Is there any point in having an oil filter at all?

[ZeeOSix]

Originally Posted By: dnewton3

I, for one, have said for some time that filters do not control wear in an engine; the oil does. Filters only have an INdirect affect on the wear. The lube has a direct affect on wear.


If you're talking about no filter vs a filter, I'd have to disagree wholeheartedly. Remover your oil filter and blank off the mounting plate so you could still run the engine. Then dump a cup of sand particles that range from 5 to 40 microns in to the engine and drive it for 5,000 miles.

What do you think the result would be. IMO, it wouldn't be pretty.

 

[ZeeOSix]

Originally Posted By: dnewton3
OEM full flow filters are basically (what I call) "chunk catchers". Thy are not efficient at the smaller particles which cause true damage (perhaps 5-15um, depending upon which study you want to place your faith in).


I wouldn't call a filter like a PureOne that is rated at 99.9% @ 20 microns a "rock cather".

Somewhere in this forum, there is data from Purolator Tech Dept that has the efficiency down to 5 microns and it was still pretty good. Sure, the oil additive package helps suspend those particles that aren't caught by the filter, but keep in mind that ALL the oil sooner or later gets pumped over and over through the engine, and eventually some of these uncaught particles will whiz through a bearing. Hopefully, the particle size is small enough to make it through the bearing clearance and not do any damage.

 

[cp3]

Originally Posted By: ZeeOSix


The longer OCI's of today actually drive the need for a good oil filter even more than ever IMO.


Good point. I think all of the pieces working together have gotten OCIs to where they are today. Take one of those out of the equation and there is a good chance it wouldn't work. At least, maybe not well.

 

[morris]

you could do like the old speed way motor cycles did. run new oil in to the engine then dump it on the ground. clean oil ALL the time.

 

[CourierDriver]

Just remember the old VW Beetles, there you go, ocis were every 3000 miles and guess what , they were very successful for a 1600 dollar car in 1967 , hmmm. no oil filters.....................

 

[dnewton3]

Guys - the point to understand is the concept of direct versus indirect relationships.

Oil has a DIRECT relationship on the longevity of equipment. Filtration has an INDIRECT relationship. (We're talking about engines here where multi-pass lubricant filtration is used. We are not talking about single-pass medium fluids such as air).

A filter simply cannot extend the life of equipment. A filter can only increase the lifecycle of lubricants. My example, while seemingly extreme, is true. Prove to me (either in real life or with a controverted example) where it fails logic, please.

Clean oil in the sump is the goal. There is more than one road to that end destination.

I have yet to see a study that proves filtration has a DIRECT affect on equipment life in regard to lubricants. In every study I've ever read, including many from the SAE files, they always show how improved filtration results in less wear. But they NEVER address how that is an INDIRECT relationship. In fact, every one I've read shows that filtration improves the cleanliness of the oil. And that cleaner oil results in less wear. That, gentlemen, is an INDIRECT relationship, not a direct one.



If filtration is so in control, then why do we not see statistically significant divergence of wear data in UOAs? Show me a string of UOAs that conclusively proves the use of an EaO or PureOne filter did "better" than some lesser alternative over a normal OCI duration when the same fluid is used. I have data from (literally) thousands of UOAs, a great many of them from automotive applications. I can find no proof of concept to show that filtration would have a DIRECT effect on equipment lifecycle.

My point is that filtration efficiency only has a large affect when these things are true:
1) contaminants must be present in sufficient quantity (occurence rate)
2) contaminants must be of a size that is capable of being captured
3) contaminants must be of such magnitude that they have otherwise overwhelmed other means of control

That in mind, it is possible that (after a series of careful studies) one could develop an OCI using properly chosen oil that would negate the effects of wear, and yet not benefit greatly from filtration. It would be finding a "sweet spot" before the oil is loaded too heavily. But, admittedly, that spot would have it's duration affected by many variable inputs; so much so that it would be difficult to predict from load to load. Which is why it is a matter of practical application to simply have the filter present, because it greatly extends the lifecycle of the lube and allows one not to worry about over-running that point.

