Atlas - 3.6 VEE-ARE 6 "fuel diluter"..turning 40Ws to 30Ws via Costco Top Tier 87 since 2018! Of course these Euro 40W oils are on the low side of 40 and a 30W Euro will be on the high side so would likely stay in-grade vs. the 40s...I should swtich next change out to some LM Top Tech 0W30 and try it!
Another "Taxi" Study: Relationship of Engine Bearing Wear and Oil Rheology 872128
- Thread starter AEHaas
- Start date
The "Fuel %" measurement in that UOA is "TR" (Trace), which means it was hardly detectable. What if it was 5%. Mabye @OVERKILL can do a quick oil+fuel viscosity mix calculation since he's shown how that works that in other threads. That viscosity loss must have been mainly from oil shearing.
It's BS so don't go off that. Low flash points and visosity in the 10s = fuel dilution. I always hear about shearing but in a NA engine like this? These oils shoudln't be shearing...I even see with HPL and that has the very high-end star VIIs.
Last edited:
OK, it's fuel dilution then, glean whatever suits your agenda. Ten people will give 10 different answers to the same question. Game theory 101.
Suits an agenda?
It’s better to be accurate on this considering the root cause is completely different as are any mitigating fixes.
Too fortune cookie for me to understand.
I'm a Scorpio if that makes any difference.
Same birthday as Hillary.
Last edited:
If i might crank this back around to the original study, I have a few questions then a an observation.
Apologies if its been mentioned but:
How many Buick V6's were run on the dyno.
How many 4.3 Caprices were run with each oil.
Were they all the same (gear ratios, transmissions tire diameter and so forth...)
Why are these 2 engines chosen.
What was the purpose of the study ( what did it set out to prove or disprove.)
To determine something like this by sampling you must first define the population (For example 4.3 engines in caprices?, 4.3 engines in general, engines in general), then what you want to sample for and then a confidence level that you want you samples to support.
So for example i want to know if 20W oil does or does not cause more wear than a 30W oil in a 4.3 caprice and i want to be 95% certain that my results accurately predict that. I don't have time to figure the numbers all out but i promise you it is not 10 or 20 cars, its more like 300 or 20W and 300 of 30W. If you want this to cross populations it's probably even more. at some point as long as to population is viable the sample stops rising, that number is probably over 400.
Also my problem with almost all these studies is that they are accelerated in some manner necessarily because if you did this with normal peoples cars it would take 10 years to get results, by which time the results might tbe meaningless. Taxi studies accelerate the testing because they are taxis but is taxi use representative of "normal use". The bus study (IIRC) did it by introducing large amounts of soot or particulate.
So what im saying is that all our UOAs and little test we do don't really mean much, for instance (and we all know this is where this is going) if Ali's Black Label does or does not blow up on 5W that's doesn't prove mine will or wont only that its might.
Now - I'm really sure i'm not about to make any friends here but a goodly portion of what i read here is "I think", "I believe" and "Should be" some of it very eloquently stated and some with some evidence to support it but still... A lot of times 'Should be" Is, and a lot of times what we "think" and "believe" is true, but sometimes not.
Apologies if its been mentioned but:
How many Buick V6's were run on the dyno.
How many 4.3 Caprices were run with each oil.
Were they all the same (gear ratios, transmissions tire diameter and so forth...)
Why are these 2 engines chosen.
What was the purpose of the study ( what did it set out to prove or disprove.)
To determine something like this by sampling you must first define the population (For example 4.3 engines in caprices?, 4.3 engines in general, engines in general), then what you want to sample for and then a confidence level that you want you samples to support.
So for example i want to know if 20W oil does or does not cause more wear than a 30W oil in a 4.3 caprice and i want to be 95% certain that my results accurately predict that. I don't have time to figure the numbers all out but i promise you it is not 10 or 20 cars, its more like 300 or 20W and 300 of 30W. If you want this to cross populations it's probably even more. at some point as long as to population is viable the sample stops rising, that number is probably over 400.
Also my problem with almost all these studies is that they are accelerated in some manner necessarily because if you did this with normal peoples cars it would take 10 years to get results, by which time the results might tbe meaningless. Taxi studies accelerate the testing because they are taxis but is taxi use representative of "normal use". The bus study (IIRC) did it by introducing large amounts of soot or particulate.
So what im saying is that all our UOAs and little test we do don't really mean much, for instance (and we all know this is where this is going) if Ali's Black Label does or does not blow up on 5W that's doesn't prove mine will or wont only that its might.
