Official S65 Bearing Specification/Clearance Wiki

[BMRLVR]

Originally Posted by BobS

You are correct, I don't know the issue. Common sense would say if it's a clearance issue, every engine or almost every engine would be effected. Clearance would be in the design of the engine and common on all s65's.

I work for a company that rebuilds industrial, marine, truck and generator diesel engines. I've seen where a engine was assembled using the wrong sized bearings (believe it or not they were labeled wrong from the factory!..ouch). The engine didn't make it off the dyno. The bearings looked like these pictured and there was metal all through the engine/pan.

I believe if this was a clearance issue, the failures would be more common and severe

Funny, I work for a company and I rebuild large Diesel engines for mining equipment. The issue with the clearances the S65 is built with is the fact that they are on the borderline. They aren't so tight that the engine fails immediately, they are as such that the lubrication of the bearing is compromised where a boundary lubrication condition is occurring. This combined with BMW's curious recommendation of a 10W60 oil in an engine with such tight clearances.

I have no doubt it is clearances, please take a look at all of the posted info and photos again and I think you will be convinced.

[kawasaki00]

Originally Posted by Yellow Snow

Yes, place a reference or inspection grade 2mm slip gauge on a good surface table. Then zero out a high quality lever type dial gauge on the slip. You can see a witness mark on the shell where they were checked at the factory.

This is a highly accurate way of measuring and surely better for your reference when comparing with other shops across the world than relying on someones mic feel.

Your photo's are interesting and seem to indicate that clearance isn't necessarily the problem here.

My theory is that the bearings are different material, hence the price difference between the two. Is it beyond the realms of possibility that some motors have been assembled with the wrong bearings top/bottom ie red top, blue bottom?

Could you check part numbers or colours on a failed set?

I know there have been a few bearing failures but percentage wise across the cars production it's a minute proportion. It's virtually unheard of in the UK.

We will soon know this answer, I have a set going to the QC lab on Monday for hardness testing and we will also know what material it is. I also have a Clevite rep coming to look at them also. I find it hard to believe they used two different types of shells for the upper and lower.

Originally Posted by BMRLVR

Funny, I work for a company and I rebuild large Diesel engines for mining equipment. The issue with the clearances the S65 is built with is the fact that they are on the borderline. They aren't so tight that the engine fails immediately, they are as such that the lubrication of the bearing is compromised where a boundary lubrication condition is occurring. This combined with BMW's curious recommendation of a 10W60 oil in an engine with such tight clearances.

I have no doubt it is clearances, please take a look at all of the posted info and photos again and I think you will be convinced.

[regular guy]

Originally Posted by e92zero

Is it possible the larger clearance is due to the bearings being worn out? Or the clearance is measured with one single set of upper/lower bearings on each of the rod after disassembly?

I also measured each bearing shell individually. Let me go back to that spreadsheet and see what it says. However, I did indirectly test that condition by swapping bearings between the largest and smallest examples. The clearance followed the rod. So I think the next step is to measure the rod bores themselves.

I have another request to measure the actual rod bore diameter. I had a feeling we would eventually get to this. So hopefully in the next two weeks, I'll be able to take those measurements. For this test, I have two full sets of factory connecting rods. So let's see what the measurements say.

Originally Posted by Yellow Snow

Yes, place a reference or inspection grade 2mm slip gauge on a good surface table. Then zero out a high quality lever type dial gauge on the slip. You can see a witness mark on the shell where they were checked at the factory.

I'll look into it. I'm always in for cross checking measurements.

My theory is that the bearings are different material, hence the price difference between the two. Is it beyond the realms of possibility that some motors have been assembled with the wrong bearings top/bottom ie red top, blue bottom?

Could you check part numbers or colours on a failed set?

Unfortunately, I've never seen any color designations on the rod bearings after the engines are disassembled. I think the oil disolves the paint on the rod bearings. All that's left are bearing shells that are all uniformly marked 088/089 or 702/703. So short of measuring them after they come out, I think it might be too hard to tell.

But alas, I did take measurements of the bearing thickness on Crankshaft-1. I'll dig those up tonight and see how they correlate with clearance.

[Yellow Snow]

Originally Posted by regular guy

Unfortunately, I've never seen any color designations on the rod bearings after the engines are disassembled. I think the oil disolves the paint on the rod bearings. All that's left are bearing shells that are all uniformly marked 088/089 or 702/703. So short of measuring them after they come out, I think it might be too hard to tell.

Can you get the 088/089 or 702/703 part numbers referenced to which way they were fitted in the rods.

[V8FunNaturally]

I could not find additional details on Porsche engines but after more research I did for the Ferrari 328 GTS/GTB which is a good high revving 3.2l V8 from the old days. I have included additional specs that might help in making an informed comparison given that this is an earlier supercar. I just realized the table I created has ferrari colors, not that I am affiliated with them in any ways .

(think metric again)

Ferrari 328 GTB/GTS engine Specs



How to interpret this table:

The Ferrari main bearing clearance is 2.5 times tighter than the S65 engine.
(in inches: 0.00051 versus 0.00115)

The Mahle-Clevite white paper guidance on rod/main clearances is completely out of agreement with Ferrari's clearance for main bearings by a factor of 3.5 X. Not the same can be said for the rod bearings. Ferrari's rod bearing clearance is slightly more conservative than MC's recommendation.

There are three different journal diameters available on the ferrari: nominal, 1st undersize and 2nd undersize. These must be paired with the corresponding bearing :nominal, 1st oversize or 2nd oversize. The result does NOT change the clearance range it only allows a thicker bearings to be used on narrower crank journals.
As we know, The S65 OEM parts catalog has "bearing shells" "yellow", "violet" and "green" for the (main) "bearing shell" and for the end "guide-bearing shell" available in both tops and downs but only a single set for the rod bearing: top blue and bottom red. Are the main shells different thickness like Ferrari's ? Are these factory hand selected for optimal pairing with journals and why do the rod bearings only have one single set ?

