Are we kidding ourselves judging motor oils by UOAs?

Status
Not open for further replies.
Somebody asked a good question a few posts back...

"If Auto RX doesn't release the trapped wear metals (as evidenced by the wear metals in the rinses not being abnormally high) then how can we say that Redline oil is cleaning out trapped wear metals? Is Redline a more capable solvent than Auto RX?"

Okay so it was me...
blush.gif
grin.gif


But I do think the question deserves some deliberation before we spiral too much further into this "verbosious cavern."
tongue.gif


Dan
 
Ummm, Gary you challenge everything everybody ever says on these boards. If I said you were one smart guy then you would challenge me on it and dare me to prove it.

Honestly, education assumes intelligence and experience and its not always true. I know more than a few chemical and mechanical engineers working in refineries who dont know as much about how motor oil works in a car as you do. Some of the people I respect the most are some of the least educated. On the flip side, some of the stupidist people that I know are the most educated and whats worse they assume education covers their ignorance so they are both stupid and even more so for not realizing it. Education has its merits but it can also be a mistaken fallacy to value the ideals of its worth based on presumptions.

Gary, You may be a silver tongued deeeeeeeeeeeeevil but I guarantee you that if our toilets stopped flushing that we would respect those things as being important inventions. Anyway, I for one learn from you and appreciate your input.

Anyway enough with the spit and polish. Lets get back to discussing the film theory and consistant UOA analysis. For a UOA to have value in predicative theory (ugh I have beeeg word Gary disease) we need to establish verifiable trends. For us to establish those trends we need to be put on UOA programs where our analysis esteems the quality of performance over variables. We need to reduce the striations of random potential. When Blackstone sends back their oil samples they compare the numbers with all the others they have received on the engine. In the Uoa world the firm with the largest database to compare with is the firm with the most valuable numbers.

Now, Ive pointed out in the past that I dont think boutique oils like Royal Purple, Amsoil, and Schaeffers get a fair shake in our analysis of them because we dont have large sample sizes on them. Terry seems to love Redline and he probably does have a large sample size as a point of reference. Thats why we listen to him.

Still there are certain cars that have popularity levels where we can start to see trends emerge. There are lots of Civics, Camrys, Accords, Taurus, F-150s on the road and when youve looked over a few dozen UOA's on them you can see trends emerge.

Gary you are on the restrictive side of the fence on the filter debate and I cant help but wonder how oil filters impact the findings. Ted is a really smart guy and he says that most particles 10 microns and under show up in the UOA but you have to wonder how a reading will vary between say a pure 1 that can filter to 10 versus a Stp that might only filter down to 20.

It would be interesting to debate surfactant cleaning versus solvent cleaning and whether oils that can coat and film such as esters do either trap or release particles that perhaps skew initial findings.

I think even Terry would agree that somebody that does 10 UOA's with 20 variables as to filter, additives, and brand and viscosity usage will see the picture more darkly than somebody that does 10 UOA's with a minimum of variables.

I do think Uoa's have value but we all need a bit more education as to how to limit our variables so that we can get proper interpretations from them. You can do scientific experiments forever but if you use contaminated dirty vials then your results will be less accurate.

In short if we are going to debate the true merits of UOA's then we must do so in a framework which predicates that people are setting up programs to maximize their use and potential.

Happy Motoring All,

cool.gif


Bugshu
 
Ted is right on. You must have a extensive combination of data base, experience,and respect the tremendous variability when using normal spectro for automtotive oil analysis.

Only draw back for Ted is that he has had to use OAI for most of that time.
wink.gif


1sttruck, I appreciate your thoughts .

I think I made it clear that Auto-RX DOES NOT raise elemental values normally associated with wear when cleaning. It carries that metallic wear particle to the filter or in a protective micelle that does not allow wear generation above normal rates and most of the time reduces it. A significant if not primary characteristic of the chemistry is that when added to its deep cleaning capability , not to mention lubricity , sets the product light years ahead of current "in-situ" solvent technology. I have not studied every BITOG posted oil analysis of RX in cleaning phase but please show me that trend of wear that has you making this incorrect conclusion.

I am not smarter than anybody on this board, I have just had a unique opportunity to procure the training and experience with a unique set of skills. Back when most of you were using straight 20 or 30 weight oils and changing every 2000 miles I was interpreting UOA's and wishing that Pb was removed from the gasoline. And why can I run this turbine oil 5606, 200 hours when my poorly engineered automotive oil can't last 3000 miles or 30 hours without gumming completely up !

The main thing that sets Dyson Analysis apart is 25+ years of proprietary data that enables a bit more accuracy in automotive lube analysis interpretation.

Doug is correct that you must have a bit more than just the raw data to get a good grip on what it means , then use it to predict, prevent, and proactively tune that automotive or truck unit.

Oh please stop saying I "love" any brand. I just look at what works in a application and apply that to your problem. Redline as a company hates me like many others that watch my online comments and let me know !


