BOQI II: The effect of HTHSV

[CR94]

Originally Posted by OilUzer
As @JHZR2 said, evaporation is not linear in either direction. if noack is %8 do you lose %16 in 120 min? ...
True, and it's not necessarily an exponential function, as I gather Gokhan assumes, either. Color me skeptical.

 

[Gokhan]

Originally Posted by drtyler
Why write Noack in all capital letters?

Ha, my bad, it's a proper name and not an abbreviation! I'll correct the capitalization.

Originally Posted by Virtus_Probi
Again, the fact that CCS is tested at different temperatures for different W grades makes the comparison across grades very shaky for me.

Yes, that's tricky for different xW grades. I have ad-hoc constants to account for CCS measured at different temperatures for different xW grades (-35 C, -30C, -25C, etc.).

 

[Gokhan]

Originally Posted by Shannow
Now you've thrown in HTHS, which as you go up in the grades (20 through 40) is the result of more polymeric VI thickeners, so now has even LESS to do with BASE OIL QUALITY.

The viscosity-index improver (VII) temporarily shears at the 1,000,000 per second shear rate at 150 C to a great extent and you're mostly looking at the base oil and not the VII effects. True, there would still be a small amount of unsheared VII but (1) the residual-VII effect is small and (2) for a given viscosity grade, the residual VII will be similar for different oils, multiplying the BOQI II by the same number very close to 1, and the approximation of neglecting it is an excellent one. Therefore, BOQI II still measures the base oil, not the VII-thickening effects.

Originally Posted by Shannow
Oh, and you forgot to divide by colour...it's truly dimensionless, and while we know that color is impacted by additives, we've ignored that entirely to date, so why start now ?

Exactly as you said, the color is an additive effect. I don't include the TBN for the same reason. BOQI II looks at the base oil, not the additives or VII, the latter of which determines the KV100. That's why I use the HTHSV, not the KV100, in the BOQI II calculation so that the VII doesn't play a role.

 

[Gokhan]

Originally Posted by JHZR2
Time in the NOACK test appears to me to be arbitrary.

Yes I agree that time is a consideration in it, but the units are % @ a specific temperature and time duration. My take is that given the reasoning that time is in NOACK, then temperature would need to be as well, but you cannot parse it to be a % / hr*deg C. The time is just to ensure that the sample is thermally equilibrated.

Originally Posted by Gokhan

For example, if NOACK is 8%, the fractional evaporation is 8% in 60 minutes, 4% in 30 minutes, and 2% in 15 minutes.

This is factually incorrect. It does not account for thermal equilibration and the distribution of light ends that may volitalize at some rate inconsistent with the bulk temperature rise. This is essentially similar in principle to a distillation process or a gas chromatography injection process, where some fraction volitalizes at some temperature, and it may not track linearly to your calculation.

Also, arbitrary time doesnt apply to the (length ^2) / time units of viscosity.

Originally Posted by CR94
Originally Posted by OilUzer
As @JHZR2 said, evaporation is not linear in either direction. if noack is %8 do you lose %16 in 120 min? ...
True, and it's not necessarily an exponential function, as I gather Gokhan assumes, either. Color me skeptical.

I discussed this before in a different thread.

(1) The Noack evaporation test is a nearly perfect exponential decay.

(2) For practical Noack values (less than 15% or so), the exponential decay can be approximated by a linear decay. Then, the decay time constant tau of the exponential decay approximately calculates to be the 1/Noack and the decay rate lambda approximately calculates to be the Noack.

So, yes, it's true that if Noack is 8%, the fractional evaporation is 8% in 60 minutes, approximately 4% in 30 minutes, and approximately 2% in 15 minutes.

Here is the actual data:

Determination of the NOACK evaporation loss of lubricants by TGA

The previous discussion in a different thread:

Noack and TGA tests as function of time

In summary:

The fractional evaporation ~ 1 - exp(-t * Noack)

Here t is in hours. There is a small error in approximating the exponential by a linear function.

If you want to be more accurate,

Evaporation rate lambda = -ln(1 - Noack)
Fractional evaporation = 1 - exp(-lambda * t)

Therefore, ideally, you would have to plug in lambda = -ln(1 - Noack) instead of Noack in the BOQI II calculation. However, it unnecessarily complicates the calculation and probably not worth it. For example, for Noack = 9.0% and 13%, you get lambda = -ln(1 - Noack) = 9.4% and 13.9% respectively. The ratios without and with the correction are 1.44 and 1.48, respectively. I can use the corrected value if people prefer a little more accuracy but the Noack test itself is not an accurate one.

