Extend UOA Test Particle Size Range?

Status
Not open for further replies.
Joined
Sep 26, 2010
Messages
9,791
We have likely discussed this ad nauseam (or not), but one cannot help but wonder why the size range in UOA tests is not extended or has a separate stage for at least two size ranges of particles. Case in point, when I drained the factory oil from my GLC, it was very pearlescent when swirled in the sun. This indicates to me there is a considerable amount of metal which was not trapped by the oil filter. For clarity, I have witnessed this scenario with every vehicle I have ever owned so it is not unique to my GLC.

I will assume that factory Mercedes-Benz oil filters have efficiency ratings inline with most other OEM filters, perhaps something on the order of 90% @ 30-40um. A further assumption is the particles that I saw in the oil are too small to be trapped by the filter, but equally too small (or too large) to be picked up by the UOA test—else the PPM quantities would be higher for the wear metals in the test results.

Based on the above, I have three questions for thought:

1. Do you agree or disagree there is more metal in the oil than the tests reveal?

2. Do you think the metal is too small to cause wear thus the current particle size range is sufficient?

3. Do you think changing the UOA test results to capture the quantities of the particles obviously present in the oil would be of value?
 
New vehicle UOAs always show higher wear metals....so some percentage of those smaller particles are detected.

To answer your questions---

1) Probably. Testing can only detect down to a certain size. There is a cost effectiveness there, costs more to detect smaller particles.

2) IMO any metal causes wear as it adds friction. Pardon the comparison, but a sandblasting effect.

3) Cost effectiveness, no. UOA tests are at a level where they provide an appropriate level of diagnoosis. I am sure that those smaller partciles could be detected but for the bulk of UOAs that level of discrimination isn't required.
 
Originally Posted By: 2015_PSD
1. Do you agree or disagree there is more metal in the oil than the tests reveal?

Yes. Without question.


Originally Posted By: 2015_PSD
2. Do you think the metal is too small to cause wear thus the current particle size range is sufficient?

As this question is stated, the answer is likely no.

But I think what you really mean is, do UOAs catch particles in the most consequential size range? If that's what you meant, then AFAIK it depends how you intend to use the UOA. As I understand, ICP spectroscopy (the most common method used in UOAs for wear metals) catches particles in the size range most closely associated with very early-stage mechanical problems -- but mechanical problems that are further along tend to chuck out bigger particles, which you'd need other analysis methods to catch (e.g. rotrode, ferrography, particle count, etc.). So, an ICP-based UOA is good for catching wear IF you start it when the engine is in good shape and get a trend -- but if you don't have that trend info and need to make an assessment with a one-off UOA, you need a different tool (and it'll still have a lot of uncertainty).


Originally Posted By: 2015_PSD
3. Do you think changing the UOA test results to capture the quantities of the particles obviously present in the oil would be of value?

Absolutely. The question, as always, is how much value for how much cost. And that depends on the situation, as well as the budget available.
 
1) Quite possibly.
2) Not enough information to answer, since we don't have the size.
3) As already mentioned, there is the issue of cost effectiveness. How much are we willing to pay?

Of course, there are always the usual qualifiers. Is an early change beneficial or detrimental or neutral? Does the condition we observe in the factory fill matter in the grand scheme of things? I wish I knew the answers, but these are the questions the keep BITOGers up at night.
wink.gif
 
First off, what you might find in the FF of any engine is of no significance.
There will be some residual machining debris as well as wear-in particles.
FF UOAs typically feature horror show metals levels which are really of no concern at all.
Remember that we're talking about parts per million here so a small absolute difference looks huge.
Now, can you get more accurate results for various particle sizes?
Yes, you can pay more to have particle counts done.
You'll see a few such UOAs in that forum. Not many, since this does involve a fair bit more money, but it can be done and those UOAs can be viewed.
Remember too that UOAs are mainly useful as a measure of oil condition and can be used to help determine appropriate drain intervals. They aren't really a good tool for measuring engine wear.
Also, some engines shed very little metal while others normally shed a lot.
If you compared a UOA from an SBC with one from a four cylinder timing belt Honda, you'd swear that the Chevy was in its death throes. It wouldn't be since it simply normally sheds more metal.
 
