Exhaust Backpressure and turbo engines
ฝัง
- เผยแพร่เมื่อ 3 ก.ค. 2024
- We take a look at EMAP's affect on volumetric efficiency while tuning two separate turbo engines on the dyno. How much is too much? Spend a few minutes with me as I examine the subject.
The crankcase pressure and coolant pressure as seen in the Infinity logs were not being used at the time so they have random values. If it were real the thing would have been blown up at that point. The DBW throttle was closing because it was a GM LS7 gold blade which apparently didnt like 30+psi. Now it has a Bosch 82mm DBW instead which has been fine at 75+ psi in our other racecars.
You hit just about everything 🤜🤛 it’s nice that you added in how different head/cams change “when” the engine starts to notice what’s going on. I always feel like that gets missed. What I normally like to do is see the N/A dyno graph. And then check it against the turbocharged version. If the turbo shifted peak torque left then the turbo’s turbine (usually) is too small. If it isn’t “stacking graphs” with rising boost. lol. Then it needs more turbo capacity.
Glad I just found this video! I'm running a variable-geometry Holset (HE351VE) on a street-driven BMW M20 (2.7L 2-valve), with a custom controller I built to modulate the vanes. Trying to get the VGT really dialed in led me to adding an EMAP sensor, and this video gives me some good insight to general EMAP/drive pressure ratio ranges for different applications, impact on power, etc.
With VGT, essentially the turbine A/R is adjustable, so especially at higher engine speeds I have a huge range of EMAP I can explore to help with spool-up. The controller is set up to run the vanes wide open until boost demand exceeds baro (taken from MS2 boost target table via CAN). Once it's time to spool, it closes the vanes with the general idea of balancing turbine drive pressure and mass flow - close them too much and you choke the engine. Once target boost is met, vanes open fully again.
Recently took a ton of data at various vane positions during spoolup to fully characterize the VGT - I found that the best boost response is at 1.7-1.8 drive pressure ratio, which impacts AFR by about 8%. More aggressive and it not only loses spool rate, but the car noticeably loses torque while it spools, amplifying any turbo lag feeling. Looking at some 4000rpm data with vanes about 79% shut during spool, gave drive pressure ratio around 1.7. Once it hit target boost and the vanes opened fully, drive pressure ratio dropped to 1.02 (EMAP dropped from 293kPa to 184kPa with 179kPa MAP). As it revved out drive pressure ratio increased to about 1.4 by redline.
Not sure if the increase in EMAP by redline is due to turbine housing or the 3" exhaust with muffler and wastegate dumping into downpipe. I know that 1.4 drive pressure ratio would not be great for a drag car, but maybe it's okay for a street car where I care about mid-range response, and can get full boost at 3000rpm with a 280deg cam at 6000ft elevation?
Lots of OE (stock) turbos will run far worse overall and generally all by redline. Depending on the overall combo as you have noted it can affect it. A colleague instrumented an STi once and saw 33psi of backpressure on 13psi of boost at 7000. That is 2.53:1 and I have seen close to that on some stock appearing turbos on Mitsubishi Evos that were still making great power, 2:1 comes to mind as an example. Thank you for sharing the data.
In regards to your question about how to go about it as what is the best solution, I find that what makes it the most streetable. In your case I happen to know ambient pressure is 80kpa or so and EMAP will stay consistent when boost will be lower. Case in point is some turbos at sea level (35R ish) will make 45-48psi and in Colorado Springs we will see 32-33 with the same 56psi of EMAP. I recently tuned a 600whp R32 GTR with a 0.92 A/r Twin scroll 8374 EFR and while it ran out at (100%wgdc) at 275kpa it spooled really well and was great between shifts. It might not have been the 700+whp the turbo could do it was GREAT fun on the street.
@@AaronONeal1 Thanks for the reply! With all the snowy weather we've been having lately I had to go back and read my comment, been a while since I played with the car. The whole VGT setup has been "fun" if you like a lot of head scratching :) Looking forward to getting back to it when the weather warms up, and again, your content is awesome.
I drive a built turbo s52 e39. My shop mate has tdi's with all different vnt turbos. Super cool to see them used on a older bmw. *hits close to home gives me the feels. Im currently trying to figure out a quiter exhaust. Currently its a 3.5" down pipe with 3" starting under rear seats. Does the bigger down pipe cost exhaust speed. How much back pressure would a chambered or turbo rear muffler add? (My car is ape flex /dual tunes on the fly ms43 dme ) 18 psi on pump 21/22 psi on e60+ map. 600cc pre compressor 50/50 water meth all the time. I digress. Fwiw I dont expect answers, just looking for more examples for reference. Ive watched these emap videos a few times sense they came out.