I do not know what is the definition of the "limit of wear" spec on the clearances, so let me know if you do.

Rod bearing side clearance 0.200-0.339 mm (0.0080 - 0.0133 inches). Ok, so I was wrong about hair size. What about 0.040 - 0.120 mm ? In which case the S65 rods side clearance would be tighter than this Ferrari by 2X. The side clearance is important because it provides a fixed restriction for oil flow on exit from bearings back to the crankcase especially if the bearings become worn with excessive diametrical clearance. If not for this restricted side clearance, one loose con-rod bearing would dramatically reduce oil pressure for the entire engine. Could this have been a design setup to maintain oil pressure in the bearing without having to over design the wet sump oil pump capacity like in the Ferrari ?

Let's look for a moment how the 328 GTS keeps lubricated and cooled. The Ferrari oil pump has a huge capacity. Engine's oil pressure increases with RPM but in the 328 GTS the pump flow is so high that it hits the oil pressure regulator limit at only 3000RPM (or even less) and pressure at the pump remain constant up to its 7700 RPM redline. This can be seen live on the dashboard gauge. Above 3000 RPM the oil flow at the pump still increases but more of it is bypassing the engine through the regulator while the pressure is maintained constant at max. This is not ideal for mileage and horsepower but certainly avoids premature engine wear. The flow of oil delivered into the bearings is still temperature dependant. When the oil is cooler, viscosity creates resistance and the non diverted flow will be less. Also If that engine is old and the bearing have sustained wear the flow will be higher as less oil is now being bypassed. With the pressure regulated at max the oil, resistance paths have to be well balanced with no easier path starving all the others. Yet if the regulated pressure is high enough it can make up for some of that imbalance.
In comparison a positive displacement oil pump has constant flow (at a given RPM) which provides less buffering against the effect of imbalance in oil paths resistance. This will starve the tightest clearance path.

Oil in a crank/rod bearing doesn't exist as a simple film. As one surface is sliding by another, the oil sort of 'rolls' between the two surfaces. It forms into microscopic rods, analogous to the needles in a roller bearing. The diameter of these 'oil rollers' is not dependant upon pressure, but viscosity. (And, of course, temperature's influence on viscosity.) Adequate pressure only assures that there will be an ample quantity of 'rollers', and that fresh oil comes in for cooling reasons. The old recommended adage for oil pressure is 10psi/1000rpm. If there is a lower path of resistance for the oil to flow that is where it will go.
Viscosity needs to be choosen such that the 'rods' are of the correct diameter, and fill up the clearance in the bearing. Conversely, the clearances could be designed to suit a particular lubricant. In many years gone past, people used to build small block Chevies with very large clearances. It's not considered proper now, but was the fad back then. You could put in 30W and HAVE PROPER PRESSURE AT ALL RPMS, but the motor wouldn't make it to the A-main. With 50W, it would last. Again, motors are being built now days with tighter clearances, so even racing engines will tend to use much thinner oil than we used to. Smaller clearances, smaller 'oil needle bearings'. Fast forward a number of years. As synthetics have less of a tendancy to thin with temp, a thinner synthetic may provide proper high temp 'oil roller' diameter.

I have put the materials for the bearing shells in the table. Not all bearings shells are equal. As to the specifics, there are tri-metal bearings (steel, copper and white metal) vs. bi-metal bearings (steel and white metal), there are lead-indium bearings (hard white metal) vs. lead-tin bearings (soft white metal), and there are different levels of precision and running clearance available. For a daily driver used for casual touring, and not expecting to rack up too many miles in the next 10 or 20 years, selecting the bearing is less critical. For a competition engine or long life you may prefer lead-indium over lead tin. For either competition or a hard running engine doing lots of miles, better precision with closer control of oil film thickness may be preferred.

For an high revving engine BMW's rod clearance choice seems arcane. There have been engine with equally tight (or even tighter) rod clearance that had no problem staying lubricated and cooled but their redline was lower than 8300 RPM. The extra clearance would prevent starving the flow of oil necessary to cooling the bearing especially with a positive flow oil pump.

I have links to the full specs for the oils in the table if interested.

[JEllis]

Interesting

[regular guy]

Originally Posted by sunsweet

I(think metric again)

Ferrari 328 GTB/GTS engine Specs

Thanks for the research on this. I collated the data with S65 actuals and put it in the following chart. I realize you came to the same (or very similar) conclusions that I summarize below.

	Jouranl Diameter	Min (mm)	Max (mm)	Nominal (mm)	Min (inch)	Max (inch)	Nominal (inch)	JIC Ratio	Oil Type
S65B40 Main (actual)	59.98000	0.02910	0.04580	0.03650	0.00115	0.00180	0.00144	0.00061	TWS 10W60
S65B40 Rod (actual)	51.98250	0.02930	0.04690	0.03250	0.00115	0.00185	0.00125	0.00061	TWS 10W60
Ferrari 328 GTS Main	62.98550	0.01300	0.05100	0.03200	0.00051	0.00201	0.00126	0.00051	Agip 10W40
Ferrari 328 GTS Rod	43.62800	0.04600	0.08900	0.06750	0.00181	0.00350	0.00266	0.00155	Agip 10W40

The table above compares nominal to nominal (which I think is the correct way to compare). If you compare smallest to smallest and largest to largest, then you're comparing apples to oranges because the Ferrari values come from a maximum allowed specifications, but the S65 values come from observations over only two crankshafts. We don't know what the maximum BMW specifications are -- making it impossible to compare actual specifications between these two motors.

