Optimal viscosity: film thickness/friction/cooling

[Gokhan]

It's never easy to choose the best viscosity for your engine. Ideally you want the thickest oil film, lowest friction, and best cooling. But these three variables can be reversely correlated, making the job difficult.

First, it's important to understand the difference between SAE viscosity and apparent viscosity. Apparent viscosity is as a function of temperature, as well as a function of shear (fast sliding action). (1) Higher the temperature, less the apparent viscosity. (2) More the shear (fast sliding action), less the apparent viscosity because of the temporary shear (viscosity loss) of viscosity-index-improver polymers under mechanical shear (fast sliding action).

Minimum oil-film thickness (MOFT): More is always better to reduce wear.

MOFT increases with apparent viscosity.
MOFT increases with engine RPM.
MOFT decreases with engine power output.

Friction: Less is always better to reduce wear and increase fuel economy.

Friction increases with apparent viscosity.
Friction decreases with good friction modifiers and good detergents.

Cooling: More is always better to decrease wear.

Cooling decreases with apparent viscosity, as apparent viscosity increases friction and decreases flow.

To summarize:

High viscosity: Potentially good MOFT for good wear protection. But then the sliding surfaces run hotter due to more friction and less flow, which is bad for wear. Also, oil temperature is higher, which may decrease the apparent viscosity and the MOFT. More friction is also bad for wear. Less oil flow is also bad for oil cooling and wear.

Low viscosity: Potentially bad MOFT, with the risk of increasing engine wear. But then the sliding surfaces run cooler due to less friction and more flow, which is good for wear. Also, the oil temperature is lower, which may increase the apparent viscosity and the MOFT. Less friction is also good for wear. More oil flow is also good for oil cooling and wear.

So, how do you optimize your viscosity. Sometimes xW-20 vs. xW-30 can make a difference, as even a 20% increase in viscosity can increase the low-RPM safety margin by 20% -- oil-film breakdown happening when the RPM falls to 1200 RPM instead of RPM falling to 1500 RPM. (Oil-film breakdown happens at low RPM and high power.) But then better cooling and less friction with lower viscosity can actually increase the oil-film thickness by cooling the contact surfaces and increasing the apparent viscosity. If you have an antique engine, chances are that your clearances are so large that choosing a low-viscosity oil won't increase your oil flow and cooling but increase wear. Also, with diesel engines, you want thick viscosities because abrasive soot particles cause a lot of wear and you want as high a MOFT as possible to separate the contact surfaces from grinding against the soot particles. But with a modern gasoline engine, how do you find the optimal viscosity for best wear protection?

 

[synthetic_crazy]

Wow, interesting post!!

I think the best approach is to not "overthink" the oil in your car--even though that is hard to do here on BITOG.

I would do a cost-benefit analysis on a sheet of paper and which ever choice (thick viscosity vs. thin) comes out with the higher benefit would be my selection. Personally, I would rather have a thin oil, with better flow, cooling properties and friction than a thick oil that would possibly be better in the wear deparment, but at the expense of friction, cooling, etc.
Remember there are plethora of people out there who do not use the proper viscosity in their engines and you don't see them on the sides of the road with their engines internally eating themselves due to higher wear.

 

[abycat]

It seems i got "a thin oil that is just thick enough" stuck in my head. I dont know if it was a quote or just all the info stuffed into a sentence.

 

[gathermewool]

Great, well thought out post!

One thing to take into consideration is that the peak load is not at low RPM's, so the chances of overwhelming the wedge, rubbing at the boundary layer and causing accelerated or excessive wear may not be as much of a concern at low RPM, as compared to high RPM, high load, high temp conditions. Even with turbo-charged engines, the peak load doesn't occur until the RPM's are much higher than 1,200 RPM.

 

[CMMeadAM]

It's pretty simple really. You follow the Manufacturer recommendations printed in the owners manual. They pay
engineers good money to develop and produce good performing
long life engines along with the recommended oil to be used.
Unless you're in their league, it's rather foolish to attempt
to second guess them. Following the owners manual works pretty
well...........if they're READ !

