ptuomov

Right now, the heads act exactly like that. The oil drain channels act as the hoses and the cavity under the valve covers acts as the canister. The pulses equate in that space, which can in some some engine configurations disturb the oil drainage.

If you could somehow run two 1" hose from the block valley end (bay 4) to the valve covers and two similar 1" hoses from the oil filler neck (bay 1) to the valve covers, I believe you would be done. That really is my preferred solution.

Alternatively, if you could connect the bay 4 to the oil filler neck directly with a manifold in the block valley, you might also be done. Finally, if someone figures out a way to connect the oil pan back with the oil pan front and/or the valve covers with a hose large enough hose to communicate the pulses, then one might also be done -- but the connection has to be above the oil level, because otherwise the hose might send oil shooting up in the wrong places. Your comment about "above oil level" is correct, I believe.

The above is speculation. The proof of the pudding is in the eating.

Alan

To me the in-valley connection between bay 1 - 4 has the added merit that its a very short path (about as short as it could be). These are very short duration pulses - to actually have mucch effect on reducing them in magnitude (vs just mitigating their downstream effects) it seems you need a path that is both large enough (cross section) and short enough to allow significant air volume movement.

Now making the in-valley path requires drilling the block (at least in bay 4), and figuring out large a large enough port to the filler base and clearance to whatever else is in the valley - so not really easy and not risk free. In stock TB config is there any hope to have room for this...?

Alan

ptuomov

The pressure pulses can travel up to the speed of sound. This of course would require an undesirably large pressure differential. But since gas is compressible, the short path is not IMO required, as long as the path has enough volume. As daveo90s4 said, one might as well hook up both ends to a canister (which by the way could be used as an oil-air separator, the container volume is the single most important factor in predicting the performance of an air oil separator). As long as one can fit large enough hoses, I think there are more than one way to skin the cat. The real trick is to somehow run large enough hoses from the rear bay 4, rest seems (in theory) doable.

With a stock intake manifold, it's probably easier to attempt the following than connecting the rear bay to front bay directly. First, move the rear knock sensor to the center fitting. Based on my background research, the exact knock sensor location is not critical. Second, enlarge the rear knock sensor hole above bay 4. Third, connect that hole above bay 4 to to both valve covers with a large flexible hose. By my eyeballing, it has some chance of fitting out of there. Valve covers need large new ports, the existing ports are too small. Fourth, connect the oil filler neck to the valve covers in the front with the same kind of large hose. I am not saying this would be easy, but it might be possible.

For someone building a wet-sump ITB motor from scratch, adding two aluminum manifolds in the valley on both sides of the knock sensors is probably a better idea. But that's unlikely to fit with the stock manifold, unless one figures out some creative intake manifold spacer arrangement.

None of this is convenient.

Then, for those of us who don't use long cams, there's Porken's suggestion. Use the (unused) rear cam bearing holes as breather ports. Connect the outer bearing holes to the front of the oil pan sump, above the oil level. This will at least allow the front corner oil drains to drain down oil with less disruption.

Strosek Ultra

For chain driven OHC engines there is a large area, where the cam chain is running, for good pressure equalization between the crankcase and the cylinder head. For belt driven OHC engines this area do not exist. When BMW 1979 introduced their first belt driven OHC six, they added a pressure equalization tube going from the lower part of the crankcase (above oil level) up via the intake manifold to the cylinder head. The inner diameter of the tube is appr. 1.125" for a 2,3 liter six and the tube can be seen in the middle of the engine on attached picture. The BMW engine is vented only from the valve cover not from the crankcase.

When Porsche designed the 928 they did not add any additional pressure equalization area but relied on the oil drains from the heads for pressure equalization between the crankcase and the heads which we all know was a mistake. What Tuomo is talking about adding large size hoses between crankcase and cylinder heads and also between bay one and bay four for good pressure equalization between crankcase and heads and between low and high pressure areas inside the crankcase is the way to go if we want the oil drain from the heads to work properly also at high rpm also for stroker engines and to reduce pumping losses for more power.

The question remains how to do this modification the best practical way. Me building an engine with an individual throttle body intake system from scratch, I cannot see any major obstacles but implementing it on an existing engine having the stock intake system need some creative thinking.

Åke

FredR

Starting to sound as though it is time to sell the 928 and buy a twin turbo! Well not really but how much ****-up should 928 owners be expected to tolerate?

