yet another great thread

I do wonder how much oil actually pumped inside the engine at any given moment.....some say 2 quarts, others say 3....but since the pump volume is directly correlated to RPM.....if 2 qts are in the engine at say 2000rpm.....does it go up with pump flow....or is it steady at 2 qts, but just moves through the engine faster due to the higher pressure and flow?

I love the picture above......now picture the same test with the mesh screen from the OB pan installed....sure it will eventually end up exactly the same...BUT the benefit of the mesh screen is it takes LONGER to climb the same distance....that little bit of time might be the difference between the pump sucking air and damaging bearings!!!!

 

And then there are differences in fill volumes that people report. I should keep better notes but I recall using 9.5 quarts to get to the full line after an engine rebuild and 8.5 for an oil change with new filter. 1 quart stuck in the cooler and passages? Sounds about right. Then what's flowing around and in passages that'll drain?

I figure there are 9.5 quarts in there and 2 or 2.5 in use. That leaves 7 quarts in the pan. Could be 6 but I figure that's a minimum. It's a lot to both cover the pick-up and to slosh out.

 

I cant find the video of Kent Jordan and I around T-hill and thought even Ryans S2000 information was around from Thunderhill. that is a sustainted 5-6 second turn at the max g loading of the track.

the "ficker" of the g meter for Carlos' 911 is only for comparison. I would run around this car, at this pace through this section of the track. again, when you are 4-5 seconds a lap faster, and its not all about power you are going to pull more gs. its pretty common for cars running in the time range we are, to be in the 1.5 g range.

Ill have to print out the telemetry of a car that I was tailing around the entire track, and show the 1.5 gs around turn 2 at Thunderhill. anyway, even if it was 1.3, it a heck of a lot more than 1 (for the best "production" cars), and it shows that I have NO oil pick up problems or pressure drops. so, over 6 seconds if there is no pressure drops, the pickup much still have access to oil.

 

I posted the oil volume in the oil cooler and lines. It was less than 12 oz.

you put a quart more in that picture of oil level, or even 2 more, plus get a more realistic configuration with the upper block above it, and I bet all things change for the better. obviosly, brian just showed some pretty high gs and no oil pressure drop, as do I no matter the track. well, in extreme cases, only a little less than one BAR drop . However, too low rprm can cause a pretty substantial drop of pressure, so there is a concern there.

 

I think the oil pump is specifically designed to deal with aerated oil. If you add more oil into an engine that is already churning the oil it is a dangerous game. You might buy additional time because the total volume to be aerated increases and this might allow the air release rate for the oil to establish a safe equilibrium with the ability of the pump to develop pressure. Higher rpms would probably push the overall rate of aeration so high that the pump cannot handle it (lower rpms would have a similar effect because of the communication slots joining the meshing teeth of the pump gears reducing the ability of the pump to generate 100% positive displacement under load). With additional oil in this situation you could expect the oil temperature to rise too (churning generates heat from friction).

 

The flickering numbers looked like noise to me. Lots of vibration and sudden bouncing in a race car. He's not breaking 1.1 or 1.2 with any consistency. Gimme the raw data!

Please do. 1g is on the best street tires. I'd expect 1.1 to 1.2 sustained for COT competition tires on a converted street car. This is in a long corner "carousel" situation. For a typical corner the g's are changing constantly magnitude and direction.

I'll repeat my theory from a few other posts: oil gets foamed on the longer straights and then sucked in on the corners. On the straights it rolls out the shallow sump and isn't flowing down the pan. Then, in the higher g corner enough liquid sloshes away from the pickup to let it gulp foam.

I'm not sure about the pump de-aerating the foam. That extra hole you've pointed to allows oil that was lubricating the pump shaft to drain back into the sump. The shaft is lubricated from the pressure side of the pump. As I understand it, the previous shaft lube was done from the low-pressure side with the oil drawn through the shaft and then sent out. The problem was that the seal could leak and allow air to be drawn in and then passed on into the engine. Probably manifested itself in a pump failure.

 

Sigh. I just spent an hour typing in a response to this and the website had a bug and the information was lost.

 

Heh, had that happen to me quite a few times. Since then I've developed a habit of selecting everything I typed (Ctrl-A) and placing it into exchange buffer (Ctrl-C) before hitting "send". This way you can just Ctrl-V it in case website bugs.

 

Glenn
You bring up an EXCELLENT point....on my lemons racer I installed the 3/8th pan spacer for two reaons:

1: Lower the oil pan to reduce windage
2: Increase total oil capacity, mostly for 2+ hour stints racing

With the spacer the oil level filled to the original full mark on the dipstick is about 11.5 qts.......I typically run no lower than 1/2 qt down on the dipstick....

So if the engine is continually cycling 2 qts through the engine...that leaves 9 qts in the pan in my car...or 7 in a stock car......

I'd like to see the same test Kevin did with 7 qts and 35' done with 9 qts....I would imagiine it would climb the wall higher, but should be closer to the pickup too.....then when you factor the plastic insert on the OB pan....the far side where it climbs the wall will ALWAYS be covered in oil.....sure the far side won't......but its better than the S4 setup that will have nothing but air

 

That's more elegant than me copying it to a text editor, which I have done. It went well for a while and lulled me into a sense of complacency.

