Originally Posted by bbhero
Isn't a UOA truly to figure out if the oil is sheared too thin or too thick from oxidation, or to demonstrate a coolant leak by K+ and Na+ being present or too much silicon suggesting a leak in the air filtration system?? That's what is truly testing for.
I have read on here numerous times that using analysis for wear is not the correct usage of this type of testing... In fact I have read where a spectrometer test cannot or will not show larger pieces of engine debris that in fact would clearly demonstrate a real problem?? That " wear" results from this type of testing has in fact looked great... Then the same exact motor literally blew up shortly after a great test analysis result. So all in all the idea of using a test of this type is not for attempting to show anything relating to engine wear. It is rather testing the oil to make it is in grade, the possibility of contaminates being present, and the possibility of outside particles being present...
If looked at from this point of view then our op did nothing all too wrong here... The oil did in fact shear out of grade in 3k miles ( which is potentially significant because of such a short run duration) ... No signs of coolant present... And no signs of a faulty air filtration system... this test was not a waste at the end if the day really... Oil out of grade is of note here.
Has far a long term trending of data... That makes sense for someone like PSD who did several long consecutive runs of Pennzoil Ultra... Then compared that to Mobil Super after several long runs consecutively. In his results he could prove Mobil Super stayed in grade just has well as the Pennzoil Ultra did. I compare this to weather computer model runs... A long term trend with model or models is far more significant than a single model run that changes drastically from the previous run. You are looking for run to run continuity or tending continuity within which to make a declarative statement. In the case of PSD2015 using the oil analysis proved he could run as long with Mobil Super has he did with Pennzoil Ultra. With each of those oils he pushed our a bit farther with each run until a certain point based upon his analysis test results. Once at a mileage 14k miles or so and then he continued to run the same exact oil in his truck for a good number of runs after that point establishing a baseline with that Pennzoil Ultra. Then he repeated the same exact process with the Mobil Super. The analysis showed the oil was still within grade and the TBN/TAN was still acceptable for both oils at those mileages. It showed how well in his application that Mobil Super was just has capable has Pennzoil Ultra was for longer runs. None of which should have technically involved wear metals being a part of that process. Because metal wear is not what UOAs are truly to be used for. It is for the condition of the oil. At least that is what I have read here from a number long time posters on here.
Your questions really are not directly related to this UOA, but UOAs in general. However I will address them here.
UOAs are tools. All tools have proper and improper uses. Despite a lot of what you've read, UOAs most certainly can be used to track wear. There are entire industries that rely on this kind of data to make smart, managed decisions in their maintenance programs. UOAs are a direct view of the fluids, and an indirect view of the equipment. But I contest that they are most certainly useful
if one knows how to use them correctly.
UOAs using spectral analysis are limited in what they can see, but that does not mean they are useless. In fact, they are very good at tracking "normal" wear. While they will not see all wear, they well see a portion of wear. This visible wear is the more common of particles; stuff at/below 5um in size. Very large particles that cannot be seen by a UOA are also quite low in occurrence rates relative to the propensity of smaller stuff; PC data confirms this. And there is some SAE study data that does at least show correlation between the UOA wear results and PC loading, indicating that as more particulate exists in it's spectrum of sensibility, the UOA wear metals also showed similar effects. People whom tell you that UOAs cannot track wear are, frankly, wrong. It is true that if a catastrophic event is eminent, the UOA may or may not see some large influx of wear metals, but that is really not a good litmus test in the first place, because catastrophic events are by nature acute and not chronic, and therefore unless you took a UOA at the EXACT right time, you'd not be able to see the issue in the first place. Also, there are most certainly examples where UOA absolutely did predict the onset of chronic major issues; Blackstone often has newsletters with just such examples. So please ignore anyone whom tells you that UOAs cannot track/predict wear; they are flat wrong. Admittedly, UOAs are not perfect, but tell me what is in life? UOAs are tools and as long as you use the tool in the proper mode, with the understanding of the benefits and limitations, we are better for using them than not using them at all.
UOAs tell us composition, but not size. PCs give us an idea of size and loading percentage, but not composition. Together, however, we can use them to get a fuller understanding of the sump condition.
