Why the low BHP Per Liter?

[JamesUK]

So I'm really curious why our motors are so poor when it comes to efficiency of generating power. When I was a boy 100 bhp/per liter was the holy grail and the sign of a good engine (I'm talking normally aspirated here). Now it’s closer to 130bhp and if we really want to talk efficient N/A combustion engines I recall our old 3L V10 motors putting out 330+ BHP Per L (ok so we had to heat them up in order to turn them over due to the Tol. of the internal parts and they only lasted 300 miles but….).

I have a stock US 3.2 that when new pushed out an amazing .....64 bhp/per L. Yes I know it’s a US motor....emissions...poor gas.....designed to do more than 300miles etc.
I'm looking at having the motor rebuilt as its getting a little tired after 30 odd years and 225k. In doing my research I keep coming back to the conclusion that basically no matter how much I spend I'll end up with an engine pushing out 80 bhp/per L.
This seems a little...disappointing. Especially when one considers the $20k+ investment involved in 3.2 (3.4/3.5)rebuild or 3.6 conversion. I’m thinking that for that kind of investment we (I) should be getting a little more than 78 bhp/per liter. With an understanding of what is setting this celling I'm having a really hard time handing over the money.

Can someone explain what it is about the design of the engine (not just "well it’s an old design from basically per war, what do you expect) that results in this condition?

[Tippy]

1. Old school combustion chamber design (snowman design some have done I believe is better)
2. Inherent design of air-cooled engines don't allow "zero-decking" (squish is a good thing)
3. 2 valves per cylinder vs 4+ valves (self explanatory)
4. Offset spark plug instead of dead center (flame front is not evenly distributed rocking piston in bore)
5. Short connecting rod lengths (longer leads to longer dwell time and less severe angle on the crank throw)

I'm sure there are many more, just some ones that popped into my head

[Matt Monson]

Honda turned the world on it's ear when they released the B16A because it was the first street car production engine to make 100hp/l. How did they do it? Variable Valve Timing. Your 3.2 Carrera is a decade too early for that sort of tech.

ps. It's a design from the early 60s. That's 20 years after the war.

[JamesUK]

Whilst I have no intention or interest in getting involved in a keyboard pissing contest I think we are all aware that other engines (including Bob Linton's very expensive 2 valve 4.0L Air cooled) have got over 100bhp/L without the aid of VTEC. As for the base of my engine I happen to believe it has more then a few areas that have more then a passing resemblance to the motor in the back of this old girl ....
But thats just my opinion and we all know what they say about those....

[gtc]

Just be glad you own an '84 911, and not an '84 Corvette.

[Tippy]

Originally Posted by gtc

Just be glad you own an '84 911, and not an '84 Corvette.

What’s wrong with 36hp/L? 😂😂😂

[Fly911]

100 to 110 HP/L is "easily" within reach for most aircooled Porsche engines. It only depends on the thickness of your wallet.
The original (Euro) 84 to 89 Carrera 3.2L (3,164cc) had 73HP/L, and with normal driving and maintenance can get 200k miles.
That engine can produce 400 naturally aspirated HP if you do the right things. I'm expecting around 380 to 410 Hp on my 87 Carrera build. But I don't expect it to last for 200k miles.

[Matt Monson]

Originally Posted by JamesUK

Whilst I have no intention or interest in getting involved in a keyboard pissing contest I think we are all aware that other engines (including Bob Linton's very expensive 2 valve 4.0L Air cooled) have got over 100bhp/L without the aid of VTEC. As for the base of my engine I happen to believe it has more then a few areas that have more then a passing resemblance to the motor in the back of this old girl ....
But thats just my opinion and we all know what they say about those....

You actually do seem to want to piss and whinge and in the process aren't reading very well. I said PRODUCTION engine. I didn't say you couldn't get there. I answered WHY weren't PRODUCTION engines making 100hp/l earlier. Power is largely made in the heads, and once they had variable timing it became pretty easy.

As another mentioned, 100hp/l is very doable with the right pocketbook. I will be doing it in the near future on a 2.5l engine. It's going to run me between $25-30k. Porsche did it 45 years ago. I'm mostly just copying their plan for it.

But to get there you have to do things that compromise fuel economy, emissions, and longevity. And once you raise the compression a certain amount, pump gas isn't even an option. So beyond tech, the answer why they didn't do it is the market. The market is limited by the above variables that production vehicles must meet.

As for your Beetle reference? Nope. Flat 4 isn't flat 6. Beetle engine is pushrod and a single cam. The engine architecture is NOTHING like a 911 engine.

[Eagledriver]

The letters, RPM.

I suspect your Honda example has a higher RPM for peak HP.

-Andy

[Tippy]

Originally Posted by Eagledriver

The letters, RPM.

I suspect your Honda example has a higher RPM for peak HP.

-Andy

Yep!!!! RPM is a huge factor, as long as cams and system airflow supports it.

[JamesUK]

"You actually do seem to want to piss and whinge and in the process aren't reading very well."

Ok I’ll just gloss over that.

"I said PRODUCTION engine. I didn't say you couldn't get there."

