VII (VM), shear, base-oil viscosity, HTHSV, friction, and wear: State of the art

Messages
5,134
Location
Paramount, California
Thread starter
What do we know about the VII (VM), shear, base-oil viscosity, HTHSV, friction, and wear in relation to each other? Recently, the combined effect of the base-oil viscosity and viscosity-index improvers [VII, also called viscosity modifier (VM)] on minimum oil-film thickness (MOFT), engine durability (wear), and fuel economy has been under my investigation. In particular, I developed a simple calculator to attack the problem, called the high-temperature, full-shear viscosity (HTFSV) calculator. The name high-temperature, full-shear viscosity (HTFSV) was something I coined. It turns out that this is state-of-the-art research because high-shear viscometers beyond the shear rate of 1,000,000 (10^6) 1/second were not available until recently because of rapid heating under high shear and the viscosity measurements under temporary shear were not possible beyond this shear rate. For those who aren't familiar with the terminology, here is a brief summary: Shear rate is the relative speed of the sliding parts separated by the oil film divided by the distance between them. Its units are inverse seconds (1/second). Shear stress (force divided by area, units of pressure) equals to the shear rate multiplied by the dynamic viscosity DV [kinematic viscosity KV (in units of cSt) times the density (in units of cP, which is pressure times time)]. Newtonian oils are such oils for which the viscosity doesn't depend on the shear rate. Non-Newtonian oils are oils for which the viscosity depends on the shear rate. In practice, monograde oils that do not contain a VII but only contain a base oil and a detergent inhibitor (DI) package are mostly Newtonian other than a small shear of the DI package and in rare cases an even a smaller shear of the base oil. Multigrade oils that do contain a VII are non-Newtonian. At low-shear rates, they are in the so-called first Newtonian phase, where the viscosity doesn't depend on the shear rate. The first Newtonian phase roughly extends to 10,000 (10^4) 1/second shear rate. The viscosity then decreases with increasing shear rate, where the VII polymer molecules go under temporary shear: "Temporary shear thinning is generally believed to result from conformational changes, such as partial alignment of the VII polymer molecules in solution under shear, that reduce the interactions between solvent/polymer and polymer/polymer molecules and thus the blend viscosity. The low-shear-rate viscosity is recovered fully after cessation of shear." The so-called high-temperature, high-shear viscosity (HTHSV) is measured at a shear rate of 1,000,000 (10^6) 1/second. Later, a second Newtonian phase is entered and the viscosity no longer decreases with increasing shear. This phase occurs roughly for shear rates greater than 10,000,000 - 100,000,000 (10^7 - 10^8) 1/second, 10 - 100 times higher than at the shear rate at which HTHSV is measured. It looks like this -- here several oils with the same base oil but different VIIs are shown. The dynamic base-oil viscosity at 120 °C is 3.30 cP, and as the shear rate increases to very high values, the viscosities of all multigrade oils containing this base oil [with no detergent-inhibitor (DI) package] and different VIIs approach this value. [Linked Image] According to the paper that will be presented in the next post here, in the second Newtonian phase, the VII fully shears and has no effect on the viscosity, the only viscosity contributions coming from the base-oil and the detergent-inhibitor (additive) package, the latter of which also goes through some temporary shear. Note that there is also the permanent shear, where the viscosity loss is never recovered after the cessation of the shear: "Permanent shear thinning or permanent viscosity loss results from the thermomechanical scission of the VM polymer chains at the high shear stresses present in lubricated contacts and is, as the name suggests, irreversible, resulting in a permanent reduction in viscosity of the lubricant." These are the shear rates encountered in an internal-combustion engine: Bearing: 100,000 - 5,000,000 (10^5 - 5 x 10^6) 1/second Piston rings: 20,000,000 (2 x 10^7) 1/second (peak rate) Valvetrain: 200,000,000 (2 x 10^8) 1/second (peak rate) Note again that HTHSV is measured and reported at 1,000,000 (1 x 10^6) 1/second shear rate, which lies somewhere near the middle of the shear-rate range for the bearings and its primary purposes are to correlate with the fuel economy and to serve as a sufficiently high viscosity for the protection of the bearings. Shear rates are taken from the following paper: Shear rates in engines and implications for lubricant design Shear rates in engines and implications for lubricant design R. I. Taylor (1) and