Flattest torque curve.

[shinn497]

After driving a friend's sti and looking up several dyno sheets, I can't help but think that the subaru boxer is very peaky torque wise. What are some methods to mitigate that? Several discussion on engine performance are only about maximum power, which is very misleading, so this information is difficult to find.

I think a big issue is the inability of the turbo to keep up at high rpms, which makes me wonder. If you had a turbo that was capable of much more power but detuned it in the midrange, could you achieve this? The only issue I would find here would be that higher power turbos are larger, with more rotating mass, and therefore take longer to spool/are less responsive.

The other method, it seems, would be to improve flow through ported heads. But this seems kind of drastic.

I think having linear, and predictable, torque would be doubly important on a lightweight car with a smaller polar moment of intertia such as this.

[D Clary]

I am new to the Subaru deal but not turbos. Turbos by design loose efficiency at high rpm. A larger turbo is better, but you have greater lag before it spools. I dealy I guess you would have two turbos a small and large in series line a Ford diesel. Second best would be two small turbos, but that adds great expense and a lot of plumbing. Your best bet is to size the turbo properly for your engine and car to attain the power and torque when you need it.

[icky]

The main problem is usdm sti heads and the geometry in the motor, try looking at the ej207 graphs.

[Silvertop]

I am new to the Subaru deal but not turbos. Turbos by design loose efficiency at high rpm. A larger turbo is better, but you have greater lag before it spools. I dealy I guess you would have two turbos a small and large in series line a Ford diesel. Second best would be two small turbos, but that adds great expense and a lot of plumbing. Your best bet is to size the turbo properly for your engine and car to attain the power and torque when you need it.

Another choice is no turbo at all. Expectations of a broader torque curve is one reason I have chosen to go NA. Of course, it means trading off some significant amount of horsepower, so most builders will not choose to go that route. But it IS a choice.:)

[Xusia]

Another choice is no turbo at all. Expectations of a broader torque curve is one reason I have chosen to go NA. Of course, it means trading off some significant amount of horsepower, so most builders will not choose to go that route. But it IS a choice.:)

Not necessarily! Let's not forget, the EZ series H6 motors fit well! No one has specfically fitted an EZ36 (yet!), but bone stock that thing puts down some decent HP and TQ figures, and being normally aspirated it should have a flatter curve and really good throttle response.

If you really need an H4, another option would be to raise the compression ratio, and run less boost on the turbo. This should get you closer to what you are looking for.

If you want to consider an H6, but don't want an EZ36, you could still raise the compression ratio on an EZ30 and get some more power out of it.

Final option would be to go EV. [Electric] Motors put out a seriously flat torque curve...

[Frank818]

There are 2 other options lol

But before, I don't think a flat curve is a goal. Ask yourself why would you want a flat curve? You already answered that and a flat curve is actually one solution to achieve that goal, not the goal itself.


But for options,

First one, an electronic turbo controller. So that it would not boost too much in the mids and then you increase boost up to redline.

That option has 2 drawbacks:

1- you still need a turbo capable of flowing well at higher rpms, which means it may lag in the mids;
2- you have to sacrifice boost in the mids, which then reduces your peak torque (most probably, but will certainly give you less torque there), but it reduces it only in the range you are boosting less.

The advantage is that it may well achieve your goal, even though you lose torque in the mids. But you are not saying the highest torque curve and also flattest, you are just asking about a flat curve. :) So I dropped the mids for you and the curve is now flat.

Not sure in your example the only concern is turbo. With a high flowing engine and good balanced cams you can achieve a flatter curve at mid-high. Talking engine in general here, I don't know for Subaru engines if that is possible but it's seems icky has the answer to that.


The second option attacks the problem differently. Traction control.

With that you can have as much torque as you want and traction control will put it to the ground to the maximum limit of what the car and tires can take. In essence it will kill the torque peak you are afraid of, cuz in the end, like I said at the beginning, what you are asking for is control over traction to prevent unwanted reactions, spins, etc., not a flat curve. I see the flat curve as a solution to control traction and more easily put all that torque to the ground, just like traction control is also a solution to your problem.

:)

[shinn497]

The main problem is usdm sti heads and the geometry in the motor, try looking at the ej207 graphs.

22442

It was actually this graph of a tuned ej207 running e85 that inspired this post.

[shinn497]

http://www.rexnet.com.au/forum/index.php?/topic/167574-jap-crap-sti-build/

Here's an example with a non stock turbo that is gorgeously flat. What did he do?!

It does look like a lot of money went into this build :/

[JeromeS13]

http://www.rexnet.com.au/forum/index.php?/topic/167574-jap-crap-sti-build/

Here's an example with a non stock turbo that is gorgeously flat. What did he do?!

It does look like a lot of money went into this build :/

The dyno graphs posted show boost, not torque.

[WRX Dave]

My 2003 WRX has a flat torque curve from 3k-6500rpm with a slight taper to redline. That's on the stock turbo running a turboback with divorced wastegate downpipe, uppipe, 3-port wastegate solenoid, and a very good tune.

[Mechie3]

A fat torque curve is easy to get by under driving a turbo setup. The downfall is your max power/torque you make suffers. Take a peaky graph and draw a line somewhere under the curve. You can detune the setup to run that flat line. Smaller motors with turbos tend to have peakier graphs simply due to the dynamics of turbos have a certain efficiency range.

go here:

http://www.cobbtuning.com/Dyno-Database-s/70119.htm

Look up the Cobb Tuning GTR. You can see how stock is a nearly flat torque curve. Once they change turbos/boost/other things it's more peaky, but still makes more torque everywhere. It becomes are you looking for maximum area under the curve or predictability.

[305mouse]

One way to address this is what Wayne had stated. He increased compression and lowered boost. That way when the turbo spools it's not the "OMG vtec kicked in yo".