20 Wt vs 30 Wt Iron Wear Levels from BITOG UOAs

[OldCowboy]

A Comparison of Iron Wear Levels Using BITOG Reported UOAs
By OldCowboy

The first 60 pages of 20 weight and 30 weight UOAs have been compiled into a spreadsheet database. This is too much data to present in a forum topic, but the data will be provided to anyone who sends me a PM.

The following criteria were used to determine if a UOA was excluded:

- Samples with unit mileage less than 20,000 were not included based on the assumption that they may still be showing high wear levels due to ongoing break-in.

- Samples with additives (MMO, Auto-RX, etc.) were not included due to unknown effects of the additive.

- Samples with make-up oil were not included. This is because I had no way of knowing what effect the addition of make-up oil had with regard to the wear metals.

- There is only one sample included per OP/engine/oil combination to ensure the data were not skewed by multiple nearly-identical samples.

- Samples that did not have iron as the highest wear metal were not considered unless the UOA indicated that this was a normal condition for that particular type engine.

- Samples that had significant fuel or water contamination were not considered due to higher than normal wear to be expected from these samples.

- Samples with less than 2000 miles on the sample were not included.

- Samples which did not have the engine identified and could not be determined via Internet search were not included. This occurred when a model had multiple engines available.

At this time there are 245 samples in this database.
- 112 samples from 4 cylinder engines (34 w/ 20 wt and 78 w/ 30 wt)
- 87 samples from 6 cylinder engines (22 w/ 20 wt and 65 w/ 30 wt)
- 46 samples from 8 cylinder engines (17 w/ 20 wt and 29 w/30 wt)

The database contains the following data fields:

Motor Oil & Viscosity
Wear Metals (Fe, Cr, Ni, Al, Cu, Pb, Sn), normalized to ppm/1000 miles
Original Poster of the UOA
Date the UOA was posted
Engine Make and Displacement
Oil Mileage
Unit Mileage

It has been stated in many documents/references that iron is the only wear metal which consistently increases as mileage increases. Therefore, this white paper will only focus on iron wear levels.


4 CYLINDER ENGINES

20 Weight vs. 30 Weight - Iron ppm per 1000 miles

Code:


20 Weight 30 Weight

Minimum Value 0.50 0.44

Maximum Value 3.24 4.76

Average Value 1.48 1.61

Standard Deviation 0.71 0.96

95% Confidence Value 0.50 0.21

95% Confidence Interval 0.98-1.98 1.40-1.81

Number of Samples 34 78


The 30 weight oils had a larger spread in iron wear levels (0.44 - 4.76 vs. 0.50 - 3.28) and a slightly higher average value (1.61 vs 1.48). However the 95% confidence value for the 30 weight oils was smaller (0.21 vs 0.50), mainly due to the larger sample size. It should be noted that the mix of engines in both the 20 weight and 30 weight UOAs was very similar.

The 95% Confidence Intervals for the 20 weight samples and 30 weight samples overlap. Thus, there is probably no statistically significant differences in the iron wear levels between 20 weight and 30 weight oils in the 4 cylinder engines reported in these UOAs. A student'sT-test run on this data indicated that the iron wear levels between the 20 weight and 30 weight oils were NOT statistically different at the alpha = 0.05 level.


Differences Among 20 Weight Oils

There were some noticeable differences in iron wear levels between the various 20 weight oils reported in these UOAs. However, many of the oils were only represented by two or three sample reports. It was arbitrarily decided to only perform statistical comparisons on oils that had five or more sample reports.

Among the 20 weight oils, only two oils had five or more samples reported. These were Amsoil and Mobil 1. The statistical data associated with these two oils are shown here:

Code:


Amsoil 20 Mobil 1 20

Minimum Value 0.50 1.02

Maximum Value 1.80 3.10

Average Value 1.10 1.81

Standard Deviation 0.52 0.70

95% Confidence Value 0.42 0.49

95% Confidence Interval 0.68-1.52 1.32-2.30

Number of Samples 6 8


There is only a slight overlap of the 95% Confidence Interval between these two oils, so a student's T-test comparing the two sample groups was conducted. According to the T-test, there was a statistically significant difference between the two sample groups with an alpha value of 0.05. Thus, it can be said that the Mobil 1 20 weight sample group shows more iron than the Amsoil 20 weight sample group with a confidence of 95%.