I created a column called "JIC Ratio" which means "Journal/Inch Clearance Ratio." This is a calculation oil clearance per inch of journal diameter.

The S65 (actual) has a JIC of 0.00061" clearance per journal/inch for both main and rod bearings. I'm actually glad these numbers match for both main and rod bearing clearance because it seems pretty logical BMW would use the same value if it's based on philosopical reasons.

The Ferrari 328 GTS uses 0.00051" clearance per journal/inch for the mains. This is 17% (1.2x) smaller than the BMW value. If we did compare min-spec Ferrari to min-observed S65, then the Ferrari main clearance is 44% (2.25x) smaller than the S65. Comparing max-spec Ferrari to max-observed S65, then the Ferrari main clearance is 11% (0.89x) larger than the S65. In my opinion, the main clearance between the Ferrari 328 GTS and BMW S65 are very comparable.

But the rod bearing clearance is what we care most about. The Ferrari rod bearing clearance tells a completely different story than the S65. In the case of the rod bearing clearance, the Ferrari is even more conservative than the Mahle/Clevite specification of 0.001" clearance per journal/inch diameter. The Ferrari 328 GTS uses a journal-to-clearance ratio of 0.00155" clearance per journal/inch diameter. That's much more conservative than the Mahle/Clevite specification. Whereas BMW S65 uses 0.00061" clearance per journal/inch diameter -- the same value they use on their main bearings as well.

So while the Ferrari and BMW mains are relatively the same, the Ferrari clearly uses even more clearance per journal/inch diameter than the Mahle/Clevite white paper recommends. And the Ferrari uses thinner oil as well.

As we know, The S65 OEM parts catalog has "bearing shells" "yellow", "violet" and "green" for the (main) "bearing shell" and for the end "guide-bearing shell" available in both tops and downs but only a single set for the rod bearing: top blue and bottom red. Are the main shells different thickness like Ferrari's ? Are these factory hand selected for optimal pairing with journals and why do the rod bearings only have one single set ?

I have seen one S65 motor disassembled with a mix of green/yellow bearing shells on the mains.

If I get time later today, I'll post information about the BMW wet sump oil system design. The BMW design sounds very similar to the Ferrari design, except the BMW design also incorporates a variable pressure bypass in addition to a relief valve.

[Ezio]

aint worried about nothin

[Gearhead999s]

Even Ferrari V8's have bottom end issues.

[V8FunNaturally]

Originally Posted by regular guy

The table above compares nominal to nominal (which I think is the correct way to compare). If you compare smallest to smallest and largest to largest, then you're comparing apples to oranges because the Ferrari values come from a maximum allowed specifications, but the S65 values come from observations over only two crankshafts. We don't know what the maximum BMW specifications are -- making it impossible to compare actual specifications between these two motors

You are right, thanks for pointing that out. By the way the more S65 clearance measurements we collect from different S65 engines the better idea we will have of the nominal value and particularly the standard deviation. From the Std Dev, mean and number of samples and using basic statistics math on normal distribution you can tell with a known % certainty within what interval the true nominal value lies. Basically the measured clearances on parts coming out of produced engines will be on a normal distribution except non compliant parts outside BMW min, max tolerances specs were already rejected and thrashed at the factory (this complicates the mathematical estimation slightly. the side tails of the distribution outside the tolerance values will not be represented in the samples we get). Furthermore if we make an hypothesis on what the minimum critical clearance should be and play with that value (assuming that this is the key issue, I am still not entirely convinced , i'd like to know more about the oil pump) then we will be able to predict what percentage of engines "pass" or "fail". If that number is 0.1 % I would not be too much worried.

Originally Posted by regular guy

I created a column called "JIC Ratio" which means "Journal/Inch Clearance Ratio." This is a calculation oil clearance per inch of journal diameter.

Yes, good idea but I would have computed the clearance/journal ratio completely and expressed the result in ppm (parts per million). This would give the same metric without inch or millimeter dependency and can also be used to compare different engines.

Originally Posted by regular guy

If I get time later today, I'll post information about the BMW wet sump oil system design. The BMW design sounds very similar to the Ferrari design, except the BMW design also incorporates a variable pressure bypass in addition to a relief valve.

I would be interested to see more than the basic marketing info on that.

[aussiem3]

I just changed the oil and after just 4k km since the last one, the oil was black. I kept some for oil analysis, and just couldn't believe the colour. It's more like the oil coming out of a diesel engine.

[regular guy]

Originally Posted by sunsweet

[Regarding the oil pump...] I would be interested to see more than the basic marketing info on that.

I know I have two different documents that describe the S65 oil system, and all I could find at the moment is the one you would probably describe as basic marketing info. I'll keep searching for the other one.

However, in a nutshell, I'm pretty sure it's a positive displacement pump. I know it has a variable pressure feedback loop to help regulate the engine's oil pressure needs. It also has a maximum pressure relief valve -- which I think hints it's a positive displacement pump. I could be all wrong about that last part, so I think it's best to let some of the engine guys chime in with more details.

[BMRLVR]

Originally Posted by regular guy

I know I have two different documents that describe the S65 oil system, and all I could find at the moment is the one you would probably describe as basic marketing info. I'll keep searching for the other one.

However, in a nutshell, I'm pretty sure it's a positive displacement pump. I know it has a variable pressure feedback loop to help regulate the engine's oil pressure needs. It also has a maximum pressure relief valve -- which I think hints it's a positive displacement pump. I could be all wrong about that last part, so I think it's best to let some of the engine guys chime in with more details.

