Porsche 993 twin turbo

[englertracing]

If you could find a 3.6,3.7, 3.8 final drive you could offset the drawbacks of larger turbos, bigger cams, etc. You can create more torque at the contact patch. Yes you will lose some top end, but the reality is, normal driving doesn't allow for speeds in excess of 150mph on a regular basis.

The two turbine hot sides under consideration are K16 and K24. Both came factory stock with this car, the European S model getting the larger K24 hot side. The compressor wheel of K16 is too small, so staying with the K16 turbos would mean installing a K24 (or similar) compressor wheel. The compressor wheel itself isn’t a big factor in lag or transient response, turbine geometry is. Compressor wheel does to a large extent determine the peak hp, however.

On the cams, I think that by increasing the duration by increasing overlap one can have one’s cake and eat it, too. This is because the increased overlap will in my opinion help across the whole rpm range above idle. The constraint here is passing the emissions at idle. Again, not planning to go overboard here, just the types of cams that the normally aspirated 993’s have. Like hydraulic 993 Euro RS emissions passing street car cams.

The third physical change would be a free-flow exhaust. Many wools say that reducing the flow restriction after turbo helps the turbo spool up. I don’t think this is true, because the mass flow of exhaust at say 2500 rpm is so low compared to 6500 rpm that I don’t see how the exhaust can be chocking off anything at 2500 rpm. Free flow turbo back exhaust will help the top end, though.

With these changes and the software calibration to complement them, I think one could add more power everywhere while keeping the lag at most at the factory S level and possibly better. The gearing changes might not be required.

what about in addition to those things an e85 conversion and a bump in CR?
also having the ability to spin the tires at any speed is something fast drivers like.....
its the drivers job to manipulate the throttle to maintain traction.

[David Redszus]

There is a large difference in design factors between NA and turbo engines. What works for one does not for the other.

Similarly, there is a big difference between aero cars and non-aero cars.

All must be tuned and driven differently.

When a high performance car is driven below its limits, you can get away with all sorts of errors.
When driven at or approaching its limits, it becomes a whole different story.

Most cars are driven well below their limits.

[ptuomov]

The gearing changes can make the car more fun to drive everywhere. It will accelerate faster and have less "dwell time in the lag zone". Porsche loves themselves some long gears. I would make that change anyways regardless of everything else. Plus you can rev higher with the longer duration cams so you may not lose that much on the top end.

Here are some GT2 Evo options:

716651B6-4FBD-42C8-A4A6-EC8B2DAE1632.jpeg

The current plan is not to change gears, at least right away. The current plan is to make more somewhat torque at both the low end (2500rpm) and high end (6500rpm) than the stock engine, so the stock gearing should be a good compromise for now.

[David Redszus]

These old heads need to take a lesson from replacement Harley hemi heads that now have a squish area to make more power .

I am skeptical. I think that even a quarter century later, these are pretty much the two-valve heads to beat for turbo application. They have just the right amount of tumble and fairly complex and well thought out squish setup with the piston actually protruding about 5mm into the cylinder head.

Those heads with large valves and twin spark plugs look very much like my 1963 Porsche GTS heads. Pistons are different.

If the intended use is to be road racing, then the engine power is adequate (if it stays together) and other factors become
important, such as track layout, torque curves, brake balance, handling, tires, and gearing selection.

Perhaps we should also include proper engine managemnt strategies, as well.
And budget strategies.

[ptuomov]

These old heads need to take a lesson from replacement Harley hemi heads that now have a squish area to make more power .

I am skeptical. I think that even a quarter century later, these are pretty much the two-valve heads to beat for turbo application. They have just the right amount of tumble and fairly complex and well thought out squish setup with the piston actually protruding about 5mm into the cylinder head.

Those heads with large valves and twin spark plugs look very much like my 1963 Porsche GTS heads. Pistons are different.

If the intended use is to be road racing, then the engine power is adequate (if it stays together) and other factors become
important, such as track layout, torque curves, brake balance, handling, tires, and gearing selection.

Perhaps we should also include proper engine managemnt strategies, as well.
And budget strategies.

These turbo heads work pretty well with just one spark plug, because the low geometric compression ratio allows for a relatively flat top piston. I believe that Porsche only put the second spark plug in when the piston crown got too big and started dividing the combustion chamber too much.

The 993 Turbo ECU is actually pretty smart in its stock form in that it takes away the punch bowl (of boost and ignition timing) when the temperatures rise. It counts knocks and monitors all sorts of temperatures, uses mass air flow (instead of manifold pressure) measurements to control "boost", etc. It appears to be a very durable engine, partly because the component protection strategies of the ECU are very robust.

I think there's a little bit of room to make more power reliably with this engine because of the following reasons. First, the climate is never that hot where the car is being used. Second, I'm putting in the second Porsche oil cooler that the factory installed in the 450hp models. Third, I have access to very consistent 98E5 brand name gasoline where the car is being driven. I'm involved with a team that races a series that mandates that pump gas and it's very consistent there. Fourth, I don't mind the slightly higher sound level and can therefore reduce the turbine outlet pressure a bunch.

Together, those four factors should allow for a bit of hot rodding without ruining the perfectly nice car. Time will tell.

[David Redszus]

Porsche race engines from that era were often boosted to produce 750-920hp. They were run in tube framed, mid-engine
FabCar chassis.

They were scary and expensive...I mean scary expensive.

[ptuomov]

Porsche race engines from that era were often boosted to produce 750-920hp. They were run in tube framed, mid-engine
FabCar chassis. They were scary and expensive...I mean scary expensive.