Now - I'm really sure i'm not about to make any friends here but a goodly portion of what i read here is "I think", "I believe" and "Should be" some of it very eloquently stated and some with some evidence to support it but still... A lot of times 'Should be" Is, and a lot of times what we "think" and "believe" is true, but sometimes not.
Agenda? Its well understood here that Blackstone (BS) infers fuel dilution from the flashpoint rather than a direct measure. If you see any numbers there in my experience with dozens of UOAs across multiple vehicles that you have more fuel than their value shows.
And beyond that there is already a standardized ASTM test for wear due to oil grade that gives statistically valid results.
I didn't know that.
Is that a laboratory test? Because we can have another argument entirely over laboratory vs real world... It probably at least is more statistically relevant that 5 or 10 taxi cabs...
API test sequence. Oil can not be API approved if they don't pass all the defined test requirements. It would be interesting to run some 35 year old oil formulations through the current tests to see how it does. Not sure if they get into journal bearing wear targets or test many different viscosities in the same engine that may have different pass/fail criteria. And not sure how the test pass/fail specs have change over the years ... I'm sure them must have along the way.
https://bobistheoilguy.com/forums/threads/testing-environment-for-api-sp-ilsac-gf-6a.330414/
Last edited:
You realize most millennials probably can drive a stick and are almost all 30s and 40s at this point, or do you just continue to categorize every generation as millennial.
It's strange that their comments section didn't say anything about the oil being out of viscosity range - or the flash point being low. They aren't out of range by much, but they usually they say something if that happens.
Was the same oil also tested at 30,460 miles?
And the oil at 6,195 miles had a lower flash point of 340*F and 1.8% fuel dilution, but it stayed withing viscosity range. Same oil again? Test results correlations seem kind of random.
Last edited:
My two were born in 79 & 82 and have driven everything from logging trucks down. I thought the boomer's kids were gen-x.
Boomers were raised by war vets and depression survivors.
edit... As I posted before, and it looks like you read it, we tried millennials or whatever the latest crop of 30s are on the railroad and they are as useless as kilts in a hurricane. One had to take 3 "mental health days" off because his parents put down the 17 year old family dog.
Last edited:
The thing is that for the most part there are very few “real world” tests that mean a darn thing. Most are useless when properly examined. The uncontrolled variables in the real world do an excellent job of invalidating test results.
It’s a cold reality that most do not wish to face.
The horror.
Experiment Reviewed: Billions of Miles in Millions of Cars Run Over Decades in Light Trucks and Cars Including Back Spec'ed Vehicles. Ali E. Haas, 2023.
Results concluded by the author: Statistically significant that 20 grade oils can be safely used in cars and light trucks.
Analysis: Historical data with no grouping nor compartmentalization, no criteria, no UOA nor tear downs, no controls, not double blind. No controlling shearing nor fuel dilution. No measurements of MOFT, HTHS. Very few vehicles using the current and best oils ever made. Side arm of vehicles using 16 grade oils. An aside, the author is a biochemist turned surgeon with no mechanical nor automotive knowledge and very little scientific abilities.
My conclusions:
Exhaustive daunting study over many years with industry wide useful results. An accolade is earned.
Results concluded by the author: Statistically significant that 20 grade oils can be safely used in cars and light trucks.
Analysis: Historical data with no grouping nor compartmentalization, no criteria, no UOA nor tear downs, no controls, not double blind. No controlling shearing nor fuel dilution. No measurements of MOFT, HTHS. Very few vehicles using the current and best oils ever made. Side arm of vehicles using 16 grade oils. An aside, the author is a biochemist turned surgeon with no mechanical nor automotive knowledge and very little scientific abilities.
My conclusions:
Exhaustive daunting study over many years with industry wide useful results. An accolade is earned.
Last edited:
Yes Honda stated that years ago when they determined that 20-grade oils gave acceptable wear.
I agree with that, however when lab test accelerate testing and do things that try to emulate real world conditions whether or not the results are relevant to joe public becomes debatable. No one is running water cooled valve covers and exhaust manifolds on street cars, no one buys their Rav 4 and runs it 24/7 on the chassis dyno and only shuts it down to change the oil or inspect it.
What a tightly controlled test can tell us is how the subject of the test performs under the test conditions and relative to other test subjects under test conditions, in this case thinner vs thicker under the ASTM test conditions.
What i'm saying is that you cant (well you can, but you should not) look at any one test in a vacuum to the exclusion of all other test, information, experience. Whether that test be a UOA, a Taxi Study, or some associations standardized test.
Similar threads
- Sticky
- Locked
