Fram XG7317 after 25000 km.

Status
Not open for further replies.
Originally Posted by ZeeOSix
Originally Posted by Davejam
Yeah, as particles build up on a filter the filtration efficiency will increase.
Pressure drop will increase and therefore flow decrease, but as long as a filter is designed for extended use you will be ok.
So yes, wear should decrease.

Pushing things beyond filter/vehicle recommendations is at your own risk. But if you know what you're doing you'll be fine and perhaps better off.


Oil filters and air filters behave differently when they start loading up and the delta-p really starts to increase. Oil filters experience way more delta-p than air filters.

Studies show that some oil filters can start sloughing off captured particulate as the delta-p becomes higher and higher from debris loading (see graph below). So in that case, the filter is actually becoming less efficient as it loads up. A filter would have to be very good at retaining captured debris and become more efficient as the delta-p becomes higher.

I wouldn't run a filter much past the rated use indicated by the manufacturer unless it was on a very clean engine. As far as the OP's use ... 25K kilometers is still within Fram's rating.

And just because the delta-p becomes higher doesn't mean you're losing oil flow. Oiling systems use a positive displacement oil pump, so if the filter becomes a bit more restrictive the oil pump still forces the same volume of oil through it. Engine oiling systems are not like a water system in your house where a clogged filter will reduce the flow out of the water faucet.


Zee - when you get a chance can you please provide a link to the study/article that produced that chart? Is it an SAE study, or what? I'd like to get a copy so I can read the entire item.
 
Originally Posted by dnewton3
Originally Posted by ZeeOSix
Originally Posted by Davejam
Yeah, as particles build up on a filter the filtration efficiency will increase.
Pressure drop will increase and therefore flow decrease, but as long as a filter is designed for extended use you will be ok.
So yes, wear should decrease.

Pushing things beyond filter/vehicle recommendations is at your own risk. But if you know what you're doing you'll be fine and perhaps better off.


Oil filters and air filters behave differently when they start loading up and the delta-p really starts to increase. Oil filters experience way more delta-p than air filters.

Studies show that some oil filters can start sloughing off captured particulate as the delta-p becomes higher and higher from debris loading (see graph below). So in that case, the filter is actually becoming less efficient as it loads up. A filter would have to be very good at retaining captured debris and become more efficient as the delta-p becomes higher.

I wouldn't run a filter much past the rated use indicated by the manufacturer unless it was on a very clean engine. As far as the OP's use ... 25K kilometers is still within Fram's rating.

And just because the delta-p becomes higher doesn't mean you're losing oil flow. Oiling systems use a positive displacement oil pump, so if the filter becomes a bit more restrictive the oil pump still forces the same volume of oil through it. Engine oiling systems are not like a water system in your house where a clogged filter will reduce the flow out of the water faucet.


Zee - when you get a chance can you please provide a link to the study/article that produced that chart? Is it an SAE study, or what? I'd like to get a copy so I can read the entire item.


Here ya go (PDF link): http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.709.8915&rep=rep1&type=pdf

Note that the graph in Fig 1 is data from a bench test (ISO 4548-12 testing with particle counters), and then in Part B they talk about modeling the media so they can replicate and explain the mechanism of decreasing efficiency with media loading that they measured in the lab.
 
Originally Posted by ZeeOSix
Here ya go (PDF link): http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.709.8915&rep=rep1&type=pdf

Note that the graph in Fig 1 is data from a bench test (ISO 4548-12 testing with particle counters), and then in Part B they talk about modeling the media so they can replicate and explain the mechanism of decreasing efficiency with media loading that they measured in the lab.


Also see factor 4 in this article. Just because an oil filter captures a particle doesn't mean it's going to be retained forever. ISO 4548-12 testing takes into account the delta-p factor as the filter loads up, but there could also be pulsations from the oil pump (depending on how it's designed) that could cause captured particles to dislodge and go back through the oiling system. The design of the media has a lot to do with how well it will retain captured particles, as much as how well it captures the particles in the first place.

https://www.machinerylubrication.com/Read/29272/oil-filtration-factors
 
That Fram Ultra held up well for you. I posted a while back showing my XG10060 after just shy of 20,000 miles. The only thing I noticed was just a bit of soot in the dome of the can. I probably won't do that again as I'll just stick to 2 OC's between filter changes once I get an Ultra back on my car.

L8R,
Matt
 
Yes, I remember that study now. Thanks guys.

