2001 Dodge Ram 2500, Valvoline High Mileage Diesel, 21k miles

Status
Not open for further replies.
Originally Posted by A_Harman
Originally Posted by gman2304
That trucks sump holds 12 qts. Even at 21,000 mile oci's, that's a lot of makeup oil. It seems to have a history of oil burning. There's a lot of fresh oil being poured into it between oci's that might be skewing the final numbers and masking the health of that engine.


This is why I correct the Iron concentration with a dilution factor that takes into account the makeup oil. The Iron wear rate is based on the corrected Iron concentration.


Just asking for a clarification A_Harmon:

In the above chart/table is this the raw data from Polaris or a spreadsheet which contains the calculations with your corrections?
 
Last edited:
This is the last post I am going to make in this thread. Mass spec does not detect "particles ". It detects atoms . You may or may know that the spark that forms the plasma is high temperature (6000K the same as the surface of the sun) but short duration and low total power (a lot of volts but milliamperes at the most). Not enough to vaporize and ionize particles of more than a few millimicrons in size. This has been confirmed by my daughter who uses mass spec for isotope geochemistry. She says it is not really reliable unless everything is in solution. She uses the second strongest acid in existence, HF. You can win this argument if you prove to me oil analysis outfits use HF on samples before mass spec. For $29 I doubt they would use this exceedingly corrosive substance. Otherwise, I allege that the Fe is virtually in solution already ( particles at most millimicron size) that do NOT "settle" and easily fit between the outer edge of the rings and the cylinders to be burned, go out the tailpipe and removed from the sump. If YOUR allegations are true it invalidates UOAs and you need to delete the section.
Do I need to have my daughter contact you or join the forum? She is a multiply published world expert on how mass spec works, she has read most of this thread and disagrees with you.
 
One last thing. I'm sure you know that an engine will shed Fe in a Gaussian distribution with total mass(not number of particles) on the y axis and size or mass per particles on the x axis. A healthy engine produces most particles in the small regime, a sick engine is skewed to the right and might real macroscopic pieces. This can't be measured by MS but the number of millimicron particles and actual Fe atoms goes up drastically. Another way of phrasing my point is that if it can be measured by MS it's small and enough to get past the rings and be burned.
 
Originally Posted by claluja
Any iron particles that reach the combustion chamber in the oil within which they are suspended are going out the tailpipe. No reasonable dispute there.
I would reasonably agree with this.

Any iron particles that leak out of the vehicle in the oil within which they are suspended are exiting the vehicle. No reasonable dispute there.
I would reasonably agree with this.

The iron particles in the oil measured by Polaris are smaller than both ring gaps and ring clearances. Its reasonable to assume that the vast majority of any particles over 10 or 15 microns will have been caught by the oil filter. No reasonable dispute there.
I would reasonably agree with this.

OP lost eight (8) quarts of oil. That's a lot (there must be some pretty big gaps somewhere!). No dispute there.
I would reasonably agree with this.

How can anyone seriously dispute A-Harman's calculation? Answer: can't be reasonably disputed. Any good engineer will calculate the worst-case scenario, because that is the most important number. That's what the OP did, and that's certainly what I would do. I want those guys designing bridges, cars, boats, and skyscrapers.


But are we assuming that the routes of exit (exhaust, or external leaks) are the ONLY means of oil volume exit?
As we know, lubes are volatile and we score them on a NOACK scale. Lube volume loss due to evaporation is to be accounted for also, right?
Does the evap portion of the volume loss carry 100% of it's solids out? Or maybe none of them? Or some portion of them?
Has anyone ever used a "catch-can" for the PCV and then taken that oil collected and had it analyzed? Can we quantify what remains in the vapor condensed lube in terms of elements? What's the ppm of metals and additive elements of the remainder? Even if we did do a UOA on the catch can contents, how do we really know if 100% of the vapors are condensed, or only a portion of them? Do we still burn "some" portion that escapes even the catch can? And if so, what portion of the vapors moving past the catch can have elemental wear particles?

Further, if we presume for a moment that 100% of what leaves a system (regardless of the method) carries with it 100% of it's elemental contamination and additives, then why are we adjusting the formula? In a manner, you're making my argument for me, I believe. Here's why ....

