Oil Filter Pressure Differential testing (PSID)

[RSMike]

Hi Team

I've conducted some testing on my car to see the different pressure drops across different filters.
The results? - It was a fizz.
When the filters are NEW, they all represent zero restriction to oil flow and pressure.
The *actual* loss shown on the graphs will be because of the change of oil flow direction, not the media itself.

Conclusion - run whatever filter you want that has reliable construction and good efficiency rates.
The filter will only become a restriction once it starts to get filled with debris. A larger filter should technically be able to hold more foreign matter before pressure drop occurs.

Oil System overview and schematic showing sensor locations:

Wix 51515XP versus Mitsubishi OEM MD356000.
The Filter filter is physically around twice the size of the Mitsi oil filter.

Just to confirm the data above, I found the smallest non-bypass filter I could.
Ryco Z442.

And the results?
The same....

Cold startup is the same.
~10-20kpa pressure differential across the oil filter.

From the testing i've done - the only reason the bypass is opening is when the filter is blocked.
If it's opening - it needs to be changed.

Enjoy!

 

[ZeeOSix]

Nice test setup ... will have to digest the details later. Just to add, but of couse the filter would bypass a lot easier if you did some cold weather start-ups and revved the engine up some.

 

[RSMike]

Nice test setup ... will have to digest the details later. Just to add, but of couse the filter would bypass a lot easier if you did some cold weather start-ups and revved the engine up some.

3000rpm with cold oil (19ºC) is still only 14kpa pressure loss.
Still a long long way away from bypass

 

[ZeeOSix]

You're using a 0W-40, so at 19C it's not really that thick. Try a 10W-xx at below 0C then you'll most likely see the filter bypass at 3000 RPM.

 

[RSMike]

You're using a 0W-40, so at 19C it's not really that thick. Try a 10W-xx at below 0C then you'll most likely see the filter bypass at 3000 RPM.

if that's the case - my question would be why are you using 10W-xx oil at freezing temps?
Why not use 0W-xx or 5W-xx instead?

 

[ZeeOSix]

if that's the case - my question would be why are you using 10W-xx oil at freezing temps?
Why not use 0W-xx or 5W-xx instead?

I'm saying 10W-xx as an example of oil that's thicker at lower temperatures. A 5W might cause some short term filter bypassing if cold enough and engine revs over 3000 RPM. Given extreme conditions, even a 0W could cause a filter to momentarily hit bypass.

Couple other guys here have done similar tests, and also concluded the oil needs to be pretty thick and/or the engine revs need to be pretty high to hit filter bypass delta-p. Of course, this is on new clean filters, but as they load up they would hit bypass easier.

It also depends on the PD oil pump volume output vs engine revs, and the pump pressure relief valve setting. A high volume oil pump with a high pressure relief setting (like on some Subaru engines) have much more potential to make oil filters hit bypass delta-p.

 

[RSMike]

Do you have any links to the previous threads?
I couldn't find anything when I searched last.

 

[ZeeOSix]

Do you have any links to the previous threads?
I couldn't find anything when I searched last.

Look in Jim Allen's threads (link below), and there's 3 or 4 related threads. Some data/graphs are missing due to how they were posted or resided at the time, but reading the text still tells most of the story.

https://bobistheoilguy.com/forums/search/688875/

 

[Job]

Looking at the picture, the oil pressure differential is only at boundaries like the element fabric. After that the oil is in another channel and since the flow stays the same the oil pressure reflects only that chamber at the pressure gauge, nothing else. The only time oil flow "disappears" is when the oil pump leaks or releases oil back to the oil pan. The bypass can open, but the oil recombines downstream and the flow is the same. Wrong? The rest is over my head.
The perfect spacing between the green yellow and blue lines would show the difference in the gallery size, and the blue has had oil flow bled off to the bearings, so is exactly that much lower.

 

[ZeeOSix]

Ch1-Ch2 is the delta-p cross the filter. Only thing that can be concluded is that if Ch1-Ch2 is less than the cracking/opening PSI of the filter bypass valve, then the bypass valve didn't open.

 

[Job]

Ch1-Ch2 is the delta-p cross the filter. Only thing that can be concluded is that if Ch1-Ch2 is less than the cracking/opening PSI of the filter bypass valve, then the bypass valve didn't open.

