MOFT, HTHSV, VII, and wear: It's complicated

Status
Not open for further replies.
Joined
Dec 29, 2010
Messages
5,889
Location
Paramount, California
There is a school of people on BITOG who think that wear is solely determined by the high-temperature, high-shear viscosity measured at 150 °C and 1,000,000 1/second shear rate (HTHSV) and the larger the HTHSV, the less the wear on the critical engine parts is.

Tribology (friction, lubrication, and wear) is very complicated and I don't like oversimplifying this very complex problem into a single parameter of the oil, namely HTHSV.

Here is an old paper from the dawn of the HTHSV idea, where they measured the minimum oil-film thickness (MOFT) and wear while varying the HTHSV and oil and viscosity-index improver (VII) type and amount. It's a very systematic study. They don't do studies like that anymore.

In conclusion, they determined that the HTHSV does not reliably predict the MOFT and wear. MOFT on the other hand directly correlates with wear and there is a minimum MOFT beyond which catastrophic wear occurs. There is also a minimum HTHSV beyond which catastrophic wear occurs but that depends on the VII type and content (but not on the additive package) and on the engine.

Some of the conclusions:

  • There is a minimum MOFT beyond which catastrophic wear occurs.
  • HTHSV is deficient in determining the MOFT and wear. MOFT can vary by more than a factor of two for different oils with the same HTHSV.
  • Increasing the viscosity beyond the required minimum actually increases the bearing wear.
  • Catastrophic wear only happens in extreme driving conditions such as wide-open throttle (WOT).
  • Only connecting-rod bearings (big-end bearings) experience catastrophic wear, not the main bearings, piston rings/cylinder liners, and the valvetrain, in extreme driving conditions such as WOT.
  • VII type is crucial in determining the MOFT and bearing wear. (I also interpret that as the VII content being crucial.)
  • KV100 is no worse than HTHSV150 in correlating with the MOFT. However, they are both deficient correlators with the MOFT and wear.
  • The data in the paper is suggesting that monogrades usually result in a higher MOFT than multigrades. (My instinct also always tells me to use an oil with the least amount of VII as possible).
  • Applicable shear rates in the bearings are 10,000,000 - 50,000,000 1/second, which is well beyond (10 to 50 times larger than) the 1,000,000 1/second the HTHSV is measured and reported at. HTHSV still decreases within this shear-rate range with the increasing shear rate, which makes HTHSV alone not sufficient to determine the MOFT and bearing wear. (I interpret this as the reason why the base-oil viscosity at 150 C plays a role in addition to HTHSV.)
  • HTHSV is a deficient predictor of an oil's load-carrying capacity and there needs to be a new ASTM Sequence "MOFT test" to characterize a commercial oil's load-carrying capacity before it can be certified. (I find this the most striking! A such ASTM test has never been implemented in the 30 years since this paper was published!)

Enjoy the paper! It's a very good controlled study. The data is exhausting to absorb.

Effect of oil rheology on journal-bearing performance: Part 4 -- Bearing durability and oil-film thickness
Published September 1, 1989 by SAE International in United States
Event: 1989 SAE International Fall Fuels and Lubricants Meeting and Exhibition
T. W. Bates -- Shell Research Ltd.,Thornton Research Centre, Chester, England
G. B. Toft -- Shell Research Ltd.,Thornton Research Centre, Chester, England
 
Last edited by a moderator:
When you plot the data in Table 3 there's more of a trend correlation between KV100, HTHS and MOFT than the summary seems to portray. Especially for testing that was done 30 years ago with whatever testing and data measuring methods they had back then, which would probably look pretty archaic in today's technology to do such studies. Have to wonder how they accurately (and how repeatable were the measurements) they measured down to 1/100th of a micron for the MOFT, and down to 1/100 of a unit on the viscosity numbers 30 years ago. Definitely an interesting test and interesting data, which seems to pretty much follow what most of us here have hashed over a few times.
 
Last edited by a moderator:
Thanks for the plots, ZeeOSix! That's very helpful.

