Is testing for Fe an indication of engine wear?

[RI_RS4]

Originally Posted By: demarpaint
Originally Posted By: edhackett
Actually it's closer to 1/4 of a dime.

PPM= mg/L
5L x 100 ppm = 500 mg =0.5g
A dime weighs 2.2g
0.5g/2.2g=0.227

Ed


Good example, to me 100 ppm of iron is a lot of wear then, especially to someone who really cherishes their vehicle and wants to keep them a long time.

Now we go full circle and ask, is the iron really from wear, or some chemical reaction that shows Fe in an analysis? Once again the only true way to accurately measure engine wear is the very impractical tear down. Or just take the analysis results with a grain of salt, and look for dirt injestion, anti-freeze, excess fuel, and bearing wear, then call it a day.


Now you can imagine why I become concerned with a 4.2L V8 that has an 8200 rpm redline with an Aluminum block and Al/Si cylinder walls shows up with 30+ ppm of Fe, in an engine that is broken in, after 6000 miles on the oil, considering that the engine has a 10L oil capacity. This is equivalent to an engine with 3L oil capacity having 100 ppm of Fe in a UOA. No matter how you look at it, that's a large amount of Fe in an engine that has glass cylinder walls.

 

[buster]

Lower the numbers are, the better the report. Sure one could argue that the Fe is so small it's insignificant and that may be the case, however, Fe is Fe and I want less in a UOA, not more.

 

[ClarkB]

Originally Posted By: edhackett
Actually it's closer to 1/4 of a dime.

PPM= mg/L
5L x 100 ppm = 500 mg =0.5g
A dime weighs 2.2g
0.5g/2.2g=0.227

Ed


I'm no chemist but I know enough to know that your math is not 100% correct. When talking ppm and solutions you need to look into moles, density, molecular weights and converting that information to volume. Your assumption is that 1,000,000 parts equals 1 L. This is not so. Your answer might be close for BITOG work, but a chemist would laugh himself silly at your assumptions.

Clark

 

[JAG]

Originally Posted By: edhackett
Actually it's closer to 1/4 of a dime.

PPM= mg/L
5L x 100 ppm = 500 mg =0.5g
A dime weighs 2.2g
0.5g/2.2g=0.227

Ed

That is correct (and concisely written, BTW) only when the element being measured has the same density as the bulk mixture in which the element exists. But when oil is the bulk mixture and the substance being measured is a metal, like iron, densities densities definitely have to be accounted for because of the large difference in densities. That's because UOA elements are measured as a mass concentration. The first line has liters in it which is a volume, so that's where the incorrectness begins.

I'd attempt to write the correct formula(s) but I've got a migraine right now, can't think too hard and don't want that to cause me to mess up.

 

[Nikola]

I am confused now. If it says in UOA 20ppm. Is it for whole sample or per liter or...

 

[XS650]

Originally Posted By: Nikola
I am confused now. If it says in UOA 20ppm. Is it for whole sample or per liter or...


Yes.....

It means that 20 out of 1,000,000 parts by weight are whatever the 20 ppm is.

It means 20ppm by weight, not volume.

 

[XS650]

Originally Posted By: JAG

That is correct (and concisely written, BTW) only when the element being measured has the same density as the bulk mixture in which the element exists.


Additionally, it is only correct when both constituents have a specific gravity of 1.000.

 

[RI_RS4]

Here are some calculations I did several years ago on the Audi V8 engine with Al/glass cylinder walls:

Hypothetical case #1
Quote:
Just for fun, lets assume that the only source of chrome in an engine is in the piston rings, and lets run some simple numbers. For the sake of argument, lets assume the following, some of which are applicable to Audi V8 engines:

V8 engine - 8 cylinders with 84.5 mm bore
Ring circumference = 265 mm
3 rings in each ring pack
24 rings total
total cumulative ring circumference = 6368 mm
ring width = 1 mm
ring depth (od/2 - id/2)/2 = 3 mm
total ring metal volume = 1mm x 3rings x 8cylinders x [(84.5/2)^^2 - (78.5/2)^^2] x 3.14
total ring metal area = 768 cubic mm = .768 cc

Assume that rings are totally composed of chromium (which is reasonably valid for outer coating layer.

