Non-destructive Ruffles® Bypass Valve testing

[Hatsuwr]

Since apparently now relief valves have ridges, I was thinking about testing them before installation.

I've seen mentioned to try putting a small amount of oil into the bottom of the filter and seeing how long it takes to leak through the bypass plate interface.

Not related to the bypass, but you can also blow air into the outlet of the filter to check for ADBV function. Seems like a low pressure would be preferred for the ADBV test, since higher pressures could improve the sealing beyond what you would see in actual use. Might also put some oil on the outlet ports to help see leaking.


Besides this (and before introducing any oil), I'm wondering if it would be possible to do a comparative test by attaching vacuum to the outlet of the filter and having a tee connected to a gauge to measure the pressure differential. Defects usually don't seem to be of the sort that increase restriction, so lower pressure differentials would indicate tears and leaks. I don't know how significant a difference you could expect from the bypass gaps we've been seeing though.

Unfortunately, my vacuum can only generate about 3 PSI max, and that's with no airflow. I'd rather be sitting around 80% of the bypass pressure. Also, the closest relevant gauge I have is 15 PSI max, so resolution would be lacking. A low range transducer would probably be ideal, since the accuracy of the reading isn't really relevant, just the precision and repeatability. Maybe actually saturating the filter media with oil first and then draining would help increase the pressure differential, and since the bypass ridges would clear of oil more easily, it could amplify the impact of those leaks.

I'll pick up a few filters tomorrow and try it out with the gauge, but I'm wondering if you all might have some better ideas.

 

[ZeeOSix]

Besides this (and before introducing any oil), I'm wondering if it would be possible to do a comparative test by attaching vacuum to the outlet of the filter and having a tee connected to a gauge to measure the pressure differential. Defects usually don't seem to be of the sort that increase restriction, so lower pressure differentials would indicate tears and leaks. I don't know how significant a difference you could expect from the bypass gaps we've been seeing though.

Yes, It would be really hard to distinguish what's causing the resulting vacuum level. Only way to maybe get an idea is if you took one of the same model and near build date and drop some epoxy in the bottom of the center tube to 100% seal the leaf spring. Then test that one as the baseline to compare ones that haven't been sealed with epoxy.

 

[Hatsuwr]

Yes, It would be really hard to distinguish what's causing the resulting vacuum level. Only way to maybe get an idea is if you took one of the same model and near build date and drop some epoxy in the bottom of the center tube to 100% seal the leaf spring. Then test that one as the baseline to compare ones that haven't been sealed with epoxy.

I think the only value in this method would be for comparative testing a batch in one sitting, since things like temperature and humidity would probably be significant factors, not to mention the state of the vacuum. So pretty much for buying a few filters and finding the best of them to install/save. No need for any baselines since it would be difficult to reproduce initial conditions and you can't really account for even small changes made to the filter construction over time.

 

[ZeeOSix]

I think the only value in this method would be for comparative testing a batch in one sitting, since things like temperature and humidity would probably be significant factors, not to mention the state of the vacuum. So pretty much for buying a few filters and finding the best of them to install/save. No need for any baselines since it would be difficult to reproduce initial conditions and you can't really account for even small changes made to the filter construction over time.

You would need a known non-leaker as a baseline (the one fixed with an epoxy seal job), otherwise you have nothing to base the comparison of the others tested on. And it would only work for the same make and model of oil filter.

 

[Hatsuwr]

You would need a known non-leaker as a baseline (the one fixed with an epoxy seal job), otherwise you have nothing to base the comparison of the others tested on. And it would only work for the same make and model of oil filter.

Yea there's no way this would work for comparing models, only for comparing a group of the same model.

If you picked up 3 filters of the same model, for example, I think it's a workable assumption to consider the restriction from the media to be roughly equivalent among them. A difference in pressure differential would then indicate a difference in flow rate through media tears, bypass plate seal failure, or other leak sources.

 

[ZeeOSix]

Yea there's no way this would work for comparing models, only for comparing a group of the same model.

If you picked up 3 filters of the same model, for example, I think it's a workable assumption to consider the restriction from the media to be roughly equivalent among them. A difference in pressure differential would then indicate a difference in flow rate through media tears, bypass plate seal failure, or other leak sources.

That's why you need a non-leaker as a baseline. Otherwise, you're just comparing 3 that leak, and have no idea what the effect of the leak path is on the vacuum level seen.

 

[CharlieBauer]

Yea there's no way this would work for comparing models, only for comparing a group of the same model.

If you picked up 3 filters of the same model, for example, I think it's a workable assumption to consider the restriction from the media to be roughly equivalent among them. A difference in pressure differential would then indicate a difference in flow rate through media tears, bypass plate seal failure, or other leak sources.

My two XG2's leaked oil through the gap at roughly the same rate so you wouldn't know.

The ISO test requires a filter to be not leaking before testing which requires the manufacturer to provide the differential pressure spec.

 

[Hatsuwr]

I guess these days it might actually be optimistic to expect at least 1/3 to not leak...

