What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Filter Test Results

DanH

Legacy Member
Mentor
The results are in.

Previous organizational thread here: https://vansairforce.net/threads/oil-filter-testing.223689/

Please support and thank our many individual donors, and our two vendor sponsors, Barrett Precision Engines and CamGuard.

There are many interesting details, and research continues. More in due course, but for now, a short synopsis...

The world standard for comparative filter test is ISO 4548-12. The 4548-12 test installation circulates a standard fluid (MIL-H-5606, here Aeroshell 4) at 100F and 6 gallons per minute while steadily feeding ISO 12103-1 test dust into the flow. A laser-based device counts particles while determining the size of each particle. The fluid then enters the filter. Downstream, another counter records the number and size of the particles which passed through the filter. For each particle size from 4 through 40 µm (microns), efficiency ratio = 1- (particles out/particles in).

As the test progresses, the filter accumulates particles, and the differential pressure measured across the element slowly rises. The rise begins slowly, but as the filter nears capacity, the rate of rise steepens significantly. The test is terminated at some specified deltaP, here being 10 psi. The total capacity of the filter is recorded in grams.

The tested filters were a Tempest 48108, a Champion 48108, a Wix 51515 (aka a NAPA 1515), a K&P S-15, and a Challenger.

Average efficiencies from inception to 10 psi deltaP, contaminent retained, and time to termination:

ScreenHunter_2187 Mar. 31 19.35.jpg

ISO 12103-1 medium test dust particle size distribution. Although the filter may capture all or none of the sizes, the 4548-12 test system is calibrated for counting the shaded range:

ISO 12103-1 Medium Test Dust.jpg

Dust.jpg

What is good enough?

The agreed standard for piston aviation oil filters is SAE ARP 1400- Rev B. The standard covers all aspects of filter construction and performance (example, burst pressure), but the key point here is filtration efficiency at two particle sizes. ARP 1400-B specifies the SAE J1858 multipass proceedure for measuring filtration, essentially identical to ISO-4548-12. The minimum required performance is >75% at 25 µm, and >90% at 40 µm, with 15 grams retention, and a terminal deltaP of 8 psi. Neither Tempest or Champion appears to meet the 25 µm requirement.

The viscosity of 50W AD aviation oil at 180F (+ or - 5) is roughly 32 cSt, vs 15 cSt for the MIL-H-5606 fluid used for ISO 4548-12 and J1658 tests. As a general rule, higher viscosity would be expected to reduce capture efficiency.

In general, we want particle size to be less than oil film thickness. For now, two textbook illustrations. The first is from Heywood's Internal Combustion Engine Fundamentals:

Heywood.jpg

And this from John Schwaner's Sky Ranch Engineering Manual:

Schwaner.jpg
 
Last edited:
So what this tells us is the only filter that actually meets the ”good enough” standard is the Wix and that the reusable filters completely suck?

wow. I’m actually kind of pissed off.

if the two reusable filters arent catching any substantial amount of particle, what is causing the delta P to rise and apparently rise very quickly given the very short test duration for these filters?
 
Last edited:
So what this tells us is the only filter that actually meets the ”good enough” standard is the Wix

True. However, there is a practical caveat. You're looking at average performance from beginning of test to termination at 10 psi deltaP. Depth media performs better when clean than later in the test when it is loaded with contaminent. The Champion and the Tempest look better if we consider some arbitrary period near the beginning of the test and ignore the efficiency values taken when loaded near the end of the test. As a practical matter, it means a 25 hour filter cleans better than a 50 hour filter, which is better than a 75 hour filter. I'll have more on that later.

and that the reusable filters completely suck?

Yep.

Well, not completely. They are catching the largest particles in the (roughly) 20% of test dust larger than 40 µm, which is why they clogged up rather quickly. Look at the retained capacity, test time to reach 10 psi, and the sizes found A3 Medium media.

Dust Bottle.jpg
 
Last edited:
Why would the reusable filters trap some of the small particles (5um ones) but not the larger particles (40um ones)?
 
Why would the reusable filters trap some of the small particles (5um ones) but not the larger particles (40um ones)?

Most likely because the smallest particles wedge in the intersections of the wire weave. It will require some magnification of a used filter to be sure. The test filters, with contaminants, are being returned to me.

