Understanding Viscosity and HTHS

Status
Not open for further replies.
I understand that, but the main problem I have is that the data therein is subject to the same concerns I had about grabbing numbers off of data sheets, much less safety sheets. We have singular data points, with no error analysis (that we can review). This isn't the same thing as grabbing numbers from someone else's raw data and working with their error analysis or doing your own. We're dealing with "typical" values, as the wording usually is, without either raw data or data dealt with by our own error analysis or one we can review. I do understand that such information isn't going to be easily forthcoming, but that doesn't change the scientific method.

Look at ordinary experimental procedure. How many times was KV40 or KV100 run by the oil company on a particular batch? I certainly don't know that. Yes, I have confidence that the value they give is typical and close enough for most purposes. Is it close enough when you're trying to establish a scientific relationship? That's where I have a problem.

Also, not to say that data sheets are useless, but there are enough errors here and there, some very glaring, that we have to watch. If there are glaring errors, how many are almost undetectable?

The predictability idea is thus, as you know. You should be able to be handed several bottles of oil, and conduct the tests - several times each - to fill in the data. Then, the calculation and predictions should hold true.

I'm the first guy to like theory and dislike the lab. That doesn't mean I can discount the value - the necessity - of experimentation. There's a reason why people once thought that heavier objects fell more quickly than lighter objects - the careful experimentation wasn't done.
 
For additional data to compare to the outputs of the tool/model, there are many motor oil patents that could be used. Some people will demand a lot of predictive accuracy and perhaps error bounds on the predictions. I think that if the tool provides values for things we are interested in and predictions are usually within 10%, it is certainly useful. Even if predictions are usually within 15%, it is still useful but less so than if predictions are more accurate. When there is no other way of estimating something, I'd rather use a tool that estimates it with 50% error (or whatever, within reason) than have no estimate at all. That's a personal preference. I know that some people would prefer to say "I have no idea" instead of saying it is X, plus or minus 50% (or whatever the error is).
 
Originally Posted by JAG
For additional data to compare to the outputs of the tool/model, there are many motor oil patents that could be used. Some people will demand a lot of predictive accuracy and perhaps error bounds on the predictions. I think that if the tool provides values for things we are interested in and predictions are usually within 10%, it is certainly useful. Even if predictions are usually within 15%, it is still useful but less so than if predictions are more accurate. When there is no other way of estimating something, I'd rather use a tool that estimates it with 50% error (or whatever, within reason) than have no estimate at all. That's a personal preference. I know that some people would prefer to say "I have no idea" instead of saying it is X, plus or minus 50% (or whatever the error is).

Thanks, JAG, I appreciate it.

You and I certainly have the same philosophy on this particular subject and you take the words out of my mouth!
 
Originally Posted by JAG
For additional data to compare to the outputs of the tool/model, there are many motor oil patents that could be used.

I'm not proposing that we require a lot of predictive precision for the calculator itself, as in a large number of significant digits. What I am asking for - and what science generally asks for - is that the relationship be firmly and accurately established before one tries to simplify or apply a model. The demonstration of the relationship must come first, then simplify and apply as needed.
 
I have, in my hands, an oil that I know for a fact has 0% VII. However, using Gokhan's equation of KV and HTHS, it shows it has 2.2% VII.
 
Originally Posted by RDY4WAR
I have, in my hands, an oil that I know for a fact has 0% VII. However, using Gokhan's equation of KV and HTHS, it shows it has 2.2% VII.

That's fine. Could you share the density, KV40, KV100, and HTHSV values? It's even better if you could share the data on the base oil including the KV40 and KV100 and the manufacturer and name if possible.

A 0.1 cP error in HTHSV results in about a 1% error in the VII content. For a monograde, since there is no VII, any error not resulting from the input data would be a result of the temperature extrapolation for the viscosity and/or density. If you share your data, we could look into it.

Once again, the calculator never claims high accuracy. It's an estimator that's developed to better understand oils. 0.1 cP error in HTHSV that you and I are talking about here is not that high. If I have a calculator that can estimate the HTHSV within 0.1 cP, I am more than happy with it.
 
Originally Posted by d00df00d
Maybe I'm saying this more for others than for you, as you have made it abundantly clear by now that you don't care. But the crux of Shannow's arguments is that you have vastly overstated the merits of this and other hypotheses you've pushed, and he is 100% correct on that.

I'd even say those overstatements border on lies, because as I said earlier, I know you know better.


Well stated and fully agree!

Originally Posted by d00df00d
nor do I think it has been productive for Shannow and others to waste so much space in threads like this when you obviously won't listen,


Which is why I don't. That said I am glad that you and Shannow help provide a perspective for those members less technically and scientifically inclined.
 
