UOA's versus Jim Fitch, et al!

[1911]

Guys,

I evidently assumed (mistakenly) that everybody would be able to understand that the Jim Fitch idea immediately implies exponentially increasing damage rates. To help clarify this, I'll state some basic principles the conditions under which this type of math is implicit. If this is not clear, please contact me.

Exponential functions apply mathematically when the rate of change of a variable (Y) is proportional to the current value of Y. If Y is the quantity whose rate of change is proportional the quantity of Y present, then Y will vary with time according to Y = C times e (the natural log) raised to the kt power or Y=CE(kt) where kt is an exponent. In this equation, T=time, C is the intitial value, and k is called the constant of proportionality.

Now does the Fitch example meet the criteria above? Yes, it defines that very criteria: If the rate of wear particle development is dependent on the concentration of wear particles (Y), then this perfectly meets the mathematical model/requirements for exponential growth.

Now remember that Stinky made an important point about exponential growth, it does not necessarily imply high values of anything (it just states that they get higher at an exponential rate). If C is low and k is low in the equation above, then you may never notice the growth especially if the sensitivity of your yardstick is not very good (This is called a UOA on BITOG)

Where does this equation apply in your everyday life....listen up now Jay!!!

1) BANK INTEREST --the amount of gain in capital per time period is proportional to the current amount there and the capital amount thus grows exponentially.

2)POPULATION GROWTH -- the rate of population increase is proportional to the current population and the population thus grows exponentially.

3) UNCONTROLLED NUCLEAR FISSION: the rate of population increase for fissuring neutrons is proportional to the amount already there and thus the population of free neutrons grows exponentially (at least until it gets a little too toasty if you smell what I'm cooking.)

4)PARTICLE INDUCED WEAR: The rate of wear particle increase is proportional to the current concentration of wear particles and thus the concentration of said particles must grow exponentially.

Now speaking with analogies here, controlled nuclear fission (as in a nuke power plant) would be analagous to having an excellent filter on your car thus stopping the consequences of unrestrained exponential growth. Perhaps current filters do remove many of the particles that would cause wear and perhaps this is one of the reasons that exponential growth rates are not generally observed. Now Jay, that is a theory and is thus sublect to debate and argument..........the mathematical invokation of the exponential growth rule is not
a theory but is implicit if you accept the Fitch premise. It can be no other way.

Now Jay, when I state theories, you can argue with me all day long and probably be right but you cannot ever be right arguing against the application of 10th grade mathematics to an obvious and clear-cut case where the math fits.

ALso, the last time I mentioned that the damage caused by particles may be dependent on lube film thickmesses, you accused me of making up that theory too. As you can see, it's a heavy theme in the Fitch article. I don't know if the general idea has ever been disputed. I don't know if 10th grade math has ever been disputed either. But if there's one here foolish enough to refer to facts as theories and to treat his own opinions as facts, it's you. You take the crown.

 

[unDummy]

Is there an increase in wear after an oil change? Is the new oil washing the engine and whatever was left behind by the previous oil?
Or, are some people installing completely dry oil filters causing all that initial wear waiting for the new filter to fill up?
Is the filter filtering better after 5k miles or when new?
Is a UOA worth looking at anymore?
Or, should TBN/particle counts be mandatory and not just an option?
Should the oil filter media be analysed?
Without a UFA(used filter analysis), is anything in the UOA accurate?
Are there any UFA providers?
Is an extended OCI a cure for those known sludge monsters?
How many engine actually 'wear out'? or die from other causes where OCI chosen wouldn't make a difference?

I personally don't see many engine last with 10k OCI's.
Sure, the engines that don't stress the oil, have large sumps, oil temp control, reasonable psi and flow, and were engineered without a econocar budget, will last.
But, where does that leave the majority? The majority:: small oil sumps, high oil temps, frequent short trips, traffic jams, high speed highway runs, bad weather conditions, low on oil(who checks it regularly)........
And we wonder why there are so many sludge monsters.

I also believe that the switch from carbs to FI to OBD2 FI, from no emissions control to TLEV/LEV/ULEV/SULEV, and the better oils available, are the main reason why have higher OCIs, and not because the engine supposedly wears less with older oil!

