Another oil thread

customboss · December 11, 2023

3 minutes ago, Grumpy Bear said:

Mud drilling?

Oil drilling, yes drill lubes are a high pressure place to test EP formulations. Keep your mind open to how formulating starts. Applications can change as the chemistries are modified for a particular application. K and Boron instead of K and asphaltic formulation.

Edited December 11, 2023 by customboss

Grumpy Bear · December 11, 2023

2 minutes ago, customboss said:

Yes drill lubes are a high pressure place to test EP formulations. Keep your mind open to how formulating starts. Applications can change as the chemistries are modified for a particular application.

This is a potassium sulfonate not a potassium borate nano particle. You will have to help me make than bridge.

I don't know "MolaKule". Just found his post on potassium running down the rabbit hole which led me to the scholarly article. He means nothing to me. I give him no gravity. No more than a man on the street who points at a quarter I may like to pick up.

Anyway, on supplemental potassium borate AW adds to ZDDP in response to lower levels of the latter; a dead giveaway would be the lack of boron and potassium in levels correct and in proportion for the chemistry. I've seen nothing in, so far, in self-oils with low zinc/phosphorus levels that would lead me to endorse some idea that potassium borate was even present.

1 hour ago, customboss said:

Pentane insolubles the only accurate testing outside of their ICP.

So, Blackstone wear/additive elemental ICP analysis is okay as well as the insolubles but the data base of averages for wear metals is not?

See, one of the things I LOVE about stats is their ability to make irrelevant the extremes of a data set. Especially very large data sets. Unless of course those averages are made up of nothing but the data from one end of the spectrum. Are you telling me the cherry pick the data they use to make up a platforms 'wear averages'? I don't know these people, but I do know something of stats. IF the input is trustable and the data set is large the math says what the math says. From the reports I've seen they seem to be forthcoming in admission of small data set engines in stating so which would give one pause.

I don't know where this is all going. I truly don't. But I do know Chevron, and if they are making a commercial blend, it is not without merit and purpose. It does not mean I'm ready to get on that bus, especially in consideration of the fact I'm currently not having a reason to pursue it.

It would seem Chevrons interest lay in the aftertreatment systems performance more so than the engines reliability. They are solving a problem, and darn few solutions come without some compromise. Right? Let's see.

I will await with bated breath the UOA's on your 'proof of concept' experiments and am more than willing to learn from whatever the data shows. This is exciting stuff is it not?! Oh, and I thank you personally for your willingness to guinea pig this chemistry.

customboss · December 12, 2023

21 hours ago, Grumpy Bear said:

This is a potassium sulfonate not a potassium borate nano particle. You will have to help me make than bridge

Starting chemisty for EP reductions. I worked on developing very advanced oil field drilling bit lubes 20+ years ago where fire and explosion danger are real. I can't remember where my data is but when I saw the oronite chevron Delo 600 add pack it made sense.

customboss · December 12, 2023

21 hours ago, Grumpy Bear said:

So, Blackstone wear/additive elemental ICP analysis is okay as well as the insolubles but the data base of averages for wear metals is not?

Yes that what I have told you over and over again. The universal averages UNIVERSE are not all encompassing and will lead the reader down the primrose path.

Unit averages mean little if your universe is that unit alone. You have to have a starting place for metallurgy, tune, fuel, oil, environment, maintenance recent, repairs and modifications.

Grumpy Bear · December 12, 2023

1 hour ago, customboss said:

The universal averages UNIVERSE are not all encompassing

I hope we are talking about the same universal average.

I'm talking about the one 'per platform' i.e. all wear data collected on the Gen 1 SBC will be different than all data collected on a CUMMINS B5.9. That just a fact.

I'm not talking about the fluid's universal elemental/physicals averages.