If we take into account that you and Charlie believe that 100% of the lost volume essentially carries out 100% of it's encapsulated holdings, then what remains must be no more or less contaminated than what left. Given that wear rates are reasonably steady (past 3k miles) in OCIs, all the way out to 15k miles, then we can essentially say that what goes "in" the oil in terms of wear metals is a constant, steady input. For every 1k miles, you'll get "X.y" ppm of Fe, Cu, etc. At any given time in the life of the sump, let's say you have drips coming out of a seal. That drip would be no more or less laden with wear metals that what remains in the sump. It's not like leaks are going to preference metals for exit. What leaves a system (via leaks or exhaust) is very likely to be extremely close, essentially identical, to what remains in the sump in terms of concentration.

And if there's some major loss of fluid, two things are happening; you're adding more lube that's (reasonably) clean with no contamination to speak of, and you're loosing lube that is contaminated.
So if the lost volume takes with it 100% of your elements it's holds in suspension, and the incoming lube carries in nothing with it in terms of wear metals, then why would the current sump volume at any given time not represent a reasonably accurate concentration irrespective of the volume added???
In short, what leave the system via different routes should carry out "X" ppm of Fe, "Y"ppm of Cu, "Z" ppm of Al .... etcetera. And that loss would represent what was "current" at the time of loss, right?
And, what remains in the system would continue to be contaminated with wear metals at the same rate as what left.

- If we were to allow a sump to drop 30% below it's marked level, then do a massive top off, and then take a UOA immediately 5 miles afterward, I'd have to agree that there's some error being induced because we've diluted the oil and taken an immediate measurement for such a large volume change.
- If we only allow a sump that holds 12 quarts to drop perhaps a quart and then get topped off (only about 8% loss), and then we top off keeping the loss to around 8% or so, AND we take a UOA perhaps 1k or 2k miles into the refreshed top-off, then it's reasonable to think we've allowed the system to normalize again; the input of wear metals has reasonably been allowed to stabilize. AND, since 8% loss is well within the NOACK score of many lubes, it's also reasonable to think that the oil loss may (not will as an assurance, but "may" as a possibility) be leaving some manner of metals behind (to a degree that I think none of us have data for).

What we know easily is how much lube it lost.
What we don't know is the manner of it's loss (leaks, burning in combustion, evap) and how much is carried out via these methods.
I would agree that external leaks and burning are likely going to take a large portion of the holdings.
But we cannot ignore loss due to vapors, and we have no data to know how much of the metals are being carried out.

Further, if what is being lost has 100% of it's wear metals going out with the host lube, then what remains should be reasonably representative of the "as current" contamination concentration, (as long as you don't do a massive sump refill and then immediately UOA; that would distort the concentrations).

We know from UOA data that wear rates are actually very consistent, barring any catastrophic event occurring. We can see engines that don't have much lube volume loss exhibit very consistent wear rates after the first few thousand miles after an OCI. Typically these rates only vary by a few tenths of ppm. Very consistent wear. And so, even if we have an engine that sees large volume loss (and subsequent make-up), there's no reason to think the wear rate of the engine changes substantially. We could debate the merit of the TCB being altered by the fresh detergents, but again, there's no study data that addresses PARTIAL OCI volume shifts, and it would be further complicated by the topic of what would be "residual" from a previous OCI ... Into the weeds we go.
What we can accept is that the engine will wear at a fairly steady rate. So the make-up oil, if replace at fairly frequent intervals, in small volumes, will not cause a condition to greatly alter the "as current" contamination concentration value.


So again I'm asking why you all are trying to adjust a mathematical formula for a condition that does not have proven data as to where ALL the wear metals go (does the evap carry or drop elements?), and the make up volume carries in nothing to the equation.
 
Originally Posted by m37charlie
This is the last post I am going to make in this thread. Mass spec does not detect "particles ". It detects atoms . You may or may know that the spark that forms the plasma is high temperature (6000K the same as the surface of the sun) but short duration and low total power (a lot of volts but milliamperes at the most). Not enough to vaporize and ionize particles of more than a few millimicrons in size. This has been confirmed by my daughter who uses mass spec for isotope geochemistry. She says it is not really reliable unless everything is in solution. She uses the second strongest acid in existence, HF. You can win this argument if you prove to me oil analysis outfits use HF on samples before mass spec. For $29 I doubt they would use this exceedingly corrosive substance. Otherwise, I allege that the Fe is virtually in solution already ( particles at most millimicron size) that do NOT "settle" and easily fit between the outer edge of the rings and the cylinders to be burned, go out the tailpipe and removed from the sump. If YOUR allegations are true it invalidates UOAs and you need to delete the section.
Do I need to have my daughter contact you or join the forum? She is a multiply published world expert on how mass spec works, she has read most of this thread and disagrees with you.


