Particle counts:
- Thread starter DuckRyder
- Start date
Volkswagen 04E 115 561 H
- 24/21/15 @ 9,472 Miles
I am the BITOG member who conducted the study of particle counts with the various oil filter brands posted by
paulri. I tried very hard to hold as many variables as I could constant while conducting that study. My conclusion was - the trends were not strong enough or consistent enough to draw a real conclusion from this. Admittedly, I only used premium filters for the study. One thing I did think was interesting is both Amsoil filters I tried showed slightly worse results than the others, but I still don't think it was clear enough to draw a conclusion - in fact, I still use Amsoil filters.
I think in order to produce clear, consistent, and repeatable results, it's going to take an engine in a laboratory setting where everything can be held constant except the filters under study.
paulri. I tried very hard to hold as many variables as I could constant while conducting that study. My conclusion was - the trends were not strong enough or consistent enough to draw a real conclusion from this. Admittedly, I only used premium filters for the study. One thing I did think was interesting is both Amsoil filters I tried showed slightly worse results than the others, but I still don't think it was clear enough to draw a conclusion - in fact, I still use Amsoil filters.
I think in order to produce clear, consistent, and repeatable results, it's going to take an engine in a laboratory setting where everything can be held constant except the filters under study.
By the chart the Micro green bypasses the larger particles and catches the smaller particles or am I mistaken?
My interpretation of the Microgreen PC is that the high efficiency "microdisk" does filter out more smaller particles below 6 microns, and the main element isn't as efficient and therefore just does average filtering of particles greater than 15 microns.
You get a more accurate sample of the working fluid. No static sump sample is ever to be considered a legitimate condition of a working fluid.
Don't get all into these ISO codes ( especially for an ICE) because without proper sampling protocols, they are basically meaningless. (not meaningless because the numbers aren't good but meaningless because the sampling port location, proper purging and running condition of the sample determine whether the sample is truly respective of the numbers)
Many people quote ( and often weaponize) the standards with no clue whatsoever what they are talking about in terms of the meaning and protocols.
- Joined
- May 28, 2014
- Messages
- 2,264
On our Escape, there's only enough clearance for the dipstick, no tube is even visible.
Last edited:
That is probably the most practical way on the average passenger car or truck.
Best - No...
What is the best way?
The ultimate ( and really only) goal of an oil sample proper is to fully represent of the condition of the lubricant in the application.
Without this, no "analysis" of the oil can or will produce any information of any value because the information in the sample is skewed.
That's a plain and simple inarguable fact.
For the sample to be accurate- it must represent what is in the working fluid.
That's easy to write an article for Noria and say that with pretty pictures but a lot more difficult to achieve in the real world when machine designers do not make it easy ( or even possible in some cases) to do.
The truth is that by the "Best Sampling Practices" taught at every level of the ML's certifications and accepted as standards, an ICE simply isn't designed or capable of getting a true sample without some modification. ( which few do)
The basic circuit goes from the pick up through the pump, filter crank ports and valves then drains back to the sump. Its not a controllable circuit like on a screw compressor nor have defined zones like a gearbox or a route like a HPU. Sometimes it takes multiple sample ports to get a true sample too.
In an ICE it generally rains down into the sump- then there is no agitation of the sump so things settle and are not picked up by a sample ( or never move during draining or move all at once- either one will skew the sample data)
All that to point this out.
Particle counts are extremely important in terms of condition assessment of the machine and the fluid- BUT- that only holds true when the sampling protocol and sample contain a true representation of what's circulating. Otherwise you are just reading numbers on a sheet that have no real meaning or significance.
Without this, no "analysis" of the oil can or will produce any information of any value because the information in the sample is skewed.
That's a plain and simple inarguable fact.
For the sample to be accurate- it must represent what is in the working fluid.
That's easy to write an article for Noria and say that with pretty pictures but a lot more difficult to achieve in the real world when machine designers do not make it easy ( or even possible in some cases) to do.
The truth is that by the "Best Sampling Practices" taught at every level of the ML's certifications and accepted as standards, an ICE simply isn't designed or capable of getting a true sample without some modification. ( which few do)
The basic circuit goes from the pick up through the pump, filter crank ports and valves then drains back to the sump. Its not a controllable circuit like on a screw compressor nor have defined zones like a gearbox or a route like a HPU. Sometimes it takes multiple sample ports to get a true sample too.
In an ICE it generally rains down into the sump- then there is no agitation of the sump so things settle and are not picked up by a sample ( or never move during draining or move all at once- either one will skew the sample data)
All that to point this out.
