dr bob

Dry sump tanks include the oil reservoir section, but more important they have the vapor separator section above the liquid level. Two gallons of oil as liquid, probably four or five by volume including entrained air. Not an expert on 928 dry sumps by any stretch, but I think at least five gallons of total tank and separator volume would be required.

Lizard928

Figure at a minimum, you want 150% of the original capacity.
So 8 Litres from the factory, you want 12 now MIN.
I would plan for around 4 though, less if you get a pump with an air oil separator.

GregBBRD

[QUOTE=Fabio421;8528356]Very nice looking workmanship.

So that oil pump is just "dry sumping" the heads? Not the whole engine? Is this due to the space constraints ( for a large oil tank ) in a street car or something else? Pretty difficult to get a tank large enough (in a street car) to both handle the oil volume and also have enough space to allow the "air" bubbles to settle out. That would require a tank of at least 3 gallons plus another gallon or two of volume to have some "air space. A space big enough for a tank of that size is tough, in an already "busy" vehicle.

Is the "outlet" of the provent plumbed back into the intake of the engine? The reason I ask is that you mentioned the crankcase evac air degrading the fresh air / fuel charge and lowering the detonation threshold. Does the provent seperate the oil out of the air so well that it can be reingested without this worry? That is what the Provent is made to do. It is basically an "age old" oil separator, with an added filter. I know that Dave (928 Specialists) worked with the Provent and made a system up to remove the oil from the breathers and thus return the air (without the oil) to the intake system. This apparently was successful (I have not purchased one or tested one) and is probably all totally "street" car really needs. I knew (from the dyno) that we were still seeing oil ejection (mist with some small drips) from the breathers at higher rpms (6500+). I made another design change to the breathers and added two Provents to Andy's engine....I didn't want to allow any oil into the intake system, after all of this work.[/QUOTE]

Notes above, in blue

GregBBRD

Yes, Edward's pulley is actually black anodized. The "pump" pictures are actually from Andy's "prototype" pump and pulley....which we had made and used instantly. We had a bit more "time" and had Edward's pulley anodized black. It looks amazingly "stock".

We made some design changes to Edward's pump...and we really had no reason to show the entire world every little detail...the people that love to copy (you know who) are going to have to do more than just a little bit of thinking to get where we are.

Since we discovered that "oil retention" in the heads was a major problem, with these engines, there are many people working on their own solutions. I'm sure that there will be many different attempts to create systems that address this problem. This solution is a bit complex, but it satisfied all of our "design" requirements (see Rob's list of requirements on Post #1)...which were pretty ambitious, considering the "history of these engines", when used in high performance applications. I feel very good that someone as simple as me, could solve a major design flaw that has been around since the creation of these engines.

GregBBRD

Someone else will need to chime in, regarding this. I'm kept "captive" in my own little world of 928's and don't pay much attention to other vehicles. I have no idea what makes any BMW work. I'd need the owner's manual to open the hood.

BC

I heard some newer MB and BMWs don't even have a dip stick.

unless you count the one driving.

Kevin Johnson

The engineers knew that the oil in the 928 was heavily aerated when the 928 was introduced in 1977. This is proven beyond a doubt by the proposed alterations to the engine found in the engine cutaway picture in the service manual.

When you restrict an oil passage in a dynamic circuit it lowers the oil pressure. In a supersatured solution air bubbles will evolve out. You certainly do not want this to occur prior to the oil supply entering hydraulic lifters which depend on oil behaving as an incompressible ideal fluid rather than a compressible mixture of free air bubbles and oil. Ergo they left out the restrictors.

The reason the oil "packed" into the heads was because it foamed upon exiting the cam bearings. Foamed oil flows more slowly and would return more slowly to the sump. Observation of Louie's video demonstrates this directly.

I think contemplation of the above will answer the issues 2, 3 and 4 below. The 968 has a superior scraper installed on the floor of the sump and baffle ergo better control of aeration.

c_span

Hehe fair enough. Serves me right for trying to say something intelligent (emphasis on trying!).

edit: actually, trying to find a bit more relevant info about the BMW approach reveals this. Anybody know the 928 owner from Boston?

GregBBRD

I might feel the same way as you, had I not seen the oil being ejected on the dyno. This "ejection" was coming out as solid slugs of very dense oil...not anything like an mixture of air and oil. Since the "foam" should rise to the very top (and the breathers are on the top), I did expect to see a very aerated product, if we were to see anything, and was somewhat surprised to see these big solid slugs of solid oil being ejected.

Ater studying Louie's video, at length, I believe that the video shows way more than was correctly interpreted. I think that early on, when you see pretty much solid oil being whipped around by the cams, the pick-up was covered and things were adequate. I think the "big foamy" parts are where the engine was sucking up air, at the pick-up, and trying to send that everywhere. It would arrive at the heads "very foamed" and we saw "foam" until the pick-up got covered and then pumped solid oil again.