The Holset HE351VE is the worst turbo ever put on a 6.7 Cummins. The actuators frequently fail, the vanes stick (from poor design), and they have horrible exhaust flow. It's designed for towing and to comply with emission standards. Why you would want to use one for some experimental project on an engine half the size is beyond me, but whatever. If you really must have a VGT, check out Diesel Power Source "The Turbonator". It's a variable vane turbine housing that uses backpressure to articulate an actuator/vanes. No need for tuning. Comes in S300 and S400 variants. 👍
@@monsterram6617 I had one available, cost me 15% of what that turbonator world cost, and it's a decent match for my engine. Also, Turbonator looks like it only has 3-stage pneumatic actuator… far less control than HE351VE (~200 steps of resolutiom). Turbonator looks like the smallest turbine wheel is 71mm which is probably wayyy too big for my application. Also there are upgraded HE351VE wheels and housings available. Finally, the car has always been a learning and development platform, which would be severely limited by a 3-step control resolution, while the variable CAN based control on the HE351VE supports quite well even if it leaves some performance on the table.
Great material
I really enjoyed this video
Glad you enjoyed it
Yeah dude. You gotta keep making these videos. Wouldn’t have had the evo tuned by anyone else.
Awesome video dude! Btw, what do you use to measure EMAP?
Really good info here Aaron, a comparison on exhaust housings would be really nice to see.
Great idea, added to the list.
Tony, I filmed the video and started editing it last night. I will have the introduction into turbine housings later this week. Thanks again for the great idea.
@@AaronONeal1 that’s awesome Aaron, looking forward to it. Hopefully these videos educate the masses!
Great video Aaron. I got round to testing my own setup yesterday .1.3L Turbo Street car running 19 psi at the intake and 37psi at the exhaust manifold. 1:1.9 Ratio. Not very good for efficiency right up the top. But like a bat outa hell in the mid range. Thanks for sharing your perspective.
This is typical values we see in stock turbo applications vs race stuff. OEM level is always about response first. Thanks for the data.
I recently have changed the exhaust housing to the next size up and in testing have seen a reduction in back pressure of 18%. I was quite happy with that. Full boost was only 150 RPM later too.
@@andrewrigg4905 did you get a chance to dyno it and see what the power increase is?
I haven’t had the chance to, I was hoping to do back to back dyno testing…. It unfortunately hasn’t worked out that way .
Great info!! This really cleared some things up for me. I've been punishing my engine with smallish turbos and high boost for years fighting misfires and killing plugs and i'm starting to think this is it. I could never understand why the AFR's were going richer with higher boost, or higher gears for that matter. I assumed it was a fuel flow/delivery issue...1st gear AFR's could be as much as half a point leaner than 3rd or 4th yet fuel pressure is same. Am I correct in assuming EMAP is lower in lower gears due to less load ?
Correct, since generally less load also means less initial boost even if it creeps towards your target by the end of the gear.
Really interesting stuff. Definitely been wanting to get a better understanding of the effects of EMAP. This is probably irrelevant, but why does TPS taper off to ~90% ffor several hundred RPM at the very top of the pull on log 1 and 3?
In that particular pull the DBW was starting to get mad at me. I made the mistake of using a GM LS7 throttle which I learned have a reputation for doing that above 30psi. I went back to a cable for the following race and now a Bosch 82mm like we run on most of our racecars.
👌🏻👌🏻👏🏻👏🏻👏🏻👏🏻
I have one spare input left on my FT550, debating to use crankcase pressure or emap????
Both at some point. EMAP is a good metric to know for your current combo. Once you do it's not like you can really alter it so then move the sensor and wire over to the crankcase
👍🏻
Hello from Sweden!
Can I have your help and guide?
Of course
what sensor you use in exhaust?
100psi AEM. It goes thru a copper coil, then a stainless line to a baffled canister.
@@AaronONeal1I'm measuring EBP with the same sensor, too. But I have a really short stainless pipe and w/o a baffled canister. My data in the Infinity software is very scetchy and I have to add the filter option near to 100% in the software to cut out the peaks and get better data.
Maybe it's particularly the baffled canister which could help me out to get better results from my sensor?
BTW, like every time, very cool content from you, Aaron!!!
Great info in here in regards to back pressure, however my one comment I’d like to put out there is this.. Your first pass before raising boost there’s no Lambda FB but it’s missing target by a mile and then some up top, it’s amazing lambda feedback wasn’t doing something unless lambda PID’s were played with from this pass to the next or lambda control was disabled. Then the next pass pulls 10% and misses target lean. 3rd pass does do closer to your point you’re trying to get across in this video as it continues to pull more but in turn is rich. Before a discussion like this, in my opinion it would be much better to provide data with a better dialed map to be able to get the knowledge across. Don’t mean this to come off like an ass, because I’m all for continuing to learn, just voicing what I see from the outside looking in. Definitely will continue to dial into the technical stuff like this.
In the end the phenomenon has occurred in the included 4B11 example, a K24, and a 4G64 as well but neither of those had EMAP. This was 2017 which was the same year as installation so I wasnt as observant. Good catch, I will make sure to keep it more accurate in the future.
@@AaronONeal1 knock control would still show a change in lambda target, white line doesn’t dip indicating knock fueling being added. However this is beside the point you’re making in this video so I guess we can just leave it at that.
What A/R are you using?
I'll have to rewatch the specific videos but these will be on the smaller side, T3 0.82 open or 1.0/1.15 T4 divided. I run the 1.28 divided on my V6