Why do 0W-X's show moderate to high wear values?

[Pablo]

As I asked. Proof? None yet.

NOT trying to be a buttalope, BUT the statement is a strong one requiring valid backing.

Good discussion, worth raising that is for sure - however it is not up to others invalidate it, it's still up to the " 0W-X's oils show moderate to high wear" team to truly prove it. (I am willing to say it's true with proof, btw)

 

[Bryanccfshr]

This is really not up to the standards I have come to appreciate here.
UOA's are definately good tools but there is alot of context that needs to be weighed when looking at raw numbers. That's why comparing UOA's in a straight drop comparo are futile.

 

[glennc]

Pablo:

"Good discussion, worth raising that is for sure - however it is not up to others invalidate it, it's still up to the " 0W-X's oils show moderate to high wear" team to truly prove it. (I am willing to say it's true with proof, btw)"

Pablo, I should clarify that I'm not claiming that 0W-xx oils don't perform. Somebody chimed in that GC was one that DID perform (as counterevidence to that statement) and I put up my numbers since I had them and they seem to me to show no particular superiority to GC. I mean, here you have 20 consecutive "normal" UOAs (these are from a couple of months ago to maybe two years ago) and at the least you can't make a case from these UOAs that GC is an exceptional oil.

I have no problem with GC, just drawing conclusions from what data I do have. It seems to work very well in some engines and I've even considered using it.

But back to the topic at hand, I don't make any claim about other 0W-xx oils and don't have any reason to think there's a trend of poor performance among them - although I'd say there seem to be some interesting data points here that at least bring it up as a valid question.

- Glenn

 

[glennc]

Vad:

"This is all speculation on your part.
Please provide some evidence."

Well, respectfully, that's what I've been trying to do, and at this point I'd ask for some backup for your own opinion if you think I'm out in left field.

- Glenn

 

[ekpolk]

Glenn:

Respectfully, there's a serious flaw in your premise that things will reliably start to average out after twenty samples. That might be the case in a simple stats class example of trying to tell the percentage of one color of gumball in a gumball machine, but this question involves literally thousands of variables, virtually none of which are controlled to any significant degree. We're all different driver's with different driving styles. We drive almost countless distinct examples of hundreds of makes and models of cars. We use different oils in engines at different points in their lives, and so forth.

Personally, I think a far more effective way of getting a hint of the differences between oils is to compare them in back-to-back or same-season runs in the same vehicle. Of course, this approach has plenty of its own flaws, but at least it isolates away a large number of the variables that your approach can't handle (at least not at the 20 sample level).

Personally, I tend to believe that there are certain sets of conditions that will result in certain patterns in UOAs. For example, as I noted in a thread a year ago, my own UOAs between M1 5w-30 and the then new-to-me GC looked strikingly similar, yet different from other members' UOAs of the same oils. This leads me to believe that as long as the oils are of the same general class, other factors, like driving habits and environmental condition, may be more important.

But back to the issue at hand. I don't see any problem with stating hypotheses or trying to compare classes of oils. But that said, there's a strong tendency, and this is a natural human thing, to overgeneralize from isolated experiences, and I think this is coming into play here. Consider how many examples you see just here on BITOG of members declaring that "GM cars" or "Civics" or you-pick-it are awful cars. When one digs a bit, we find that this is a conclusion founded upon one single bad ownership experience. Would it be fair go judge all of us men based upon the character and behavior of Ted Bundy?

 

[ekpolk]

I know I'm going to sound like a GC partisan here (who, me???), but I have one more nit to pick with the GC comparisons. If you've dug back into this oil's "ancient history" here, you'll see that the earlier batches contained high iron levels in their VOAs (there was one with Fe around 10, right out of the bottle, for example). This virgin iron thing seemed to drop off in the later batches. I don't think it's there at all with the gold GC. Did you, per chance, try to factor out "pre-existing" iron in your comparisons?

 

[BlueWorld]

I suspect the wildly varying GC reports are a lot like the Redline and Motul UOAs.