I think that Stinky hit the nail head on concerning a UOA. Reread his post above: abnormal wear or issues..... contamination...shortcomings.....improper operation....
Seems pretty simple.

 

[Jay]

1911, it would be much easier to debate you if you'd state, specifically, what you think oil analysis can and can't do and why.

 

[Stinky Peterson]

1911 - I think your ideas would be more accepted if you didn't insult the intelligence of this board's members.

Filters, additives, low wear rates, and oil changes prevent the exponential growth from happening. No fancy math required to understand this. Take an unfiltered gear compartment and don't change oil and you will see something different.

One last thing - you don't need to talk down to people to get your point across, show a little respect and you'll get along a lot better.

 

[1911]

Originally posted by Stinky Peterson:
[QB] 1911 - I think your ideas would be more accepted if you didn't insult the intelligence of this board's members.

Filters, additives, low wear rates, and oil changes prevent the exponential growth from happening. No fancy math required to understand this. Take an unfiltered gear compartment and don't change oil and you will see something different.

Sorry about the double post, having some computer issues here. I appreciate your comments here but I have some important things to say. I was giving a free math lesson, not insulting the board Stinky. If someone understands that math, they can ignore the post, if they don't, they need to read it as I gave a nice free math lesson and should be complimented
for it.

Let me explain why I get a little but upset.
It's scary that can state that 3+2 equals 5 on this board and get argued with.
I simply made a very simple mathematical observation that should have been obvious --(especially to all the alledged scientists here). I did not really state an opinion but left an open question to the experts or alledged experts. I seeked an explanation.

Forgive me for saying this but it bothers me that you refer to 10th grade math as being "fancy". It should be mundane and intuitive, especially for a scientist. I also get upset with Americans in general in that nobody seems to know 10th grade math including many of the "I am a scientist types here". If one does not know 10th grade math, they are note not a scientist. If someone cannot see and understand what the criteria is for exponential growth, they are not scientists but imposters fooling the masses.

I'm not here to make friends but to stimulate thought (and clearer thought) and to get the experts to tell come out of the woodwork and speak the truth. I was successful in doing that here and hopefully, some people (including the oh so smart scientists) also got to learn some 9th or 10th grade math that in reality should be extremely intuitive to them.

 

[Stinky Peterson]

quote:

---

Originally posted by unDummy:

Is there an increase in wear after an oil change? Is the new oil washing the engine and whatever was left behind by the previous oil?

Or, are some people installing completely dry oil filters causing all that initial wear waiting for the new filter to fill up?
Is the filter filtering better after 5k miles or when new?

Is a UOA worth looking at anymore?
Or, should TBN/particle counts be mandatory and not just an option?

Should the oil filter media be analysed?
Without a UFA(used filter analysis), is anything in the UOA accurate?...

---

Hi Un,

You ask some good questions. I cannot say much about engines and higher wear rate after oil changes. If Ted has seen this then I have no reason to doubt it as he a straight shooter. In most of the engines we work with the oil is sampled when it is changed so we don't get to watch them over time. In hydraulic systems when the oil might be used for 10,000 hours and we have 30 samples I have seen this and believe it is due to the new oil being dirtier than the old oil. Under our contamination control program we filter new oils before they can be used but most people don't do this and assume the oil is clean.

In our industry oil filters are not prefilled and I do not think this is detrimental. I know others disagree but I think this is one of those areas where is it hard to measure the effects. A few months ago I was arguing pretty strongly against prefilling filters because of the contamination we were seeing in new oils. Since then people have sent me about 40 quarts of new oils for testing and I have revised my opinion after observing that most of the oils in quart container are pretty clean. For bulk oils I still do not recommend prefilling filters unless the oil has been prefiltered. I also do not think it is necessary to fill the filter.

UOA are worthwhile and can be useful in determining oil change interval and finding problems before they become catastrophic. The method we use to determine change interval is to first establish a trend by taking 3 or 4 samples at the normal interval. Then if the recommended drain is 4000 miles take a sample at that time but don't change oil. Run another 1000 or whatever you are comfortable with and take another sample. If all is well continue doing this until you observe an sharp increase. Once you have established the break point then back off by 500 miles or so.