Density of chromium = 7.3 g/cc
Total weight of rings = 5.6 g

Density of oil = .877 g/cc
Engine oil capacity 10L
total weight of oil = 8770 g

IF all of the rings were to dissolve into the available oil in this engine, then the chromium impurity rate would be 638 ppm. This, of course, would be total engine destruction.

So, what would be a tolerable total wear on the rings? Lets assume that the engine can tolerate 1 mil of ring wear in the cylinder bore.

1 mil = 0.0254 mm change in total ring diameter

As a 1st approximation, this is .42% of the total ring diameter and therefore of the area of the rings (0.0254/6= 0.0042)

Thus, our maximum allowable wear would be

0.42% x 638 ppm = 2.7 ppm


Hypothetical Case #2
Quote:
If the engine is primarily built from iron and steel, and wear is sort of happening uniformly throughout all those metal surfaces, there's an awful lot of iron wear that can occur before anything catastrophic happens. My previous computation was on the rings. But if you do the same calculation for the cylinder walls, and realize that the bore circumference is swept along the entire stroke, then if the wear is only cylinder wall wear, you have one heck of a lot of metal that can wear before you put a big dent in the cylinder wall tolerances. But, if your assumption is wrong, and iron wear is coming from the cam lobes, the cam followers, or some part of a bearing surface, then all bets are off.

For the same engine, if the cylinders were iron,then:

bore = 84.5 mm the
circumference = 265 mm
total circumference of all cylinders = 6368 mm
stroke = 92.8 mm
total swept cylinder area = 591 square cm

Density of oil = .877 g/cc
Engine oil capacity 10L
total weight of oil = 8770 g

Total iron in oil measured with 100 ppm Iron

iron/oil = 100 ppm

iron = oil weight x 100ppm
iron = 8770g x .0001 = .877 g

So, if .877 g of iron were removed evenly from the walls of all cylinders, how thick would the wear be?

iron density = 7.86 g/cc

.866 g of iron/ 7.86 g/cc = .11 cc of iron removed

Total swept area = 591 square sm.

thickness = 0.11 cubic cm/ 591 square cm
wear thickness = 0.00018 cm
wear thickness = 0.0018 mm
wear thickness = 0.073 mils
wear thickness = 1.8 microns


So, assuming that all of the wear occurs evenly across the cylinder walls, this engine would see 1.8 microns of wear per 100 ppm. That doesn't sound like much. But, if you wanted your cylinders to have less than 1 mil of cylinder wall wear, then your total iron wear could not exceed:

100 ppm/0.073 = 1370 ppm total over the life of the engine.

If you define 1 mil cylinder bore wear as being the acceptable wear limit for the expected life of the engine then total engine life would depend on the following rates

100 ppm per 5K miles = 68.5 K miles engine life expectancy
100 ppm per 10K miles = 137 K miles engine life expectancy
100 ppm per 20K miles = 274 K miles engine life expectancy

Understand that I know nothing about engine wear limits. I think I'm reasonably close with a 1/1000 ths wear tolerance for modern high performance engines. If all the wear is coming from the cylinder walls, then these all seem like reasonable numbers. But, if that level is wear is coming from the smaller bearing surfaces, or from cam lobes, I'd say we'd be in for a heap of trouble. But, even with those cylinder wear numbers, I think Exxon Mobil's definition of running a very long time is different than yours or mine?


