Rethinking idea that Fram Ultra is the best filter

[doitmyself]

Guys and gals, Charlie presented a hypothesis or two and asked for peer review. He stated that he was exploring new ideas, but wanted to make a conclusion that would lead to a decision. No one has presented any help to Charlie to aid his decision because it is not possible.

Maybe if we challenged Mr. Buckley in a more respectful way, it might lead to more answers. What other filter manufacturer comes here to play with us? We no longer hear from PeteC of Wix and they made their efficiency data basically useless.

In the end, what CT8 just stated above is correct. But, it is fun to ponder things and is the OCD nature of many here.

 

[kschachn]

Originally Posted By: CharlieBauer
What do you do in the absence of data that is pertinent to your decision criteria?

Well you certainly don't then go forward and try to draw a conclusion based on imagined data.

I worked as a research technologist for a number of years out of college and my job was to design and implement tests that would result in statistically valid data when analyzed by the PhD mathematicians on our staff. I then had to present my results to the entire research group on our Friday seminars and be open to criticism and discussion. So maybe that experience has me living a bit far on the "what is actual valid data" side of things, but I just don't see any of what you've proposed can ever be considered as statistically valid data. In fact there's little to nothing ever posted on this site that meets that criteria, at least in terms of tests that individuals perform on isolated vehicles with little to no isolation of the variables. None of these tests even has a control variable, so what on earth can you compare it against? People often say we need "real-world tests" but the dark reality is that in most instances there is no such thing as real-world tests. It is exceedingly difficult to gather valid data from numerous disparate results with no controls on the process. Nearly always the only place that statistically valid data can be collected and turned into meaningful conclusions is the laboratory. There, all significant variables can be isolated and the results can be mathematically analyzed for validity.

People don't like to hear that but as of yet no one has convinced me otherwise.

 

[paulri]

Here is a later version of Btanchors' UOA/particle counts, that does include a Fram Ultra. The Fram Ultra is on the left side of the particle count section.

What particularly impressed me with the performance of the M1 was that this was in the early miles of the engine, which you'd normally expect a LOT of particles in the oil from.

This is one of the bits of information that lead me to think that the higher efficient filters are better at filtering out particles that aren't going to cause engine wear, anyways. The numbers from the particle counts in this image vary widely, but the wear metals in the upper section of the UOA don't appear to move by as much. For example, compare the RP filter under the 76,xxx column with the Ultra at 86,xxx miles. The oil in the Ultra had less than half the number of particles as the oil with the RP filter, but look at the wear metals--after 5000 miles, both UOAs were roughly the same. The cleanest oil was with the M1 filter at 30,xxx and 35,xxx. But the wear metal rates were about the same as with the other filters, weren't they?

Question--I have heard that blackstone particle counts are extrapolations. Does anyone know which number is actually counted out, and which are extrapolated from that?

 

[MParr]

More good points. “Controlled laboratory conditions “ do not apply once that filter design leaves the lab and goes to the production line. Quality control isn’t 100%. Distribution and handling will have an effect on an unknown number of those filters. Installer negligence can be an issue as well. Like I said before, what happens in the lab, stays in the lab.

 

[bullwinkle]

Fact is, ANY wire backed synthetic media filter is going to be more resistant to tearing and capable of holding far more particles than any cellulose filter will. Depends on what you're trying to do-if you change at 5K or less, any DECENT (again, Puro, looking at you) cellulose filter will get the job done. If you're going 10K, or more, including multiple OCIs, a filter is needed that can handle it.

 

[Trav]

Fram Ultra is a good well made filter its reasonably priced and readily available, it is a quality product. For long OCI and multiple cycles its large loading capacity may well make it the best readily available filter. If it works well on your vehicle I say why fix what aint broke.

If you are like me someone who performs 5K OCI or OLM then its a total waste of money, OE, Wix/Napa, German Mann, Mahle, Delco, ST. etc are more than good enough, the difference in engine wear will be immeasurable.

 

[Yah-Tah-Hey]

Just eleven months and one day and I'll be back in the oil filter conundrum.

 

[DoubleWasp]

There is no conspiracy here.

The source of this re-thinking is literally assumption stacked on assumption, and then multiplied.

It would be one thing if there were some other data to continue this line of thinking, but there isn't.