The pump in the S65 is indeed a positive displacement vane pump (all engine oil pumps have to be positive displacement or pressure would not be very high or consistent at lower RPM's......... The same way a centrifugal SC makes little boost at low RPM). The pump is also variable displacement which means it is able to vary oil flow. The flow control is accomplished by oil pressure fed back to the pump through a signal passage right off of the main oil gallery. This signal pressure will act on spring pressure and pressure at the outlet of the pump and when spring pressure overcomes the oil pressure in the signal passage, the pump "strokes up" and oil flow increases until the pressure is balanced with spring pressure again and the pump "de-strokes" and oil flow decreases. In a pressure compensated variable displacement hydraulic system like this, pressure control will mainly be controlled by the pumps flow control circuit however a relief valve will be there to prevent system damage from over pressure due to cold oil or malfunctioning pressure compensating/flow control circuit

[V8FunNaturally]

How many actually documented cases of failed engine are found on stock S85 M5 ? Is it an even rarer occurence than on the S65 ?

The M5 S85 oil pump is a high-pressure pump because the VANOS oil is supplied by it and those VANOS need a high pressure. The S65 VANOS have considerably stronger switching moment and only require low-pressure. BMW replaced the high pressure pump from the S85 with a flow controlled hinged-valve oil pump with a feed capacity adjusted to suit the VANOS low-pressure system. Another difference was in splitting the scavenger and main pump locations. There is no space in the S65 to fit the scavenger pump and the main oil pump in the same housing like in the S85. So they use separate drive. The cranckshaft gearwheel drives the scavenger pump first and the scavenger pump shaft drives the main pump via chain drive.

Based on the info I already had the oil pressure regulation line feedback and the variable volume flow hinged valve main oil pump achieve a constant pressure at the pump output which is lower than in the S85 based on VANOS requirement.

In fixed-displacement oil pumps, designers typically oversize the pumps to handle the harshest engine operating conditions (Ferrari). Most of the time, they consume more power and deliver significantly higher oil pressure than needed. Or is that really not needed for the main/rod bearings ?? They contain pressure-relief valves as a crude, cost-effective, and reliable way to avoid excessively high oil pressures.
Variable-displacement oil pumps eliminate excess oil flow (the volume of displaced oil inside the pump per pump rotor rotation is adjustable which is what the feedback line controls) significantly reducing the parasitic load on the engine crankshaft, and ultimately saving fuel. But variable-displacement pumps can reduce the oil pressure to as a low as 1-2 bars compared to the 4-6 bars from standard fixed-displacement pumps. They can also provide much lower flow rates. The chain primary drive's hydraulic tensioner is affected by oil feed pressure. Maintaining a lower pressure lowers the running tension in the chain drive. During warm-up, this more efficient pump heats the oil less, producing oil that is cooler (not good for a low clearance ?). Less overall heat is transferred to the oil around the engine's piston and bore regions. But reduced oil flow also increases combustion chamber temperatures and reduces piston cooling, which heats the metal cylinder liners (+2°C to 6°C was measured in a industry consortium study on a single engine using this type of pump) .

The lower crankcase of the S65 is constructed using die-cast aluminium. Grey cast iron inlays are used to reinforce the bedplate construction. These also limit crankshaft bearing clearances over a greater temperature range and thus supposedly have a positive effect on the oil flow rate.

It would be interesting to collect oil pressure data on a running engine test for the S65. Is there an OEM oil pressure sensor in there and could the data be accessible from the MSS60 Engine control system via OBD ? If not is there a way to install a custom sensor ?

[regular guy]

Originally Posted by sunsweet

How many actually documented cases of failed engine are found on stock S85 M5 ? Is it an even rarer occurence than on the S65 ?

The M5 S85 oil pump is a high-pressure pump because the VANOS oil is supplied by it and those VANOS need a high pressure. The S65 VANOS have considerably stronger switching moment and only require low-pressure. BMW replaced the high pressure pump from the S85 with a flow controlled hinged-valve oil pump with a feed capacity adjusted to suit the VANOS low-pressure system. Another difference was in splitting the scavenger and main pump locations. There is no space in the S65 to fit the scavenger pump and the main oil pump in the same housing like in the S85. So they use separate drive. The cranckshaft gearwheel drives the scavenger pump first and the scavenger pump shaft drives the main pump via chain drive.

Based on the info I already had the oil pressure regulation line feedback and the variable volume flow hinged valve main oil pump achieve a constant pressure at the pump output which is lower than in the S85 based on VANOS requirement.

In fixed-displacement oil pumps, designers typically oversize the pumps to handle the harshest engine operating conditions (Ferrari). Most of the time, they consume more power and deliver significantly higher oil pressure than needed. Or is that really not needed for the main/rod bearings ?? They contain pressure-relief valves as a crude, cost-effective, and reliable way to avoid excessively high oil pressures.
Variable-displacement oil pumps eliminate excess oil flow (the volume of displaced oil inside the pump per pump rotor rotation is adjustable which is what the feedback line controls) significantly reducing the parasitic load on the engine crankshaft, and ultimately saving fuel. But variable-displacement pumps can reduce the oil pressure to as a low as 1-2 bars compared to the 4-6 bars from standard fixed-displacement pumps. They can also provide much lower flow rates. The chain primary drive's hydraulic tensioner is affected by oil feed pressure. Maintaining a lower pressure lowers the running tension in the chain drive. During warm-up, this more efficient pump heats the oil less, producing oil that is cooler (not good for a low clearance ?). Less overall heat is transferred to the oil around the engine's piston and bore regions. But reduced oil flow also increases combustion chamber temperatures and reduces piston cooling, which heats the metal cylinder liners (+2°C to 6°C was measured in a industry consortium study on a single engine using this type of pump) .

The lower crankcase of the S65 is constructed using die-cast aluminium. Grey cast iron inlays are used to reinforce the bedplate construction. These also limit crankshaft bearing clearances over a greater temperature range and thus supposedly have a positive effect on the oil flow rate.