Put a good ester base oil into a dirty engine for the first time and you get some cleaning action that will skew the UOA on the first or second OCI.

Until you you get a UOA on an oil that's been in an engine for an for a few OCIs, I think there are far too many interaction variables to draw any meaningful conclusions.

 

[Audi Junkie]

"Did you, per chance, try to factor out "pre-existing" iron in your comparisons?"

 

[glennc]

GC doesn't seem to show particularly good wear numbers on average. Summary info from 20 sequential UOAs on this site:

GC Al Fe Pb Cu
10 1 7 2 2
5 1 11 1 10
5 2 6 5 18
7 3 15 1 5
10 1 11 2 10
8 3 9 4 3
5 2 9 21 4
6 2 4 0 6
6 4 29 1 4
6 2 9 5 11
5 3 20 10 32
3 2 10 5 44
7 2 20 17 3
10 3 22 29 10
5 3 19 3 2
4 1 7 1 3
5 1 14 7 40
4 4 11 4 42
4 3 10 6 60
7 2 12 2 2
122 0.37 2.09 1.03 2.55

(First column is k miles on oil.

Some of these are great, but on average they are... average.

- Glenn

 

[TC8]

ekpolk, I have to make some comments here which may touch some nerves. Nothing personal. I own both M1 and GC. But seems to me that the GC users are more fanatic. Is that the elf influence?

So does any GC user do a VOA before each UOA? Blackstone must love that. Otherwise, why even bother with any GC UOA since there could be a "significant" skew of your results anyway? BTW, the Cu also seems higher in Glen's data. How much should Glen substract that?

There is a law of large numbers. How confident you are about your average reflecting the true behavior is only a function of sample size and the distribution spread, not how many variables involved. Now no one has sanity checked Glen's raw data. But throwing all those possibilities out without counter data showing GC to be better is just pure subjective opinions. Typical scientific approach will assume the data to be unbiased until shown otherwise. For example, you have to show convincing data that GC users abuse the oil more? If they run Chevron, they will generate more wear compared to the control case? Are their engines typically in a worse shape? Note, arguing "qualitatively" does not work well even though you may have strong convictions on this.

Final disclaimer, I'm not taking position regarding whether 0w-x shows higher wear or not. I'm not taking position whether GC is better or not, either. I'm only commenting about how to reason/differentiate information, false information and non information.

 

[vad]

You're not taking a position?

I think you just did by accepting without challenging the speculative raw data that has been followed by unsubstantiated claims.
And then by placing the burden of proof on the "fanatical" GC users.

 

[ekpolk]

TC:

No nerves touched -- at least if one was, the intoxicating effect of being under the influence of GC has deadened the pain. . .

It's been ~20-plus years since I ran through a sequence of stats and adv stats in college, but I do recall a long series of discussions in adv stats regarding sampling strategies and type/number of variables involved. If we had a large "master database" then we could do what you suggest productively. Do you suggest that any of the factors I mention are NOT factors that bear upon the vitality of motor oil? In fact, on a couple earlier other occasions, I've plowed through our data looking for one thing or another, but I've ended up realizing that each "datapoint" comes with some "qualifier" (oh, this was taken from a car with an intake leak, or sample taken in Alaska in January, or Phoenix in July, etc.). In reality, given the total lack of control of sampling and the other many variables, Glenn's data is no less subjective than the "opinions" which I've mentioned, which afterall, are based largely upon the same info (and of course, some prior learning too).

One last thing, as to the iron skew thing, no I don't do VOA every time. This was an issue noted with early GC, and not seen with similar consistency in any other oil I've seen here. Obviously, it can only skew UOA results in the unfavorable direction (no VOAs with negative iron...). Thus, any attempt to compare UOA results of GC with other products in which you say, "look at GC's higher wear #s", and those include iron, is suspect unless you've made some attempt to determine which GC you have and whether it already had x ppm of iron at the outset. Sure, yes, this heavily complicates any comparison. I don't see any indication that Glenn accounted for this.