Pressure-Viscosity Coefficient vs Temperature for Base Oil Groups

[RDY4WAR]

I'm looking for information about the effect of temperature on the pressure-viscosity coefficients of different base oil groups. My understanding of this is that mineral oils will tend to have a higher P-V coefficient up until the point that it can no longer take the heat, and then synthetics start to shine. If this is correct, how does it differ between each mineral oil group I, II, and III and how does aromatic (solubility) and sulfur content effect the P-V coefficient? (if at all) The same for synthetics with groups IV and V.

 

[JAG]

There is a nice plot of the data in this thread:
https://www.bobistheoilguy.com/foru...yota-0w-20-sn-made-in-heaven#Post3303301

I expect aromatics to have a relatively high P-V coefficient based on the Group I and napthene coefficients. I doubt sulfur has a direct role because it's not a fluid. It is bound up in the various fluids and is most prevalent in the fluids that have high P-V coefficients so it does seem to have an indirect role.

Esters tend to have relatively low P-V coefficients.

 

[RDY4WAR]

Thank you, Jag. That information is invaluable.

I've read the first link to the article on machinerylubrication. It shows the P-V coefficient advantage of mineral oils fading away as temperature increases. The graph posted in that thread is only for 100*C. I would love to see how they compare at 40*C, 80*C, 120*C, and 150*C as well. Say for a drag racing engine where the oil never exceeds 180*F and often 120-150*F during the run itself, would the higher P-V coefficient of a group I/II blended oil offer benefits in wear protection at that lower running temperature? Would the improved solubility and surface affinity help with spreading ZDDP and other additives around the engine? The oxidative stability wouldn't be a concern since these engines dump the oil every 15-20 passes.

 

[JAG]

Check out this book: https://books.google.com/books?id=g...sure-viscosity%20coefficient&f=false

It's a challenging optimization problem partly because there are so many factors and many of them are competing. For example, fluids with high P-V coefficients tend to have low VI. It's easy to find a relatively viscous oil (ex. SAE 50) but it might cavitate at relatively low temperatures and high RPMs. This problem is too complicated for me to be able to solve. I can just give you some resources that may get you on your way toward beginning to solve it. At some point, testing/experience has to be done/gained because theory can only take you so far.

 

[RDY4WAR]

Thanks for the link. I'm kicking back and reading the one Gokhan posted in the other thread, and then I'll move on to that one.

Calculations and theory cannot replace real world experimentation, that I completely agree with. I'm just trying to grasp the theory at the moment. The experimentation will come later.

 

[Gokhan]

At the end of the day, it's the viscosity that determines the wear. Sure, the PVC modifies the viscosity under high contact pressure. However, given that it's usually unknown and there are other factors that determine the viscosity, such as the temperature and viscosity index, this is what I would optimize for least wear in boundary and mixed lubrication:

(1) Choose a base oil with the highest possible base-oil viscosity. 15W-40 oils are the king in this respect. Of course, there is also 20W-50.

(2) Choose a base oil with the highest possible base-oil viscosity index (BO VI).

Now, unless you're going boutique like the Amsoil ACD 10W-30/SAE 30, (1) and (2) are hard to get at the same time. In any case a 15W-40 will probably be the winner for wear protection in boundary and mixed lubrication unless you want to go boutique and find something like the Amsoil ACD.

Note that the viscosity modifiers have no effect in boundary lubrication and have little or no effect in mixed lubrication. You need to look at the base oil alone unless you are looking at the hydrodynamic lubrication in the bearings.

Patents usually don't deserve much credit as they are not supported by scientific papers. With the stated risk, here is a patent by Chevron that claims they discovered less camshaft and tappet wear with sooted oil with GTL base oil than with Group II base oil because the PVC of the GTL base oil decreases less with the increasing temperature. (It has a higher PVC index.) However, I think the reason for less wear with GTL base oil is because of my points (1) and (2) above. In other words (1) the GTL base oil had a higher KV100 viscosity and (2) the GTL base oil a higher viscosity index (BO VI). I don't think the PVC index had anything to do with it.

https://www.bobistheoilguy.com/foru...-higher-pvc-a-lot-less-wear-says-chevron

 

[Gokhan]

Note that my points (1) and (2) together mean a very high base-oil viscosity at the operating oil temperatures such as 120 °C and 150 °C. Also note that we are talking about the base-oil viscosity, not the finished-oil viscosity affected by the viscosity modifier.

The HTFSV/base-oil viscosity tables I posted in the white papers section have estimates for the dynamic base-oil viscosity at 150 °C for popular oils. In order to reduce wear, picking an oil with a higher HTFSV/base-oil viscosity may help. Of course, the additives play an important role as well.