Wire weave.jpg
 
Dan, your line
Depth media performs better when clean than later in the test when it is loaded with contaminant.
could use a little context. I work with water filtration, so while the contaminants and viscosity differ, the filtration aspects are quite similar. In general, I'd encourage interested people to dig a little into depth versus surface filtration.

Depth filtration tends to achieve progressively higher removal efficiency as it develops a surface cake of contaminants. The trapped materials make it yet more difficult for the fluid to reach the "clean" side of the media. The cost for this improved performance is higher pressure drop. Depth filters don't typically have a defined minimum particle size that'll pass, as your testing demonstrates. Whether any given particle will be trapped depends on myriad factors beyond simply particle size, such as particle shape, differential pressure, filter uniformity, etc. For this reason, they're customarily given nominal filtration ratings, though these may or may not have a particular test methodology associated with the claim.

Surface filtration, like woven screens, tends towards absolute filtration. The openings are highly uniform. Still, large particles may pass if they're not spherical. Threads, like metal fragments for examples, could be millimeters long and pass through a fine mesh if they happen to hit just right (or wrong, as it were). Surface filters also improve in removal efficiency with the development of a surface cake, but unlike depth filters they tend to have a much lower solids storage capacity. A particle either stops at the surface, or it doesn't. Once a given hole through the mesh gets blocked, pressure rises as the remaining open area decreases. Same basic trend as a depth filter, but can be dramatically more pronounced.

It's easy to visualize the contrast. Imagine shaking dirt through a metal screen versus shaking that same dirt through a suspended piece of shag carpet. Both will allow very small particles to pass, but the shag will hold a lot more dirt before it's completely blocked as compared to a screen. This assumes that the screen and the shag have the same basic opening sizes in the weave.

What's not visible in your removal table is that the particles larger than 40 microns rapidly clog the screen filters, leading to a rapid pressure rise. Sure, it's only about 15-18% of the dust by volume, but once the dust particles are bigger than the mesh, they're not going through and plugging occurs incredibly fast. These same particles have a lot more media volume available in the disposable filters, hence the longer run times.
 
At first glance, the numbers for the reusable filters are disappointing, to say the least.
 
That settles it. I for one will ONLY allow dust that meets ISO standard 12103-1 into my engine (sarcasm font off).

Been involved in plenty of similar testing in my past. Easy to poke a hole in any testing process but this one is thorough, widely recognized/accepted, The main thing is, it makes products comparable. Much thx to all involved.
 
Dan, your line

could use a little context. I work with water filtration, so while the contaminants and viscosity differ, the filtration aspects are quite similar. In general, I'd encourage interested people to dig a little into depth versus surface filtration.

Good coverage Dave.

Depth filtration tends to achieve progressively higher removal efficiency as it develops a surface cake of contaminants. The trapped materials make it yet more difficult for the fluid to reach the "clean" side of the media. The cost for this improved performance is higher pressure drop.

That was my assumption until I saw the complete test results, and then had a nice conversation with the lab director. He tells me he has seen it go both ways, depending on several factors. Actually, efficiency for some of the filters in this test went down initially, then rose a bit toward the end. None were better after loading than they were when clean.

The critical difference between water filtration and engine oil filtration (which includes the 4548-12 equipment) is probably that the oil systems use constant displacement pumps...the flow rate does not change much regardless of downstream resistance to flow. As the resistance to flow goes up, upstream pressure rises, as does velocity through the media, so it carries more contaminent to the downstream side. My understand of water systems (which is not much) suggests they are essentially constant pressure on the upstream side of a media, so as the media loads, flow rate drops.

Anyway, here's an example in numbers, the data table for a single particle size, Champion filter.

Champion 25 micron data table.jpg

BTW, note the caveat I wrote in a previous post regarding average efficiency vs efficiency in the early life.

What's not visible in your removal table is that the particles larger than 40 microns rapidly clog the screen filters, leading to a rapid pressure rise. Sure, it's only about 15-18% of the dust by volume, but once the dust particles are bigger than the mesh, they're not going through and plugging occurs incredibly fast. These same particles have a lot more media volume available in the disposable filters, hence the longer run times.

Agree!
 