Particle counts are extremely important in terms of condition assessment of the machine and the fluid- BUT- that only holds true when the sampling protocol and sample contain a true representation of what's circulating. Otherwise you are just reading numbers on a sheet that have no real meaning or significance.
Although I apprehend your premise, I struggle a little with some aspects.
The composition of the oil in the sump must necessarily correlate with what is circulating through the lube system, because it IS what’s circulating. If there are things which drop to the bottom of the pan once they rain down out of the engine, and they don’t wash out, how can it be said those things are “circulating?”
Perhaps the question, then, is how strongly a bulk sump sample‘s composition correlates with fluid composition at a given point in the engine. As a corollary: What are the specific differences, if any, and how can bulk sump samples not be correlated?
You assert that bulk sump samples aren’t good or useful. That’s fine, but not particularly helpful information. Would you like to elaborate on exactly what one would do to achieve what you believe to be a ‘good’ sample? At a minimum, that would distinguish you from Noria’s articles. (-:
To be clear: I have said here before and say again that engine oil UOAs are not cost-effective on personal transportation vehicles, and don’t tell anything like what many seem to think. They’re for hobby or entertainment only. Nevertheless, since most here have some interest, your elaboration could be helpful.
The composition of the oil in the sump must necessarily correlate with what is circulating through the lube system, because it IS what’s circulating. If there are things which drop to the bottom of the pan once they rain down out of the engine, and they don’t wash out, how can it be said those things are “circulating?”
Perhaps the question, then, is how strongly a bulk sump sample‘s composition correlates with fluid composition at a given point in the engine. As a corollary: What are the specific differences, if any, and how can bulk sump samples not be correlated?
You assert that bulk sump samples aren’t good or useful. That’s fine, but not particularly helpful information. Would you like to elaborate on exactly what one would do to achieve what you believe to be a ‘good’ sample? At a minimum, that would distinguish you from Noria’s articles. (-:
To be clear: I have said here before and say again that engine oil UOAs are not cost-effective on personal transportation vehicles, and don’t tell anything like what many seem to think. They’re for hobby or entertainment only. Nevertheless, since most here have some interest, your elaboration could be helpful.
Sure, I'll be glad to drill down a bit
Not by a long shot. In reality in liquids containing solids (and we are talking about both particle size as well as density distribution) the "hang time" is a combination of things like weight, heat, fluid properties and like with all things gravity- they will eventually settle.
Once they settle, unless you have proper agitation, they are going to stay settled- the only variable is time. ( I do lots of mixing studies and agitation and there is no "mixing" or "agitation" in any standard sump)
2 critical things about those particles ( when suspended in the stream)- I take the stream sample, I have a "real time" indication of the particle generator ( leads toward machine condition assessment) in terms of what is wearing and the frequency of the wear.
When in a sump, I am getting an agglomerated combinate collage of everything- useless for any legitimate data driven assessment of a machine.
you don't want "circulation" in a sump- you want it to settle an stay there- but when a sample is drawn from the drain (where the velocity may be ample to pick them up and carry them like a current over a silt bed)- it is not a true estimation of the running system condition.
This is why all the ML- trainings tell you to purge the sample ports, mini-mess valves, Luneta or whatever to keep from skewing the sample.
That's the problem. They cannot be without substantial modification. If this were an HPU or similar device I would sample at:
Post filter- measure performance
post crank- bearing wear
top drain- valve train wear
Sump- General chemistry
A plan similar to that would give me a reasonable idea of critical item wear in the major areas ( as good as any OA program can give)
A ICE simply isn't set up like that to allow for a single representative sample.
I don't need distinguishing from Noria, I have bested them many times. I have great respect for Jim but he is not the first or the best in the field but he was at the right place at the right time to make Noria what it is.
Lets rephrase that a bit. They are not accurate or conducive to any legitimate assessment of mechanical wear any more than the "bitter dregs" are a representation of the wine..
You have to define "good" in terms and context with what you ( the machine owner) wants because there is more than one "correct" way to do this.
Are you interested in the condition of the oil and a change frequency?
Are you interested in the readings for use in evaluating or zeroing in on wear in the early stages?
That's 2 different condition monitoring strategies and methods to execute.
To a large degree, your statement is correct in the general but they do provide some value based information.
People trend a lot of "stuff" and swap numbers but it rarely stops a catastrophic event ( or detects one) and the wear metal analysis at best is "holistic". The chemical trend are much better.