There is definitely a correlation between the MOFT and HTHSV. However, as your plot also shows, for some HTHSV values, MOFT can vary by a factor of two, which is too large of a variation to determine the MOFT from the HTHSV alone.

As far as the accuracy of the HTHSV measurements, I don't know, but I'm guessing they did more than one measurement to make sure their results are accurate.

In order to determine the MOFT, they built a capacitor out of the bearing and connecting rod by electrically insulating the bearing from the connecting rod. Then, it's straightforward to determine the MOFT by measuring the capacitance, as the value of the capacitance depends on the MOFT (minimum oil-film thickness) that separates the "plates" of the capacitor, which are the bearing and connecting rod. They can easily measure the MOFT as a function of the crankshaft angle with a one-degree resolution using this simple built-in apparatus. They said the lead wires attached would periodically break in operation but were easy to replace.
 
Originally Posted by Gokhan
Thanks for the plots, ZeeOSix! That's very helpful.

There is definitely a correlation between the MOFT and HTHSV. However, as your plot also shows, for some HTHSV values, MOFT can vary by a factor of two, which is too large of a variation to determine the MOFT from the HTHSV alone.

If you look at your plots, which are very useful, it shows that the highest scatter in MOFT is for the MOFT 3 measurement, which has the highest RPM and lowest MOFT. Since the shear rate is the speed divided by the MOFT, MOFT 3 measurement corresponds to the highest shear rates.

This is suggesting that the simple relation between the MOFT and HTHSV -- the latter measured and reported at 1,000,000 1/second shear rate in a standard fashion -- is failing at higher shear rates, which were reported in the paper to be 10 - 50 times higher than that. To me this is suggesting that different VII types are resulting in different HTHSV values at higher shear rates and the standard HTHSV value reported at 1,000,000 1/second shear rate is not sufficient. Of course, there could be other factors in play as well.

Therefore, I stand by the importance of the base-oil viscosity at 150 C in determining wear, in addition to the HTHSV viscosity. For that reason, for a given HTHSV, I prefer oils with the smallest amount of VII possible if wear is a concern.
 
TLDR!!!
lol.gif
 
I agree that there is more of a correlation between KV100 & HTHS with MOFT than is being made out in the summary.

It really surprises me that in this day and age where more and more cars are having lower viscosity oils specified from the factory that monogrades haven't become more common place. SAE20 should be good for below freezing, how many of us really need anything more? Same goes for SAE16. My Defender which doesn't really get used during the winter could quite easily get away with a SAE30. I've said it before on this forum, I wish I could get Amsoil ACD 10w30 here for a reasonable price. I'd run it in all my vehicles without fail.
 
Table 3 footnotes show that the only difference between the MOFT 2 and MOFT 3 operating conditions was the engine RPM (2500 vs 3000).

One characteristic of journal bearings is the MOFT will increase due to RPM - ie, the hydrodynamic film wedge thickens with increased shaft speed if the viscosity inside the bearing stays constant (which it doesn't).

The oil shearing inside the bearing also increases with increased RPM, which causes more localized heating and viscosity decrease inside the bearing. On a side note, this is why higher viscosity oil is recommended for high RPM track use to help keep the MOFT to an adequate/safe level for bearing protection - you can see that in the plotted data for MOFT 3. In order to increase MOFT to keep the bearings safe you need thicker oil with more HTHSV as the engine RPM and sump temperature increases.

Since the whole trend line of the MOFT 3 data shifted down, then apparently the decrease in the oil viscosity inside the bearing due to shear heating was the larger factor over the film wedge increase due to the RPM increase.
 
Originally Posted by ZeeOSix
Table 3 footnotes show that the only difference between the MOFT 2 and MOFT 3 operating conditions was the engine RPM (2500 vs 3000).

One characteristic of journal bearings is the MOFT will increase due to RPM - ie, the hydrodynamic film wedge thickens with increased shaft speed if the viscosity inside the bearing stays constant (which it doesn't).