Originally Posted By: fdcg27
First off, what you might find in the FF of any engine is of no significance.
There will be some residual machining debris as well as wear-in particles.
FF UOAs typically feature horror show metals levels which are really of no concern at all.
Remember that we're talking about parts per million here so a small absolute difference looks huge.
Now, can you get more accurate results for various particle sizes?
Yes, you can pay more to have particle counts done.
You'll see a few such UOAs in that forum. Not many, since this does involve a fair bit more money, but it can be done and those UOAs can be viewed.
Remember too that UOAs are mainly useful as a measure of oil condition and can be used to help determine appropriate drain intervals. They aren't really a good tool for measuring engine wear.
Also, some engines shed very little metal while others normally shed a lot.
If you compared a UOA from an SBC with one from a four cylinder timing belt Honda, you'd swear that the Chevy was in its death throes. It wouldn't be since it simply normally sheds more metal.
For clarity, I am thinking past the FF, but it is a great example of the theoretical weakness of a UOA and yes, I am aware they cannot be reliably used to measure wear, but what I am suggesting is that with a few fundamental changes could they be? Also, the PPM reading is based upon certain sized particles--the ones which may actually need to be counted may not be in the PPM reading. I have done several particle counts and I am not sold on that as a way to measure wear either since they simply place a sharper lens on the same results already taken. I am not saying you are wrong by any stretch, but I am thinking quite a bit out of the box.
 
Originally Posted By: Garak
1) Quite possibly.
2) Not enough information to answer, since we don't have the size.
3) As already mentioned, there is the issue of cost effectiveness. How much are we willing to pay?

Of course, there are always the usual qualifiers. Is an early change beneficial or detrimental or neutral? Does the condition we observe in the factory fill matter in the grand scheme of things? I wish I knew the answers, but these are the questions the keep BITOGers up at night.
wink.gif

Agreed; but I am also thinking past the FF such that if there were a couple of ranges of particle sizes tested we MAY have a better tool to measure wear, but I also concede there are a zillion variables to it. For the cost, I think it would depend if all of the metal that is seen is smaller or larger than what is already captured in the test. Smaller is likely more cost whereas larger may not be. One would think in this day and age of computer controlled everything, that we would be able to improve the UOA past oil health.
 
It has been noted that some engine just shed more metal than others. It’s not a sign of a bad engine, just a characteristic. Some manufacturers may have different materials or alloys, different casting and production methods.

This throws another aspect into the mix.
 
I've posted this graph before, it is lifted from a Machinery Lubrication article I ran across. Note the upper limit of size detectable by emission spectrography and where it falls on their threat graph.

When I ran AA and KF on oil analysis in college for our automotive engineering and fluid power departments, we were often told that for the most part anything that showed up in a standard (non-digested) analysis was not in and of itself damaging to the equipment since the particle size that was detectable was too small to cause damage. They instead used the analysis to look for trends or for spikes to very high levels to indicate impending failure. I'm no expert, but I do chuckle here when I see people saying M1 causes "wear" when an analysis shows 18ppm of iron instead of 6. Some of the bad samples we tested would show wear metals in the several thousand ppm range. These days with ICP the detected range might be different than with AA, I don't know. I do know ICP has a much hotter plasma and a higher residence time than with the old AA so is likely to be able to vaporize larger particles.

We always performed an acid digestion on our samples, that gave a rough indication of larger and potentially damaging particles in the samples. We'd run the analysis on both the as-received sample and the digested one and compare the results. If you got a much higher reading on the digested sample that indicated a high incidence of relatively large particles.