I have to take a pessimistic view of the applicability of conclusions drawn; probably does not surprise any of you. I am being specific here. It's not that I doubt the study data or results. It's that I doubt it matters in the real world. The data is good, but biased, as are most HALT simulations. The bias is clearly stated right in the abstract ...
"By interspersing filter efficiency simulations with many more particles than would actually arrive at the media during life time simulations, the quality of the filtration statistics of media scale simulations can be improved such that a quantitative comparison with experimental data becomes possible."
Specifically this: "... with many more particles than would actually arrive at the media during lifetime simulations ..."
IOW - if they didn't overload the media, they'd not be able to tell a difference in filter efficiency effects.
This is a self-fulfilling prophesy. It goes like this ...
We can't tell how well a filter is working, so let's significantly turn up the contamination loading to induce a difference. Only then will we be able to reveal a disparity in performance.
Oh - by the way ... the rate we use is not representative of normal contamination rates in an engine, because nothing gets that dirty in real world use during a normal OCI.

This is not unlike the infamous GM filter study. To induce an effect they could measure (actually distinguish), they had to grotesquely load up the sump with such a burden that the differences in component wear would become apparent. Here in the M/H study, they essentially do the same thing.

Again - I state this for all to read. HALTs are tools to reveal disparity in products or components, but they do NOT necessarily represent the reality of how things react in the real world. HALTs are purposely made to bring forth differences, but do so in conditions that border on the absurd. I deal with HALTs all the time in our testing at work. I often have to warn others of drawing improper conclusions, and I do so now with you all.

I appreciate what HALTs can tell us, but I also understand that the conditions HALTs operate in do not exist in my garage. I agree that the M/H study shows filter efficiency decreasing over time, but that's because they are loading it to a level that is not realistic. My engine, your engine, pretty much any engine made in the last couple decades, does not run that dirty. Efficiency speaks to the potential of a filter to trap particles. But it does NOT assure that the actual contamination rate will be high enough to illuminate the difference between media choices.

Think of this as a conditional "If... Then..." statement.
IF the contamination loading is monstrously high enough, THEN we can discern a difference in filtration efficiency.
But when the initial condition does not exist, it's not an assurance that the expected result will be present.
What we know is that when M/H heavily doses the fluid stream with particulate, the filtration efficiency will degrade slightly over time.
What we do not know is how that media effect presents itself when the media is NOT overly saturated with unrealistic burden.
Could it not be possible that lighter loading results in better retention?

Think of this example. Imagine a toll gate area with three booths, each with an operator that can pass 1000 cars per hour, with a wait time of only 30 seconds per occurrence; that would represent the design intent of the planners for the anticipated peak processing time of day (rush hour). That represents the realistic expected rate of passing vehicles and wait times. There will be a disparity of natural human condition that exhibits some differential in processing at each booth, but because the system was designed as a whole to process the expected flow, there's little ability to discern the difference in toll gate efficiency overall. But now induce a vehicle presence such that it is 10x greater than reality. Try to put 10,000 cars through that same toll gate system in an hour, and then measure the wait time per occurrence. Some operators may become quicker, some slower, because the pressure on the system is now well beyond any design intent. Because the gate system was not designed to accommodate the greatly increased loading, the disparity of result becomes unnaturally influenced. The example of the toll gate system illuminates the incongruity of conclusions from non-standard conditions. When you dump too many cars into the toll gate system, the wait time per vehicle greatly increases because the processing system does not get greatly changed; humans won't go 10x faster just because there are 10x more cars present.

That is how HALTs can induce differences; they operate in a condition that is unexpected in "normal" life. Until you bias the inputs, there will not be a reasonable disparity in the results which you could measure. Attempting to quantify a result by unnaturally testing a system really does not give any indication how the system operates in the "normal" conditions.

Most simply put, both the GM filter study (speaking to wear relative to media efficiency) and the M/H study (speaking to particle retention relative to time) must rely on bombarding the media with loading so heavy it induces the result they seek. If that precursor is not present, they cannot tell a difference in the media effects. Because engines don't actually run as dirty as the study conditions, the conclusions drawn from the study are very unlikely to manifest into differing results in your garage.

The M/H study proves that when a system is overloaded past a reasonable expectation of normality, it will behave in a certain manner.
But it does not prove that same result is likely when "normal" conditions are present; that would be an improper conclusion to draw.
 