This is the same info I opted not to quote from BMW literature yesterday because it sounded too marketing driven to me.

It would be interesting to collect oil pressure data on a running engine test for the S65. Is there an OEM oil pressure sensor in there and could the data be accessible from the MSS60 Engine control system via OBD ? If not is there a way to install a custom sensor ?

I haven't yet found one through OBD or DCAN. I'm still checking with some contacts to see if one exists and how to access it. I've been looking for this for a while...but nothing so far.

I've got some local cars and toying with the idea of instrumenting data-logging oil pressure gauges.

[B767capt]

[QUOTE=sunsweet;14741714]How many actually documented cases of failed engine are found on stock S85 M5 ? Is it an even rarer occurence than on the S65 ? The M5 S85 oil pump is a high-pressure pump because the VANOS oil is supplied by it and those VANOS need a high pressure.




There are 2 pumps in the S85. The vanos pump is gear driven directly off the crank and can produce up to 1500+/- psi for vanos operation. There is a chain on the vanos pump that runs over to the engine oil pump and that is how the engine oil pump is driven. I just did my rod bearings and had both of them out. I assume the oil pump puts out similar to the S65.

[BMRLVR]

Originally Posted by sunsweet

How many actually documented cases of failed engine are found on stock S85 M5 ? Is it an even rarer occurence than on the S65 ?

The M5 S85 oil pump is a high-pressure pump because the VANOS oil is supplied by it and those VANOS need a high pressure. The S65 VANOS have considerably stronger switching moment and only require low-pressure. BMW replaced the high pressure pump from the S85 with a flow controlled hinged-valve oil pump with a feed capacity adjusted to suit the VANOS low-pressure system. Another difference was in splitting the scavenger and main pump locations. There is no space in the S65 to fit the scavenger pump and the main oil pump in the same housing like in the S85. So they use separate drive. The cranckshaft gearwheel drives the scavenger pump first and the scavenger pump shaft drives the main pump via chain drive.

Based on the info I already had the oil pressure regulation line feedback and the variable volume flow hinged valve main oil pump achieve a constant pressure at the pump output which is lower than in the S85 based on VANOS requirement.

In fixed-displacement oil pumps, designers typically oversize the pumps to handle the harshest engine operating conditions (Ferrari). Most of the time, they consume more power and deliver significantly higher oil pressure than needed. Or is that really not needed for the main/rod bearings ?? They contain pressure-relief valves as a crude, cost-effective, and reliable way to avoid excessively high oil pressures.
Variable-displacement oil pumps eliminate excess oil flow (the volume of displaced oil inside the pump per pump rotor rotation is adjustable which is what the feedback line controls) significantly reducing the parasitic load on the engine crankshaft, and ultimately saving fuel. But variable-displacement pumps can reduce the oil pressure to as a low as 1-2 bars compared to the 4-6 bars from standard fixed-displacement pumps. They can also provide much lower flow rates. The chain primary drive's hydraulic tensioner is affected by oil feed pressure. Maintaining a lower pressure lowers the running tension in the chain drive. During warm-up, this more efficient pump heats the oil less, producing oil that is cooler (not good for a low clearance ?). Less overall heat is transferred to the oil around the engine's piston and bore regions. But reduced oil flow also increases combustion chamber temperatures and reduces piston cooling, which heats the metal cylinder liners (+2°C to 6°C was measured in a industry consortium study on a single engine using this type of pump) .

The lower crankcase of the S65 is constructed using die-cast aluminium. Grey cast iron inlays are used to reinforce the bedplate construction. These also limit crankshaft bearing clearances over a greater temperature range and thus supposedly have a positive effect on the oil flow rate.

It would be interesting to collect oil pressure data on a running engine test for the S65. Is there an OEM oil pressure sensor in there and could the data be accessible from the MSS60 Engine control system via OBD ? If not is there a way to install a custom sensor ?

Your information is wrong on many accounts.

First: The S85 has just as high a rate of failure in stock form....... In fact the big kicker for us over here was a link to an M5 board thread on an S85 bearing failure. There were some posters on there that had done bearings on many S85's and they saw the same thing as well as posted pictures. The S85 uses the exact same bearings, rods and journal diameters as the S65 so it is no surprise that there is an issue with it as well since it too has the same clearance issues.

Second: The S85 doesn't use a high pressure pump for the entire engine due to it's VANOS system......... To use the relief valve for the engine reduce the VANOS pressure of 115 BAR (~1670 PSI) down to engine oil pressure of about 5 BAR (~73 PSI) is not practical. Since there is no flow control or pressure reducing valve pictured in the S85 lubrication circuit the relief valve would be the only way to drop the pressure. To anyone who knows hydraulics, they know that using a relief valve to reduce pressure creates a lot of heat, and that heat generated is increased a great deal the higher the system pressure is going over relief. If the system were set up in this way the oil would be oxidized beyond any usable condition in a matter of a few hours. It was simpler and easier and more reliable for BMW to have separate pumps for the two systems.

The S85 actually uses five oil pumps:

1) One variable displacement vane pump for engine lubrication.

2) One duo centric pump that transfers oil from the front sump to the rear sump (both the vane and duo centric pump are within one housing)

3) One high pressure radial piston pump complete with an accumulator for the VANOS system (The pressure side of this system is completely independent from the engine lubrication system).

4) Two electric scavenge pumps in the cylinder heads to return oil to the sump during high cornering forces. These pumps are activated using the ECM and information on G forces from the YAW sensor.

Please re-read your S85 lubrication system documentation again, because it is not as you are describing.

Third: The S65 uses two oil pumps, a scavenge pump and the main oil pump. The scavenge pump is just like the S85 but it sits where the VANOS high pressure pump sat in the S85. The Scavenge pump is gear driven off of the crank and the main oil pump is chain driven off of it.

