Filter efficiency over time

[experimentator]

Hello,

New user here with a question. I am wondering if there is a general rule regarding filter efficiency over the length of service. I have heard the "caking theory" on this forum, that a filter increases efficiency as it gets plugged, until the bypass opens.

This SAE paper, the only one I could find and from 1965, says the opposite is true and that the cake theory is wrong (read from the last paragraph of the first page) http://papers.sae.org/650866/

Given the evidence I'm inclined to believe the SAE paper but given its date I was wondering if there were any other studies done on the topic I could review.

Thank you for any insight.

 

[Revving2Redline]

NAPA Platinums are long duration filters for 10-15k. They only shine better than Napa Gold after 10k. Even air filters do better. I had a KN Filter that I never cleaned, tapped it out and ran again. Even your HVAC at home will filter better with some dust/dirt. In general applications I think changing too soon isnt gonna do you any good and fills a wasteland prematurely. Good question.

 

[ZeeOSix]

Originally Posted By: Jonzobot
Hello,

New user here with a question. I am wondering if there is a general rule regarding filter efficiency over the length of service. I have heard the "caking theory" on this forum, that a filter increases efficiency as it gets plugged, until the bypass opens.

This SAE paper, the only one I could find and from 1965, says the opposite is true and that the cake theory is wrong (read from the last paragraph of the first page) http://papers.sae.org/650866/

Given the evidence I'm inclined to believe the SAE paper but given its date I was wondering if there were any other studies done on the topic I could review.

Thank you for any insight.


Hard to say what that SAE paper really says just from the short summary. Did you have access to the whole paper?

I emailed Purolator a few years ago, and the engineer there said the efficiency started out high, went down some with use, then went back up again towards the end of life, but not as high as initially. Said the curve would look like a hockey stick shape.

 

[experimentator]

Thank you both for the replies.

You can click "Preview Paper" on the right side of the page to read what I'm referring to.

I agree the environment and unnecessary waste should be a concern.

 

[ZeeOSix]

Originally Posted By: Jonzobot
You can click "Preview Paper" on the right side of the page to read what I'm referring to.


Interesting, as this test showed that the longer you used the filter the more engine wear was detected. Of course back then engines ran pretty dirty and sludged up easily, so maybe those filters were being used well past their nominal life span causing too much delta-p with caused debris to get pushed past the media and/or the bypass valve was opening way too much.

 

[Shannow]

I think that part of the problem is that they ran the oil the same length of time too...they also seem to be making a strawman that it's when the filter is in bypass at normal operating temperature is when the suggestion is to change the filter...not sure I've seen that recommendation.

Here's DrDave's results posted the other month on particle counts

https://bobistheoilguy.com/forums/ubbthreads.php/topics/4053638/Re:_2013_Sportster_1200_5000_m#Post4053638

Originally Posted By: DrDave
I went back to my UOA reports with particle count. These are for a Toyota Tacoma with the 2.4L.

>4u
11,383 miles 102
7507 miles 435
7453 miles 4500
6966 miles 113,507

>6u
11,383 miles 55
7507 miles 237
7453 miles 2451
6966 miles 13,710

These are cumulative miles without changing the filter. It appears that the filter really starts to look good at about 15k miles. The top numbers are the most recent particle count. I wish I knew how to post the actual report.

Dave


Originally Posted By: DrDave
Thank you:

On the report I just got a couple of days ago the numbers actually started coming back up. The >4u number jumped to 435 particles. That is the level I saw at the 20k mile point. I'm thinking that I am running in the partial bypass mode. It's time to change the filter. The filter is at the 41k mile point. I'm going to send the filter off for analysis. I'll share those when they come back.

Dave


"cake" theory works for all filters, however there's the filter bypass in the mix as well...the bypass is typically to protect the structure of the filter, and stop it either breaking and releasing stuff, or jamming particles through the filter.

When the DP gets higher, it CAN mobilise debris again, and when a partially plugged filter is in startup, with cold oil, it can drag stuff around.

 

[UltrafanUK]

Originally Posted By: Jonzobot
Hello,

New user here with a question. I am wondering if there is a general rule regarding filter efficiency over the length of service. I have heard the "caking theory" on this forum, that a filter increases efficiency as it gets plugged, until the bypass opens.