B. R. de Kraker (2) (1) Shell Global Solutions (UK), Manchester, UK (2) Shell Global Solutions US Inc., Houston, TX, USA March 1, 2017 It's been about 25 years since HTHSV has been implemented into the SAE J300 viscosity specifications, starting in February 1991 and then gradually being established in the next several years. Others may have a more concrete historical timeline. Nevertheless, it looks like multigrade oils are still not fully understood when it comes to how their viscosity behaves in different parts of the engine, in other words how they temporarily shear. As it was said in the beginning, this was partly due to the fact that measuring the viscosity at shear rates beyond 1,000,000 1/second was not possible until recently. Hence, we have a state-of-the-art paper to discuss here. Thankfully, it's open-access and everyone can read it. In order to carry out their research, they utilized a new viscometer, which is capable of measuring the viscosity at shear rates up to 10,000,00 (10^7) 1/second. These shear rates had never been studied before. They tested about 18 different 15W-40 test oils containing different types of VIIs with or without a DI package. Here are some of their main conclusions:
  • Amount of raw (solid) VII (VM) polymers used in oils are only about 1 - 2% for all VII types except the polymethacrylate (PMA) VII, which is 5 - 6%. Note that the VII is often sold as already dissolved in an oil and the percentages for that product will be much higher. Regardless, at the end, the base oil acts as the ultimate solvent that dissolves the VII into a solution.
  • Viscosity increases linearly with VII concentration. (This is also what I assumed in my calculator.)
  • The first Newtonian phase roughly extends to 10,000 (10^4) 1/second. However, this range increases with the increasing temperature.
  • In the second Newtonian phase (at shear rates beyond 10,000,000 - 100,0000,000 (10^7 - 10^8) 1/second, the VII fully shears and the viscosity is solely due to the base oil and detergent-inhibitor (DI) package. In fact, DI also shears to some extent.
  • Regarding different VII types, perhaps the only unusual type is the polymethacrylate (PMA) type. It requires a much higher raw (solid) treat rate, which is probably not desirable. However, it differs from other VIIs in that its thickening power, which is the percentage viscosity increase over the base oil, increases with the increasing temperature, wheraas for other VIIs it's either roughly constant or slowly increasing or slowly decreasing. As a result, oils containing PMA VII can have an extremely high viscosity index (VI). I wonder if oils like TGMO 0W-20 SN use PMA VII, which would explain their high VI.
  • For all VII types, thickening power is proportional to the lack of temporary-shear stability. In other words, if a VII has a higher thickening power, it goes under more temporary shear. (This is actually good as far as my calculator is concerned because the one and only adjustable parameter in it depends on the ratio of the thickening power to the shear instability, which cancels the effect of varying VII type to some degree.)
  • Most important conclusion: HTHSV is not sufficient to describe the viscosity of an oil under shear, as the temporary shear does not stop at 1,000,000 (10^6) 1/second. Both the fuel economy and wear is influenced by both the HTHSV and base-oil viscosity at 150 °C and possibly other factors as well, such as KV100 [kinematic (low-shear) viscosity at 100 °C] and viscosity index (VI).
So, as I have been advocating, base-oil viscosity at 150 °C matters in addition to the HTHSV. Enjoy the articles (Part I and Part II). The first one characterizes the viscosity of the oils under shear. The second one studies friction (fuel economy). I found the second one somewhat inconclusive because the base oil and HTHSV are identical for different oils studied and the only remaining variables are KV100 and VI. As one would expect, a higher VI and lower KV100 results in less friction (better fuel economy). Shear thinning and hydrodynamic fri...g oils. Part I: Shear-thinning behaviour Shear thinning and hydrodynamic friction of viscosity-modifier-containing oils. Part I: Shear-thinning behaviour Nigel Marx (1), Luis Fernández (2), Francisco Barceló (2), and Hugh Spikes (1) (1) Imperial College, London, UK (2) Repsol Technology Centre, Madrid, Spain June 21, 2018 Part II: Hydrodynamic friction of viscosity-modified oils in a journal-bearing machine Part II: Hydrodynamic friction of viscosity-modified oils in a journal-bearing machine Sorin-Cristian Vladescu (1), Nigel Marx (1), Luis Fernández (2), Francisco Barceló (2), and Hugh Spikes (1) (1) Imperial College, London, UK (2) Repsol Technology Centre, Madrid, Spain September 6, 2018 By the way, long live Professor Emeritus Hugh Spikes, who has been a pioneer in tribology.
 
