost leak test, will be doing one regardless but looking for solid information.


Posted this on ECMTuning.com and haven't gotten any real responses.

Well we all know that boost leaks are bad regardless of how you are reading airflow. Spool up time will suffer, turbo will work harder etc.

On a MAF setup a boost or vacuum leak post sensor will cause idle issues, stalling, bogging, poor throttle response, rich and or lean conditions etc. Usually easily detectable prior to testing.


Using speed density, from what I'm aware of, non of those issues should or would apply compared to a MAF setup, at least outside of boost.

So what kind of symptoms would I see or anyone for that fact that is running speed density (aside from slower spool)? What should I be looking for to give me an insight to a potential leak that would be visible on a log?

I'm on a new turbo and unfamiliar on how it should be running in each RPM and how it should be boosting. Figured there may be some logable data that could or would conclude that there is in fact a leak. Say BoostEst is greater than actual boost numbers or some other set of data. That's the info I'm looking for.

 

[black_gst]

it wont show up on a log as far as i know besides loss of power of loss of peak PSI or a laggy feeling

 

[LiquidX]

You would think there would be something visual on a log while in boost anyway. The reason why I stated BoostEst and actual boost numbers via MAP sensor is they are off and I would think from the reading I've done is that may suggest a boost leak. Looking for other ways on a log that may suggest a leak as well since they are hard to realize on a SD setup without regularly testing.

 

[99gst_racer]

There won't be any of the usual symptoms with SD. SD measures pressure at the intake plenum, so if there's a leak before the intake manifold, your MAP sensor won't know about it. A leak is only ever detrimental to the engine's tune if it's a leak after the point where airflow is measured.

 

[boostNmetal]

I'm going to get into the habit of doing a quick BLT at every
oil change for the reasons listed above. I would think that anything other than a large leak would be pretty transparent to the driver. That's good and bad I suppose.

 

[knochgoon24]

There really aren't any symptoms other than slower spool. That's the beauty of SD. And any change in intake temps from a harder working turbo would pretty much be negated by the intercooler.

 

[boostNmetal]

any change in intake temps from a harder working turbo would pretty much be negated by the intercooler.

Good point. An IAT before the intercooler as well as after would be useful for trending to indicate a possible boost leak cropping up. Hmmm, I might try this out.

 

[LiquidX]

I'm aware of slower spool and that would be given regardless but I was hoping for something that you could visually see and be "oh that's odd, perhaps I have a leak". Again, the only that that I was aware of would be BoostEst vs. real boost and the discrepancy there.

Don't get me wrong, I love how SD works I just would like a way to know if something went wrong instead of having to constantly check. It would be more of a hassle on a turbo you have no knowledge of but regardless anytime you mess with the air track you should do a leak test.

After all, this is one of the reasons why I went to SD as boost leaks wouldn't have the standard effect on the motor.


So Paul, you feel that there is no data that is logable that would give a hint that there would be a boost leak at all?
(Hoping Tom or Dave chime in either on here or ECMLink's forum, if so, I'll post quote onto here if on other)

 

[97gst_spyder]

What if you added another map sensor to the IC track? Like on the J-pipe. Then log the extra map signal and compare it to the map up by the tb?

 

[LiquidX]

I could see how that would work but I'm not going to go and spend another $80 or so for a new map sensor and then figure how to wire it in at that location. Doing this not only for myself but for anyone else that is running speed density and looking for this information.

 

[Sam@RRE]

BoostEst is a boost estimate derived from airflow. With a MAF, airflow numbers will shoot up because more air is being PULLED through the sensor to accommodate for the leak, thus leading to a high airflow and BoostEST value. In SD, airflow is calculated through a function of RPM (speed) and Air density (pressure and temp). So if the turbo is compensating for a leak and spinning faster, and pulling in more air through the intake, there isn't a MAF measuring more air flowing, pressure remains constant, and as stated, temps will be compensated for by the AIT or negated by the intercooler. So there is no loggable value that will indicate a boost leak in SD. MAFs count airflow directly, and SD calculates airflow. With a leak, same amount of airflow makes it into the engine, whether its SD or MAf, the difference is how much fuel the ECU calculated to spray.

 

[LiquidX]

With a leak, same amount of airflow makes it into the engine, whether its SD or MAf, the difference is how much fuel the ECU calculated to spray.