Last edited:
The results really aren't that surprising to me. I'm as far from an engineer as you can get but I never understood why anybody would use a re-usable filter in any engine and especially in an airplane engine. That has always been a mystery to me and the test results just confirm what I always suspected.

Thank you to all who helped support this testing.
 
These results are consistent with my personal experience. I have an oil analysis done by Blackstone after every oil change. I switched to one of these reusable filters. The Blackstone report was very significantly worse with the K&P. I immediately switched back to the Tempest that I had been using and my numbers went back to their normal trends prior to the K&P experiment.

Thanks for the confirmation!
 
Wall Street Journal is reporting Champion Aerospace, LLC of Liberty, South Carolina is seeing a sudden spike in oil filter sales. A spokesperson for the company said they were completely caught off guard and have no explanation for the unprecedented demand. :)
 
An earlier comment caught my eye, one which highlighted the performance of the Wix filter. If the numbers seem to show it to be a very good filter, why wouldn't we all be flocking to the Wix unit? I know many will say, "I only fly genuine aircraft filters" - I understand this bias. From a purely technical standpoint, what part of the Wix 51515's performance (either mechanically or in its filtration function) makes it less than the best choice?

I very much appreciate seeing these results as presented so far.
 
Excellent report and thanks Dan and everyone else. So, this goes along with the flakey Xring that I don't like and the problems a few of the guys have had. I still have a brand new S15 sitting in its tube for anyone that wants.😉 Thankfully I still have 2 Champ filters to use.

I can't imagine that the old style Lycoming screen was any better.
 
That was my assumption until I saw the complete test results, and then had a nice conversation with the lab director. He tells me he has seen it go both ways, depending on several factors. Actually, efficiency for some of the filters in this test went down initially, then rose a bit toward the end. None were better after loading than they were when clean.

Definitely true that it can go both ways. Most of my work is with gravity filtration, where driving pressure is quite low in comparison to an oil system (typically less than 1 psi differential over the filter). To the point of the backwash initiation, effluent quality usually improves but higher pressure will begin to force material through that was caught initially. In water systems, we're cleaning on the fly, and tend to see the "worst" filtrate when relatively high driving head sees the clean media coming out of the backwash process.

It's my hunch that screens in this application are fine. If you begin to see internal destruction, it's going to go one of two routes. One, It'll be gradual enough, and you should pick it up in pressure changes, oil analysis, or just visual inspection at the oil change. Two, it'll be quick, and it'll present with a whole lot of other signs of trouble. For either path, the difference in solids storage capacity between a screen and disposable isn't liable to make a whole lot of difference in the end.

I appreciate the time and effort you put into this study. Good stuff.
 
Well...I'm ditching the K&P and going back to Wix. No disputing these results.

I can only wonder why K&P filter screens are so successful? Maybe the low delta P and good flow, when new/clean, means initial HP gains? Good for high performance racing applications where oil change intervals are short and every HP matters...bad for long term oil and engine longevity? Makes sense that Lycoming changed interval from 25 to 50 hrs when converting from screens to paper cartridges/canisters.
 
Keep in mind that Lycoming allows no filter with 25 hour oil changes. I assume they expect a typical engine to reach tbo without a filter and 25 hour oil changes?
I would be curious to know what size, qty, and type of particle is in the oil after 25 hours of normal use for a “typcial” engine and set up. Anybody doing oil analysis with no filter and 25 hour changes? Still might not show.
Do we need to filter all particles? All sizes? Perhaps the reusable filters aren’t that bad in “real” life scenarios. (Doesn’t look good for them regardless).
I only used the S15 on the RV for a short time when paper filters were scarce and switched back as soon as I could get them again. I had a new one on the shelf for Dan because I did not have the clearance to get any kind of torque wrench on the Bückers engine. When I torqued the S15 on the RV, it went from loose to “torqued” with a very small amount of rotation. I was not comfortable, nor could I determine, how to torque by hand like you can with a normal filter. So, Dan, you got the new S15 intended for the Bucker. This contributed to my lack of comfort using the S15 on the RV before these tests came out.
 
An earlier comment caught my eye, one which highlighted the performance of the Wix filter. If the numbers seem to show it to be a very good filter, why wouldn't we all be flocking to the Wix unit? I know many will say, "I only fly genuine aircraft filters" - I understand this bias. From a purely technical standpoint, what part of the Wix 51515's performance (either mechanically or in its filtration function) makes it less than the best choice?