Jim - even the studies you and I recently viewed and discussed do NOT show filtration has a direct effect; they show an INdirect effect. In all the studies I've ever seen, the loading of oil with contaminants is the manipulated constant and wear is the measured variable result. What if they tried a "different" approach? What if the DOE was not to test the contaminat loading, but to hold the wear rate as a constant, and manipulate the methodology of contaminant control? IOW - design an experiment that says "we want a wear rate measured over 100 hours that results in 10ppm of Fe, 3ppm of Cu, 2ppm of Cr and 3ppm of Pb". Now, what means of control could generate that wear rate? You could then manipulate the sump cleanliness by means of either OCI or filtration. ONLY by that type of DOE would you be able to say conclusively what level of filtration would equate to some level of OCI duration. That response curve may be linear or parabolic; I don't know because I have NEVER seen such a study. And I don't any of you have either, because to the best of my knowledge, it's neve been done and posted to any reputable site. And even that "reversed" testing would still not show that filtration has a direct effect; it would only prove the indirect relationship from the opposite direction.

Do not confuse correlation with causation. That is a dangerous mistake to make and leads to many poor and flawed conclusions.

Filters do not extend the life of equipment when it comes to lubes. They do extend the lifecycle of the oil. That, in turn, extends the lifecycle of the equipment. But until that oil is overcome by contamination to a point where it cannot adequately control the contamination, filtration has very little (if any) true affect on lifecycle. And I would challenge anyone to point to a study that conclusively shows this as a DIRECT cause/effect relationship.

I never said it was a great idea to run without a filter. I was explaining the scientific rationale of differentiating between correlation and causation with regard to direct and indirect relationships. Perhaps I did a poor job the first time; perhaps this will help.

 

[4IsWEm]

You accept that clean oil is the goal. I agree that there are two ways to keep it clean, change it or filter it.

In today's world where most people have to travel a lot more miles than they did 40 years ago, is it really viable that oil be changed so frequently that a filter isn't needed; Isn't it a moot point?

 

[Jim Allen]

Ok, indirect vs direct. I get it. All we really need to know is dirty oil results in more wear. Clean oil results in less. Dirty oil needs to be changed sooner and clean oil lasts longer.

To me, the filtration level is the point to debate. We know the ideal contamination state of lubricating oil is 0 mg/liter but where is the practical level of contamination? Where is the contamination point where wear starts to go up and oil life is impacted by the debris? I think the industry standard (what we see on the OE) is a bit lower (higher level of acceptable contamination) than it should be right now, especially considering the excellent tools we have available... i.e. a simple spin-on filter change.

Speaking of which,back years ago, the trans tests by Eleftherakis and Khalil showed that the optimal level of trans oil contamination was under 25mg/l with around 10mg/l being ideal. The average level of contamination they encountered in transmissions (passenger cars and light trucks mostly) with about 70K mils of operation and no service was 263 mg/l. The cleaner levels were and are achievable but there was little incentive for the OEs to improve. End users, particularly in the commercial/industrial world, soon adopted enhanced filtration to extend equipment life and reduced maintenance and repair downtime. The OE has followed suit slowly and only by absolute necessity.

 

[dnewton3]

Certainly OEM response to market demands and improvements are in play.

Not that long ago, warranty coverage was a simple 3yr/36k mile. Now 100k miles is not uncommon. "Better" filters are a resonse to two things:
1) longer OCIs
2) longer warranty periods


To the OPs question, it's not unlike about any other topic here; the asnwer is "it depends".

Does one "need" a filter if they are going to OCI every 3k miles, and only drive 8k miles a year in the midwest where ferrous cancer (rust) will consume the vehicle first? No - they could probably get away with not using a filter and the frequent OCIs would make the engine last long enough to get the car to it's deathpoint.

Does one "need" a filter if they live in the arid southwest, and drive 30k miles a year, will run 7.5k mile OCIs, with the desire to get 300k+ miles life expectancy out of the rig? Yeah - probably so.

As typical of most posts, a poorly defined question results in a less than desirable set of answers; they are either unusably broad or confined in such a manner that they are almost useless to but one specific condition. Further complicating the topic is the fact that his thread title is actually a bit different that his stated question. One is "is the filter necessary" and the other is "is the wear performance reduction worth the cost of a premium filter". But even those need to be "better" defined ...

