20 Wt vs 30 Wt Iron Wear Levels from BITOG UOAs

[rjacket]

Firstly, thank you very much for the effort in this study.

A quick question on the figure being chosen to measure wear.

You took the absolute iron wear and compared it across all the samples.

But isn't it the case that all the UOA's have an iron wear number and an average iron wear number for that engine.

Wouldn't we be more interested in knowing the performance difference against the average iron levels?

It strikes me that comparing the absolute number introduces the bias of the actual engine that was being used. M1 might have been used in engines with high iron wear tendencies, PP might have been used in engines with low iron wear tendencies.

The engine, engine age / condition, OCI, driving style, weather are all the factors unique to the results of a UOA. Shouldn't we be trying to minimize their influences as much as possible? It seems that engine, engine age, OCI are all factors that are excludeable from the data you collected in the spreadsheet.

 

[steve20]

Great info OC-thanks very much for sharing
Steve

 

[MolaKule]

Quote:
this is deeply flawed.


I disagree.

While you may diasagree with the sampling methodology or the data used, OC did a great job with the data available, using accepted Statistical Methods.

 

[Topo]

Great Work. Great Conversation. I appreciate the level of Knowledge on this site. I couldn't even begin to ask questions or debate on your findings, but never less I was very intrigued by your data.

 

[azsynthetic]

Originally Posted By: MolaKule
Quote:
this is deeply flawed.


I disagree.

While you may diasagree with the sampling methodology or the data used, OC did a great job with the data available, using accepted Statistical Methods.


I disagree

The data used is flawed therefore the analysis is flawed. Now, to make this a little more interesting. My VW 1.9L TDI originally spec-ed for 5w40 and were ran with various 40WT, then I went to 10W30, then 5W30, then 5W20 and currently running 0W5 Neo synthetic. The wear numbers are basically the same between the 40Wt and 30WT. The 20WT oil is being analyzed and will be posted shortly. I will do the same with the 5WT when time comes.

 

[tig1]

Originally Posted By: azsynthetic
Originally Posted By: MolaKule
Quote:
this is deeply flawed.


I disagree.

While you may diasagree with the sampling methodology or the data used, OC did a great job with the data available, using accepted Statistical Methods.


I disagree

The data used is flawed therefore the analysis is flawed. Now, to make this a little more interesting. My VW 1.9L TDI originally spec-ed for 5w40 and were ran with various 40WT, then I went to 10W30, then 5W30, then 5W20 and currently running 0W5 Neo synthetic. The wear numbers are basically the same between the 40Wt and 30WT. The 20WT oil is being analyzed and will be posted shortly. I will do the same with the 5WT when time comes.


Keep us informed. You're useing 5-20 and now 0-5 in a VW TDI? Interesting.

 

[GMorg]

There are no perfect experiments. There are no perfect data sets. However there are appropriate analyses of the available data. Thankyou, very much OldCowboy.

I hope that we could all find a weakness in the data or the analysis. However, those weaknesses would not distract from this effort. I think that it is an interesting observation that is worth some pondering. It clearly won't end any arguments here, but it will propagate some thought for those capable continuous re-evaluation.

Iron in this analysis IS a surrogate for wear, a marker. It is not a perfect marker, but since tear-down measurements aren't available in this data, it approaches silly to suggest that it be use. You evaluate what you can.

@friendly_jacek
Given the part of your background that you have chosen to share, I would hope that you have some appreciation of biological models (not mathematical models) and draw some parallels to this analysis.

 

[Quest]

Thank you Old_Cowboy for your work on this subject matter. I'm sure your work and dedication to BITOG knowledge is highly appreciated by many of us readers.

(** I'm in agreement with other posters remarks i.e. GMorg and Molakule...things that make good sense to me **)

Q.

 

[3311]

I wonder how filters going into to bypass affect iron levels? In theory, all other things equal, wouldn't a 20wt have less/lessor bypass events than a 30wt, thus reduce unfiltered oil and potentially affecting iron level comparisons? Just a thought I've had bouncing around in my noggin for while with respect to the 20wt vs 30wt debate.

Edit: Have any of the oem's that have moved to specifying 20wt ever published their test data/tear down analyses etc. on the subject of 20wt vs 30wt? It would certainly be interesting.

Thanks for the efforts OldCowboy.

 

[Exhaustgases]

The true test is to start with 2 factory new engines.
Run one on 20 and the other on 30, from break in to many hours of loaded operation over a varied speed range.
Then do it over again on 2 more.
Its not a fare test when one of the engines has coated aluminum cylinders and the other does not.

 

[fpracha]

Originally Posted By: MolaKule
Quote:
this is deeply flawed.

I disagree.

While you may diasagree with the sampling methodology or the data used, OC did a great job with the data available, using accepted Statistical Methods.


+1. Thank you Old_Cowboy

Your effort is indeed useful BITOG knowledge.
It is better to consult this data than to make a recommendation based on opinions only!

 

[Exhaustgases]

Something to think about. As far as a lubricant suspending iron in itself, I can fully understand why a lower viscosity oil would contain less. It simply drops out of it faster than a thicker or heavier viscosity oil. As far as better at protecting the bearing surfaces, that would depend on the additives, clearances, loads, temps, speeds and things like that. The higher viscosity will always better support a load, and is the reason for higher viscosity oils used in highly loaded, high thermal environment machines.

 

[fpracha]

Originally Posted By: Exhaustgases
Something to think about. As far as a lubricant suspending iron in itself, I can fully understand why a lower viscosity oil would contain less. It simply drops out of it faster than a thicker or heavier viscosity oil. As far as better at protecting the bearing surfaces, that would depend on the additives, clearances, loads, temps, speeds and things like that. The higher viscosity will always better support a load, and is the reason for higher viscosity oils used in highly loaded, high thermal environment machines.
May be , or may be not... if the iron particles are of small size both a thin oi well. It is not so simple and more complex engineering science is involved here rather than conjectures based on personal feelings only !

 

[Gabe]

Thanks for the hard work