My fault I did not type production into my response. I was thinking production when I wrote my reply (perhaps the confusion comes from mentioning Bob Linton’s engine). My mistake as this is my whole point. Many other production engines are able to achieve much more efficient power figures then ours, some are not even what you would consider to be performance car (Rover K series).

"Power is largely made in the heads, and once they had variable timing it became pretty easy.
As another mentioned, 100hp/l is very doable with the right pocketbook. I will be doing it in the near future on a 2.5l engine. It's going to run me between $25-30k. Porsche did it 45 years ago. I'm mostly just copying their plan for it.
But to get there you have to do things that compromise fuel economy, emissions, and longevity. And once you raise the compression a certain amount, pump gas isn't even an option. So beyond tech, the answer why they didn't do it is the market. The market is limited by the above variables that production vehicles must meet."

Ok so now we are getting somewhere. If we agree that most of the efficacy can be gained from the heads (Which I believe is the base of most of “Tippy’s” comments what are you planning to do to your 2.5 to gain the power? I dont think your planning on going 4 valve or VTCC so Id love to know how? This is the core of my question after all. I personal am not too concerned with MPG but I do live in CA so passing SMOG would be nice and it lasting enough time between expensive rebuilds as to not result in my wife kicking me out of the house would be great so may I ask what you and “Fly911” foresee your life cycle to be? To that point “Fly911”, you sound like your building a fire breathing monster, care to share how?

As for your Beetle reference? Nope. Flat 4 isn't flat 6. Beetle engine is pushrod and a single cam. The engine architecture is NOTHING like a 911 engine.

Understood. I just happen to think (key word is “I”) that no engineering firm forgets what came before (but thanks I now understand what those other two spark plugs are for) . I see similarity’s, you don’t. Tomayto/Tomahto

[Fly911]

To that point “Fly911”, you sound like your building a fire breathing monster, care to share how?

I have been in the process of building an engine based on the original matching numbers 3.2L for my 1987 Carrera. The matching numbers G50 gearbox is finished, with shorter 3-4-5 ratios, GT3 LSD and LWF and clutch.

The 3.2L case has been machined to fit 102mm LNE light weight pistons, balanced to +/- 0.1 gram. The crank is a 997 GT3 (76.4mm) that has been knife edged and balanced. The con rods are 997 GT3 titanium from Pankl, also balanced to +/-0.1 gram.
Cam shafts are from Elgin, inlet duration is 322 degrees, outlet is 318 degrees, high lift. Rockers modified to fit cams.
Fuel injection/ECU is from Canems (in the UK), with 48mm Jenvey ITB's. Heads are modified with twin plugs, 993RSR valves and ceramic coated.
Headers are Fabspeed RSR with 997 GT3 titanium muffler.

The car itself looks original on the outside, but has Bilstein coil overs and 930 Turbo brakes. The front control arms are lengthened by 20mm. Camber is negative 3.8 degrees on all four.

[Tippy]

Ah yes, ITB’s help too. Cylinders don’t have to fight for air due to a common plenum.

[Geronimo]

So Fly911, with that build you are shooting for 270-300CHP? Wondering, how does that build cost stack up to a stock 993 3.6 vario?

[schleven]

Originally Posted by Geronimo

So Fly911, with that build you are shooting for 270-300CHP? Wondering, how does that build cost stack up to a stock 993 3.6 vario?

I would assume he's shooting for more, as he's got ~3.75L of displacement.

[Fly911]

Originally Posted by Geronimo

So Fly911, with that build you are shooting for 270-300CHP? Wondering, how does that build cost stack up to a stock 993 3.6 vario?

My engine is very different from a 3.6L VarioRam. All internal parts are 997 GT3, crank, oil pump, con rods etc. It's also 3.75 liters (Porsche calls it 3.8), with a red line over 1,000rpm above the 993 3.6L. Target is 400Hp, but anything above 380 is acceptable.

My goal was to see how far you can stretch a 3.2L Carrera engine. It is possible to stretch it even further with a 997GT3 4.0L crank (80.4mm) and LNE 102.3mm pistons/cylinders. That would make it 4 liters. But with the smaller spigot spacing than the 3.6L (964/993) engines, the 102mm P/C kit is already a stretch.
I'm "religious" on keeping it a matching numbers car, so a 3.6L based engine was never in the picture.

I am however looking into changing the G50/01 internals to a G50/20, and re-gear 2nd, 3rd 4th and 5th, and keep the old 5th gear as 6th. The gear housing and differential is pretty much identical between the G50/01 and G50/20, only the front cone for 6th gear would needed to be modified.

It would be awesome with a matching numbers '87 Carrera with a 400Hp 3.8L and 6 speed gearbox!

And on the outside, no one would be able to tell the difference (except the dual center exhaust outlet from the GT3 muffler).

[Geronimo]

Whoa geeez, so the engine costs more than my house. LOL

[Jeff Alton]

Fly911's motor should reach his goals. Be interesting to see if the 48mm Jenvey's are big enough (we use that size on our 3.4 builds). Don't know much about that cam but it does have plenty of duration! Keep us posted on the build and tuning!