Not that I think it is worth anything, but is this problem why Porsche tried the balance line between the banks on the GTS? If so it would be woefully small even if it was capable of working.

Regards

Fred

UpFixenDerPorsche

Very disappointing behaviour.

Here in Australia it's called "misleading advertising", for which there are severe penalties.



.

UpFixenDerPorsche

Firstly, thanks to the many contributors on this topic, as well as those of the 2/6 bearing failure and oil aeration issues.

I've spent the better part of a few nights ... zzzzz zzzz .... crawling through dozens of posts (and dogfights) and gained a great deal of insight into not only the original issues but also the issues faced in finding a good solution (to which I hope to possibly contribute over time).

But for the moment: oil return: a few thoughts:

Basic concept:- Fit an extension 'pipe' in the base of each return shaft to extend the return oil flow to below the level in the sump, effectively sealing the shafts from crankcase 'updraft' pulses.


IMHO this would stabilise the oil level (no sudden drop c/- high rpm accumulation in the heads), and reduce the amount of oil to be flung about by the crank (reducing aeration).

Ideally the 4 drainpipes on each side should combine into one (each side) and run forward to the bottom centre of the sump.

Could be done internally or externally.

(1): The internal option is preferred, but I'll have to leave it the wizards to see if there's enough space.

2 - If there's insufficient internal space, the shafts could be interconnected externally (tapping into the base of the shafts), with the single drain from each side leading back to the bottom of the sump.

Plugging the shafts:

(a) fix a bung into the base of each shaft from inside the crankcase
or
(b) slide a rectangular tube (snug fit) down the shafts from the cylinder head. Bottom of the tube closed off.

(c) Then drill/tap through the outer wall into the tubes inside and fit the large external drainpipe.

Use of a crankcase vacuum pump would help the flow in either case.

BTW: does anyone know the volume of oil pumped to the heads?

Your thoughts?

Cheers.

UpFixen.

.

atb

Wouldn't running a vaccuum pump like Colin does, pulling vacuum out of the chimney and pumping it back into the valve covers, accomplish the same thing? Seems like if you ported Bay 4 as discussed above, and ran a hose to "y" in with the chimney on the way to the vaccuum pump, you'd be good. Because it's pulling a vaccuum, you could use smaller diameter hose?

Bigfoot928

Adam,
I was thinking about adding this between the blanking plate I'm installing for the oil filler neck and making some custom fittings for the mount points where the V block plugs reside. Originally I had some fittings made to screw into these locations for vacuum/pressure readings.... I guess I could still do that...

ptuomov

In my opinion, I don't think one should extend the drain channels all the way to below the sump oil level, unless one makes it completely sure that there's an ample path for gas to flow from the crankcase to the head. If there's no such ample path, then it's possible that the crankcase will pressurize and the oil will be pushed up to the heads. If one can figure out how to create such alternative large breathing channel, one has probably already solved any possible oil-drain problems from the heads.

I believe that a vacuum pump can help the issue in two ways. First, by lowering the crankcase average pressure, the vacuum pump lowers the gas density. Lower density gas means that there is less energy in the piston pumping pulses. Second, if the breathing channels get plugged by oil, one can possibly alleviate some existing high-pressure zones by selectively scavenging some parts of the crankcase to encourage the plugged oil to flow the right way. For example, if one scavenges the crankcase proper and not the heads, that's likely to "nudge" the flow in the right direction when the oil channels are plugged with oil.

One idea that is perhaps under appreciated is picking one of the two center drain channels in each head and then blocking the oil flow (but not gas flow) to that channel. According to the simulation data that I have, the center channels really want to flow a lot of gas between the heads and the crankcase. One idea is to extend the center oil drain in the heads to above the expected oil level. This way, the center drain channel can't be filled with oil and instead it will always equate the pressure between the heads and the crankcase, allowing the corner oil drains (most important from the oil drain perspective) to flow oil down with less disruption. The 928 block oil drain channels are massive, so as long as the pressure is equalized they should drain all the oil in the world with just three channels.

namasgt

The chimney hole under the oil filler neck on the 928 block is pretty small compared to for example the 6 holes the coyote engine has in the valley. On the 928 You would think the chimney hole to the crankcase is the same size as the base of the oil filler neck but it actually tappers down to about 1.5"X1.5". Does anyone have picture of this opening or the measurements of it?
I would think that the opening is too small to cure the problem completely, given what other manufacturers have done. I think the vacuum pump Colin is using would help in this situation to suck the gases out of the crank case rather than letting the gases push through that hole. Thoughts?

daveo90s4

To state the obvious, the cam cover voids and the drain back holes are fulfilling two purposes, one being oil related (oil entering cam covers under pressure and draining to sump via drain back holes under gravity) and air related (acting as the canister and tubes referred to previously). So the drain back holes are at the same a) at high revs carrying high volumes of high speed pulsating air and b) responsible for allowing oil to passively return to the sump.

