Making Boost Faster: An oddball idea.

[Exarch]

I've read a few threads here and there about making boost faster (lower RPM), and have a rather interesting idea to throw out there.

The idea comes from the VW turbo implimentation, which runs a K03 sport (on the MY2003 which I'm familiar with) being pushed by a 1.8L 20 valve I-4. It begins making usable positive boost at 2000 rpm. Part due to the size of the turbo (tiny) and partially due to a fancy-pants system which is the focus of this idea.

The 1.8T uses an ECU controlled BPV so as not to lose metered air and all that fun jazz.

When you put your foot into it the ECU will partially open the BPV and vary it's amount of opening as the boost builds, resulting in a closed BPV by the time boost gets up above a certain point (10 lbs if memory serves). What this serves to do is lessen the resistance on the compressor wheel so it picks up RPM faster and is capable of generating boost quicker.

The question here, is would it be viable to set up a similar system which is driver operated on a Subaru. Namely a small solenoid operated valve valve venting from the pressurised to atmosphere-post-MAF side of the intake which is controlled by a simple button. I'd also like to explore theories of wether or not it would have any noticable benefit, and wether or not it would cause the ECU to throw all manner of codes/errors.

The only possible problem that could be run into would be if the ECU meteres fuel based on throttle position, airflow, RPM and o2 sensor readings without taking into account actual manifold pressure. Would one of the Subaru-Savants roaming around be up for commenting on this particular potential problem?

It seems to me that if temorairly partially bleeding pressure to spin the wheels up faster were the greatest thing since the discovery of gasoline, everyone would do it; However I could just be very pessimistic.

Anyone have any ideas either for or against this rather out-of-the-box idea? Or is it more common than I know about and used all over the place?

I'd be very inclined to give it a try, and assuming I'm still curious, when I pick up a Subaru of my own, I'll give it a whirl. I would assume the worst that can happen is that it makes it run rich for a few seconds, which won't kill it... And if it sucks as a mod, removing it should be a big deal either...

[oldfartracer]

It's called anti-lag, the rally cars run this and so do a lot of the fastest drag cars. I know the Hydra EM has this and I think the UTEC can be setup with it also. This can also eat turbos if not careful.
You would better to keep a small turbo and P&P everything to help spool.

[wagonracer]

I always like to see creative thinking, and am prone to experimenting too. I'll be watching this thread for feedback from the "experts"

[dumdum]

That doesn't make any sense?At boost onset if the BPV is opening then the turbo has to work harder to make positive pressure.I could see VW using that to control max boost pressure but not help with spool up.Generally BPV only help with spool up when you shift.They kkep the compressor spinning in a no load condition.

antilag is generally not found on oem cars.It has to do with retarded timing and very rich fueling to ignite gases in the up-pipe,causing an explosion to spin the turbo durring shifting.
Antilag system - Wikipedia, the free encyclopedia
Turbo Anti-lag systems - A technical description

[Exarch]

@oldfartracer

I thought antilag was as dumdum mentioned, a method of over-riching and igniting in the exhaust system to spin the turbo up. That would also, I would think, be part of the reason Anti-Lag can eat turbos.

@wagonracer
I can't say this is an original idea, the Germans put it in my old Jetta 1.8t. See below for some times to spool in various situations. I used to have the videos, as there was a very large misconcetption on the VW forums I was part of that the reason a 1.8t is spooled (and making max torque) by 2000 RPM was only because the turbo was so small.

The other solid pointer is that folks who, with no other mods at all, move from the ECU controlled BPV to a manual or non-ecu electronic BOV with the BPV completely disabled/replaced usually slow their quarter mile and 60-foot times down right off the bat.

Sadly I can't say I'm a Subaru expert yet, however I can say that I know my turbochargers!!!

@dumdum
It makes perfect sense, the goal is to get the turbo's spindle speed up into the optimal range as fast as possible. By venting a -small- amount of pressure from the pressurized side of the intake, for -part- of the spooling process, you're getting the speed up. If you look at a compressor map, you'll notice areas of rpm which are less efficient. The goal is to bypass these and get straight to the spindle speeds where the turbo is more efficient, then not close the valve so fast that you stall the compressor or end up with it decelerating too much between when it's using kinetic energy to compress the air and when that compressed air is burned. On the 1.8t, from a roll-on in first starting at 1800 rpm, the system has the BPV in one or another stage of opening for around the first half-second after you put your foot into it. Engaging the BPV spooling system while running at full boost on the 1.8t will drop the pressure in the intake by around 5psi.

We're dealing with a set of turbines here, and turbines are a very complicated items when you start getting down to actual functionality. Glancing at a compressor map will point this out quickly.

