Discuss Subaru EJ Motor Builds For Road Racing

[ssssly]

Wayne,

The units of measure can make a huge difference in understanding power. The confusion about HP and TQ is that people call them different things. So they think they are different things, when they are actually the same thing. Which is why I suggested calculating both as NM of force. It eliminates the confusion. HP is TQ/Time. They are the same thing, referred to as different things depending on where you plot them on a timeline.

The power at the crank is of course identical regardless of the rest of the drive train. But we are talking about measuring wheel horse power. Which means we must take into account the gearing.

The express purpose of gears is to multiply or divide force. So since you are measuring the power at the wheels, after it has gone through the gears, the ratio of the gear you are in is directly proportional to the amount of power/time. People do 4th gear pulls in a Subaru because it is the gear closest to 1:1, the actual crank power minus drivetrain loss. The total amount of power/time will absolutely be different depending on the gear, its what the gears are there for.

Power loss in a drivetrain is caused by changing the direction of the force, the power required to rotate the mass of the components and heat. Not by the gear ratio.

Gears change the % of power delivered at a given time in reference to the rotational speed of the final output (the wheels).

The crossover will absolutely change dependent on the amount of time elapsed to reach a certain RPM. Its what the crossover measures.

If the volume of the system remains the same, an increase in CFM input will increase PSI and the converse. As RPM rises, in order to maintain an equal PSI you must increase CFM because the engine is taking more volume from the system.

And the CFM:PSI in a compressor map is a direct correlation to piping volume. The smaller the volume of the piping, the higher the pressure at a given CFM.

If the piping is too small, the required PSI to flow the same CFM increases too much. The turbo then has to spin faster to overcome the existing pressure to meet the same CFM. Creating unwanted heat and pushing you out of the efficiency range.

If the piping is too large, the turbo must flow more CFM to maintain positive pressure relative to atmosphere and the volume consumed by the engine. Again having to spin too fast and creating excess heat pushing you out of the efficiency range.

The CFM:Power ratio in a forced induction engine is not 1:1. The engine only operates at 100% efficiency under boost. If you are under boost you are not at lambda, you are rich. Its generally closer to 1:1.45 depending on how rich you are running at full boost.

And yes you would want to detune it through the mid range. If you are trying to achieve HP (X) at at 9000RPM you need a turbo that will flow that CFM. If the turbo is capable of flowing CFM for HP (X) at 9000RPM it will flow considerably more at 5500 RPM, where your HP/TQ number are most likely going to cross. Hence if you can spool it up that fast, you are going to well exceed your desired peak HP number. You will need to bleed of WG gas to keep the power at the level you want from the crossover to redline. IE detune it through the midrange 38-5500 RPM. If you don't start slowing spool down a few 1000 RPM before your goal, you will overshoot unless you have an external WG.

If you choose a turbo that has a max CFM equal to the power that you want at crossover, you will be horribly disappointed. If it is flowing its max efficient CFM at 5500 RPM it will never hold that CFM to 9000 RPM. It will hit the desired number and trail off rapidly in direct proportion to RPM.

And it is way easier to deal with a power curve that builds and then stays even to redline than one that has a massive spike at 3800 RPM and then drops at 5500. The last thing I want to drive is a light, rear engine car that makes way more power prior to crossover and then drops like a rock. Is a good way to downshift, break the tires loose and kill yourself.

All HPDE that I'm aware of is based on time. But none I know of other than a handful of obscure drift categories have pw/wt requirements. And they generally curb allowed power via restrictors. But I certainly don't know all the rules for all the classes of every kind of racing.

But giving up max high end power for quicker low and midrange power is normally a pretty good tradeoff for things like gymkhana and rallycross.

[wleehendrick]

The confusion about HP and TQ is that people call them different things. So they think they are different things, when they are actually the same thing. Which is why I suggested calculating both as NM of force. It eliminates the confusion. HP is TQ/Time. They are the same thing, referred to as different things depending on where you plot them on a timeline.

I'm not going to write a dissertation here, since the HP/torque debates pops up all the time on car forums with lots of misconceptions. Don't take this the wrong way, but I will say that before correcting Wayne, you might want to get your physics down; a Newton is a unit of Force, a Newton*meter is a unit of Torque (or Work, but then it's called a Joule); there's no such thing as a 'NM of force'.

Although interrelated by time and distance, Power, Torque and Force are different, they have different units and it is improper to suggest thinking of them as the same; you're mixing them up and using the wrong units which just adds confusion to a topic that a lot of folks don't completely understand.

Of course, if you know the torque curve, the power curve can be derived (and vice-versa).

The crossover will absolutely change dependent on the amount of time elapsed to reach a certain RPM. Its what the crossover measures.

