The Myth of "Flow Over Efficiency"

[Hohn]

Don't want to side track the tread, but I think the term "high flow" used by many in the context of oil filters stems from the misconception and myth of "flow over efficiency". Even if the context was defined as "which filter flows the most at the same dP", they are all going to be flowing pretty close to each other unless there is a big outlier in the mix, just like when comparing their dP vs flow performance. On a hair splitting level, there will be some flow difference, but it will be too small to really matter when the pump is out of pressure relief unless the oil pump is way mis-matched to the engine (pretty unlikely).

As far as Honda OEM filters, it would be nice to see some dP vs flow numbers for them, and also between the two OEM versions (Japan and Fram USA). I'm betting the OEM version by Fram is made to match closer to the Japan OEM version than to the Fram EG.


Some filter makers do list just a flow value (ie: 9 GPM) with no qualifying associate information, so it's a pretty nebulous spec. I'd like to know what that spec is based on.

I'm replying to this in this thread instead of the Honda thread, I think this is worth discussing.

A flow value without a restriction value means nothing. And a restriction value without a flow value means nothing. For convenience, shorthand values are developed the make the relevant companion spec implicit instead of explicity.

For example, the valve industry uses parameters like Flow Factor and Flow Coefficient (Kv and Cv respectively). They assume some standard pressure drop and then furnish the "factor" as a volumetric flow rate.

Unfortunately, unless you have the ability to see a flow rate and know (by reference) what the pressure drop is at which that flow rate is measured, then the flow rate value is meaningless. It cannot be compared to anything. It's apples/oranges.

It's like how people say "This has a 750cfm carburetor on it." Well, it is, but only when the pressure drop across the carburetor is 1.5"Hg. How many people know the 1.5"Hg part?

Or what about a "1200cfm air filter"? Well, air filters are rated at different pressure drops. Some rated at 8" of water, some at 10", some as low as 6". But 1200cfm at 10" is a lot less real world flow than 1200cfm at 6". But if you don't know the actual pressure drop going with the "flow rating" the flow rating is worthless marketing drivel. Sometimes you get lucky and they tie it to an ASTM or SAE procedure with at least a discernible reference pressure drop.

But quite often even using a standard reference doesn't help you. For example, the ISO4548-12 that we all love has no reference flow rate in the standard. A Fram Ultra and a Purolator Boss can both run the test, but the flow rates for the tests are selected by the OEMs, it is NOT part of the test standard. Fram might run it at 10L/min and Purolator at 8L/min and it's perfectly acceptable to the standard. Which means you cannot compare pressure drops of one filter to another from the ISO test because you have no idea which flow rate was used when those pressure drops were measured.

Heck, the initial pressure drop for the test isn't even dictated. All that the test shows is how pressure drop increases at an arbitrary flow rate as a function of particle mass accumulation. It measures particle efficiency and *change* in restriction. But you have no idea what the baseline is what the context of that change is.

It's a glaring shortcoming in this test to not index the flow rate to, say, media area. At the very least, they could categorize filters and have different reference flows for "passenger car" vs "light truck" or whatever. As long as the filter maker can select different flows for different models and you as the customer have no idea what the reference flow rate is, the flow and restriction aspects of 4548-12 are worthless for comparison purposes. All this test shows is dirt removal and capacity. It does NOT show the flow vs restriction curve of the filter! It's just giving you the offset from some mystery curve as a function of particle removal. But you don't know the baseline curve or where you are on that curve, so?!?!

My wish is that every consumer of data would reflexively ask "at what pressure drop?" when given a flow rate, or ask "at what flow rate?" if given a pressure drop. Without both pieces of information, the value is undefined and thus worthless.

At least in the Brands Ranked testing they used the same flow rate range for all filters they tested. But that's not because the standard required them to!

 

[ZeeOSix]

I'm replying to this in this thread instead of the Honda thread, I think this is worth discussing.

A flow value without a restriction value means nothing. And a restriction value without a flow value means nothing. For convenience, shorthand values are developed the make the relevant companion spec implicit instead of explicity.

For example, the valve industry uses parameters like Flow Factor and Flow Coefficient (Kv and Cv respectively). They assume some standard pressure drop and then furnish the "factor" as a volumetric flow rate.

Unfortunately, unless you have the ability to see a flow rate and know (by reference) what the pressure drop is at which that flow rate is measured, then the flow rate value is meaningless. It cannot be compared to anything. It's apples/oranges.

It's like how people say "This has a 750cfm carburetor on it." Well, it is, but only when the pressure drop across the carburetor is 1.5"Hg. How many people know the 1.5"Hg part?

Or what about a "1200cfm air filter"? Well, air filters are rated at different pressure drops. Some rated at 8" of water, some at 10", some as low as 6". But 1200cfm at 10" is a lot less real world flow than 1200cfm at 6". But if you don't know the actual pressure drop going with the "flow rating" the flow rating is worthless marketing drivel. Sometimes you get lucky and they tie it to an ASTM or SAE procedure with at least a discernible reference pressure drop.

Yes, that was my point with the statement of: "Some filter makers do list just a flow value (ie: 9 GPM) with no qualifying associate information, so it's a pretty nebulous spec". Now if they said: "9 GPM at 5 PSI dP and 11 cSt oil viscosity" then it would make sense. Another thing seen often is simply an oil filer efficiency percentage with no associated micron size. The carburetor and air filter flow "spec" are also good examples, as they are also similar incomplete "specs" that are somewhat nebulous with the required associated information missing that would make it clear and accurate.