Backup_200405_oil-fig3.jpg
 
Originally Posted By: JLTD
New vehicle UOAs always show higher wear metals....so some percentage of those smaller particles are detected.

To answer your questions---

1) Probably. Testing can only detect down to a certain size. There is a cost effectiveness there, costs more to detect smaller particles.

2) IMO any metal causes wear as it adds friction. Pardon the comparison, but a sandblasting effect.

No, emission spectrography (ICP or AA) can only detect up to a certain size. In fact, it is the single atom that is detected. It's the large particles that require further processing of the sample prior to analysis.

And very small particles do not add friction or cause wear, they just exist in the oil like any of the other metallic atoms present in formulated motor oil.
 
KS, I hadn't seen that article before, but I didn't even think about SEM-EDX testing for oil analysis. We have it done when something ends up on our final product and don't know what it is, but it makes sense. The lab usually tries to wash it off into solution with alcohol or another solvent, and then transfers the solution into the machine and reads it out. I'd imagine doing this test on oil would help skip a couple steps since it's already liquid. Used in this manner, SEM-EDX is quite limited in its scope because it gives you elemental breakdowns and you must have a reference sample for them to compare and see if it is the same compound; I will have to ask the lab guys if they can actually do particle counts on it. I may send them a sample from my next UOA just to see what all they can tell me.

And for the cost-benefit side, when we have to send out an SEM-EDX sample, it's likely $600-700 total by the time you get the analysis and their report on what is or isn't in your sample. So no, I don't see any casual UOA springing for that type of analysis, even if it tells you all kinds of neat information. That cost represents a lifetime's worth of oil changes for some vehicles!
 
Originally Posted By: 2015_PSD
when I drained the factory oil .... it was very pearlescent when swirled in the sun. This indicates to me there is a considerable amount of metal which was not trapped by the oil filter - - - I have witnessed this scenario with every vehicle I have ever owned so it is not unique...


I have also noticed this multiple, MULTIPLE times.

It is one of the main reasons I added a bypass oil filter.
 
Originally Posted By: Linctex
Originally Posted By: 2015_PSD
when I drained the factory oil .... it was very pearlescent when swirled in the sun. This indicates to me there is a considerable amount of metal which was not trapped by the oil filter - - - I have witnessed this scenario with every vehicle I have ever owned so it is not unique...


I have also noticed this multiple, MULTIPLE times.

It is one of the main reasons I added a bypass oil filter.


Same here and I will always be a 1k on FF owner …
On the subject of UOA … who would you bet has the more sophisticated lab and personnel:
1) Blackstone
2) CVX, RDS, XOM etc …
 
Originally Posted By: 4WD
Same here and I will always be a 1k on FF owner …
On the subject of UOA … who would you bet has the more sophisticated lab and personnel:
1) Blackstone
2) CVX, RDS, XOM etc …
Horizon (Polaris) seems to one of the most sophisticated and carries many certifications and modern test methodologies:

https://polarislabs.com/polaris/our-promise/
 
Originally Posted By: 2015_PSD
We have likely discussed this ad nauseam (or not), but one cannot help but wonder why the size range in UOA tests is not extended or has a separate stage for at least two size ranges of particles. Case in point, when I drained the factory oil from my GLC, it was very pearlescent when swirled in the sun. This indicates to me there is a considerable amount of metal which was not trapped by the oil filter. For clarity, I have witnessed this scenario with every vehicle I have ever owned so it is not unique to my GLC.

I will assume that factory Mercedes-Benz oil filters have efficiency ratings inline with most other OEM filters, perhaps something on the order of 90% @ 30-40um. A further assumption is the particles that I saw in the oil are too small to be trapped by the filter, but equally too small (or too large) to be picked up by the UOA test—else the PPM quantities would be higher for the wear metals in the test results.

Based on the above, I have three questions for thought:

1. Do you agree or disagree there is more metal in the oil than the tests reveal?

2. Do you think the metal is too small to cause wear thus the current particle size range is sufficient?

3. Do you think changing the UOA test results to capture the quantities of the particles obviously present in the oil would be of value?