I think TBN has it's place but not for routine sampling. If you are trying to stretch you changes by 2-3 times then it is a good idea to do some spot checks to see how it is holding up. In most cases if you have a TBN issue it will show up as elevated wear elements, viscosity change, etc. You will get more benefit by doing more frequent sampling than from TBN. We use FTIR to monitor oil condition and this works well for our purposes. When people start adding additives and mixing oils then the FTIR reading will contain error.

On the topic of filter analysis this is where most people are missing the boat. There is as much information in an oil filter as there is in an oil analysis! The problem is it doesn't have numbers so it's a little harder to understand. We use a special filter cutter that is not unlike a pipe cutter. It cleanly cuts the base off the filter so you can remove the canister and inspect the paper pleats. The trick is to do this every time you change filters so you will learn what a normal filter looks like. You will see some metal particles but they should be small and almost 2 dimensional.

As for particle count, it would definitely provide some valuable information. We us PQ in our lab so we are know when large iron particles are present but most labs don't have this capability. We will be adding a new instrument in the near future that will let us do particle counts on engines and in fact quantify the type of wear particles be generated. Until then it is available for an extra charge but here again I don't recommend it for routine analysis (because of the cost). The particles of interest will generally settle out of the oil so before you send in your oil samples turn the bottles upside down on the lids while they are still hot and let them set for a couple of hours. Then turn them over, remove the lid, and see what's there. Believe it or not this is what we do when our particle counter is broke down and we find just as many abnormal samples.

 

[Stinky Peterson]

1911 when you say things like: "...alledged scientists..." and "...the oh so smart scientists..." you are insulting members of this board. Don't think for one minute that I don't understand the math just because I don't present a mathematical argument. My job requires me to communicate daily with customers who have little or no training or understanding of science or mathematics so I try to do it in a way that does not intimidate or make them feel inadequate. When I say "fancy" math I am only referring to math that most people do not use on a daily basis.

 

[labman]

Wait a minute. 5 micron particles aren't going to rip 5 micron chunks out of your bearings. When I attack a piece of steel with a 24 grit sanding disk, I can hardly see the particles I am cutting away. Kind of a dark gray film accumulates, but nothing approaches the 710 micron grit abrasive. So 5 micron contaminates are not going to accumulate exponentially. We agonize over filters here, but even the best of the full flow filters give their efficiencies at 15-20 microns.

I have never had to replace an engine or bearings due to wear. There just isn't much payback to reducing wear. Has the EPA missed an important way of reducing emissions and fuel usage? If UOA's measure total metals, even sub micron particles and ions, etc., then they should be a good guide to wear. Of course, if metals are low because the smaller particles were left behind in the Charmin, bypass filters may look better than they are for wear, but could still reduce emissions and fuel usage. Pre oilers are not going to have much effect on engine life, but could reduce the formation of problem particles. Again, does anybody have data showing pre oilers increase gas mileage?

This topic certainly has shaken up the troops, 2 huge threads over night.

 

[1sttruck]

From a previous post. Looking closer at the data used to justify longer change intervals it appears that in fact it justifies shorter change intervals. I may not have developed the best way to determine rates for each interval, but I do know that you can't just average the rate at whatever the current interval is; 10.5 does not equal 6.3 :^)


http://theoildrop.server101.com/ubb/ultimatebb.php?ubb=get_topic;f=1;t=010116#000016

Using the the synthetic oil study at http://neptune.spacebears.com/cars/stories/mobil1.html , and looking at the Mobil 1 5W30 Super-Syn Formula data, we have for the first six 1k samples:

miles, iron, make up oil

1k 10 0
2k 12 0
3k 14 .5
4k 19 .5
5k 23 .5
6k 26 .5

It's evidently a 5.7l engine with about a 6 qt sump. Looking at the iron rate normalized for sump + make up oil, engine displacement and mileage, we have the rates shown in the second column below. We need to look closer though as the rate at 2k states 6.3 but that's different than the 10.5 rate at 1k, so we need to determine what the rate was for the second 1k miles, then the next 1k miles, etc., which is what we have in the third column. Since there is a big jump going from 3k to 4k, the data suggests that it's reasonable to change it at 3k. I'll take a closer look at how to determine the rates for each sample interval, as so far I've done it this way just because it's easy to do.

