Lets talk about Turbos

[bu11dogg2]

I think a lot of people will think that they will be able to slap a GT35R or a Rotated Kit on their 818.

I think Brandon has a great point:


yeah, the stock Wrx wastegate and turbo dynamics tables in the ECU allow you to control and target boost really easy based on throttle% and rpm.

However turbo spool is somewhat a function of engine load, 1st and 2nd gear don't often see 100% of your max target boost because the engine is swallowing air since it increases revs faster than the turbo can. The STI ECU allows for gear-specific wastegate values to counter some of the lack of spool in lower gears and over-boosting risk in higher gears. But each turbo has it's boost 'signature' which is pretty difficult to change even with all sorts of trickery in wastegate duty to make numbers.

With this car being so lightweight, I'm interested to see how the turbo selection process works out. Shooting for 375-400 hp in this car might make for a pretty laggy torque curve given the car simply won't have the inertia to build turbo spool in lower gears. To make big power you need a big turbo. But a big turbo takes time to spool. Something the size of an 18G or so would probably be about as laggy as most people would want in a car this light I imagine. Above that and people might end up having to sacrifice some of the boost/torque smoothing tactics in order to try and make 375hp+ worth of boost before slapping into the rev-limiter in gears 1-3, doing so will give you the twitchy wall of torque effect.

So, with that in mind, a smaller turbo that hits target boost sooner but might not have as much top end would be easier to tune for. Having a snail that CAN hit target boost levels of 18+psi in the low to mid 3k range would have enough time for a factory WRX ECU to "smooth" the torque delivery and reduce the power surge that risks spinning this car out of control.

I guess this brings back the H6 discussion to some extent. This Turbo lag and turbo power surge issue with just RWD will become a giant topic of debate and discussion. An N/A H6 will become even more attractive as people get competitive in this platform.

Just my take, I hate turbo lag but some guys don't mind waiting until 4000-4500 rpm for power. I prefer it much....much sooner than that.

I'm keeping my Forced Performance HTA68 and adding Methanol Injection.

here's an overlay of my car with a stock turbo vs HTA68

Mitsubishi TD04L-13T
(390cfm at 14.7psi, 200-275whp, Bolt-On)
This is the standard equipment turbocharger used on the USDM Subaru Impreza WRX MY02-07.
Expect to achieve full boost with the proper mods and a quality tune between 2500-3000rpms.


Forced Performance 68HTA
(47 lb/min, 350-470whp, Bolt-On)
Everyone has seen or heard about Evo3 16g turbos and TD05H 18g turbos for the Subaru 2.0 and 2.5 engines, our newest turbo is the FP 68HTA™Turbocharger for the Subaru STi/WRX unit which out flows both the E316g and 18g wheels and allows for higher PR operation without increased compressor outlet temperatures like the existing 16g and 18g wheels. Mass flow is increased to a whopping 47lb/min which is several lb/min greater than either the E316g or the 18g wheels are capable of. Rapid spool and large torque numbers are guaranteed by the TD05H turbine section.

http://efilogics.com/dyno/graph.php?gb=0&hp=1&torque=1&rpm=1&sl=1&sln=1&runid1=792&rgb1=000000255&runid2=1142&rgb2=204000000

Boost and RPM

http://efilogics.com/dyno/graph.php?gb=0&hp=1&torque=1&boost=1&rpm=1&sl=1&sln=1&runid1=1221&rgb1=000000255&runid2=1142&rgb2=204000000

Similar down low, waaaaay bigger up top.

Here's the skinny on some stock turbos: (yes, I stole this from NASIOC)


IHI VF39
(250-325whp, Bolt-On)
This is the standard equipment turbocharger used on the USDM Subaru Impreza WRX STI. It can be found on all model years from 2004-2006. The VF39 utilizes a thrust bearing design and the P18 exhaust housing.
Expect to achieve full boost with the proper mods and a quality tune between 3000-3500rpms. 2002-2005 WRX owners will need fuel upgrades for this turbocharger and proper engine management is highly recommended for all vehicles that utilize this turbo aftermarket. Though they are prone to cracking, VF39’s can be had for very cheap if bought used.


IHI VF43
(250-325whp, Bolt-On)
This is the standard equipment turbocharger used on the MY07 USDM Subaru Impreza WRX STI. It can be found on both base STI's and STI Limited's. The VF43 utilizes a thrust bearing design and the P18 exhaust housing. The primary difference between the VF43 and the VF39 used previously on STI's seems to be the stiffness of the wastegate. The VF43 has a stiffer wastegate designed to reduce boost creep issues.
Expect to achieve full boost with the proper mods and a quality tune between 3000-3500rpms. 2002-2005 WRX owners will need fuel upgrades for this turbocharger and proper engine management is highly recommended for all vehicles that utilize this turbo aftermarket.


IHI VF48
(250-325whp, Bolt-On)
This is the standard equipment turbocharger used on the USDM Subaru WRX STI MY08-10. The primary difference between the VF48 and the VF43 used previously on STI's is the more advanced bearing design.
Expect to achieve full boost with the proper mods and a quality tune between 3000-3500rpms. 2002-2005 WRX owners will need fuel upgrades for this turbocharger and proper engine management is highly recommended for all vehicles. This turbocharger is not a direct bolt-on for the MY08+ WRX.

IHI VF52
(250-325whp, Bolt-On)
This is the standard equipment turbocharger used on the USDM Subaru WRX MY09-10. This turbocharger is similar to the VF48 but utilizes the USDM Legacy GT-style housing.
Expect to achieve full boost with the proper mods and a quality tune between 2800-3300rpms. This is NOT a direct bolt-on for 2002-2007 WRX, but will fit the 2008 WRX and 2005+ Legacy GT without modification.

[bbjones121]

The FP 68HTA seemed to be overhyped in the legacy community once people started getting some actual dyno numbers out. I would recommend BNR as a pretty good option, especially if you are going to run the flanged intercooler (AVO or Perrin, obviously not the stock intercooler).

[16g-95gsx]

Turbine housing selection will eliminate the load issue. A .63 AR turbine GT35R will spool far different than a .68 T4 version. I personally think limiting your turbo selection to "bolt-ons" with a car like the 818 would be foolish. The stock manifolds are going to be too heavy and should therefore be tossed out anyway in favor of a lighter tubular option, not to mention better performance. The great thing about turbos is that the options are endless.

