Mobil AP, strange HTHS?

[fdcg27]

Almost every oil has a PDS that gives this figure at 40 and 100C.
I'm unaware of any source of KV at 150C, so you can only calculate it by an extrapolation along a curve of the figures at lower temperatures.
You cannot find a KV number for 150C unless you calculate it which necessarily involves making assumptions.

 

[Gokhan]

Originally Posted By: CharlieBauer
Has anybody collected the actual KV150 of oils and compared them to the theoretical KV150 that this index is calculated on?

Would not seem to be very scientific to suggest an index has some sort of validity if a theoretical number is the most important part of it's calculation and the validity of the calculation of that theoretical number has not been explored.

The biggest source of error in these calculations is not the theoretical models for the temperature dependence of the kinematic viscosity (KV) but the accuracy of the numbers reported in the datasheets. ExxonMobil has no legal responsibility to have any accuracy in their reported numbers whatsoever.

For example, TGMO 0W-20 SN is reported to have KV40 = 36.1 cSt and KV100 = 8.5 cSt, corresponding to VI = 225, but the WearCheck results from my VOA sample are KV40 = 36.16 cSt and KV100 = 8.79 cSt, corresponding to VI = 236. Of course, WearCheck results also have a nonzero accuracy.

I used the calculator by Widman International SRL.

http://www.widman.biz/English/Calculators/Operational.html

They also have a calculator for the viscosity index (VI):

http://www.widman.biz/English/Calculators/VI.html

I think the double-exponential model used for the temperature dependence of the kinematic viscosity (KV) is really good for the small temperature range here. There is a Wikipedia article on this:

https://en.wikipedia.org/wiki/Temperatur...matic_viscosity

Last but not least, here is the key scientific review article by Christopher J. Seeton on the models for the temperature dependence of the kinematic viscosity (KV) that goes over the formulas in the Wikipedia and beyond:

https://link.springer.com/article/10.1007/s11249-006-9071-2/fulltext.html

 

[1JZ_E46]

I am aware of the difference between significant figures and decimal places. But there are still issues with the A_Harmon calc: (1) we are using an estimate for viscosity and density at 150C (there needs to be some sort of "+-" error range applied to those calcs, which would flow through to the final value) and (2) many manufacturers either don't give HTHS or they give the minimum number per the ACEA spec (Castrol, Liqui-Moly). These change the confidence in the final value significantly. Given the amount of estimation and the fact that HTHS provided is only to one decimal place, I'd be hesitant to rely much on the index.

Edit: I would also assume some sort of range of values provided on the product data sheets due to fluctuations between batches, quality control. Something like +- 3-5%

 

[CharlieBauer]

Originally Posted By: Gokhan
Originally Posted By: CharlieBauer
Has anybody collected the actual KV150 of oils and compared them to the theoretical KV150 that this index is calculated on?

Would not seem to be very scientific to suggest an index has some sort of validity if a theoretical number is the most important part of it's calculation and the validity of the calculation of that theoretical number has not been explored.

The biggest source of error in these calculations is not the theoretical models for the temperature dependence of the kinematic viscosity (KV) but the accuracy of the numbers reported in the datasheets.


I've found one example of the theoretical calculation being compared to the actual numbers. It was provided by user bobbydavro who works for Castrol and is I believe an oil formulator.

The calculation estimated the KV150 was 3.88. But per bobbydavro, the actual KV150 was 4.1.

He also said the oil is shear stable. The index said it is 0.835.

 

[Gokhan]

Originally Posted By: CharlieBauer
I've found one example of the theoretical calculation being compared to the actual numbers. It was provided by user bobbydavro who works for Castrol and is I believe an oil formulator.

The calculation estimated the KV150 was 3.88. But per bobbydavro, the actual KV150 was 4.1.

He also said the oil is shear stable. The index said it is 0.835.

KV40 and KV100 both have at least a few percent measurement error. When you combine errors in KV40 and KV100, they usually combine as the root mean square of the percentage errors, which makes it worse. KV150 also has at least a few percent measurement error. Therefore, the theoretical value is within the measurement errors of the experimental value. It doesn't really mean that the theoretical model is off by -5%.

 

[buster]

Is there a way to tell whether AP 0w20 has more or less VII's than AP 5w20? I know for EP 0w20, that oil contains approximately 70% PAO vs the 5w20 EP which is somewhere around 35% PAO. (based on MSDS)

I've been using the 0w20 bc I feel the majority PAO base oil in the 0w20 is less prone to shear and needs less VII's than the 5w20, but I could be mistaken.

If the 5w20 does in fact use less VII's than the 0w20 within the AP line, I may just go with that.

 

[Gokhan]

Originally Posted By: buster
Is there a way to tell whether AP 0w20 has more or less VII's than AP 5w20? I know for EP 0w20, that oil contains approximately 70% PAO vs the 5w20 EP which is somewhere around 35% PAO. (based on MSDS)

I've been using the 0w20 bc I feel the majority PAO base oil in the 0w20 is less prone to shear and needs less VII's than the 5w20, but I could be mistaken.

If the 5w20 does in fact use less VII's than the 0w20 within the AP line, I may just go with that.

Yes, M1 0W-20 EP and M1 5W-20 EP have about the same amount of VII according to the A_Harman index.

However, according to the A_Harman index, M1 AP 5W-20 has almost no VII and M1 AP 0W-20 has some moderate amount of VII (though not excessive but more than in M1 EP 0W-20). The difference in the amount of the VII is also evident from the stark contrast between the VIs of the two oils -- 172 vs. 141 for M1 AP 0W-20 and M1 AP 5W-20, respectively.