Messages
24,130
Location
PNW
Originally Posted by Gokhan
It looks like this -- here several oils with the same base oil but different VIIs are shown. The dynamic base-oil viscosity at 120 °C is 3.30 cP, and as the shear rate increases to very high values, the viscosities of all multigrade oils containing this base oil [with no detergent-inhibitor (DI) package] and different VIIs approach this value. [Linked Image]
The two PDF links look like pretty intense reading ... will have to digest them when time permits. Was the 120 C oil temperature before the oil was sheared, or was the resulting oil temperature while shearing always controlled to 120 C in all these cases? What's interesting is right around the 10^6 [1/sec] shear rate (the standard for the 150 C HTHS measurement), all the viscosities are nearly the same.
 
Messages
5,134
Location
Paramount, California
Thread starter
Originally Posted by ZeeOSix
Was the 120 C oil temperature before the oil was sheared, or was the resulting oil temperature while shearing always controlled to 120 C in all these cases?
The measurements between 500,000 - 10,000,000 (5 x 10^5 - 1 x 10^7) 1/second shear rate were done in an ultrashear viscometer (USV), which is able to make the measurements in only 0.1 seconds before the oil can heat due to shear and any small heating is compensated by the USV in the subsequent measurements. Then, three cycles with temperature compensation is performed and the average result is reported. Such a device wasn't available until recently and they were unable to measure the viscosity beyond 1,000,000 (10^6) 1/second shear rate, which is where the HTHSV is measured and reported at.
Originally Posted by ZeeOSix
What's interesting is right around the 10^6 [1/sec] shear rate (the standard for the 150 C HTHS measurement), all the viscosities are nearly the same.
This is by design. The VII content in all test oils was adjusted to have HTHSV = 3.70 cP, with only small deviations, and HTHSV is measured at 1,000,000 (10^6) 1/second shear rate and 150 °C. The plot is for 120 °C and the viscosity will be higher and there should be slight differences between different oils due to viscosity-index and temperature-dependent shear-stability variations. Regarding the common intersection in the plot, it appears that it shifted to a shear rate slightly lower than 1,000,000 (10^6) 1/second because the temperature is lower than 150 °C, at which all have 3.70 cP viscosity at the shear rate 1,000,000 (10^6) 1/second by design.
 
Messages
2,371
Location
pa
surely over my head, but reading anyhow hoping to learn something, as always thanks for posting above average tech!
 

JAG

Messages
5,320
Location
Fredericksburg, VA
Thanks for the links. It looks like they have some data you can use to run through your tool to see how well it predicts. I don't know all of the needed inputs so I don't know if every input is provided.
 
Messages
5,134
Location
Paramount, California
Thread starter
Originally Posted by JAG
Thanks for the links. It looks like they have some data you can use to run through your tool to see how well it predicts. I don't know all of the needed inputs so I don't know if every input is provided.
Yes, I have already run the calculations for all the test oils. I am getting a 7% root mean squared (RMS) error for the base-oil viscosity at 150 °C, which is the main output of the calculator. This is actually pretty good. One of the oils with an OCP VII has a 15% error and most are under 5%. I did recalibrate though, as my previous calculation using the Exxon Mobil blending guide had ignored the unknown effects of the DI pack but we have data without a DI pack now. With the old calibration, the error is a little higher. One thing that didn't get checked: They didn't have the density values but they had the Vogel parameters for the dynamic viscosity as a function of the temperature. So, I directly used the DV150 values resulting from these parameters. This means that neither the ASTM D341 viscosity - temperature extrapolation nor the density extrapolation got checked. It would have been nice to know the density values to see how close they are. I could in principle email the authors. So, the checks are quite encouraging and for almost all VIIs, you get good results. This is because the adjustable calibration parameter only depends on the ratio of the VII viscosity-boost rate to the VII shear-instability rate. Since they are usually proportional (the less the shear stability, the more the viscosity boost), changing the VII type does not have a drastic effect and you could live with a single calibration parameter. In the extreme case with the very shear-stable OCP VII, I only got a 15% error, which is so - so but not bad. I will discuss these results later in detail.
 

