Mike Simard

I found that my dry sumped race car would fill up the left head with oil but only after a long straight, not cornering.
I suspect that the drains on the left side are a problem , they exit right at the crankshaft center and are exposed to windage unlike the right side where they are shielded.
My fix was to install some drain extenders on the left side to protect the drain exit from windage.

About aeration, the 2 ways to get air out of oil are:
1-A long path for oil to flow down like in AO's example.
2-Vacuum. Nothing is as good as vacuum at making bubbles disappear.

ptuomov

AO: I mostly agree with you.

I've kind of come a full circle around this. Currently, I think that the dumbest thing that one could do is to block off the filter neck and vent the air out from all four cam cover ports. In fact, who could be so dumb? ...well, two years ago, I was planning to do just that by adding a low-flow capacity air-oil separation device to the oil filler neck and venting out from all four cam cover ports! Myself today thinks myself two years ago was an idiot. What myself two years from now will think is an interesting question... ;-)

This spring, my breather system will be modified as follows:

- The cam cover ports will be be connected together per side with t-fittings. There will be two very low cracking pressure stra-val check valves making sure that air can only go in and not out from each cam cover and that the two cam covers don't communicate. The hoses coming from the two stra-val check valves will be joined again and the single hose will go to the pitot tube facing the flow in the low-pressure side of the turbo plumbing (that pitot tube was already installed by John). So from the pitot tube, the flow divorces to two cam covers, then goes thru a check valve and then distributes thru the two cam cover ports. In this arrangment, nothing comes out of the cam covers, the air may only go in.

- The magnesium filler neck will have a huge fitting welded to it. This fitting will connect to a hose that leads to the ProVent (already installed, but suboptimally used). From the ProVent, the separated oil will flow to the sump. The separate air will be sucked into the low-pressure side of the turbo, with a slash cut tube that (already installed in there by John). In addition, the filler neck will continue to provide air to the brake booster venturi pump under deceleration. (Louis Ott's smog pump option 666 has been checked long time ago.)

The _theory_ of the operation is as follows. The pitot tube facing flow creates a small pressure. At low rpms and head pressure, air will flow into the cam covers. At high rpms, the cam covers will have a higher pressure than what the pitot tube produces, and the check valve will stop the flow from reversing. The check valve makes head pressure a friend, not an enemy! The slash cut tube in the compressor intake will provide a small vacuum. This small vacuum will suck crankcase fumes out from the ProVent which in turn will suck them out of the crankcase (and not the heads) even at low rpms. As the crankcase pressure increases with rpms, the flow increases as well and the slash cut tube has only a minor part of the pressure differential the main force being the pressure int he crankcase.

Let's see if this works or if the future me will laugh at this post year from now!

AO

Mike,

You're killing me here. Right and left? When you say left, do you mean passenger side? That would make sense because of the rotation of the crank would fling oil up against the drain passages.

AO

It's kinda of funny, maybe ironic, or possibly moronic, you pick. The Twin-Screw 85 Euro I picked up recently (granted it's a 16v, but the principle is the same) has the oil filler completely blocked off and has 4 vents, 2 from each cam housing, going through dual air/oil separators then through a scavenge system. Umm... long story short... it doesn't work. Oil is fricken everywhere. SO I can confirm that is the WRONG way to do it.

My current system is almost like you describe except I do not have check valves at the cam vents. Obviously I don't have the turbo either, so that's different too. But I think we are roughly on the same page. But mine is brute force trial and error. You seem to put a little more forethought into it.

Edit: I have thought about plumbing the Provent's air discharge to the intake to create a little vacuum, but am afraid of oil mist causing detonation. Maybe I shouldn't be scared.

ptuomov

I would hypothesize that you had already solved many problems with your dry sump at that point by
- scavenging oil from the crankcase so it doesn't surge and hit the crank and be thrown up the right side oil drains
- scavenging gas from the crankcase so the pressure differential between the heads and the crankcase isn't enough to support oil plugs in the drain channels.

At that point, you had solved all the problems except the crankshaft hitting oil back up the exposed drain channels on the left hand side (or creating some weird pressure zone).

Does what I write above make sense to you? Or am I on the wrong track?

I would add couple:
Kinetic energy separator (unfortunately in use in the stock girdle oil passages!!!) and centrifugal separator. In any setting where oil and air flow fast in one direction, oil is much more reluctant to make turns than less dense air. This I think is the operating principle behind Thien baffles, hydrocylones etc.

On thing that might be cool to make would be a very thin Thien baffle that srews between the oil filler neck and the cap. Haven't figured out any plumbing or anything else yet, just thinking out loud.

ptuomov

Very kind of you to call my procrastination "forethought"...