Allow me to quote myself from my former post:
"Do you NEED an oil filter? It depends upon the length of you OCI plan."

That is still a VERY accurate statement.

 

[SwampSurvivor]

So, if the oil filter helps the oil stay "clean" which will better protect the engine ...

Is it detrimental to just keep changing the oil filter and never change the oil?

 

[ZeeOSix]

Originally Posted By: dnewton3
A filter simply cannot extend the life of equipment. A filter can only increase the lifecycle of lubricants. My example, while seemingly extreme, is true. Prove to me (either in real life or with a controverted example) where it fails logic, please.


I have to disagree here if you are talking filter vs. no filter - which is the subject line of this thread. Like I said earlier, take a cup of sand with particles that range from 5 to 100 microns and dump it in the oil on an engine with no oil filter. Run the oil for 10K miles and see how the engine does. I guarantee it will have lots of wear in the bearings if there is nothing to catch those sand particles and those particles while they go through the bearings over and over for 10K miles.

 

[RiceCake]

I sincerely hate topics about oil like this one because people get two opinions that are so vastly wrong and flawed they both give me a headache watching people willingly bring them up.

The oil filter filters out debris from the oil, debris large enough to cause severe engine damage, regardless of its composition. Fine powdered metals in the oil make it basically a sort of grinding paste. Stuff like that can happen, as well as larger bits of metal. If you think oil filters are a scam because "well I never see big bits of metal in my filter material" then you're completely oblivious to the fact these materials can and do exist in your engine. Running without a filter is an invitation for any stray particle to happily recirculate through your motor and contribute to bearing and parts wear over and over and over again.

At the same time the oil filter doesn't "clean" your oil at all. You need to change your oil. If you sincerely think a filter is there to clean the oil out entirely, what about water, or fuel in the oil, or anything liquid for that matter? How about all the additives in the oil that get lost?

The only problem here is people believing the oil filter either purifies their [censored] oil completely to magic spring water, or that its useless because wear particles somehow can't exist inside an engine at large enough dimensions to cause any wear, so filters are a scam.

Did it ever dawn on anyone a filter cuts down the chance for major debris to get in your bearings and flow throughout the engine repeatedly, and the smaller particles that are not a concern simply collect in the oil and are removed when the oil is changed?

That is what your filter is for. The filter screens out particles that are definitely big enough to cause problems. They can, and will happen, in any motor. If you argue "I can't see them on my filter though" you're being dumb because these particles are microns in size. Get yourself a tube of rubbing compound and try to see the grit in there. Or toothpaste even. These can easily polish materials but the grit in them is impossible to see.

I consider this a done topic. Every time I come back to it someone has some new crackpot theory to promote it seems. Figure it out. The filter is there for a reason. Not to purify or clean your oil, its there to catch chunks of crud that may or may not get into your engine, and keep them from being forced into the bearings.

 

[k1rod]

Originally Posted By: RiceCake
I sincerely hate topics about oil like this one because people get two opinions that are so vastly wrong and flawed they both give me a headache watching people willingly bring them up.

The oil filter filters out debris from the oil, debris large enough to cause severe engine damage, regardless of its composition. Fine powdered metals in the oil make it basically a sort of grinding paste. Stuff like that can happen, as well as larger bits of metal. If you think oil filters are a scam because "well I never see big bits of metal in my filter material" then you're completely oblivious to the fact these materials can and do exist in your engine. Running without a filter is an invitation for any stray particle to happily recirculate through your motor and contribute to bearing and parts wear over and over and over again.

At the same time the oil filter doesn't "clean" your oil at all. You need to change your oil. If you sincerely think a filter is there to clean the oil out entirely, what about water, or fuel in the oil, or anything liquid for that matter? How about all the additives in the oil that get lost?

The only problem here is people believing the oil filter either purifies their [censored] oil completely to magic spring water, or that its useless because wear particles somehow can't exist inside an engine at large enough dimensions to cause any wear, so filters are a scam.

Did it ever dawn on anyone a filter cuts down the chance for major debris to get in your bearings and flow throughout the engine repeatedly, and the smaller particles that are not a concern simply collect in the oil and are removed when the oil is changed?