But quite often even using a standard reference doesn't help you. For example, the ISO4548-12 that we all love has no reference flow rate in the standard. A Fram Ultra and a Purolator Boss can both run the test, but the flow rates for the tests are selected by the OEMs, it is NOT part of the test standard. Fram might run it at 10L/min and Purolator at 8L/min and it's perfectly acceptable to the standard. Which means you cannot compare pressure drops of one filter to another from the ISO test because you have no idea which flow rate was used when those pressure drops were measured.

Heck, the initial pressure drop for the test isn't even dictated. All that the test shows is how pressure drop increases at an arbitrary flow rate as a function of particle mass accumulation. It measures particle efficiency and *change* in restriction. But you have no idea what the baseline is what the context of that change is.

It's a glaring shortcoming in this test to not index the flow rate to, say, media area. At the very least, they could categorize filters and have different reference flows for "passenger car" vs "light truck" or whatever. As long as the filter maker can select different flows for different models and you as the customer have no idea what the reference flow rate is, the flow and restriction aspects of 4548-12 are worthless for comparison purposes. All this test shows is dirt removal and capacity. It does NOT show the flow vs restriction curve of the filter! It's just giving you the offset from some mystery curve as a function of particle removal. But you don't know the baseline curve or where you are on that curve, so?!?!

My wish is that every consumer of data would reflexively ask "at what pressure drop?" when given a flow rate, or ask "at what flow rate?" if given a pressure drop. Without both pieces of information, the value is undefined and thus worthless.

At least in the Brands Ranked testing they used the same flow rate range for all filters they tested. But that's not because the standard required them to!

If you read ISO 4548-12, the test flow rate is based on the estimated holding capacity of the filter. That typically comes from the filter designer, and if they can't give a good number then the ISO spec says the holding capacity would need to be determined by actually testing an example of the specific filter model before actually running the ISO efficiency test. And the dust loading rate is also determined from the estimated holding capacity, or actually measured holding capacity if that's required. This makes some sense because the test conditions are trying to be matched to the filter holding capacity so the test duration time (recommenced to more than 30 minutes) is about the same regardless of how small or large the filter element is. You wouldn't want to test a small filter with a too fast of oil flow and dust loading, and you wouldn't want to test a large filter with a too slow of oil flow and dust loading rate. When Ascent did the ISO testing on the 5 filters, they were all basically the same size (all speced for the same engine) and the holding capacity was estimated to be close to their average so the same oil flow and dust loading rates were used.

And the dP vs flow that Ascent did was basically what BR attempts to do, which is to measure the dP as the flow rate increases over an oil flow range while holding the oil viscosity constant. ISO test standards basically require that, otherwise you couldn't generate a dP vs flow curve at a specific viscosity.

 

[carl2591]

You let it idle till the temps get to operating temps?? I think the issues is fuel dilution in oil for gas and with a newer EPA diesel the amount of soot that creates in the intake system.. I have seen pictures of cummins engine with the intake especially at the grid heater almost completely clogged from excessive idling much like what you suggest..

the general consensus among not only oil experts,, Lake Speed Jr, and engine experts suggest in excessive cold,, like -10F or so to start and let engine idle for 30 to 60 seconds, put vehicle in gear and accelerate at a below normal pace for some miles till the engine warm up naturally..

the info many of the average joe self taught think are facts are actually are just that myths.. when you know you know..

% of this stuff I don't even comprehend. I'm nothing more than you're average " Joe" self taught mechanic. But, as a kid, 12-13 yrs old, I was always told " if it's cold outside, let the engine idle for awhile so the oil can warm up". It can get pretty cold in NY ..10-15 degrees in the winter. I always let the car warm up to get fluids up to near operating temps. Taught my wife to do the same. No blown radiator hoses, transmission leaks, or engine problems. Yeah, you waste some fuel, but big whoop.

 

[50yr Street Mech]

You let it idle till the temps get to operating temps?? I think the issues is fuel dilution in oil for gas and with a newer EPA diesel the amount of soot that creates in the intake system.. I have seen pictures of cummins engine with the intake especially at the grid heater almost completely clogged from excessive idling much like what you suggest..

the general consensus among not only oil experts,, Lake Speed Jr, and engine experts suggest in excessive cold,, like -10F or so to start and let engine idle for 30 to 60 seconds, put vehicle in gear and accelerate at a below normal pace for some miles till the engine warm up naturally..

the info many of the average joe self taught think are facts are actually are just that myths.. when you know you know..

I agree with you regarding diesels. I drove diesel work trucks from 1986-2007 supplied by my employer, Con Edison Utility Company in NY. None had turbos. Motor pool mechanics were instructed to start them, and let them idle for 15-20 minutes in the winter months before the start of next shift. The cloud of blue smoke was unbelievable when the next shift drove off. When I retired in 2022, the same procedure was in place, totally disregarding new laws about idling vehicles. But, only 1 engine blew in all those years that I know of.. That was due to the driver not doing their vehicle pre-check, which included checking engine oil level. He got suspended for 2 weeks, after the mechanics finished their investigation on the engine, and found engine to be 4 qts low on oil. The mechanics never started gasoline powered pick ups or cars prior to a shift in the winter. Personally, I'll keep warming up my NA gas beaters. So far, no problems with the engines or transmissions in the winter months. I'll keep doing what I was taught as a young mechanic. My personal preference.

 

[ZeeOSix]

the general consensus among not only oil experts,, Lake Speed Jr, and engine experts suggest in excessive cold,, like -10F or so to start and let engine idle for 30 to 60 seconds, put vehicle in gear and accelerate at a below normal pace for some miles till the engine warm up naturally..

A cold start-up at -10F is going to get a bit more warm-up time than 30-60 seconds on my machines. I typically wait until the high idle settles down to under 1000 RPM, so if it was -10F it's going to take longer than 60 seconds.