ILSAC thin oil & fuel economy -- half truths

[Gokhan]

In US and Japan, when you buy motor oil, chances are that it will have the ILSAC starburst symbol on it, which implies among other things that you are getting the lowest viscosity allowable for the SAE viscosity grade you are purchasing. This is in sharp contrast to oils sold in Europe, where most oils are in the thick range. This is because there is a widespread belief that thinner the oil you use, you would get better fuel economy. This has been driven mostly by US and especially Japanese OEMs' quest for squeezing every possible fractional gain in fuel economy. The simplistic motivation is that for hydrodynamically lubricated systems, internal frictional energy loss inside the oil decreases with decreasing viscosity.

However, in reality, this is not as simple. First, only the bearings and parts of the cylinders and rings are hydrodynamically lubricated. Second, the driving conditions can put them outside the hydrodynamic region. Third, as oils are getting thinner and thinner, they are being put outside the hydrodynamic-lubrication region and into the mixed-lubrication region.

The real evidence comes from the official whitepaper of ILSAC on their fuel-economy test -- Sequence VID (link).

They studied various viscosities with and without different kinds of friction modifiers and for different operating conditions (stages), the latter of which are summarized in the first table. The results are summarized in the second table. If Oil B - Oil A is positive, B has better fuel economy than A, and if it's negative, B has worse fuel economy than A and so on. Yellow highlighting means statistically significant result.

The undisputed conclusion is that both moly (inorganic) and organic (nonmetal) friction modifiers (FM) increase the fuel economy substantially. We also know that these compounds decrease wear. So, they are win - win additives in engine oils.

What is not clear is what the viscosity does. In many cases, thinner oil actually results in worse fuel economy (with the same FM additives) and thicker oil results in better fuel economy. It also widely depends on operating conditions (stages).

In many other cases, the results are not statistically meaningful.

So, what's going on? The answer lies in the fundamental curve of lubrication -- the Stribeck curve, which plots the friction as a function of viscosity , RPM (v), and load pressure (P). There are three regions: boundary lubrication (metal-to-metal contact, highest friction, highest wear), mixed lubrication (some metal-to-metal contact with widely wearing friction), and hydrodynamic lubrication (no metal-to-metal contact at all, no wear at all, but friction increases with viscosity). What's happening is that OEMs have been trying to reduce the viscosity to the area where mixed- and hydrodynamic-lubrication regions meet, which has the lowest friction. However, since different parts of the engine has different load pressures (P), they can't all be in the same region. Moreover, the load pressures vary widely with operating conditions such as RPM and power output. Therefore, when you lower the viscosity, you also run the risk of increasing the friction and wear by going toward further leftward on the Stribeck curve, into the left part of the mixed lubrication and into the boundary lubrication, with increasing metal-to-metal contact.

Moral of the story: do not outright assume that the OEM viscosity recommendation will either result in better fuel economy or less wear. You may get neither benefit but worse fuel economy and increase wear. Your driving conditions should also be a major factor in choosing the right viscosity for your engine. (High speeds, high loads [such as uphill driving], towing, etc. require thicker oil.) Also, base-oil quality and friction modifiers affect the fuel economy greatly, with quality synthetic oils (PAO, GTL, Group III+, etc.) loaded with quality inorganic (trinuclear moly etc.) and state-of-the-art organic friction modifiers giving the best fuel economy.

 

[Panzerman]

I am with you. But it doesn't matter how you explain it, how you argue it or even if you prove it. We are a quickly becoming a doctrine society and people are going to do whatever it says in the owners manual, no matter what. That doesn't even depend if it was written for the same engine in Detroit or Paris.

 

[Pontual]

Maybe Most driving in the world is short tripping, city driving. Grocery gabbing is a real bad fuel economy condition, with 30 to 40% more fuel consuption in the first couple of miles. So EPA target this kind of usage, where most trips are made with cold to mildly warm oil, so the fuel economy with thinner oils, comes in that circunstances, with low oil consuption and wear, since the oil is thick enough. If higher average engine oil temperatures and rpm and loads, are seen, then a higher grade is advisable.

 

[bbhero]

I agree with you as well. Also I greatly appreciate all of your input on here too. Very, very insightful, information packed and smart. Thanks a lot.

 

[Jake777]

so assume higher viscosity does = better protection.... I can't seem to convince myself to use 5w30 in my older odyssey van where it calls for 5w20. I did switch to pennzoil high milage and it is a bit thicker than other typical 5w20 oils.