That is what your filter is for. The filter screens out particles that are definitely big enough to cause problems. They can, and will happen, in any motor. If you argue "I can't see them on my filter though" you're being dumb because these particles are microns in size. Get yourself a tube of rubbing compound and try to see the grit in there. Or toothpaste even. These can easily polish materials but the grit in them is impossible to see.

I consider this a done topic. Every time I come back to it someone has some new crackpot theory to promote it seems. Figure it out. The filter is there for a reason. Not to purify or clean your oil, its there to catch chunks of crud that may or may not get into your engine, and keep them from being forced into the bearings.


That's nice. And it is all very reasonable sounding. But where is the data? Without data all you have is opinions and conjecture. I personally tend to think you're correct. But so far I have never seen any proof of it. Without any real world data, both of us may be wrong. The only data point I do have on this (and it certainly qualifies as statistically insignificant), is an Arens Snowblower I ran for 25 years and gave it away still with full compression. It had no oil or air filter. And in New Hampshire with 150 ft driveway, it got a lot of use.

 

[ZeeOSix]

Originally Posted By: k1rod
That's nice. And it is all very reasonable sounding. But where is the data? Without data all you have is opinions and conjecture. I personally tend to think you're correct. But so far I have never seen any proof of it. Without any real world data, both of us may be wrong.


I think there have been a few good scientific tests/studies done on what sized particles cause the most engine wear/damage. Can't reference a link or report at this time, but maybe a Google search would find something. It's talked about once and awhile on this board. Anyway, a good oil filter will no doubt remove many of the particles that can cause engine damage. Oil filters would never have been invented if that wasn't true.

 

[ZeeOSix]

Did some searching ...
http://www.amsoil.com/techservicesbullet...gine%20wear.pdf

Conclusions
The SAE paper summarizes the test results with
the following conclusions:
“Abrasive engine wear can be substantially reduced
with an increase in filter single pass efficiency.
Compared to a 40u filter, engine wear
was reduced by 50% with 30u filtration. Likewise,
wear was reduced by 70% with 15u filtration.
“Controlling the abrasive contaminants in the
range of 2 to 22u in the lube oil is necessary for
controlling engine wear.
“The micron rating of a filter, as established in a
single pass efficiency type test, does an excellent
job in indicating the filter’s ability to remove
abrasive particles in the engine lube oil system.”

More ...
http://www.youtube.com/watch?v=phDfmD-JsKc

http://www.hepofilters.com/gm-study.html

Lots of stuff shows up if you Google "engine wear vs particle size".

 

[Virtuoso]

Originally Posted By: RiceCake
I sincerely hate topics about oil like this one because people get two opinions that are so vastly wrong and flawed they both give me a headache watching people willingly bring them up.


You certainly don't have to participate. We wouldn't want you to hurt yourself

 

[dnewton3]

Originally Posted By: ZeeOSix
Did some searching ...
http://www.amsoil.com/techservicesbullet...gine%20wear.pdf

Conclusions
The SAE paper summarizes the test results with
the following conclusions:
“Abrasive engine wear can be substantially reduced
with an increase in filter single pass efficiency.
Compared to a 40u filter, engine wear
was reduced by 50% with 30u filtration. Likewise,
wear was reduced by 70% with 15u filtration.
“Controlling the abrasive contaminants in the
range of 2 to 22u in the lube oil is necessary for
controlling engine wear.
“The micron rating of a filter, as established in a
single pass efficiency type test, does an excellent
job in indicating the filter’s ability to remove
abrasive particles in the engine lube oil system.”




The study that link above is referring to is the study by GM and is SAE #88-1825.

And the premise of that study is often misunderstood by most who read it, and is almost assuredly misunderstood by those who only read the synopsis shown above.

Here is what that study did:
They took two lab test engines and HEAVILY loaded the sumps with fine dust on an hourly basis to create a grossly exaggerated contaminant rate, thereby creating a very accelerated wear rate. They acknowledge this fact. Then they tested the ability of various filters (with ever smaller pore structure) to see how reducing contaminants with filtration would affect wear rates. I have no problem with this; they had a DOE protocol and they stuck to it.