MolaKule

Staff member
Messages
21,321
Location
Iowegia - USA
Some clarification here as to the base oils needed to prepare the OCP VII:
Originally Posted by Gokhan
...Amount of raw (solid) VII (VM) polymers used in oils are only about 1 - 2% for all VII types except the polymethacrylate (PMA) VII, which is 5 - 6%. Note that the VII is often sold as already dissolved in an oil and the percentages for that product will be much higher. Regardless, at the end, the base oil acts as the ultimate solvent that dissolves the VII into a solution...
Some larger blending facilities may take the solid "bales" of OCP and grind them up, and mix them with hot oils such as Group I or II and include a small amount of Naphthenic oil or Cumene (isopropylbenzene, 2-phenylpropane, or 1-methylethyl benzene) solvents to further assist or 'dissolve' the solid chunks of ground-up OCP resulting in a liquid. That liquid is kept heated and finally mixed in a hot tank with the base oils and DI additive package to produce the finished product. Most blending facilities use the liquid forms of VIIs since they may not have a grinder and is more convenient for processing. The liquid forms of VII are already mixed with Group I or II base oils and may include a small amount of Naphthenic or Cumene solvents. So we need to be clear as to what base oils are being discussed in this context.
 
Last edited:
Messages
5,134
Location
Paramount, California
Thread starter
Originally Posted by JAG
That's overall quite good prediction on base oil viscosity! Good job.
Originally Posted by MolaKule
Some clarification here as to the base oils needed to prepare the OCP VII:
Originally Posted by Gokhan
...Amount of raw (solid) VII (VM) polymers used in oils are only about 1 - 2% for all VII types except the polymethacrylate (PMA) VII, which is 5 - 6%. Note that the VII is often sold as already dissolved in an oil and the percentages for that product will be much higher. Regardless, at the end, the base oil acts as the ultimate solvent that dissolves the VII into a solution...
Some larger blending facilities may take the solid "bales" of OCP and grind them up, and mix them with hot oils such as Group I or II and include a small amount of Naphthenic oil or Cumene (isopropylbenzene, 2-phenylpropane, or 1-methylethyl benzene) solvents to further assist or 'dissolve' the solid chunks of ground-up OCP resulting in a liquid. That liquid is kept heated and finally mixed in a hot tank with the base oils and DI additive package to produce the finished product. Most blending facilities use the liquid forms of VIIs since they may not have a grinder and is more convenient for processing. The liquid forms of VII are already mixed with Group I or II base oils and may include a small amount of Naphthenic or Cumene solvents. So we need to be clear as to what base oils are being discussed in this context.
The VIIs used in the study are all liquid forms, in other words the solid polymers dissolved in a solvent containing other base oils and chemicals. This makes me realize that I need to include the contribution of the VII solvent to the main base oil when I estimate the actual base-oil viscosity values to be compared with my calculator's output. I will attempt to estimate/calculate the true base-oil viscosity using their viscosity vs. shear rate extrapolations at high shear rates. I had taken 2.07 cP for the true base-oil viscosity when comparing to my calculator, but yes, the solvent of the VII will change things a little and hopefully it will improve the agreement with my calculator.
 
Messages
5,134
Location
Paramount, California
Thread starter
Originally Posted by Gokhan
I will attempt to estimate/calculate the true base-oil viscosity using their viscosity vs. shear rate extrapolations at high shear rates. I had taken 2.07 cP for the true base-oil viscosity when comparing to my calculator, but yes, the solvent of the VII will change things a little and hopefully it will improve the agreement with my calculator.
In fact, the two oils (#5 and #8) that my calculator seemed to overestimate the base-oil viscosity most have a much higher viscosity in the second Newtonian phase, where it should be the base-oil viscosity, if I look at Figure 8. I will try to see if I can get this 120 °C data from the plot along with the 60 °C data from another plot and then get an estimate for the true base-oil viscosity at 150 °C that also includes the VII solvent. It's certainly very encouraging that my calculator seems to be capturing that.
 
