Specific power output increase - How ?

[thevojtiseks]

I have a early 2.2S 911 which has a 2.2 liter engine and 180hp at 7200rpm. This gives a specifc output of around 82 bhp per litre. I'm looking at the latest Porsche engine offering and I see specific outputs around 93 bhp per liter at 6,600rpm in the carrera S and 87.5 in the Boxster S at 6,200. My question i how do they do it ?? I had always assumed higher rpm's = higher bhp, but this seems to be going the other way. I assume the vvt and variable intake don't do much for peak hp but help spread out the power over a wider rpm range. Thanks !

[Steve@Rennsport]

Hi:

These engines have 4-valve heads that can flow a LOT more than the average air-cooled 2-valve heads can and through a much wider range of valve lifts.

Add high-compression, VVT, and variable geometry intake systems and you can achieve 100 HP/litre and still remain drivable and smog legal.

[jluetjen]

Higher RPM = Higher HP is the "easy" way to do it. Getting more HP without increasing the RPM is the better way to do it. The F1 guys have gone in the same direction as Porsche has too, For the last few years their engines have stayed at about 19,000 peak RPM while their HP has increased (in general terms ) from just short of 900 HP up towards 950 HP. This is inspite of having longer engine lives.

In the case of Porsche, when comparing the air-cooled to the water cooled engines, they are almost "chaulk and cheese". The big improvement over the air-cooled engines are the 4-valve, penta-roofed heads. Copied from the Cosworth DFV, these heads have or allow the following benefits:
- Increased CR: More HP across the rev range, better combustion.
- As Steve said, better flow, especially at low lifts. So this also increases the flexibility and HP across the rev range.
- Increased turbulance by using the port design to cause "tumble" as the air flows into the chamber. This causes the mixture to be more consistant across the combustion chamber.
- The compact combustion chamber increases the thermal efficiency and reduces the requirement for spark advance.

[thevojtiseks]

So, in the case of this 2.5l giving a claimed 285hp, this likley be generated by the use of higher RPM's, higher CR and better breathing ??

http://cgi.ebay.com/ebaymotors/ws/eBayISAPI.dll?ViewItem&rd=1,1&item=4615316162&sspagename=STRK%3AMEWA%3AIT

Has anyone attempted to make 4 valve heads for an Air cooled engine ??

[jluetjen]

Interesting engine. Basically the performance of the 2.5ST motors falls pretty much on a straight line between the earlier 906 engine, and the later RSR motors (and BTW the 908 and 917 motors which in many respects are nothing more then 8 and 12 cylinder versions of a 911 motor). In the past I've posted data on the ST motors, both the long-stroke and the short-stroke versions. Generally the performance per liter of those engines is pretty much all on the same level. You listed the factors that provide that extra HP (not to mention the changes made to the engine to ensure long life at 8000+ RPM). For comparision, the 2 liter Cosworth BDG engine (think of it as half of a DFV that has been bored and stroked) made 280 HP way up at 9250 RPM, and 175 lb-ft of torque at 7000 RPM compared to the ST's 270 HP at 8000 RPM and 192 lb-ft at 5300 RPM.

Porsche discovered when they developed the 4-valve heads for the 956/962 that the 4 valves don't leave enough room to flow the air required and provide the surface area to cool the heads. So they developed water-cooled heads for the 956/962's in the 4-valve configurations.

[cnavarro]

Although not a 911, the Apfelbeck (sp?) Project has been developing aircooled three-valve heads for the Porsche 914 and I'm pretty sure there are running prototypes. They worked out the cooling problems with getting air around the valves and behind the combustion chamber for cooling. Pretty neat stuff.

[Porschekid962]

While the valves may be the main bottleneck in pumping more power out of early air cooled engines, keep in mind that as you add rpm's you need to update or upgrade just about everything else in the motor. Rods, rod bolts, pistons for lighter weight, valves, valve springs, retainers, guides, bla bla bla. Any reciprocating mass needs to be minimized in order to minimize wear and tear mostly on the valvetrain.

The last 3.0litre F1 motors were spinning above 20k rpm, all that crap you see on tv is bunk. They use specially designed pneumatic valvetrain systems with hollow stem Titanium valve springs that are super short in length. Plus some of the teams that have lots of money dont even use cams, everything is electronically controlled. Aside from spinning the motor up around 10k+ rpm the only other way to increase power realistically is with a turbo or forced induction system. If youve read the book on the building of the McLaren F1 the intake scoop on the roof was good for about 25 hp above 170mph.

Just drop weight from the car, its a lot cheaper than bolting some one off 4 valve heads onto the motor, if you have the bank get some 959 heads and have at it or send them my way and lets see what can happen!

For ultimate NA power I think we might be seeing more and more water cooled transplants into earlier chassis, its logical, it makes sense and the power is phenominal if you have the dough.

[jluetjen]

You're on the right track Kid962, but I have to comment on a couple of points...

Originally posted by Porschekid962

The last 3.0litre F1 motors were spinning above 20k rpm, all that crap you see on tv is bunk.

What crap? With the extended life engines the peak HP engine speed seems to have stabilized just short of 20K RPM. They can spin above that, but it shortens the life of the engine which results in pretty significant penelties if the engine doesn't make it for 2 weekends. The interesting point is that there are ways to increase the HP by minding the details even without increasing the rev's. The F1 guys have done it, and I have no doubt that guys like Steve and Henry know how to do it for air-cooled 911 engines.

They use specially designed pneumatic valvetrain systems with hollow stem Titanium valve springs that are super short in length.

Actually I think that the last engine supplier to use metalic springs was Asiatech a few years ago. Nowadays everyone is using pneumatic valve springs and finger followers. The finger followers (like 911's use) allow a more radical cam profile then bucket followers due to the mechanical advantage.

Plus some of the teams that have lots of money dont even use cams, everything is electronically controlled.

That's been rumored, but the case has been made by a number of engine designers that the weight (high-up in the engine) of such a system is prohibitive. The old fashion mechanical camshaft system is actually pretty efficient.

Aside from spinning the motor up around 10k+ rpm the only other way to increase power realistically is with a turbo or forced induction system. If youve read the book on the building of the McLaren F1 the intake scoop on the roof was good for about 25 hp above 170mph.

Just drop weight from the car, its a lot cheaper than bolting some one off 4 valve heads onto the motor, if you have the bank get some 959 heads and have at it or send them my way and lets see what can happen!

For ultimate NA power I think we might be seeing more and more water cooled transplants into earlier chassis, its logical, it makes sense and the power is phenominal if you have the dough.

I'm with you here.