But here is what most of you "wanna be" engineers and statisticians don't understand ...

There are three variables that generally negate wear in a typical ICE:
1) filtration
2) oil add pack
3) OCI duration

To test the effect of filtration, they had to negate the other two contributors.
- They purposely overloaded the oil with a particulate load so heavy that they overwhelmed the oil. They did so on purpose! The goal of this study was to test the effects of various filtration efficiencies at various pore sizes of filters. To achieve this, they must over-load the oil additive package to a point where they can exclude the oil's ability to deal with contamination. It is a necessity to do so. If they did NOT overload the oil, they would have multiple variables (oil add-pack and filter capability). By purposely overloading the oil add-pack, they excluded the oil's ability to deal with contamination, thereby reducing variables. If they had not purposely overwhelmed the oil, they could not discount its affects, and would not be able to attribute any change to the desired variable of filtration.
- Further, they eliminated fresh oil as a variable; they never changed oil once the test started. Again, they must eliminate variables to attribute any effects to the one controlled variable.



The problem is that Joe BITOGer does not understand the DOE protocol, because he ONLY reads that synopsis as shown in the above link. It leads one to believe that ONLY filtration can control contaminants. But that is NOT what the study set out to prove, that is NOT what the study showed, and that is NOT the right conclusion to come to!

Read these direct quotes from the study:
"Although it is recognized that there are many factors that contribute to engine wear, this paper deals only with the characteristics of abrasive contaminants and their effect on engine wear. By varying the level of filtration, different levels of sump contamination were achieved. From this, a relationship was established between filtration efficiency and engine wear rates."

"Engine wear was accelerated to minimize test time by adding 50 grams of AC Fine Test Dust to the crankcase in slurry form every hour with a total test duration of 8 hours."

"It is important to note that this analysis is used only to compare relative wear rates. Used oil analysis from engines in the field will not typically show such a clear correlation since wear metals generated between oil changes will be at much lower concentrations."

That second quote is paramount to understanding not only what is included in the summary conclusion, but what should NOT be credited in the conclusion!

That study was created ONLY to compare/contrast the effect of filtration on wear rates. Period. It does NOT address the capability of the additive-package elements to deal with wear rates or the OCI to reduce contamination. They PURPOSELY over-dosed the lube with contamination to negate the effects of the add-pack, and did not change oil so as to reduce contamination.

And then the final nail in the coffin is the last sentence of the 3rd quote ...
"Used oil analysis from engines in the field will not typically show such a clear correlation since wear metals generated between oil changes will be at much lower concentrations". IOW - real world analysis of wear in typical use will never show this kind of disparity in wear reduction. Why?
1) contamination never reaches such high levels where the performance differences are so clearly exhibited under "normal" operational conditions
2) such contamination is negated at low levels by add-packs
3) such contamination is negated at high levels by OCI duration

And our UOAs show this to be true time and time and time again. Wear metal counts in "normal" UOAs show that filter selection is a moot point. And in context, it is possible to select a OCI duration that would effectively negate filter use all together.

Further, what did the test study NOT include? It did NOT use OCI as a tested variable. IOW - they did not consider (it was not desirable to them as part of the protocol) to see how OCI duration would affect wear rates. They held the OCI as a constant (they never changed the oil after the test started). If they wanted to test how OCIs affect wear, they would have to hold filtration as a constant, and vary the OCI. If they did that as a separate test, they could then correlate an OCI duration to a specific filtration pore size and efficiency via wear rate data. But they didn't, so we'll never know the answer to that question relative to that testing protocol.


Again, I am NOT advocating the elimination of filters. But I am trying to get you folks to understand that the studies you see need to be interpreted correctly, and that the data in front of you (UOAs) shows that filtration selection is generally a moot point because it is the OIL that controls most contamination in small particulate size, and the frequency of large particulate is fairly low relative to other issues. The only time filtration makes a major shift in wear rates is when the particulate is too large for the additive package to control and/or the OCI to negate.

I realize this comes off as rude, but many of you don't understand what you are reading. It is completely incorrect to presume that ONLY filtration controls contamination, and it is completely wrong to use that SAE study to profess it to be so.

