I like to think of hydrodynamic lubrication like a boat planing. Yeah, simple minds think in simple ways.
5w-20 and engine longevity
- Thread starter z917990
- Start date
- Status
Wear is significantly increased whenever you have metal to metal contact, ie "boundary lubrication" conditions, as opposed to hydrodynamic or "full film" lubrication. A thinner oil decreases oil pressure and oil film thickness between all moving parts of ANY lubricated component. Hence you have a lower margin of safety with a thinner lubricant - regardless of how good the basestock and/or additive package is. For example, if the oil thins due to polymer shearing or fuel dilution, or from high temps/loads, you are more likely to see excessive wear with a thinner oil.
This concept does not rely on limited UOA data; it's a very basic principle of tribology and is NOT open to debate.
This concept does not rely on limited UOA data; it's a very basic principle of tribology and is NOT open to debate.
So basically what you all are saying is that pressure of oil itself it NOT the main factor, but the quantity of oil that is pushed to the desired point of lubricating film is the reasone that the 5w20 or 0w20 viscosity oils will lubricate today's close tolerance engines even over a 5w30 or 10w30.
Is my logic right so far?
Is my logic right so far?
For an engine of a given design - including oil pump output, oil temps and tolerances - oil pressure IS a direct function of viscosity, or more specifically high temp, high shear rate viscosity.
However, engines can be designed to work well with very low viscosity oils. This is primarily done by:
1) increasing oil pump output (increases oil pressure for a given oil viscosity)
2) increasing the contact area of the crank/rod/cam bearings (decreases the loads between parts)
3) incorporating "roller" elements in the drivetrain components (eliminates sliding friction) and using more rounded cam lobe profiles, which decreases the load between the lobe and the lifter ( through increased contact area).
Changes can also be made to the piston/ring assembly to ensure proper lubrication and better cooling. For example, some engine now use "oil jets" to help cool the pistons.
However, engines can be designed to work well with very low viscosity oils. This is primarily done by:
1) increasing oil pump output (increases oil pressure for a given oil viscosity)
2) increasing the contact area of the crank/rod/cam bearings (decreases the loads between parts)
3) incorporating "roller" elements in the drivetrain components (eliminates sliding friction) and using more rounded cam lobe profiles, which decreases the load between the lobe and the lifter ( through increased contact area).
Changes can also be made to the piston/ring assembly to ensure proper lubrication and better cooling. For example, some engine now use "oil jets" to help cool the pistons.
I like data, so I made some charts of DarkDan's spreadsheet. The first one compares "Weighted Al/FE" with the various oil weights. Here it is;
Looks like a really good correlation of wear to weight, right? Well, I took a closer look and saw that there were some samples that really skewed the analysis. There was a direct correlation between OCI and wear. So, I reduced the data to only include samples where the OCI was greater than 3000mi and less than 10000 miles. This left me with only a couple samples for some weights. So, I grouped the results by XW20 and Xw30. Here is the graph of that data.
Funny how a closer look can change the results.
Remember, with these numbers a higher value means less wear. The weighted xx is determined by miles/ppm.
Looks like a really good correlation of wear to weight, right? Well, I took a closer look and saw that there were some samples that really skewed the analysis. There was a direct correlation between OCI and wear. So, I reduced the data to only include samples where the OCI was greater than 3000mi and less than 10000 miles. This left me with only a couple samples for some weights. So, I grouped the results by XW20 and Xw30. Here is the graph of that data.
Funny how a closer look can change the results.
Remember, with these numbers a higher value means less wear. The weighted xx is determined by miles/ppm.
Please post your second graph...!
You dont see two graphs? I do. BTW, here is the data;
So if I'm seeing the graphs and reading the data right, then the clear winner in the wear metals department is 0w20. Since I have a Honda, I could use this particular weight year 'round, and have better than enough protection, provided I don't decide to hit the racetrack and do 100 laps or so at 6500 rpm?
I just couldn't believe that something that pours like water could offer better protection than an already pretty thin 5w30.
I just couldn't believe that something that pours like water could offer better protection than an already pretty thin 5w30.
On Firefox/Linux, I don't see any second graph,
only a dash -
Going over to Windows/InternetExplorer, I see a
box with a red X in it... dead reference.
only a dash -
Going over to Windows/InternetExplorer, I see a
box with a red X in it... dead reference.
outstanding makes a simple graph and shows >20 are better? at least as AL goes.
bruce
Hmmm.. Let me explain again what I see from the data.
