Dyno result for the risky business Porsche

[englertracing]

sheet metal intake manifold
or call the fellas that make the itb kit, and ask them how much cheaper for one without throttle shafts, that you can bolt a pair of plenums on top of

Why? I’d much rather have the turbos spool up at low rpms with the stock intake manifold and then ramp up the boost at high rpms than have the turbos not spool up at low rpms with a short runner intake. And there’s an additional thermal bonus at high rpms when running pump gas.

Here’s how I believe the ramping up of boost at high rpms is going to work in my favor on pump gas. My street car is well intercooled, so it will pull out a ton of heat from the charge, especially when the boost is ramped up at high rpms. Then, importantly, the cooled charge will lose about 25% of the absolute pressure at 8000 rpm because of the intake manifold Helmholtz resonator being out of tune. This will cool the gas that actually makes it in the cylinder, and that's what matter. Even though the compressor outputs horribly hot charge, the cylinder actually gets not too hot, less dense charge after Intercooling and anti-tuned intake. Then one can give the engine more ignition advance on pump gas (compared to engine with a manifold tuned to increase the charge density like in NA engine).

The gain comes from the ability to run very hot charge thru the intercooler and have the inter cooler take out more heat from the whole intake system. Porsche is using this strategy in the four cylinder turbo models now in 2020 also after having switched to this strategy in the 2008 Carrera Turbo.

See what I'm saying?

Only way this "backfires" if the exhaust pressure gets too high and messes up the gas exchange.

a sheet metal intake manifold can be any size and length you want it to be
can you extrude hone your existing manifold?
create a larger version of your existing manifold?
can you cut your existing manifold apart port it and weld it back togeather? - or acquire a second one for this purpose

[englertracing]

Looking at your intake manifold, it doesn't appear it would be all that hard to construct from mandrel bent aluminum tubing.
whats inside the plenums? simple radii, or pronounced bells? does it have any type of diffusion and balancing system in the plenums?

[ptuomov]

Some info on the stock 928 S4/GT/GTS intake manifold:

viewtopic.php?p=432489#p432489

The objective is to create an intake manifold that has monster torque normally aspirated in the 2000-3000 rpm range to spool the turbos but doesn’t choke off the engine too badly in the 6000-8000 rpm range so that it can be “force fed” without excessive exhaust back pressure. The LESS torque it makes normally aspirated in the 4000-6000 rpm range the better.

[ptuomov]

Here’s a short-runner ITB intake that works great producing top end power with a normally aspirated engine and high octane fuel:

B3D6DBCC-CB9A-4A0F-9447-212B7AB2066C.png

90134D57-E0E5-46CC-A988-1DCD2CACBE6A.jpeg

That’s John Gill’s car. Starting point was E85 fuel, Colt Stage 2 cams (largest that are still perfectly streetable with a plenum manifold), 11.5:1 static compression, headers, race exhaust. The only change was swapping out the stock intake manifold to AT Power 48mm individual throttle bodies. The result was about 30% pickup in peak power, which means about 100+ rear wheel horsepower in the 6000-7000 rpm range.

For a turbo street car with pump gas, I’m thinking that this sort of short runner intake is _not_ the way to go...

[ptuomov]

Still no video...

BD50E7C0-FF31-4824-A6F1-E92F93F6BE37.jpeg

It's kind of dumbfounding how accurate the simple formula is for 6000 rpm and under for this engine: the rear wheel torque in lbf-ft is absolute manifold pressure in psia times 19.7 (lbf-ft/psia). 33 psia gives that 650 lbf-ft that we observe above.

With this new piston dish shape that reduces compression a bit and promotes tumble, there’s no knocks in the mid range rpms with 33 psia / 18.3 psig boost running the engine on 93 octane E10 pump gas from a random gas station. That bodes well for us being able to run the engine efficiently instead of in “party-mode” and still meet our goals.

[englertracing]

Its like your using the the dreaded gm tpi manifold...
What if you went long runner "cross ram" style and ditched the 180 manifold, for simplicity and straight runner sake, could get plenty of length and flow

[englertracing]

How about boxing something like this

unnamed (3).jpg

[ptuomov]

How about boxing something like thisunnamed (3).jpg

That would be great for a normally aspirated engine, but for a turbo street engine, it would generate the most torque in the mid range rpm, which is precisely where I don’t need it. I need monster torque at 2000-3000 rpm and then the torque not falling off too fast above 6000 rpm. Between 4000-6000 rpm, it can have a hole — the turbos can fill any hole there.

[ptuomov]

Here a fuller view of how the torque per absolute manifold pressure falls after 6000 rpm with this stock intake manifold:

.

CE0DE7A9-931E-4050-8D67-A30726A7F718.jpeg

.

Some observations:

First, Niklas @ JDS figured out how to get ancient EZ-K to run above 7100 rpm with a chip, a socket, and some "my dad's got a great set of TV repair tools from back in the day" soldering job. Now we’re at 8000 rpm and nowhere close to the new EZK limits. None of this would be possible without JDS.

Second, the base engine really wants to make peak torque below 6000 rpm and peak power slightly above 6000 rpm. This shows up in dyno graph produced with the flattish boost curve. It’s making that about 600 rwhp or better in the whole 6000-8000 rpm range but the power is not going up with rpm.

Third, my simple formula predicts that it should make 388 lbf-ft at 8000 rpm and 27.13 psia. And it makes almost exactly that torque in practice. The S4 intake manifold (mostly the culprit) results in the 625 lbf-ft torque at 6000 rpm dropping to about 390 lbf-ft at 8000 rpm. This means that we can and must ramp up the boost by about 10 psi over that range while hoping for the best on the exhaust back pressure.