Maybe you're referring to something else, but as Wayne mentioned, the crossover point, when power (in HP) is equal to torque (in lb-ft) is at 5252RPM. Always. There's nothing you can do to the motor to change that. (However, the crossover RPM would be obviously change if different units were used, kW and N-m for example)

[ssssly]

The term Newton Meter can be used to measure two separate things.

"A Newton Meter of Energy" or a "Joule" which describes energy and is defined as: When a force of 1 Newton is applied to an object for 1 meter.

"A Newton Meter of Force" which describes torque and is defined as: When a force of 1 Newton is applied at a distance of 1 meter from a pivot at a right angle to a radius.

The difference is a semantic and dependent on local diction. I learned the phrase irregarldess from my teachers as well. Appologize if it caused confusion.

The crossover point is irrelevant and exactly the point. It makes zero sense to say I want to make X torque and X HP. When at 5251 RPM you want 400ftlbs and at 5253 RPM you want 400hp.

5252 RPM is the crossover point because it is the conversion factor for ftlbs to HP. Which is exactly why it is arbitrary in reference to determining power. The time is based off of RPM and can be calculated at any point. There is no place where your HP number is independent of your TQ number.

For every single plot on the readout from a dynamometer, the HP number is determined by measuring ftlbs of torque, multiplying it by RPM and then dividing that number by 5252. It is never an independent measurement of its own. The only way to increase HP anywhere from 1 to 10,000 RPM is to increase torque.

Your max torque will be where your engine hits max efficiency, on a turbocharged car, where your turbo is fully spooled. Lets say thats 4000 RPM at at 400ftlbs at a 1:1 gear ratio. At 4000 RPM that will be 304hp. If you were producing the same torque at 6000 RPM you would be making 456HP. At 8000 RPM 609hp. But at every plot on that line you would still be making 542 N-m.

So if you are trying to calculate via HP, and you are class restricted, you have to calculate everything backwards. Whatever max hp is at max RPM, figure out what the torque is at that RPM, then interpolate that backwards to max torque. Making sure not to overshoot it to begin with.

If via poor wording I implied that 400hp was = to 400ftlbs other than at 5252 RPM I apologize.

That is however exactly why it makes it a poor unit of measure, particularly when compared and contrasted to torque which it is a time dependent factor of.

[apexanimal]

bored...

[metalmaker12]

Blah blah boring

[wleehendrick]

The term Newton Meter can be used to measure two separate things.

"A Newton Meter of Energy" or a "Joule" which describes energy and is defined as: When a force of 1 Newton is applied to an object for 1 meter.

Hi ssssly, correct.

"A Newton Meter of Force" which describes torque and is defined as: When a force of 1 Newton is applied at a distance of 1 meter from a pivot at a right angle to a radius.

Nope... That's a Newton-meter of Torque, not Force. A Newton is the metric unit of Force. Informal semantics can be a bit of an issue, and not to harp on it, but being a scientist and engineer, and having seen a million dollar program over-run due to different interpretations of a poorly worded requirement that could have been avoided with proper technical terminology, I'm kinda sensitive to it!

Most of what you said I pretty much agree with, but I can't reiterate enough that Power and Torque are very different quantities (despite being interdependent). Power is the ability to do a given amount of work in a given amount of time and if the goal a given acceleration, it's power that's important; it's irrelevant whether that's made by high torque at low RPM or low torque at high RPM, it's the resultant power (under the curve, not peak, of course) that determines acceleration. Gearing changes torque at the output, but not power. This is why power is an important metric, and why power (not torque) to weight ratio is used for classification.

Anyway, enough theoretical ramblings, let's get back to discussing EJ's where you experts have much more to contribute than I!

[Sgt.Gator]

Here's an example of what starting with 30 psi and tapering to 21 psi will do, 500+ Hp and 500+ Tq, and a small fortune spent to get there:

http://forums.nasioc.com/forums/show....php?t=2076135

[Wayne Presley]

Well back to the original question, here's what I'd do for a power to weight class (specifically STR2).
2.5 short block with 9.5 CR pistons
cams with 270-274 duration 10.2-10.5mm lift
GT2871 or 20g turbo
750cc injectors
Dry sump
ARP head studs

Set up the tune to make 256 RWHP from 3500 to 7000 rpm using the ECU to vary the boost. That way when ever you step on the gas you the max allowed HP on tap. Now the real thing to do with the power to weight class is to improve the drag coefficient on the car, rotational enertia and reduce rolling resistance.

[Scargo]

Thanks, Wayne, for that input. What heads would you use? AVCS or DAVCS or none? Stock, bowls smoothed only, ported, big valves? As I've said earlier, big torque has been made with stock cams in a DVACS motor.


I like the suggestion that drag and rolling resistance should be looked at. For ST2, I wonder if FFR tried a smaller or different wing while doing the wind tunnel aero study?