I think the HTA68 would be well matched on a car like this due to it's quick spool and decently high power potential. It also is cheaper than many other options. FP has had it on sale in the past for only 799 bucks, which is a decent price for a small 47lb/min turbo. Old school 50trims flow similarly but would be heavier and have more lag. Their pricing would be slightly cheaper but I don't think they'd be as good of a match.

[bbjones121]

I agree. There are so many tubular options to be limited to using the LGT or 08+wrx style turbo. You might be able to keep things more compact though?


Turbine housing selection will eliminate the load issue. A .63 AR turbine GT35R will spool far different than a .68 T4 version. I personally think limiting your turbo selection to "bolt-ons" with a car like the 818 would be foolish. The stock manifolds are going to be too heavy and should therefore be tossed out anyway in favor of a lighter tubular option, not to mention better performance. The great thing about turbos is that the options are endless.

I think the HTA68 would be well matched on a car like this due to it's quick spool and decently high power potential. It also is cheaper than many other options. FP has had it on sale in the past for only 799 bucks, which is a decent price for a small 47lb/min turbo. Old school 50trims flow similarly but would be heavier and have more lag. Their pricing would be slightly cheaper but I don't think they'd be as good of a match.

[riptide motorsport]

great post, learning everyday..............Steven

[bu11dogg2]

The FP 68HTA seemed to be overhyped in the legacy community once people started getting some actual dyno numbers out. I would recommend BNR as a pretty good option, especially if you are going to run the flanged intercooler (AVO or Perrin, obviously not the stock intercooler).

The 68HTA is better with the 2.0, that is why there was limited success with the Legacy crowd (2.5).

I only posted stock turbos to start. I'll put up more aftermarket later.

[BrandonDrums]

Turbine housing selection will eliminate the load issue. A .63 AR turbine GT35R will spool far different than a .68 T4 version. I personally think limiting your turbo selection to "bolt-ons" with a car like the 818 would be foolish. The stock manifolds are going to be too heavy and should therefore be tossed out anyway in favor of a lighter tubular option, not to mention better performance. The great thing about turbos is that the options are endless.

I think the HTA68 would be well matched on a car like this due to it's quick spool and decently high power potential. It also is cheaper than many other options. FP has had it on sale in the past for only 799 bucks, which is a decent price for a small 47lb/min turbo. Old school 50trims flow similarly but would be heavier and have more lag. Their pricing would be slightly cheaper but I don't think they'd be as good of a match.

I'm sure there are plenty of ways to get around load issues but there is no free Lunch so to speak. Going with a smaller housing might eliminate some spool time but it will limit the top end.

Now, putting on equal length headers and doing an externally gated rotated mount setup will of course eliminate many of lag issues I brought up but at much cost and custom fabrication.

For the most part I was talking about bolt-on turbos and factory manifolds.

Another thing that might get around the lack of spool in lower gears is to swap for taller ratios. I'm typing with my thumbs but there are several reasons you'd want taller gears in the 818 than just turbo spool.

[Jmcd]

I think you should put the spool issue in relative terms, vs dyno comparison/numbers: will a bigger turbo spool even less quickly in a lighter car in lower gears? Yes.

Even if you could spool an 18g/68hta in lower gears-could you put the power down, considering torque multiplication? Could you use that on the street?

A 7k 1-2 shift puts you around 3850 rpms, 2-3 is 4800 or so, and those look like fun places to be on that boost curve. Based upon your numbers on a 2.0 engine, it would seem that anyone more concerned with the lower end would be able to compensate for reduced load-based spool with a 2.5 block, or go with tuning options for torque (e85 comes to mind)

I have a feeling that many people, however will be happy, with a td04 (which, tuned will be a handful), or be plenty happy having a bolt on worth 50% more hp, with small perceptible lag (mostly felt while cruising anyway) in exchange for an incredible top end.

[16g-95gsx]

I agree 100%, I believe that the lag concerns are overdone. A 2.5L block will have no issues spooling any of these turbos if it is properly setup. Even a modest 300whp on a car as light as the 818 would be extremely potent considering the mild amount of work put into the engine. The bigger concern IMO would be actually delivering that power to the pavement, and ultimate tractive ability as that will be the determining factor of whether or not the car actually IS quick.

[crobin4]

I agree 100%, I believe that the lag concerns are overdone. A 2.5L block will have no issues spooling any of these turbos if it is properly setup. Even a modest 300whp on a car as light as the 818 would be extremely potent considering the mild amount of work put into the engine. The bigger concern IMO would be actually delivering that power to the pavement, and ultimate tractive ability as that will be the determining factor of whether or not the car actually IS quick.

Which is precisely why I've been crying for room for tires.

[BrandonDrums]

I agree 100%, I believe that the lag concerns are overdone. A 2.5L block will have no issues spooling any of these turbos if it is properly setup. Even a modest 300whp on a car as light as the 818 would be extremely potent considering the mild amount of work put into the engine. The bigger concern IMO would be actually delivering that power to the pavement, and ultimate tractive ability as that will be the determining factor of whether or not the car actually IS quick.


I agree 100%, I believe that the lag concerns are overdone. A 2.5L block will have no issues spooling any of these turbos if it is properly setup. Even a modest 300whp on a car as light as the 818 would be extremely potent considering the mild amount of work put into the engine. The bigger concern IMO would be actually delivering that power to the pavement, and ultimate tractive ability as that will be the determining factor of whether or not the car actually IS quick.

You're exactly right! I was talking about the guys wanting to put 500+ hp in this car. a 18G or Big 16g wouldn't be a big deal. However, slapping a garret class giant turbo looking to get 450whp in a 1800lb car with these engines is going to have:

[lag]
continue
[lag\]

[wheelspin]


But yeah, all the turbos listed above will pretty much be fine for this car. You're still looking at more lag than you would in a heavier car and with the reduced traction you'd have to be a bit more conscientious with the foot but the OEM STI turbos and 16g turbos will be responsive enough to deal with.

If you wanted to put a much bigger turbo in the car though, we're talking taller gears, headers, external wastegate etc. Taller gears to reduce torque multiplication plus making more use of the powerband (taller overall but still closely matched in ratios). Otherwise it's going to be a fight to keep the power down and enjoy the benefits of being MR and 1800lbs.

[Monkey Tricycle Flip]

The Blouch 19T on a 2.0L would likely be a great match for this application. It doesn't have quite the peak numbers of a VF39 or VF48, but it the area under the curve is larger and pushed to the left.