The oil shearing inside the bearing also increases with increased RPM, which causes more localized heating and viscosity decrease inside the bearing. On a side note, this is why higher viscosity oil is recommended for high RPM track use to help keep the MOFT to an adequate/safe level for bearing protection - you can see that in the plotted data for MOFT 3. In order to increase MOFT to keep the bearings safe you need thicker oil with more HTHSV as the engine RPM and sump temperature increases.

Since the whole trend line of the MOFT 3 data shifted down, then apparently the decrease in the oil viscosity inside the bearing due to shear heating was the larger factor over the film wedge increase due to the RPM increase.

They claim to have fixed the oil-gallery temperature to 130 °C between MOFT 2 and MOFT 3.

Yes, the torque is the same for these two cycles and you would think the latter would have a larger MOFT.

However, the way they define the MOFT is over a 720° crankshaft angle and you're getting the minimum during the exhaust cycle. (See the figure.) So, I don't know how the RPM is affecting it.
 
Another paper reaching the same conclusion as this one.

Abstract:

Bearing oil film thickness (BOFT) measurements were obtained in the front main bearing of a 3.8-liter engine. Engine speed, engine load, and oil temperature were varied to determine the effect of these parameters on BOFT. For single-grade engine oils, the minimum oil film thickness (MOFT) correlated with the Sommerfeld number (Speed x Viscosity/Load). In addition, it was also determined that the type of dispersant package used in a particular single-grade oil did not affect the oil's MOFT values. For multigrade oils (SAE 5W-30, 10W-30, 10W-40, and 15W-40) MOFT values could not be related to the Sommerfeld number with a high degree of correlation. Although high-shear, high-temperature (HSHT) viscosity was found to be important, the viscosity index (VI) improver type was also a factor in determining MOFT. The contribution of VI improver to MOFT was found to be dependent on the SAE grade and VI improver type.

https://www.jstor.org/stable/44471538

The Bearing Oil Film Thickness of Single and Multi-Grade Oils—Part 1: Experimental Results in a 3.8L Engine
Asoke K. Deysarkar
SAE Transactions
Vol. 97, Section 3: JOURNAL OF FUELS AND LUBRICANTS (1988), pp. 335-348
 
OEM 0w game = 0w20 = inherently high quality base stocks …
Works in all climates … fixed number in after action analysis …
 
Originally Posted by Gokhan
In conclusion, they determined that the HTHSV does not reliably predict the MOFT
And yet their own graph showed a nearly straight line. .... Granted, HTHS doesn't take into account the effects of high polarity surface active polymer esters, moly, tungsten, ZDDP, or any other barriers when the film gets very small. That's where the new 0w20 oils can be used, as they have low viscous drag when hydrodynamic, and the additives needed to form a barrier.




mofthths.JPG
 
Last edited:
Quote
Just a reminder that posting copyrighted material in its entirety on BITOG is not permitted.

If you find something interesting on CNN, for example, you may post a link to it. But DO NOT copy and paste the material into your post unless you have permission from the copyright holder to do so.

Posting brief quotations from copyrighted material is okay and falls within the Fair Use provisions of US copyright law.

Thanks.
 
there are real synthetic straight 30's meeting the 10w30 requirement with no viscosity improvers as advertised, of course there are NOT the fake synthetics selling for $5 a qt in 4 or 5 qt containers. this info points to the use of common 15-40 oils for great protection on the cheap unless its very cold temperatures. nice paper by the bye
 
Originally Posted by Gokhan

For single-grade engine oils, the minimum oil film thickness (MOFT) correlated with the Sommerfeld number (Speed x Viscosity/Load). In addition, it was also determined that the type of dispersant package used in a particular single-grade oil did not affect the oil's MOFT values. For multigrade oils (SAE 5W-30, 10W-30, 10W-40, and 15W-40) MOFT values could not be related to the Sommerfeld number with a high degree of correlation.


Does this include multi-grade synthetic oils that contain no VII such as 10w-30 PAO/POE oils?
 
Originally Posted by Gokhan
Originally Posted by ZeeOSix
Table 3 footnotes show that the only difference between the MOFT 2 and MOFT 3 operating conditions was the engine RPM (2500 vs 3000).