1) Thank you for explaining your viewpoint; this is what I'm looking for.

2) This is not the way I've had it explained to be previously. However, I'm not convinced you are wrong; only in opposition to what I've heard in the past. It may well be that you are correct and my basis for understanding the MS process is wrong. I need to do some more research. At least you've given me a place to start other than what I already have preconceived from other sources. I would welcome a contact from your daughter; if you'll PM me I can give you my contact info.

3) If we assume that the loss of volume carries with it the vast majority (if not all) of the elements, then would not my vision of the "as current" volume concentration be correct? (see my thread above and continuation below)

We cannot really know (at least not yet) what portion of the vapor loss does or does not contribute to residual concentrations. I've not seen data on this, nor any study. Does the vapor loss carry 100% of it's holdings? None? Some portion?

What I disagree with is how the AHarmon method is attempting to adjust for a condition that is likely not really in need of adjustment.
I will, for the sake of the conversation, acquiesce to your position; 100% of the lost volume takes with it essentially 100% of the elements held in suspension.
So if we are introducing new volume that carries in near-pure (no wear metals) lube, and we allow for that lube to be contaminated for 1k or 2k miles by an engine that has very steady wear rates, then why in the world would we try to adjust a formula? It should be this simple:
- make all your top-offs reasonably frequent; this will decrease dilution effects and also greatly reduce any TCB shift
- take your UOA samples perhaps 1k or 2k miles after the top-off
- what remains in the sump would be very representative of the true nature of the wear trends; the ppm concentration should be very near "true"


In all my data (over 15k UOAs of all manner of engines), I can find essentially no correlation between oil consumption and wear rates. (There are a precious few examples, but they are known to be from engines with high failure issues and do not represent the vast majority of reliable engines). UOAs that have high consumption (say 50% sump volume or even more) over the OCI duration, do not have wear rates much different than those which have very low oil consumption (say 20% or less over the OCI). I have UOAs from engines that have high consumption and those with low consumption. In OCIs where there was no volume added, the perceived shift in wear rates is lower than the standard deviation. IOW, in engines that do have higher consumption, but don't get volume added (the OCI was short enough that the net sump level was not to a danger point prior to the OCI), the wear rates were not appreciably different than engines with low consumption. Real data does not exhibit statistical difference between engines with high and low consumption, even when there is no volume topped off.
Example: Several UOAs taken from the same engine, at different times, under different conditions
- UOA sample has an Fe wear rate of 1.8ppm/1k miles. It has low consumption over a 7k mile OCI (the level on the dipstick goes down very little; less than 10% change in volume and no oil added).
- Next UOA sample taken later has an Fe wear rate of 1.9ppm/1k miles. It has higher consumption over a 7k mile OCI (the level goes from "full" to "add"; perhaps a 20% volume shift; volume added incrementally).
- Later yet again, another UOA sample taken indicates 2.7ppm/1k miles of Fe. There was a volume loss of that same 20%, but no oil was added. This higher wear rate was due to extremely heavy loading; throttle at 75% or more most all the time, with EGTs at 1200-1300F all the time (max of 1350F by OEM design).
Macro data shows that this engine family will have an average of 2.3ppm/1k miles at 7k miles exposure, with a normal distribution variation sigma of .4ppm; hence anything up to 2.7ppm/1k miles is "normal" for the first standard deviation. The UOA sample where 20% volume was added contrasted to the UOA where 20% volume was not added is not appreciably different because of oil added; it's because of the extreme use factor applied in one of the OCI cycles. And yet ALL of the wear rates were within "normal" distribution. My point is that wear rates really don't change much, regardless if you add oil or not; certainly not in a manner that can be credibly discerned by statistical analysis. The potential changes in wear rates due to oil volume addition is smaller than the natural variation, and therefore we cannot claim to have an "ah-ha!" moment that would make us believe we can attribute an output to an input. Regardless of the make-up factor, wear rates don't change much, even in examples where the same volume loss may or may not have enjoyed the addition of make-up oil. Much of this depends upon the nature of the individual engine performance (micro data). Since no one here has proven to me that they indeed have solid micro data, we rely on macro data for proof.

You've not convinced me that a correction factor should exist and be implemented. If we accept that what leaves a sump carries out 100% of it's holdings, then we can presume that there is no increase in concentration. And if we're careful not to be taking samples after large sump top-off additions, the wear rates will be very steady. So why adjust a formula for concentration when what you claim takes all the stuff with it, and what remains is not altered to any statistical degree by the wear rates?