This SAE paper, the only one I could find and from 1965, says the opposite is true and that the cake theory is wrong (read from the last paragraph of the first page) http://papers.sae.org/650866/

Given the evidence I'm inclined to believe the SAE paper but given its date I was wondering if there
were any other studies done on the topic I could review.

Thank you for any insight.



OMG, that paper was sponsored by FRAM, the makers of the Orange can of death!

In very basic terms, most cheap full flow filters let less dirt through when they get dirty, BUT that is not true of synthetic media filters. With a synthetic filter there is no real difference between a new and an old one.

Some of the research results fail to point out when the air filter was changed, as they are far more efficient when dirty. How efficient the air filter is makes a big difference to the life of the oil.

Finally there is one school of oil filter gurus who are fans of longer filter change intervals and long life filters that point out the main risk of too many filter changes involves the extra wear from dry starts, oil leaks from installation finger trouble and debris that was already in the new filter.
I never ever change an oil filter until it reaches its max recommended change interval limits (2 years not 1), although that does mean I do 2 for 1 (Oil at 10K km and oil & filter at 20K km).

 

[gfh77665]

Originally Posted By: UltrafanUK


Some of the research results fail to point out when the air filter was changed, as they are far more efficient when dirty. How efficient the air filter is makes a big difference to the life of the oil.



Glad you posted that! I once read extensivly about air filter replacement. Over changing air filters actually hurt engines. No one wants to believe it, but a partially loaded air filter is MUCH better than a new one out of the box. (Of course I am not talking about one severely caked up).

 

[UltrafanUK]

One obvious point that makes a real joke of sponsored oil filter study results is the issue of when the filter goes into bypass. I missed that point and it is an important one for those doing a lot of short trips.

In the real world I get a fractionally better result from my second use of an oil filter than the first, although the figure I got from the longest extended OCI was even better in Fe per mile terms.

Trying to figure out if the results of extending or reducing an OCI are due to changes in filter efficiency or the effects of changes in the anti wear or friction layer deposited by the oil is very difficult.

 

[WellOiled]

Thanks for the post. The paper is an interesting read.

 

[kschachn]

Originally Posted By: Shannow
Here's DrDave's results posted the other month on particle counts

https://bobistheoilguy.com/forums/ubbthreads.php/topics/4053638/Re:_2013_Sportster_1200_5000_m#Post4053638


How do you make sense of that data without knowing the rate at which those particles are being generated? And what are the particles?

 

[kschachn]

Originally Posted By: UltrafanUK
In the real world I get a fractionally better result from my second use of an oil filter than the first, although the figure I got from the longest extended OCI was even better in Fe per mile terms.


How do you know the particles in the oil are iron and what is the size distribution? For the most part (greater than 3 microns or so) iron in an ICP analysis is not from the "particles" in the oil unless you do an acid digestion.

 

[camrydriver111]

Originally Posted By: ZeeOSix
Originally Posted By: Jonzobot
Hello,

New user here with a question. I am wondering if there is a general rule regarding filter efficiency over the length of service. I have heard the "caking theory" on this forum, that a filter increases efficiency as it gets plugged, until the bypass opens.

This SAE paper, the only one I could find and from 1965, says the opposite is true and that the cake theory is wrong (read from the last paragraph of the first page) http://papers.sae.org/650866/

Given the evidence I'm inclined to believe the SAE paper but given its date I was wondering if there were any other studies done on the topic I could review.

Thank you for any insight.


Hard to say what that SAE paper really says just from the short summary. Did you have access to the whole paper?

I emailed Purolator a few years ago, and the engineer there said the efficiency started out high, went down some with use, then went back up again towards the end of life, but not as high as initially. Said the curve would look like a hockey stick shape.


That 1965 paper also shows a hockey stick curve.

It sounds like two different methods were used to generate the curve.

The preview only gives you 5 out of 15 pages.

I would look at some more recent literature. By this point all this has to be well-studied and well-known.

 

[goodtimes]

Very good research in finding the article. That 1965 testing was 1000 times more scientific, unbiased, and thorough than sending a small oil stream sample into a lab to measure "particles." I for one don't want any cake on my oil filters, like them plain. There was a 10K Camry cartridge filter shown here recently, that's what I want to see, looked barely used. Air filters are a totally different ball game.