 

[Solarent]

Well you certainly got my attention; I re read the post 3 times to make sure I was following your train of thought correctly...

It seems like where you are going is that higher viscosity oils should give better wear protection than lower viscosity oils if viscosity only was the consideration. In modern engines however, lower viscosities are being spec'd and used with confidence by many of us which could be problematic in that the minimum oil thickness may not be enough to protect from wear. So how do you balance the low viscosity equation while keeping wear to a minimum?

My answer (and yours too) is additive technology
Quote:
Friction decreases with good friction modifiers and good detergents


I think if we were to summarize the entire shift to synthetics and new oils BRIEFLY we could say something like: We've changed oil technology to allow first for mulitgrade oils, then synthetic technology and now to additive technology to provide the optimal balance between film thickness/friction/cooling. We also need to factor in environmental impact in there too which is the driving force for new developments in additive technology.

Example ZDDP is fantastic for engine wear and fluid film strength however not so great for catalytic converters or the environment. so we have to find new ways to use thinner oils (which allow for engineering advances - getting more power from smaller engines) while still protecting our engines with oil where viscosity alone can't provide the protection we need. AND it has to be good for the environment.

welcome to the conundrum that is the life of oil formulators.

 

[CATERHAM]

It's very simply, install an oil pressure gauge, their is no other way I know of to determine the optimum operational viscosity in an engine. And what is the optimum operational viscosity? Generally it coincides with the minimum hot oil pressure at elevated rev's that's specified by the manufacturer. That could be 50 psi or 60 psi or even as high a 75 psi (Porsche). And since the HTHSV rating of an oil correlates with oil pressure, it's that viscosity that you will be (usually)lowering until you get close to the optimum OP your seeking.

Secondly, if you do a lot of cold starts and short trips choose an oil with the highest VI you can find. This will provide the lowest viscosity on start-up and in effect shorten the warming up process.

 

[Bruce T]

Caterham, have you found that the optimal operational viscosity by an oil pressure gauge is typically in agreement with the Owner's Manual? I know it's often said on BITOG that the OM will favor fuel economy over engine wear, but I don't necessarily think they're mutually exclusive.

 

[JOD]

Originally Posted By: Gokhan


Minimum oil-film thickness (MOFT): More is always better to reduce wear.

MOFT increases with apparent viscosity.


This isn't necessarily true. Here's a relevant quote from this PDF Link Truth is, the viscosity both at the bearing and at the top of the piston skirt is potentially thinner with a heavier-viscosity oil.

"Bearing durability is also an area of concern, although it should be remembered that there
are three important physical effects which help ensure bearings survive. One is that typical
automotive lubricants have viscosities that are very sensitive to pressure (the commonly
used Barus equation suggests that viscosity increases exponentially with pressure), and so
as oil film thickness decrease, pressures rise, leading to higher oil viscosities, which help
support the bearing loads. Secondly, the squeeze term in Reynolds’ equation (which is often
neglected) helps ensure thicker oil films. Thirdly, and perhaps most importantly, bearings
deform when pressures are too high, again helping to sustain oil film thicknesses.

Over the years it has also been postulated that the inherent viscoelasticity of multigrade oils
bestows a load bearing benefit on bearings. Okrent34 suggested that at higher eccentricity
ratios, the elasticity of a multigrade oil (which arises due to the polymer additives in the oil)
gives a larger load bearing capacity that would be the case for an equivalent viscosity oil
that did not have any elastic behaviour. Such an effect has been confirmed experimentally by
Williamson et al35.

In our laboratory, it has been observed that in a modern gasoline engine, well designed
automotive bearings can be lubricated with oils as thin as 2.3 mPa.s without any observable
wear on either con-rod or main bearings.