Differences Among 30 Weight Oils

Five oils in the 30 weight oil group had five or more UOAs reported. The summarized data is presented below:

Code:


Amsoil Castrol Edge Castrol Syntec

Minimum Value 0.59 0.73 0.67

Maximum Value 4.37 2.95 2.77

Average Value 1.80 1.56 1.72

Standard Deviation 1.15 0.88 0.80

95% Confidence Value 0.75 0.71 0.64

95% Confidence Interval 1.04-2.55 0.85-2.26 1.08-2.36

Number of Samples 9 6 6





Mobil 1 Pennzoil Platinum

Minimum Value 1.15 0.44

Maximum Value 4.39 3.67

Average Value 2.32 1.28

Standard Deviation 1.08 0.91

95% Confidence Value 0.64 0.56

95% Confidence Interval 1.68-2.96 0.71-1.84

Number of Samples 11 10


There is significant overlap in the 95% Confidence Intervals of all of these sample groups. However, there is the least overlap between Mobil 1 and Pennzoil Platinum. A student's T-test shows that the Mobil 1 sample group shows higher iron wear levels than the Pennzoil Platinum sample group at an alpha of 0.03. This is a statistical significance at the 97% level. None of the other sample groups showed a significant difference as compared to Pennzoil Platinum. Additionally, none of the other sample groups showed a significant difference as compared to Mobil 1.

The 30 weight sample groups can be ranked in the following order, from lowest to highest iron wear levels:

Pennzoil Platinum < Castrol Edge < Castrol Syntec < Amsoil

Comparison of Amsoil and Mobil 1 20 weight vs 30 weight

Since both Amsoil and Mobil 1 had appropriately large sample groups in both 20 weight and 30 weight viscosities, the obvious question is which viscosity produces the least iron wear. This comparison is shown below:

Code:


Amsoil 20 Amsoil 30

Minimum Value 0.50 0.59

Maximum Value 1.80 4.37

Average Value 1.10 1.80

Standard Deviation 0.52 1.15

95% Confidence Value 0.42 0.75

95% Confidence Interval 0.68-1.52 1.04-2.55

Number of Samples 6 9





Mobil 1 20 Mobil 1 30

Minimum Value 1.02 1.15

Maximum Value 3.10 4.39

Average Value 1.81 2.32

Standard Deviation 0.70 1.08

95% Confidence Value 0.49 0.64

95% Confidence Interval 1.32-2.30 1.68-2.96

Number of Samples 8 11


Both oils show lower average wear levels in the 20 weight sample group than the 30 weight sample group. A student's T-test shows that neither of these 20 weight oils is statistically different than their 30 weight counterpart at an alpha = 0.05 level.


6 CYLINDER ENGINES

20 Weight vs. 30 Weight - Iron ppm per 1000 miles

Code:


20 Weight 30 Weight

Minimum Value 0.60 0.61

Maximum Value 6.57 19.02

Average Value 1.82 2.72

Standard Deviation 1.20 3.00

95% Confidence Value 0.76 0.66

95% Confidence Interval 1.06-2.59 2.06-3.38

Number of Samples 22 78


The 20 weight sample group had significantly lower iron levels as measured by a student's T-test at an alpha = 0.05 level. The mix of engines in the 30 weight sample group differed from the 20 weight group in that the 30 weight group had engines from GM and Chrysler that were not present in the 20 weight group. These engines tended to have higher iron levels than the Ford, Honda, Mitsubishi, Hyundai, Kia, Toyota, and GM 3.6L engines in the 20 weight group. Therefore, I stripped all Chrysler and non-3.6L GM engines out of the 6-cylinder 30 weight sample group and reexamined the 20 vs 30 weight question with these results:

Code:


20 Weight 30 Weight

Minimum Value 0.60 0.62

Maximum Value 6.57 11.83

Average Value 1.82 2.15

Standard Deviation 1.20 1.95

95% Confidence Value 0.50 0.56

95% Confidence Interval 1.32-2.32 1.58-2.71

Number of Samples 22 46


When the Chrysler and non-3.6L GM engines are removed from the 30 weight samples, there is no significant difference between the 20 weight and 30 weight oils. This adjusted 30 weight sample group will be used for the remaining 6 cylinder statistical comparisons. It should be noted that both the range of values and the average value dropped when Chrysler and non-3.6L GM 6 cylinder engines were removed from the sample group.

