Just to say that you have a fantastic webzine, I have enjoyed reading all the articles. I am attempting to do a project, but have just recently changed my mind about how to achieve what I want. Basis is a Subaru Liberty RX with DOHC and 115kW. I want to get to about 180kW without forced aspiration. What should I do first? I only have an STi exhaust on at the moment.
It's not an uncommon email - in fact this is one that we got only a few weeks ago. And while Greg's obviously quite genuine in his request, it's what he doesn't understand which is so important.
The Starting Point
First up, let's look at what the Subaru engine already does. The 2.5-litre DOHC flat four develops its 115kW peak power at 5600 rpm, and its peak torque of 221Nm at 4000 rpm.
Now since kW = (Nm x rpm) / 9550, we can use that equation to work out that at the peak torque rpm of 4000 rpm, the engine is also developing 93kW. And at the peak power rpm of 5600 rpm, the engine is twisting 196Nm.
(bold = peak figures)
So, Greg wants to lift peak power from 115kW to 180kW - a 57 per cent increase. And there are only two ways he can do it - he can retain the peak power torque at 196Nm - but rev the engine 57 per cent harder. Doing this would therefore give max power at 8800 rpm.
Or, he can do it another way. How? By leaving the peak power revs the same (ie at 5600 rpm) and lifting the torque instead. So, if torque at max power rpm was increased by 57 per cent (to 308Nm) then the engine will also make 180kW, but this time at just 5600 rpm.
Two Paths to Power
So, the power gain of a modified engine comes from either increasing the torque at peak power rpm, or increasing the rpm while still developing the same peak power torque.
These are the only two ways of increasing engine power!!
But Greg, we gotta tell you that getting the Subaru engine to stay together at 8800 rpm is going to require damn good internals - otherwise it's pretty likely that the pistons will fly out of the engine as the rods let go! And of course the engine is still going to have to develop a pretty major 196Nm at those 8800 rpm. Hmmmmmmm.
"So," says Greg. "No problemo. I don't much like the thought of revving my little Subaru engine to 8800 rpm anyway - I'll go the increased torque route, thanks."
OK, now the torque developed by an engine is pretty well proportional to its breathing. An increase in torque of, say, 20 per cent will require an extra 20 per cent more air (and fuel) into the engine at those revs. How much air an engine actually breathes, compared with how much an engine could in an ideal world, is called its volumetric efficiency. So at 100 per cent volumetric efficiency, a 2.5-litre engine like Greg's will breathe in 2.5 litres of air every two rotations of the crank (ie every single four-stroke cycle). And in this engine at peak torque - that's 4000 rpm, remember - it's almost certain that the Subaru flat four will already have a volumetric efficiency of very nearly 100 per cent. These days, at peak torque most engines do.
Therefore, what you're looking at with this engine is a max torque possibility anywhere in the rev range of around 221Nm.
So Greg, you don't want to increase the max rpm of the engine. But with your max torque capability of 221Nm at 5600 rpm, you're looking at a peak power potential of 130kW.... That's not a whole lot more than standard, is it?
"OK, OK, but don't be so pedantic", says Greg. "What about a mix of increased rpm and increased high-rpm torque? Let's say I rev it to 6500 rpm - what increase in torque will I need then at those revs to get my 180kW?"
Well, you'll need 264Nm - an increase of 35 per cent in torque over what's now being developed at peak power. Can you get 35 per cent more torque at 16 per cent higher revs?
But you'll need to do a helluva lot of very expensive work. The valve springs or hydraulic lifters might start causing problems at the higher revs, you'll certainly need a pretty radical set of cams, the heads will need to be done, and the intake and exhaust will need to be tuned to the max. Like, as in probably needing a completely new fabricated intake and exhaust system. The rpm at which peak torque is developed will move up a heap, and the car will be lumpy and pretty horrible at idle. And cost? Oh maybe $15,000.
"But hold on, hold on!" says Greg. "The Honda S2000 develops 176kW - and I've got an extra half a litre of engine capacity!"
But the S2000 develops that power at 8300 rpm. At those revs, to develop that much power, it needs to have 'only' 202Nm. You'll need an extra 62Nm over the Honda S2000 to get your desired 180kW at 6500 rpm - and in an engine that hasn't got variable valve timing! So you can expect the Liberty to be an almost undriveble beast at lower revs.
"So let me get this straight. To get 180kW I can get the same amount of air into the engine every intake stroke as it does now - but increase revs to 8800 rpm. Or, I can keep peak power revs as they are - and be happy with 130kW. Or, I can go a bit both ways and get 35 per cent more air into the engine at 6500 rpm."
"But even that last compromise will still give me an engine almost undriveable on the road - the horrible cams and that? "
"Oh shit. There's stuff that I didn't understand here. What do you reckon I should do?"
Well, it's easy Greg.
You force the engine to breathe more air. Getting an extra 57 per cent more air into those cylinders at 5600 rpm shouldn't be much of a challenge - not if you turbo or supercharge it.... And you'll also gain very substantially in mid-range grunt, and the car will still be beautifully driveable, and economy will still be pretty good, and.....