I had a 19T with GrimmSpeed ported headers, up-pipe and a full TBE and on my 9-2X. The thing was a spool monster.

[bromikl]

For the turbo newbies:

STS Turbo? How is that different from a standard turbo? How does it reduce spool lag?

On a separate matter, is it possible to always run on pressurized intake? What would happen if you ran your turbo all the time to constantly run at say, 3 psi?

[bbjones121]

On a separate matter, is it possible to always run on pressurized intake? What would happen if you ran your turbo all the time to constantly run at say, 3 psi?

You would become a part time resident at the gas station.

[thebeerbaron]

On a separate matter, is it possible to always run on pressurized intake? What would happen if you ran your turbo all the time to constantly run at say, 3 psi?

I am not a turbo expert* but I do know that if you had a turbo that was small enough to pressurize at idle, it would be far too small at the top end.

There are other terminal problems with this idea, but this is at least one show stopper.



* I should really learn a bit more since I seem to say that quite often now

[Twinspool]

On a separate matter, is it possible to always run on pressurized intake? What would happen if you ran your turbo all the time to constantly run at say, 3 psi?

You don't have to be a turbo expert to remember from Otto cycle 101 that the engine operates under vacuum except for at wide open throttle. Show me how you drive with your throttle wired full open all the time and then I'll show you how to make 3psi at idle. :)

[Gollum]

You can only cruise with as much vacuum/boost as your load allows, if you're trying to maintain a speed at least. I've never seen a setup that was under boost when cruising, unless you're talking about cruising at 150+mph, or a puny little engine in a huge vehicle.

Regarding reducing the turbine size for quicker spool, it's not a bad idea. In theory it doesn't limit the compressor side at all, it just changes the dynamics of where you end up on the compressor map across the RPM and boost range. In many cases you can lower the turbine AR and still see the same max power potential from the turbo in an application. There are exceptions of course, but you get my point. The bigger issue is that if you were already on the brink of surge you're throwing yourself into the surge area of the map because the turbo will spool much lower down, and the engine might not be able to ingest enough air to keep the turbo happy. But if we're only talking about a compressor that will only flow roughly 400hp worth of air, this probably won't be a problem.

I've seen people do the opposite. Go from a mid sized turbo with a turbine AR too small for the application, and they have surge issues. They upgrade to a LARGER turbo with a larger compressor, but also a larger turbine AR and the surge goes away. They can now run higher boost without surge, but the downside is that it's probably lag city.

Having a compressor that really fits the way your engine breaths is a big part of a good turbo selection. Or you get real fancy and do a F1 style BOV that allows you to run extreme size offsets. But that would be a custom built turbo in most applications. You can usually get an off the shelf turbo that will do what you want with a conventional wastegate setup.

All one more reason to lower the front diff ratio.... Would give you more traction in the lower gears. Would allow for the turbo to load up a bit easier. And would give you tamer cruise RPM. Sometimes loosing a few hundredths to 60mph and a couple thousandths in the quarter mile is worth it.

[Twinspool]

Even if you're well away from the surge lines, downsizing the turbine A/R or housing will always increase the pressure ratio and negatively effect cylinder evacuation. Don't forget that the larger compressor wheel comes with more inertia, the same damn turbine suddenly can't spool the larger wheel and what do people do? They say use a smaller turbine to improve the spool time. Doh!

[bromikl]

Has anyone developed a means to store pressurized air to prevent the need to spool up? What if you had a pressure-regulated tank of air that releases its store instantly when throttle position calls for it, and recharges when the turbo would normally spool down?

[mn_vette]

Has anyone developed a means to store pressurized air to prevent the need to spool up? What if you had a pressure-regulated tank of air that releases its store instantly when throttle position calls for it, and recharges when the turbo would normally spool down?

It has been thought about. The problem is that you need such a large tank to feed an engine at boost that it becomes prohibitive.

About swapping compressors, the turbine side is what flows out the exhaust. So if you are running an internal wastegate there is a limit to how much horsepower you can flow through the turbine housing. Keeping them balanced is a must. If you switch to an external wastegate then things free up a bit more.

[Gollum]

The best way to keep the turbo spooled is through excess fuel in the exhaust fumes. This can be done to mild levels on street cars, and then there's the radical levels of anti-lag that rally cars use which dumps loads of fuel causing huge explosions of fuel in the exhaust as fuel hits the hot turbine. This will shorten the life of the turbo considerably, but can be worth it if you're racing competitively.

And despite what many think, you usually CAN get a turbo to spool off idle with no load. It just takes some time combined with the right ECU control. Even my ancient 80's T3 can get me to max psi at idle on an engine that would only have a whopping 140hp without a turbo. And that's with an ancient ECU at that! With something like megasquirt I could spool a GT30 at idle, and probably a GT35R.

And also, regarding pressure ratio, you're right twinspool, but you also have to remember that once you've reached your boost/rpm target your wastegate will start to open, changing the dynamics of cylinder evacuation. And also, lack of cylinder evacuation isn't the big reason you hit surge to begin with. If you were running a radical cam with tons of overlap they would be more related, but still not entirely. I personally wouldn't recommend going to a smaller turbine to fix lag issues. There's many other things you can do, and just resulting to a different turbo can be a costly experiment.

Let's also remember (for the sake of those not too familiar with turbos), that turbine AR isn't the turbine size. AR is just the inlet ratio which greatly affects gas velocity, not quantity. If you're running a large enough wastegate for your application, you really shouldn't have an issue with upper RPM breathing unless you're also having a hard time keeping the turbo up to target boost (which really shouldn't be a problem). People complain about flow capacity of a smaller AR, but I think that's a very backwards mindset. We WANT a turbo to spool as soon and as easily as possible, and then match the compressor to the engine and boost curve we're looking for.

So to condense this into simple statements:

Turbine size and AR - The size of the turbine should match your compressor size, and the AR should match your engine to the turbine size.

Compressor size and AR - Compressor size and wheel design are the most important factors for the compressor. The AR is relatively minor on this site of the turbo, and thus not usually a selectable spec of a turbo. But when you look at a compressor map, you need to remember that has nothing to do with the hot side of the turbo, only the compressor side. You'll see a wide variety of compressor map shapes and ranges, all for different reasons/applications. So while a turbine will be directly related to the engine capacity and volumetric efficiency, the compressor is directly related to HP desired and at what pressure and pressure ratio will be required for a given engine.


