There are four general ways to measure wear which are common:
- electron bombardment on the parts
- component weight
- physical attribute gauges (micrometer; calipers)
- UOAs
The first three methods each have pluses and minuses. The biggest problem is that none of us have the time/money to tear apart an engine/diff/tranny to do these kinds of measurements. And to what end? Are you going to do a teardown and rebuild every 10k miles to see if your oil is really doing it's job??????? And, every time you would "reassemble" your engine, you introduce huge error concerns because you would NEVER get all parts torqued to the exact same value as the previous time ... Any analysis of wear involving teardown is a one-time-only venture. And the R&R is horrible if the teardown is attempted in repeat cycles.
That leave UOAs. UOAs are not perfect; I don't think anyone who is sane would say otherwise. But they are a very quick and inexpensive look at the wear metals, and you take inference from those values to make an overall impression of wear. UOAs most certainly have shown good-to-excellent correlation in measuring wear shifts relative to the other aforementioned methods.
What I don't understand is how anyone can cling tightly to a UOA's info regarding the inputs of wear (vis, FP, elemental additives), but ignore the outputs?
Let's think of something else for a moment ... Think of making a cake.
- The recipe calls for batter mix, eggs, milk, salt, sugar, and things unique to the desired taste such as cocoa, vanilla, etc. These are all "inputs" to the cake. Each ingredient has a specific measure, and at times a specific process such as mixing, beating, cooling, etc.
- Then these ingredients are put into a process; they are baked at some temp, for some duration of time
- then finally, you eat the cake.
How would you ever know if too much salt or too little sugar were used, if you never tasted it????? Not one single ingredient has meaning until someone first "measures" the success or failure by eating it. The "output" (the taste of the finish product) is what matters most. You can learn to tweak the inputs ... more of this, less of that, but in the end, the taste is what matters. This is why I don't understand why BITOGers get so wrapped up in VOAs, or ignore the wear data from a UOA. There are dozens upon dozens of formulations which give good wear results. Think of all the brands and grades of lubes on the market, and most of them do a very admirable job of controlling wear. When people say "This is low on boron" or "This has a really nice load of Ca and Mg", to what end does that prove anything? Want to know how well lube does it's job? Measure the wear. Anything else is just guessing at inputs. Anyone can say "ZDDP" reduces wear. OK - how did we learn that? We to try it and measure the result. Measuring inputs allows one to recreate success, whereas measuring outputs allows one to judge success.
Why would you care about the inputs, if the outputs were desirable. If I gave you a gallon of lube and had it run in your car engine, and the wear dat was excellent, would it matter if it were Mobil 1 AFE, or a 50/50 mix of goat milk and dog urine? If the WEAR is GOOD, then the inputs ONLY MATTER in terms of recreating the lube again. The OUTPUT is what tells you the INPUTS are acceptable; it's NOT the other way around.
Wanna know how well a lube does it's job? You have to measure wear. And the ONLY quick, inexpensive way to get a reasonble (not perfect) sense of what's going on is to take a UOA. Period. VOAs only hint at the potential for success. UOAs tell you what actually happened in use.
Which tells you how well a ball team played the game? Knowing the starting line-up, or looking at the score as the clock goes to zero? Knowing who's starting the game only gives a hint at what MIGHT happen. Seeing the box score tells you how well the team actually played.
Saying a UOA is not useful to track wear is, IMO, very short-sighted and uninformed. It's a proven methodology that is very easy to employ. It is also very easy to misunderstand, and misuse. And most assuredly, it's the only way a gear-head like Joe Bitog can do so from his garage.
Here's what you need to know about a UOA:
- Singular UOAs may be able to discover a developing issue, but they are not assured to do so.
- Singular UOAs are, in no way, shape or form, able to tell you if one lube is "better" than another; this is a fools errand.
- Singular UOAs are able to show you if contamination is present; this should be cause for further investigation if so.
- Multiple UOAs MUST be taken in large quantities (30+ samples) to have any hope of finding averages and standard deviations sufficient to reduce the sigma variation to a plausibly believable range
- Multiple UOAs can be done in either macro or micro conditions, but you cannot interchange those willy-nilly and believe you have a useful output
- Multiple UOAs do take variation into account; that's the entire purpose of the statistical sample analysis in the first place
- Any UOA is going to be your only viable, fiscally attainable means of tracking wear in your garage.