Charlie - I'm not trying to be argumentative. I'm trying to debate this from a point of logic. Whereas I don't agree with you at this point, I do respect your opinions and knowledge. If you want to convince me you're right (and you may well be), then show me the math and help me understand why it's necessary to adjust a formula for concentration of a metal when the "as current" sample should be very true to the overall condition. If you wish to exit the conversation, that would be a loss, but I'll offer the apology if I've made you mad. I have no intent to irritate you; only challenge you to defend your position. My position is that it is unnecessary to adjust for sump volume loss (presuming the conditions I state for the sample methodology) in terms of wear metals.

My position is that, as long as you top-off reasonably frequently, and do so in smaller quantities, and take your UOA sample after the sump is given a change of homogenize, then the natural variation of wear trends will far outpace any wear induced by the top-off. And if you grossly alter your formula (as you and AHarmon suggest), then you're falsely accrediting a shift in wear trends when no adjustment is necessary in the first place. Even if you do this, you have very little likelihood to be able to "prove" your formula is needed when a typical 3-sigma distribution will be far larger than the correction you put in.

I believe you are correcting unnecessarily, and even when you do, your correction cannot be discerned from the typical variation of normal everyday expectations.
 
Last edited:
Very interesting discussion.

One comment: Modern engine oil additives contain a thiadiazole-type chemistry, a multi-functional component which acts as an Ashless-High Temperature Antioxidant, Antiwear/Antiscuff, Friction Reducer,Corrosion Inhibitor, Extreme Pressure, and Metal Deactivator. that contributes anywhere from 1 to 3 ppm of iron alone.

I have proven this when I was making my own additive formulae. I had pondered why I was getting 1-3 ppm of iron in a clean HDPE mixing drum and the thiadiazole-type compound came in 5-liter brown glass bottles.

The base oils were PAO's, esters and other oils which came in 55 gallon steel drums, but when those virgin oils were analyzed, there was zero ppm of iron found. The prevailing theory at the time was that the iron came from the steel drums. I could never accept this because the drums were lined with a thin protective film to keep base oils from reacting with the steel.

I made up a number of finished 100mLiter lubricants with and without various compounds including the thiadiazole chemistry, and wola! Analyses showed the samples without the thiadiazole chemistry had zero ppm of iron but the samples with thiadiazole chemistry showed 1-3 ppm of iron.

So this chemistry may skew any computations involving ferrous analyses.
 
Last edited:
Originally Posted by MolaKule
Very interesting discussion.

One comment: Modern engine oil additives contain a thiadiazole-type chemistry, a multi-functional component which acts as an Ashless-High Temperature Antioxidant, Antiwear/Antiscuff, Friction Reducer,Corrosion Inhibitor, Extreme Pressure, and Metal Deactivator. that contributes anywhere from 1 to 3 ppm of iron alone.

I have proven this when I was making my own additive formulae. I had pondered why I was getting 1-3 ppm of iron in a clean HDPE mixing drum and the thiadiazole-type compound came in 5-liter brown glass bottles.

The base oils were PAO's, esters and other oils which came in 55 gallon steel drums, but when those virgin oils were analyzed, there was zero ppm of iron found. The prevailing theory at the time was that the iron came from the steel drums. I could never accept this because the drums were lined with a thin protective film to keep base oils from reacting with the steel.

I made up a number of finished 100mLiter lubricants with and without various compounds including the thiadiazole chemistry, and wola! Analyses showed the samples without the thiadiazole chemistry had zero ppm of iron but the samples with thiadiazole chemistry showed 1-3 ppm of iron.

So this chemistry may skew any computations involving ferrous analyses.





Did you ever try using a thiadiazole-type compound that came in a container other than a glass bottle? Lots of iron in the glass.
 
Originally Posted by claluja



Did you ever try using a thiadiazole-type compound that came in a container other than a glass bottle? Lots of iron in the glass.


I appreciate your "cause-affect" question and yes as a matter of fact, I did have a 1-Liter bottle delivered from a different chemistry house in a HDP2 bottle with the same showing.

But then we have to look at the atoms of iron sulfide in terms of diffusion theory in tightly-bound atoms in glass crystals. It is mainly iron sulfide that gives rise to the amber color of glass. So if iron sulfide atoms could diffuse into the thiadiazole compound, it should have also raised the sulfur content in the analysis as well, which it didn't.