The assumption that lower viscosity lubricants automatically give rise to thinner oil films in
key lubricated contacts in a gasoline engine is also open to question, particularly in the case
of piston rings. Laser Induced Fluorescence measurements have found that, in a Nissan
gasoline engine, the mid-stroke top ring oil film thickness was greater for an SAE-5W/20
lubricant than it was for an SAE-15W/40 lubricant. These effects were also observed in our
laboratory for monograde lubricants. Similar effects have been observed by S.L. Moore of
BP36. Figure 20 illustrates the observations. A qualitative explanation of such an effect
could be as follows : There are two routes by which lubricant reaches the top piston ring.
Route #1 (the “conventional” route) is that oil is left on the liner by the passage of the
preceding ring. The higher the oil viscosity, the larger will be the oil film thickness left on
the liner. Route #2 involves oil being transported to the top piston ring via the ring gaps
(such flows have been observed by Nakashima et al37), and this is thought to favour lower
viscosity lubricants. The precise balance between oil transported by the two routes will
determine whether the oil film thickness under the top ring is greater for a lower viscosity
oil or not."

Lastly, keep in mind that when discussing wear, you've disregarded the area of wear that is far more important to the every-day driver: engine oil seal wear. Valve stem seals, cam seals, main seals, valve cover seals, etc all wear due to either friction torque or excessive pressure, both of which increase as viscosity increases. It's unlikely that someone will have an engine failure due to inadequate lubrication; it several orders of magnitude more likely that they'll start having consumption problems or leaking cam/main seal on a high-mileage engine.

I'm with Caterham on the oil viscosity question; as long is it's thick enough to provide adequate oil pressure under all of the driving conditions the engine will see, it's "thick enough".

 

[JOD]

Originally Posted By: Bruce T
I know it's often said on BITOG that the OM will favor fuel economy over engine wear, but I don't necessarily think they're mutually exclusive.


I think this false dichotomy is one of the main reasons for some folk's strict adherence to the "thicker is better" mantra: "they're doing it for fuel economy, it must be bad".

 

[A_Harman]

Originally Posted By: CMMeadAM
It's pretty simple really. You follow the Manufacturer recommendations printed in the owners manual. They pay
engineers good money to develop and produce good performing
long life engines along with the recommended oil to be used.
Unless you're in their league, it's rather foolish to attempt
to second guess them. Following the owners manual works pretty
well...........if they're READ !


Drat! That's what I was going to say! Gokhan's post summarizes the balance of considerations that engineers go through when specifying an oil. When you follow the owner's manual recommendation, you are putting your trust in the engineers that developed the engine.

 

[Solarent]

Thanks for the link... it was a good read - a little dated, but still very applicable to many considerations today.

I also agree that oil pressure is a good gauge to make sure you have the adequate viscosity. However I still point out that I don't think we give additive technology enough credit in the roll it plays in controlling wear. As discussed in your comments:
Quote:
"the elasticity of a multigrade oil (which arises due to the polymer additives in the oil) give a larger load bearing capacity".


When selecting an oil for the "best wear protection" as Gokhan initially postulated viscosity alone will never get you to that "best" echelon - you can get good protection, even better protection but the best possible protection comes from good base stock and a strong additive package that gives you the elasticity, film strengths and stability to do the job specific to your engine, the work it does and your designated maintenance program.

 

[Solarent]

Originally Posted By: JOD
"they're doing it for fuel economy, it must be bad".


IMO thinner oils ARE being done for fuel economy reasons which is primarily driven by environmental impact - not just for the sake of using thinner oils, but more because we are designing smaller engines that output the same or more power then their predecessors which in turn produces tighter tolerances and higher pressures - hence thinner oils - and it's not bad - it's a good thing.

I do agree that fuel economy vs engine wear is a false dichotomy.

 

[tommygunn]

If 90% of wear occurs at start up I think your worried more about the numbers on the right side of the W, when you should be more concerned about the # on the left side of the W. I went from 5W 50 to 5W 30 and couldn't be happier! Don't try to out think the engineers who design these complex engines, they know better than we do.