 

[wemay]

http://www.bobistheoilguy.com/forums/ubb...mp;#Post3929965

Originally Posted By: Solarent

...CAFE targets in the next 10 years will go to almost double the current standard. Essentially since the late 70's we've seen a roughly 40% increase in fuel economy - so 30 years to get from 20 mpg to 27 mpg and about a 30% increase in the last 5 years since 2011. The current 2025 target puts the requirement to an additional 95% increase in 10 short years. That's a huge amount and the OEM's are leaving no stone unturned.

This is where my perspective changed a little. We all know that thinner oils can directly contribute to fuel economy. It's fairly well documented that lower HTHS (with the right additives to control wear) can directly impact the fuel economy in standardized tests. We see that in the Sequence VID test as well as in other on-road tests for fuel economy improvements. So the fact that ULV lubricants directly contribute to fuel economy should come as no surprise to anyone. However there is a great deal of debate to whether this matters to the average user. I think many BITOG enthusiasts would say it doesn't and that thinner oils won't produce a significant ROI especially if the durability of the engine is sacrificed in any way.

But what about using engine oil to enable new types technologies that are by their very nature more fuel efficient? What if instead of being concerned about the fractions of a percent that a lubricant may directly contribute - we were more concerned about bringing together the tribological and rheological properties of advanced ULV fluids with new engine materials, higher power densities, increased turbo-charging, smaller displacements, more complex and efficient transmissions, axles and other parts of the powertrain? What if by their very nature a highly engineered ULV fluid allows the use of more complex technology than before because it is specifically targeted to this new tech and no longer is held to be compatible with engines that were designed 50 years ago.

Here is an example - Turbochargers are known to be tough on oil - specifically they can shear oils out of grade which then risks the durability because the minimum oil film thickness may be sacrificed. Typically the response has been to avoid ULV oils and revert to a higher HTHS or higher viscosity fluid. BUT What if the need for a heavier viscosity grade was eliminated by the use of a more robust additive layer where the oil is more about the chemistry (additive package) then it is about the carrier (base oil viscosity). All of the sudden you can turbo-charge with higher pressures because the chemical protective layers are what protects the bearings and the base oil just makes sure the chemistry gets to the right place. If that is happening then you want the base oil to be as thin as possible to reduce any kind of drag because it no longer required to maintain any kind of durability. crzy crzy shocked2 What I know you think I'm crazy right!

Consider this quote about GDI engines:
Quote:
Power densities (Average HP/liter) have increased rather steadily as engine hardware has advanced. With GDI engines now beginning to dominate the market, Lubrizol believes that engine oil formulation has reached a tipping point in which new performance demands will require higher concentrations of specific additive components when compared to ILSAC GF-5.

Set for licensing in Q1 2018, ILSAC’s GF-6 engine oil specification aims to better suit GDI engine designs by improving lubricant robustness, increasing cleanliness and durability, and producing higher fuel economy throughout the entire oil change interval. Among the many changes in the performance requirements of engine oils from GF-5 to GF-6 (all of which are proposed), the most notable will be improved viscosity, deposit, and oxidation control, as well as very significant increases in fuel economy improvement (FEI). Protecting against the occurrence of engine oil-caused, low-speed pre-ignition (LSPI), along with increased wear protection for various engine components such as timing chains and valve trains, will also be the primary focus of GF-6 lubricants.

...So for all you naysayers saying that these ULV lubes are probably going to backfire - I'd be willing to bet you are wrong. What is happening instead is a whole new range of engineering possibilities are opening up because the new oils enable these technologies rather than continuing to rely on viscosity as the primary defense against wear. When I came to the realization that thinner oils aren't about cold-flow or tiny gains in fuel economy at start up (which is what most of the threads devolve into when we talk about thick vs thin), it completely changed my perspective. This is why certain OEM's like HONDA and TOYOTA are so keen on 0W8, 0W12 and 0W16 - because they have abandoned the thought that viscosity is required for durability and are more focused on what a highly engineered fluid can do - and what new technology can be developed when we throw all the old design constraints out the window.

At the presentation I attended, the topic came up about just how thin can we go and how far can developments in the molecules really do enable new engineering designs. Lubrizol seems highly confident that they can build new additive packages that can deliver improved performance on every level, durability included.