One characteristic of journal bearings is the MOFT will increase due to RPM - ie, the hydrodynamic film wedge thickens with increased shaft speed if the viscosity inside the bearing stays constant (which it doesn't).

The oil shearing inside the bearing also increases with increased RPM, which causes more localized heating and viscosity decrease inside the bearing. On a side note, this is why higher viscosity oil is recommended for high RPM track use to help keep the MOFT to an adequate/safe level for bearing protection - you can see that in the plotted data for MOFT 3. In order to increase MOFT to keep the bearings safe you need thicker oil with more HTHSV as the engine RPM and sump temperature increases.

Since the whole trend line of the MOFT 3 data shifted down, then apparently the decrease in the oil viscosity inside the bearing due to shear heating was the larger factor over the film wedge increase due to the RPM increase.

They claim to have fixed the oil-gallery temperature to 130 °C between MOFT 2 and MOFT 3.

Yes, the torque is the same for these two cycles and you would think the latter would have a larger MOFT.

However, the way they define the MOFT is over a 720° crankshaft angle and you're getting the minimum during the exhaust cycle. (See the figure.) So, I don't know how the RPM is affecting it.


I would expect the MOFT as defined in Figure 2 to occur around the same point in the crankshaft rotation as RPM increases. Yes, the MOFT occurs basically right after the power stroke shortly and about 1/3 the way into the exhaust stroke - I think I've seen similar MOFT vs crank angle graphs for 4-stoke engines that look similar to this one.

So if the only difference in operating conditions is the engine RPM, then a decrease in MOFT due to increased RPM could be caused by both engine RPM (increased inertial force into the rod) and increased localized heating/shearing of the oil in the bearing due to increased RPM.

If you look at the plotted trends of the data in Table 3, the basic conclusion (barring the data scatter on the lower end) is that MOFT increases as the HTHSV increases. Also, it shows that MOFT decreases with increased oil supply temperature (less viscosity) and MOFT decreases with increased engine RPM (causing more viscosity decrease due to more shearing inside the bearing and increased inertial loads in the rod). No surprise since the hydrodynamic wedge (MOFT) inside a journal bearing is highly dependent on the localized oil viscosity.

MOFT 1 @ 2500 RPM, 100 Nm Torque, 100 C Oil Temp
MOFT 2 @ 2500 RPM, 100 Nm Torque, 130 C Oil Temp
MOFT 3 @ 3000 RPM, 100 Nm Torque, 130 C Oil Temp

HTHSV vs MOFT Trends.JPG
 
Originally Posted by RDY4WAR
Does this include multi-grade synthetic oils that contain no VII such as 10w-30 PAO/POE oils?

Most base oils are Newtonian, including PAO, POE, and AN. The non-Newtonian behavior is mostly caused by the VII.

You will probably see the cases where the MOFT doesn't have a simple square-root dependence on the HTHSV for only non-Newtonian oils (oils with VII). The more the VII, the more the deviation from this relationship will be.
 
Originally Posted by ZeeOSix
If you look at the plotted trends of the data in Table 3, the basic conclusion (barring the data scatter on the lower end) is that MOFT increases as the HTHSV increases. Also, it shows that MOFT decreases with increased oil supply temperature (less viscosity) and MOFT decreases with increased engine RPM (causing more viscosity decrease due to more shearing inside the bearing and increased inertial loads in the rod). No surprise since the hydrodynamic wedge (MOFT) inside a journal bearing is highly dependent on the localized oil viscosity.

MOFT 1 @ 2500 RPM, 100 Nm Torque, 100 C Oil Temp
MOFT 2 @ 2500 RPM, 100 Nm Torque, 130 C Oil Temp
MOFT 3 @ 3000 RPM, 100 Nm Torque, 130 C Oil Temp

Thanks for reposting your nice plot after it was deleted! I prefer it to the one in the paper because it plots the MOFT for the three engine conditions separately.

We all agree that the MOFT is strongly correlated with the HTHSV for a given type of oil with a fixed type and content of VII.