It is the pressure difference at the exact reading points only. Flow is assumed constant. No other activity in between makes any difference. Pressure is 100% determined by the size of the pipe the gauge is reading. There is inlet pipe, then media and bypass valve, then outlet pipe. Ch1-Ch2 is not the delta-p across the filter, it’s across the inlet and outlet pipes. The delta-p across the filter is across the media fabric thickness regulated by a bypass valve to a maximum value. All oil flow goes into the outlet pipe, it’s past over.

 

[ZeeOSix]

It is the pressure difference at the exact reading points only. Flow is assumed constant. No other activity in between makes any difference. Pressure is 100% determined by the size of the pipe the gauge is reading. There is inlet pipe, then media and bypass valve, then outlet pipe. Ch1-Ch2 is not the delta-p across the filter, it’s across the inlet and outlet pipes. The delta-p across the filter is across the media fabric thickness regulated by a bypass valve to a maximum value. All oil flow goes into the outlet pipe, it’s past over.

Look at the OP's test setup, not the schematic. Red circles are the inlet and outlet pressure sensors on the filter mount, and measure the oil pressure at the very inlet and outlet of the filter ... no external tubing going into our out of the filter is involved. This is the same way the delta-p is measured when a filter goes through an ISO 4548-12 test. No test setup will measure just the delta-p across only the media on an assembled filter. The added delta-p from the inlet base holes and the center tube holes is minuscule, so measuring the delta-p across the entire filter assembly is good enough for a delta-p measurement.

 

[Job]

Look at the OP's test setup, not the schematic. Red circles are the inlet and outlet pressure sensors on the filter mount, and measure the oil pressure at the very inlet and outlet of the filter ... no external tubing going into our out of the filter is involved. This is the same way the delta-p is measured when a filter goes through an ISO 4548-12 test. No test setup will measure just the delta-p across only the media on an assembled filter. The added delta-p from the inlet base holes and the center tube holes is minuscule, so measuring the delta-p across the entire filter assembly is good enough for a delta-p measurement.

View attachment 59185

Not to start an endless argument, but each fitting on the sensors must be a “pipe.” Pressure changes right at the boundary at the fitting. I thought pressure could be measured on the inlet side, then watch if the pressure goes up with restriction increasing with rpm or other reason, then drops, signifying bypass opening. Or if the filter is loading like in the dirt test, inlet pressure gradually builds to the bypass valve rating.
If all the oil is passing through both media and/or bypass, how would such a sensor like above see any difference between them?

 

[ZeeOSix]

^^^ You obviously don't understand this test. Where the pressure sensors are located is as close as possible to the inlet and outlet of the filter. There is no "flow" in the small pressure sensing ports in the filter mount that go to the sensors, it's just a pressurized column of oil. So it's measuring the delta-p across the filter assembly as good as any setup could.

I already explained how you'd have to use the data to determine if the filter bypass valve opened or not.

 

[Job]

^^^ You obviously don't understand this test. Where the pressure sensors are located is as close as possible to the inlet and outlet of the filter. There is no "flow" in the small pressure sensing ports in the filter mount that go to the sensors, it's just a pressurized column of oil. So it's measuring the delta-p across the filter assembly as good as any setup could.

I already explained how you'd have to use the data to determine if the filter bypass valve opened or not.

When you start with the personal stuff it’s done.

 

[ZeeOSix]

Good, you never believe anything I ever say and always want to argue, so you'll have to go find out for yourself. The statement stands, you don't understand the test setup.

 

[dnewton3]

Excellent efforts! Love to see members here actually do some real experimentation and well documented results (as opposed to all the typical bench-racing that goes on here) ...

I'd like to see some data come from long duration use; run those filters out to 10k miles and see what happens as they mature. I generally decided long ago from JAs data that dP in filter is moot; using thinner lubes in modern engines with decent quality filters means bypass is a RARE thing, and the filters generally are never at risk of internal destruction. His data showed that internal bypass events are nearly non-existent, and you have to practically abuse the engine/filter (high revs on thicker oils in a cold engine) to ever get the filter to BP anyway.

I do applaud the work done here. We need more of it! We can benefit from real experiments with well documented results.
The rest of you bench-racers using only theory please take note.

 

[ZeeOSix]

The rest of you bench-racers using only theory please take note.

Some bench racers' have "theories" that match the real data, but some have also seen similar data in other sources and read a lot of papers/studies, and also have engineering backgroungs. So it's easy to see who's a bench racer or not.

Member @Ascent Filtration Testing here, who owns a test lab and does ISO 4548 testing, has also shown some really good testing data. Something you would hardly ever see at that level from any filter manufacturer.