 

[vinu_neuro]

Originally Posted By: CATERHAM
It's very simply, install an oil pressure gauge, their is no other way I know of to determine the optimum operational viscosity in an engine. And what is the optimum operational viscosity? Generally it coincides with the minimum hot oil pressure at elevated rev's that's specified by the manufacturer. That could be 50 psi or 60 psi or even as high a 75 psi (Porsche). And since the HTHSV rating of an oil correlates with oil pressure, it's that viscosity that you will be (usually)lowering until you get close to the optimum OP your seeking.

Secondly, if you do a lot of cold starts and short trips choose an oil with the highest VI you can find. This will provide the lowest viscosity on start-up and in effect shorten the warming up process.


HTHS lubrication at the cam/rocker and ring/wall interface is not reflected in line pressure.

 

[Jim Allen]

Ahhh... a classic, top drawer BITOG discussion!

JOD, thanks for that white paper... very interesting.

I agree that the OEM recommended viscosity it the first place to start and for most of us, it will be the right recommendation.

Like CATERHAM, I've spent a long time observing oil pressure and oil temperature gauges and making viscosity choices based on them. I'm just as willing to go thick according to indications as to go thin. For the most part, with my equipment, the indications have been to go thinner.

"As thin as possible, as thick as necessary."

 

[Jim Allen]

Originally Posted By: vinu_neuro


HTHS lubrication at the cam/rocker and ring/wall interface is not reflected in line pressure.


But the indications are that a thinner oil is better at the ring wall area in many circumstances, are they not? With the cam, it's as much boundary lubrication as anything too, so aren't the additives doing most of the heavy lifting there (flat tappets for sure)?

 

[CATERHAM]

Originally Posted By: Bruce T
Caterham, have you found that the optimal operational viscosity by an oil pressure gauge is typically in agreement with the Owner's Manual? I know it's often said on BITOG that the OM will favor fuel economy over engine wear, but I don't necessarily think they're mutually exclusive.

No, with fresh specified oil, even a 0W-20 grade, your oil pressure will be well above the optimal operational viscosity.
That difference is the safety margin to allow for any number of causes that can reduce the oil's viscosity during service including but not limited to oil shear, fuel dilution and unexpected high oil temp's.

When a manufacturer specifies a 20wt oil instead of a what was previously a 30wt, effectively all they're doing is reducing the viscosity safety margin or cushion. In reality, nothing much has really changed terms of typical engine operational viscosities, it's more a recognition that motor oils today are much more shear stable than they used to be consiquently you don't have to factor in as big a safety margin for oil shear as the engineers once needed to.

If you're into motor oil, installing an OP gauge has got to be the most eye opening thing you can do. What could be more appealing than knowing every second an engine is running what the operational viscosity is? It eliminates the need for a viscosity safety margin, but you'll undoubtedly find you've got one anyway even if you're running the lightest 0W-20 oil you can find; it's actually very difficult to run an oil that's too light.
And if something does happen that drops the oils viscosity unexpectedly such as fuel dilution from a wonky injector you'll know right away and can take immediate action.
Driving conservatively dramatically drops the viscosity demands of an engine so no engine damage will result from running an engine with oil that's lighter than optimum.

 

[CATERHAM]

Originally Posted By: Jim Allen
Originally Posted By: vinu_neuro

HTHS lubrication at the cam/rocker and ring/wall interface is not reflected in line pressure.

But the indications are that a thinner oil is better at the ring wall area in many circumstances, are they not? With the cam, it's as much boundary lubrication as anything too, so aren't the additives doing most of the heavy lifting there (flat tappets for sure)?

Exactly. The limiting factor on how light an oil you can run is just before the onset of bearing wiping and you're OP will tell you that. HTHS viscosity is sometimes referred to as "bearing viscosity" since that viscosity measure under stress closely correlates with OP.
With fresh oil, how light an oil you can run is 100% temperature related. keep the oil temp's under control and you can run very light oil indeed; even a 5wt HTHSV 1.7cP oil in a highly stress race engine under maximum load. Hence the use of 5wt qualifying oil in NASCAR for a few laps as long as the oil temp's don't exceed 180F or so.

 

[mechtech2]

Which oil? The mfrs. recommendation, or up or down one grade.
Not much to it.
And modern engines are not blowing up on the street.