The first chart averages the the wear metal data from all samples of the same weight oil. It shows a clear trend that heavier weight oils show fewer wear metals per mile of usage. Right?
The problem is that there are a few samples that skew the data. For the heavier weight oil there are quite a few samples with an OCI over 10000 miles. There were even a couple with over 20000 miles. These super high mileage samples showed incredibly low numbers for ppm/mile of wear. Conversely the thin oils had quite a few samples with OCI's less than 3000 miles. These "low OCI" samples showed very high ppm/mile numbers. Thus, these "low OCI" samples made the 20 weight oils look bad and the "high OCI" samples made the 30 weight oils look good. As we have seen in many UOA's wear metals do not increase linearly with oil age. So, I threw out all of the "low OCI" samples and the "high OCI" samples and only used the samples with and OCI between 3000 and 10000 miles. That is what we see on the second chart. To me, the second chart shows no significant difference in wear metals between XW20 and XW30 oils.
Once again, the second chart shows that there is no correlation between wear metal ppm and oil weight when you use samples of a similar OCI.
The first chart averages the the wear metal data from all samples of the same weight oil. It shows a clear trend that heavier weight oils show fewer wear metals per mile of usage. Right?
The problem is that there are a few samples that skew the data. For the heavier weight oil there are quite a few samples with an OCI over 10000 miles. There were even a couple with over 20000 miles. These super high mileage samples showed incredibly low numbers for ppm/mile of wear. Conversely the thin oils had quite a few samples with OCI's less than 3000 miles. These "low OCI" samples showed very high ppm/mile numbers. Thus, these "low OCI" samples made the 20 weight oils look bad and the "high OCI" samples made the 30 weight oils look good. As we have seen in many UOA's wear metals do not increase linearly with oil age. So, I threw out all of the "low OCI" samples and the "high OCI" samples and only used the samples with and OCI between 3000 and 10000 miles. That is what we see on the second chart. To me, the second chart shows no significant difference in wear metals between XW20 and XW30 oils.
Once again, the second chart shows that there is no correlation between wear metal ppm and oil weight when you use samples of a similar OCI.
Maybe this chart will help explain it.
The lower wear metal ppms are correlated to the OCI not the weight of the oil.
The lower wear metal ppms are correlated to the OCI not the weight of the oil.
Ok, now I understand. Final verdict=pick the flipping weight of oil that you choose for your particular taste and style, right?
Hello,
I hope personal viewpoints are not seen as absolute!
Ted - you said;
"In the same engine under the same conditions, a significantly thinner oil will reduce oil pressure..."
I would add however, that a low viscosity lubricant will also typically reduce average oil temps. The difference in dropping down one SAE grade is normally a 5F-10F reduction in equilibrium oil temps. So the difference in oil pressure between say a 5w-30 and a 5w-40 is not as great as the difference in HT/HS viscosity - which is measured at the same 150C/302F temp.
The temperature "equivalency" between each successive SAE grade in terms of oil thickness is about 20F/10C. For example, the following "mid-grade" SAE samples would yield approx the same oil pressure:
xw-20; running @ 190F
xw-30; running @ 210F
xw-40; running @ 230F
xw-50; running @ 250F
All things being equal, the thinner oil running at the lower temp is preferred. Significantly lower fluid temps will reduce wear and also extend the life of elastomeric seals/gaskets...."
Ted these broad statements appear simply theoretical, are they? - sadly, they may not be too accurate!
An engine measured to the same criteria at the Manufacturer's prescribed datum point of say 4000rpm will in all probability show the same OP, be the oil a 0w-20 or a 15w-50! Because of the pumping losses alluded to by Bruce very few if any engine makers will provide a test number for OP at idle! The engine's condition - age, wear etc, may have some effect too!
I have done a lot of oil temperature testing in live water cooled engines Ted and I have not experienced a significantly meaningful reduction in residual oil temperatures in the oil pan when using less viscous lubricants! Aircooled engines are a different story!
Certain engine components may run cooler due to more linear oil flow characteristics if the change in viscosity is also from a mineral to a synthetic lubricant! Measuring the change would however require sophisticated test equipment
Generally, other factors have a greater influence on residual oil temperature than viscosity does!
Ted - you said:
"So the difference in oil pressure between say a 5w-30 and a 5w-40 is not as great as the difference in HT/HS viscosity - which is measured at the same 150C/302F temp."