On the exhaust side, it takes considerably more force (read: engine-work) to get the turbo from low efficiency RPM up to high efficiency RPM than it does to keep it in the higher efficiency RPM band. (given a linear-climbing load)

On the compresor side, it takes considerably more work to compress air at a low efficiency RPM while accellerating to the high-efficiency RPM than it does to begin to start compressing air at the high-efficency RPM.

What this does is delay some of the load until the turbo has spooled into a more efficient range. By not trying to load the turbo while spooling it at the same time, you end up not having to go through the crap end of efficiency for both wheels. The end result is overall more efficient use of hardware, which results in faster spool times or faster time-to-power between when you demand and when it puts out. In some marginal cases, such as making boost at lower RPM, or spooling faster while winding up, this idea is sheer money.

A very sumarized example from my old 1.8t days:
1st gear roll on @ 1800 rpm:
With spooling system working: 13 lbs of boost in around 7-tenths of a second at 1950-2000 RPM
With spooling system disabled: 13 lbs of boost in around 3 seconds at 2300-2400 RPM.

5th gear roll on @ 2000 rpm
With spooing system working: 13 lbs of boost in around 2 seconds at 2050-2100 RPM
With spooling system disabled: 13 lbs of boost in around 7 seconds at 2250-2300 RPM

1-2 shift at redline (6700 rpm), getting back to full boost:
With spooling system working: Full boost in 0.25-0.4 seconds following shift.
With spooling system disabled: Full boost in .75 to 1 second following shift.

Think of it as poor-mans variable geometry turbocharging, it achieves roughly the same result through a completely different, and less complicated/expensive , means.

I think about the only potential way this could backfire would be if the BPV closes too fast and causes enough of sudden load shock to the compressor and stalls it, which would just be a waste of time. Then again it would be simple to build a circuit which would allow for a slower close, either that or just use a rheostat or vari-sistor, manually controll it, then automate through simple circuitry...

It would be easy enough to find out what diameter of valve to use based off some CFM of the engine, which is pretty simple to calc out. Time to make a spreadsheet!!!!

[dumdum]

Originally Posted by Exarch

@dumdum
It makes perfect sense, the goal is to get the turbo's spindle speed up into the optimal range as fast as possible. By venting a -small- amount of pressure from the pressurized side of the intake, for -part- of the spooling process, you're getting the speed up. If you look at a compressor map, you'll notice areas of rpm which are less efficient. The goal is to bypass these and get straight to the spindle speeds where the turbo is more efficient

I understand turbos pretty well too.I see this a robbing Peter to pay Paul.You causing a boost/intake leak to try and get the turbo to spin up faster around or before the boost thresh hold by rerouting that air to try and pinwheel the compressor.Kinda see how I see it?Maybe VW has configure it to work well on there systems but I can get the same effect by changing timing durring low load/low rpm areas to raise my EGT's to help spool up,or port and polishing my exhaust manifolds and turbo,installing a catless up-pipe,adding a turbo back exhaust,adding a 3 port BCS in interrupt mode,upgrade to a ball/roller bearing turbo.Some of those ideas I mentioned are more $$$ than others but they work pretty well.On the stock TD04L-13T turbo I can make 17 psi in first gear around 5500-6000rpm and 20psi around 2900rpm in 3rd gear(I could make more but thats what I cap it at)I'm not discounting your idea and I would love to see somebody try it but I'm sure the air venting back would have to be configured in a way that it is most efficient.The BPV vent pipe on our cars come back into the inlet pipe 3-4 inches away from the turbo at a 90 degree angle.I would think the vented air would need to either hit the turbo compressor straight on or at an angle from the opposite side that would be pushing the turbo in the way it is already spinning.

[Exarch]

Always nice to run across a fellow snail-head!

It is in fact robbing Peter to pay Paul, the only trick is that you can end up paying Paul more than you robbed Peter for. I do understand your point of view though, and it is sound. I tend to be one of those that caters to some rather wierd ideas when it comes to these things; Unlimited curiousity!!!

I'll agree that VW could have used some rather arcane trickery to get theirs to work and provide benefit. The big differences between the Suby and VW engines are displacement and turbo size along with possibly head flow. The VW is a smaller engine (1.8L), and uses a very dinky turbo, the venerable K03 Sport. On that last front, the 1.8L I-4 has a 20-valve head with 3 exhaust valves per cylinder, so I'd wager it can pass a lot of gas in a short amount of time.