A comparison to make sense of all this, is to think about a small Honda B16, versus my old Datsun L28ET, which has more than a liter over the Honda. Both engines though produce about the same power in NA form. The Honda produces peak power at around 8k (or redline if you will), and the Datsun makes peak power around 6k (just under in most cases). Both engines work great with a T3 turbine at street power levels. The B16 will prefer a larger turbine AR (as high as .75) to allow for better top end breathing (again, good wastegate design could fix this on a smaller AR), and will still create full boost by 3krpm, giving this engine a huge full boost range since it revs all the way to 8k. The datsun though, will work better with a smaller AR (as low as .45), as this engine produces it's peak power at a lower rpm. This will give the engine full boost sooner (bellow 2k in most gears) and the engine won't really have breathing issues, even with an internal wastegate, since the engine has a naturally choked deign that runs out of breath up top.

On most T3 compressors, both of these engines will produce around 250-300hp without fancy tricks. The datsun though will be a little lower PSI, and run up the compressor map much more horizontally, while the B16 will be higher PSI and run up the map more vertically. With both engines though, the T4 compressor size will be better suited to reaching 300hp and a bit beyond, as the T3 will be outside of it's efficiency range by a bit at those power levels. Note though, that they will out on outer edges of the map in different regions, due to the required pressure ratios required.

Hopefully that explains some things for some people. I don't really expect to come across as the "teacher" here. I'm just sharing what I know, and I hope that some of you other turbo experts and chime in with your knowledge and experience and give your take on the complex subject of turbos so that the rest of us can further our understanding.

[mn_vette]

I have one additional thing to add about spooling. Its not the engine load or rpm that causes the turbo to spool. Its the volume and duration of exhaust on the turbine. Which turns out to be a combination of throttle position, rpm, and time.

If you are at idle the amount of air flowing in the exhaust if very small. And if you are in neutral and holding the revs to 6000 rpm, you still aren't flowing much exhaust air. At WOT and spinning 3000 rpm you will be moving half the exhaust than at 6000 rpm. Now once you are moving enough exhaust into the turbo to get it to spool you need to keep it there for a time to get the turbine accellerated up to speed to create the correct amount of boost. And once you start getting a little bit of boost it increases your exhaust flow to give you more spool.

Keeping these things in mind, the first gear of the WRX is going to go pretty quickly. Assuming that we may even be able to hit WOT after the launch we are going to hit redline pretty quickly limiting the time that the exhaust energy has to work on the turbine. At what rpm does a certin level kick in? Now put the car in 5th gear at 30mph and try the same thing again. Go to WOT and see at what rpm level the boost level is reached. In a higher gear it has less rpm to work with, but more time.

I had an ECU on my turbo corvette that had a boost builder mode. On the flip of a switch it would retard the timing as you increased RPM to a set level. This would cause more and more throttle to be required to maintain that RPM until you were at WOT. With the set RPM, WOT, and a few seconds it would create boost while the car was still in neutral. Its a nice trick for drag racers, but I don't think the 818 would benefit from a bunch of boost off the line.

[16g-95gsx]

You guys are all on the right track, in the end the concerns over spool are very overrated. It's great that we all come from differing backgrounds and have different inputs on the subject but ultimately we speak the same note. I personally used to measure backpressure at the collector in order to properly select AR. I feel that the standard E16g would be a great choice for the price, weight, and size. Other options would be the HTA68, and for the higher powered cars GT30R's with a tial turbine housing to keep things lighter with enormous power potential. In the end it will be up to the user, but the point is that spool/lag is over rated in most cases. With a larger turbo 1st gear will be worthless, but there is more that can be done to solve that as well. If you want the ultimate in lagless shifts and setups an auto trans is the way to go, but you'll add weight, and inefficiency, which in a car not necessarily meant for a straight line is likely to hurt the spirit of what the car should be.

[readymix]

The problem with the "lag" discussion is determining what you consider to be "laggy." I'm running a GT35R with the .82 hotside on a 2.5L, and on pump gas I've got it set up to 22 psi for the max. I can reach this mark by 4000 rpms and hold till redline. What alot of people fail to consider is that while, yes, I'm not hitting full boost till ~4000 rpms, at 3000 rpms I'm making 200+hp at the wheels. You need to look at the curves and where power starts to become usable. For 4000rpms I have more horsepower than a WRX with minor exhaust mods and a TD04 can muster at peak. The car is entirely driveable...at least in Impreza form. Not sure how it'll react on an 1800lb car. Everyone who has driven my car in traffic and around town finds the power delivery very enjoyable but manageable.

For me, my 818 purchase will be based on standard road use, and I find the more freeway friendly top end power more suitable than low RPM short burst play. Of course, I'm a cheap *******, and if I had to start from scratch, I'd likely just get something smaller. But it is what it is, and I will simply transplant what I have into the 818 and call it a day. I say all this now, but who knows, maybe I'll change my tune later on and want a smaller, more low-end torquey snail.

[BrandonDrums]

Well, there's lag and throttle response. You might be making 200 hp at the wheels at 3 grand but what if you have to modulate the go-juice mid corner and have to wait .5 seconds to get the power down before spinning out.

I think that's mainly the concern here, not straight-line wot power curve numbers.

[Gollum]

That's the thing though, with a WELL DONE recirculation system, correctly sized setup, with minimal piping you can create a setup that's VERY capable of modulation in the corner. Usually what makes a "hot rodded" street car with a turbo a complete pig in the corners is a poor bypass system and a terrible ECU tune that doesn't know how to handle transients well.

Also, if you're going from zero to full throttle in an instant odds are you weren't prepared for the corner, and with that careful roll on of throttle in a corner you should receive controllable power. The problem comes when you have a sub 2 liter engine making 400+hp and you get some guy with no experience driving it stabbing the throttle to full right at the apex of a turn. He ends up backwards and then going on a rant saying turbos are "poo poo" for racing.

Rally drives have been doing it for years, and on course a lot harder to remember than a track. ;-)

[Benji]

Yes but then don't rally drivers also use anti-lag because they don't care what happens to the turbos so much.

[bromikl]

I have a turbo idea worth investigating, but it's not directly related to the 818.