 

[Gokhan]

Originally Posted by JAG
The units matter is a can of worms and I think it's unnecessary since this variable is a figure of merit, not some variable derived from scientific first principles. Figures of merit don't need to have units that fit any rules.

I can see why the following might be a decent figure of merit: HTHSV / (CCS x NOACK)

Ideally, the numerator is a variable that we want to maximize, while both variables in the denominator are what we want to minimize. All of the variables are correlated, which is not a problem. Actually, it's more correct to say that we want to maximize the ratio HTHSV / CCS and minimize 1 / NOACK. For example, if the formulator changes nothing but decreases base oil viscosity and increases the amount of VIIs such that the KV100C is held constant, the HTHS increases, the CCS decreases, and the NOACK increases. One would have to know the actual values to determine whether the figure of merit would go up or down. I would be interested in seeing this figure of merit for many oils (base oils and fully formulated).

I don't understand the rationale for the new figure of merit (BOQI II) being:
BOQI x HTHSV - HTHSV / (CCS x NOACK)
, rather than simply:
HTHSV / (CCS x NOACK)

Oh, it's not the difference.

BOQI ~ 1/(CCS * Noack)
BOQI II ~ HTHSV/(CCS * Noack)
BOQI II = HTHSV * BOQI

 

[Shannow]

Originally Posted by Gokhan
Originally Posted by Shannow
Now you've thrown in HTHS, which as you go up in the grades (20 through 40) is the result of more polymeric VI thickeners, so now has even LESS to do with BASE OIL QUALITY.

The viscosity-index improver (VII) temporarily shears at the 1,000,000 per second shear rate at 150 C to a great extent and you're mostly looking at the base oil and not the VII effects. True, there would still be a small amount of unsheared VII but (1) the residual-VII effect is small and (2) for a given viscosity grade, the residual VII will be similar for different oils, multiplying the BOQI II by the same number very close to 1, and the approximation of neglecting it is an excellent one. Therefore, BOQI II still measures the base oil, not the VII-thickening effects.


No, you are wrong, and assuming that the VII effect becomes zero...evidence ?....for 0W20, 0W30, and 0W40, the basestocks are getting progressively thinner...so clearly, the increasing amounts of VII are having an increasing effect in the HTHS

 

[Gokhan]

Here are the equations for BOQI II:

BOQI II = c * HTHSV / (CCS * Noack)

c = 3,500,000 for CCS @ -35 C (0W-xx oils)
c = 2,000,000 for CCS @ -30 C (5W-xx oils)
c = 1,160,000 for CCS @ -25 C (10W-xx oils)

For higher accuracy, use:

BOQI = c * HTHSV / (CCS * 100 * lambda)
with lambda = -ln(1 - [Noack / 100])

Noack is the percent evaporation in 60 minutes. lambda is the evaporation rate in units of 1/hour. HTHSV and CCS are in units of cP. The units for BOQI II are time.

Here are some representative oils:

Code
Oil Constant HTHSV CCS Noack lambda BOQI II* (*using lambda)



CE 0W-20 2015 3500000 2.60 5537 12.7 0.1358 121

CE 5W-20 2017 2000000 2.70 5851 12.8 0.1370 67

CE 5W-30 2013 2000000 3.00 5544 11.1 0.1177 92

CGTX 5W-20 2013 2000000 2.60 5136 14.0 0.1508 67

CGTX 5W-30 2011 2000000 3.00 5653 13.7 0.1473 72

GC 0w-30 3500000 3.50 5880 8.3 0.0866 240

M1 0W-20 2017 3500000 2.70 4182 10.7 0.1132 200

M1 0w-40 SM 3500000 3.60 5600 9.0 0.0943 239

M1 5W-30 2013 2000000 3.10 3937 10.1 0.1065 148

M1 AP 5W-30 2018 2000000 3.00 5175 8.5 0.0888 131

PPPP 0W-20 2017 3500000 2.70 5884 10.3 0.1087 148

PPPP 5W-30 2014 2000000 3.00 4080 9.1 0.0954 154

PYB 5W-20 2016 2000000 2.70 5339 13.1 0.1404 72

Amsoil SS 0W-20 3500000 2.67 5122 8.5 0.0888 205

Amsoil SS 0W-30 3500000 3.07 5372 8.8 0.0921 217

Amsoil SS 0W-40 3500000 3.76 6062 7.7 0.0801 271

Amsoil SS 10W-30 1160000 3.11 4278 4.1 0.0419 201

Amsoil SS 5W-20 2000000 2.67 4385 5.8 0.0598 204

Amsoil SS 5W-30 2000000 3.11 3968 6.7 0.0694 226

Amsoil SS 5W-50 2000000 4.45 5108 6.1 0.0629 277

CE, CGTX, and GC are Castrol Edge, Castrol GTX, and German Castrol. PPPP is Pennzoil Platinum with PurePlus. PYB is Pennzoil Yellow Bottle. Amsoil SS is Amsoil Signature Series.