Turbo and supercharger are basically the same on the cold side, so I think we're in the exact same situation.

You also mentioned routing the separated air from ProVent back to the intake before the compressor. I think you can test this by directing the ProVent separated air outlet to another catch can and see how much oil remains in the separated air. If very little, hook away to the intake. If a lot, then "Go Back To Start, Do Not Pass Go, Do Not Collect £200".

Mike Simard

Established law of where things are has the left of the car=driver side in US. That's also why they call that "left hand drive" and British and others are "right hand drive" cars. It's chisled in a stone tablet somewhere.
I would use 'starboard' and 'port' but I don't know what those mean.

Mike Simard

You're right, there are more than 2 ways.

This is getting impractical but just for fun here's yet another:

Ultrtasonic vibrating. I've seen this used in epoxy mixing which is an area that gets alot of effort put into de-aeration.
In the carbon fiber parts I make, it's really tricky to get the air bubbles out that are inevitably there after mixing. I've seen a trick of placing an ultrasonic ram into the mixture and the air bubbles dissapear. I haven't had luck with that because of the special epoxy I use so my process is to place the mixture in a vacuum chamber before the mold goes together. Maybe a thinner mixture is able to have air vibrated out.

The idea of shaking air out is used passivly in motorsports such as F1. The oil tanks are mounted to the engine, not the chassis.

FredR

I think Louie's work on this problem is reasonably definitive for wet sumped stock motors. Combination of high rpm's and left hand bends causes oil to stay in the camshaft gallery all frothed up. Whether other factors can also cause similar symptoms remains to be seen.

I find Mike's comment that on his dry sump motor he saw oil in the left camshaft gallery even more intriguing in that there is no obviously logical reason for this [that I can think of].

I have been working my motor quite hard recently whilst sharktuning and have some high rpm issues with pinging and in the process I have been going through oil at what seems to be a very high [alarming?] rate. I have no obvious signs of distress and this motor [I have owned it for 14 years in my two 928's] has always had a tendency to high oil consumption when "worked". I am linking the knock to oil passing through the rings due to high crankcase pressure despite efforts to vent it.

I have carried out a number of experiments to see if anything modulates the problem and found nothing obvious to date. I eliminated recycled oil vapours but that made no difference. The most interesting point I noted was that I vented the rear passenger side vent direct to atmosphere expecting to see some oil blowout but there were no signs of oily residue in the pipe or splatter in the general area.

For the oil filler neck vent [GTS style] I took it to a temporary catch can [a plastic milk bottle] by the side of the radiator and found nothing in that.

I am wondering if the baffle plate in the base of the filler neck is in fact resticting vapour flow coming from the crankcase thus causing an overall increase in crank case pressure. I am hoping to try a test to measure crankcase pressure this weekend utilising a temporary connection to the oil filler pipe.

That cylinders 6 and 2 are the predominant cylinders for onset of knock I find interesting considering the 2/6 journal is the one prone to failure - whether there is a cross link is another matter.

There seems to be little doubt that reducing crank case pressure is the best way to go if the pump is long term reliable for street use and Colin's approach I find interesting. Doubtless the breather system can be improved significantly as well but to get vapours to flow out of the crankcase requires a small positive pressure to establish a flow.

I also agree with Tuomo's observation that it does not take much of a dynamic to arrest oil flow descending from the passenger side cam bank especially if the bank is being educted into the air intake [slight vacuum?] or vented to atmosphere.

Another observation I made is on my balance pipe across the cam banks. I use a home brew silicon rubber pipe that is translucent and I have never seen a drop of oil in that pipe.

I rather suspect that BMW solved some of these problems with their E39 M5 motor design that pumps oil from the cam gallery [as Doctor Brown now does].

One has to wonder just how much Porsche were in control of their engine designs considering all the problems seen not to mention early GTS problems specifically.

I cannot eliminate a mechanical wear problem somewhere at the moment but for sure I would not like to force feed my motor unless and until I can solve these issues.

And we like these cars?

Regards

Fred

ptuomov

FredR --

If the crankcase pressure and the head pressure have equated, it's unlikely that oil will eject from the heads in an engine placed straight up on a dyno.

I think (but do not know) that the most certain way to eject oil from the heads is to allow the pressure differential between the crankcase and the heads to exist and grow. My current view is that anything that lowers the pressure in the heads and/or increases it in the crankcase is counterproductive. As an extreme example of a counterproductive setup, AO described above buying a car that had the oil filler neck plated off and all four valve cover ports vented out. To quote AO "Oil is fricken everywhere. SO I can confirm that is the WRONG way to do it." Only procrastination prevented me from replicating this mistake, when remembering (European) high-school level physics should have been sufficient. ;-)

If you don't see any oil ejected from any of the cam cover ports, then the logical conclusion from that is that in your engine the crankcase doesn't pressurize to a higher pressure than the heads. There could be two reasons for this. First, either the crankcase pressure is lower than in problem cases. Second, or the head pressure is higher than in problem cases.