I very much appreciate seeing these results as presented so far.

The first two items leaping to mind are can burst pressure and a rather low bypass valve opening pressure. Both are of lesser concern for owners in warmer climates, and using multi grade.
I have looked at a lot of physical specs and found better choices within the Wix lineup...but that's a subject for another time.
 
Last edited:
Very interesting information - I'm happy we have some data. Can't deny that it's a bit disappointing since the cleanable filters seemed to be an elegant solution to filter inspections, cost savings, and the lower environmental impact. Can't easily get Wix here, but we do have the MANN-FILTER W 940/4.
 
Timely. Just cleaned my PC Racing PCS4 for the second time.

Benched it in lieu of a Wix 51068 as I was not a fan of the gasket based on others' issues.

It was easy to inspect and clean, but a bit small. I was hoping it was "better" than the factory screen and would support 50 hours and solid filtration and better flow than a cellulouse filter.

Clearly a nope. $13 for the Wix. I was previously a K&N HP2004 user, but they changed country of origin from USA to Mexico to S. Korea.

The 5151 Wix and K&N HP3003 are too tall at 5.3" for my RV-6 O-320 with the filter horizontal pointed aft to the firewall.
 
Last edited:
I’m curious what theses cleanable filter manufacturers would say about these results?
K&P says, which I don't agree with, that a cleanable filter uses a different ASTM-F316

"From the K&P web site."..

4) How does this type of filtration compare to paper oil filters?
We use ASTMF316 testing procedures which eliminate many of the user variables found in the SAE procedures. Basically, the filter media is pressurized from one side, and when the media starts passing particles, that is the micron rating. We sent filter media from several common brands of paper filters to the lab to be run through the ASTM test. We sent the media to the lab with no names, just numbers for identification so they wouldn't have any idea what brand filter they were testing. The results for the paper filters ranged from 48 microns for the best filter to over 300 microns for the worst filter. Our tests were right in line with other testing results we have researched that have paper media filters passing particles anywhere between 50 and 90 microns. What does this mean? Paper filters are rated on averages, percentages of efficiency (also known as beta ratios) and multiple passes, so a 10 micron rated paper filter (as advertised on the packaging) may be letting particles 50 microns and larger through. The medical grade stainless steel cloth that we use is consistent across the entire media surface and is rated at 35 microns, meaning nothing larger than 35 microns should pass through the material. The bottom line is we meet or exceed the filtration performance of OEM filters, eliminating any warranty issues."
https://kandpengineering technical-information testing
 
...with a caveat. You're looking at average performance from beginning of test to termination at 10 psi deltaP. Depth media performs better when clean than later in the test when it is loaded with contaminent. The Champion and the Tempest look better if we consider some arbitrary period near the beginning of the test and drop the efficiency values taken when loaded near the end of the test. As a practical matter, it means a 25 hour filter cleans better than a 50 hour filter, which is better than a 75 hour filter. I'll have more on that later.
Anyway, here's an example in numbers, the data table for a single particle size, Champion filter.

Champion @25 um
Minutes deltaP. Efficiency %
10 0.01 psid 85.15
20. 0.02 psid 81.34
30. 0.04 psid 77.43
40. 0.08 psid 73.74
50. 0.14 psid 69.91
60. 0.26 psid 66.3
70. 0.47 psid 61.85
80. 0.64 psid 58.56
90. 1.09 psid 54.31
100. 2.43 psid 48.92
110. 10.16 psid 37.14
110.1. 10.31 psid 31.75

It would be interesting to know what is the deltaP and Retained Capacity after 50 hours of use on an average Lycoming. Too many variables engine to engine so probably would create a large scattering of results.
This leads me to want to determine it for my engine. How could I go about measuring the deltaP and/or Retained Capacity of my filter after 50 hours? Can it be done easily and cheaply in my own shop/hangar? I could then maybe use the above data to estimate, for my engine, how effective my filter is when I remove it at 50 hours. Then know if I can wait til 60 hours or should be changing filter at 40 hours.
 