You can see that all viscosities of Amsoil Signature Series did similarly but 0W-40 and 5W-50 got an extra bump -- perhaps some VII effect there.

 

[Gokhan]

BOQI II is distinguishing greatly between conventional oils and some synthetic oils. The known PAO-based oils Amsoil SS scored excellent and the half-PAO-based M1 AFE 0W-20 scored well. GTL-based PPPP scored decent.

The PAO-based German Castrol scored excellent.

 

[ZeeOSix]

Originally Posted by CR94
Originally Posted by OilUzer
As @JHZR2 said, evaporation is not linear in either direction. if noack is %8 do you lose %16 in 120 min? ...
True, and it's not necessarily an exponential function, as I gather Gokhan assumes, either. Color me skeptical.


Only way to know is conduct the test for a few hours longer beyond the 1 hour test point, collect data every 10 min or shorter time period and plot out the curve.

 

[Gokhan]

Originally Posted by ZeeOSix
Originally Posted by CR94
Originally Posted by OilUzer
As @JHZR2 said, evaporation is not linear in either direction. if noack is %8 do you lose %16 in 120 min? ...
True, and it's not necessarily an exponential function, as I gather Gokhan assumes, either. Color me skeptical.
Only way to know is conduct the test for a few hours longer beyond the 1 hour test point, collect data every 15 min and plot out the curve.

Evaporation rate lambda = -ln(1 - Noack)
Fractional evaporation = 1 - exp(-lambda * t)

For Noack = 8.0000%, lambda = 0.083382.

Then, the fractional evaporation as a function of time is:

Code
Time Evaporation



0 hr 0.0%

1 hr 8.0%

2 hr 15.4%

3 hr 22.1%

4 hr 28.4%

5 hr 34.1%

6 hr 39.4%

7 hr 44.2%

8 hr 48.7%

9 hr 52.8%

10 hr 56.6%

11 hr 60.0%

12 hr 63.2%

13 hr 66.2%

14 hr 68.9%

15 hr 71.4%

16 hr 73.7%

17 hr 75.8%

18 hr 77.7%

19 hr 79.5%

20 hr 81.1%

21 hr 82.6%

22 hr 84.0%

23 hr 85.3%

24 hr 86.5%

 

[OilUzer]

@Gokhan,
Maybe you've already done this but can you give us a high level summary please.

If I'm buying oil, is there a number (index) below which I should avoid the oil ... or can for example reduce or extend oci due to a given index number, etc.
Trying to get a better feel for these numbers ...

It will also be interesting if you include an approximate price per quart column next to each oil and index. Curious how prices track the index.

only if you had unlimited access to all uoa data and could tie the results to each oil & index

 

[bbhero]

Taking bets on how long this goes on for interns of pages....

I will say 15 pages

 

[1JZ_E46]

How do we find BOQI II of an oil without a stated CCS? Estimate? For example, M1 10W-30 HM. HTHS of 3.5, Noack of 7.6%.

 

[Gokhan]

It looks like BOQI II around 150 or higher is Group III+++, GTL, or has some PAO in the mix. BOQI II around 200 or higher is PAO. BOQI II around 100 or less is lowest-grade Group III and BOQI around 70 is Group II+.

The add pack makes a big difference in the overall oil quality as well. I ran into a paper that looked at the SJ 5W-40s. They figured out the amount of PAO, ester, and Group III in them. Surprisingly, some PAO-based oils did poorly in oil-life (oxidation) tests because of their add pack. I can post it later.

You cannot base your oil selection entirely on BOQI II but it can give you an idea of the base-oil type.

 

[Gokhan]

Originally Posted by 1JZ_E46
How do we find BOQI II of an oil without a stated CCS? Estimate? For example, M1 10W-30 HM. HTHS of 3.5, Noack of 7.6%.

Unfortunately, you can't. You need the CCS. The MRV is controlled by the pour-point depressant.

Sometimes, Oil-Club Russia will test for CCS, Noack, etc.

 

[Gokhan]

According to the ExxonMobil MSDS, M1 HM 10W-30 is mostly Group III with about 5% PAO and 10% GTL. It obviously has a thick base oil, hence high HTHSV and low Noack, which comes at the expense of higher CCS.