Differences Among 30 Weight Oils

Only three 30 weight oils had the requisite number of samples for statistical comparison; Mobil 1, Pennzoil Platinum, and Pennzoil Ultra:

Code:


Mobil 1 Penn Plat Penn Ultra

Minimum Value 0.62 0.89 0.79

Maximum Value 8.00 3.69 3.66

Average Value 3.01 1.44 1.59

Standard Deviation 2.80 0.93 1.18

95% Confidence Value 2.24 0.64 1.03

95% Confidence Interval 0.77-5.25 0.80-2.08 0.56-2.62

Number of Samples 6 8 5


Given the major overlap in the 95% Confidence Intervals of the three sample groups, it comes as no surprise that student's T-tests comparing the three sample groups shows no significant differences.

Comparison of Pennzoil Platinum 20 weight vs 30 weight

Pennzoil Platinum is the only oil with enough samples in both viscosities to allow a statistical comparison:

Code:


Penn Plat 20 Penn Plat 30

Minimum Value 1.12 0.89

Maximum Value 2.20 3.69

Average Value 1.53 1.44

Standard Deviation 0.39 0.93

95% Confidence Value 0.31 0.64

95% Confidence Interval 1.22-1.84 0.80-2.08

Number of Samples 6 8


Because the 95% Confidence Interval for the 20 weight Pennzoil Platinum samples is totally contained within the 95% Confidence Interval for the 30 weight Pennzoil Platinum samples, there is no significant difference between the iron wear levels of these two oils.

8 CYLINDER ENGINES

20 Weight vs. 30 Weight - Iron ppm per 1000 miles

Code:


20 Weight 30 Weight

Minimum Value 0.63 0.65

Maximum Value 7.00 7.60

Average Value 2.45 2.73

Standard Deviation 1.66 1.98

95% Confidence Value 0.79 0.72

95% Confidence Interval 1.66-3.24 2.01-3.45

Number of Samples 17 29


The 8 cylinder 20 weight sample group consisted of engines from Ford and Chrysler. The 8 cylinder 30 weight sample group consisted of engines from Ford, Chrysler, Toyota, GM, and Nissan.

Both the overlap of the 95% Confidence Intervals and a student's T-test show that there is no significant difference between the 20 weight samples and the 30 weight samples.

Comparison of Amsoil and Mobil 1 30 weight

The only oils which had enough samples to compare were the Amsoil 30 weight and the Mobil 1 EP 30 weight:

Code:


Amsoil 30 Mobil 1 EP 30

Minimum Value 0.65 0.76

Maximum Value 5.00 6.73

Average Value 2.35 3.31

Standard Deviation 1.52 2.94

95% Confidence Value 1.21 2.57

95% Confidence Interval 1.14-3.56 0.74-5.88

Number of Samples 6 5


Both the overlap of the 95% Confidence Intervals and a student's T-test show that there is no significant difference between the Amsoil 30 weight samples and the Mobil 1 EP 30 weight samples.

Comparison of 20 weight vs 30 weight oils - Ford and Chrysler V8 Engines

Because the 30 weight group had a different mix of engines than the 20 weight group, the Toyota, GM, and Nissan engines were removed from the 30 weight group and the comparison was repeated:

Code:


20 Weight 30 Weight

Minimum Value 0.63 2.00

Maximum Value 7.00 7.60

Average Value 2.45 4.25

Standard Deviation 1.66 2.09

95% Confidence Value 0.79 1.37

95% Confidence Interval 1.66-3.24 2.88-5.61

Number of Samples 17 9


When Ford and Chrysler V8 engines are compared using 20 weight vs 30 weight oils, the 20 weight sample group showed lower iron levels at the alpha = 0.04 level. This is a 96% significance level. It should be noted that the average iron wear value increased when the Japanese and German V8 engines were removed from the sample group.


GENERAL OBSERVATIONS

- 20 weight oils consistently showed lower average iron levels than 30 weight oils. The only time this difference was statistically significant was with Ford and Chrysler V8 engines.

- Mobil 1 oils tend to show higher iron wear levels than some of the other oils in the database. In most cases, the difference is not statistically significant, although there was one case where Pennzoil Platinum showed significantly better iron wear levels than Mobil 1.

- There is adequate data to support the high regard BITOG members have for Pennzoil Platinum oil. In those cases where a direct comparison could be made, Pennzoil Platinum showed reduced iron levels compared to other oils, although these differences were not always statistically significant.