Fourth: I am going to answer the following quote directly:

Originally Posted by sunsweet

In fixed-displacement oil pumps, designers typically oversize the pumps to handle the harshest engine operating conditions (Ferrari). Most of the time, they consume more power and deliver significantly higher oil pressure than needed. Or is that really not needed for the main/rod bearings ?? They contain pressure-relief valves as a crude, cost-effective, and reliable way to avoid excessively high oil pressures.
Variable-displacement oil pumps eliminate excess oil flow (the volume of displaced oil inside the pump per pump rotor rotation is adjustable which is what the feedback line controls) significantly reducing the parasitic load on the engine crankshaft, and ultimately saving fuel. But variable-displacement pumps can reduce the oil pressure to as a low as 1-2 bars compared to the 4-6 bars from standard fixed-displacement pumps. They can also provide much lower flow rates. The chain primary drive's hydraulic tensioner is affected by oil feed pressure. Maintaining a lower pressure lowers the running tension in the chain drive. During warm-up, this more efficient pump heats the oil less, producing oil that is cooler (not good for a low clearance ?). Less overall heat is transferred to the oil around the engine's piston and bore regions. But reduced oil flow also increases combustion chamber temperatures and reduces piston cooling, which heats the metal cylinder liners (+2°C to 6°C was measured in a industry consortium study on a single engine using this type of pump) .

When using fixed displacement oil pumps manufacturers didn't oversize them because they wanted to ensure the pumps would handle the "Harshest engine operating conditions". The manufacturers oversize the pumps because they have no choice. At low RPM's a fixed displacement pump that was sized for the max flow an engine would need wouldn't come close to supplying the oil that it needed at less than max engine speed. The issue with a fixed displacement pump is in it's name, FIXED. It pumps the same volume of oil with every revolution. The way one sizes a fixed displacement oil pump for an engine is that he wants the pump to maintain the minimum oil pressure for hot oil at a low engine RPM, usually at idle or about 2000 RPM (the points where most manufactures rate the oil pressure in their vehicles) and then it increases as RPM's increase up to the relief valve setting. Some engines will hit the relief valve by 3000 RPM, and some even less. Once you are on the relief valve you are not only wasting energy, you are creating heat too. The combustion process and internal friction of an engine makes more than enough heat to warm the oil, adding heat with a relief valve is not wanted in any way, shape, or form also many times going over relief can cause the oil to aerate which is another unwanted trait. Another issue with a fixed displacement pump is that it can not provide consistent system pressure. The pressure it makes is going to vary depending on oil temperature, engine speed and wether it is over relief or not.

With a variable displacement pump the pump is also oversized to the max flow that is necessary, however it will match it's flow characteristics with what pressure is required in the system. A properly sized variable displacement pump usually operates at somewhere around 50% duty under normal operation, giving it the ability to increase flow by 50% if necessary. Variable displacement pumps are also known as pressure compensated in my industry, Heavy Equipment. Pressure compensated variable displacement hydraulic systems are replacing fixed displacement hydraulic systems in almost every application now-a-days. The benefits to them are many; Constant pressure and flow regardless of engine speed, reduced load on the engine due to going over relief continuously when requesting max hydraulic power, cooler oil temps, and reduced pump size and weight. The only drawbacks to a variable displacement system is reduced ability to deal with contamination, additional complexity, and increased cost.

So lets relate my above description of pressure compensated variable displacement hydraulic systems to the lubrication system in the S65. The oil pressure in the S65 should be close to continuous across the RPM range (at operating temperature) regardless of RPM once it comes off of Idle, this is even more desirable for the VANOS system since it will work best with a constant pressure. I would like to see oil pressure data across the RPM range to see how the pressure reacts to RPM. The variable displacement pump is not going to cause less flow in the engine all by it self, it is trying to maintain flow at a point where pressure stays constant. Also of note the pressure desired by BMW in it's service documentation is 4-6 BAR so that dispels your fears that the variable displacement pump may cause pressure to drop to 1-2 BAR compared to a fixed displacement pump.

And finally Fifth: The crankcases of the S65 and S85 as well as bearings, crankshaft, connecting rods, and pistons are the same. These engines were developed side by side and are nearly identical save a few details and the difference of two cylinders and 1L of displacement.

I hope I covered everything to your satisfaction!

[aussiem3]

[V8FunNaturally]

Originally Posted by BMRLVR

Your information is wrong on many accounts.

Yes I was aware it might be, it helps to have the documentation like you do (not to mention the service documentation ). I knew posting whatever info I had and trying to fill in the blanks for what is missing would trigger someone's ego to complete and correct it. Thank you for stepping up. I am not a specialist on engines topics like VANOS hydraulics at all. But I knew the S65 does not have a high pressure pump for VANOS. The S65 VANOS hydraulic motors run from the low pressure main pump. I did not have a chance to read about the S85 oil pumps.

Originally Posted by BMRLVR

Second: The S85 doesn't use a high pressure pump for the entire engine due to it's VANOS system......... To use the relief valve for the engine reduce the VANOS pressure of 115 BAR (~1670 PSI) down to engine oil pressure of about 5 BAR (~73 PSI) is not practical. Since there is no flow control or pressure reducing valve pictured in the S85 lubrication circuit the relief valve would be the only way to drop the pressure. To anyone who knows hydraulics, they know that using a relief valve to reduce pressure creates a lot of heat, and that heat generated is increased a great deal the higher the system pressure is going over relief. If the system were set up in this way the oil would be oxidized beyond any usable condition in a matter of a few hours. It was simpler and easier and more reliable for BMW to have separate pumps for the two systems.

The S85 actually uses five oil pumps:

1) One variable displacement vane pump for engine lubrication.