 

Hi Glen,

The difference in pump shaft diameter to bore diameter yields ~ 1.25mm^2

The drilling sizes appear to be 4mm in diameter which yields ~12.56mm^2

This would present the bearing clearance as the metered orifice (~.04mm or .0015").

The location of the drilling on the pressure side is where gases would normally be centrifugally pushed to by the rotation of the pump. This was already well known behavior and employed in deaerator mechanisms prior to the design of the 928 (I presented the patent citations earlier on this).

It is possible that an engineer at Porsche accidentally designed it this way -- I don't believe that.

I think the pump that I pulled was original to the engine. When I prised out the shaft seal I noticed that is was coated with an adhesive. Perhaps some rebuilt pumps lack this? Or pumps that have generic type seals installed?

I can envision other possible drilling variations with the same casting that would set up oiling from the low pressure side of the pump. (For example, shortened drilling of the angle drain area in the casting and an intersecting vertical drilling from the marginal area of the annular, beveled ring at the internal terminal end of the shaft bearing.)

But this is just speculation given that I have not examined a part meeting this construction.

 

Glen,

Yes, there is some noise, but in any standard racing turn,you get some peaks and valleys, except for what you say, is a constant radius turn that hangs it out for a full 180 degrees. So, keep in mind, my car is 5 seconds faster a lap than this car and would be pulling more gs. So, whether its 1.2 or 1.4gs, it doesnt matter, im not seeing ANY issues with oil pressure and the proof is in the pudding. the better question, is WHY have I had sucess running 110% of the fastest cup car in the world type times, in a 928 with no mods and have done so for over 10 years with several different engines. (no failures, or even hints of failures?) Brian, and all his recent experience is still 10 seconds slower a lap, so its not the G loading factor, because if it was, you would think that I would have a hint of an issue. I can run brian's times without brakes (seriously) and in one gear. (not because of my driving, but because the car is capable enough with its power and grip)

Now, as far as G loading, it is pretty easy to see the carocels I run, plug in the time and speed (or distance) and figure out the G loading. you cant run the times I run with a car that doesnt pull 1.4gs on most turns and decels.

-M

 

Glen, I just found the motec output from Kent jordans evo when he and I were running back in 2005 (may) at thunderhill we were neck and neck. he beat me as he got pole and was in the lead all race, with me trying to find a way around him for 30min. the times were slightly slower than we run now, (around 3 seconds) but the track is faster and both of us have a lot more hp. the times around the famed, turn 2, was 76mph and it is a full carocel 180degree turn from a previous top speed straigth of 114mph the main straight before that of 131mph. His Motec prints out 1.4g's for that turn and the duration of the turn is 10 seconds. I dont need motec for my car, because I was on his tail for every step of the way. the only thing he had on me, was a slight pull down the main straight. I also have the video for this race and this motec output lap "4".

Anyway, Im sure I could get the information from sears point, the only other track that has a true carocel for more than 6 seconds, but you will see the same things. the key thing here, is that my rpm is near 4500, not near 6k as would might kill the engine and the lap time.

 

1.4? I still think that's high but OK.

Something we've discussed before, and here agian, is driving style. Your style is to use lower RPMs in corners. This is surely part of your unique data point in engine longevity.

Kevin: So they could be blowing the foam out through the pump shaft. Dual purpose! Might be a coincidence of the design, but let's go with intention.

 

It appears that there may be enough surface on the bevel outside the shaft bearing (towards the pulley) to drill and then tap it for some sort of carb jet. A plan might be to increase the exit area by 50% -- it would be like adding a smaller second bearing journal.

The failure mode seems to be when the percentage and volume of entrained air overwhelms the ultimately fixed capacity of the pump to separate out air and leave mostly oil.

With MK pulling 1.4 G he is really whipping the oil full of air. Look at the differential between coolant and oil temperatures. The oil temperature is not rising because it is transporting rejected heat but rather because its presence is generating heat through friction.

I think MK's driving style is (a guess here) generating an understeer attitude to the car. This would shift the bulk of the oil forward into bay one where its presence would virtually guarantee an endless frothy bath of oil over the pickup (a general circular pattern with the rotating assembly). This hypothesis addresses his other data of seeing the pressure drop when passing through turns at lower than his normal rpms. The design of the pump is to communicate on the pressure side (and on the suction side) in order to force as much air out as possible (and hopefully towards a deaerator vent). It is sort of like slippage in a torque converter to me (I don't know if the physics coincides but the gross behavior strikes me as similar). In turns with low rpms his oil pump hasn't locked up yet -- with locking up being providing a sufficiently "solid" flow of oil.

This could be a synergy between Mark and the general design of the 928 with its front to rear balance. It would be interesting if his driving style would work in a car with decidedly unconquerable oversteer in turns -- that forces the bulk of the oil to the rear of the engine.

Back to work...