Originally Posted by Gokhan
Shannow, why are you trying to derail/troll a good discussion as you often do? From looking at your latest posts, the only thing you do lately on this board seems to be talking about general topics and your life and family and then troll me



No, I try to bring to the understanding of others that what you are offering is not what you are making it out to be.

In the mean time, you have disparaged my qualifications, career and profession, the content of my technical posts, and similarly other posters who actually worked in the industry...ANY industry.

As to my recent posts, you have noticed that they are becoming fewer..definitely a trend that will continue...
 
Originally Posted by Shannow
No, I try to bring to the understanding of others that what you are offering is not what you are making it out to be.

No, you're bringing your opinions that you portray as absolute facts.

Originally Posted by Shannow
In the mean time, you have disparaged my qualifications, career and profession, the content of my technical posts, and similarly other posters who actually worked in the industry...ANY industry.

That was never my intention but on the contrary, I feel that that's how you treat me.
 
Originally Posted by Gokhan
Originally Posted by d00df00d
Among those who have commented, the correlation between degree of relevant expertise and how warmly they view what you're doing is clearly inverse, and quite stark. The only outlier is you.

Not true at all, some experts (MolaKule) found the idea interesting and others (JAG) useful...


What I said was this is an interesting, unverified Hypothesis that without Empirical testing and verification, it remains an unverified Hypothesis. I never said I found it useful and cautioned that it should not be used for recommendations.

Originally Posted by Gokhan
For example, relativity and quantum mechanics was created by understanding the limits at which Newtonian mechanics failed.


Are you saying your spreadsheet/calculator is on par with the Metrics of Special and General Relativity or the postulates of Quantum Mechanics?

Originally Posted by Gokhan
As a more specific example, if the tool fails grossly for a specific oil or yields an unusual results, that could help one figure out an unusual ingredient in that oil.


What unusual ingredients would there be in a formulation that your spreadsheet/calculator could discover or predict?
 
Last edited:
Originally Posted by MolaKule
What I said was this is an interesting, unverified Hypothesis that without Empirical testing and verification, it remains an unverified Hypothesis. I never said I found it useful and cautioned that it should not be used for recommendations.

That's fine if that's your philosophy. JAG and I share a different philosophy on this:

Originally Posted by JAG
For additional data to compare to the outputs of the tool/model, there are many motor oil patents that could be used. Some people will demand a lot of predictive accuracy and perhaps error bounds on the predictions. I think that if the tool provides values for things we are interested in and predictions are usually within 10%, it is certainly useful. Even if predictions are usually within 15%, it is still useful but less so than if predictions are more accurate. When there is no other way of estimating something, I'd rather use a tool that estimates it with 50% error (or whatever, within reason) than have no estimate at all. That's a personal preference. I know that some people would prefer to say "I have no idea" instead of saying it is X, plus or minus 50% (or whatever the error is).

It's much better to have some HTHSV/base-oil estimator than to be totally in the dark. I'm quite happy with what it can predict.

Originally Posted by MolaKule
Are you saying your spreadsheet/calculator is on par with the Metrics of Special and General Relativity or the postulates of Quantum Mechanics?

Of course, not. This is what I said to Garak:

Originally Posted by Gokhan
It's only an estimator and has the caveats that were mentioned.

It's certainly much better than being totally in the dark, and please spare the nonsense on the cold-fusion analogy and other things you said. It's a simple calculator based on simple math and physics and not claimed to be a scientific breakthrough to say the least.

Originally Posted by MolaKule
What unusual ingredients would there be in a formulation that your spreadsheet/calculator could discover or predict?

For example, it calculated the base-oil viscosity index for the M1 HM oils, especially the 5W-30 grade, to be unusually low, which possibly indicated that these oils have a much higher concentration of the 4.7 cSt Synessticâ„¢ 5 AN base stock with the VI = 74. AN base stocks, especially the 4.7 cSt base stock with excellent solvency, can be used as seal swellers and they increase thermal stability. This is all in check with the M1 HM not having Resource Conserving rating due to seal swelling, as well as Mobil 1 specifically saying it has more thermal stability than the vanilla M1, despite M1 HM having less PAO and GTL according to the MSDS. POE and AN base stocks are not disclosed in the MSDS because they are not considered hazardous. I find this prediction from the calculator to be remarkable. Could it be false? Perhaps, but again, it's better than to be totally in the dark.
 
Originally Posted by Gokhan
For example, it calculated the base-oil viscosity index for the M1 HM oils, especially the 5W-30 grade, to be unusually low, which possibly indicated that these oils have a much higher concentration of the 4.7 cSt Synessticâ„¢ 5 AN base stock with the VI = 74. AN base stocks, especially the 4.7 cSt base stock with excellent solvency, can be used as seal swellers and they increase thermal stability. This is all in check with the M1 HM not having Resource Conserving rating due to seal swelling, as well as Mobil 1 specifically saying it has more thermal stability than the vanilla M1, despite M1 HM having less PAO and GTL according to the MSDS. POE and AN base stocks are not disclosed in the MSDS because they are not considered hazardous. I find this prediction from the calculator to be remarkable. Could it be false? Perhaps, but again, it's better than to be totally in the dark.