 

[Jim Allen]

Originally Posted By: RiceCake
I sincerely hate topics about oil like this one because people get two opinions that are so vastly wrong and flawed they both give me a headache watching people willingly bring them up....



RiceCake: You sure have a different interpretation of this thread than I do. While it did go pretty far into the "what if" category with regards to no filter, I don't remember see anything that gave the impression anyone thought oil filters would turn used oil into anything akin to "pure." Maybe I came the closest by stated the ideal oil contaminant level is 0 mg/l, I made no allusions to that actually being possible.

Don't know how long you've been reading BITOG. I see only 24 posts on your counter but you could be a reader for much longer. Anyway, we go pretty far into the theoretical here sometimes. Sometimes way into the ozone. Just the way it is. If you were as irritated as you came across at this light duty trip into the stratosphere, I foresee some BITOG-induced cranial explosions in your future. Watch your caffeine intake before BITIG ( : < ). BTW, there are oil filters (bypass) that can remove water and fuel.

ZeeOSix: Good references and there are others.

Dave: Yes, I see your points, and you really opened my eyes on a couple of elements of these tests, but the tests show a basic truth that cleaner oil is better and filtration is a way to keep oil cleaner longer.

Cleaner oil also has a longer service life and if you are doing statistical analysis of UOAs, that would be a good thing to factor in if possible. I'm not sure how you can factor in filtration levels into your statistical analysis anyway because most often you don't have the details on that to factor in. I think such things would require separate testing to understand fully, even if we do understand the basic concepts.

 

[dnewton3]

I think you're beginning to understand, Jim.

Go back and read many of my comment in this thread, and others.

CLEAN OIL is what controls wear. There are multiple methods to achieve any specific level of cleanliness. In addition, certain additives can reduce wear as well.

I completely agree that filtration reduces contamination; never said anything different.

But I completely abhor it when people point to that study (and others like it) and do NOT understand how true scientific studies are designed, and what conclusions can and cannot be made from the data.

As I've said many times before, clean oil is what keeps wear low. Filters do NOT reduce wear; filters clean oil. That is the difference between a DIRECT and INDIRECT relationship.

The GM/SAE study did NOT prove that filtration directly reduces wear. That study proved that filters reduce contamination; that was the direct relationship. That study also proved that cleaner oil reduces wear; that is also a direct relationship. The INDIRECT relationship is that filtration is but one of three means to reduce wear.

In effect, the GM study was actually somewhat misleading in that regard. They tested TWO separate issues; one was effects of filters on contamination, the other was the effect of contamination on wear. That is why it is an INDIRECT relationship.

And because the study purposely (and grossly) eliminated other wear control techinques (add pack and OCI), we have no fair, reasonable conclusion to draw from THAT study about how those two things reduce wear. And what they do acknowledge in a very small manner, but has a HUGE impact on how to interpret this study, is that REAL WORLD UOAs are not going to experience this kind of wear reduction because:
1) contamination never gets that bad to show filter disparity
2) OCIs help negate contamination
3) Add packs help control wear

And this is where most people go way off course because they PRESUME the study to show things that were never tested as part of the DOE protocol.

I realize I'm being somewhat rude and blunt, and I hope that I can apologize for any misgivings that I create. But folks need to quit trying to point to a few words in a synopsis and then think they understand the whole world of filtration and lubricants. It just ain't so, and if I bruise an ego or two along the way, so be it. I'd rather have the truth be fully understood. I am sorry if I hurt feelings, but I think it's more important to understand the concepts. I fully apologize to anyone, and you specifically, Jim, that I'm irritating or offending with my attitude.

 

[Jim Allen]

I wasn't irritated to any significant degree, Dave, so there's no need for an apology to me anyway. It was beginning to come across a little pedantic until I really saw the point. Pedantic is the BITOG MO, so there is (or should be) a high level of tolerance for it here. In this case, you stuck with it long enough that it struck home with this hardhead at least, so it was pedantic-with-a-purpose-and-a-result. Plus, this would seem to be one of your "Niagara Falls" topics. We all have them.

You didn't reply on my comment about factoring in the filtration levels into the UOA stats you are compiling.