 

[Panzerman]

It doesn't matter if this has been posted before. This is one debate that will never end and neither side will give in on. Kind of like the war between the free thinkers and the manual thumpers.

 

[doitmyself]

You state:
Quote:
The undisputed conclusion is that both moly (inorganic) and organic (nonmetal) friction modifiers (FM) increase the fuel economy substantially. We also know that these compounds decrease wear. So, they are win - win additives in engine oils.


The report conclusion states:
Quote:
While there was not enough statistical evidence to support the conclusion, it appeared that friction modified oils offer a very slight fuel economy benefit over non-friction modified oils.

The actual sequence VID ASTM test suggests that in order for SN/ILSAC GF5 oils to meet spec., they must pass the minimum fuel economy % change. The % change implies that xw-20 oils result in less fuel used than xw-30 and 10w30 oils.

See table at bottom of page one (Pass/Fail Criteria): http://www.swri.org/3pubs/brochure/ae/pdf/SequenceVID.pdf


So, while your assertion that viscosity might affect fuel usage different than what we are being fed, the final oil product must pass the sequence VID test which indicates xw-20 oils result in less fuel used than xw-30 oils in THESE lab conditions.

It seems that your "moral of the story" is simply not to assume that lab tests equate to real life results. Nothing new there. Just a bunch of "mights". But, we need to start somewhere and that is the objective of trying to develop quantifiable tests.

 

[Gokhan]

Originally Posted By: wemay
Lubrizol seems highly confident that they can build new additive packages that can deliver improved performance on every level, durability included.

Of course, they will. Selling additives packages is what they do. Have you ever seen an elixir salesman who is less than highly confident?

 

[OVERKILL]

Which is why I'm more interested in Infineum's take, as they are just a joint venture between XOM/SOPUS.

 

[wemay]

I guess time will tell. Especially with some new turbo applications being spec'd to use xw20.

And for the record, if a choice HAS to be made, i fall on the protection side. But is there really a need to choose?

 

[Shannow]

The only "zero wear" area is in full blown hydrodynamic lubrication...in times past, it was referred to as the zero wear area.

When that's gone, you are relying on additives and tribofilms, and while in that regime, you don't have "zero wear", there's always something happening. You can't develop an additive that provides an improvement over parts never coming into contact. (*)

It might be slow enough that you will never find the end point in 200,000 miles of vehicle ownership, and thus be a moot point. Like the mileage improvement that you can't prove as an end user, but in reality IS there

(*) just on that, if boundary is the new operating point, you need some BIG filtration improvements. there's no room for any particle between (say) shafts and bearings.

 

[Gokhan]

Yes, the actual Sequence VID laboratory test uses a very specific engine (2009 3.6 L V6 GM) and a very crude baseline oil (20W-30, no FM, no VII, probably high-friction mineral base stocks). The test engine never produces more than 30 HP at 2,000 RPM; so, the load pressures are very low and RPM is high enough for the given load. This means that the bearings run almost entirely in the hydrodynamic region, where thinner oil is useful in reducing friction without causing wear.

Does this apply to real-life driving conditions? Little if at all. Real-life driving conditions are more in the gray area (mixed lubrication) where the wear concerns and fuel-economy benefits of thinner oils are not clear.

 

[fdcg27]

I suspect that the OEMs making their grade recommendation were aware of all of what you've posted.
The reality is that modern engine management systems will deliver pretty good fuel economy on any grade of oil and any grade of oil will also deliver 200K+ life with most engines.
The engine still outlasts the rest of the car even if a 0W-20 grade oil is used.
There's also a little more involved than the research you're quoting depicts.
No manufacturer recommends a twenty grade with the intent of reducing fuel economy and most manufacturers know their engines very well.

 

[Rand]

Gokhan:
I have quite a few issues with your original post.

almost no 0w20 oils are near the thin end of the viscosity allowed.
Many other oils are near the midpoint in the allowed range.

Excess additives can have many bad side effects.. the most common excessive deposits.

Many times euro oils are also formulated for extremely long oil change intervals as well.. You cant just focus on one thing and call your opinion a factual conclusion.

 

[53' Stude]

Very well said sir

 

[BrocLuno]

Originally Posted By: Gokhan
Yes, the actual Sequence VID laboratory test uses a very specific engine (2009 3.6 L V6 GM) and a very crude baseline oil (20W-30, no FM, no VII, probably high-friction mineral base stocks). The test engine never produces more than 30 HP at 2,000 RPM; so, the load pressures are very low and RPM is high enough for the given load. This means that the bearings run almost entirely in the hydrodynamic region, where thinner oil is useful in reducing friction without causing wear.