Terry

[ March 08, 2005, 01:04 PM: Message edited by: Terry ]
 
Dan said Verbosius !!!!!!

lol.gif


Hehehehehehehehehehe

Did his posts on gasoline and filling up at the pump rule or what?

worshippy.gif


Happy Motoring All,
 
Auto-Rx is interesting stuff. It seems to dissolve sludge/varnish and then precipitate (?) the stuff in the oil filter, thus avoiding the clogging of oil lines and such. Some have shown images of a lot of granular sludge in the oil filter after using AutoRx, even though the owners had been using synthetic oil. By precipitating as much of the dissolved sludge and varnish in the oil filter you don't see it showing up in UOAs, but there still seem to be some higher wear numbers here and there. There also seems to be some higher wear numbers when using AutoRx in a transmission, possibly because the dissolved sludge remains suspended in the oil instead of precipitating somewhere, which is a good thing.
 
This is a very good thread.

1sttruck, Terry and Doug made excellent points. Ultimately a tear down is your best measure, but that is just too impractical. I've been very interested for a long time now on Redline's chemistry and how it reacts to certain conditions. I also find it interesting that they do NOT think oil analysis is a good way to compare oil A to oil B. According to them, these PPM #'s are insignificant. Yes, that is the word they used.


I think keeping the engine clean with periodic RX treatments and oil analysis using any oil is good enough. We've seen enough proof in my mind that many oils can do the job.

My issue with RL is that we have seen numbers drop when switching to other oils, such as Bror's Schaeffer UOA and others. These engines were also clean. And we have also seen RL run for consecutive drains and the #'s never droped. Those are the cases that have made me skeptical of it.
 
Summarizing some items..... When attempting to use wear metals in UOAs as an indicator of wear it seems reasonable to want to keep track of all of the metals. If we leave wear metals behind, in the form of sludge, varnish, or 'surface active films, we might be missing some problems and under reporting wear. In addition the films left behind might confound other items like using a different oil. Filter debris can consist of quite a bit of material, and especially when cleaning an engine with something like AutoRx or using a fine filter one should measure the material collected and add it to the other ppm data.
 
Question:

If the engine truly is "scraping" (metal to metal wear, resulting in the metal in the UOA's)...

Wouldn't you also see some more rapid than normal depletion of the zinc and phosphorous numbers?

Zinc and phos are "used up" as metal tries to hit metal. Right?

Or how about moly? Isn't it also "used up" as the metal parts "crush" though the moly layer?

So if all we're really seeing is old metal from "sludge" being stripped out of the engine--the EP add packs should not be abnormally depleted.

Am I thinking right here?

Dan
 
Just thinking out loud here but maybe some oils chemistry acts a little like AutoRX and helps deposit the wear metals in the filter. If that is the case we could have an oil grinding the heck out of the motor but if it helps keep the evidence in the filter we may never know.
 
Let me see if I'm getting the idea. You want to pick out some properties that can be read in a uoa. Lets pick iron for an example. You have a reading of 10ppm (please, just for the sake of an example). Then you tear the engine down and measure the wear. Now you do the same for 20, 30, 40, 50 and so on. Now that you have iron number and actual wear you create a curve and call this a calibration for x engine, y oil, z use. If you numbers create a reasonable curve with little standard deviation you may have a correlation coefficient that can say, for this one engine using this one oil under this one set of conditions that an iron reading of 28ppm (just an example for discussion) that you have x amount of wear. Yiiiikes, I see why everyone is useing judgement and guessing instead of facts.
 
I like your style, Larry ..but I didn't see anyone suggest that you could plot the wear of any or all engines (the same one) in that manner. At least I didn't see it.

This question has evolved over the many posts on this topic.

First it was "Does it come from wear?"

Sure. Call it decay if you will ..and attribute it to ancillary this or that ..but the bottom line is that stuff was where it should be/was ..and is where it is now. It's "gone" from the resident bodies that contained them.

This the vast majority of the planet calls "wear". It's like the lint you find in your dryer lint trap. The more you have ..the faster your clothes wear out. Can you predict how long your clothes are going to last based on the amount of lint in the trap? No ..but an increase in lint of 50% would surely lead many ..many to figure that whatever the ACTUAL wear rate is ...it has just doubled and that whatever component of the clothing that is wearing out ...will last one half as long ...however short or long that may be.


I've yet to see anything here that would alter this adapted opinion.
 
Thanks Gary. My description was only just a little toungue in cheek. We build an engine/generator package for severe cold and we are acually working on some indicators that can be read and matched with an actual physical wear measurement that can be taken any time without tearing an engine down. We acually match the pairs, real time measured target with physical measured wear and that leads to a calibration. That leads to a curve with a good correlation number. We can predict when an engine is going to dump on us. But this situation is easy because the engines run at 3.6k at full load all the time. Some day we will get smart enough to plot the three dimensional curve of a street engine in normal day to day use, but a correlation with three or more vars from multiple moving targets is going to be tough to figure out The scatter plot is going to look like an explosion in a ping-pong ball factory and the movement along a plot line will look like a jail break. But the generator plots work so well it's at least worth talking about. Lets come up with three wear indicators in a uoa and at least two types of wear that can, oh by the way be measures without a major tear down. Then lets pick a set of time, distance, load, rpm and temperature readings, throw them in the mix for a calibration and in there somewhere will be our answer(s). Sounds like a project to me.
 