- American designed and manufactured 6 cylinder and 8 cylinder engines tended to show higher iron levels than Japanese and German engines.

 

[volk06]

Wow thats a lot of information and an interesting read! Thanks for all the hard work!

 

[dparm]

tl;dr: Mobil 1 causes higher iron levels



Seriously though, this is excellent. Suddenly bringing back fond memories of AP Statistics in high school.

And for clarification, was every motor run on the mfg recommended viscosity? What about motors that were backspec'd?

 

[2oldtommy]

Thank you OldCowboy! You deserve an award for this major work of pulling this data together. WELL DONE ! --Oldtommy

 

[friendly_jacek]

Very impressive work indeed.

A couple of constructive criticisms.
1. You noticed that iron levels tend to vary between different engine makers. Furthermore, makers decide what oil to recommend (20 vs 30). Thus the study results have that recommendation bias build in. For instance, if honda recommends only 20 oils and honda engines tend to show low iron, then the study will show lower iron results in the 20 group.
2. There is a phenomenon of increased engine wear right after oil change. It's reflected by an initial spike in Fe/1000 miles. You did the right thing rejecting
Finally, lets examine what the results mean. If we assume that Fe/1000 miles is a good measure of wear (which unfortunately it is not), the results don't convince me as argument that 20 is better than 30. It just shows engines that use 20 are not worse or better than engines that use 30. I have 0w20 oil in my prius and believe it's a low wearing engine with a very sound 0W20 recommendation. It would be unfair to compare this engine to one of the Detroit offerings that requires/required 30 oil.

I do agree with the Mobil 1 issue. Since I joined BITOG, I've heard numerous complains of Mobil always having higher Fe numbers. This doesn't surprise me and some Mobil competitors advertised that their products produced less wear than Mobil 1 in API tests. That wouldn't stop me from using Mobil 1 though.

And finely, engine wear is not a big deal these days. From all the cars I have/had, only one experienced problems related to engine oil and that had nothing to do with wear but rather oil coking from high temp leading to oil consumption. This is why I did the 350F oil cooking study that you didn't like much.

 

[OldCowboy]

Hi friendly,

Because this is a retrospective study, there are probably a lot of inadvertent biases built in.

A couple of points:

- The data I recorded had both the unit mileage and the OCI mileage. I'd be glad to send it to you.

- Everything I've read indicates that iron is the only wear metal which correlates directly with mileage. Please provide a reference which shows an initial spike in iron values after an oil change.

- Please provide a reference which shows that Fe/1000 miles is not a good measure of wear.

- I did not say that 20 weight oils were better than 30 weight oils. I said that the 20 weight oils showed lower iron levels in Ford and Chrysler V8 engines at a 96% significance level.

- My only complaint about your oil cooking study was that it was poorly controlled. I'd love to see you repeat it with consistent oil volumes, temperatures, air flow, etc.

- I agree with you that engine wear is not a big deal today. My main reason for conducting this study was to have something quantitative to point at when someone recommended a 30 weight oil for an engine which the manufacturer recommended a 20 weight.

 

[friendly_jacek]

Originally Posted By: OldCowboy

- Everything I've read indicates that iron is the only wear metal which correlates directly with mileage. Please provide a reference which shows an initial spike in iron values after an oil change.

- Please provide a reference which shows that Fe/1000 miles is not a good measure of wear.


Please see http://neptune.spacebears.com/cars/stories/mobil1.html
In the first 1,000 miles Fe/1000miles=10
In the all 18,000 miles Fe/1000miles=2.6

http://neptune.spacebears.com/cars/stories/amsoil.html shows similar effect.

The choice is yours: either there is increased wear with fresh oil or Fe/miles is not as useful as you thought. My take is both of the above.

Originally Posted By: OldCowboy

- I did not say that 20 weight oils were better than 30 weight oils. I said that the 20 weight oils showed lower iron levels in Ford and Chrysler V8 engines at a 96% significance level.


Compared to what, other engines that require 30 oils? That doesn't tell much as you compared apples to oranges.

Originally Posted By: OldCowboy

My main reason for conducting this study was to have something quantitative to point at when someone recommended a 30 weight oil for an engine which the manufacturer recommended a 20 weight.


Unfortunately, you did not study Fe levels in engines specifying 20 oils that were run on 30 oils.

Don't get me wrong, you did a great job and put a lot of time in it. But you have to be careful what conclusions you can draw from this if you want to stay with science.

