- Joined
- Oct 26, 2022
- Messages
- 40
Thoughts and opinions on this “oil filter testing” video
“Testing oil filters”
- Thread starter Chevy454z71
- Start date
- Status
I just watched this.
I am very surprised the K&N performed better than the Fram?
I know K&N oil filters are made by Champ labs who also makes Royal Purple… so it’s believable… but I don’t know if I believe the testing just because he didn’t say if it was done by a certified facility etc
I am very surprised the K&N performed better than the Fram?
I know K&N oil filters are made by Champ labs who also makes Royal Purple… so it’s believable… but I don’t know if I believe the testing just because he didn’t say if it was done by a certified facility etc
The 30micron size of the aluminum particles would be on the order of coarse test dust.
Typically medium test dust would be used for oil filters which has a median particle size ~20micron.
Then I question the pump used. Why are we getting the same flow at the same pump speed in part 2, but different pressure in part 1.
Something's not quite right.
The fact that a lot of material got caught by the Fram ADBV shows something was very wrong with the test set-up.
And the loading is crazy high in all cases.
Typically medium test dust would be used for oil filters which has a median particle size ~20micron.
Then I question the pump used. Why are we getting the same flow at the same pump speed in part 2, but different pressure in part 1.
Something's not quite right.
The fact that a lot of material got caught by the Fram ADBV shows something was very wrong with the test set-up.
And the loading is crazy high in all cases.
I don't think these tests mean squat, adding big clumps of aluminum powder to such a small sump, and what do they mean by "constant pump power," are they using a positive displacement type pump? Purposefully limiting the power of the pump for more restrictive filters is not representative of real operating conditions where a positive displacement pump would just become a higher parasitic load but still force the same level of oil flow. Also did he say in the begining that more media is more restrictive, when in fact the restriction actually decreases by the square of the area increase. How is the industry standard test actually performed?
In real life, how often does an engine fail solely because of 1 bad oil filter? I could see this if there was repeated use of a horrid filter, but that's gotta be extremely rare.
I wonder if they took into account the bypass spec of the filters?
Per ISO 4548-12. Ascent did ISO testing per the official ISO 4548-12 test standard that has been in use around the world since 1999.
In the ISO test, the flow rate is held constant with a PD oil pump.
Would you all like to see ISO 4548-12 Oil Filter Lab Testing Comparison, Efficiency & Capacity, Pressure vs Flow, Bubble Point, and Burst?
I have this as a “watched” thread, so if you do happen to start a new one, could you put a note or link in this one to direct followers to the new thread? Thanks!! Yes, no problem I can do that.
bobistheoilguy.com
I mean what's the testing procedure?
Yes, he was totally wrong making that claim. Pretty basic that more flow area will result in less flow restriction - ie, delta-p across the filter.
If you read that whole tread there is a lot of insight on how the ISO 4548-12 test is performed. If you search for ISO 4548-12, there use to be a full PDF download in a Google search a while back. Hard to find, because ISO wants people to buy the procedure.
He didn't have any way to measure the actual dP across the filter, so the filters certainly could have been in bypass at some point as they loaded up. All he was measuring was inlet pressure and flow rate through the filter.
Also, in the beginning where he cranked the pump up to "90" to get a pressure on the inlet side, he never showed what the flow rate was through each filter. If the flow rate was not the same, then the input pressure doesn't mean much. If the pump was really a postitive displacement pump, then the input pressure a full RPM would be valid. But all he was measuring was pump power, which may not directly equate to pump RPM under different loads.
There was a certain member on here years ago that did a similar test and it did not go over well. This person also did a test where he got UOA data from a run with an oil filter and without an oil filter, and the UOA's looked the same.
Speaking of UOAs, like mentioned by @Brian123 above, and also mentioned in the "test video" ... using a UOA for wear particles isn't the best way to determine the filtering performance because UOAs only measure particles that are around 5 microns and smaller. So pretty much all full-flow filters will not filter much at 5 microns or smaller. Results can be skewed some by the accuracy of the test itself. Now if a super duper efficient bypass filter was used, then you might see a much larger impact on the UOA wear metal ppm detected.