Does this apply to real-life driving conditions? Little if at all. Real-life driving conditions are more in the gray area (mixed lubrication) where the wear concerns and fuel-economy benefits of thinner oils are not clear.



That is patently not a real world engine test. I use maybe somewhere between 12 and 20 hp to maintain speed limit highway speeds. But to maintain that same speed going over Donner Summit, I need way more HP... Or bucking a big headwind on I-5, or going over Tahachepi Pass, etc.

And if I only used 30 HP to accelerate, I'd be at 40 mph by the time I got to San Francisco, maybe ...

They should run that test against load to at least 100 HP and maybe a bit of lugging ... How would it work then - not very well.

5W-30 for winter and 10W-30 or 40 for summer is all I'll give them. And my vehicles are expected to last well beyond 200,000 miles before I push them out the door ... I don't have salt roads in winter, so the bodies are usually fine, even when the motors die

 

[A_Harman]

Originally Posted By: BrocLuno
Originally Posted By: Gokhan
Yes, the actual Sequence VID laboratory test uses a very specific engine (2009 3.6 L V6 GM) and a very crude baseline oil (20W-30, no FM, no VII, probably high-friction mineral base stocks). The test engine never produces more than 30 HP at 2,000 RPM; so, the load pressures are very low and RPM is high enough for the given load. This means that the bearings run almost entirely in the hydrodynamic region, where thinner oil is useful in reducing friction without causing wear.

Does this apply to real-life driving conditions? Little if at all. Real-life driving conditions are more in the gray area (mixed lubrication) where the wear concerns and fuel-economy benefits of thinner oils are not clear.



That is patently not a real world engine test. I use maybe somewhere between 12 and 20 hp to maintain speed limit highway speeds. But to maintain that same speed going over Donner Summit, I need way more HP... Or bucking a big headwind on I-5, or going over Tahachepi Pass, etc.

And if I only used 30 HP to accelerate, I'd be at 40 mph by the time I got to San Francisco, maybe ...

They should run that test against load to at least 100 HP and maybe a bit of lugging ... How would it work then - not very well.

5W-30 for winter and 10W-30 or 40 for summer is all I'll give them. And my vehicles are expected to last well beyond 200,000 miles before I push them out the door ... I don't have salt roads in winter, so the bodies are usually fine, even when the motors die

Fret not, BrocLuno, the Federal Fuel Economy test procedure is not the only test that OEM's do. Not by a long shot. But it is the test that tells what fuel economy numbers go on the window sticker, so it is important. But there are plenty of duty cycle tests that expose engines to full throttle, high power operation using the same oil used during the fuel economy test.

 

[Gokhan]

Sequence VID is an ILSAC oil test, not the US EPA fuel-economy test.

You can always design a test that favors thinner oil (by choosing the operating conditions so that the engine mostly works in the hydrodynamic-lubrication region) but does it apply to real-life driving at all? In fact, does an oil-fuel-economy test always run in a specific GM engine apply at all to a Japanese, Korean, or even a different GM engine?

Here is a better and somewhat rhetorical question: If the idea is to push the engine into the boundary between hydrodynamic lubrication and mixed lubrication by lowering the oil viscosity, why not instead keep the viscosity high but increase the load pressure (P) to push the engine into the same region? (See the Stribeck curve above.) This could easily be done by changing the bearing and cylinder geometries. It will have identical results for both fuel economy and wear considerations. Stribeck curve doesn't care whether you are changing the viscosity or bearing/cylinder geometry (which changes the load pressure P). In fact, keeping the viscosity high also lowers the friction in the valvetrain and parts of the cylinder and rings, which work in the boundary- or leftward mixed-lubrication regions; so, redesigning the geometry is a win - win option, resulting in even bigger fuel-economy gains than lowering the viscosity.

The answer to my somewhat rhetorical question: Japanese OEMs want you to believe that they are giving you the best fuel economy by employing newer technologies, such as thinner and thinner oil. On the other hand, redesigning the geometries has no such advertising value, even though it's a better way of improving the fuel economy.

 

[Merkava_4]

What is hydrodynamic lubrication ? If I set a crankshaft down into its main bearing journals and bolt the main caps on and give the crankshaft a spin by hand, is that hydrodynamic lubrication ?