I understand what the rest of your quote said but this line made me think.

If you have a boost leak your true boost numbers would be off of what the estimated value would be. If the ECU is spitting out numbers in which there would be 100% effeciency with airflow (no leaks) then shouldn't the estimate and real boost match up when your values/tune is set up properly?


Maybe I'm just trying to think this out way too much than I should or perhaps it's the beer.

The airflow readings are read from the MAP and IAT together and therefor will still allow your to drive without issues regardless of the leak. However, because of that if there is a leak post MAP wouldn't your trims be off due to the lack of "density" and if it's pre sensors your gauge would be off in relation to what it should be.

Hope I made that understandable

 

[knochgoon24]

If there was a leak post-MAP, the MAP would be reading the lower pressure and would compensate accordingly.

You would still be injecting the same amount of fuel as when you didn't have the leak.

Now if you somehow had a leak post-injector (not that you really can) and were blowing some of that fuel out, you'd notice a lean condition, but by that point you'd smell the fuel.

 

[the_mork]

True enough in theory (for small leaks or tuned out leaks/pressure drops) but it doesn't work that way with post-MAP leaks because they're so close to the IM. The ugly leak for SD cars is the injector isolators which caused me some real issues when I was trying to dial in my 1600s, and yes the MAP probably did work correctly in that case and there wouldn't have been a problem if I had a single cylinder engine but because of the divided runners on the IM I ended up with some very weird behavior. OF course that behavior will probably be similar with any air metering setup, but further exacerbated with an MAF system.

Liquidx, the results from estimated boost are exactly what is described, estimated boost. If you determined your VE curve for RPM and MAP and then adjusted the estimated boost effectively then yes, estimated boost should be equal to real boost at the IM, and that is the important part. Your issue is that there is no way to read what is going on any earlier than that.

There are two ways to detect fairly large leaks but they're not useful for small leaks and by the time the leaks are large enough for either of these methods to be useful then you'll notice them anyway.

The first is to pay attention to your fueling and realize that if you're going richer then you probably have a lower VE, but you'll notice the large popping sound of intercooler piping flying apart and lack of power before you notice a change in fueling and you'd still have to check just to make sure that your tune doesn't suck.

The second way might actually be useful depending on your setup. Use either a manual boost controller or set your wastegate solenoid to a static duty cycle, and if your peak logged boost falls and remains lower for a a while then you likely are developing a boost leak. Of course atmospheric conditions can also affect this so it requires watching this sort of thing over time. It wouldn't be useful for a failing bov or a blown off vacuum nipple but it might alert you to a corroding IC or a pin hole in a coupler.

The short answer in the end is that there isn't an effective way of determining boost leaks through logs on an SD system. Even trying to do this on an MAF based system is sketchy IMO compared to just getting the tester out.

 

[crash89]

I know this is quite old, oh well.

There won't be any of the usual symptoms with SD. SD measures pressure at the intake plenum, so if there's a leak before the intake manifold, your MAP sensor won't know about it. A leak is only ever detrimental to the engine's tune if it's a leak after the point where airflow is measured.

Now with that said, Ive been having a rich issue in my spyder for some time now and never even thought it could be because of a leak.. My previous engine when I was tuning it started reading a lot of knock so we kept trying to tune it out.. Well turns out it was rod knock. Now, with this "new" engine I knew I was probably running rich due to the amount of fuel we gave the tune to try and get rid of the knock we were reading..

This past weekend I wanted to hit the test and tunes at the shootout and had Mr Peepers help me out with a tune.. Took out a ton of fuel and still running pig rich as soon as I hit full boost and reading a bunch of knock.. Upon letting it idle one morning I listened to the motor and thought I could hear what appears could be a leak at the intake gasket.

Now, Judging by what Paul said the map sensor has already done its measuring and by a leak at the gasket the air that has been measured is leaking.. This could cause my issue if Im not mistaken correct??

I havent made a boost leak tester to fit my holset yet, hence me asking as Im not positive I have a leak.

 

[LandSpeed-DSM]

A significant Post-TB leak could do that.

If you are logging turbine shaft speed that would tip you off as well, I realize you are on a Holset and we don't have provisions for that but throwing it out there for future searchers with different setups.

Only way to know for sure is BLT and some soapy water.