 

[Shannow]

As I've said in quite a few threads over the years on PV coefficient...and in some threads where people have even suggested changing oil spring pressure to increase the viscosity...there's no difference at those sorts of pressures.

The Pressure viscosity co-efficient is really very relevant in gear teeth and rolling element bearings, where a first approximation is point (ball) or line (roller) contact, where any load applied leads to an infinite surface pressure, which would be instantaneous failure.

Hertzian stresses occur at these points, where both surfaces deform to provide an actual bearing load point. The roller squishes, the race cups, and the load is taken...however, at that point surface fatigue is inevitable, as the material is worked (hopefully) elastically, as plastically failure is small numbers of revolutions.

The PVC of a lubricant then matters very very much...under the extraordinary surface pressures, the lubricant becomes near crystaline (rather than squeezing out of the gap), and spreads the load over an even greater surface area, reducing surface deformation, and increasing fatigue life (roller bearing catalogues can calculate life pretty well based on lad, speed, temperature and viscosity - so does gear design - bear in mind these are 100,000 hour replacement items in industry, about 11-12 years of 24/7 operation).

Consider this pic...

bearings and pistons are in the dead flat part of the curves...50-100MPa is the compressive strength of these components, let alone with an adequate fatigue life...consider that 250 grade rebar is 250MPa yield strength, not fatigue trength, and it's still very flat...maybe lifter surfaces on flat tappets, and some links in chains...haven't looked that deep.

When you look at the viscosities shown in that curve, and apply it to a journal bearing, the viscosity increases available would increase friction, and burn it down quickly if they were there...a rolling element bearing just rolls over the wave of super thick "fluid"

Now to the machinery lubrication article...be careful with the chart...

Note that it's a normalised curve, to the 80C mineral example being given the value of unity...the table gives the actual minimum thicknesses...again, in the hertzian stress type range of EHD.

When I read the chart with 30 years of making stuff last those hundreds of thousands of hours, and failure costing $10M's, not to mention a half mil per day of downtime...

If the design point is 80C...
* any oil provides a higher film thickness at a lower temperature
* the synthetics provide a higher film thickness at temperatures above design.

Note again, that the Noria article is around rolling element and gear interfaces, and the nominal zero point/line contact...and in both they are a "rolling" contact rather than sliding, hydrodynamic contact.

(I'll post an anecdotal issue that I dealt with in the '90s that might help)

 

[Shannow]

Re the pressure viscosity co-efficient, and rolling elements...

Back in the '90s, I inherited a plant shutdown with all the engineering having already been done and parts on site...there was a 2,500hp fan that had been re-engineered to allow the 30' long shaft to be unthreaded through the bearings rather than have the impeller removed and bearings taken off either end...it was huge.

So in order to achieve this, they had to have progressively larger bearings and tapered assembly tapered bushes (removing clearance from 14" bearings by pulling up a tapered sleeve adaptor is pretty special)...

Anyway, the increase in bearing size, reduced the specific load on the bearings, which caused the rollers to skid...there was not enough surface pressure to deform the surfaces, force the PVC thickening, and force the rollers to roll...they were more to the hydrodynamic end of EHD...and slipped a little...and got ridiculously hot.

Ended up having to put a PAO oil in it just to get the oxidation resistance, and accept the higher operating temperatures (only other synthetic I used at the power station has been SHC series grease)

 

[RDY4WAR]

That's great info, Shannow. Thank you. I've got more to think about.

 

[buster]

Thanks Shannow great information.

 

[Gokhan]

Demystifying the pressure - viscosity coefficient (PVC)

n = n0 * exp(alpha * P)

n = viscosity under pressure P
n0 = viscosity under atmospheric pressure
alpha = pressure - viscosity coefficient
exp( ) = natural exponential function

Typical values for alpha = 9 - 20 1/GPa (0.00006 - 0.00014 1/psi).

The peak pressures in the bearings are somewhere around 40 MPa = 0.04 GPa. If you plug in this into the formula, the viscosity under a 40 MPa peak pressure in the bearing is multiplied by 1.43 and 2.23 for alpha = 9 1/GPa and 20 1/GPa, respectively.

Reference on peak pressures in a bearing:

Analysis of oil-film generation on the main journal bearing using a thin-film sensor and elastohydrodynamic lubrication (EHL) model
Masatsugu Inui -- Nissan Motor Co., Ltd., Makoto Kobayashi -- Nissan Motor Co., Ltd., Kensaku Oowaki -- Nissan Motor Co., Ltd., Takayoshi Furukawa -- Nissan Motor Co., Ltd., Yuji Mihara -- Tokyo City University, and Michiyasu Owashi -- Tokyo City University
SAE article 2013-01-1217
https://www.jstor.org/stable/26272802
(Need to create free account to read)

 

[Shannow]

Bear in mind how little of the bearing is exposed to those maximum pressures...by definition the bearing ends are at the pressure of the environment, as that's where the oil escapes, and the rapid increase and precipitous decrease of pressures leading up to and after the point of MOFT.

