Chemical vs Adhesive wear

Status
Not open for further replies.
Joined
Nov 16, 2002
Messages
38,279
Location
NJ
Quote:
Unfortunately, oil analysis is not very good at distinguishing wear between different formulations. Emission spectroscopy has a particle size limit of 3 to 5 microns, which means that particles larger will not be detected. Unfortunately, most serious wear issues generate wear particles in the range of 5 - 15 microns. Oil analysis only measures about 15-20% of the particles in the oil, and changing form one formulation to another is likely to change the particle size profile. Usually formulations with more antiwear additive will more aggressively react with the metal surface and when rubbing occurs will produce smaller particles. Generally, more antiwear additives will give greater iron spectrochemical numbers, even though the total iron can be lower. There are other techniques such as ferrography, which looks at the wear particles under a microscope, but now we are talking about analysis many times more expensive than spectrochemical analysis. The oils with the better spectrochemical numbers will be much less chemically active on the metal surface, so they will be less able to handle more severe loads. There is always a trade-off between chemical wear and adhesive wear. Chemical wear is the very small particles and soluble metals which is identified in the spectrochemical analysis, while adhesive wear is many orders of magnitude greater than the chemical wear, but much is not identified in spectrochemical analysis. But if you were using spectrochemical analysis as a maintenance tool and started seeing a deviation over the baseline, then you would know something was wrong.
 
Looks like Mobil 1 trying to justify there high iron counts.
wink.gif


I admit it is more complex than simple wear and I am learning more and more about this process through some difficult head nodding reading. That Trbofilms actually utilize iron oxides in their formation. That certain chemical reactions of additives that are meant to produce both a protective and a friction reducing tribofilm can also react with metals.

Aa a neophyte non profesional average Joe who tries to interpret Tribological papers I suspect that interpreting mobil 1's high FE numbers as wear may be an oversimplification.
 
Actually it was an email I received from Redline awhile back. During the days when everyone was scratching their head at all the strange RL UOA's.
 
Originally Posted By: mechtech2
The last sentence of the original post seems to have a good handle on the value of an analysis.

Based on the original quote, it would seem that meaningful comparisons would also need to be with the same oil, or between oils using the same or very similar additives and chemistry.
 
Last edited:
This definitely gives a very different perspective on M1's Iron numbers....... Perhaps GM actually knows what they are doing using, recommending and certifying the 5w30........
 
This is where a PQ count helps, as it measures the total Fe present in a sample regardless of size, not just the amount in a standard spectroscopic test.

Look at the Fe numbers vs the PQ index on these samples vs the stem cap failure.
http://www.bobistheoilguy.com/forums/ubb...true#Post587904
BTW, as a post script to that, I found three other stem caps to be failing. (badly brinelled)
Most all labs here in Australia use PQ, but it doesn't seem that common in NA as standard.
 
PQ and ferrography definitely can help provide a more complete look at what is going on.
 
Exactly!!
thumbsup2.gif
The use of proper statistical analyses of a stable and defined process is the key. Switching oil formulations, testing laboratories and filtration set ups on a routine basis destroys the notion of a stable process because your process inputs are changing constantly. Oil analysis wear rates are best trended via EWMA control charts. Here's a quick primer on what an EWMA does vs. the traditional trending methods.

http://www.itl.nist.gov/div898/handbook/pmc/section3/pmc324.htm
 
Last edited:
EWMAt = Yt + ( 1- ) EWMAt-1 for t = 1, 2, ..., n. zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz
shocked2.gif
 
Last edited:
Quote:
Switching oil formulations, testing laboratories and filtration set ups on a routine basis destroys the notion of a stable process because your process inputs are changing constantly.



I call it "switch hitting". It's where you alter at least one variable repeatedly ..and wonder why you get unexpected results. You can't draw too many conclusions on a series of one.
 
We call that OFAT (one factor at a time) experimentation in my world...it's usually slow and the results are so confounded in the end you are left with less process knowledge than what you previously had due to all the group's conjecture. Most people get angered after this process then go into the phase of, "Me like tribal knowledge better stop expensive testing ugh ugh!!".
06.gif
 
On the original subject, if the conclusion is that superior oil MMM produces more fine particle wear while cheaper conventional oil PPP produces more larger particle wear, could a home test like this provide some evidence?

- Use both brands in a comparable OCI with drain plug magnets.

(A) Compare UOA iron level. Higher level indicates more fine particle wear.

(B) Compare the harvest on drain plug magnets. Greater harvest indicates more larger particle wear.


If the assertion is correct then:
- Oil MMM use will produce higher (A) and lower (B)
- Oil PPP use will produce lower (A) and higher (B)
 
I disagree with the entire premise...

The HDEO's have high levels of very reactive ZDDP, yet they consistently generate lower iron levels than Mobil 1 - this includes Mobils' excellent Delvac 1, HD synthetic diesel formulation.
 
Originally Posted By: TeeDub
I disagree with the entire premise...

The HDEO's have high levels of very reactive ZDDP, yet they consistently generate lower iron levels than Mobil 1 - this includes Mobils' excellent Delvac 1, HD synthetic diesel formulation.


Yeah, I'm not 100% in agreement with it. It's still wear if more metal is in the oil than when you put it in. So you want lower UOA metals regardless I would think...

However, there are different types of wear, as he stated and unless you look at the particle size or tear the engine down, you can't always assume a UOA will catch the type of wear taking place, unless you maybe spend $500 a sample like JGR supposedly does.

Tear downs are still the gold standard. There is also a possibility that some oils handle wear in different situations and different temperatures better than others. That is basically what Redline was saying....imo.

For 8 years now Mobil 1 has shown higher Fe and RL higher Pb. That tells me most likely it's not an issue.
 
Hi,
buster - You again highlight an issue I raised when I first joined BITOG many years ago - single pass UOAs are "interesting" and normally only provide an accurate "snapshop" of the lubricant's condition! And especially the TAN/TBN aspects

Trending the UOAs within an engine "family" on a certain lubricant is really the only way wear metals uptake rates can be plotted with reasonable results. In this regard PQI data is important for appraising some metallurgical content's wear in the engine's structure

Some people on a Porsche Forum and elsewhere have promoted single pass UOAs as a measure of a lubricant's wear prevention performance. Largely the data accumulated is interesting - but not conclusive in any way. It is estimated (Cummins) that to be meaningful at least 800k would need to be covered in each circumstance before UOAs produced such relavant data. Even then such results may be inconclusive

In the case of Porsche 911 engines, modifications are so common that even the "metallurgical mix" may not equate when comparing what appears to be the "same" engines within the family!
 
Last edited:
I hope to have acid digested sample results here to show the board soon. It should (I don't know yet) show the entire sample reduced to the readable particle size. This assumes that the equipment isn't off track @ 100's of ppm. Even this will not tell you "good" or "bad" since it's a sample of one and pretty much totally unqualified in a group to find what "typical" is.
 
Status
Not open for further replies.
Back
Top