 

[Old Mustang Guy]

I like OE because I know of some fairly rare cases in which there are some problems with certain filters, like with the Ford 4.6 motors I had. If I wanted to do a long OCI I might look for something better, perhaps more capacity.

 

[oil_film_movies]

Originally Posted By: Old Mustang Guy
I like OE because I know of some fairly rare cases in which there are some problems with certain filters, like with the Ford 4.6 motors I had. If I wanted to do a long OCI I might look for something better, perhaps more capacity.
Like Ford OE motorcraft oil filters that are famous for getting holes in them? OE Ford is probably the worst choice. As long as Purolator makes Motorcraft, they will continue to be a poor choice.

 

[Rand]

People Get too hung up on "best"

Good filter for the money or OEM here.

example the subaru blue oem are basically bottom tier frams but with the proper bypass spec.
Wix is almost the only aftermarket filter with the high bypass spec.

Hyundai OEM are actually quite nice. Thick Can. ptfe round gasket.

I didnt like the fram ultra in my 3.2 pentastar was extremely hard to get seated on the endcap and remove when used(canister filter not spin on)

 

[goodtimes]

I am starting to think the other direction from many of my previous posts, maybe it is the best for me.

One reason is I am betting on the 99%@16 micron number. That's just a really good number if true. Plus the stainless steel wire mesh and gasket on the bypass leaf retainer. Plus no louvers.

 

[oil_film_movies]

Originally Posted By: goodtimes
I am starting to think the other direction from many of my previous posts, maybe it is the best for me.

One reason is I am betting on the 99%@16 micron number. That's just a really good number if true. Plus the stainless steel wire mesh and gasket on the bypass leaf retainer. Plus no louvers.

No louvers, forgot about that nice part. Yes best construction. I've been part of the Ultra-Borg Collective for a while.
Another mind bows to Ultra.

 

[ZeeOSix]

Originally Posted By: HangFire
Sorry Charlie, thinking is NOT allowed on this sub-forum. You must submit to the opinion of the most rabid poster here.


You got any technical facts to add to this discussion? I don't think so, just personal attacts. Nice way to discuss this.

 

[ZeeOSix]

Originally Posted By: 2015_PSD
So...you take:

1. Filters from different manufacturers
2. Filters with and without filter magnets
3. Test data from two different labs
4. Input data from different vehicles/drivers/maintenance schedules

"Analyze" it and from this you cobble together a "conclusion"? I assume you realize that nothing has been proved or disproved and this would not even qualify as a SWAG.


Yep

 

[ZeeOSix]

Originally Posted By: zrxkawboy
Originally Posted By: HangFire
Sorry Charlie, thinking is NOT allowed on this sub-forum. You must submit to the opinion of the most rabid poster here.


Well said, HF.

Charlie, I completely understand what you're saying. You're not claiming to have all the data/answers, just using what you have, and exploring new possibilities. Unfortunately, you've triggered the forums Fram reps, and as you can see, that's not a pretty sight.

If someone is going to try and prove a theory the info and data should backup that theory pretty solidly, not cause more questions than answers.

Now the Fram haters have come out like zombies in the dark because the words Fram and Ultra was mentioned.

 

[ZeeOSix]

Originally Posted By: kschachn
Originally Posted By: CharlieBauer
What do you do in the absence of data that is pertinent to your decision criteria?

Well you certainly don't then go forward and try to draw a conclusion based on imagined data.

I worked as a research technologist for a number of years out of college and my job was to design and implement tests that would result in statistically valid data when analyzed by the PhD mathematicians on our staff. I then had to present my results to the entire research group on our Friday seminars and be open to criticism and discussion. So maybe that experience has me living a bit far on the "what is actual valid data" side of things, but I just don't see any of what you've proposed can ever be considered as statistically valid data. In fact there's little to nothing ever posted on this site that meets that criteria, at least in terms of tests that individuals perform on isolated vehicles with little to no isolation of the variables. None of these tests even has a control variable, so what on earth can you compare it against? People often say we need "real-world tests" but the dark reality is that in most instances there is no such thing as real-world tests. It is exceedingly difficult to gather valid data from numerous disparate results with no controls on the process. Nearly always the only place that statistically valid data can be collected and turned into meaningful conclusions is the laboratory. There, all significant variables can be isolated and the results can be mathematically analyzed for validity.

People don't like to hear that but as of yet no one has convinced me otherwise.