2) One duo centric pump that transfers oil from the front sump to the rear sump (both the vane and duo centric pump are within one housing)

3) One high pressure radial piston pump complete with an accumulator for the VANOS system (The pressure side of this system is completely independent from the engine lubrication system).

4) Two electric scavenge pumps in the cylinder heads to return oil to the sump during high cornering forces. These pumps are activated using the ECM and information on G forces from the YAW sensor.

Please re-read your S85 lubrication system documentation again, because it is not as you are describing.

Third: The S65 uses two oil pumps, a scavenge pump and the main oil pump. The scavenge pump is just like the S85 but it sits where the VANOS high pressure pump sat in the S85. The Scavenge pump is gear driven off of the crank and the main oil pump is chain driven off of it.

Your information is wrong on many accounts.

^ this info is from the BMW S85 documentation. All I have seen is some sparse side references to the S85 from a marketing document on the M3 that talks about the philosophy for changes from S85 to S65 .

Originally Posted by BMRLVR

Fourth: I am going to answer the following quote directly:

When using fixed displacement oil pumps manufacturers didn't oversize them because they wanted to ensure the pumps would handle the "Harshest engine operating conditions". The manufacturers oversize the pumps because they have no choice. At low RPM's a fixed displacement pump that was sized for the max flow an engine would need wouldn't come close to supplying the oil that it needed at less than max engine speed. The issue with a fixed displacement pump is in it's name, FIXED. It pumps the same volume of oil with every revolution. The way one sizes a fixed displacement oil pump for an engine is that he wants the pump to maintain the minimum oil pressure for hot oil at a low engine RPM, usually at idle or about 2000 RPM (the points where most manufactures rate the oil pressure in their vehicles) and then it increases as RPM's increase up to the relief valve setting. Some engines will hit the relief valve by 3000 RPM, and some even less. Once you are on the relief valve you are not only wasting energy, you are creating heat too. The combustion process and internal friction of an engine makes more than enough heat to warm the oil, adding heat with a relief valve is not wanted in any way, shape, or form also many times going over relief can cause the oil to aerate which is another unwanted trait. Another issue with a fixed displacement pump is that it can not provide consistent system pressure. The pressure it makes is going to vary depending on oil temperature, engine speed and wether it is over relief or not.

With a variable displacement pump the pump is also oversized to the max flow that is necessary, however it will match it's flow characteristics with what pressure is required in the system. A properly sized variable displacement pump usually operates at somewhere around 50% duty under normal operation, giving it the ability to increase flow by 50% if necessary. Variable displacement pumps are also known as pressure compensated in my industry, Heavy Equipment. Pressure compensated variable displacement hydraulic systems are replacing fixed displacement hydraulic systems in almost every application now-a-days. The benefits to them are many; Constant pressure and flow regardless of engine speed, reduced load on the engine due to going over relief continuously when requesting max hydraulic power, cooler oil temps, and reduced pump size and weight. The only drawbacks to a variable displacement system is reduced ability to deal with contamination, additional complexity, and increased cost.

So lets relate my above description of pressure compensated variable displacement hydraulic systems to the lubrication system in the S65. The oil pressure in the S65 should be close to continuous across the RPM range (at operating temperature) regardless of RPM once it comes off of Idle, this is even more desirable for the VANOS system since it will work best with a constant pressure. I would like to see oil pressure data across the RPM range to see how the pressure reacts to RPM. The variable displacement pump is not going to cause less flow in the engine all by it self, it is trying to maintain flow at a point where pressure stays constant.

Good insights. That is what I was asking one page back. But if both the S85 and S65 have the same variable displacement main pump and both engines have the same failures then that does not factor out the pump from the equation. The facts are the Ferrari has no such issues and had a fixed displacement pump. Just taking notice of that..
Just like I said I want to see oil pressure data on a live engine. I hope regular guy put wheels on that project.

Originally Posted by BMRLVR

"The S85 uses the exact same bearings, rods and journal diameters as the S65 so it is no surprise that there is an issue with it as well since it too has the same clearance issues".

That is only speculation until the verification work is done to prove it and you have nothing in the way of proving that link at this point. From an engineering standpoint the following applies to any field of engineering. You cannot draw conclusions without going full circle. Here is the process :

1) You think the clearance is linked to the engine failures (rationaly unproven hypothesis)
2) Apply one way to fix the suspected source of the problem. For example Find an engine showing the premises of bearing failures and replace them with bearings with looser clearances. I know there are no bearings currently available on the market so take the crankshaft out and grind the journals to increase clearances and put new shells.
3) use that car for 10,000 more miles, including some track driving. open the engine again and check the state of wear of the bearings shells
4) If the shells shows sign of premature wear again, you've got nothing, zilch your initial hypothesis is most likely wrong.
5) If the bearings show reduced but still unusual wear, you have not figured the whole picture yet. It is not purely about clearance. There are other elements at play with the issue but you found a partial workaround.
6) If the bearings show NO sign of premature wear only then have you verified the cause of issue.

Until then you've got nothing but suspicion. It might be a very educated guess coming from experience but it still is just that. You cannot call the specified clearance "the issue".

If it was not for following that process most manifacturers would waste their time applying false cures, introducing further variables into a production issue and eventually never finding the true explanation and the failure mechanism at work.

[BMRLVR]

Originally Posted by sunsweet

Yes I was aware it might be, it helps to have the documentation like you do (not to mention the service documentation ). I knew posting whatever info I had and trying to fill in the blanks for what is missing would trigger someone's ego to complete and correct it. Thank you for stepping up. I am not a specialist on engines topics like VANOS hydraulics at all. But I knew the S65 does not have a high pressure pump for VANOS. The S65 VANOS hydraulic motors run from the low pressure main pump. I did not have a chance to read about the S85 oil pumps.