Yes, I knew that was tongue in cheek
wink.gif


That would be a nice project for someone with that level of need. You could have the salesman show up with xx.xx hours of useable life left in the gen set to sell the new one ..or negotiate the refit. What more could you ask for?
dunno.gif



You, more than maybe anyone, are aware of 1sttruck's "bathtub curve". My simple thought is that I don't need to know the length nor the amplitude of that nearly flat "bottom". The higher it is, one would surely reason, the shorter it will be. This component isn't needed for naviagation.


Engine signitures can certainly be complex and certainly you can't predict the life cycle of most of them (perhaps none) with UOA. But I have to look no further than my simple jeep 4.0 and see the telltale Fe numbers ...and the informal stats that the engine will routinely last 200k or more ..and virtually never has any symptoms ..except blow by in thousands of obseved instances ..to connect the dots and that the Fe is cylinder/ring wear. This only "weak link" is clearly evidenced in UOA.

Just like you and I will probably live to our genetic life expectancy with minor variations....the condition that we're in when we croak can vary quite a bit.

So, I don't have any problem with the way I view UOA. Nor do I have a problem with 1sttruck's. I can't plot wear. I can only plot and attempt to alter what I consider wear indicators.

For people who are in such apparent agreement ...we've surely been playing tug-o-war for a good long time.
smile.gif
 
On engine life vs UOA being a wear indicator, the discussions to date have been mostly around the reliability of UOAS as wear indicators. With a typical situation that we see around here, wondering about the benefits of some different oil or change interval at some point in low, flat, middle portion of the wear curve, it seems reasonable to do; a clean, flush, do several of the oil change intervals with the new oil with UOAs as desried, also do particles and filter debris to add to the UOA wera metal counts, and at the conclusion of the test do another clean with another UOA, particle and filter debris analysis to add to the wear metal amounts. AutoRx seems like good stuff to use for cleaning, but that is another item that needs more discussion. Now you've made a good effort at collecting wear metals for comparisons.

If concerned about engine life it seems reasonable to also track data like compression, oil consumption, and maybe another wear indicator like Larry has suggested. The concern here is that apparent small changes in wear metals might result in big changes in engine life as wear can obviously be very localized. Being able to correlate UOAs/compression/ and ? to engine lfe is really the desired goal.
 
I do believe that UOA is a predictor of wear. The secret is having an expert interpret the testing data. An expert already knows what normal wear metal numbers should look like. Because he has a data bank of probably 2000 similar engines, may be even many running the same filter and oil brand. Then there is trending, UOA of your own car over many miles. I think if I saw a material change in my reslting data, I would go looking why. May be at this point I would disect the oil filter and flush accumulatuion out of the filter paper with mineral spiits and look for sizeable wear metal sizes.

There have been many cases whereby a person has died shortly after they passed their physical exam, with flying colors. No science is perfect.
I still see a doctor regularly.I still trust the experts.
 
I agree Rick20. It is my experience that UOA are extremely accurate if interpreted properly.
TooSlick also makes a great point in that you must isolate all variables before looking at brands.
 
A uoa is a good tool, but it is just one half or one third of the equation. Without a correlation it's a snap-shot. The perfect example, as stated above, is the guy that has a complete physical, walks out the door, has a heart-attack and dies. Your uoa is only as good as the lab, the guy that reads it and the trend he sees from all the previous uoa's. To think you will be lucky enough to do a uoa, just as a major event unfolds is wishful thinking. Having said that, you should either do uoa with some kind of pattern in mind or don't do them at all.

Some day we will have or z,y,z's in our plot and a computer will be able to give you a timeline. Imagine what that would be worth. If someone has some money, maybe Gary and I can do a project. I've got the first step completed...how hard can it be (lol)?
 
Hello.

A thought just went through my mind. You are discussing the value of uoa's and some don't think they are important, and some do. Let's take a moment and compare the uoa to let's say, an analysis of the human body's blood. There are tests that can be done on the blood to find out if you are deficient in iron, one of the b vitamins, or calcium. On the other hand, the test may indicate to much calcium. Each of these deficiencies or excess amounts say something about what's going on in the human body.

Example, To many sugars in the blood may indicate a possible problem with the pancreas, or conversely, excess homocysteine and a reduced amount of folic acid in the blood can indicate future problems with the heart and arteries.

We need these tests, because they tell us a story that we otherwise might never know until it were too late to take preventative measures. When we find out that our oil keeps having a significantly reduced amount of zinc, this might indicate extra bearing wear taking place. Antifreeze in the oil indicates a leak in a gasket surface. And so on... So if we listen to the results of the tests, we can then follow a path with purpose and take appropriate action, instead of reacting to a serious problem down the road at a later time when we might not expect to.

Just a few thoughts I had on this subject.


BtB

[ May 08, 2005, 06:36 AM: Message edited by: Bob The Builder ]
 
Status
Not open for further replies.
Back
Top Bottom