 

[Klutch9]

::laughs in cartridge filter::

 

[dnewton3]

You don't need one filter for a baseline value; you need 30 of them.
Then you need 30 of the tested characteristic ones as well.
Remember ... small sample sets lend themselves to highly misleading answers because of poor stdev accuracy.

I suspect the differences are going to be fairly small; hence you'd also need a fairly accurate vacuum gauge (not something from the local auto parts store).
And you'll need to have a good idea of the R&R for that gauge; most folks don't know how to design/conduct a decent R&R trial.


IOW - this is just a venture into the realm of YT "science", because it's unlikely the OP is able to adequately satisfy the criteria for a reliably credible answer to his quest.

 

[CharlieBauer]

You don't need one filter for a baseline value; you need 30 of them.
Then you need 30 of the tested characteristic ones as well.
Remember ... small sample sets lend themselves to highly misleading answers because of poor stdev accuracy.

I suspect the differences are going to be fairly small; hence you'd also need a fairly accurate vacuum gauge (not something from the local auto parts store).
And you'll need to have a good idea of the R&R for that gauge; most folks don't know how to design/conduct a decent R&R trial.


IOW - this is just a venture into the realm of YT "science", because it's unlikely the OP is able to adequately satisfy the criteria for a reliably credible answer to his quest.

Right. With 30 measurements and filter modelling software, you would be able to determine the worst case hit to efficiency.

Or you could just pour oil down and any leak means don't use the filter.

 

[Hatsuwr]

I'm thinking there there might be some misunderstanding of the purpose of a test like this. It is not to compare some random filter to a known good filter, it is to compare a batch of filters of the same model to identify differences in the amount of leakage present.
The one assumption it really relies on is the assumption that the media of the filters is consistent enough that leakage paths are a larger factor in differences in restriction. At that point, if you have a significant airflow source and ability to measure the small differences in pressure, it seems like there would be a pretty clear correlation between the pressure differential and the size of the leaks.

The gauge I have on hand is a 4.5" 15 PSI Winters gauge. I imagine it would work well if I was able to generate a pressure differential closer to 10 PSI. I'll be testing that out today, but I don't expect much. I have a 7.25 PSI transducer added to my next Digikey order, but testing with that will have to wait until I get back from a trip I have coming up.

 

[Reloader59]

Just do the test the way you want.

 

[ZeeOSix]

I'm thinking there there might be some misunderstanding of the purpose of a test like this. It is not to compare some random filter to a known good filter, it is to compare a batch of filters of the same model to identify differences in the amount of leakage present.

If your goal if to be able to measure the effect of the air leakage through the gaps between the leaf spring and end cap, you will need a baseline where you know there is zero leakage at that interface. Otherwise, the test is nebulous and proves nothing.

The one assumption it really relies on is the assumption that the media of the filters is consistent enough that leakage paths are a larger factor in differences in restriction. At that point, if you have a significant airflow source and ability to measure the small differences in pressure, it seems like there would be a pretty clear correlation between the pressure differential and the size of the leaks.

Once you have a baseline on a non-leaker, then you can compare all the other filters (same brand and model of course) to the baseline results. How are you going to rate them if you don't have a known non-leaking leaf spring to end cap baseline?

 

[Ronn]

You would need a known non-leaker as a baseline (the one fixed with an epoxy seal job

Yes
Just ask BR to send over the Endurance they particle tested, definitely a “NON-LEAKER”

 

[ZeeOSix]

Just ask BR to send over the Endurance they particle tested, definitely a “NON-LEAKER”

No proof ... only a wish in your imagination.

 

[Ronn]

No proof ... only a wish in your imagination.

Show me any Endurance that has passed a light leak test….let alone ANY of the “clones” made from 2023 on….RP…Amsoil…Supertech

 

[Ronn]

No proof ... only a wish in your imagination.

Got it backwards you’re the one wishing it wasn’t a “LEAKER”, because if it was , you’ve got some “splainin’ to do” lol

 

[ZeeOSix]

Got it backwards you’re the one wishing it wasn’t a “LEAKER”, because if it was , you’ve got some “splainin’ to do” lol

I could care less if it was a leaker or not, because I'm not a die hard Endurance fanboy and don't have one installed on any vehicles, lol. My claim is simply that nobody knows if the one BR tested was a leaker or not because there's zero proof one way or the other. And besides, their test data ranking is inconsistent with official ISO efficiency rankings of the same filters. So bottom line is you can't take their testing as the golden standard because it's not.

 

[Ronn]

could care less if it was a leaker or not, because I'm not a die hard Endurance fanboy and don't have one installed on any vehicles, lol.

This isn’t about me being a “fan boy”…
You’re implying that the leak substantially impacts performance, which in all probability, is contrary to a “non sanctioned“, albeit well conducted test. Of course …”no proof” it was a leaker, but highly likely and test Isn’t a “gold standard”. However, BR was a well constructed test “guided” by ISO standards using ISO particles….meticulously cleaning all equipment which had been upgraded as well. So, even though it’s not “the ultimate” standard, the results point to superior particle filtration, especially when compared to others using the same methods. No, I’m not talking about filter “ranking” either.
“Walks like a duck…Quacks like a duck”, but it we can’t call it one because we need more proof. Got it.