Of course not Ted although M1 0w-40 and GC 0w-30 at 100C are near, and that is the point - and this is why the ACEA's HTHS Quality rating system is so valuable
This was the fallacious comment made to Buster by "the Amsoil guy" that prompted my first post;
"....the film thickenss isn't much different between a 20wt and a 30wt oil so you won't see much of a difference."
It referred to "thickness" not OP between an SAE20 and an SAE30 lubricant!
Ted I have trouble understanding this................;
"For an engine of a given design - including oil pump output, oil temps and tolerances - oil pressure IS a direct function of viscosity, or more specifically high temp, high shear rate viscosity."
Can you explain it please? - I'm have some trouble.....?
Ted this comment is meaningful;
"Changes can also be made to the piston/ring assembly to ensure proper lubrication and better cooling. For example, some engine now use "oil jets" to help cool the pistons."
In 1960 Mercedes used at least four compression ring designs and three oil ring designs that used any of four wall contact springs. Oil jets (piston squirters) for under crown piston cooling have been around for several decades too Ted.
Many German engine makers tried oil jets in production petrol engines two decades or so ago and many discarded them quickly too. Porsche used them in MY87 on some engines but only for a small production run. Oil foaming and etc is sometimes their product! Engines that continually operate at high revs have more difficulty with oil jets than those operating at low revs
(Oil jets have been successfully used in heavy diesel engines for several decades and with somewhat lessor lubricant issues. Many of these engine have flight revs at 2200rpm and operate at around 1600rpm with 1800rpm maximum)
Ring pack position on the piston and ring design have been changed in recent times to meet emission demands. Many now run very much closer to the "fire" area than a decade ago and this has introduced many lubricant issues. The average time that oil spends in the ring pack area is directly influenced by viscosity. There is some clinical evidence that lighter oils spend less time there and lubricate better than heavier viscosities (say 0w-20 against 15w-40)
In item 1) you said:
"increasing oil pump output (increases oil pressure for a given oil viscosity)"
Ted increasing the oil pump's output will only increase the volume! It will have no effect on pressure!
In you item 3) on valve trains you did not mention lighter valve spring pressures which have significantly reduced cam wear - especially with radical cams. You probably simply forgot! Many engine designs now use up to three lobes to actuate one valve!
Bruce381 - Pressure versus Flow! - This issue was discussed very well some years ago on here and Bob's comments, reports and diagrams are accurate. Flow and pressure are often confused. Pressure is as you know a product of the resistance to flow. As you know the more linear flow of (certain) synthetics (no name mentioned or I shall be judged as biased) can sometimes manifest itself as a lower OP especially at very low revs - usually under 1000rpm.This is of little consequence unless it is at/near zero!
AEHaas - you said;
"Oil pressure is insignificant in automotive lubrication. The separation (of parts) pressure of oil in a bearing is several thousand pounds verses the feed pressure of tens of pounds. Flow is however important, regardless of the pressure.
I beg to differ and this convoluted comment is conceptually wrong! How will you get flow without pressure? IMHO if you instead had said: "Flow is however important to journal bearings, almost regardless of pressure" I could have agreed!
Oil pressure (OP) is VERY SIGNIFICANT and in many valve trains it is the prime actuator in cam timing changes especially where vane drivers are used and clearances are very small to avoid noise and etc. In others it is the prime actuator for electrical phasers etc that control such things and in many other hydraulically controlled situations within the engine too
In fact OP is now much more significant than ever before in evolving engine designs.
Now, if we were talking about a Blue Flame 6 then, that's another story!
Personally I would have no hesitation in using an Approved and Listed 0w-20 synthetic oil in an engine designed for it (I did - in 1996 in a new V6 Explorer). After all in MY1941 the most common oil recommended was SAE20 for temperatures up to 90F! Lubricants have come a long way since then!
This has been a meaningful thread with many of us thinking we are correct and probably being incorrect - and vice versa!
The beauty of BITOG no doubt!
Doug
I hope personal viewpoints are not seen as absolute!
Ted - you said;
"In the same engine under the same conditions, a significantly thinner oil will reduce oil pressure..."
I would add however, that a low viscosity lubricant will also typically reduce average oil temps. The difference in dropping down one SAE grade is normally a 5F-10F reduction in equilibrium oil temps. So the difference in oil pressure between say a 5w-30 and a 5w-40 is not as great as the difference in HT/HS viscosity - which is measured at the same 150C/302F temp.