Besides, these thiadiazole molecules are very large and would have to diffuse into and in between the very small crystal lattices of glass in order to exchange any other atoms.
 
Last edited:
Originally Posted by MolaKule
Originally Posted by claluja



Did you ever try using a thiadiazole-type compound that came in a container other than a glass bottle? Lots of iron in the glass.


I appreciate your "cause-affect" question and yes as a matter of fact, I did have a 1-Liter bottle delivered from a different chemistry house in a HDP2 bottle with the same showing.

But then we have to look at the atoms of iron sulfide in terms of diffusion theory in tightly-bound atoms in glass crystals. It is mainly iron sulfide that gives rise to the amber color of glass. So if iron sulfide atoms could diffuse into the thiadiazole compound, it should have also raised the sulfur content in the analysis as well, which it didn't.

Besides, these thiadiazole molecules are very large and would have to diffuse into and in between the very small crystal lattices of glass in order to exchange any other atoms.






Sounds like you've got that covered. Btw, guessing about half the iron in that glass is bonded to O, both ferric and ferrous. Other half to S. Rough estimates.
Not sure if that matters, as I'm not familiar with your compounds. Materials leach out from glass quite a bit depending in circumstances, with Na being the most well known culprit.
 
So let me see if I have this straight... you run Synthetic RT6 for many years with much shorter OCI's and then switch to a conventional oil (VHMD) and go 21,000 miles on it?

I guess it really doesn't matter because at the rate that truck is using oil, you're constantly replenishing the sump! Does it leak or burn oil? both?
 
Originally Posted by MolaKule
Originally Posted by A_Harman
Originally Posted by gman2304
That trucks sump holds 12 qts. Even at 21,000 mile oci's, that's a lot of makeup oil. It seems to have a history of oil burning. There's a lot of fresh oil being poured into it between oci's that might be skewing the final numbers and masking the health of that engine.


This is why I correct the Iron concentration with a dilution factor that takes into account the makeup oil. The Iron wear rate is based on the corrected Iron concentration.


Just asking for a clarification A_Harmon:

In the above chart/table is this the raw data from Polaris or a spreadsheet which contains the calculations with your corrections?


Sorry about taking a few months off of this thread. The image is a snip of the Excel spreadsheet where I have entered the raw data from Polaris, then embellished on it by adding my own calculations in the rows titled "Dilution Factor", "Iron (corrected)", and "Iron wear rate".
 
Originally Posted by racin4ds
So let me see if I have this straight... you run Synthetic RT6 for many years with much shorter OCI's and then switch to a conventional oil (VHMD) and go 21,000 miles on it?

I guess it really doesn't matter because at the rate that truck is using oil, you're constantly replenishing the sump! Does it leak or burn oil? both?


No, the chart requires some explanation. In the "Truck Miles" row, if the cell fill from column to column is the same color, that means I took an oil sample without changing the oil. The first samples I was taking in the 2012-2013 time frame was when I was trying to find an OCI that was long enough to take advantage of the long-drain intervals of synthetic without getting crazy. The shortest OCI I ever ran on RT6 is 24K miles. In the pre-history of UOA's on this truck, I ran VPBC faithfully on 1-year oil changes from 2003 to 20012. In those years, that was about 18k mile OCI's.

Then I fell in with the BITOG crowd, and started getting UOA's...

The truck both leaks and burns oil, but it probably leaks more than it burns. It's always been pretty leaky, and I have tried to dry it up over the years, but let's just say the frame is still rust-free because of oil leakage, not because I was taking it to Krown. The main culprit is the oil breather on the gearcase cover. It also sends some oil out the exhaust if it idles for more than an hour. The first couple of runs through the gears generates blue smoke out the exhaust, I think because of a leaky turbine seal.
 
Is this the oil you used? Did it help with oil consumption?

The PCMO Maxlife oils state that they have seal conditioners. The Heavy Duty Maxlife does not. I wonder what makes it "high mileage". My only guess is its thicker viscosity and possibly higher HTHS.

Screenshot_20190515-202015.jpg


Screenshot_20190511-075942.jpg
 
Originally Posted by SavagePatch
Is this the oil you used? Did it help with oil consumption?

The PCMO Maxlife oils state that they have seal conditioners. The Heavy Duty Maxlife does not. I wonder what makes it "high mileage". My only guess is its thicker viscosity and possibly higher HTHS.


Yeah, that was the stuff. It didn't help out with the oil consumption rate.
 
Status
Not open for further replies.
Back
Top