Messages
43,638
Location
'Stralia
Seriously ? The Viscosity modifiers play ZERO part in the viscosity in the second newtonian phase ? Where do they go ? They have molecular weights in the thousands, so they can't become infinitely thin planes... BTW Selby was doing 10^7 in the late '80s...
 
Messages
5,134
Location
Paramount, California
Thread starter
Shannow, The principle investigator of the study, Professor Emeritus Hugh Spikes, has multiple medals in tribology and multiple other honors in addition to his over 300 peer-reviewed papers on tribology, with ten of them having received the best-paper award. https://en.wikipedia.org/wiki/Hugh_Spikes I also saw that the first author Nigel Marx, sponsored by Shell at the time, now works at Ferrari as an F1 engine development engineer after having received his PhD. That said, to answer your questions, this is what they said about temporary shear of the VIIs (VMs), which I already quoted in my summary of the paper: "Temporary shear thinning is generally believed to result from conformational changes, such as partial alignment of the VM polymer molecules in solution under shear, that reduce the interactions between solvent/polymer and polymer/polymer molecules and thus the blend viscosity. The low shear rate viscosity is recovered fully after cessation of shear." According to what they are saying, when the shear rate becomes very high, these interactions become so weak that the VIIs no longer contribute to the viscosity. Of course, they don't go anywhere -- the loss of interaction is due to their alignment under shear. In any case, their extensive data for about 18 different oils is fully consistent with a full shear of the VIIs in the second Newtonian phase. In addition, note that the results in the paper show that the 10^6 s^-1 shear rate for the HTHSV measurement lies more or less in the middle of the non-Newtonian phase -- you seemed to have believed that it was already in the second Newtonian phase according to your exchanges with me, which can't be any further from the reality. Regarding former studies, they say: "As will be shown in this study, the temporary shear thinning behaviour of VM-containing engine oils occurs typically over the shear rate range 10^4 to 10^8 s^−1 and for many years it was not possible to reach shear rates above ca 10^6 s^−1 in high shear viscometers. This meant that the shear thinning behaviour of VM solutions could not be fully explored. Typical flow curves of engine oils up to 10^6 s^−1 can be found in [8, 9, 10, 11]. This limitation has recently been addressed by the development of the PCS ultrashear viscometer (USV) that is able to reach 10^7 s^−1. This enables almost entire shear thinning curves to be obtained for VMs representative of those used in engine oils at realistic concentrations and temperatures. Full flow curves of two engine oils obtained in this way are described by Taylor [12]. To date, this approach has not been used to explore and compare the shear thinning properties of different VMs. Such information is, however, important for designing low friction engine oils. This paper therefore describes a systematic study of the temporary shear thinning properties of a range of commercial VM blends, both in simple solutions and in fully formulated engine oils. The measured shear thinning behaviour is then used in conjunction with a conventional isothermal hydrodynamic lubrication model in the companion paper to explore how and the extent to which shear thinning influences film thickness and friction in a journal bearing." This is an excellent recent paper and a very systematic study of the VII shear, which is actually what I have been working on myself on this board the last couple of months without knowing its existence. I strongly suggest that you study it and learn a wealth of new information. My own work adds more to the knowledge. https://link.springer.com/article/10.1007/s11249-018-1039-5
 
Messages
43,638
Location
'Stralia
Only asked where they went, and if the authors clearly stated that at the ultra high shear rates they had "zero" effect.
 
Messages
5,134
Location
Paramount, California
Thread starter
Originally Posted by Shannow
Only asked where they went, and if the authors clearly stated that at the ultra high shear rates they had "zero" effect.
Yes, alignment of the VII molecules under shear reducing their interaction with each other and oil molecules is what they told as the generally believed mechanism of the temporary shear of the VII. Yes, in multiple places in both the Part I and Part II papers, they imply/say/assume the VII has no effect in the second Newtonian phase, after about 10^8 1/s. This is not to say the effect is entirely zero but either negligible or too small to measure/detect. For shear rates above 10^7 1/s, it's extrapolation. In addition, with the DI pack, the second Newtonian phase has a significantly lower viscosity than the low-shear viscosity of the base oil and DI blend (without the VII) for some oils and this is due to the aggressive temporary shear of some DI packs. I will digitize their plots at some point in order to compare them to my base-oil viscosity calculator. At that point, we will have more quantitative information on their numbers.
 