What if the turbo ran on waste heat, rather than exhaust pressure? You could run it all day long without burning any extra gas. Your exhaust header wouldn't have the restriction, and your turbo would be available any time or RPM you wanted. Here's what I'm thinking:

Exhaust gasses after the catalytic are about 500F. If one were to use that heat to boil water, one could generate up to 30 PSI (or more) steam to run the turbo. Two exhaust pipes - one smaller one inside a larger one - would be an effective cross-flow heat exchanger. Exhaust gasses would be in the larger pipe. Distilled water from small reservoir (1-2 Gal) is pumped in by a fuel pump (or similar pump) into the downstream end of the smaller pipe.This would create a cross-flow heat exchanger and generate significant pressure any time the water pump is running. The steam would spin the turbo, and exhaust from the car.

No exhaust restriction. No spool time - or more accurately, turn it on for hard driving, off when you don't need it. Constant boost over engine RPM range. No extra fuel burned.

Comments?

[readymix]

I have a turbo idea worth investigating, but it's not directly related to the 818.

What if the turbo ran on waste heat, rather than exhaust pressure? You could run it all day long without burning any extra gas. Your exhaust header wouldn't have the restriction, and your turbo would be available any time or RPM you wanted. Here's what I'm thinking:

Exhaust gasses after the catalytic are about 500F. If one were to use that heat to boil water, one could generate up to 30 PSI (or more) steam to run the turbo. Two exhaust pipes - one smaller one inside a larger one - would be an effective cross-flow heat exchanger. Exhaust gasses would be in the larger pipe. Distilled water from small reservoir (1-2 Gal) is pumped in by a fuel pump (or similar pump) into the downstream end of the smaller pipe.This would create a cross-flow heat exchanger and generate significant pressure any time the water pump is running. The steam would spin the turbo, and exhaust from the car.

No exhaust restriction. No spool time - or more accurately, turn it on for hard driving, off when you don't need it. Constant boost over engine RPM range. No extra fuel burned.

Comments?

The problem is volume and size. While the post-cat temps are 500*, they likely aren't consistant throughout the rpm ranges and loads a car uses. But even if we assume they were, you would need alot of water in a constantly boiling state...remember, 500 degrees will heat the water to boiling, but the act of transferring that heat to the water will cool the heat source...so another concern is maintaining that heat level. And the system would need to be entirely closed, which means you would need the steam in a looped system, that would then condense and recollect the water, return it to the reservoir for re-boiling.

It's entirely too complicated of a system to be economicly feasible. You might get some more low end grunt if you could pull this off, as you'd have boost on tap at all times (assuming that the system is efficient at all) but you would have that boost at the cost of the weight increase required to house the system in your car.

Also, the note about "extra fuel burned." You don't burn extra fuel because you are spinning the turbo with your exhaust. You are burning the extra fuel because your cylinder air volume and density requires you to provide more fuel to avoid detonation. If you add more air, you need more fuel. That's the very basics of engine operation. So, in your example, a turbine that constantly provides higher than atmospheric air pressure to the engine, you will need to use MORE fuel than a turbo powered car, because in a turbo powered car, you aren't pressurizing the engine till the turbo starts rolling. Most of the time your car is at vaccum and running at stoich. This is why turbo cars get decent freeway MPG when you aren't stabbing the gas pedal. Set the cruise control and go. A constantly boosted system would require alot of added fuel.

[bromikl]

readymix,

I wasn't planning to recover the steam after it's used. I figure a couple gallons of water is enough for an hour or more of running the turbo. The water would be atomized (like fuel injection) in the heat exchanger and instantly flashed to steam. So no boiling pot of water. Unless the turbo is running, the inner pipe is dry. Lots of stored heat in there. The max pressure would be dependent on exhaust temp, but I don't expect the exhaust pipes to cool quickly while there's still hot exhaust running through them. And the pressure needed to run the turbo isn't going to be any where near the max pressure possible at 500F.

Complicated? Yes, but not unimaginably complicated. Programming boost and waste gate etc. is a different kind of complicated. And yes, it would weigh more. But for low end boost and no exhaust restriction, might be worth it.

[readymix]

readymix,

I wasn't planning to recover the steam after it's used. I figure a couple gallons of water is enough for an hour or more of running the turbo. The water would be atomized (like fuel injection) in the heat exchanger and instantly flashed to steam. So no boiling pot of water. Unless the turbo is running, the inner pipe is dry. Lots of stored heat in there. The max pressure would be dependent on exhaust temp, but I don't expect the exhaust pipes to cool quickly while there's still hot exhaust running through them. And the pressure needed to run the turbo isn't going to be any where near the max pressure possible at 500F.

Complicated? Yes, but not unimaginably complicated. Programming boost and waste gate etc. is a different kind of complicated. And yes, it would weigh more. But for low end boost and no exhaust restriction, might be worth it.

I could maybe see it as a race only application, but for constant road use which this kit is designed for, it would be highly impractical. And without a recapture, you run in to a few issues. For one, the water in my area is extremely hard. I could cast fake teeth with all the calcium we scrape off the humidifier in our bedroom every month. This would cause all sorts of problems, and would require me to purchase distilled water. Which leads to another issue, you would likely only get a few hours of use out of the system, which is less driving than one tank of gas will afford you. That means that not only do you need to fill the water tank during a gas fill, but you would need to carry along a few gallons of distilled water so you could pull over and refill the system at some point. And you would most assuredly need to keep it filled, because the ECU programming required to switch the fuel and timing curves from a constantly boosted programming to a standard curve N/A setting isn't really possible with the standard Subaru ECU. You would need to run some kind of standalone computer. Then likely run some sort of tank level sensor, like a fuel level sensor, in your water tank to alert the standalone to a low water level. Assuming you can even find a standalone that will sense a low value in a liquid tank as an input, and will use that input to change mappings, the cost of that will be prohibitive to the daily driver/weekender types. Again, race only application.

[thebeerbaron]

What you're trying to do is run a steam turbine (http://en.wikipedia.org/wiki/Steam_turbine) using heat from the exhaust gasses. You use exhaust heat to boil water, then steam turns the turbine which is connected to a compressor and you get boost.

Read the section of the wiki on Thermodynamics of Steam Turbines (http://en.wikipedia.org/wiki/Steam_turbine#Thermodynamics_of_steam_turbines). My short layman's summation is: the power to turn the turbine comes from the pressure change of the steam from high pressure to low. Your design that doesn't include a boiling pot of water doesn't have this - all your power is lost when your atomized water "flashes" to steam. At that point, it's just wet air and you're not going to get any work out of it.