^ this info is from the BMW S85 documentation. All I have seen is some sparse side references to the S85 from a marketing document on the M3 that talks about the philosophy for changes from S85 to S65 .


Good insights. That is what I was asking one page back. But if both the S85 and S65 have the same variable displacement main pump and both engines have the same failures then that does not factor out the pump from the equation. The facts are the Ferrari has no such issues and had a fixed displacement pump. Just taking notice of that..
Just like I said I want to see oil pressure data on a live engine. I hope regular guy put wheels on that project.


That is only speculation until the verification work is done to prove it and you have nothing in the way of proving that link at this point. From an engineering standpoint the following applies to any field of engineering. You cannot draw conclusions without going full circle. Here is the process :

1) You think the clearance is linked to the engine failures (rationaly unproven hypothesis)
2) Apply one way to fix the suspected source of the problem. For example Find an engine showing the premises of bearing failures and replace them with bearings with looser clearances. I know there are no bearings currently available on the market so take the crankshaft out and grind the journals to increase clearances and put new shells.
3) use that car for 10,000 more miles, including some track driving. open the engine again and check the state of wear of the bearings shells
4) If the shells shows sign of premature wear again, you've got nothing, zilch your initial hypothesis is most likely wrong.
5) If the bearings show reduced but still unusual wear, you have not figured the whole picture yet. It is not purely about clearance. There are other elements at play with the issue but you found a partial workaround.
6) If the bearings show NO sign of premature wear only then have you verified the cause of issue.

Until then you've got nothing but suspicion. It might be a very educated guess coming from experience but it still is just that. You cannot call the specified clearance "the issue".

If it was not for following that process most manifacturers would waste their time applying false cures, introducing further variables into a production issue and eventually never finding the true explanation and the failure mechanism at work.

I have the pleasure of working with engineers on a daily basis and I can tell you that while many of them are brilliant individuals and are good at what they do, nothing can replace real world experience when it comes to some issues. I can tell you with nearly 100% certainty that the problem with the S65 lies strictly in the clearances, nothing else. We don't need to re-invent the wheel here, we just need to set the clearances in this engine to something closer to what has been known to work in internal combustion engines since their inception........... No need to over-think it any more than that.

There is no issue with the lubrication system, if their were an issue there would be wear issues in other areas of the engine since the main and rod bearings are the first components to get oil. If the oil pumps were the culprit the top end would most likely starve for oil and camshaft bearings, camshaft lobes and valve lifters would be prematurely wearing not to mention that the VANOS system would not work properly. Every other area of the engine looks fine which leads me to believe that the tight clearances are the problem we are dealing with here.

I Build engines for a living, Kawasaki00 builds race engines for a living, regular guy is friends with one of the most respected engine builders in North America (and probably the world) and all of us are screaming that these clearances are WAY TOO TIGHT!!! Another thing to add to this is that Dinan opens up the clearances on its S65 and S85 strokers to 0.0019" mains and 0.0022" on the rods.......... They too obviously seen issue when they pulled apart their first S85 and S65's and felt it needed to be corrected.

So my friend, I understand your reluctance to believe in this but trust me when I say that opening up the clearances on the S65 and S85 will fix the issue, and I am speaking as a professional not some yahoo on the internet trying to stir up shit.

[BobS]

I just don't understand how people are going 75,000-100,000 or more miles and not having any issues. It just seems to me if clearance was the issue there would be a lot more failures

[regular guy]

Originally Posted by sunsweet

Just like I said I want to see oil pressure data on a live engine. I hope regular guy put wheels on that project.

I will give it my best shot. The local guy I have in mind is a great candidate because his engine will soon go under the knife and get re-sized journals and larger displacement. So it would be really nice to have this data before and after journal size changes.

That is only speculation until the verification work is done to prove it and you have nothing in the way of proving that link at this point. From an engineering standpoint the following applies to any field of engineering. You cannot draw conclusions without going full circle. Here is the process :

1) You think the clearance is linked to the engine failures (rationaly unproven hypothesis)
2) Apply one way to fix the suspected source of the problem. For example Find an engine showing the premises of bearing failures and replace them with bearings with looser clearances. I know there are no bearings currently available on the market so take the crankshaft out and grind the journals to increase clearances and put new shells.
3) use that car for 10,000 more miles, including some track driving. open the engine again and check the state of wear of the bearings shells
4) If the shells shows sign of premature wear again, you've got nothing, zilch your initial hypothesis is most likely wrong.
5) If the bearings show reduced but still unusual wear, you have not figured the whole picture yet. It is not purely about clearance. There are other elements at play with the issue but you found a partial workaround.
6) If the bearings show NO sign of premature wear only then have you verified the cause of issue.

I'd agree with you up to the point of 10000 miles. I don't think that's a long enough cycle to see the difference. I think it needs to be about 30-40k miles to see conclusive evidence. Not sure you know this, but you need to disassemble the entire motor to get the crank out on the S65; it doesn't just drop out the bottom like other motors.

Until then you've got nothing but suspicion. It might be a very educated guess coming from experience but it still is just that. You cannot call the specified clearance "the issue".

If it was not for following that process most manifacturers would waste their time applying false cures, introducing further variables into a production issue and eventually never finding the true explanation and the failure mechanism at work.

The most important takeaway I got from the Ferrari comparison is the Ferrari's conformance to the journal/bearing clearance ratio school of design. They didn't just follow it, they were even more conservative, and they still used thin oil. The theory here is BMW's failure to follow that guideline coupled with the thick oil requirement leads to the bearing failures that all conform to the symptoms (and photos) found in the Clevite white paper. So even if it's just an unproven theory, the small clearance is an issue that should be addressed during the first engine rebuild.