But I know what you are saying. Your argument is when manufacturer specifies 20 oils, it's safe to do so. We all knew it already as most 5W30 oils quickly shear down to 5W20 and engines tolerate this with no issues.

Now, there are sometimes reasons to run heavier oils in cars specifying xW20. Toyota and Subaru clearly state that in their manuals and many other car makers say that to their Australian owners.

 

[OldCowboy]

Hi Friendly,

Thanks for your comment, but I have to respectfully disagree with some of your contentions.

As concerns your first link, http://neptune.spacebears.com/cars/stories/mobil1.html, one thing you may not have noticed was that this study was done on a relatively new car. The study started when the vehicle had 9,939 miles, so it was still breaking in and iron wear levels are expected to be high. It ended when the vehicle had 27,960 miles. During the first 12,025 miles, they added 5 quarts of oil. At 12,025 miles they changed the filter, and continued to run the oil for another 6,000 miles. Did they have to add the oil because there was a seep/leak somewhere? If so, then some of the wear metals were leaking out with the oil. I'm also disturbed that the iron level took a huge jump between 16,000 and 17,000 miles on the oil with no apparent explanation.

I don't think this study says anything about increased iron with fresh oil.

The second study (http://neptune.spacebears.com/cars/stories/amsoil.html) again doesn't indicate to me that there is more iron wear with a new oil. Like the first study, the "0" mile reading is a virgin oil out of the bottle. We all know that an oil change never gets all of the old oil out and the value after 1000 miles is new wear metal plus wear metal left due to incomplete change of the oil. Had they taken a sample after 10 miles, we might well have seen an iron level of 6-7 ppm just from an incomplete oil change.

Again, this study doesn't say anything about an initial spike in iron values after an oil change.

As concerns the lower iron levels with 20 weight oils in Ford and Chrysler V8 engines, that statistically significant finding compared Ford and Chrysler V8 engines running 30 weight oils with Ford and Chrysler V8 engines running 20 weight oils. All as reported by BITOG members. So I did, in fact, compare apples to apples.

Concerning your comment about "...engines specifying 20 oils that were run on 30 oils." Many of the engines in the 30 weight groups were, in fact, engines for which the manufacturer recommended 20 weight oils. Likewise, many of the engines in the 20 weight group were engines for which the manufacturer recommended 30 weight oils.

May I send you the raw data?

 

[friendly_jacek]

Look, I gave the links to spacebears study to open your eyes for things you were not aware about. Instead learning new things you keep arguing how you know things better when you don't. To blame 10 ppm Fe on incomplete oil change is nuts. We don't know what the initial Fe was, but assuming 20-30 ppm like at the end of Amsoil study that would mean 1/3-1/2 residual old oil left during the oil change process.

Are you saying that many scientists were incorrect when they published many SAE papers on this topic? This is just one of many, feel free to search for more:
http://papers.sae.org/2007-01-4133

 

[OldCowboy]

Hi Friendly,

Please do not attempt to lecture me on what science is. I have a PhD in Immunology/Virology and 15 years experience in biomedical experimental design and analysis. I also have 20 years experience as an Operations Research Analyst. The bottom line is that I am only interested in hard, reproducible facts. Opinions, hearsay, poorly designed experiments, and marketing fluff mean nothing to me.

Let's examine the spacebears study (http://neptune.spacebears.com/cars/stories/mobil1.html). The main problem with this study is that it involved a single test subject. This means the study has zero degrees of freedom and can only tell us about the behavior of a single entity. It cannot be extrapolated to describe the behavior of a population. In other words, it tells us something about that single LS1 engine. It tells us nothing about LS1 engines as a population and nothing about iron wear levels in other engine families.

I am not a triboligist and am new at the study of engine oils and wear analysis. I don't know how much motor oil is retained when an LS1 engine is drained. Do you? Also, I do not know the margin of error for iron levels present with the analytical laboratory this study used. Do you? I do find it interesting that the lab reported 2 ppm iron in virgin Mobil 1 oil. This means that the maximum iron wear level after 1000 miles is 8ppm (10 -2). But we know nothing about the iron wear level on the previous oil change. If we presume the vehicle was running on a factory fill for the previous 9,939 miles and knowing it is a new vehicle, it is not unreasonable to expect 40 - 60 ppm iron due to new vehicle break-in. Now if 10% of the previous fill was left in the engine that means 4-6 ppm of iron would be carried over into the next oil change. If you add the 2 ppm of iron the lab reported for the virgin oil with 4 ppm of iron carry -over, 6 ppm of the 10 ppm noted at 1000 miles would be accounted for. I'm not saying that's what happened. I'm saying we don't know. And if we don't know, we don't know.