Also, you're going to lose a lot of efficiency in your air-to-water heat exchange in the tail pipe, especially if you use the relatively low surface area of the outside of the pipe. There's a reason intercoolers and radiators are as dense as they are!

I like the thinking outside the box, but if you really want a turbocharger that has boost whenever you want it, maybe you want a supercharger?

[thebeerbaron]

How often do I get a chance to actually use this clip in a relevant setting?


http://www.youtube.com/watch?v=vWxZm8WjlI8

[Gollum]

Technically a turbo already IS using the heat of the exhaust to create power! (hence the video above).

Turbochargers aren't just an air turbine, but a thermodynamic device. This is one of the big reasons STS turbo systems are so horribly inefficient. The further downstream you put the turbo the cooler the gases get, and the more energy is lost. In order to understand this principle you must understand some basics of physics in that kinetic energy is kinetic energy is kinetic energy . Can I say that more?

It takes energy to get your car moving, and then you apply the brakes and as much energy as your car has stored in it's movement goes into heat in the brakes. A tire can only grip as much ENERGY as it can withstand until it reaches it's contact heat point. I know tires work better at higher temps usually, but it still comes down to how well the tire can work as a heat transfer device, or heat sink. It's basic physics in the end, and the complexity is understanding how to work around the laws of physics to better a design.

And when it comes to turbos, heat is GOOD. There's a lot of good conversation on the topic, so I won't bore us all to death.


Benji ~ Regarding anti-lag systems. Yes rally uses anti-lag measures, but those don't help them with throttle modulation and such. Watch any of ken block's videos and you'll notice that's is a VERY rare thing now days to hear that anti-lag system getting into true backfires to keep the turbo up to speed. It takes serious throttle downtime to require the turbo to be brought back up to speed. The rest of the time the turbo is kept spooling by slightly (not overly) rich decel mixtures that increases the weight of the exhaust gases. And thanks to modern bearing technology that turbo will stay going for quite a while on with just a little bit of help.

Example video:

http://www.youtube.com/watch?v=4TshFWSsrn8

We don't hear major backfiring until late in the video. At 3:50 you have a good example of minor anit-lag that can be done to some measure with most turbos and manifolds. This is no different than the gurgling you get from a carb'ed V8 when decelerating in gear. In fact even a lot of OEM EFI systems gurgle quite a bit once you remove the CAT and restrictive mufflers in order to hear it.

But that at 4:10 we hear some REAL backfiring. That's the stuff that will blow things up. Notice it was triggered by extended full throttle and the rev limit, and then he left off in order to gain grip to get going straight again. This might not even be full anti-lag as much as circumstance. But at 5:50 we hear definite anti-lag measure. He was braking down from probably well over 100mph down to what looks like 30mph or so entering a spin. MAJOR gurgling here, but still not what vintage rally cars used to need to do, with the load POP....POP....POP.... almost in perfect timing as the turbo speed sensor sensed the turbo spooling down.

Remember that a lot of these people here talk about how lag isn't an issue with most turbos, and they're talking about journal bearing turbos! Get a ball bearing and it's even LESS of an issue. Lag is way overblown, and can always be learned around by any skilled driver.

[BrandonDrums]

^^^thats why I'm saving up for the new GTX3071r in the ATP Subaru specific housings.

I HATE lag but I'm really wanting more go-sauce and the stats on the new GTX suggest that the responsiveness will be there.

I think the moral of this story is, for useable high hp in a RWD MR ultralight such as an 818, you have to spend a bit more to get the proper equipment.

In a wrx or sti, having AWD alleviates many of the woes a "binary" power curve will produce in a tail-happy car.

In the 818, tune and setup will matter more to keep the car on the road with these big turbos.

[thebeerbaron]

In the 818, tune and setup will matter more to keep the car on the road with these big turbos.

I think the key to keeping an 818 on the road is going to be driver "training", for lack of a more PC word. The Subaru engine is so easy to mod and 227HP sounds like so little in a world where even a Honda Odyssey has 248 horses. There are going to be a lot of high-power engines dropped into 818s just because it seems like the fun thing to do. Hanes and Fruit of the Loom are going to sell a lot of new underwear to newly-minted 818 drivers who overestimated the amount of power they needed. :)

In all seriousness though, I think you've hit it right on the nose. Smooth power delivery and responsiveness will be the name of the game for 818s that can be driven fast by someone other than The Stig. From what I've heard about the WRX, I think the stock motor will deliver that, but I think it will be very interesting to see what people come up with and what they find drivable.

[readymix]

We could bypass the turbos and steam turbines altogether with this:

http://www.eugeneleafty.com/DEI/Nebraska%20GM%20Modern%20Muscle%20%20View%20topic% 20-%20Turbo%20Questions%20%28pg%201%29.htm

[Twinspool]

^^^thats why I'm saving up for the new GTX3071r in the ATP Subaru specific housings.

Can I get an amen! That's precisely the unit I was eyeballing as the one to beat. You have good taste Brandon.

Bromikl, you really should go down to the local Jr. College and sign up for at least the first semester of Physics and then thermo. You'll get all your questions answered and can avoid looking like a buffoon in the future.

[thebeerbaron]

Bromikl, you really should go down to the local Jr. College and sign up for at least the first semester of Physics and then thermo. You'll get all your questions answered and can avoid looking like a buffoon in the future.

I don't mind looking like a buffoon, I'm quite good at it. But some people might take offense to the comparison.

It's cool to putz around outside your sphere of knowledge, that's how you pick up new interests. If Bromikl really wants to pursue this idea, then I think the Junior college classes would be a great start. But to take a college class just so you don't ask a "stupid" question? That's a bit extreme.

Question away, but be ready for tough-love answers.

[bbjones121]

The more ideas people have and the more questions people try to answer, the better off society will be. There are lots of things we take for granted and a lot of them came from people generating radical ideas and testing them. Thomas Edison anyone? Never knock down someone who questions the way something has been done in the past, that might not be the way something is done in the future.