The short answer is that a OA program by itself ( meaning not coupled with a stout PM and technical assessment program) is of very limited value and a legitimate argument can be made that the cost/effort is not worth the benefit.
Not by a long shot. In reality in liquids containing solids (and we are talking about both particle size as well as density distribution) the "hang time" is a combination of things like weight, heat, fluid properties and like with all things gravity- they will eventually settle.
Once they settle, unless you have proper agitation, they are going to stay settled- the only variable is time. ( I do lots of mixing studies and agitation and there is no "mixing" or "agitation" in any standard sump)
2 critical things about those particles ( when suspended in the stream)- I take the stream sample, I have a "real time" indication of the particle generator ( leads toward machine condition assessment) in terms of what is wearing and the frequency of the wear.
When in a sump, I am getting an agglomerated combinate collage of everything- useless for any legitimate data driven assessment of a machine.
you don't want "circulation" in a sump- you want it to settle an stay there- but when a sample is drawn from the drain (where the velocity may be ample to pick them up and carry them like a current over a silt bed)- it is not a true estimation of the running system condition.
This is why all the ML- trainings tell you to purge the sample ports, mini-mess valves, Luneta or whatever to keep from skewing the sample.
That's the problem. They cannot be without substantial modification. If this were an HPU or similar device I would sample at:
Post filter- measure performance
post crank- bearing wear
top drain- valve train wear
Sump- General chemistry
A plan similar to that would give me a reasonable idea of critical item wear in the major areas ( as good as any OA program can give)
A ICE simply isn't set up like that to allow for a single representative sample.
I don't need distinguishing from Noria, I have bested them many times. I have great respect for Jim but he is not the first or the best in the field but he was at the right place at the right time to make Noria what it is.
Lets rephrase that a bit. They are not accurate or conducive to any legitimate assessment of mechanical wear any more than the "bitter dregs" are a representation of the wine..
You have to define "good" in terms and context with what you ( the machine owner) wants because there is more than one "correct" way to do this.
Are you interested in the condition of the oil and a change frequency?
Are you interested in the readings for use in evaluating or zeroing in on wear in the early stages?
That's 2 different condition monitoring strategies and methods to execute.
To a large degree, your statement is correct in the general but they do provide some value based information.
People trend a lot of "stuff" and swap numbers but it rarely stops a catastrophic event ( or detects one) and the wear metal analysis at best is "holistic". The chemical trend are much better.
The short answer is that a OA program by itself ( meaning not coupled with a stout PM and technical assessment program) is of very limited value and a legitimate argument can be made that the cost/effort is not worth the benefit.
That's why sampling with a tube stuck down the dipstick tube shortly after engine shutdown and taking a sample only an inch or two from the top surface of the sump level (well above the bottom of the oil pan) is about as good as it gets with oil sampling an ICE. Draining from the pan isn't ideal because of possible settled debris flushing out with the oil stream.
The oil pump pick-up is well above the bottom of the oil pan in most instances, so anything that is heavy enough to settle in the bottom of the pan and stay there shouldn't be circulating in the oil volume through the system, and shouldn't be in an oil sample taken near the top of the sump level.
Last edited:
1. Thank you. (-:
2. I agree that settling can be helpful to prevent unnecessary wear. To say that a drain stream washes out that which has settled is to acknowledge that the information we seek is there. I certainly agree that identifying the source, time, and reason can be extremely difficult without more info. I also agree that this may make correlation weaker, perhaps to the point of uselessness.
3. I agree that those would be wonderful places to have sampling ports available, and that engines are not set up for that and it would not be cost-effective to set them up in passenger vehicle engines.
4. I agree and began to learn this some years ago now, via fleet and production facility experience. Fleet experience demonstrates to me that catastrophic engine failures generally happen without prior UOA warning, and also that even if one has prior warning the UOA data do not provide useful info with respect to preventing it (again, speaking of automotive engines).
Used across a large fleet, I have found that engine oil UOA can be useful to maximize return on lube oil cost by allowing one to run the fluid until it is chemically 'used up,' but that's about it. On other types of machinery, I find UOA _can_ be very useful.
The people who are actual mechanics here may have experience with what oil pans look like inside on a number of engines over the years. Are they dirty or clean looking on the bottom most of the time? The couple I have taken off I don’t remember.
Yet you think my samples taken in just such a manner are atypical?
Highly variable. Generally an engine that has been subjected to reasonable care with modern oils is pretty clean, some of them can get pretty gross though.
Similar threads