And this was the original thread:

Acceptable wear metal ranges

 

[Greaser]

Originally Posted By: buster
Lower the numbers are, the better the report. Sure one could argue that the Fe is so small it's insignificant and that may be the case, however, Fe is Fe and I want less in a UOA, not more.
Just what I was going to say

 

[demarpaint]

Originally Posted By: Greaser
Originally Posted By: buster
Lower the numbers are, the better the report. Sure one could argue that the Fe is so small it's insignificant and that may be the case, however, Fe is Fe and I want less in a UOA, not more.
Just what I was going to say

Exactly, the lower the iron #'s the better the report, unless it can be proven that iron in a UOA comes from the add pack and is essential in protecting an engine from wear. I find that extremely tough to believe. All the math makes my head spin, lower numbers mean less wear. We can debate what the numbers should be, and what might or might not be harmful, but...............

 

[edhackett]

You are right.

PPM is a poor way to report results. Unfortunately it is a convention that is adopted by many industries, but has different meanings depending on the convention used. PPM is often reported meaning mg per Liter volume and it also = mg/Kg.

So, when a lab reports PPM do they mean mg/L or mg/Kg?

I called Blackstone and their standards are in mg/Kg(likely the industry standard). You do need to correct for the specific gravity of oil when calculating the total mass of the analyte in a certain volume of solute.

In the original case, assuming a specific gravity of 0.880, five liters of oil would weigh 4.40 KG and contain 440 mg of Fe. That's 1/5 the weight of a dime.

Ed

 

[Gary Allan]

I have nothing against Amsoil products, but their marketing is somewhat suspect IMO. [/quote

Fair enough. Do you support the notion of "90% of engine wear occurs during start up?" ..knowing that it's implying that "start up" is the first few moments of operation, while start up is defined as "steady state"?

Baths of sludge pouring down upon your car while at a drive through?

A "shredder" where your engine gets disintegrated and "reforms" by itself?


All top notch bona fide companies doing questionable marketing in terms of the impression that they communicate vs. what really occurs.

..but carry on...you have plenty of company in your criticism.

 

[Mark888]

Originally Posted By: RI_RS4
mark

Interesting that you ignored my posting, debunking the suspension theory and the theory that:

Quote:
That assumes that the internal parts of an engine have significant amounts of metal where free Iron (Fe) atoms become unbound from their molecular structure that exist in whatever steel alloy is used for the engine part (not to mention that many engines are primarily aluminum alloy or other alloys these days). I believe that assumption is false.


It seems that all you want to do is argue against Amsoil or "xxx" oil company marketing. That's fine, but do not use lousy assumptions to denigrate the usefulness of oil analysis as one means to track wear - among other useful things. In the right hands, and with the right lab and equipment, much can be done to diagnose and fine tune an engine.

Sorry, I did not mean to ignore you post. I believe that the bottom line is this:

There are too many variables related to where the Fe came from, how much of the engine is actually made from Fe, how well the Fe is suspended by the oil, what was in the crankcase before the last batch of oil was put in, etc, etc.

The most logical and straight forward (and therefore scientifically best) way to test for engine wear it to use a testing procedure where the parts can be examined and the wear can be measured. I realize this cannot be done by the average consumer on their own cars (at least not more than once in most cases), but I am not concerned about convenience, I am concerned about truth and accuracy, and engine tear-down is the only way to get that. Such tear-downs do not have to be done by every consumer, they could be done by independent labs, which have strict control methods to make sure the tests are fair.

As I said previously, I have nothing against Amsoil, they make good products. But I don't like the advertising claims, especially from some of the distributors. Also, the best Amsoil products are not API certified, which is problem for me and many other people who have a car still under warranty.

 

[Boomer]

One thing I have not seen in all of these discussions on iron in oil is the oil filter. These are made from cheap stamped steel with a steel plate with threads. I knw that on many new filters you can rub a black film off of the bottom of the filter which I have always taken to be an oil/iron dust mix resulting from rapid and non-exacting machining. Now could it be that if a new filter was "extracted" and the iron measured in the extract that we would get a high number? For example, is there a correlation between iron in UOAs with a particular oil and the filters used with that oil? Could add packs in some oils be better at washing the filter clean and raising the measured iron in the oil?

 

[RI_RS4]

you do realize that oil analysis has been correlated to engine wear for many years now. The question is not whether increased elementals means increased wear or "bad stuff", it's rather a question of where it came from and whether or not it is important.