So, I'll just stick to the nature of statistics. The entire branch of mathematics exists for the purpose of evaluating the universe from a 'sample' of that universe whose size is large enough to predict that universe with a knowable probability. It doesn't need to be all encompassing to be highly accurate. 68–95–99.7 rule - Wikipedia

As this rule applies to those motor platforms, I find no reasonable explanation to doubt them without doubting mathematics itself. I'm not ready to make that jump. A 99.7% probability is close enough for any work I do.

My point was that Blackstone is good about saying, "We just don't have enough data on that platform yet".

Now. if I missed my guess and you were indeed talking about the fluids then, agreed. Within this context. PER CHEMSITY. That is saying the Mobil 1 of 1970 is NOT the same chemistry as the Mobil 1 sitting on the shelf at Wally World today, and we never ever know when they change. Which is why I pretty much insist on running an unused sample from the batch used if I am comparing anything outside the SAE designated parameters of the fluid. Those things chemically defined or by law restricted. In which case the 'entire universe' is in that single production lot. Could be over hundreds of lots but I'd never know tweaks that are not part of the 'normal and natural' distribution contained within the set.

customboss · December 12, 2023

4 hours ago, Grumpy Bear said:

I'm talking about the one 'per platform' i.e. all wear data collected on the Gen 1 SBC will be different than all data collected on a CUMMINS B5.9. That just a fact.

Grumps .. their UNIVERSAL averages for wear values are meaningless because they don't discriminate tight enough across the metallurgy in new engines or transmissions, or gearboxes to have an accurate starting point. Jamming together disparate data under a brand name is not knowing good enough data in, to get good statistical data out.

You've got multi sets of data that are disapparate and unrelated so that a universal average is nearly meaningless.

They don't have data bases on reference oils or chemistry to even attack that. Agreed

customboss · December 12, 2023

Don't overthink it.

Grumpy Bear · December 13, 2023

23 hours ago, customboss said:

Grumps .. their UNIVERSAL averages for wear values are meaningless because they don't discriminate tight enough across the metallurgy in new engines or transmissions, or gearboxes to have an accurate starting point. Jamming together disparate data under a brand name is not knowing good enough data in, to get good statistical data out.

According to Blackstone: ["....also compare/contrast your results to those of “universal averages”. Those averages come from equipment just like yours; engine to engine, transmission to transmission, etc."]

We are just going to disagree on this. I cannot imagine a company that would intentionally undermine its own business by not doing what cost nothing to do and is literally part of the process to make the report. Data entry.

customboss · December 14, 2023

4 hours ago, Grumpy Bear said:

According to Blackstone: ["....also compare/contrast your results to those of “universal averages”. Those averages come from equipment just like yours; engine to engine, transmission to transmission, etc."]

We are just going to disagree on this. I cannot imagine a company that would intentionally undermine its own business by not doing what cost nothing to do and is literally part of the process to make the report. Data entry.

https://bobistheoilguy.com/forums/threads/universal-averages-on-used-oil-analysis-uoa.53258/

customboss · December 14, 2023

4 hours ago, Grumpy Bear said:

According to Blackstone: ["....also compare/contrast your results to those of “universal averages”. Those averages come from equipment just like yours; engine to engine, transmission to transmission, etc."]

We are just going to disagree on this. I cannot imagine a company that would intentionally undermine its own business by not doing what cost nothing to do and is literally part of the process to make the report. Data entry.

The aviation data base is consistent and accurate but the UNIVERSE of CIVIL light aviation units is easily databased. Similar metallurgies, similar fuels, similar lubes, similar issues. Variability and incorrect input of automotive data that Bklabs universe is not standardized as aviation.