1. Yes, I think there is almost always more than the tests reveal.

2. Nope, I think that the SAE and most third parties nail it when they say its the smaller particle that aren't caught that cause the most wear and that current full flow filters won't catch them. especially in a heavily loaded device - like towing, a motorhome, a boat - those sparticle gets squeezed in journals really hard during sustained output.I think the tiniest particle also get into timing chain pin bushing and accelerate wear which is why we see a lot of timing chain replacement in trucks using long life maintenance minders.

3. Depends on the machine itself and its use- if that machine being down holds back your work (like a dump loader or big genset you need) I feel it's worth paying a little more to for a particle count to ensure uptime and accurately determine OCI intervals.

If its a lightly loaded 8-year-old car and you have 2 backups I don't think its worth the money.


On the subject of #1 and 2 unless one can buy and fit a 2 stage with a bypass (people trust Cummins - but not microgreen) your only choices for further removal are

1. Dump the sump more often clearing out the sub-20-micron particle load.

2. Install a plumbed in bypass - expensive and effective

3. Put magnets in strategic locations - a drain plug and or one on the filter or some in the oil flow itself.
The magnets do the most for the least to help with this problem and I ALWAYS find debris on them.

Next time you have a pearlescent sample drag a neodymium through it and see what comes off on the mag.


UD
 
Last edited:
What I find interesting is that we [blanket statement]replace our vehicles before the engines wear out. Our maintained transmissions rarely wear out, they break. There of course will be engine failures from bad parts, assembly problems over heating etc. I will not include lack of proper maintenance because I am sure we service our vehicles.
 
Originally Posted By: UncleDave
Dump the sump more often clearing out the sub-20-micron particle load.
There would be so many here who would dogpile on a statement like this, but (IMHO) this is the "cheapest" way to remove the metal. Before, I am "dogpiled", I understand about extending OCIs (I have done it myself), the statement that frequent oil changes are not seen as "cheap insurance", or that the metal does not matter since there are plenty of OEMs who recommend 10K OCIs on factory fills.
Originally Posted By: UncleDave
Next time you have a pearlescent sample drag a neodymium through it and see what comes off on the mag.
I had not thought about doing that, I guess I will have to try and remember the next time that I have a new engine and am dumping the factory fill.
 
Back when I was in college we didn't have an SEM, much less one that could perform EDX. When I got my first real job out of college we had one but I had long since stopped performing oil analysis by then.

It's a tedious method for performing a bulk analysis and is far more suited for examination of specific particles, both in size and composition. As you mention you'd wash out the particles in a sample and examine them, and be able to get an idea of the composition of specific targeted particles. I'm assuming that's why they title it "Singe Particle Analysis" in that chart, I don't think you'd use it to get an overall idea of the amount and type of metals in a sample.

Originally Posted By: SubieRubyRoo
KS, I hadn't seen that article before, but I didn't even think about SEM-EDX testing for oil analysis. We have it done when something ends up on our final product and don't know what it is, but it makes sense. The lab usually tries to wash it off into solution with alcohol or another solvent, and then transfers the solution into the machine and reads it out. I'd imagine doing this test on oil would help skip a couple steps since it's already liquid. Used in this manner, SEM-EDX is quite limited in its scope because it gives you elemental breakdowns and you must have a reference sample for them to compare and see if it is the same compound; I will have to ask the lab guys if they can actually do particle counts on it. I may send them a sample from my next UOA just to see what all they can tell me.

And for the cost-benefit side, when we have to send out an SEM-EDX sample, it's likely $600-700 total by the time you get the analysis and their report on what is or isn't in your sample. So no, I don't see any casual UOA springing for that type of analysis, even if it tells you all kinds of neat information. That cost represents a lifetime's worth of oil changes for some vehicles!
 
Status
Not open for further replies.
Back
Top