The SAE paper quoted by both you and the spacebears study is entitled, "The Effect of Oil Drain Interval on Valvetrain Friction and Wear." Please note it says nothing about the effect of oil drain interval on iron wear levels. Have you paid for, downloaded, and critically examined this paper? I have and the results are not what you might expect from reading the abstract. The study used new engines and changed the oil at 3000, 5000, 7500, 10,000 and 15,000 miles. Here are their iron results in ppm from engine oil analysis:
Code:


3K 5K 7.5K 10K 15K

Oil #1 39 18 17 20 33

Oil #2 57 155 120 34 32

Oil #3 86 26 18 21 28


In terms of ppm/1k miles, this is what their study looks like:

Code:


3K 5K 7.5K 10K 15K

Oil #1 13 9 6.8 8 6.6

Oil #2 19 77.5 48 13.6 6.4

Oil #3 28.7 13 7.2 8.4 5.6


So what do these results tell us? Once again, each oil represents a single entity and thus can tell us nothing about anything but that entity. The only conclusion that can be drawn is that iron levels per 1000 miles are extremely variable during new engine break-in. The iron levels did not start becoming regular until the 10,000 to 15,000 mile run, after significant break-in had occurred. But, this study in no way says that the iron levels are higher in the first few hundred miles than in the remainder of the OCI.

 

[SteveSRT8]

Nice work. Great presentation.

Probably wouldn't characterize it as revelatory, but it was extremely interesting reading.

Really want to thank you guys for the efforts here.

 

[Pablo]

Good work. Interestingly some of the generalizations do fit the mindset of some folks here.

I do have a major concern. Your title "A Comparison of Iron Wear Levels Using BITOG Reported UOAs" is very incorrect, and you did mention what I'm getting at - you absolutely should not have used the word "wear" - just call it "A Comparison of Iron Levels Using BITOG Reported UOAs"

Thanks!!

 

[friendly_jacek]

Originally Posted By: OldCowboy

Please do not attempt to lecture me on what science is. I have a PhD in Immunology/Virology and 15 years experience in biomedical experimental design and analysis. I also have 20 years experience as an Operations Research Analyst.


Too bad this degenerated into bragging about credentials. I also happen to a be a scientist and college professor in biomedical sciences. And this means squat as far as tribology is concerned.

There are many things you are just learning on this forum but your arrogance is amazing.

Originally Posted By: OldCowboy

So what do these results tell us? Once again, each oil represents a single entity and thus can tell us nothing about anything but that entity. The only conclusion that can be drawn is that iron levels per 1000 miles are extremely variable during new engine break-in. The iron levels did not start becoming regular until the 10,000 to 15,000 mile run, after significant break-in had occurred. But, this study in no way says that the iron levels are higher in the first few hundred miles than in the remainder of the OCI.


They did 3 oil experiments with multiple measurements and if you plug the Fe/mile numbers from 3k and compare to Fe/mile at 15k you will see a statistically significant difference. This is what I tried to tell you several posts above that you cannot compare Fe/mile without controlling for length of OCI for crying out loud!

Why is it so difficult to understand?

Is this the "show me state" thing?

 

[OldCowboy]

Hi Friendly,

This dialog began when you asserted, "There is a phenomenon of increased engine wear right after oil change. It's reflected by an initial spike in Fe/1000 miles." I asked for references and you gave me three.

Let's return to the SAE study. You seem to think that the oil was left in the vehicles for the entire 15,000 miles. As I stated in my previous posting, "The study used new engines and changed the oil at 3000, 5000, 7500, 10,000 and 15,000 miles." That means that the ppm/1000 miles you see after 3000 miles (13-28 ppm) is larger than the iron level for the last oil change in the series which is a 5000 mile run from 10,000 to 15,000 miles. Since they started with new engines, it is to be expected that the first oil change would have higher iron levels than the last oil change. This is what happens during break-in and is why I excluded UOAs with engine mileage of less than 20,000 miles.