Of course, MSDS lists a lot more PAO in M1 AP 0W-20 than in M1 AP 5W-20.

Code:
Oil Density KV40 KV100 KV150 HTHSV VI DV150 A_Harman index



M1 0W-20 0.841 44.80 8.70 3.91 2.70 177 2.91 0.928

M1 5W-20 0.852 49.80 8.90 3.89 2.75 160 2.93 0.938

M1 5W-30 0.855 61.70 11.00 4.75 3.10 172 3.59 0.862

M1 EP 0W-20 0.839 44.90 8.60 3.85 2.70 173 2.86 0.944

M1 EP 5W-20 0.850 49.60 8.90 3.90 2.75 161 2.93 0.937

M1 EP 5W-30 0.851 59.80 10.60 4.57 3.00 169 3.44 0.872

M1 HM 5W-20 0.856 50.00 8.60 3.72 2.70 150 2.82 0.958

M1 HM 5W-30 0.856 72.00 12.10 5.09 3.10 166 3.86 0.804

M1 AP 0W-20 0.840 45.69 8.70 3.89 2.60 172 2.89 0.899

M1 AP 5W-20 0.850 48.88 8.20 3.53 2.60 141 2.66 0.979

M1 AP 5W-30 0.851 63.33 11.70 5.09 3.00 183 3.83 0.783

 

[Shannow]

Good point, so now you can't decide what goes in your engine based on data sheets either.

(and they don't tell you what happens in the engine either, with the exception of the API/SAE engine tests...which AREN'T your engine, in it's current condition, so they aren't representative either are they)

At least you know the colour of the bottle, and which label you find attractive.

 

[Shannow]

Originally Posted By: CharlieBauer
Has anybody collected the actual KV150 of oils and compared them to the theoretical KV150 that this index is calculated on?

Would not seem to be very scientific to suggest an index has some sort of validity if a theoretical number is the most important part of it's calculation and the validity of the calculation of that theoretical number has not been explored.


The calculation is an ASTM standard...it's explained here, along with a whole lot of pther lubricant related stuff...

http://www.mne.psu.edu/chang/me462/handouts/viscosity_property.pdf

Interesting that they have the gall to suggest that you can state the temporary viscosity loss as a percentage

 

[Shannow]

Here's the ASTM standard behind the WIDMAN calculators...

ftp://185.72.26.245/Astm/1/Section 05/ASTM0504/PDF/D7152.pdf

....next

(before you start on the density side of things, the ASTM density calculation is internationally accepted as the temperature correction when delivering and receiving volumes of liquids)

 

[1JZ_E46]

Originally Posted By: Shannow
Originally Posted By: CharlieBauer
Has anybody collected the actual KV150 of oils and compared them to the theoretical KV150 that this index is calculated on?

Would not seem to be very scientific to suggest an index has some sort of validity if a theoretical number is the most important part of it's calculation and the validity of the calculation of that theoretical number has not been explored.


The calculation is an ASTM standard...it's explained here, along with a whole lot of pther lubricant related stuff...

http://www.mne.psu.edu/chang/me462/handouts/viscosity_property.pdf

Interesting that they have the gall to suggest that you can state the temporary viscosity loss as a percentage


I could see the KV150 estimation being close with a mono-grade oil (no VII), but less so with a multi-grade oil in the presence of VIIs. I would think VIIs would alter how an oil loses viscosity from a "standard" pure oil.

 

[Jetronic]

I agree that the VII impact between KV40 and KV150 might not be a line that intersects with KV100 for every oil, but I also think (from what I've seen) that it's very nearly a line for 99% of the oils. The calculation is not off by 10% unless the KV40 or KV100 numbers are quite off

 

[Shannow]

Originally Posted By: Shannow
And yes, if you are comparing the impact on VIIs on the finished product, then looking at the Harman Index for CITGO SAE30 (1), Ravenol 0W16 (0.99) and TGMO (0.85) tells you an awful lot about the temporary shear thinning of an oil...and by inference the possibility of permanent.

Comparing M1 0W30 versus 5W30 not so much.

 

[CharlieBauer]

Originally Posted By: Shannow
And yes, if you are comparing the impact on VIIs on the finished product, then looking at the Harman Index for CITGO SAE30 (1), Ravenol 0W16 (0.99) and TGMO (0.85) tells you an awful lot about the temporary shear thinning of an oil...and by inference the possibility of permanent.

Comparing M1 0W30 versus 5W30 not so much.


While you confirm the idea that the calculation tells you if something has VII or not, you seem to be both claiming a comparison can be made and that it can't.

Specifically, you say that 0.85 "tells you an awful lot about the temporary shear thinning of an oil".

Various M1 5w30's have similar numbers to TGMO's 0.85. What can be inferred from those numbers alone about M1 5w30?

 

[Shannow]

ratio of hths to newtonian viscosity...how much variance (i.e. VII impact) the oil has from a Newtonian bases case.

Surely you can admit that even with your concerns there's a difference between 0.84, 0.92, and 1.00.

 

[CharlieBauer]

Originally Posted By: Shannow
ratio of hths to newtonian viscosity...how much variance (i.e. VII impact) the oil has from a Newtonian bases case.

Surely you can admit that even with your concerns there's a difference between 0.84, 0.92, and 1.00.


Totally agreed and admitted!

An oil with a calculation of 1 or near 1, has little or no VII.

An oil with a calculation of 0.x has VII.

But I don't know what I can infer is the performance difference between an oil calculated at 0.84 and an oil calculated at 0.92.

Hence my statement that the calculation has validity, but calling it an index suggests comparability beyond the binary "Does the oil contain VIIs?" question that is answered by the simple calculation.