 

[Gokhan]

Originally Posted by Shannow
Bear in mind how little of the bearing is exposed to those maximum pressures...by definition the bearing ends are at the pressure of the environment, as that's where the oil escapes, and the rapid increase and precipitous decrease of pressures leading up to and after the point of MOFT.

Nice schematic, thanks.

I was telling RDY4WAR the same thing but when I think more about it, I am not convinced.

The Nissan paper is showing that the pressure is around 40 MPa over a wide crankshaft-angle range. It could be even more dramatic in RDY4WAR's racing engine. That causes a pressure-induced boost in the viscosity by 40 - 120% depending on the pressure - viscosity coefficient (PVC) of the base oil, which would change the minimum oil-film thickness (MOFT), friction, and wear.

 

[Shannow]

As I've said in many of these threads, the PVC in an engine bearings is limited by pressure...you can't have PVC without pressure.

And the pressure (bearing dimensions) is limited purely by the fatigue strength of the bearing materials, so there's a limit (and it's not very high - not 100MPa) to the surface pressures that you can allow the oil film to apply. Same with piston skirts and the like.

As to the pressure profile...by definition it HAS to be zero pressure at the side leakage points...an engine bearing has cyclic loading, around the 360 degrees of arc...there will be pressures extant all around the periphery, just not at the same time...

 

[RDY4WAR]

For me, bearings aren't a concern. Take for example an engine with solid roller lifters with high intensity cam lobes, pushing the envelope of what's physically possible, with almost a full inch of lift, with so much inertia and jerk in the valvetrain that it requires triple valve springs with near 1,000 lbs of spring force to keep it stable when the engine is spinning 9000+ rpm. That 1,000 lbs spring pressure is multiplied across a 2:1 ratio rocker arm as well. I'm not sure exactly how much pressure that translates to at the contact point of the roller on the cam lobe when jerk and inertia is factored in, but it has to be pretty high.

 

[ZeeOSix]

Originally Posted by Shannow
As I've said in quite a few threads over the years on PV coefficient...and in some threads where people have even suggested changing oil spring pressure to increase the viscosity...there's no difference at those sorts of pressures.


Changing the oil spring pressure ... ?? You talking about guys changing the pressure regulation spring in the oil pump?

 

[Shannow]

Originally Posted by ZeeOSix
Originally Posted by Shannow
As I've said in quite a few threads over the years on PV coefficient...and in some threads where people have even suggested changing oil spring pressure to increase the viscosity...there's no difference at those sorts of pressures.


Changing the oil spring pressure ... ?? You talking about guys changing the pressure regulation spring in the oil pump?


Yep...no-one in this thread, but a thin is in proponent some years ago was explaining to others that increasing the oil pressure would increase the viscosity....technically yes, approaching zero%

 

[ZeeOSix]

Originally Posted by Shannow
As I've said in many of these threads, the PVC in an engine bearings is limited by pressure...you can't have PVC without pressure.

And the pressure (bearing dimensions) is limited purely by the fatigue strength of the bearing materials, so there's a limit (and it's not very high - not 100MPa) to the surface pressures that you can allow the oil film to apply. Same with piston skirts and the like.

As to the pressure profile...by definition it HAS to be zero pressure at the side leakage points...an engine bearing has cyclic loading, around the 360 degrees of arc...there will be pressures extant all around the periphery, just not at the same time...


4-cycle engine of course, showing big end con-rod data. Don't recall the source of the graph, but gives an idea of how the oil film pressure and MOFT changes during a 2 revolution 4-cycle.

 

[ZeeOSix]

Originally Posted by Shannow
Originally Posted by ZeeOSix
Originally Posted by Shannow
As I've said in quite a few threads over the years on PV coefficient...and in some threads where people have even suggested changing oil spring pressure to increase the viscosity...there's no difference at those sorts of pressures.

Changing the oil spring pressure ... ?? You talking about guys changing the pressure regulation spring in the oil pump?

Yep...no-one in this thread, but a thin is in proponent some years ago was explaining to others that increasing the oil pressure would increase the viscosity....technically yes, approaching zero%


Yeah, considering that 100 MPa = 14,504 PSI, it's not hard to realize that increasing the oil pump pressure output 20~30 PSI isn't going to do anything. Per the charts posted, you have to get above about 250 MPa pressure region to see any (and slight at that) viscosity increase due to pressure.