Messages
24,130
Location
PNW
Originally Posted by Gokhan
That said, to answer your questions, this is what they said about temporary shear of the VIIs (VMs), which I already quoted in my summary of the paper: "Temporary shear thinning is generally believed to result from conformational changes, such as partial alignment of the VM polymer molecules in solution under shear, that reduce the interactions between solvent/polymer and polymer/polymer molecules and thus the blend viscosity. The low shear rate viscosity is recovered fully after cessation of shear." According to what they are saying, when the shear rate becomes very high, these interactions become so weak that the VIIs no longer contribute to the viscosity. Of course, they don't go anywhere -- the loss of interaction is due to their alignment under shear.
This link talks about that in Figure 2, and the text about Figure 2 on page 39. And this was back in 2004. https://www.stle.org/images/pdf/STL...20and%20Friction_tlt%20article_Oct04.pdf
 
Messages
5,134
Location
Paramount, California
Thread starter
Here is the theory behind my HTFS formula: [Linked Image] [Linked Image] The average error in HTFS for different VII types is only 6%. However, I didn't study the error in ASTM D341 and density extrapolation, as the paper had the dynamic-viscosity values directly but not the density values. Here is the comparison of my formula to the test oils in the paper. The oil (type of the base oil, DI, and VII) is described in the first column. I couldn't compare most DI-containing oils except Oil #11 because they didn't have high-temperature data on them. Note that the DI (detergent inhibitor) package shears as well, in some case substantially. As a result, in the second Newtonian phase (ultrahigh shear rates, full temporary shear), you're left with the base oil + unsheared part of the DI + VII solvent but the VII polymer has no effect. PMA VII is unusual in that the viscosity-boost rate increases with the temperature, greatly enhancing the viscosity index (VI). Some ultrahigh-viscosity-index 0W-20 oils may be using a PMA VII. The downside is that it requires a lot more polymer than other VII types, which could increase the deposit formation. Nevertheless, the formula still works for the PMA VII, as far as HTFS is concerned. Note that the calculation of the VI, which is a secondary calculation, is not accurate, especially for VIIs with a strongly temperature-dependent viscosity-boost rate. [Linked Image] https://docs.google.com/spreadsheets/d/1gnOrQxsbymULx1s6_uBQi8zNHfJXg7lwwQpwzLSIWQI/edit?usp=sharing
 
Messages
5,134
Location
Paramount, California
Thread starter
This is the summary of how key viscosity quantities relate to the VII content: L = low-shear (HTLS) viscosity H = high-shear (HTHS) viscosity F = full-shear (HTFS) viscosity (base oil and sheared additives) V = VII content A = A_Harman index b = VII viscosity-boost rate ~ 10.5 (typical value but varies) s = VII temporary-shear rate ~ 2.0 (typical value but varies) c = b/s ~ 5.25 (typical value but varies) A = H/L V = (1/s) * (1 - A) = (1/s) * (1 - H/L) L/F - 1 = c * (1 - A) = (b/s) * (1 - A) = (b/s) * (1 - H/L) = b * V L/F - 1 represents how pronounced the non-Newtonian (S-shaped region) is. When it gets higher (the S-shaped region in the curve gets deeper), the fuel economy may improve but the wear may increase. [Linked Image] Also: 1 - A = 1 - H/L = s * V 1 - A represents how much the oil shears at 10^6 1/s, which is the shear rate at which the HTHS viscosity is measured. When it gets higher, the high-shear (HTHS) viscosity gets lower with respect to the low-shear (HTLS) viscosity. It is directly related to the VII content and how much the VII temporarily shears at the 10^6 1/s shear rate. Note that part of the additive package temporarily shears as well, appearing as a VII even if the oil is a monograde with no VII but only an additive (detergent inhibitor [DI]) package. In addition: H/F = (1 + b * V) * (1 - s * V) Here is a plot of H/F as a function of V (HTHS/base-oil viscosity @ 150 °C as a function the VII content). Note that the VII content is in arbitrary units, as the commercially solved VII comes in various solvents and also the actual treat rate depends on the VII type. Nevertheless, the range of HTHS/HTFS in the plot (1.00 - 1.80) represents the range for most commercial oils and the shape of the curve should be similar for other b and s values: [Linked Image]
 
Top