Last edited:
WOW those things are like TANKS... I have always wondered with all that filter media does it put more strain on the oil pump? I know Honda VTEC is run by oil pressure and that's why many Honda guys like the A01s.

I mean there is just so much filter media that oil has to push through... Now don't get me wrong.. Im sure its filtering the oil very well... There has to be some cost... Like I like FLOW.

I will never forget my 1997 Buick Lesabre .. I put a PUREONE oil filter and at idle the oil pressure gauge was nearly in the red and as soon as I hit the gas it would go to the middle where it was when I was using the AC PF47

I remember getting a new PuroOne oil filter and swapping it and it did the same thing and it really scared me.. Now my Gramps gave me that car and it only had 38K miles on it.. very nice engine 3800 V6 Series II.

Anyways.... People were telling me I needed a new oil pump and others were saying it was the oil pressure gauge and or switch....

Anyways... I took the 2nd PuroOne oil filter off and put on a new Delco PF47 and as soon as I started the engine the needle was perfect in the middle of the gauge... TO THIS DAY I WILL NEVER FORGET THAT.

My grandpa laughed his butt off at me and said I have always told you A/C Delco Parts...

My Gramps had the coolest job for G.M back in the Good ole days before computer simulations. He used to be in R&D and his tests were REAL WORLD....

He loved it!

anyways back to the point. Is there any problem with the oil having to go through all that filter media???
 
For the record...GM 3800 Series II did not have oil pressure issues with aftermarket filters unless they were otherwise damaged.

The only time i have seen them with oil pressure issues is when they would hydro lock early in life to GMs faulty EGR passage in the intake. Every one we ever did went like this. Motor locks up out of the blue at fairly low milage. Intake replaced and motor disassembled enough to un hydro lock it. Re-assemble with redesigned dorman (later OEM) parts and experience hot idle low oil pressure.

They still ran forever. Only way to band aid it our personal cars was to run 50wt oil. Worked like a charm. My father ran one on 50wt due to wiped bearings for over 200k.


Most went before 40,000 miles and GM was congratulated with a class action lawsuit.

Of our personal cars this happened to a Park Ave and a LeSabre. The park ave bent a rod and we had to replace the engine.
 
Last edited:
Originally Posted by dnewton3
Most simply put, both the GM filter study (speaking to wear relative to media efficiency) and the M/H study (speaking to particle retention relative to time) must rely on bombarding the media with loading so heavy it induces the result they seek. If that precursor is not present, they cannot tell a difference in the media effects. Because engines don't actually run as dirty as the study conditions, the conclusions drawn from the study are very unlikely to manifest into differing results in your garage.

The M/H study proves that when a system is overloaded past a reasonable expectation of normality, it will behave in a certain manner.
But it does not prove that same result is likely when "normal" conditions are present; that would be an improper conclusion to draw.


I've snipped a lot of your reply, and agree for the most part. I think the take away here is for the guys who want to run an oil filter for way past what it's been rated for, it might not be a good idea. And for the guys who always claim that "an oil filter gets more efficiency with use" mantra that it's probably more false than true. I do believe for some filters, depending on the media design and debris holding as a function of delta-p, that running them way past their rating could cause debris shedding and therefore a decrease in efficiency as shown in the M+H/Purolator graph. There was also another study (can't recall by who, but it was a filter manufacture) that did pulsation testing and found depending on the media design some filters shed lots of captured particles when hit with pressure pulsations (ie, essentially short duration high delta-p events).

As I've mentioned before, I have not seen one study that proves that oil filters get more efficient with loading, regardless if it's loaded fast or slow. IMO, if you slowly loaded an oil filter with debris until it was close to going into bypass then the delta-p will be enough to shed some captured particles, and therefore decrease the filters efficiency to some degree as the M+H/Purolator graph shows. Some medias may be less sensitive to this than others. So running $3 cellulose media filters for 25K miles might not be a great idea.
 
Originally Posted by dnewton3
Again - I state this for all to read. HALTs are tools to reveal disparity in products or components, but they do NOT necessarily represent the reality of how things react in the real world. HALTs are purposely made to bring forth differences, but do so in conditions that border on the absurd. I deal with HALTs all the time in our testing at work. I often have to warn others of drawing improper conclusions, and I do so now with you all.

This is all very helpful and informative.

My view has always been that better efficiency is great, with all other things being equal or relatively so. A high efficiency filter over a short OCI isn't worth triple an "ordinary" filter. A high efficiency filter over a short OCI with an identical price to the "ordinary" filter might be. As you've pointed out many times here, the average filter out there (and lube) is vastly underutilized, aside from cases of neglect.
 