We know that the air flow trumps oil return flow. If it did not then the heads would not pack with oil at high revs and we would not be having these discussions.

Excusing, or largely excusing, the drain back holes from either function would seem the logical solution.

Two main options.
1. Extra oil evacuation from the cam covers, preferably by active scavenging back to the sump. This has been done successfully. Using the unused rear cam journal 'holes' looks promising, these being at the lower points.
2. Adding extra crankcase fore and aft airflow. It might well be problematic to do this internally by enlarging webbing holes. But externally, above the oil level, is probably doable.

Doing both of the above might well be the optimal solution.

But this is all conjecture. Greg has already solved this with his proprietary system, hasn't he? Or was that for naturally aspirated engines only? If so not sure why a solution for na engines would differ markedly from a solution for boosted engines ( same components doing same dual function).

Cheers

Dave

ptuomov

A couple of observations, based on somewhat applicable simulation results and conjecture:

- Each oil drainback flows and pulses gas differently. If the engine is vented from the cam covers, then at high rpms some of the drainbacks pulse with the direction of the gas flow changing and some of them just flow gas up. The driver and passenger sides also flow differently.

- We don't know that air flow trumps the oil flow. It's possible and in fact likely that the channels sometimes flow oil down and gas up simultaneously. I don't believe that oil plug is the normal situation in those drain channels, but that's just my belief.

- What (if anything) is done in terms of a windage tray and similar contraptions inside the crankcase has a surprisingly large impact on how the drain channels work. One thing that has been shown to work with the 928s is the drainback covers that Mike Simard has made. They are elegantly designed pieces, but require you to split the crankcase halves so they are really only an option as a part of a complete engine overhaul.

- Not all engines pack their heads with oil, even under high rpms. 5.0L engines with good ring seal seem to drain oil pretty well, at least based on our experience. Increasing displacement, especially by increasing stroke, makes the problem worse. That's one reason why GTS has so many worse problems.

- The active scavenging of the heads has two problems. First, if you scavenge actively from the heads, then you're hurting any natural drainback. That's a bit inelegant, but that's in the eye of the beholder. Second, if you really want to scavenge all four corners of the heads, you need four pump stages, which amounts to an expensive and physically large pump.

- It would be really interesting if someone would take a magnesium oil filler neck and valve covers, weld two ports to the neck and one to each cover, and then just connect them with two 1" hose sections. Leave everything else stock. That alone might solve a lot of problems, but we don't know before someone tries it.

Alan

We know more than that... on later model cars one of the breathers (of two primary but still undersized ones) evacuates out of the passenger side cam cover - so there is exit blow-by flow that must travel up the passenger side drains (and perhaps in the GTS with its head to head balance tube - some up the driver side drains too).

I think eliminating any net upward flow into the heads is helpful to drainback even if there is pressure equalization pulsation movement.

If there is a way to create more deliberate down flow from the heads to the crank (other than the normal camshaft oil flows) - that would seem to be helpful to drainback too.

Ways to do this:

1) In a vacuum pumped system evacuate from the crank vent and recirculate back to the cam covers - a 'virtuous cycle'.

2) Whenever blow-by flow rates allow - also flush clean air into the cam covers (from the post filter airbox bottom) via vacuum limiting valves.

That the vacuum pump & AOS loop will also help reduce entrainment, and somewhat limit pressure pulse flows anyway is a side benefit. Its a simpler to install solution than a multi-stage oil scavenging system for the heads - but still has some packaging challenges. Probably still too much effort for most to implement but for a GTS with major oil consumption issues there aren't any simple options.

Alan

Strosek Ultra

"One thing that has been shown to work with the 928s is the drainback covers that Mike Simard has made".

Tuomo, any pictures?

Åke