Fourth, there is no oil out the breathers at all, cumulatively, from all the experiments with the new short block. This may not sound significant to most of you, but without effective countermeasures this engine model has oil drain and ejection problems above 5500 rpm for many engine builders. There are many “solutions” out there that are designed based on superstitions and not tested at high rpms. I’m happy to say that our breather solution works at least for now — this may change as we subject the car to high g-forces at 8000 rpm.

[ptuomov]

Inching up with the mapping process:

19026CE8-BB26-4F18-8BD9-7130A3195019.jpeg

4283A307-17F9-4FA0-AA54-D4B41AC0DCCB.jpeg

We are at a reasonably good point on the compressor map based on a quick back of the envelope calculation. It doesn’t seem like a terrible idea to move northeast on the compressor map with more engine rpms, that is, to ramp up the boost to compensate for the falling VE of the underlying engine.

I sometimes spend too much time trying to improve on simple back of the envelope calculations. This got me thinking about how we can locate the engine on the compressor map as precisely as possible:

First, mass air flow. We need to log the MAF signal and use that and the hpx-e specification table for the specific housing to get the first estimate of mass air flow. We can also get a backup measurement of the mass air flow from manifold absolute pressure, manifold absolute temperature, and assumed VE curve as the second estimate. We can use the measured and logged AFR and fuel pressure and injector pulse width computation to estimate the mass air flow as the third estimate. The fourth and crude backup measurement can be computed from the measured rear wheel power.

Second, the corrected pressure ratio. We need the corrected pressure ratio for the compressor to get the y-axis on the map. To get that, we need the compressor inlet pressure and compressor outlet pressure and temperature. The inlet temperature is likely very close to ambient. These will get us the corrected pressure ratio with some computations.

Third, we can simply log the turbocharger shaft speed.

Fourth, and this is something that people don't often think about, we can use the observed pressures and temperatures and the mass flow to compute the compressor efficiency. We can get a fourth way to locate the engine on the map from this measured compressor efficiency. Because of the smallest compressor cover that we are using, the map shifts by about 10% to the left.

I don't think any of this sophistry will change the conclusion. It’s just be fun to compute all this as precisely as possible, even if it doesn't ultimately make much of a difference.

Still no video.

[ptuomov]

A low boost teaser video:

https://www.youtube.com/watch?v=JEuDrTd ... e=youtu.be

[ptuomov]

The A/C compressor, clutch, and the belt are having a hard time keeping up with 8000 rpm...

[MadBill]

The Gordon Murray T.50 revs to 12,100 RPM. He had to go with a 48 v. electric AC compressor...

[ptuomov]

The Gordon Murray T.50 revs to 12,100 RPM. He had to go with a 48 v. electric AC compressor...

This engine is almost stock and we want to keep everything that we can as stock as possible, so trying to figure out pulleys etc.

Latest video:

https://youtu.be/UaZRwmSDz78

And this is where we are now, still needs MOARRR rpm:

https://www.instagram.com/p/CEL7A4NA43s ... s6rt4xpqlc

[ptuomov]

The 8000 rpm Porsche 928 S4 engine in the car uses this windage tray/scraper/baffle system to control oil: http://www.crank-scrapers.com/Porsche_928_pictures.html. It's made by Kevin Johnson of Ishihara-Johnson Crank Scrapers.

Now, some of you may know that I've had various heated debates with Kevin Johnson over the years on various engine and oiling related topics. Before getting into details about my opinions about scraper systems, let me just say that his system currently installed in my car works extremely well, so congratulations to Kevin for the design. It's not puking oil out of the breathers in the dyno at 8000 rpm, which is a common problem for other 928 S4 engines even at lower rpms. Although I haven't driven the car in this very latest configuration yet, John has driven it pretty hard, we've also observed no problems on the road.

If we were just running this engine on a stationary dyno, the only pieces that would be required for the engine to work really well are the oil-pan spacer, possibly the early 928 sump mesh cover or something analogous/functionally equivalent, and some drain-blocking pieces from Kevin's kit. Of course, we are not just running this engine on a stationary dyno. Instead, the engine is in a car and will be subject to all sorts of accelerations. The rest of the kit is helpful in keeping the oil from escaping the deep part of the sump, uncovering the pickup, and flowing into the rotating assembly.

Two concerns have been voiced about adding such general windage tray/scraper/baffling devices in the 928 crankcase by the Porsche 928 crowd.

The first is whether these devices directly hinder oil drain. This is (per my impression) the more popular concern among certain 928 enthusiasts. This concern is in my opinion mostly misguided. The oil is going to drain very quickly thru even small openings on a vibrating engine, as long as crankcase gas flows don't prevent this oil return. Kevin’s kit in particular is thoughtful about oil drain.

The second concern is whether the scraper/windage tray devices hinder crankcase air flows between the bays. There's the generic concern that not enough gas flow area between crankcase bays will increase pumping losses and reduce power. This is not a concern to me, because we've got enough power. However, the lesser flow area between bays may force the piston pumping pulses to equalize between bays in an undesirable way. This may indirectly hinder oil return. This undesirable effect is what one should be on the lookout for when evaluating alternative windage tray/scraper/baffling systems for the 928.

We just ordered a couple more systems from Kevin to put on various engines, one of which will go to the sister engine of the current engine that will have the same 8000 rpm rotating assembly.