Originally Posted by ZeeOSix
Originally Posted by dnewton3
Most simply put, both the GM filter study (speaking to wear relative to media efficiency) and the M/H study (speaking to particle retention relative to time) must rely on bombarding the media with loading so heavy it induces the result they seek. If that precursor is not present, they cannot tell a difference in the media effects. Because engines don't actually run as dirty as the study conditions, the conclusions drawn from the study are very unlikely to manifest into differing results in your garage.

The M/H study proves that when a system is overloaded past a reasonable expectation of normality, it will behave in a certain manner.
But it does not prove that same result is likely when "normal" conditions are present; that would be an improper conclusion to draw.


I've snipped a lot of your reply, and agree for the most part. I think the take away here is for the guys who want to run an oil filter for way past what it's been rated for, it might not be a good idea. And for the guys who always claim that "an oil filter gets more efficiency with use" mantra that it's probably more false than true. I do believe for some filters, depending on the media design and debris holding as a function of delta-p, that running them way past their rating could cause debris shedding and therefore a decrease in efficiency as shown in the M+H/Purolator graph. There was also another study (can't recall by who, but it was a filter manufacture) that did pulsation testing and found depending on the media design some filters shed lots of captured particles when hit with pressure pulsations (ie, essentially short duration high delta-p events).

As I've mentioned before, I have not seen one study that proves that oil filters get more efficient with loading, regardless if it's loaded fast or slow. IMO, if you slowly loaded an oil filter with debris until it was close to going into bypass then the delta-p will be enough to shed some captured particles, and therefore decrease the filters efficiency to some degree as the M+H/Purolator graph shows. Some medias may be less sensitive to this than others. So running $3 cellulose media filters for 25K miles might not be a great idea.


Agreed.

I think there are some filters that do get more efficient with time exposure. Air filters do, as I understand. Also, depth media oil and hydraulic filters. But most all oil filters I'm aware of for common vehicle applications are surface media loading, not depth loading. Surface loading particle retention efficiency is probably a bit of a hybrid condition. When new, the first few hours will increase in effiency because the loading is truly creating a tighter pore effect, but as it matures, it's shedding on the backside. The question becomes this:
When is that tipping point occurring? When does it go from getting more efficient to less efficient?
The M/H study does not really give us a true understanding, because their loading burden is grossly exaggerated.
It it quite possible that for "normal" applications, the filters at least sustain a reasonable effect for the intended duration of use.
I would agree also that using a filter for a duration (in terms of flow exposure, not time on a calendar) far longer than its design intent is not a good idea.
 
Originally Posted by dnewton3
The question becomes this:
When is that tipping point occurring? When does it go from getting more efficient to less efficient?
The M/H study does not really give us a true understanding, because their loading burden is grossly exaggerated.
It it quite possible that for "normal" applications, the filters at least sustain a reasonable effect for the intended duration of use.
I would agree also that using a filter for a duration (in terms of flow exposure, not time on a calendar) far longer than its design intent is not a good idea.


The filter in the M+H graph shows the 20u efficiency dropping pretty early in the loading period. Even though the loading is much faster than in real use, I'd think that engineers doing this study (and all the engineering committees reviewing this information) wouldn't even consider this data if they didn't think it was somewhat realistic, even though it's accelerated loading. Sure, a filter in real use on the road may only load up to 25% of the total possible loading performance level of the filter. In that case there would be less decreasing of efficiency compared to allowing the filter to totally load up causing the delta-p to become very high and causing more debris shedding.

Keep in mind that the ISO 4548-12 test resulting efficiency number is the average of the start to end efficiency measurement. So in the M+H graph example the 20u particles start out at ~90% and end up at ~60%, which gives an average 20u efficiency of 75%. If a filter has a high ISO 4548-12 efficiency at 20u, then by definition it can't be shedding many particles in order to achieve that ISO efficiency rating.
 
With a GDI engine my OCI's are approx. 4K miles with a Fram Ultra (which I will run for two OCI's) ... GDI engines produce a fair amount of soot so I figure if I get a couple of reasonable OCI's out of a $10.00 Fram Ultra then I got my money's worth out of it (as compared to the inferior OEM filter costing approx. $6.00 ~ $7.00 ea.)
 
Status
Not open for further replies.
Back
Top