Last edited:
The first two items leaping to mind are can burst pressure and a rather low bypass valve opening pressure. Both are of lesser concern for owners in warmer climates, and using multi grade.
I have looked at a lot of physical specs and found better choices within the Wix lineup...but that's a subject for another time.
Dan - you've clearly been busy digging deeper into these matters. I will happily stand by for whatever information you might care to share with respect to a better choice from the Wix lineup.

Again, thank you for pursuing this very helpful data.
 
K&P says, which I don't agree with, that a cleanable filter uses a different ASTM-F316
Agree with you.

They use an ASTM that does not measure filtration. They are using "Smoke and Mirrors" to get people to look at something their so called filter is good at because it does NOT filter well.

I want a good filter at a reasonable price.
 
Explain why Challenger has a STC on their reusable filter. I assume they have proof that it is adequate.

paper filters filter better as they plug up, I dont believe that "extra" filtration due to plugging up is part of the planned efficiency.
 
Explain why Challenger has a STC on their reusable filter. I assume they have proof that it is adequate.

paper filters filter better as they plug up, I dont believe that "extra" filtration due to plugging up is part of the planned efficiency.
The stainless surgical mesh filters are finer openings than the OEM screen. Measured in Kitplanes a while back-
"A Permanent Solution
Oil filters you don’t have to replace. By Paul Dye -October 15, 2022"

If durable, they do no worse, hence STC territory. "Does no harm" does not have to be the best solution to gain the approval.

"An STC will be issued only if:
the pertinent technical data have been examined and found satisfactory,
all necessary tests and compliance inspections have been completed, and
the alteration has been found to conform with the technical data."
 
Last edited:
Explain why Challenger has a STC on their reusable filter. I assume they have proof that it is adequate.

Are STC applications public documents? The STC number appears to be SE02352CH. Responsible office is AIR-7C0: Chicago ACO Branch. Does this group know anyone there?

paper filters filter better as they plug up,

Gary, I'm working with data listing the efficiency of three filters, for 16 different particle sizes recorded at 10 minute intervals. They do not filter better as they load, although some columns spike a bit just prior to termination at 10 psi deltaP. Please see post #11 for an example.
 
Last edited:
the STC paperwork comes with the filter, so as public as it gets. If you mean the data used for approval, not sure how to find it. as far as the application approv al goes: https://drs.faa.gov/browse/excelExternalWindow/9080A4682257DC0086257D09006E6C5F.0001

these have been used for more than a decade, nearing its second one, so revelations at this late stage seem odd. so wouldnt the question be, do they provide adequate protection?

an STC is not an easy test. I work on avionic certification.
 
Explain why Challenger has a STC on their reusable filter. I assume they have proof that it is adequate.
Im not sure that the acceptance of an STC requires any performance guarantees other than “will not do harm”, and I doubt it’s even that high a standard. Considering the original configuration is “no filter”, then the bar is pretty low.
 
these have been used for more than a decade, nearing its second one, so revelations at this late stage seem odd. so wouldnt the question be, do they provide adequate protection?
The tested screen filters didn't catch any significant quantity in the range defined by ARP1400, the piston aviation filter standard. Please see post #1.
 
The tested screen filters didn't catch any significant quantity in the range defined by ARP1400, the piston aviation filter standard. Please see post #1.

Ok, the question is if it did/does not, is there any significant harmed being done that will prevent the engine from reaching TBO? In other words are these particles of any significance?
 
Ok, the question is if it did/does not, is there any significant harmed being done that will prevent the engine from reaching TBO? In other words are these particles of any significance?

Would you be willing to pour five grams of A3 Medium down your oil filler tube?
 
Last edited:
So Lycoming came out with a service circular, SSP-885 dated April, 1986 that converted the screen to an oil filter. As part of it it claims the following benefit:
Installing an oil filter on an engine not equipped with a filter can increase the oil change interval from 25 to 50 hours.

The plane we have wont make it to 50 hrs on an oil change, with or without a filter as the oil is worn out by then anyway. By 25 hours it is long gone. Plus we do not have an air filter. So if you are extending the oil changes to 50 hrs, a paper filter is probably best. I think Challenger specifies 25 hrs oil changes but I dont see it right now.
 
So Lycoming came out with a service circular, SSP-885 dated April, 1986 that converted the screen to an oil filter. As part of it it claims the following benefit:
Installing an oil filter on an engine not equipped with a filter can increase the oil change interval from 25 to 50 hours.

