FredR

In post No42 I was expressing agreement with you that the concept you paraphrased is absurd- I trust you understand that. What I do not understand is which post did you see such written? For sure I have never written such jibberish.

When you use the term "leakage" I presume you refer the amount of oil that flows through the bearings as opposed to the amount of oil that flows through the pump- you may also conceivably use that term to define the amount of oil you understand is passing through the relief valve but I doubt that is the case.

Now, on your latest post we are getting to the heart of the matter. The very reason I asked GB if he had any analogue pressure data between 3k and 6k rpms was to prove the point that oil pressure remains more or less constant between those engine speeds and that is exactly what is demonstrated in the data GB presented give or take a few psi. For the relief valve to have opened the pressure has to exceed the set point of 8 barg and that did not happen. So yes- above 3k rpms the oil pressure is staying more or less constant and no oil is being recycled across the pump whatsoever.

This tells me that the bearing is acting as a pumping ring and in so doing is in effect reducing the system hydraulic resistance as the rpm's increase thus keeping the discharge pressure of the oil pump more or less constant above 3k rpms. Whether by design or coincidence I know not and for sure I do not have any specific knowledge of the bearing hydrodynamics but the proof of the pudding is in Greg's last post- it confirms what I had deduced about what is actually going on. If you have any friends who are into bearing design maybe they can confirm this.

If cavitation were an issue for these pumps we would have impellers with a Swiss cheese appearance- never seen anyone post pictures of such. Similarly the only way gas is going to get into the suction is if the oil level in the sump is too low- a distinct possibility if, as and when a shed load of oil were to be held captive in the cam chamber. I suspect this has a lot to do with 2/6 failure but to be clear, the 2/6 failure probably occurs first because it is the weakest link in the system- aka the first bearing to be impacted when the lube pump suction is compromised and thus a symptom of a bigger problem- not the root cause- thus why racers adopted the accusump approach to give some back up if and when oil pressure faltered. Not really a problem on examples used purely for road use although I have read of Autobahn engine grenades in ze Fatherland!

You have stated previously that you are of the opinion that some of the oil pump flow is is diverted through the relief valve- I have stated that I believe that is just not the case based on the evidence I have seen and GB's data supports that. What you choose to believe is down to you- just sharing my thoughts and why I stick to a 20W50 oil..

GregBBRD

Not sure where the specification of "8 bar" for the pressure relief valve to open comes from....but what happens if this is inaccurate and the valve opens at 7 bar? I can see my results matching up to this, pretty easily. I'd guess that at higher rpms, the small "ports" in the pressure relief system becoming overwhelmed a the pressure rising, even with the valve wide open.

FredR

The 8 barg spec I have seen referred to on a number of occasions but the first time I saw it in print was in the technical data sheets Tuomo posted earlier in this thread. The relief valve discharge ports are not likely to get overwhelmed or blocked with a 7 bar differential pressure across them and one presumes by the tenor of your post that you have not seen that happen - assuming that is the case then I doubt anyone else would have.. Of course, the oil at that point has not been through the oil filter and crap has been found in engine sumps from time to time but the fact that it is there [and nowhere else?] tells us something!

Spring operated relief valves are very dependable and in the process industries they are the last line of defence. That being said we have to test them on a regular basis [every two years] and to a specific protocol. The critical measure is the popping pressure- i.e. the pressure at which they start to open and all such valves have to be fully open by the time the accumulated system pressure hits 110% of the popping pressure. Although I cannot be 100% certain I am pretty sure the relief valve in the engine will follow the same practice but rarely if ever will it get to a fully open situation. I have no clue as to what capacity the relief valve is sized for. It is nigh on impossible for the system to be completely blocked but in theory I suppose it is possible if the problem occurred in the discharge port- in such circumstance the relief valve should be sized for the pump capacity at max rpms.

As our cars age it is highly unlikely that the relief valve will fail even if it is never pulled apart and cleaned but I would think it is highly likely that the spring resistance will tire some and the set point may well reduce as a consequence. Should this happen the consequences for the most part will be irrelevant. Your data on Andy's engine tells us that at max rpms the pressure is around 106 psig or 7.3 barg- that on an engine that was rebuilt to the highest standards, doubtless with a new oil pump and spot on bearing clearances that under test clearly exceed the Porsche oil pressure minimum specifications. Over time those numbers will invariably drop slightly as the bearings bed in and wear but either way, it is unlikely that the relief valve set pressure will drop below the oil pressures currently generated. Do the same pressure test on my motor that is now 29 years old and covered 100k miles and I am pretty sure my oil pressure will be noticeably lower but, as long as it is greater than 5 barg at 4.8k rpms, it would still meet the Porsche minimum spec and I would like to test for this one of these days but on the positive side, when I fitted the earlier two pin sender I did a calibration check and the display values aligned correctly at 2 barg and 5 barg.

On older engines, should oil pressure drop below the specified benchmark, then my guess is that accelerated wear in the cam area will take place as that is more likely to be susceptible to lower oil pressure [or so I would think]. On the GTS motor you are currently restoring it would be interesting to know what has caused the cams to wear out the way they did as it does not look like a mileage related thing.

Now, if one goes by the Porsche spec that the oil pressure should be higher than 5 barg at 4800 rpms when fully warm, that may be a good performance indicator to advise the owner that a major strip down is needed to replace bearings and/or that a new oil pump is needed as a minimum. However, before resorting to nay drastic action I would pull the relief valve and replace the spring with a new one to see if that modulates the condition at all.

ptuomov

It's a spring operated valve. It will flow more at higher pressure, and will need a higher pressure to flow more. If it cracks at pressure of X, it'll be fully open at some pressure c*X with c>1. Constant c is determined by the spring length valve closed and spring length valve fully opened. (Size of the opening, which is the main determinant of the bypass flow, depends on the geometry of the valve. There's also the direct effect of the same size opening flowing more at higher pressures by square root law, but that's a minor effect, I believe.) There's no mystery that the oil pressure needs to be slightly higher when the pump bypass needs to flow 30% of the pumped oil vs. slightly lower when the pump bypass needs to flow 5% of oil.

I believe that the 928 S4 rod bearings do indeed consume more oil at 6000 rpm than at 3000 rpm. The issue is complex because on the one hand the pumping action at the rod journal reduces flow resistance while the centrifugal force at the main bearings increases flow resistance. At very high rpms, the centrifugal force wins and the flow resistance actually starts growing with rpms again. The rod bearings are just one leak in the engine, though, and the flow resistance doesn't drop (and flow at a given pressure doesn't increase) that much with rpms.

Here's a generic (but believable) example from the internet, I believe it's originally from Ricardo paper (but I'm not sure):




As a side note, if it wouldn't be for the rod journals being oiled from the mains (say instead oiling from the crank nose), the 928 S4 engine wouldn't need anything close to 5 bar oil pressure. Only reason why I'd be concerned about low oil pressure at low rpms is that it may be indicative of some other problem than just the oil pressure.