 

[Shannow]

Originally Posted By: kschachn
Originally Posted By: Shannow
Here's DrDave's results posted the other month on particle counts

https://bobistheoilguy.com/forums/ubbthreads.php/topics/4053638/Re:_2013_Sportster_1200_5000_m#Post4053638


How do you make sense of that data without knowing the rate at which those particles are being generated? And what are the particles?


Particle count goes down with filter use...pretty simple...then as the filter loads up, eventually the particle count goes up.

And new oil comes PACKED with particles.

In the industry, "particle count" is an important measurement of your hydraulic system's health. (Certain systems we have to filter the oil on the way in). In engines, Caterpillar have some pretty stringent Particle count requirements for new and used oil, without any commentary on the source or identification.

In the OP's paper, PARTICLES of 10-20um were introduced to the oil with the intent that the filter removed them...

 

[Shannow]

Originally Posted By: goodtimes
Very good research in finding the article. That 1965 testing was 1000 times more scientific, unbiased, and thorough than sending a small oil stream sample into a lab to measure "particles." I for one don't want any cake on my oil filters, like them plain. There was a 10K Camry cartridge filter shown here recently, that's what I want to see, looked barely used. Air filters are a totally different ball game.


If your engine, operating and maintenance practices are sound filters are only along for the ride...you can probably go a dozen OCIs and not fill it.

You simply don't want that stuff in your engine...but the filter is there to make sure.

 

[Jetronic]

Originally Posted By: ZeeOSix
Originally Posted By: Jonzobot
You can click "Preview Paper" on the right side of the page to read what I'm referring to.


Interesting, as this test showed that the longer you used the filter the more engine wear was detected. Of course back then engines ran pretty dirty and sludged up easily, so maybe those filters were being used well past their nominal life span causing too much delta-p with caused debris to get pushed past the media and/or the bypass valve was opening way too much.


Or, the wear particles that are too small to be caught cause wear, and extra wear particles which also are to small to be caught increase the wear rate exponentially. It occurs to me they did more a wear rate test than a filter efficiency test.

What would happen if they changed the filter, but not changed the oil (keeping the very high particle count).

 

[Jetronic]

Originally Posted By: Shannow
I think that part of the problem is that they ran the oil the same length of time too...they also seem to be making a strawman that it's when the filter is in bypass at normal operating temperature is when the suggestion is to change the filter...not sure I've seen that recommendation.



I've seen that recommendation for a Volvo engine I was responsible for (coach). The filter was to be changed when the max oil pressure (hot) dropped 15 psi from new.

 

[ZeeOSix]

Originally Posted By: Jetronic
Originally Posted By: ZeeOSix
Originally Posted By: Jonzobot
You can click "Preview Paper" on the right side of the page to read what I'm referring to.


Interesting, as this test showed that the longer you used the filter the more engine wear was detected. Of course back then engines ran pretty dirty and sludged up easily, so maybe those filters were being used well past their nominal life span causing too much delta-p with caused debris to get pushed past the media and/or the bypass valve was opening way too much.


Or, the wear particles that are too small to be caught cause wear, and extra wear particles which also are too small to be caught increase the wear rate exponentially. It occurs to me they did more a wear rate test than a filter efficiency test.

What would happen if they changed the filter, but not changed the oil (keeping the very high particle count).


Yep, good point. The filters back then were probably much less efficient than the top tier filters of today, and that would cause the majority of wear particles below the 20 micron range to just keep piling up and recirculating through the motor and never really get caught by the filter. That could be one of the main reasons they detected increased engine wear as time went on, and more and more wear particles were being produced and added to the oil as mileage increased.

It would be interesting to do a test like that today and compare a low efficiency to a a high efficiency oil filter (keeping all other factors the same) and see of the engine wear curve changed.

 

[experimentator]

Appreciate the many responses. I haven't had a chance to go through them all, but on this point in particular I think was addressed in Figure 2 of the report.

I agree it's dated and has to be looked at with a critical eye, perhaps it only applies to cellulose media as applicable to dino oil.

What I'm looking for in particular are other (preferably recent) scientific studies on the subject; everything else I think is theory and/or anecdotal. I was hoping someone could point me in that direction.