The plane we have wont make it to 50 hrs on an oil change, with or without a filter as the oil is worn out by then anyway. By 25 hours it is long gone. Plus we do not have an air filter. So if you are extending the oil changes to 50 hrs, a paper filter is probably best. I think Challenger specifies 25 hrs oil changes but I dont see it right now.
What do you mean , “ worn out by then anyways.”?
 
What the marketing blob says about my Flo PCS4 . Me thinks I go back to paper.

  • Made from laser cut, medical grade, 304 stainless steel micronic filter cloth, this filter provides 200 percent more filter area in many cases
  • The Flo stainless filter catches items down to 35 microns, which is about 3 times better than most good paper or brass filters
  • Unlike glued paper filters, the Flo Oil Filter pleat seam is welded and able to withstand up to 600 degrees
  • This filter maintains consistent flow under all conditions including extreme heat, the presence of water, and cold start ups where paper filters can flow so poorly that they often cause the bypass valve to open and allow unfiltered oil to enter your engine
 

Attachments

  • IMG_0813.jpeg
    IMG_0813.jpeg
    79.3 KB · Views: 21
Dan, can you post the full report? I read through your post a few times and didn't see a link, did I miss it?
 
You didn't miss it. By agreement, I may not publish or otherwise distribute the full report as delivered, or identify the source of the report. Same for the names of the engineers who have given considerable time to our project. However, I may derive data and re-format, so that's the plan. The fundamentals are all in post #1. The bulk is mostly tables with efficiency data taken at 10 minute intervals for 16 particle sizes. Happy to address specific questions.

The facility tests for practically everyone in the filter industry. I toured the place a few weeks ago, and the boxes are right there on the shelves. They are very careful about maintaining neutrality.

It was the price of working with a highly regarded lab. The fuels and lubricants department is a only one small part of the organization. I am not willing to burn any bridges there. Way too much value in the contacts.
 
Last edited:
the STC paperwork comes with the filter, so as public as it gets. If you mean the data used for approval, not sure how to find it.
Compliance data for an STC is virtually always proprietary; I've never seen otherwise in my experience. The FAA and/or supporting DERs have it else no STC would have been issued, but they have no authority to release it so any cert data would have to come from Challenger.

Having said that, look at FAR Part 33 - the engine cert regs - and see what they say about oil. That will give you a good idea of the standards that Challenger had to meet for the STC.

HTH

Dave
 
You didn't miss it. By agreement, I may not publish or otherwise distribute the full report as delivered, or identify the source of the report. Same for the names of the engineers who have given considerable time to our project. However, I may derive data and re-format, so that's the plan. The fundamentals are all in post #1. The bulk is mostly tables with efficiency data taken at 10 minute intervals for 16 particle sizes. Happy to address specific questions.

The facility tests for practically everyone in the filter industry. I toured the place a few weeks ago, and the boxes are right there on the shelves. They are very careful about maintaining neutrality.

It was the price of working with a highly regarded lab. The fuels and lubricants department is an only one small part of the organization. I am not willing to burn any bridges there. Way too much value in the contacts.
Bummer that so much of the source and full report are confidential, if sort of understandable. Kitplanes being able to incorporate it into an updated review of the reusables would be a valuable (and maybe necessary) service given their previous endorsement. Hopefully they’ll find a way to do it.
 
Bummer that so much of the source and full report are confidential, if sort of understandable. Kitplanes being able to incorporate it into an updated review of the reusables would be a valuable (and maybe necessary) service given their previous endorsement. Hopefully they’ll find a way to do it.
Already have that commitment. And again, we have the data. I can publish data.

Understand something. VAF is a great place for discussion. Kitplanes is a great place for a comprehensive narrative. I spoke with both Doug and Marc before I came to the group, with the idea of leveraging the best of both media for mutual benefit.

So, please be patient. This is not a finished product.
 
... look at FAR Part 33 - the engine cert regs - and see what they say about oil. That will give you a good idea of the standards that Challenger had to meet for the STC.

Good call Dave. Explains a lot...recip engines don't appear to have a filter requirement...zip, nada, none:

ScreenHunter_2202 Apr. 13 16.12.jpg

The turbine folks get lubrication standards:

ScreenHunter_2201 Apr. 13 16.10.jpg
 
Back
Top