[BrandonDrums]

I think the key to keeping an 818 on the road is going to be driver "training", for lack of a more PC word. The Subaru engine is so easy to mod and 227HP sounds like so little in a world where even a Honda Odyssey has 248 horses. There are going to be a lot of high-power engines dropped into 818s just because it seems like the fun thing to do. Hanes and Fruit of the Loom are going to sell a lot of new underwear to newly-minted 818 drivers who overestimated the amount of power they needed. :)

In all seriousness though, I think you've hit it right on the nose. Smooth power delivery and responsiveness will be the name of the game for 818s that can be driven fast by someone other than The Stig. From what I've heard about the WRX, I think the stock motor will deliver that, but I think it will be very interesting to see what people come up with and what they find drivable.

Yeah, 227 hp is PLENTY. But having driven the stock motor with the factory tune, it's quite binary in it's torque delivery. I remember being afraid to pull into traffic because of the lag issues. Under 3,500 the stock 2.0 has NO steam then it's whoosh. Even a completely stock motor will need to be tuned to smooth out the power delivery in my opinion.

**Edit, the stock TUNE is what makes the power delivery so bad.

[05xtsy]

^^^^ Exactly ^^^^ Solution, accessport and tubro back. IMO the stock setup, that is if your donor's turbo doesn't have too much mileage on it, is going to be an absolute blast and reliable. I'm more interested in brakes, suspension and tires. My wrx is stage 2. I honestly cant imagine it OVER 1000 LBs Lighter.

[bromikl]

I'm a pro at 'looking like a buffoon.' The steam turbine idea is sound, but impractical, as readymix pointed out. I seriously underestimated the feedwater needed to run the turbo. About 12L per minute is as close as I can figure. The heat exchanger needed to transfer that much heat would be too heavy, and recapturing the condensate would become necessary.

I've come to realize the current application is the most effective available. Too bad about all that energy blowing out the tailpipe! :confused:

[Gollum]

The stock WRX turbo's response can definitely be improved upon in many ways, and that's where a lot of people get confused about the whole "lag" issue. They drive one or two cars with a bit of lag and assume all turbo cars are like that.

Open up the exhaust, remap the fuel and ignition, make some other tweaks, and you're off to having a completely different feeling vehicle.

I don't think I can stress enough, that like aerodynamics it's about the PACKAGE, not just a single part in the system.

[chrisarella]

I see someone copy/pasted off of NASIOC. ;)

I've contributed to that thread myself a few times over the years. I've been modifying Subies of almost 10 years and can say from experience avoid the IHI turbos like the plague. I've had 4 different kinds including the long sought-after JDM twinscroll turbos and can say with certainty they aren't worth the headache. I've had nothing but bad luck with them. Stick with tried and true Garrett or Mitsubishi turbos.

[BrandonDrums]

The stock WRX turbo's response can definitely be improved upon in many ways, and that's where a lot of people get confused about the whole "lag" issue. They drive one or two cars with a bit of lag and assume all turbo cars are like that.

Open up the exhaust, remap the fuel and ignition, make some other tweaks, and you're off to having a completely different feeling vehicle.

I don't think I can stress enough, that like aerodynamics it's about the PACKAGE, not just a single part in the system.

The stock turbo is actually just about the least laggy turbo you can put on these engines. It's just not capable of big power. I think most of this discussion about lag and turbo selection is about the challenges of bolting on a turbo capable of 300+whp. Modders of WRX's and STI's put some giant giant turbos on their cars which take until 4k+ to spool then lay down extremely high numbers which on a lightweight rear-drive and mid-engined car would be a nightmare to control. Perhaps "lag" isn't the best terminology, it's a lack of throttle response and absence of a smooth linear torque curve that's the challenge with big turbos.

A TD04 (the stock wrx turbo) won't be laggy on an 818 with such a shortened exhaust system. It's just not going to give you 300whp either. You WILL need to get rid of the stock tune though, THAT is the source of all the lag on a stock WRX. NOT the turbo.

[BrandonDrums]

I'm a pro at 'looking like a buffoon.' The steam turbine idea is sound, but impractical, as readymix pointed out. I seriously underestimated the feedwater needed to run the turbo. About 12L per minute is as close as I can figure. The heat exchanger needed to transfer that much heat would be too heavy, and recapturing the condensate would become necessary.

I've come to realize the current application is the most effective available. Too bad about all that energy blowing out the tailpipe! :confused:


I'm a pro at 'looking like a buffoon.' The steam turbine idea is sound, but impractical, as readymix pointed out. I seriously underestimated the feedwater needed to run the turbo. About 12L per minute is as close as I can figure. The heat exchanger needed to transfer that much heat would be too heavy, and recapturing the condensate would become necessary.

I've come to realize the current application is the most effective available. Too bad about all that energy blowing out the tailpipe! :confused:

Actually, you're honestly not far off with your steam idea. It's actually something that is being done already...at least on one side of your H20 thinking http://www.aquatune.com

I really think aqua injection should be on all cars these days, at least all turbo and performance cars. The chemistry and physics IS there to make a strong case for it's use. It allows to some extent for boosted cars to run leaner fuel mixtures by letting the water mist cool and "calm" the fuel charge in high compression situations. Dumping extra fuel in an engine during high boost is done to prevent detonation and cool the burn to ease the stress on the internal parts. Using aqua injection saves gas, improves c02 emissions in addition to preserving the engine and improving turbo spool by replacing unburnt fuel. Water is more dense than gasoline so it provides more mass to the exhaust charge which increases turbo spool, water also aides in the combustion of the existing fuel since it's a polar molecule and disperses the fuel particulate more evenly within the cylinder.

However, so does meth injection. You don't get a power boost with water injection when you compare it vs. meth and you might even see some performance losses if you compare the combustion equations but it would allow you to be more aggressive then you otherwise would with boost, a/f, timing etc.

Perrin had a twin turbo H-6 STI where they used Aquamist which got me looking into it.
http://www.perrinperformance.com/pages/show/32

Don't be so hard on yourself, you aren't a buffoon but a GENIUS!

[bromikl]

I can't say I considered injecting water into the intake. That really is a great idea. It would help cool the intake as the mist evaporates, preventing detonation. The water vapor would further expand in the cylinders increasing exhaust volume to speed up spooling. Genius. Wish I had thought of it.

I never realized people were dumping fuel into a cylinder simply to add expansion mass, knowing that it wouldn't all get burned. Doesn't seem right.

[chrisarella]

The stock turbo is actually just about the least laggy turbo you can put on these engines. It's just not capable of big power. I think most of this discussion about lag and turbo selection is about the challenges of bolting on a turbo capable of 300+whp. Modders of WRX's and STI's put some giant giant turbos on their cars which take until 4k+ to spool then lay down extremely high numbers which on a lightweight rear-drive and mid-engined car would be a nightmare to control. Perhaps "lag" isn't the best terminology, it's a lack of throttle response and absence of a smooth linear torque curve that's the challenge with big turbos.