The race car version of this car, 993 GT2 Evo, had 34.2mm compressor inlet restrictors. The official max power level with those restrictors was 342kW/465hp at 5750rpm and torque level 670Nm at 5000rpm. The engine long block was internally basically completely stock with a deeper exhaust valve relief cut into the stock piston for the GT2 Evo M64/83 camshafts and some head oil channels modified for the mechanical lifters associated with these cams. I don't think this particular car/engine model ever raced under factory banner at 750-920hp levels, but I might be wrong. Purpose built race cars of course did.

[ptuomov]

A popular and in my opinion sensible modification for a street car is to install the K24 compressor wheel from the 450hp model into the K16 turbo of the 408hp model.

Here’s one result from such a modification with some supporting changes. First, the compressor maps. Then a German tuner dyno result. I’ve hand plotted the stock torque curve on the dyno curve:

29756565-ACC6-4885-9AE5-0421BF8079F8.jpeg

6DAF50DD-5A9C-4456-8871-09CCEB6704B6.jpeg

0B00CDB7-2157-462D-9E69-661DAA9C7D3E.jpeg

[ptuomov]

One of those GT2 Evo engines from another group:

860B7F1B-AB2C-42A3-AC4C-C01DD6AC8B08.jpeg

4CC3C3E6-5ED8-4AA6-9A16-769CD95F5F1A.jpeg

76EC9036-FE6C-4594-92E1-51D1D62FE73D.jpeg

510378ED-39A0-4BA7-B59D-B80DCE3D4B28.jpeg

E40DA9ED-D1AE-4DBA-A53C-2A276F0BA96D.jpeg

[BLSTIC]

That's an amazingly flat power curve. Looks like it would hardly be worth changing gears in a lot of situations

[ptuomov]

That's an amazingly flat power curve. Looks like it would hardly be worth changing gears in a lot of situations

I’m thinking that this must be due to the compressor inlet restrictors mandated by the rules. The restrictor turbo engines seem to have that kind of power curves.

[ptuomov]

This is a sensible article that in my opinion captures the lag issue well:

https://www.caranddriver.com/news/a1534 ... -the-test/

“The lesson? Turbocharged engines may claim fat torque curves starting from low rpm, but they’re significantly more responsive if you drive them as you would a naturally aspirated engine: Downshift, then accelerate.“

.

88686004-7629-48D1-A548-84840A716635.jpeg

9109680C-2ED2-4A53-96A7-109BB346F836.jpeg

CBB1F02F-1529-45EC-A5A9-76EE40188267.jpeg

[ptuomov]

I wonder how much “free” noise reduction one could get by combing the pulses between banks? Each bank fires its cylinders 240 crank degrees apart, with banks firing 120 degrees apart from each other. The pipe itself is filled pretty well with the typical 36mm diameter primaries and exhaust cam duration of 220-260 degrees depending on the lift point used to measure duration.

The noise reduction “Z-pipe” would go somewhere like here:

.

6123EE8E-3355-4C54-BD86-28B80EC3E3D7.jpeg

0C4702FF-32E3-4AA1-9D11-71809FE47194.jpeg

DB4219ED-3B16-48C9-92B3-C66ED7B6284A.jpeg

[hoffman900]

From a different thread:

It uses DI. The Audi R4 TFSI DTM engine makes > 570bhp from 6500-9000rpm, and 590bhp at peak. Boost limited to 3.5 bar and fuel flow limited to 90kg/hr. It is allowed a 100kg/hr fuel flow rate push to pass that boosts power to 644hp. They run these engines at AFR of 16:1 to 20:1.

Interesting. 20:1 AFR would be difficult without stratified charge and/or some pretty serious flame speed enhancement. Do you know it doesn't use turbulent jet ignition? F1 use DI and TJI.

I get 21:1 AFR (2,016 kg/hr airflow) with pretty conservative numbers:
2 litre
3.5 bar MAP
8,000 rpm
110% VE
Intercooling to ambient + 30*C

perhaps they taper the boost from 6,500 - 9,000 to limit AFR.

It’s just DI, no TJI, and with a normal spark plug. Rules limited to 5000psi injector pressure but the design can go to 7500psi. They're doing this on 102 octane spec pump gas.

They run a fixed boost level with varying lambda. They found it is faster around a race track this way due to the ability to be back at full power after a shift more quickly than varying boost and a having fixed lambda (like most people do). This was worth 0.17” at Hockenheim. They have graphs showing power % / boost / etc (with the x-axis being time ) showing the improvements during / after a shift. Over this 1" window, average power was 40hp more with constant boost / varying lamda vs. varying boost / constant lambda. They realized more time from driver confidence / driveability.

Static compression: 11.4:1

They also have graphs showing lambda, ignition delay, coefficient of variation of IMEP, and crank angle duration of 5-90% mass fraction burned for 6500-8500rpm. Ignition delay is 0.31-0.34 ms (ignition to 5% MFB), 5-90% MFB has a rising trend from 30-34* CA, and IMEPconv / % is lowest at 1.20 and over 1.50 at peak. This issue was worth it: https://www.highpowermedia.com/Product/ ... -issue-136

That issue fo RET is a good read.

[ptuomov]

I love the way the fuel flow rules limited engines, like the Audi DTM engine and Formula 1 engines go with constant boost or near constant boost, and varying lamba. It also makings for a faster engine on the race track / dynamically as well.

For air cooled Porsches, it’s the heat shedding capacity that limits the power. I think the 993 Turbo runs at 10.5 AFR under sustained full load from the factory. When anything gets heat soaked above a limit, it turns off the power.