The temperature "equivalency" between each successive SAE grade in terms of oil thickness is about 20F/10C. For example, the following "mid-grade" SAE samples would yield approx the same oil pressure:
xw-20; running @ 190F
xw-30; running @ 210F
xw-40; running @ 230F
xw-50; running @ 250F
All things being equal, the thinner oil running at the lower temp is preferred. Significantly lower fluid temps will reduce wear and also extend the life of elastomeric seals/gaskets...."
Ted these broad statements appear simply theoretical, are they? - sadly, they may not be too accurate!
An engine measured to the same criteria at the Manufacturer's prescribed datum point of say 4000rpm will in all probability show the same OP, be the oil a 0w-20 or a 15w-50! Because of the pumping losses alluded to by Bruce very few if any engine makers will provide a test number for OP at idle! The engine's condition - age, wear etc, may have some effect too!
I have done a lot of oil temperature testing in live water cooled engines Ted and I have not experienced a significantly meaningful reduction in residual oil temperatures in the oil pan when using less viscous lubricants! Aircooled engines are a different story!
Certain engine components may run cooler due to more linear oil flow characteristics if the change in viscosity is also from a mineral to a synthetic lubricant! Measuring the change would however require sophisticated test equipment
Generally, other factors have a greater influence on residual oil temperature than viscosity does!
Ted - you said:
"So the difference in oil pressure between say a 5w-30 and a 5w-40 is not as great as the difference in HT/HS viscosity - which is measured at the same 150C/302F temp."
Of course not Ted although M1 0w-40 and GC 0w-30 at 100C are near, and that is the point - and this is why the ACEA's HTHS Quality rating system is so valuable
This was the fallacious comment made to Buster by "the Amsoil guy" that prompted my first post;
"....the film thickenss isn't much different between a 20wt and a 30wt oil so you won't see much of a difference."
It referred to "thickness" not OP between an SAE20 and an SAE30 lubricant!
Ted I have trouble understanding this................;
"For an engine of a given design - including oil pump output, oil temps and tolerances - oil pressure IS a direct function of viscosity, or more specifically high temp, high shear rate viscosity."
Can you explain it please? - I'm have some trouble.....?
Ted this comment is meaningful;
"Changes can also be made to the piston/ring assembly to ensure proper lubrication and better cooling. For example, some engine now use "oil jets" to help cool the pistons."
In 1960 Mercedes used at least four compression ring designs and three oil ring designs that used any of four wall contact springs. Oil jets (piston squirters) for under crown piston cooling have been around for several decades too Ted.
Many German engine makers tried oil jets in production petrol engines two decades or so ago and many discarded them quickly too. Porsche used them in MY87 on some engines but only for a small production run. Oil foaming and etc is sometimes their product! Engines that continually operate at high revs have more difficulty with oil jets than those operating at low revs
(Oil jets have been successfully used in heavy diesel engines for several decades and with somewhat lessor lubricant issues. Many of these engine have flight revs at 2200rpm and operate at around 1600rpm with 1800rpm maximum)
Ring pack position on the piston and ring design have been changed in recent times to meet emission demands. Many now run very much closer to the "fire" area than a decade ago and this has introduced many lubricant issues. The average time that oil spends in the ring pack area is directly influenced by viscosity. There is some clinical evidence that lighter oils spend less time there and lubricate better than heavier viscosities (say 0w-20 against 15w-40)
In item 1) you said:
"increasing oil pump output (increases oil pressure for a given oil viscosity)"
Ted increasing the oil pump's output will only increase the volume! It will have no effect on pressure!
In you item 3) on valve trains you did not mention lighter valve spring pressures which have significantly reduced cam wear - especially with radical cams. You probably simply forgot! Many engine designs now use up to three lobes to actuate one valve!
Bruce381 - Pressure versus Flow! - This issue was discussed very well some years ago on here and Bob's comments, reports and diagrams are accurate. Flow and pressure are often confused. Pressure is as you know a product of the resistance to flow. As you know the more linear flow of (certain) synthetics (no name mentioned or I shall be judged as biased) can sometimes manifest itself as a lower OP especially at very low revs - usually under 1000rpm.This is of little consequence unless it is at/near zero!
AEHaas - you said;
"Oil pressure is insignificant in automotive lubrication. The separation (of parts) pressure of oil in a bearing is several thousand pounds verses the feed pressure of tens of pounds. Flow is however important, regardless of the pressure.
I beg to differ and this convoluted comment is conceptually wrong! How will you get flow without pressure? IMHO if you instead had said: "Flow is however important to journal bearings, almost regardless of pressure" I could have agreed!