"ONCE WE GOT IT ON THE COMPUTER IT SIMPLIFIED THE WHOLE THING. WE ALWAYS RAN TWO COMPUTERS BECAUSE WE COULD NEVER AFFORD TO BE DOWN, SO WE ALWAYS HAD OUR DATABASE RUNNING ON ONE, AND THE REPORTS RUNNING ON ANOTHER. WE UPDATED OUR DATABASE ABOUT ONCE EVERY TEN DAYS, AND IT JUST GREW AND GREW. MY WIFE AND I RAN THE BUSINESS FOR ABOUT THE FIRST 25 YEARS WHILE I WORKED FOR DANA, AND THEN THE LAST TEN YEARS OF THE COMPANY I RAN IT COMPLETELY. I RETIRED AND SOLD THE COMPANY TO BLACKSTONE LABS IN APRIL 2002. "

https://www.avweb.com/features/howard-fenton/

Grumpy Bear · December 14, 2023

Nothing quoted changes my mind because nothing quoted changes the math's ability to find center tendency nor shift the "empirical rule" far enough to raise an eyebrow. Stats exist to make sense of the fact PERFECT DATA SETS DO NOT. Stats are just a means to find USEFUL relevance in imperfection.

Stats are a lie detector. A branch of stats called "Process Capability". We used it all the time in manufacturing. Pull retains, a few random retains. Measure, record and process this incomplete and random data to find the equipment's ability to reach a process specification the bean counters were demanding to ascertain if that goal was even possible on that machine. And if possible, under what conditions. Does the machine, process and raw materials have to be perfect to make that spec or just horseshoe close? Is it impossible? If not, what changes will make success 90% probable?

It is truly amazing how few data points are required to FIND the center and RANGE to a 99.7% probability of the same answer one would get by measuring, recording and crunching the numbers of 100% of all samples. It is its purpose, and it is darn good at it.

One has to ask to what purpose is PERFECTON a requirement. How close does a hand grenade have to be, to be effective?

UNIVERSAL AVERAGES are not a few dozen statistically processed random samples, which would be VERY accurate I might add; they are what they say they are. The mathematical average of all data collected to the last data entry. More complete than a statistical processing result and yet by not enough to matter. Even if it is weeks behind. We are not talking about a few dozen samples run but HUNDREDS and sometimes THOUSANDS of data points. Only takes about 13 points to be over 99% sure.

I will repeat. We will not agree on this. I'm not tossing my education and my experience to the curb because there is someone is tossing some mud on the floor.

You don't like Blackstone, don't use them. Pretty simple stuff. Right? :dunno: And if this is for some other reason, then you are selling something. Maybe not in the traditional sense but sales pitch non the less.

customboss · December 14, 2023

7 hours ago, Grumpy Bear said:

Nothing quoted changes my mind because nothing quoted changes the math's ability to find center tendency nor shift the "empirical rule" far enough to raise an eyebrow. Stats exist to make sense of the fact PERFECT DATA SETS DO NOT. Stats are just a means to find USEFUL relevance in imperfection.

Stats are a lie detector. A branch of stats called "Process Capability". We used it all the time in manufacturing. Pull retains, a few random retains. Measure, record and process this incomplete and random data to find the equipment's ability to reach a process specification the bean counters were demanding to ascertain if that goal was even possible on that machine. And if possible, under what conditions. Does the machine, process and raw materials have to be perfect to make that spec or just horseshoe close? Is it impossible? If not, what changes will make success 90% probable?

It is truly amazing how few data points are required to FIND the center and RANGE to a 99.7% probability of the same answer one would get by measuring, recording and crunching the numbers of 100% of all samples. It is its purpose, and it is darn good at it.

One has to ask to what purpose is PERFECTON a requirement. How close does a hand grenade have to be, to be effective?

UNIVERSAL AVERAGES are not a few dozen statistically processed random samples, which would be VERY accurate I might add; they are what they say they are. The mathematical average of all data collected to the last data entry. More complete than a statistical processing result and yet by not enough to matter. Even if it is weeks behind. We are not talking about a few dozen samples run but HUNDREDS and sometimes THOUSANDS of data points. Only takes about 13 points to be over 99% sure.

I will repeat. We will not agree on this. I'm not tossing my education and my experience to the curb because there is someone is tossing some mud on the floor.

You don't like Blackstone, don't use them. Pretty simple stuff. Right?  And if this is for some other reason, then you are selling something. Maybe not in the traditional sense but sales pitch non the less.