Good post, and to add along those lines, the SAE bus study clearly showed that oil filters that tested to be more efficient in the laboratory also wete the filters that resulted in cleaner oil when used in the field.

Many people think the ISO 4548-12 test doesn't mean anything, but per the bus study correlation it certainly shows a good correlation between lab test data and real world test data.

 

[ZeeOSix]

To add ... here's data from the SAE Bus Study showing the correlation between filter efficiency measured in the lab compared to how clean the oil measured in field use. Same bottom line that is always concluded ... the more efficient oil filter gives cleaner oil.

Out of filters A, B & C the most efficiently measured filter in the lab (Filter C) also gave the cleanest oil in the field testing. Filters A & B follow the same correlation down the line. There wasn't some magical 'switcharoo' that showed the least efficient filter somehow became an efficient filter in field use.

 

[CharlieBauer]

Maybe it's good to recap.

Data I looked at

Reasonably solid data: Dr Dave's finding that the high efficiency Amsoil synthetic filter becomes very very efficient somewhere between 14.4k miles and 22.2k miles.

Reasonably solid data: btanchors particle counts on same vehicle mostly at 5k intervals with varying filters. Multiple times on Amsoil full synthetic and M1 filters. Mostly found that the Mobil filter was 1 cleanliness code cleaner than Amsoil. Remember 1 code count less means halving the particles so it is not insignificant.

Less solid data: btanchors findings with other filters or at other mileage. 1 run with Fram Ultra to 5k as good as Mobil 1. 1 run with Amsoil and M1 to 10k show no improvement over 5k.

What I interpreted from the above: At mileages under 14.4k, Mobil 1 filter produced better cleanliness codes than Amsoil full synthetic.


Interpretation:

- Amsoil full synthetic is more efficient according to ISO testing. It also has more capacity and lasts longer than the M1 cellulose blend filter.
- But it looks like it will only produce better cleanliness codes around the 20k mile mark.
- For mileages under that specifically at the 5k mark, the M1 filter is producing better cleanliness codes.
- If I were doing 5k miles a year, then it could be 4 years before the Amsoil gives me cleaner oil than a M1 filter changed every 2 years. The M1 filter will be giving me cleaner oil earlier and overall for a longer time.


Extrapolation & Assuming:

- Amsoil is full synthetic high capacity filter 98.7% efficient at 20 microns. M1 is cellulose blend currently 99% efficient at 30 microns, seem to remember 98.7% efficient at 25 microns a while back.
- Fram Ultra is more efficient than the Amsoil and also full synthetic and high capacity. Tough Guard is more efficient than the M1 and also cellulose blend and similar rated life in mileage.
- Single particle count by btanchors on a Fram Ultra gave same cleanliness code as M1 filter
- Possible that higher efficiency cellulose blend filter rated for half the life of a Fram Ultra could be similarly better on cleanliness codes just as M1 was vs Amsoil.
- Rather than waiting potentially 3 to 4 years to achieve better cleanliness codes, could be better for 5000 miles a year to run Tough Guard (or equivalent) 99% at 20 micron vs Fram Ultra 99%+ at 20 microns. Possible that oil will be cleaner sooner.
- Also, ISO rating is already so high and so close, there doesn't seem to be any downside.

Hence "Rethinking"!

 

[CharlieBauer]

Originally Posted By: ZeeOSix
To add ... here's data from the SAE Bus Study showing the correlation between filter efficiency measured in the lab compared to how clean the oil measured in field use. Same bottom line that is always concluded ... the more efficient oil filter gives cleaner oil.

Out of filters A, B & C the most efficiently measured filter in the lab (Filter C) also gave the cleanest oil in the field testing. Filters A & B follow the same correlation down the line. There wasn't some magical 'switcharoo' that showed the least efficient filter somehow became an efficient filter in field use.









Thanks. These are the results one would expect to see.

From what I can see, although I don't have all the information, so you can clarify here as necessary, it differs from what I'm exploring in two respects.

Firstly, the differences in filter ratings in the SAE test are significant vs me looking at Amsoil vs M1 with a view to switching from Fram Ultra to Toughguard.

Secondly, I'm talking about the time taken until the ISO ratings lead to better cleanliness code and how that differs between a high capacity full synthetic filter and a lower capacity cellulose blend filter. Did the bus study provide any particle counts at different mileages? Not suggesting that a hugely lower efficiency filter is going to better but that a similarly efficient filter may be better sooner if it has less capacity.