 

[ZeeOSix]

Originally Posted by Gokhan
Originally Posted by ZeeOSix
Originally Posted by CR94
Originally Posted by OilUzer
As @JHZR2 said, evaporation is not linear in either direction. if noack is %8 do you lose %16 in 120 min? ...
True, and it's not necessarily an exponential function, as I gather Gokhan assumes, either. Color me skeptical.
Only way to know is conduct the test for a few hours longer beyond the 1 hour test point, collect data every 15 min and plot out the curve.

Evaporation rate lambda = -ln(1 - Noack)
Fractional evaporation = 1 - exp(-lambda * t)

For Noack = 8.0000%, lambda = 0.083382.

Then, the fractional evaporation as a function of time is:

Code
Time Evaporation



0 hr 0.0%

1 hr 8.0%

2 hr 15.4%

3 hr 22.1%

4 hr 28.4%

5 hr 34.1%

6 hr 39.4%

7 hr 44.2%

8 hr 48.7%

9 hr 52.8%

10 hr 56.6%

11 hr 60.0%

12 hr 63.2%

13 hr 66.2%

14 hr 68.9%

15 hr 71.4%

16 hr 73.7%

17 hr 75.8%

18 hr 77.7%

19 hr 79.5%

20 hr 81.1%

21 hr 82.6%

22 hr 84.0%

23 hr 85.3%

24 hr 86.5%


Do you have actual test data from a 24 hr Noack test to validate your calculations out to 24 hrs? That's what I was getting at.

 

[Shannow]

Originally Posted by bbhero
Taking bets on how long this goes on for interns of pages....

I will say 15 pages

I've only got a couple of posts left in me...seriously...there's too much "wannabelieve" on BITOG these days...but here goes...

 

[Shannow]

Originally Posted by Shannow
Originally Posted by Gokhan
Originally Posted by Shannow
Now you've thrown in HTHS, which as you go up in the grades (20 through 40) is the result of more polymeric VI thickeners, so now has even LESS to do with BASE OIL QUALITY.

The viscosity-index improver (VII) temporarily shears at the 1,000,000 per second shear rate at 150 C to a great extent and you're mostly looking at the base oil and not the VII effects. True, there would still be a small amount of unsheared VII but (1) the residual-VII effect is small and (2) for a given viscosity grade, the residual VII will be similar for different oils, multiplying the BOQI II by the same number very close to 1, and the approximation of neglecting it is an excellent one. Therefore, BOQI II still measures the base oil, not the VII-thickening effects.


No, you are wrong, and assuming that the VII effect becomes zero...evidence ?....for 0W20, 0W30, and 0W40, the basestocks are getting progressively thinner...so clearly, the increasing amounts of VII are having an increasing effect in the HTHS


OK, supporting my post above, now that I'm home and access to my compy, and the spreadsheet quoted in the OP's post.

First Item... the above statement that the HTHS is a measure of the base oil HTHS, as the polymeric viscosity modifiers are sheared, and thus ineffective.

There are hundreds of resources on line which explain the Newtonian range, and the non Newtonian range quite clearly and succincltly...and not one of them claims that the effect of Viscosity modifiers is "zero" at the HTHS high shear regime...it's Newtonian again, and it's thicker than the base oils.

Now I'll support my above statement (quoted), in refuting Gokhan's statement that the HTS is reflective of base oil viscosity.

Here's my spreadsheet modified to include the base oil HTHS (taken from Mobil data sheets, and using the Widman mixing tool)…

Note : just like I stated, moving from 0W20 to 0W40, not only does the base oil viscosity become thiner...but (and it's bleedingly obvious) so does the HTHS...BECAUSE THE BASE OIL IS THINNER....

 

[Shannow]

Now what I find interesting about this thread, and the changes to the BOQI that apparently I prompted.

In the past, I have (quite rightly) criticised the BOQI for a couple of reasons.
* that using M1 Spectrasyn, it gave different "quality indices" for a range of oils blended with exactly the same family of basestocks.
* that using the 100% PURE basestocks, it gave different BOQIs - how could SS4, 6, and 8 all have different numbers if the index worked ???

Firstly Gokhan disclaimed that Mobil data sheet that he is now relying on as an indicative advertisement only, and not reflective of a real lubricant. Now that mashing the numbers together comes up with something...something meaningless, but still something...it's valid.

Secondly, the fact that the index didn't work for the first three members of the spectrasyn range has been ignored....that's important, as the origin of the BOQI was as a BLENDING TOOL...to estimate whether a blend would reach a target spec before breaking out the beakers.

So below is what the BOQI and the BOQI state about Spectrasyn 4, 6, and 8...note the wild variance in "quality" between 3 PAOs of the same process/family.