 

[dnewton3]

Some bench racers' have "theories" that match the real data, but some have also seen similar data in other sources and read a lot of papers/studies, and also have engineering backgroungs. So it's easy to see who's a bench racer or not.

Member @Ascent Filtration Testing here, who owns a test lab and does ISO 4548 testing, has also shown some really good testing data. Something you would hardly ever see at that level from any filter manufacturer.

I would agree. I guess when I speak of "bench racers", in regard to filters and dP effects, et al., I'm not referring to anyone who's theory is supported with decent data sources or well documented experiments (even home-grown ones such as this, which I applaud simply for the joy of personal learning). Of course, I guess I'd have to say that if one's "theory" is supported with well documented studies and engineering principles, it ain't really "theory" at that point, is it? Rather, it's "rationale" based on facts, and not "theory". I've read a lot of garbage on this site (specifically in the filter sub-forums) that never gets any support from credible study data, or well designed and applied personal experiments.

To delineate in simple layman's terms ...
- To theorize is to make projections or speculations based on assumptions in the absence of facts and credible data
- To use rationale is to make conclusions based on direct proof from identical (or very similar) results of other circumstances/conditions

There's nothing wrong with using theory. But to accept theory as fact is bad science. Often, the "theory" we read about here is just so poorly reasoned that it hurts to read at times. Some folks here don't even understand the basics of engineering principles in regard to hydraulics. But they'll surely try to convince us that XYZ happens, and not have anything in the barrel to back it up.

Experiments at home, such as this one and Jim's, at least exhibit some form of commitment and glean data (results) to prove/disprove a theory. Professional labs such as the one you linked also have significant merit.
Those are a far cry from some folks on here which vomit up some poorly thought out dribble and then walk away as if nothing is worthy of challenge.

(Zee - I had not intended to include you in the masses; if you took offense then I apologize. But I stand my ground; there are too many "hit and run" comments made here about filters and filtration effects with little to no data to back them up).


Oh, and RSMike, PLEASE run some of the same dP experiments with used filter (perhaps out to 10k miles). Let's see what happens when real life is included in the mix.

 

[Job]

I would agree. I guess when I speak of "bench racers", in regard to filters and dP effects, et al., I'm not referring to anyone who's theory is supported with decent data sources or well documented experiments (even home-grown ones such as this, which I applaud simply for the joy of personal learning). Of course, I guess I'd have to say that if one's "theory" is supported with well documented studies and engineering principles, it ain't really "theory" at that point, is it? Rather, it's "rationale" based on facts, and not "theory". I've read a lot of garbage on this site (specifically in the filter sub-forums) that never gets any support from credible study data, or well designed and applied personal experiments.

To delineate in simple layman's terms ...
- To theorize is to make projections or speculations based on assumptions in the absence of facts and credible data
- To use rationale is to make conclusions based on direct proof from identical (or very similar) results of other circumstances/conditions

There's nothing wrong with using theory. But to accept theory as fact is bad science. Often, the "theory" we read about here is just so poorly reasoned that it hurts to read at times. Some folks here don't even understand the basics of engineering principles in regard to hydraulics. But they'll surely try to convince us that XYZ happens, and not have anything in the barrel to back it up.

Experiments at home, such as this one and Jim's, at least exhibit some form of commitment and glean data (results) to prove/disprove a theory. Professional labs such as the one you linked also have significant merit.
Those are a far cry from some folks on here which vomit up some poorly thought out dribble and then walk away as if nothing is worthy of challenge.

(Zee - I had not intended to include you in the masses; if you took offense then I apologize. But I stand my ground; there are too many "hit and run" comments made here about filters and filtration effects with little to no data to back them up).


Oh, and RSMike, PLEASE run some of the same dP experiments with used filter (perhaps out to 10k miles). Let's see what happens when real life is included in the mix.

I'll bite. Pounds per square inch varies according to the size of the "pipe" with the same flow rate. An oil filter has an inlet pipe and an outlet pipe. There is no other way in or out of an oil filter. I don't need a test setup to know this, it is not theory, it's very basic physics, units of measurement, math. Wrong? Why? I don't mind being wrong, and don't need to be right.
People leave a discussion as it devolves into personal attack bickering, which is bullying. Because of someone needing to "win", when it's no longer about the science, and they are the wrong ones? No wonder some very good resources in people have chosen to leave this site. The bullying to "win" needs to stop. People should feel free to ask and ponder without being personally judged, that's good discussion. I guarantee you there are people here who think twice about asking something or challenging something because they know what is coming at them if they do. That's wrong.
Board racer