Further, in terms of UOA particulate that is "seen", are we to assume that when a major issue is impending, it ONLY sheds large particles? That if your timing chain guide on a 4.6L Ford engine is getting chewed up because the chain tensioner is bad, that it ONLY throws off large Fe and Al chunks and nothing below 5um? That's absurd. When high wear issues develop, they often throw an entire spectrum of elevated particle sizes; large and small alike. What a UOA cannot see is a con-rod that is about to snap, or a bolt/roll-pin that is about to shear. Those kinds of acute failure cannot be predicted by spectral analysis. But items like bearings going out, chain guides wearing out, etc will typically show an escalation of metals all across the PC range, and therefore some amount of warning might be available. NO - not always, but let's not pretend like bad things ONLY throw big chunks and no small stuff. Even if an issue were to toss out particles larger than 5um, it won't be long before those 5um particles would be ground down into smaller particles by the interaction with other tight clearances while circulating in the fluid stream, etc. And subsequently those large particles have the potential to be made into smaller ones; often small enough to then be picked up in a UOA.
UOAs track both outputs and inputs. Inputs such as vis, FP, TBN/TAN, etc are all inputs to the wear equation. And many times, we see these variables change, and yet nothing substantial happens to wear. How many times have we seen EB engine UOAs that have a lot of fuel dilution, but no change in wear rate? How many times have we heard the resounding gong of people warning about TBN being low, and yet nothing abnormal happens in the wear, and the engines chug along perfectly fine? I'm so over the TBN/TAN issue that I don't even pay for those any longer, and I pay them little if any heed. This is because those inputs have NOT shown good correlation to the wear; the inputs are not directly affecting outputs. This is a fundamental misunderstanding folks have. Without correlation there can be no causation! It's that simple. What these inputs are good for are PREDICTORS of possible shifts in wear as the conditions continue. If you TBN goes low, it does NOT assure that wear will suddenly spike; rather, it is a clue to pay closer attention in the next UOA, and possibly UOA a bit sooner, as the wear MAY be affected (as opposed to WILL be affected). Inputs are predictors of the potential for change; outputs are the results that show what actually happened!
Additionally, what alternatives are there to understand wear if we don't use UOAs? Well there's tearing an engine down, right? Let's look at the pro's and con's of tearing into a piece of equipment to track wear; what do "tear downs" offer us:
- timely? Not by a long shot. It takes hours to pull an engine or trans, rip it apart, measure clearances, etc.
- cost? Nope - tear downs certainly are not cheap, are they?
- accuracy? measurement techniques and tools often show that this approach does not have a good score in gauge R&R (repeatability and reproducibility), and so it's not as wonderful as one first thinks
- repeat applications? tearing into a piece of equipment also induces errors in reassembly as well; will you reuse bearings? new bearings? what about fasteners? All these induce man-made variation
So, it's not exactly a slam dunk in terms of calling a tear-down any more useful or accurate compared to a UOA. In fact, it's really a last-ditch effort to understand why something went wrong in a piece of equipment; no one tears into a perfectly good engine or trans because they think things are totally fine, do they? The only time a tear down happens is when something is expected to have gone horribly wrong. Some catastrophic event has already happened anyway; so how's this any more affable than the insult that often gets hurled at a UOA in that some people think it's not useful for understanding major problems???
In the view of long term trending of data, rather than write it all out here, I'll just say read this:
https://www.bobistheoilguy.com/used-oil-analysis-how-to-decide-what-is-normal/
This discusses the whys, wherefores, etc of how to use the data in a pragmatic manner, and clearly tells about what we can and cannot glean from analysis. Macro and micro data streams are so grossly misunderstood on this site it's almost painful to read at times. Many people here are either ignorant or arrogant in regard to the topic of "normal" variation; they either are not aware of it, or they are but choose to pay it no heed.
UOAs are not perfect, but they do indeed exhibit proven correlation between their information and other means of tracking issues such as particle count analysis. UOAs are reasonably quick. UOAs are fairly inexpensive. UOAs have certainly shown that they can many times (but not always) indicate impending doom. You can poo-poo UOAs if you want, but for me, I'll use them in the manner in which they are best utilized, understanding their benefits and limitations, and I'll happily trod along knowing I've got a system that gives me info as to inputs and outputs, inexpensively with reasonable accuracy.
I now return you to your regularly scheduled programming.