A TD04 (the stock wrx turbo) won't be laggy on an 818 with such a shortened exhaust system. It's just not going to give you 300whp either. You WILL need to get rid of the stock tune though, THAT is the source of all the lag on a stock WRX. NOT the turbo.

Exactly. Like all decisions the factory makes turbo selection is a compromise. It isn't exactly equal parts, but there is an exchange of power to lag. Really the biggest variable in the factory WRX's lag was the Three catalitic converters in the system to pass US (moreso California) emission standards. The td04 was selected to counteract that. But there are better options out there. There are much better wheel designs to leverage in recent years, and twin scroll designs reduce the lag even more. As I tell everyone, they absolutely need to decide what their goals are for the car (cause it can't be great at everything), read the turbo FAQ on NASIOC to narrow down which turbo matches the characteristics of their goals, read reviews of people who actually OWN those turbos, then make your selection.

As far as tune goes, there are a lot of options regarding tuning, ranging from piggy back units like the Utec, to complete stand-alone ECUs like the Hydra. In between (and the most popular), Cobb Tuning has created what they call an Accessport which reflashes the ECU (amongst having the ability to do other things like launch control and data logging). They also offer off-the-shelf maps that were designed around specific configurations of modifications, but those won't likely work well with the 818 since so much will be different. You're likely looking at getting an Accessport and having a tuner (that knows Subarus) do the tuning. There are dosens of good ones around the country, and even some like Clark Turner who travels doing dyno tuning days all over North America.

We can load these forums up with a ton of info around Subarus, but I can't stress enough how much info already exists on NASIOC. Just reading the FAQs alone will keep you busy reading for weeks. I'm not trying to discount this forum but save you guys a lot of time finding your answers. Start on NASIOC, and if you don't find it there or it is a question specific to the 818 then this is the place. No disrespect at all. ;)

[Gollum]

Water injection is a great addition to any street car that's pushing the limits of what their turbo and/or intercooler can do. But as stated, you can get a lot of the same benefits from meth, and also consequently E85.

Overall the idea really is sound and has already really proven itself. The only hurdles will be how to control it and how to make sure you either don't run out of water, or that your system doesn't blow up because your water tank went dry.

[VF48WRX]

I wouldn't even think about running any GT series turbos with a 5 speed trans. I haven't seen how the chassis on the 818 is set up, but I doubt pulling the trans will be a walk in the park like it is on a wrx chassis. I am currently running a VF48 turbo on my 2 liter WRX with an RS trans and rear diff.. I think this setup would be absolutely violent on an 818 that weighs very little.

[2FastCobra]

How many WRX motors would it take to spool my BW 83mm T6 turbo??? Just joking, I would like to build an 818 for my next car. Good info here!! http://i91.photobucket.com/albums/k301/2fastCobra/dynomagshoot017-Copy1024x768.jpg

[FinishlineWRX]

What about the newer style twin scroll 20g? Seems like something that might offer quick spool and decent power.

All seems to be dependent on how much room FF leaves us around the engine to be able to fit aftermarket components in.

[Smitty911]

I'm a pro at 'looking like a buffoon.' The steam turbine idea is sound, but impractical, as readymix pointed out. I seriously underestimated the feedwater needed to run the turbo. About 12L per minute is as close as I can figure. The heat exchanger needed to transfer that much heat would be too heavy, and recapturing the condensate would become necessary.

I've come to realize the current application is the most effective available. Too bad about all that energy blowing out the tailpipe! :confused:

Ahhhh, it's getting used. Ever hear of Global Warming, Opps sorry, Climate Change. :p ;)

Smitty

[soul strife]

How many WRX motors would it take to spool my BW 83mm T6 turbo??? Just joking, I would like to build an 818 for my next car. Good info here!! http://i91.photobucket.com/albums/k301/2fastCobra/dynomagshoot017-Copy1024x768.jpg

Seriously Drool. I am with you though. I am used to V8's running Supers/Turbos. It's interesting reading all this as the science makes sense but, it's in a different application. I love reading threads like this. IMHO, the 818 and subbie motor are a perfect match.

[818dennis]

Their is a lot of info about Turbos on this page that I don't under stand. But I think I'll go with the turbo that comes with the engine I choose. I would like to try also adding Hydrogen gas to it, http://www.youtube.com/watch?v=0mA3-c9AiWw . Then when I have the money maybe supercharger, http://www.xcceleration.com/engine-packages.html , http://www.raptorsc.com.au/models.php?id=6 . Dennis

[metalmaker12]

Vf37 twin scroll for me, 325-350 whp in 1600-1800lbs is plenty

[philly15]

Vf37 twin scroll for me, 325-350 whp in 1600-1800lbs is plenty

nice ive been lookin for a vf36 setup but for the price of setting up a twin scroll i think i decided against it decided to do a ported, polished, and coated vf43, (however if i find a good deal on a vf36 it might still happen!) and going with an ewg and e85 tune shooting for the 350 whp range should be attainable fairly easily with the lessened drivetrain loss and the e85

[longislandwrx]

I have a ported and coated vf48 which I think will be a perfect turbo for this car. I havent read much about twin scroll on a ej20 does it work well?

[JeromeS13]

I have a ported and coated vf48 which I think will be a perfect turbo for this car. I havent read much about twin scroll on a ej20 does it work well?

I will be going this route, as well, on an EJ255/7.

[philly15]

I have a ported and coated vf48 which I think will be a perfect turbo for this car. I havent read much about twin scroll on a ej20 does it work well?

ive heard mixed results based on the flow of the heads it seems this is where all the 2.5 blocks as well as the jdm 2.0's do well they have better flowing heads as well as AVCS so spool is very quick im going to be running the ej255 from the 06+wrx and it has the single AVCS should be a blast on my vf43

[philly15]

Vf37 twin scroll for me, 325-350 whp in 1600-1800lbs is plenty

that thing is gonna rip a friend of mine has 30+ plus hours into his vf37 tune on his ej207 version 7 and i think he said it was in the neighborhood of 350 hp + tq to the wheels which is just insane for that car take the less drivetrain loss and 1800 lbs!? would be a lot of fun almost down right scary