My bet is on the second explanation. One specific potential cause for your symptoms (and lack thereof) you should rule out is worn valve guides. If the intake valve guides and/or seals are worn, then the engine will consume some oil as oil gets sucked in thru valve guides. This would explain both the detonation as oil lowers fuel octane and oil consumption even if the breather lines or the intake manifold don't show any oil. If the exhaust valve guides are worn, then the high-pressure exhaust gas will leak to the valve cover thru the exhaust valve guide, pressurizing the valve covers. As a positive byproduct of the worn exhaust valve guides, the pressurized valve cover prevents oil from packing into the heads.

So my guess would be that if you replace your exhaust valve guides, you can have your motor injecting oil into the intake like many the other GTS engines in no time! ;-)

Best, Tuomo

Strosek Ultra

Here is some thoughtful information of how BMW solved the problem with the "oil plugs" for their V10 M5 engine. Note the similarity in design of the BMW V10 and the 928 engines.

http://autospeed.com/cms/title_BMWs-...4/article.html

Four oil pumps provide the engine with lubricating oil. During extreme cornering, centrifugal forces push the engine oil to the cylinder bank facing the outside of the bend, thereby preventing the natural return of oil from the cylinder head, which might lead to inadequate oil supply in the sump. Should the worst happen, the oil pump could suck air. To prevent this situation, the engine features a transverse force regulated oil supply system. This system incorporates two electrically-operated duo-centric pumps which pick up oil from the outer cylinder head and transport it to the main oil sump if lateral acceleration exceeds 0.6 g. A lateral-g sensor transmits signals to the pumps.

Ake

Lizard928

Mike,

It is interesting that you found the left (5-8 bank) to fill with oil...

My solution for the heads filling up was to add a vacuum pump. I then run it to a custom made air oil separator (didnt like the provent). The bottom catches the oil and then I have a small OD tube with goes to the 1-4 head currently. This then helps to limit the vacuum the pump can generate, as well as sucks the oil back into the engine, but naturally creates a flow of air from the head down. I suck the oil out of the filler neck (stock, with baffle removed). So this also helps to eliminate the oil cloud being too big as well. So far I have noticed my oil consumption drop substantially.

GregBBRD

For whatever it is worth, my steady state dyno testing, on Andy's engine, showed that "both" heads filled and ejected oil at virtually the same instant. Mark Anderson's dry sumped race engine would "suck" the oil tank down, at high rpms, more than the volume of one head....much closer to the volume of both heads.

Both the "race" system that I made for several engines (active removal pump and many other changes) and the passive Ventilation Kit that I made for the "street" vehicles (multi-goal engineering and results) have both solved any oil returning to the intake. And neither system has had any "issues" with the oil pick-up being uncovered. And we have seen no evidence of any aeration. People keep "trying" to engineer other solutions...but it all seems a bit redundant, to me.

Tough for me to figure out why people still are trying to "solve" this problem....but what the hell. I'm sure we need multiple products to solve the same problem, for the tiny amount of these vehicles that are out there.

All wet sump engine have aeration. One of the main benefits of most dry sump systems is that the volume of the oil is large enough and sits in the tank for long enough to allow the bubbles to "come out" of the oil. Aeration is the nature of the beast. Different oils aerate less and there have been many methods to remove the bubbles which have been attempted over the years of engine design.

However, there's a given rule that no one is ever going to escape: When the volume of the oil gets low enough in the pan or the dry sump tank....there are going to be bubbles sucked up....and the engine bearings are going to suffer.....regardless of what else one does.

The main idea, for me, was to get more oil to return, faster to the pan. Both systems have been shown to do that, without any issues, in their individual applications.

ptuomov

I understand that this is proprietary information about a customer's engine, so perhaps you can't / don't want to answer this, but I'll ask anyway: Did Andy's engine that exhibited this oil ejection behavior happen to have the oil filler neck shut close with a plate or somehow else modified such that crankcase gasses can't evacuate via that route? Just interested because this would be one data point to either support or contradict my theory about pressure differential.

AO

Tuomo,

You may have not seen this during your absentia. I don't recall seeing anything about modifying the oil filler area, but my short-term memory is crap these day. Now what was I doing....

https://rennlist.com/forums/928-foru...8-engines.html