Oil pressure (OP) is VERY SIGNIFICANT and in many valve trains it is the prime actuator in cam timing changes especially where vane drivers are used and clearances are very small to avoid noise and etc. In others it is the prime actuator for electrical phasers etc that control such things and in many other hydraulically controlled situations within the engine too
In fact OP is now much more significant than ever before in evolving engine designs.
Now, if we were talking about a Blue Flame 6 then, that's another story!
Personally I would have no hesitation in using an Approved and Listed 0w-20 synthetic oil in an engine designed for it (I did - in 1996 in a new V6 Explorer). After all in MY1941 the most common oil recommended was SAE20 for temperatures up to 90F! Lubricants have come a long way since then!
This has been a meaningful thread with many of us thinking we are correct and probably being incorrect - and vice versa!
The beauty of BITOG no doubt!
Doug
Doug,
The difference in oil temps with thinner lubricants is very easy to explain with an analogy....
Consider the amount of energy it takes to stir a glass of hot water thickened with corn starch, vs a glass of plain water. One of the basic laws of thermodynamics states that total energy in a closed system cannot be gained or lost, only transferred. So the extra energy required to shear, stir and deform a more viscous oil goes directly into heating that fluid. This is easily observed in differentials if you compare fluid temps between an SAE 75w-90 and an SAE 75w-140 gear oil. Since there is no combustion taking place, the difference in temps is strictly an function of "intrafluid" friction as well as the rate of heat transfer. It should be mentioned that the lower the viscosity of a working fluid, the more efficient the heat transfer - in this case to the engine block and oil pan.
Of course in a thermostatically controlled, water cooled engine, MINIMUM oil temps are maintained to a certain degree. However running a significantly thicker lubricant will increase MAXIMUM oil temps under the same conditions, for the reasons stated above....
TS
The difference in oil temps with thinner lubricants is very easy to explain with an analogy....
Consider the amount of energy it takes to stir a glass of hot water thickened with corn starch, vs a glass of plain water. One of the basic laws of thermodynamics states that total energy in a closed system cannot be gained or lost, only transferred. So the extra energy required to shear, stir and deform a more viscous oil goes directly into heating that fluid. This is easily observed in differentials if you compare fluid temps between an SAE 75w-90 and an SAE 75w-140 gear oil. Since there is no combustion taking place, the difference in temps is strictly an function of "intrafluid" friction as well as the rate of heat transfer. It should be mentioned that the lower the viscosity of a working fluid, the more efficient the heat transfer - in this case to the engine block and oil pan.
Of course in a thermostatically controlled, water cooled engine, MINIMUM oil temps are maintained to a certain degree. However running a significantly thicker lubricant will increase MAXIMUM oil temps under the same conditions, for the reasons stated above....
TS
Winston - Great observation and perspective.
"..However, engines can be designed to work well with very low viscosity oils. This is primarily done by:
3) incorporating "roller" elements in the drivetrain components (eliminates sliding friction) and using more rounded cam lobe profiles, which decreases the load between the lobe and the lifter ( through increased contact area)."
Aways back when I was riding motorcycles Honda tended to run plain breaing cranks while the other makes tended to run roller bearing cranks, at least in the bigger bikes. Honda also probably had the larger design group, and often seemed to make the bikes 'good enough', compared to the over engineering that other makes often exhibited. The plain bearing cranks needed less volume of oil at higher pressures to stay reliable, but the big difference between the two was that the stock roller cranks were often used up to 2x to 3x the stock power output in modified bikes, while the plain bearing cranks seem to need lots of attention for similar loads.
Both tended to run some 40w oil as I recall.
3) incorporating "roller" elements in the drivetrain components (eliminates sliding friction) and using more rounded cam lobe profiles, which decreases the load between the lobe and the lifter ( through increased contact area)."
Aways back when I was riding motorcycles Honda tended to run plain breaing cranks while the other makes tended to run roller bearing cranks, at least in the bigger bikes. Honda also probably had the larger design group, and often seemed to make the bikes 'good enough', compared to the over engineering that other makes often exhibited. The plain bearing cranks needed less volume of oil at higher pressures to stay reliable, but the big difference between the two was that the stock roller cranks were often used up to 2x to 3x the stock power output in modified bikes, while the plain bearing cranks seem to need lots of attention for similar loads.
Both tended to run some 40w oil as I recall.
- Status
Similar threads