That's all great speculation, but little more. The first statement is something that can be tested, albeit not all that readily. Untested, it's speculation. The second statement about GF-5 rating is more speculation. There are a few reasons why an oil could preclude GF-5 certification. One would be elevated HTHS. Another would be that they don't target that, as a high mileage oil. Another would be your suggestion. Without an answer from Mobil, and I suspect they wouldn't answer that, it's not even a testable speculation.
 
Originally Posted by JAG
Some people will demand a lot of predictive accuracy and perhaps error bounds on the predictions. I think that if the tool provides values for things we are interested in and predictions are usually within 10%, it is certainly useful. Even if predictions are usually within 15%, it is still useful but less so than if predictions are more accurate. When there is no other way of estimating something, I'd rather use a tool that estimates it with 50% error (or whatever, within reason) than have no estimate at all. That's a personal preference. I know that some people would prefer to say "I have no idea" instead of saying it is X, plus or minus 50% (or whatever the error is).

Completely agree.

And without experimental validation, the error bars are just as hypothetical as the prediction. That means the prediction in any given case is X, plus or minus [??]%.

That's not better or worse than having no idea. It IS having no idea.
 
Yes, verify the relationship and have a complete error analysis, particularly to ensure that the magnitude of the error bar isn't greater than the magnitude of any of the values in the first place, or that the relationship doesn't unfortunately fall within the error bars themselves.
 
Everytime I read one of these threads it makes me do this:
33.gif
 
Originally Posted by d00df00d
Originally Posted by JAG
Some people will demand a lot of predictive accuracy and perhaps error bounds on the predictions. I think that if the tool provides values for things we are interested in and predictions are usually within 10%, it is certainly useful. Even if predictions are usually within 15%, it is still useful but less so than if predictions are more accurate. When there is no other way of estimating something, I'd rather use a tool that estimates it with 50% error (or whatever, within reason) than have no estimate at all. That's a personal preference. I know that some people would prefer to say "I have no idea" instead of saying it is X, plus or minus 50% (or whatever the error is).

Completely agree.

And without experimental validation, the error bars are just as hypothetical as the prediction. That means the prediction in any given case is X, plus or minus [??]%.

That's not better or worse than having no idea. It IS having no idea.

How do you agree with JAG? He's saying the exact polar opposite of what you're saying.
 
Originally Posted by Gokhan
Originally Posted by d00df00d
Originally Posted by JAG
Some people will demand a lot of predictive accuracy and perhaps error bounds on the predictions. I think that if the tool provides values for things we are interested in and predictions are usually within 10%, it is certainly useful. Even if predictions are usually within 15%, it is still useful but less so than if predictions are more accurate. When there is no other way of estimating something, I'd rather use a tool that estimates it with 50% error (or whatever, within reason) than have no estimate at all. That's a personal preference. I know that some people would prefer to say "I have no idea" instead of saying it is X, plus or minus 50% (or whatever the error is).

Completely agree.

And without experimental validation, the error bars are just as hypothetical as the prediction. That means the prediction in any given case is X, plus or minus [??]%.

That's not better or worse than having no idea. It IS having no idea.

How do you agree with JAG? He's saying the exact polar opposite of what you're saying.

The part I quoted, taken literally, is a completely sensible broad-strokes discussion that does nothing whatsoever to validate your hypotheses or the way you push them.

The thrust of it is that, given a model that is known to correspond with reality in a predictable way, it's probably better than a blind guess even if the error bars are huge. Of course it is. That's almost a tautology.

JAG is wrong to suggest that that argument validates your approach. The reason he is wrong on that point is that your hypotheses are not known to correspond with reality and do not have known error bars. In other words, his argument does not apply because two of the fundamental criteria are not met. There's nothing wrong with the argument itself, though.
 
Originally Posted by d00df00d
JAG is wrong to suggest that that argument validates your approach.

It's fine if you don't agree with JAG or me, but when you say he is wrong or I am wrong, that's inappropriate.

In the light of the engineering and scientific discussion, you know JAG is an engineer and I am both an engineer and a research scientist. Perhaps give us a little slack here on our engineering and scientific intuition?
 
Just because you're an engineer doesn't mean you're always right. An engineer designed the Mustang II after all.

If anything, it's more unsettling that an engineer is willing to push what is essentially psuedoscience.
 
Status
Not open for further replies.
Back
Top