Side Note: To better illustrate the size of the turbo above, my 1.8T would push 12.7 lbs of boost from ~2000 - 5000 RPM, at which point the boost would drop down to around 9 lbs by 6500 RPM. According to my handy-dandy CFM/ lbs-minute spreadsheet, the 1.8 would be gobbling 12.1 lbs/minute of air at redline. For context: The STi 2.5 will be chewing through 16.8 lbs/minute of air. As far as I know, the stock VF39 turbos don't have a problem holding stock boost to redline. Ergo, the VF39 is enormous when compared with the K03.

Side Side Note: On the compressor map for the K03 Sport, at 6500 rpm and 12.1 lbs/minute, it is amazing it does anything constructive at all, as that flow rate is off the sheet of paper I had the map on... Good thing the turbo isn't big enough for boost creep, else many VW folks would have a 1.8handgrenade.

Back on track: I would think the only way to find this out would be to install and give it a try. The worst that could happen is a waste of some time and a relatively small amount of money. If it worked as I think it might, it would be a pretty cool trick.

I see what you mean about he BPV dumping back in at a very odd place. The BPV on the Jetta was at least a foot up the intake from the turbo. Personally I'd not feel comfortable at all dumping any quantity of air through the stock BPV attach point while the engine is under load and trying to spool the turbo. I'd be affraid of massively disturbing any type of laminar flow headed for the compressor, when you do that you lose some VE on the intake side, complicate fueling, and just generally can have a poor day.

While I'm thinking about that, I wonder if there are any gains to be made by (if this is possible) moving that BPV dump point further away, or placing it at an angle towards the inlet of the turbo. When you shift and the BPV dumps pressure in there, it's gotta create turbulence, which is going to [probably] slow the turbo down more than if it were just ingesting a smooth flow through the intake. Maybe there could be some small merit in working a way of dumping the BPV'd air into the compressor at a better angle? Maybe it would allow the turbine to keep up some RPM and not have to be accelerated as much following the shift. Who knows!!!

If you've read any of my other threads, such as in the newb section, you know that I'm in the market for a Subaru of my very own (and trying to sell an Accord) and am hoping to have one in hand in about 5 months. A lot of this theorycrafting will be a lot simpler when I actually have the car to look at... Untill then, I'll theorycraft till I'm blue in the face!

I cut my teeth on the VW, and would still have it if everything aside from the drivetrain didn't fall apart (was assembled in mexico ). I can't wait to have a forced induction car with some horsepower again.

[dumdum]

Originally Posted by Exarch

While I'm thinking about that, I wonder if there are any gains to be made by (if this is possible) moving that BPV dump point further away, or placing it at an angle towards the inlet of the turbo. When you shift and the BPV dumps pressure in there, it's gotta create turbulence, which is going to [probably] slow the turbo down more than if it were just ingesting a smooth flow through the intake. Maybe there could be some small merit in working a way of dumping the BPV'd air into the compressor at a better angle? Maybe it would allow the turbine to keep up some RPM and not have to be accelerated as much following the shift. Who knows!!!

Unfortunately untill you have really looked at the engine bay in person of the WRX you can kinda see why they did it.Things are kinda packed in there and space is a problem.Also where the BPV is located creates the necessity for it's entrance into the pipe.I really don't see turbulance being to much of a problem since the valves barely open at low boost and at high boost so much turbulent air is moving through the twists and turns of the stock turbo inlet anyways.I agree it could defiantly could be a better design.

[Exarch]

Ah, yeah, I can see how the BPV positioning can make things a little problematic. From what I last recall peeking under the hood of a WRX, the intake tract is quite short, meaning it would be rather difficult to move the BPV inlet further up. In the end though, with it flowing that much air, one could see no difference at all even if it were moved or angled differently.

I need to do some digging and update my spreadsheet for displacement CFM and lbs/minute. I want to figure out the velocity of the incoming air through the intake, should be a pretty simple formula, finding it could be exciting though. I think after being able to see how much air is rolling through the intake would give a better WAG (wild ass guess) on the whole bit.

On another side note, when dealing with turbocharged vehicles and intake velocities, and since CFM and such is calculated from rpm vs displacement, would you need to take the effective compression from the compressor map and run it the other way?

Such as, if an engine is injesting 200 CFM at redline/WOT, and the compression ratio for the given turbo and CFM is 2:1, does that mean the intake will be pulling at 400 CFM? Guess I have some more digging to do tonight. That works too, as it should be a very quiet 12 hour shift, and it's my friday

Any flud dynamicists around? One of them would be very handy to have in pocket for some of these questions.

The thing that dawned on me with the "spool valve" idea as I woke up this morning (or evening, I'm a night shift type) is this: Does slightly enriching the mixture cause any situation on the exhaust side which would help spooling? How about if you're enriching by reducing air and not increasing fuel?

Once again, paging the local turbine engineer or flow junkie to help out with this...