Liking Bklabs has nothing to do with what I’m sharing. As a matter of fact I am friends and colleagues of Jim ( RIP ) and his kids who run the lab.

If you run incorrect inaccurate raw data into a statistical analysis and don’t have much accurate your statistics are flawed.
I’ll end with this gentle reminder that……

Grumpy Bear · December 14, 2023

Now why did you have to go and make it personal with such a juvenile move? You have trouble with other's boundaries it seems.

I don't process data with opinions. I process it with procedure. They are telling me the wear data is collected and segregated by motor 'family'. I cut a pasted that earlier. I.E. All Ford FE in one pile. All Ford 385 series in a pile. SBC LS in one pile. Gen 1 in another. BBC Mark IV in another, and so on. That is close enough for the process to do its job.

The measured results I have received from other labs on my equipment over decades dovetail perfectly with their averages. Pretty good indicator the system isn't corrupt beyond usefulness. What, did you think you were my first?

As long as the data they input is the data they collected. They could input only half the data and it wouldn't change a thing.

Are you saying they make up data and enter that? That they can't run a D5185? Really? Because that would be an unusable data set and a claim of incompetence. Liable if it can't be proven.

Your claim is either false or they are liars. Are you calling them out as liars? I'm not. My opinion isn't an opinion, it's ASSSURED faith in the math and my willingness to believe the answer from the source. You are A source but not THE source.

If I want to know if my grandmother makes good cookies, I try one. I don't ask the kid across the street.

We've focused in on Blackstone. There are literally dozens of other labs, and you have issues with all of them but the one you started.

Personally, I haven't found Mr. Perfect yet. I have found many a Mr. Useful. You may ramble on if you wish. I got other things to do.

Edited December 14, 2023 by Grumpy Bear

customboss · December 14, 2023

2 minutes ago, Grumpy Bear said:

Now why did you have to go and make it personal with such a juvenile move? You have trouble with other's boundaries it seems.

I don't process data with opinions. I process it with procedure. They are telling me the wear data is collected and segregated by motor 'family'. I cut a pasted that earlier. I.E. All Ford FE in one pile. All Ford 385 series in a pile. SBC LS in one pile. Gen 1 in another. BBC Mark IV in another, and so on. That is close enough for the process to do its job.

The measured results I have received from other labs on my equipment over decades dovetail perfectly with their averages. Pretty good indicator the system isn't corrupt beyond usefulness. What, did you think you were my first?

As long as the data they input is the data they collected. They could input only half the data and it wouldn't change a thing.

Are you saying they make up data and enter that? That they can't run a D5185? Really? Because that would be an unusable data set and a claim of incompetence. Liable if it can't be proven.

Your claim is either false or they are liars. Are you calling them out as liars? I'm not. My opinion isn't an opinion, it's ASSSURED faith in the math and my willingness to believe the answer from the source. You are A source but not THE source.

If I want to know if my grandmother makes good cookies, I try one. I don't ask the kid across the street.

We've focused in on Blackstone. There are literally dozens of other labs, and you have issues with all of them but the one you started.

Personally, I haven't found Mr. Perfect yet. I have found many a Mr. Useful. You may ramble on if you wish. I got other things to do.

Hey it was a joke. Ease up Brother. You post the most ridiculous memes and pics. Enjoy me returning the favor. We disagree. Move on.

Grumpy Bear · December 16, 2023

2 pieces of popcorn +2 pieces of popcorn =4 pieces of popcorn. The math is not dependent on the quality of the popcorn.

Occom's razor. To have a statistically different outcome Blackstone would have to run the numbers on the entire population of results. Calculate the standard deviation. Find all data between the second and third sigma or greater on but one side of the entire distribution. Delete all data not within that parameter and call that the universe averaging only that data. Then the results could be skewed enough to matter. Given that level of effort :idiot: Really? Doesn't past the stink test.