Yes, another oil thread.
 

Its all Gelbster's fault. He posted this in another thread and I was immediately drawn in.

The apostacy is very strong in this article. We must either convert to the author's oil religion or burn him at the stake.

First, we were always taught NEVER use an oil additive, but the authors data shows that the best results were from oil doused with Prolong.

And, looking through the data, all of our sacred cows for oil choice, Mobil 1 0-40, Castrol 5-40, 5-40 Joe Gibbs, Red Line etc. Don't come even near to the best performing oils in this study.

And the other interesting result of the study is that thinner is better. All the highest ranking oils were 30 wt.

Don't flame me, I'm only the messenger.

SECTION 1 – MOTOR OIL

Its good reading, I have to admit.

In terms of "what does this mean to me?", the remaining question in my mind is how the oil pressure test results (in psi) relate to the actual pressures between components in a real engine?

For example; if the "pressure" in a bearing is only 50,000 psi then every oil tested will adequately meet the pressure required to avoid serious wear.

On the other hand, if engine friction surface "pressures" vary from 50K psi to 100K psi, then the data implies that using a different oil may preclude wear.

The author never tells us what the MINIMUM pressure is to preclude wear. Is higher better? Probably. But what pressure level is "good enough" versus "not good enough"?

I would say the engine pressures vary with load. Full throttle certainly puts more pressure on the rod and main bearings, but I don't know what the actual peak pressures would be in our cars.

real science is peer-reviewed, not published in a personal blog.

his testing was in a clean environment, not a real-life engine environment, so it was easy for him to denigrate the value of detergents.

he states that most engine damage occurs at start-up, but his testing is in a steady-state environment, with no effort to rationalise his testing results against his statements regarding oil weight and start-up performance.

he states that some oils produce noisier engines than others, but that this is not an issue. any engineer (no idea about tribologists) knows that noise is bad. noise is energy going in the wrong place.

oil manufactures state that zddp works by bonding to the wear surface, hence the issue with detergents which compete with zddp for access to this surface. he states that this bonding does not occur, but still holds to the statements about detergents having a negative effect on wear performance by 'tying-up zddp at the wear surface'.

and finally, he states again and again that this is not a test of how oil performs in an engine, but rather how oil performs at an extreme, artificial limit. he states that his test results mirror what is seen in race engines, but only in American race engines which see a lot more mechanical action at the cams than we do in our overhead cam engines. which also see much more frequent oil and filter changes so detergents are not required. which do not see the same number of cold start-ups as a street vehicle (note that race oils often do not have a cold weight) etc.

listen, anyone can write something and put it on the web - adding numbers to it doesn't make it science.

Ps, don't flame me either, i'm only the devils advocate.

Great- we are all seeking knowledge from a credible source. I just hope we find him/her before the opinionated types ruin this Thread :-) No contributors are so far in that bad category !
I am fascinated by the extent to which oil recommendations that are valid for conventional engines are less valid for the M96. That is why I went hunting through the Subaru world. I have been doing this water-cooled flat engine thing longer than Raby - my first one was a 1965 Lancia Flavia.
Lancia Flavia: What I Drove Yesterday For 570 Miles Through A Snow Storm

I checked over at bobistheoilguy.com and the armchair experts there ripped the author a new one.

He does come off as a shill for Prolong, but he did publish tests that use of Prolong with some oils actually reduced their antiwear properties.

golly, I hope i'm not the opinionated one, as I expressed no opinion in my post, just identified discrepancies in the referenced 'publication'. truly, if that were the best response one could bring to my arguments then can I consider my cause as won? my read of the original post was to discuss the merits of the article, not just the implications.

otherwise, by design there is less non-bearing metal-on-metal action in an overhead cam engine than in a conventional pushrod engine, with the exception of the chains (considered a wear item). and the chain failure we see in the m96 is associated with (a) stretching, (b) plastic deterioration, and (c) the gosh-darned ims, none of which is related to the performance characteristics tested in the 'study'. not fascinating, just factual.

I love a good discussion! but make it a discussion based on the facts (or lack thereof) presented and not do-away with troublesome arguments with a blanket label of 'opinion' or 'faith' and consider it a valid rebuttal. really, i am of the opinion that it takes a lot of faith to assume that the testing as presented has any application in a real-life engine environment. however, if all that is expected is a bunch of people to line up and say that they are going to use 10w30 from now on because they read it on the internet, then i'll stow my arguments (opinions?).

Some may find the UOA thread useful?
http://986forum.com/forums/general-discussions/55411-need-some-help-deciphering-blackstone-lab-report.html