The first spacebears study (http://neptune.spacebears.com/cars/stories/mobil1.html) began when the engine (LS1) had 9,939 miles on it and proceeded for 18,000 mile. As would be expected, the iron level per 1000 miles dropped during the course of the study. Again the engine was still going through break-in during the early part of the study.

The second spacebears study (http://neptune.spacebears.com/cars/stories/amsoil.html) intrigues me and I'd like to see it repeated. They preceded the test with a 3000 mile OCI before beginning the test and the engine had almost 31,000 miles on it at the beginning, so the 1000 mile iron finding is probably pretty accurate. I'd especially like to see a repeat of the part between 2000 and 6000 miles where the iron level remains virtually constant. I can see how you would extrapolate this as proof that there is greater wear immediately after an oil change. And that may, in fact, be the case. However, this is a study of a single vehicle, and as a scientist, I'm sure you would agree that proof of this hypothesis requires more than a single vehicle. I'd love to repeat this study myself, but my current vehicle only has 7000 miles on it and likely not prove anything due to continuing break-in.

I owe you an apology. I was arrogant.

 

[tig1]

Originally Posted By: Pablo
Good work. Interestingly some of the generalizations do fit the mindset of some folks here.

I do have a major concern. Your title "A Comparison of Iron Wear Levels Using BITOG Reported UOAs" is very incorrect, and you did mention what I'm getting at - you absolutely should not have used the word "wear" - just call it "A Comparison of Iron Levels Using BITOG Reported UOAs"

Thanks!!


+1. The metal numbers of a UOA doesn't have anything to do with actual engine wear. Often many GM engines show much higher numbers than other makes, but as we know GM engines will last just as long. Real wear, as Doug has pointed out, can only be determined by tear down, as his experience has shown us.

 

[OVERKILL]

Originally Posted By: tig1
Originally Posted By: Pablo
Good work. Interestingly some of the generalizations do fit the mindset of some folks here.

I do have a major concern. Your title "A Comparison of Iron Wear Levels Using BITOG Reported UOAs" is very incorrect, and you did mention what I'm getting at - you absolutely should not have used the word "wear" - just call it "A Comparison of Iron Levels Using BITOG Reported UOAs"

Thanks!!


+1. The metal numbers of a UOA doesn't have anything to do with actual engine wear. Often many GM engines show much higher numbers than other makes, but as we know GM engines will last just as long. Real wear, as Doug has pointed out, can only be determined by tear down, as his experience has shown us.


Exactly.

100ppm condemnation points in 30L sumps over 1.2 million Km's with 90,000Km OCI's and everything still spec'd "as new" when checked upon tear-down.

 

[ueberooo]

this is deeply flawed. cars rated for 20w oils tend to be newer with new anti wear technology such as molybdenum lined cylinders and rings.

you would have to somehow correct for this, though i'm not sure if it s possible

 

[Jeff_in_VABch]

A couple of things:

Do 20W engines generally have larger sumps so the contaminant load is diluted? I only know a lot of Dodge sumps are 7 quart capacity.

Since american engines are almost always larger displacement with more iron swept area, the cylinders, would this mean all american engines should have more iron in the oil?

So many variables here, it's a tough call on what numbers mean.

 

[Gene K]

Originally Posted By: Jeff_in_VABch
A couple of things:

Do 20W engines generally have larger sumps so the contaminant load is diluted? I only know a lot of Dodge sumps are 7 quart capacity.

Since american engines are almost always larger displacement with more iron swept area, the cylinders, would this mean all american engines should have more iron in the oil?

So many variables here, it's a tough call on what numbers mean.


Most 4.6L Fords and Dodge 4.7L are 6 Quart. They did not change the size when they changed the recommendation to 5W20.

I believe all Hemis both 5W20 (MDS Models) and 5W30 (2500 and pre-MDS) are 7 Quart.

 

[fsskier]

Tig1 said it well....Fe numbers may be interesting, but adding the word "wear levels" is totally incorrect. As a Tribology report from Redline (Or was it Royal Purple, please correct me) made note:

Some of the best antiwear additives form chemical alloys with the iron, in the range best detected during oil analysis. With these additives, the higher Fe levels in the used oil are associated with very low wear levels in the engine.

My best memory of their additional comment..."Of course you could make an oil without them, get lower Fe numbers for advertising.... while your customers engines are wearing rapidly....."

That is likely the reason that most Motor oil Sequence approval tests measure the actual wear, and not just the purity of the used oil going to the recycler!!