1k 10.5 10.5
2k 6.3 2.1
3k 5.3 3.3
4k 5.8 7.4
5k 6.1 6.9
6k 6.1 6.2

 

[1sttruck]

What got me started on questioning the value of UOAs in determining wear were articles noting that in order to properly determine wear rates in diesels one needed to use 'conditioned oil', which is used oil. The reason is that the wear rates were noticeably higher than fresh oil. The proposal was to have NIST develop a standard conditioned oil for wear testing, starting with adding abrasives to new oil. Since these types of tests were monitoring actual wear on parts as opposed to ppm levels in a UOA, one needs to look closer at what is actually being measured in a UOA. As a consequence I started promoting a 'films theory' to try to explain why UOAs may not always be reliable.

I also noticed that UOAs didn't seem to be used much for wear testing, although they're considered an essential part of a maintenance program, and are obviously useful for determining the condition of oil.

 

[pscholte]

It's easier for me just to listen to the Elves. Selecting the correct motor oil does not require an advanced degree in either Physics, Mathematics or even 16th Century Tibetan Literature. If it did that would be the end of my enjoyment.

 

[bulwnkl]

quote:

---

Originally posted by labman:
Wait a minute. 5 micron particles aren't going to rip 5 micron chunks out of your bearings. When I attack a piece of steel with a 24 grit sanding disk, I can hardly see the particles I am cutting away. Kind of a dark gray film accumulates, but nothing approaches the 710 micron grit abrasive. So 5 micron contaminates are not going to accumulate exponentially.

---

Thank you labman, you beat me to it.

{Hoping my tone/approach will sound familiar to the object thereof:}
1911, you don't seem to grasp anything beyond simplistic 9th or 10th grade math. If you did, you wouldn't compare exponential growth in systems where there is only 1 "particle" or "particle size" to a system in which the possible particles are infinite. You would also not ignore or discard the effects of the filtration system, film thickness, additive system, different particle hardnesses, etc, in an effort to satisify your beligerence. Furthermore, you haven't actually advanced a point in this thread. You've made several attacks but have made no effort to construct anything.

Stinky Peterson has been trying to give you a free lesson in the most basic, very far pre-9th-grade courtesy and manners and you just don't get it. First he tried by example and then by adressing you and the problem politely but directly and you still can't seem to grasp this basic principle.
{End assumed tone/approach}

Perhaps you could construct and advance a coherent hypothesis in a way that will invite discussion without being derogatory, keeping in mind the wide variation of the audience on this website.

 

[pscholte]

Quite frankly, 1911's form of argumenation reminds me VERY STRONGLY of another member (wish I could remember his user name) who appeared, disturbed the waters, apparently did or did not achieve his desired effect and so, as quickly as he appeared, disappeared. I wonder....

 

[buster]

I'm getting lost in the mumbo jumbo....what are we debating here?

Here is what I think it comes down to:

1. IF under the same EXACT conditions Oil A shows less wear across the board to oil B, then yes, you can assume that oil A is providing better protection.

2. This does NOT mean your engine will wear out faster on oil B simply bc engines die from other reasons like Dirt.

3. What makes all this difficult though, and can lead to confusion/debates or whaterver, is that your dealing with Chemistry. Chemicals react differently with different metals, hence the different UOA's in different engines with different oils.

Redline for example reacts with bearings of certain metals, however it's not WEAR from friction. It's RL's Chemistry reacting with the bearing materials, layer down it's protective layer and dispersing small particles.

These are my observations and opinions as of 4/11/05. That could change anyday.

Seriously though, unless your seeing HUGE differences in wear from one oil to the next, it really doesn't matter. I'd be willing to bet Patman's Vette lasts no longer with GC then M1/Amsoil. But why not go with the lower of the two for peace of mind and to a very, very small degree, better protection.