However, the point according to the two papers posted here is that, when you change the VII content or type, for smaller MOFT values, the MOFT can change by a factor of two for the same HTHSV value. This is equivalent to changing the HTHSV by a factor of four, as the MOFT goes as the square root of the HTHSV. In other words, by simply changing the VII type or content, you can make an HTHSV = 4.0 cP 5W-40 oil behave like an HTHSV = 1.0 cP oil, which is by the way the viscosity of water at room temperature. This could be more than enough to take you into the catastrophic-wear regime.

This is why the papers are calling the HTHSV150 (and KV100) a "deficient" correlator with the MOFT for oils containing a VII.

Now, in practice, most oils seem to use an olefin copolymer (OCP) VII these days because of their resistance to form engine and turbocharger deposits. The main remaining variables are (1) the base-oil viscosity and/or VII content, (2) base-oil viscosity index, and (3) base-oil pressure - viscosity coefficient. Variable 3 is probably both hard to control and hard to measure; therefore, it's hard to know its value. Variables 1 and 2 are somewhat related. If I am concerned with connecting-rod bearing (big-end bearing) wear or even piston-ring/cylinder-liner wear, in addition to picking an oil with a high HTHSV, I would look for an oil with a high-quality base oil (high base-oil viscosity index) and low VII content.

Looking at the oil in the paper for which catastrophic wear happened, I understand why. It's a Group I SAE 20 monograde with HTHSV150 = 2.0 cP and VI = 100. That would translate into an HTHSV170 = 1.6 cP at 170 °C, which is probably a typical bearing temperature during extended WOT. Other oils either have a higher HTHSV150 or a higher VI because they are a multigrade. That also cautions one against relying on Group I monogrades as they have a very low VI and the HTHSV can fall dramatically at extreme temperatures, especially if the starting HTHSV is small (in the modern day, no less than 2.6 cP is allowed for SAE 20).

Now, here is the million dollar question: Can an almost-VII-free HTHSV = 3.0 cP 5W-30 oil or even an almost-VII-free HTHSV = 2.6 cP 5W-20 oil made from a high-VI base oil protect against catastrophic wear better than an HTHSV = 4.0 cP oil with a very high VII content and a base-oil with a so - so viscosity index? I think the answer is that it's quite possible. In any case, the paper in the original post is suggesting that you would get less wear and better fuel economy with oils having a minimum HTHSV = 2.5 cP than with oils with a higher HTHSV. That pretty much puts any modern SAE 0W-20 or SAE 5W-20 oil as an optimal oil for any modern car, except those with very high power outputs and/or high engine revs or diesel engines that have naturally high torque and low rev, resulting in a smaller MOFT.

It is very complicated indeed. That's probably why many people don't even bother to think about it and only consider the HTHSV as a protection against wear.
 
The other problem to remember is that we have people with blogs that have very odd ways of measuring MOFT or finding a proxy thereof. HTHS is generally fairly easy to predict within grade, particularly in a 30 grade, whereas MOFT is little more problematic.
 
Originally Posted by Gokhan
Now, here is the million dollar question: Can an almost-VII-free HTHSV = 3.0 cP 5W-30 oil or even an almost-VII-free HTHSV = 2.6 cP 5W-20 oil made from a high-VI base oil protect against catastrophic wear better than an HTHSV = 4.0 cP oil with a very high VII content and a base-oil with a so - so viscosity index?


So one reason I chose Valvoline Advanced 5W-30 full synthetic is because the HTHSV is 3.2 cP, KV100 is 10.2 cSt, VI = 158 and Noack 9.3%. The VI of 158 and KV100 of 10.2 seems to be on the low end for most 5W-30 full synthetics oils that I compared (ie, Castrol, Pennzoil, Mobil 1). So how does Valvoline Advanced 5W-30 get a relatively high HTSHV with a relatively low VI and KV100 compared to those other 3 brands which all have a HTHSV less than 3.2 and a VI greater than 158.
 
Status
Not open for further replies.
Back
Top