The correct test of the filtering performance would be an ISO 4406 particle count test.
The correct test of the filtering performance would be an ISO 4406 particle count test.
Last edited by a moderator:
I'd like to see the oil up at operating temp 100*c too.
A better feel for the flow resistance comparison can be seen when they do the start of each holding capacity test. They all start out with 9 L/min of flow, so the same flow rate is going through each filter. We only know the filter inlet side pressure to achieve that flow rate, and don't know what that pressure would be if there was no filter on the mount - ie, the flow resistance of the flow circuit (hoses, flow gauge, etc) down stream from the pressure gauge.
And as jetman eludes to, this oil is at room temperature so it's relatively thick and will therefore produce much more dP across the filter than if the oil was at 200F.
Best I could read on the pressure gauge for each filter with the flow at 9 L/min (2.4 GPM) was:
K&N = 6 PSI
Fram = 6 PSI
SuperTech = 5.5 PSI
Mobil 1 = 5 PSI
So there was really only ~1 PSI difference between them all with thick oil at 2.4 GPM of flow.
The oil pump was always set to power level "23" as the filters were being loaded with test dust, but the flow went down, so the oil pump can't be a positive displacement pump, or if it was the pump RPM decreased as the filters became more restrictive. If the pump was PD and didn't change RPM then the flow should have basically remained constant and the dP across the filter would increase - that's how it works on an engine with a PD pump. As mentioned before, he didn't have a pressure gauge on the outlet side of the filter, so really can't say what the dP was across each filter.
And as jetman eludes to, this oil is at room temperature so it's relatively thick and will therefore produce much more dP across the filter than if the oil was at 200F.
Best I could read on the pressure gauge for each filter with the flow at 9 L/min (2.4 GPM) was:
K&N = 6 PSI
Fram = 6 PSI
SuperTech = 5.5 PSI
Mobil 1 = 5 PSI
So there was really only ~1 PSI difference between them all with thick oil at 2.4 GPM of flow.
The oil pump was always set to power level "23" as the filters were being loaded with test dust, but the flow went down, so the oil pump can't be a positive displacement pump, or if it was the pump RPM decreased as the filters became more restrictive. If the pump was PD and didn't change RPM then the flow should have basically remained constant and the dP across the filter would increase - that's how it works on an engine with a PD pump. As mentioned before, he didn't have a pressure gauge on the outlet side of the filter, so really can't say what the dP was across each filter.
Last edited:
Statistical error for the 30 micron particle capacity... what statistical error? What was the sample size for each filter? Most likely one. Fram Ultra was the only one that claimed 99% efficiency at 20 microns, whereas the rest were at 30 microns.
This video is a small step above something from Project Farm.
Real engineers do testing according to standardized procedures, unless no procedure exist, then you have to make one up
This video is a small step above something from Project Farm.
Real engineers do testing according to standardized procedures, unless no procedure exist, then you have to make one up
I'd say a step or two below. Not much can be gleaned from the test except that they all have about the same dP (within 1 PSI of each other) at a flow setting of 9 L/min (2.4 GPM) on the flow gauge (calibrated for water) with the same oil at room temperature.
The "holding capacity" test was way too uncontrolled. The contaminated oil "clean-up" test would have been better if they would have contaminated the test oil with the test dust to a reasonable (not too crazy) level, or better yet with test dust with a distribution of particle sizes, Then run the clean-up test with the same flow rate for the same time through each filter, so the same total volume of dirty oil was passed through each filter. Then send out all the oil samples for an ISO 4406 particle count to see which filter did the best clean-up job of the contaminated test oil. Using a UOA that measures the wear metals ppm is not going to say much because it only measures particles 5 microns or smaller, and full flow filters don't filter much at 5u or smaller.
Also need a pressure gauge on the outlet side of the filter to measure dP across the filter if that is of any interest. If there is a dP and a flow measurement, don't really need a PD pump because you can set the flow and then see the resulting dP.
Also need a pressure gauge on the outlet side of the filter to measure dP across the filter if that is of any interest. If there is a dP and a flow measurement, don't really need a PD pump because you can set the flow and then see the resulting dP.
Last edited:
- Status