 

[Steve S]

Amsoil seem to value uoa's as an indicator of wear.

 

[Doug Hillary]

Hi,
IMHO UOA's are given a supra position on here for monitoring "wear" and probably without too much justification but for many BITOG posters they are a fun "event" and should be viewed that way. Enjoy!
Variance of one or two ppm (for example) on one sample over another, or when using one oil against another are really not too meaningful in the scheme of things - or in the life of an engine

For many of us who use UOAs professionally they are but a maintenance management tool and an accurate snapshot of the oil's condition at a point in time
To be meaningful the component Manufacturer's limits on the wear elements they hold critical need to be known or wear tracers installed that are known qualities
Oil condemnation limits covering wear metals and oil condition factors are available for some engines and they have been for the last several decades! They are meaningful and helpful

As technologies move on, the real place that UOAs have will become increasing evident and this IMHO could range from obsolescence to critically important - we shall see!

As to the accurate measurement of wear the real play is in using Radioactive Tracer Technology (RATT). Using the two accepted test methods, Surface/Thin Layer Activation (S/TLA) or Bulk Activation(BA) much can be learnt in real time. The true effects of varying metallurgy, engine design issues, speed, load, heat, filtration and etc etc. can be faithfully replicated and to levels never before possible. Simultaneous wear measurement of individual components like rings, bores, bearings, valve gear and etc. is possible using RATT

And importantly for us all to realise the effects of an individual lubricant's chemistry can be accurately measured the same way!

Real time wear/destructive measurement on whole vehicles/engines/etc when used in an accurately simulated operating environment from anywhere in our World is old hat and has been around for the last two decades or so
RATT speeds the process enormously and I can see where all of this will have/has had an obvious effect on the Approved oils Lists of the major vehicle Manufacturers. I believe that this is but ONE reason why some Approved lubricants are only available via the Manufacturer's network

So mathematical wear modelling using UOAs, unless the huge variable called "the total operating environment" can be accurately quantified, will never be too meaningful IMHO

Regards
Doug

 

[1sttruck]

UOAs might be an indicator of wear, but since we are dealing with a lot of variables it might not. We've seen:

1. UOAs that on exhibit high wear metals but the wear metals drop to low levels once the engine is cleaned. This suggests that the wear metals were probably coming from sludge and varnish.

2. UOAs that looked good on engines that were sludging up. The typical UOA doesn't seem to pick up oil breakdown very well.

3. Lots of varnish on some engines even when synthetics were used, and some sludge in others when dino was used, with good UOAs over a long period; one really needs to dissolve the sludge and varnish and add the wear metals back in to the previous UOAs.

4. UOAs as they're commonly used telling us that the longer the change interval the better the oil gets, but wear studies telling us that to make used oil you could add abrasives to new oil.

etc., so 'wear' in a UOA is not always what it seems.

 

[Jay]

As far as bypass filters delivering increased gas mileage is concerned--I don't think you need one to test that theory. Does gas mileage steadily decrease with a conventional full-flow filter from the beginning of an OCI to the end? That is, do you notice better gas mileage when you first install new oil, and poorer gas mileage toward the end of the OCI (if the oil doesn't thicken)?

I sure don't.

 

[labman]

Good point Jay. Jim Fitch brought up some interesting ideas. Before I make any changes to the upkeep of my car and truck, I want to see how they look in the light of real world facts. Perhaps I never remember noticing much difference do to my fanatical new dino every 3 months. According to his ideas, people running 10K+ intervals should see their mileage in the tank at the end.

 

[rob-the-oil-nut]

I emailed Terry Dyson and suggested he have a look at this thread. He spent a few minutes only, he is super busy, (something about a ZOil Mass spec test), he might post something Monday or Tuesday.
He replied, however, "One thing you might post for me is that when Dyson Analysis interprets an analysis we are using a data based on info NO one else has access to because of the years of testing, mostly proprietary, we do."

Keep in mind most commentators on BITOG are NOT experts, [they are un-trained...or worse; like me -Rob] so to use that site as a example for how folks can’t properly determine meaningful data is poor science at best.

FWIW / FYI
Rob