Previously, someone mentioned building/tuning for shorter venues or fast, long straight tracks. I have always thought that was normally the job of gearing. Realistically, it seems beyond the reach of the amateur racer to be frequently changing gearing, though swapping out a whole transaxle would be straightforward.
I am wondering what others think about this as it relates to engine configuration decisions. Limit the RPMs to 7K? Can a track motor be reliable at 8K when it only has to deliver 260WHP? I've already been pushing 7.5 with my stock rod/crank motor in the STi and it's still doing well. This with bigger oil pump and all the trimmings.


I have chosen to go down a path of 2.34L, destroked screamer. I don't have the heads or cams yet. Still scheming.

[Sgt.Gator]

My race car has completed one season of Grand Am Cup with the stock turbo, one season of Time Attack with a FP 18G turbo and Tarzan Yamada at the wheel, and now almost one season of ICSCC ST class and ICSCC Enduro racing (fingers crossed) with the same 18G and stock internals. No rebuilds or internal mods. It's an EJ255 or 257 (never bothered to check which) with external mods that were updated by Cobb - Surgeline last spring, the mods list is below. I have plenty of power across the RPM band. I have my tachometer warning lights set to go amber at 6200, red at 6500, rev limiter at 6800. I generally shift at 6400-6500. You can see from my Dyno there's no point in going higher than 6700, and it will last much longer that way. With LGT gearing I hit 156 mph at the end of the Spokane front straight, and I still have a few more rpm to go, the tach lights are in the amber.

If you have the usual WRX/STI gearing than yes, you'd probably need to build your engine to run higher rpms or you may run out of top speed. You can figure all that out with the M Factory online gear calculator. See my post in the Transmissions sticky for which trannys have what gear ratios and the calculator.

I intend to use my 5 speed LGT tranny in the 818 and put a JDM Forester STI 6 speed in the LGT race car (super gearing!) . So my 818 build engine will be much like the one I have now, lower rpm max, more low end power.

Cobb-Subaru%u00252BLegacy%u00252BWagon%u00252B16psi.jpg

The exact mods Cobb did, keeping the 7 year old Forced Performance 18G:

COBB Subaru 1000cc Top Feed Injector Set
Motive Autowerks Composite TGV Deletes- Metal Intake Manifold
Aeromotive EFI Bypass Regulator
COBB Lightweight Pulley
COBB Tuning Electronic Boost Control Solenoid for '08+ STi/LGT
Most all lines replaced with Aeroquip
AEM 3.5 Bar MAP Sensor
COBB Accessport V3
SPAL cooling fans.
A ton of little fittings, hose ends ect.

[Wayne Presley]

You can always get more flexibility with the dual AVCS heads but I would think you could get plenty out of a non AVCS motor given bigger cams.

Thanks, Wayne, for that input. What heads would you use? AVCS or DAVCS or none? Stock, bowls smoothed only, ported, big valves? As I've said earlier, big torque has been made with stock cams in a DVACS motor.


I like the suggestion that drag and rolling resistance should be looked at. For ST2, I wonder if FFR tried a smaller or different wing while doing the wind tunnel aero study?


Previously, someone mentioned building/tuning for shorter venues or fast, long straight tracks. I have always thought that was normally the job of gearing. Realistically, it seems beyond the reach of the amateur racer to be frequently changing gearing, though swapping out a whole transaxle would be straightforward.
I am wondering what others think about this as it relates to engine configuration decisions. Limit the RPMs to 7K? Can a track motor be reliable at 8K when it only has to deliver 260WHP? I've already been pushing 7.5 with my stock rod/crank motor in the STi and it's still doing well. This with bigger oil pump and all the trimmings.


I have chosen to go down a path of 2.34L, destroked screamer. I don't have the heads or cams yet. Still scheming.

[Scargo]

For my R, I'm building a 2.34 destroker with DAVCS, stock cams and only modding the heads for high RPM work. I'm trying to maximize torque but limit the power to around 300 for ST-2 or 8:1 P/W ratio.
I'm looking at a Borg Warner EFR 7064 twinscroll without internal wastegating. BorgWarner only has a 1.05 A/R for that arrangement. From 6500 RPM up, I will need to waste 50 to 70%. ssssly seems to indicate that doing so is not a big deal; even a plus for keeping the turbo spooled. Can an IWG setup be as efficient and accurate at wastegating or will I be frying it with so much exhaust through the wastegate? I see there are water-cooled EWGs...
I can go down to a .92 A/R with an IWG. With an IWG I can go to a .92 A/R. But that's internal and seems like it would stay very hot. Would it be reliable?
The way the way ssssly is talking about tuning the turbo is new to me.
Additionally, I also don't know why I'd want a manual boost controller and might that be illegal?