Zee -
I know we agree in principle on most things. But the thing your comments leave undefined is the very crux of the problem in these discussions. Having other things "constant" does not help define the parameters; it only equalizes them.
You have to "equalize" things, meaning look at the effect of only one factor by holding all other factors constant, or the effect of the variable will never be understood. Hold all other effecting parameters constant, and only focus on the one variable for the resulting effect. In this case the only variable should be the filter efficiency while holding all other possible effecting factors as constant as possible. This is the basic way the standardized ISO 4548-12 test works, or any other SAE/ASTM, etc test works. It's designed to test the performance of filters all ran under the same operating conditions.
If a sump is heavily laden with particulate, as a result of a long OCI in a dirty running engine, then a higher efficiency filter that also has higher capacity is going to shine above the others. That I would agree with. But if a sump is fairly fresh, and held has a constant with low contamination from frequent OCIs with a good lube, the high eff and high capacity won't much be able to distinguish itself above other choices.
Same thing I've always said. The longer the OCI, the more important it is to use a higher efficiency oil filter to keep the oil cleaner over that longer OCI. Because wear from particulate in dirty oil is basically proportional to the cleanliness level of the oil times how long the oil has been circulated through the oiling system.
As we've agreed before, the difference between a 95% filter and a 99% filter (at 20um) over a 5k mile OCI isn't going to really manifest into a tangible difference in terms of wear control. But those same two filters over perhaps 15k miles with a dirty running engine might make a very distinct difference. Or, if a 95% filter were placed against an 80% filter, the better would certainly prevail.
Of course ... I've always said the same in every one of these discussions. Obviously, the closer that two filters are in filtering efficiency, the less difference they will make in use. We just say the same things in different ways, been that way for years.
Holding something "constant" only puts it in a state of not being a variable, but it does NOT define the condition as to whether it's inherently high or low, relative to the overall considerations.
Still has to be done if you really want to see the effect of only the variable parameter you're interested in. If you want to see the effect of every variable, then you'd have to repeat the same testing and hold each other variable constant and look at the effect of whatever other variable you're interested in. If you had a system where 4 variables had an effect, you'd have to run the test 4 different times, while holding everything constant besides variable 1, then variable 2, 3 and 4 to see how each one effects the outcome. Just basic test logic.
When things are lab tested, they are often put to extremes to make the desired differences become more dramatic. But that does not mean that all real world situations will manifest into the same results. If the conditions are reasonably mundane, the large disparity in performance choices may well never materialize.
True ... but as an example, the Cummins field study showed there was a clear correlation between oil filtration and engine wear. Every wear study shows that cleaner oil results in less wear. Nobody can prove otherwise. The "level" of difference is certainly there, and people can argue that all day long, but the bottom line is that better oil filtration results in cleaner oil, which results in less wear. That's good enough for me to use filters that are more efficient than not, regardless of my OCI or use conditions. I cover all bases and unknown "ifs" by just using filters that have higher rated efficiency. It's a no brainier and doesn't require your own test program and science project to try and out think test information that's already there.
Again - if shorter OCIs, good air filtration and a decent oil are in play, the oil filter is generally along for the ride because of two things:
- there's not a lot of contamination in the first place (the quantity of particulate present is low)
- what does exist is on the smaller end of the scale (the size of the particulate is small)
You either have to assume or test to verify that those things are going on, or just cover it in the simple way ... by using a relatively high efficiency filter. I'd rather spend a few bucks more for a filter than spend lots of money doing my own "test program".
Also, all the wear studies basically say the particles 20u or smaller cause the most engine wear. So using a higher efficiency filter will remove more of those sub 20u particles than a much lower efficiency filter. If someone wants to go a level above that, then they'd have to go with a bypass filtering system.
So, when you say "hold all other things constant", you cannot say that without ALSO defining what the constant state is ... is it constantly clean or constantly dirty? I'm not saying oil filtration isn't important; it most certainly is. It's just that its importance is a variable based on other inputs to the overall equation. If you have a heavily contamination sump, the oil filter capacity and efficiency are very important. If you keep a clean sump with frequent changes, the oil filter becomes far less of a factor.
Anytime you have multiple variables effecting a system, if you can hold all but one variable constant you're going to see the effect of that variable. You have to hold a constant state of the system, except for the variable your looking at. This is how the ISO 4548-12 is meant to work. Of course, doing testing in the field is much more difficult to do, but if done reasonably well it will show the effect of one variable changing, just like shown in the in-field Cummins filter testing.
Of course if a controlled test where the sump was always pretty clean, it would be harder to see the difference of the filter efficiency that if the sump was really dirty. But regardless, it's still a fact that a higher efficiency filter will always keep the oil cleaner than a lower efficiency filter - regardless of the resulting difference level.
If someone wants to spend lots of time and money doing their own "test program" to try and determine what the level of difference is between filters to decide what filter to use, then go for it and please post the test results here for all to see. Looking at particle counts in UOAs posted here show that the ISO particle count data difference between 99% @ 20u vs 99% @ 40u (or 50% @ 20u) filters is pretty significant. How much difference in wear does that result in? ... certainly not enough to "blow-up" the engine, but it could make a difference over the long run. Of course the car might rust out or get T-boned and totaled, so who cares about oil and filters?