Several comments:

1) Yes, everyone would agree that tear down and inspection is best.

2) In lieu of this, trended oil analysis works extremely well.

3) With trended analysis the differences between engine operating conditions and oil type can be easily observed.

If you know your engine, have enough samples, and have used a consistent, accurate lab, then you can tell the difference between oils. On the engines I know best, we've done that. Our data base on the Audi RS4 Aluminum block V8 now has 100 samples over 21 vehicles. We know conclusively which oil performs best.

 

[JAG]

Originally Posted By: RI_RS4
...oil analysis has been correlated to engine wear for many years now.

I'm sure it has but it has also shown so-so or even lousy correlations for one or more reasons other than mistakes.

This book has some good information.
http://books.google.com/books?id=ufT_1_a...result#PPA93,M1

 

[demarpaint]

Here is another thing that puzzles me with a UOA. Lets say you are in search of the best oil for "your car". What I have learned is a good oil in one car might not be as good for another car. I also learned that you need to use the particular oil you want tested more than once to get accurate results for that oil. So it could take the average driver years, thousands of miles, and a lot of cash spent on testing before he lands on the right oil for his application.

Based on that I think I would only use a UOA if I suspected a problem, EG: dirt ingestion, or a coolant leak. Finding the best oil would be too time consuming and costly for me. Its now 34 years of driving and God only knows how many miles, I've never had one done, yet. JMO

 

[Mark888]

Originally Posted By: demarpaint
Here is another thing that puzzles me with a UOA. Lets say you are in search of the best oil for "your car". What I have learned is a good oil in one car might not be as good for another car. I also learned that you need to use the particular oil you want tested more than once to get accurate results for that oil. So it could take the average driver years, thousands of miles, and a lot of cash spent on testing before he lands on the right oil for his application.

Based on that I think I would only use a UOA if I suspected a problem, EG: dirt ingestion, or a coolant leak. Finding the best oil would be too time consuming and costly for me. Its now 34 years of driving and God only knows how many miles, I've never had one done, yet. JMO

Yes, there is a cost to having perfect information, and there is a benefit. One has to weigh the costs versus the benefits, because it is very easy to spend more money and time getting perfect information that it is worth (or more money that it would have cost to just get the best quality product at the outset).

 

[Mark888]

Originally Posted By: RI_RS4
you do realize that oil analysis has been correlated to engine wear for many years now. The question is not whether increased elementals means increased wear or "bad stuff", it's rather a question of where it came from and whether or not it is important.

Several comments:

1) Yes, everyone would agree that tear down and inspection is best.

2) In lieu of this, trended oil analysis works extremely well.

3) With trended analysis the differences between engine operating conditions and oil type can be easily observed.

If you know your engine, have enough samples, and have used a consistent, accurate lab, then you can tell the difference between oils. On the engines I know best, we've done that. Our data base on the Audi RS4 Aluminum block V8 now has 100 samples over 21 vehicles. We know conclusively which oil performs best.

I am certainly not saying oil analysis is worthless, but I do wonder about its relevance for measuring engine wear, especially when comparing two similar products.

With regard to whether oil analysis has proven to have a high correlation with engine wear, how was that determined? Did someone do an oil analysis and then tear down the engine to verify the oil analysis correlation?

My suspicion is that when significant engine wear is occurring, you might be able to detect that in an oil analysis. But when comparing similar high quality oils (such as the equivalent synthetic oils from M1, Castrol, Amsoil, etc) the amount of engine wear is so small that it would be difficult to determine which had the most engine wear even with a tear-down, much less correlate that to an oil analysis.

 

[Gary Allan]

Quote:
I am certainly not saying oil analysis is worthless, but I do wonder about its relevance for measuring engine wear, especially when comparing two similar products.


I'd say that it's a fair indicator of an un-indexed rate of wear. You'll have no idea what it translates out to in miles left or .000's removed ..or from where, but it will, assuming all other elements being somewhat equal, show the change in rate.