ptuomov

In my opinion, the 928 S4 has a big problem of spraying too much oil to the pistons and bore walls. All efforts should be taken to reduce the amount of oil spraying all over the place in that crankcase. These include rings better suited to removing oil from the bore walls, increased pan volume with a spacer, pumped vacuum in the crankcase or at least a good breather system, baffling for the oil sump, limiting rpms, and various devices to manage the piston pumping pulses, etc. Personally, I believe that with a mildly-modified stock bottom end with its wet sump and all, you can't remove too much oil from the bore walls and the pistons.

At least with the stock S4 pistons and stock width rings running on a stock open-deck Alusil cylinder towers, the thermal inertia and heat flow capacity are very high. Piston cooling doesn't seem to be a high priority compared to dealing with all that excess oil everywhere. If you're concerned about cooling, a much higher priority item would be to figure out how to get more air flow thru the radiator. The stock bottom end doesn't seem to have any problems getting the heat into the coolant.

After thinking about is for a couple of years, I'm not surprised that the factory eliminated the squirters ftom S4 early during the first production year. Now, if you go to 350g pistons with razor-blade rings running steel sleeves, and four stage dry sump pulling 20 inches of vacuum in the crankcase, then maybe it's a different tradeoff. Maybe you do want to wire-EDM the rods for wrist-pin lubrication and to add piston oil squirters. At this point, I know of exactly one engine that even remotely resembles that.

bertram928

Agreed.
However in a lower RPM turbo application the benefits of cooling underside of piston crown and wrist pin would outweigh disadvantage of resulting excess windage from absence of oil control measures.

Roy928tt

Nah, You use piston squirters to cool pistons. The fug floating around inside a crankcase does nothing to cool pistons, best to get rid of it to ease the job of the oil control rings.

As noted by Tuomo the alusil block transfers heat to the coolant very well.

bertram928

It is my contention that by going to the above kind of non OEM piston and rings (and or steel liners) one looses a substantial amount of direct heat conduction to water jacket (thinner rings and less skirt area).
Compounded in case of steel liners the added resistance to heat conduction would likely raise piston temperatures to point of cooking oil in lands, resulting carbon build up from high top land temps would perhaps lower pre-ignition threshold in a high specific output turbo motor.
It would be most beneficial to employ piston squirters in that scenario to provide an additional heat path for cooling piston crown resulting in lower top land temps as well.
Goes without saying oil windage control needs to be optimally applied commensurate with intended rpm range as well.

Not sure about benefits of wire-EDM oil feed passage to wrist pin in case of piston squirter being employed.
Mike I trust executed this in expert fashion mitigating risk of introducing stress risers when passage runs to close to surface (challenging when dealing with rod beam limited cross sectional area).
The net cost vs. benefit may not make this the route to take.

bertram928

Yes agreed
was building a case for using piston squirters and instead controlling the resulting additional "fug"

AO

Interesting Knock Run data. I found the the 2/6 (but mainly 2) tends to knock more often than others especially at that 3500-4000 transition to the upper RPMs. Kind of a low level errant knock, and it appears yours does too. I've tried many things to get rid of that, but I think its either a false positive, or just something else in the overall design of the motor. But you certainly seem to have a bit too much advance from about 5k up.

I need to look at the new ST2 SW and see if it makes sense to tweak my GT... Now that I think about it, It's been way too long since I've driven that thing. Hmmm...

V2Rocket

consider that the 944 turbo was produced for 5 years and never had squirters even at 100hp/L (1988 turbo S and 1989-91 standard turbo).
as i recall the only piston ring set currently available for purchase is "944 turbo spec" even for the 928 engines so those rings must be pretty good at de-oiling a wall.

on the four-bangers they didn't have squirters until the 968 in 1992. i think that was due to the big piston being very thin in the middle section to keep weight down, so they needed to control piston temperature better.
even today most of the "big builds" approaching or even exceeding 200hp/L don't use squirters, just big oil coolers with stock radiators.

ptuomov

The high-rpm (6000+ rpm) knocks come mostly from the exhaust configuration and the pressure wave from the 90-degree offending cylinder. In a turbo engine with the (cycle average) intake port pressure lower than the exhaust port pressure and exhaust manifold (not long-tube headers), this is an important effect. In a belt-driven supercharger motor that has long-tube headers, this is not an important effect, because the pressure differential goes in the different direction and because the long-tube header separates the exhaust blowdown pulses.

Similar to a turbo engine, with a normally aspirated motor with stock exhaust manifolds, I'd expect to see the 1&6 knocking at high rpms. A normally aspirated engine with headers, I wouldn't expect to see that.

The mid range 4000-6000 rpm knocks that happen between the low rpms (where the 180-degree exhaust blowdown interference causes knocks) and the high rpms (where the 90-degree exhaust blowdown interference causes knocks) are due to some cylinders filling better than others due to the intake manifold design. It's the long runners that are straight that fill the best, and they also knock the most in the mid range rpms.

Reading too many books and looking at too much sensor data has caused me to believe that "things happen for a reason". ;-)

The knocks in that graph were induced intentionally and the knock retard left on to see what cylinders knock at what rpms. It will not knock in the tune we actually run.

PorKen

Put restrictors in those runners and blow up that boost, brah.

ptuomov

And then step out of the car and uninstall the restrictors as the engine accelerates to 6000 rpm, before shifting, and going back out and reinstalling them back?

PorKen

Just a touch to bring the flow rate of the straightest runners closer to the kurvy runners.

Or:

ptuomov

One could make more power either with cylinder specific ignition trim and/or with a fabricated intake manifold. Right now, it's a bit of an experiment how far the stock computers and the stock intake manifold can go. I'm just curious what's the limit for those, so we might see.

V2Rocket

cylinder-specific ignition trim, or fueling adjustment, would be better?
more filling would need more fueling, no?

ptuomov

As far as the high-rpm knock problem goes, it's not that 1&6 fill better with new air. They don't. It's that the hot exhaust gas doesn't get out of the cylinder because 3&5 blow into the exhaust ports during the valve overlap, and that exhaust gas heats up the new charge such that it knocks. So those cylinders don't really need more fuel, they just need a retarded ignition timing.

For the mid-range rpm knock that is not caused by exhaust but instead by some cylinders having better filling intake runners, fueling and ignition trims both would be useful.

In my case, however, that mid-range imbalance problem isn't nearly as consequential as the high-rpm imbalance problem, so cylinder-specific ignition trim would be a higher priority than cylinder-specific fuel trim.

PorKen

^^^^ that is a great summation.

Individual cylinder fuel trim would be nifty, but I'm not sure it would do more than be more efficient (at least N/A). Too bad, because I was just visualizing how to program a dual batch fire, using the vapor recovery solenoid output, to fuel some cylinders (EG. inner/outer) at a different rate.