This really makes sense, when you read on and understand that the engine, in the video, was "scrap metal" very soon after this test.

I'm trying to keep the heads from filling up with oil, in the first place, thus keeping the pick-up covered. I'm not sure you have followed this entire "dyno testing on Andy's engine", from the beginning, but on another thread, you can clearly see the oil pressure drop on Andy's engine, when the pick-up sucked air.

Kevin Johnson

You were pondering before why Porsche did not put the restrictor in the 928. I explained why it was not put in the stock engine. It was in fact suggested for the late race engines. There is a letter floating around from Porsche explicitly stating this. The 944S restrictor valve limits oil pressure in the head to 3 bars.

It was established 5 or 6 years ago now that oil ejection and aeration in the 928 engine are two distinct issues. Sigh.

We simply disagree on the interpretation. My interpretation allows for the aerated oil seen splashed up the vents time to drain down and make its way to the pickup and be drawn in. Yes, that is drawing in entrained air but oil as the outerphase as well. My interpretation allows for the massive oil pump in the 928 to try to compress that mixture -- by design -- and place that air into solution at the rate of 9% per bar pressure. My interpretation explains the delay in the appearance of foam from the cam bearings while yours does not. A superior hypothesis accounts for all available data. It's that simple.

If the oil was being ejected in large quantities then it is hardly surprising that the cyclic rate of the pump would soon find a sump level too low. Correct?

I am not saying you did not address the problem but what I am seeing is a drifting back to earlier views about the oil issues. Aeration and ejection -- association is not causation.

BC

Most of the aeration is already solved in the sump with push down of the sump and separation of the crank from the sump.

That would account for the solid oil going out the breathers- as the aeration is not happening. So it is a separate problem and Greg has solved this symptom- which is oil in the intake, which causes detonation and engines to go kerplunk. An entirely separate issue could be investigated now on why the crank case is pushing pressure. In a closed system
With no input it would
Not have pressure.

928mac

Guys I thinK the product by the OP is great and for the modified engines may be a requirment.
Personally I think that this is a solution to the problem, but
If the oil dose not drain down,
then why?

I think it is due to blow by pressure travaling to the top of the engine because that is where the venting happens.

Is this correct?

On a normal single cam V block this is not as bad of an issue because the cam valley is open for dranige but it can still be an issue.


So on the 928 OHC engine, If you remove that pressure at the crankcase and draw air in through the top the problem is solved. Correct?

I realize that it may not be that easy to evacuate case presure without getting oil but I think it should be possible to design into the original oil fill, for at least the more stock engines.

I really enjoy all your guys experiance and by no way am I suggesting that I am right.
just my 2 cents.

john gill

I hope this thread continues the robust technical discussion and doesnt degenerate into a personel clash .
Kevin I have seen this oil ejection from my own projsect in my own dyno, it is not aerated oil, in the heads it becomes a solid stream of oil from the filling the provent from the rocked cover vents until it overflows onto the floor.
If you place clear tube in the vent line you can see the stream of oil.
My original plan was to provide the sump with an engine vacuum source for this provent, from either a pump or exhaust venturi , but quickly despatched the idea after the volume of oil was so massive it would overwhelm anything in the vacuum line and probably either plugg it up or have it flow out the exhaust, either way the result is catastrophic
I used the 4 vents in the heads to the provent, once the engine reached 6000 rpm the oil output from these became scary, so any mapping beyond this was a waste of time, for all you with a stock 32 valve reving beyond this regularly, if you saw what happened in my dyno cell you wouldnt want repeat it .
Kevin this particular engine has one of your scapers installed, whilst I am sure it has helped it does not fix the inherent design flaw this engine posesses . In addition to this I installed the pressure limiting valves from Porsche that you mentioned , which again suggests to me that the issue is not pressure related, it is flow and the inability to return the oil to the sump.
If you perform a basic volume calculation you can see that both heads can contain more volume than the sump does if the oil does not return .
My next project engine will again contain your scrapers and the valves, but it also have the head scavenge system installed to complete the loop.
I beleive that the best system is to dry sump the engine , because this system will scavenge all areas of the engine I know that it is best , but with all the time that I have spent trying to fit find a place for the tank (I believe it needs to be 20 litres) I cannot locate a place and still maintain the stock look, which is important maintainance of the value of the GT .
My theory will be well tested when this system goes into the club race car .

blown 87

Same thing as Porsche, money.

jorj7

That's been my solution so far, multiple crank case vents to relieve pressure to
help the oil drain back. Of course I haven't maintained over 6,000 rpm for long
distances yet. I've been in the 5500-5700 rpm range, which is the 190-200
mph range. If I get to the 210-215 mph range, then I'll be in the same boat as
Greg encountered on the dyno and will have to readdress this issue.